Pathophysiology of ischemic stroke

INTRODUCTION — The term ischemic stroke is used to describe a variety of conditions in which blood flow to part or all of the brain is reduced, resulting in tissue damage. Although in some cases this may be a chronic condition, most strokes occur acutely. Research over the last four decades has resulted in a significant expansion of our knowledge and understanding of the molecular and cellular processes that underlie ischemia-induced cellular injury.

The goal of this review is to provide an overview of the underlying factors, such as the hemodynamic changes and molecular and cellular pathways, which are involved in stroke-related brain injury. A better understanding of these processes may help in the development of new therapies that are needed to treat this devastating disease.

STROKE SUBTYPES — The etiology and clinical classification of ischemic stroke subtypes is reviewed here briefly and discussed in greater detail separately. (See "Stroke: Etiology, classification, and epidemiology", section on 'Brain ischemia' and "Clinical diagnosis of stroke subtypes".)

Acute ischemic stroke subtypes are often classified in clinical studies using a system developed by investigators of the TOAST trial, based upon the underlying cause (table 1) [1]. Under this system, strokes are classified into the following categories:

●Large artery atherosclerosis

●Cardioembolism

●Small vessel occlusion

●Stroke of other, unusual, determined etiology

●Stroke of undetermined etiology

Ischemic strokes are due to a reduction or complete blockage of blood flow [2]. This reduction can be due to decreased systemic perfusion, severe stenosis, or occlusion of a blood vessel. Decreased systemic perfusion can be the result of low blood pressure, heart failure, or loss of blood. Determination of the type of stroke can influence treatment to be used. The main causes of ischemia are thrombosis, embolization, and lacunar infarction from small vessel disease. Ischemic strokes represent approximately 80 percent of all strokes. (See "Stroke: Etiology, classification, and epidemiology", section on 'Epidemiology'.)

●Thrombosis refers to obstruction of a blood vessel due to a localized occlusive process within a blood vessel [2]. The obstruction may occur acutely or gradually. In many cases, underlying pathology such as atherosclerosis may cause narrowing of the diseased vessel. This may lead to restriction of blood flow gradually, or in some cases, platelets may adhere to the atherosclerotic plaque forming a clot leading to acute occlusion of the vessel. Atherosclerosis usually affects larger extracranial and intracranial vessels. In some cases, acute occlusion of a vessel unaffected by atherosclerosis may occur because of a hypercoagulable state. (See "Stroke: Etiology, classification, and epidemiology", section on 'Thrombosis'.)

●Embolism refers to clot or other material formed elsewhere within the vascular system that travels from the site of formation and lodges in distal vessels causing blockage of those vessel and ischemia [2]. The heart is a common source of this material, although other arteries may also be sources of this embolic material (artery to artery embolism). In the heart, clots may form on valves or chambers. Tumors, venous clots, septic emboli, air, and fat can also embolize and cause stroke. Embolic strokes tend to be cortical and are more likely to undergo hemorrhagic transformation, probably due to vessel damage caused by the embolus. Emboli from venous sources such as a deep venous thrombosis (DVT) can also cause stroke if the emboli are able to migrate to the arterial system through a patent foramen ovale (PFO) or an arteriovenous (AV) shunt such as pulmonary AV fistulae. (See "Stroke: Etiology, classification, and epidemiology", section on 'Embolism'.)

●Lacunar infarction occurs as a result of small vessel disease. Smaller penetrating vessels are more commonly affected by chronic hypertension leading to hyperplasia of the tunica media of these vessels and deposition of fibrinoid material leading to lumen narrowing and occlusion [2]. Lacunar strokes can occur anywhere in the brain but are typically seen in subcortical areas. Atheroma can also encroach on the orifices of smaller vessels leading to occlusion and stroke. (See "Lacunar infarcts".)

●Nonatherosclerotic abnormalities of the cerebral vasculature, whether inherited or acquired, predispose to ischemic stroke at all ages, but particularly in younger adults and children. These can be divided into noninflammatory and inflammatory etiologies. The following list, though not exhaustive, highlights the major nonatherosclerotic vasculopathies associated with ischemic stroke:

•Arterial dissection (figure 1)

•Fibromuscular dysplasia (see "Clinical manifestations and diagnosis of fibromuscular dysplasia")

•Vasculitis (see "Overview of and approach to the vasculitides in adults" and "Vasculitis in children: Incidence and classification")

•Moyamoya disease (see "Moyamoya disease and moyamoya syndrome: Etiology, clinical features, and diagnosis")

•Sickle cell disease arteriopathy (see "Acute stroke (ischemic and hemorrhagic) in children and adults with sickle cell disease")

•Focal cerebral arteriopathy of childhood (see "Ischemic stroke in children and young adults: Epidemiology, etiology, and risk factors", section on 'Focal cerebral arteriopathy')

●Decreased systemic perfusion due to systemic hypotension may produce generalized ischemia to the brain [2]. This is most critical in the borderzone (or watershed) areas, which are territories that occupy the boundary region of two adjacent arterial supply zones. The ischemia caused by hypotension may be asymmetric due to preexisting vascular lesions. Areas of the brain commonly affected include the hippocampal pyramidal cells, cerebellar Purkinje cells, and cortical laminar cells discussed below. (See "Stroke: Etiology, classification, and epidemiology", section on 'Systemic hypoperfusion'.)

Despite extensive testing to identify the etiology of the stroke, no clear cause is found in approximately 25 percent of patients with ischemic stroke at hospital discharge (cryptogenic stroke). In rare cases, the pathophysiology of the ischemic infarct may not even be vascular. Examples include mitochondrial disorders such as mitochondrial encephalopathy with lactic acidosis and stroke-like episodes (MELAS) in which the pathophysiology is a failure of the energy production system (mitochondria) rather than a problem with blood delivery.

CEREBRAL AUTOREGULATION — Under normal conditions, the rate of cerebral blood flow is primarily determined by the amount of resistance within cerebral blood vessels, which is directly related to their diameter [3]. Dilation of vessels leads to an increased volume of blood in the brain and increased cerebral blood flow, whereas constriction of vessels has the opposite effect. Cerebral blood flow is also determined by variation in the cerebral perfusion pressure.

Cerebral autoregulation is the phenomenon by which cerebral blood flow is maintained at a relatively constant level despite moderate variations in perfusion pressure. The mechanism by which autoregulation occurs is not well understood and may involve multiple pathways. Evidence suggests that the smooth muscle in cerebral vessels can respond directly to changes in perfusion pressure, contracting when pressure increases and relaxing when pressure drops. Reductions in cerebral blood flow may also lead to dilation of blood vessels through the release of vasoactive substances, although the molecules responsible for this have not been identified. The endothelial release of nitric oxide also appears to play a role in autoregulation.

Maintenance of cerebral blood flow by autoregulation typically occurs within a mean arterial pressure range of 60 to 150 mmHg. The upper and lower limits vary between individuals, however. Outside of this range, the brain is unable to compensate for changes in perfusion pressure, and the cerebral blood flow increases or decreases passively with corresponding changes in pressure, resulting in the risk of ischemia at low pressures and edema at high pressures (figure 2).

Cerebral autoregulation during stroke — Cerebral autoregulation is impaired during some disease conditions, including ischemic stroke [3-5]. As cerebral perfusion pressure falls, cerebral blood vessels dilate to increase cerebral blood flow. A decrease in perfusion pressure beyond the ability of the brain to compensate results in a reduction in cerebral blood flow. Initially, the oxygen extraction fraction is increased in order to maintain levels of oxygen delivery to the brain. As the cerebral blood flow continues to fall, other mechanisms come into play (figure 3).

Inhibition of protein synthesis occurs at flow rates below 50 mL/100 g per minute. At 35 mL/100 g per minute, protein synthesis ceases completely, and glucose utilization is transiently increased. At 25 mL/100 g per minute, glucose utilization drops dramatically with the onset of anaerobic glycolysis, resulting in tissue acidosis from the accumulation of lactic acid. Neuronal electrical failure occurs at 16 to 18 mL/100 g per minute, and failure of membrane ion homeostasis occurs at 10 to 12 mL/100 g per minute. This level typically marks the threshold for the development of infarct (figure 3).

In hypertensive individuals, autoregulation has adapted to occur at higher arterial pressures. Reduction of blood pressure to normal levels could actually exacerbate the derangement to autoregulation that occurs during stroke and lead to a further decrease in cerebral blood flow (figure 2).

CONSEQUENCES OF REDUCTION IN BLOOD FLOW DURING STROKE — The human brain is exquisitely sensitive and susceptible to even short durations of ischemia. The brain is responsible for a large part of the body's metabolism and receives approximately 20 percent of the cardiac output although it is only 2 percent of total body weight [3]. The brain contains little or no energy stores of its own, and therefore relies on the blood for their delivery. Even brief deprivation can lead to death of the affected brain tissue. During stroke, reduction of blood flow to a portion or all of the brain results in a deprivation of glucose and oxygen [2].

Most strokes are caused by focal ischemia, affecting only a portion of the brain, typically involving a single blood vessel and its downstream branches. The region directly surrounding the vessel is the most affected. Within this region, cells in a central core of tissue will be irreversibly damaged and die by necrosis if the duration of ischemia is long enough. At distances farther from the affected vessel, some cells may receive a small amount of oxygen and glucose by diffusion from collateral vessels. These cells do not die immediately and can recover if blood flow is restored in a timely manner. The central core of tissue destined to die, or containing tissue that is already dead, is called the infarct. The region of potentially salvageable tissue is known as the penumbra.

Mechanisms of ischemic cell injury and death — Brain ischemia initiates a cascade of events that eventually lead to cell death; including depletion of adenosine triphosphate (ATP); changes in ionic concentrations of sodium, potassium, and calcium; increased lactate; acidosis; accumulation of oxygen free radicals; intracellular accumulation of water; and activation of proteolytic processes [2,6-8].

As a consequence of the electrical failure that occurs during ischemia, the release of the excitatory amino acid glutamate at neuronal synapses is increased [2]. This leads to the activation of glutamate receptors and the opening of ion channels that allow potassium ions to exit the cell and sodium and calcium ions to enter, which has a number of physiologic effects. The primary glutamate receptor subtype involved in ischemic damage is the N-methyl-D-aspartate (NMDA) receptor. In addition, the alpha-amino-3-hydroxy-5-methyl-4-isoxazoleproprionic acid (AMPA) and metabotropic glutamate receptors are thought to play a role. Activation of these receptors leads to membrane depolarization and increased calcium influx.

Numerous cellular signaling pathways respond to calcium levels, and the influx of calcium resulting from glutamate receptor stimulation leads to their activation. These pathways have both beneficial and detrimental effects. The influx of sodium ions is balanced by the influx of water into the cell, leading to edema. Sodium influx also causes reversal of the normal process of glutamate uptake by astrocyte glutamate transporters, resulting in increased glutamate release [9-12]. As a result of its increased release and decreased uptake, glutamate accumulates to excessive levels and leads to continuous stimulation. This condition is often referred to as excitotoxicity.

Another effect of NMDA receptor activation is the production of nitric oxide [13]. The activity of nitric oxide synthase (NOS) and the total amount of nitric oxide present in the brain are increased following exposure to hypoxia [14].

Nitric oxide is an important signaling molecule within the body and can be beneficial at normal physiologic levels. As an example, endothelial nitric oxide synthase (eNOS) leads to the production of low levels of nitric oxide that cause vasodilation and increase blood flow [15]. However, neuronal nitric oxide synthase (nNOS) and inducible nitric oxide synthase (iNOS) result in larger amounts of nitric oxide that may lead to brain injury. Nitric oxide is a free radical and reacts directly with cellular components to damage them. Nitric oxide can also react with another free radical, superoxide, to produce the highly reactive peroxynitrite. Peroxynitrite causes single strand breaks in DNA [16]. This results in the activation of DNA repair enzymes, which consume vital energy needed for other processes. DNA damage also may activate the process of apoptosis, leading to cell death.

The production of reactive oxygen species, a normal byproduct of oxidative metabolism, is also increased during ischemia. Like nitric oxide, they can react with and damage cellular components. Injury to the plasma membrane of a cell can lead to the inability to control ion flux, resulting in mitochondrial failure. Reactive oxygen species, as well as calcium influx and other factors, can also permeabilize the mitochondrial membrane [17]. This leads to metabolic failure as well as the release of initiators of apoptosis and DNA damage. Metabolic failure results in the depletion of cellular ATP levels. ATP is required for nuclear condensation and DNA degradation in the final stages of apoptosis [18]. In the absence of ATP, cell death occurs by necrosis rather than apoptosis. (See 'Necrosis and apoptosis' below.)

The release of byproducts from cellular damage and death by necrosis activates components of the inflammatory pathway [19]. The role that inflammation plays during ischemia is mixed, having both positive and negative effects [20]. On the one hand, inflammation results in an increase in blood flow to the ischemic region, which may deliver vital glucose and oxygen to cells. On the other hand, increased blood flow may also deliver more calcium to the area, resulting in increased tissue damage.

Inflammation also results in the migration of activated leukocytes to damaged tissue [21,22]. Although these leukocytes may remove damaged and necrotic tissue, they also release cytokines to attract additional inflammatory cells. Under severe inflammatory conditions, these cytokines can accumulate to toxic levels.

Necrosis and apoptosis — Cell death following cerebral ischemia or stroke can occur by either necrosis or by apoptosis. The process of necrosis is not well understood. In early stages, cellular chromatin becomes uniformly compacted, the endoplasmic reticulum is dilated, and ribosomes are dispersed [23]. In later stages, swelling of the cell and mitochondria is followed by rupture of the nuclear, organelle, and plasma membranes, leading to the release of cellular material into the surrounding environment. This release of material results in the stimulation of inflammatory processes within the brain.

Apoptosis is highly regulated and has been studied in more detail than necrosis. As in necrosis, the chromatin begins to condense during early stages of apoptosis. Instead of cellular swelling, however, the contents of the cytoplasm also condense, and the mitochondria and other organelles remain intact. In later stages, the nucleus is broken into discrete fragments and the entire contents of the cell are divided into membrane bound bodies that are subsequently phagocytosed by macrophages.

There are three known pathways by which apoptosis can be initiated [24]:

●Mitochondrial permeabilization

●Death receptor (Fas) pathway

●Endoplasmic reticulum stress

The most well-known pathway involves permeabilization of the mitochondria and release of cytochrome c into the cytoplasm. Activation of membrane-bound Fas, the so called "death receptor," and the accumulation of misfolded proteins at the endoplasmic reticulum during stress, can also lead to apoptosis. These initiators all lead to the activation of caspases that cleave cellular proteins and eventually cause cell death. Caspase-independent mechanisms of apoptosis have also been proposed.

The pattern of cell death after cerebral ischemia, as seen in animal models, depends on the nature of the insult to cerebral tissue [25]. In global cerebral ischemia, such as occurs after cardiac arrest and resuscitation or transient severe systemic hypotension, the entire brain is exposed to ischemia. Formation of infarct is not immediate, but rather occurs after a delay of 12 hours to several days. Cell death is limited to those regions of the brain that are particularly susceptible to ischemic damage, such as the CA1 and CA4 regions of the hippocampus, the striatum, and cortical layers two and five. Cell death in these regions occurs primarily by apoptosis.

Focal cerebral ischemia is a more common pattern than global ischemia in human stroke. In animal models of focal ischemia, changes in cell morphology are visible microscopically as early as two to three hours after the insult, and the infarct develops rapidly over a period of 6 to 24 hours. Cell death occurs by necrosis in the core, with apoptotic cells located on the periphery [6]. In addition to the type of insult, the duration of ischemia also affects the pattern by which cell death occurs. Longer ischemic insults produce greater damage to cerebral tissue, resulting in an increased proportion of necrosis and decreased proportion of apoptosis.

There have been few studies of apoptosis in the brain following stroke in human patients. However, accumulating evidence suggests that apoptosis is involved [26-29], as illustrated by the following observations:

●In a neuropathology study that compared specimens from 27 patients who had cerebral infarction with specimens from rat brains subjected to experimental transient forebrain ischemia, the patterns of cell death were similar in human and animal brain tissue and included both morphologic and histochemical findings typical of apoptosis [26]. In the human stroke specimens, apoptosis was apparent during the subacute stage, but was not seen in acute or chronic stages.

●In another neuropathology report that compared 13 cases of fatal ischemic stroke with three patients who died of non-neurologic causes, histochemical and morphologic changes indicative of apoptosis were seen in cells throughout the brain of both patients and controls [29]. The morphologic changes were more advanced in the peri-infarct region and infarct core of the patients with stroke. Apoptotic cells were located primarily within the peri-infarct region, consisting of up to 26 percent of all cells. Increased ischemic damage and neuronal necrosis was associated with a decrease in the percentage of apoptotic cells.

The deciding factor in determining whether cells undergo necrosis or apoptosis seems to be the level of energy available in the form of ATP, which is required for formation of the apoptosome. Apoptosis is unable to proceed in its absence. When energy levels are limiting, cell death therefore occurs by necrosis rather than apoptosis. The role of ATP in the mechanism of cell death has been investigated primarily using cell culture models. Cultured neurons depend on the presence of serum in the culture medium for survival [24]. If the serum is removed, the cells die by necrosis. In serum-free media with added glucose, however, the cells die by apoptosis.

Levels of ATP in the brain are decreased during stroke due to the lack of glucose and oxygen required for normal cellular metabolism. Glucose metabolism is decreased by approximately 50 percent in both global and focal ischemia models of stroke. As a consequence, ATP levels may fall to 10 percent of normal in global models or 25 percent in the infarct core in focal ischemia models. ATP levels in the penumbra, however, only drop to 50 percent to 70 percent of normal [30].

ATP levels in the brain may also be decreased by mitochondrial damage or failure; activity of DNA repair enzymes, such as poly ADP-ribose polymerase (PARP); and neuronal depolarization related to glutamate excitotoxicity. In stroke, therefore, low levels of ATP within the core infarct are insufficient to support apoptosis, and cell death occurs by necrosis. In the penumbra, ATP levels are sufficient enough that cell death by apoptosis can occur. As the duration of ischemia increases, however, ATP levels are eventually depleted and the proportion of cells that undergo necrosis is increased, with a decrease in the number of apoptotic cells.

Loss of brain structural integrity — Cerebral ischemia and infarction leads to loss of the structural integrity of the affected brain tissue and blood vessels [6]. This process of tissue destruction and neurovascular disruption is mediated in part by the release of various proteases, particularly the matrix metalloproteases (MMP) that degrade collagens and laminins in the basal lamina [7,31]. The loss of vascular integrity leads to a breakdown of the blood-brain-barrier and development of cerebral edema. Catastrophic failure of vascular integrity is postulated to cause hemorrhagic conversion of ischemic infarction by allowing extravasation of blood constituents into the brain parenchyma [32].

Cerebral edema — Cerebral edema complicating stroke can cause secondary damage by several mechanisms, including increased intracranial pressure, which may decrease cerebral blood flow, and mass effect causing displacement of brain tissue from one compartment to another (ie, herniation), a process that can be life-threatening.

Two types of cerebral edema can occur as a consequence of ischemic stroke [2,6,32,33].

●Cytotoxic edema is caused by the failure of ATP-dependent transport of sodium and calcium ions across the cell membrane. The result is accumulation of water and swelling of the cellular elements of the brain, including neurons, glia, and endothelial cells.

●Vasogenic edema is caused by increased permeability or breakdown of the brain vascular endothelial cells that constitute the blood-brain barrier [34]. This allows proteins and other macromolecules to enter the extracellular space, resulting in increased extracellular fluid volume.

Roughly 10 percent of ischemic strokes are classified as malignant or massive because of the presence of space-occupying cerebral edema that is severe enough to produce elevated intracranial pressure and brain herniation. (See "Malignant cerebral hemispheric infarction with swelling and risk of herniation", section on 'Malignant hemispheric infarction'.)

GENETICS OF STROKE — Many of the known risk factors for stroke are variable traits influenced by multiple genes, making it difficult to sort out the genetics behind them. The study of stroke genetics is also impaired by interactions between different risk factors that modulate their effects. It is widely accepted, however, that there is a genetic component to stroke that can lead to increased or decreased risk. Outside of the monogenic disorders discussed below, it is probable that many alleles with small effect sizes contribute to the risk of ischemic stroke [35,36]. Much of the evidence for this comes from studies of twins and from families with a history of stroke [37].

●Earlier studies of twins have been troubled by low sample numbers and poor classification of stroke type [37]. However, these studies indicate that stroke-related death in one sibling is associated with a higher risk of stroke-related death in the other sibling among monozygotic (identical) twins versus dizygotic (fraternal) twins. This observation suggests that genetic factors shared by the monozygotic twins played a role in their strokes. As an example, in one of the larger twin studies that evaluated 990 same-sex twin pairs, stroke death affecting both siblings was twice as likely among monozygotic twin pairs compared with dizygotic twin pairs (10 versus 5 percent), and the difference was statistically significant [38].

●A family history of stroke is associated with an increased risk of stroke among the offspring [39]. This has been observed for offspring with maternal and paternal histories of stroke [37], and among individuals having a sibling with a prior stroke [40].

Additional insights into the relationship of genetic variants and the risk of ischemic stroke come from genome-wide association studies (GWAS):

●A 2012 meta-analysis of GWAS that analyzed data from over 12,000 subjects of European ancestry with ischemic stroke and 60,000 controls identified three loci (PITX2, ZFHX3, and HDAC9) with genome-wide significance for ischemic stroke [41]. Importantly, each locus was associated with a specific stroke subtype:

•PITX2 and ZFHX3, previously identified as risk factors for atrial fibrillation [42-44], were associated with cardioembolic stroke [41]

•HDAC9 [44] was associated with large vessel ischemic stroke [41]

●Findings from a 2013 European GWAS of genetic factors related to coagulation suggested that ABO gene variants are associated with large vessel and cardioembolic stroke subtypes [45], and a systematic review and meta-analysis testing the association of APOE genotype with MRI markers of cerebrovascular disease found that APOE epsilon 2 carrier status was associated with an increased risk of ischemic stroke [46].

●In accord with prior results, a 2016 meta-analysis of 12 GWAS with over 10,000 stroke cases and 19,000 controls found genome-wide significance for four loci [47]:

•PITX2 and ZFHX3 for cardioembolic stroke

•HDAC9 for large vessel disease

•ABO for all ischemic stroke

Ethnic differences may also contribute to stroke risk. Although differences in lifestyle may be partly responsible for increased or decreased likelihood of stroke, genetic factors also play a role. As an example, individuals of Black African descent have a significantly higher rate of stroke than White populations, even after adjusting for differences in nongenetic risk factors [37,48]. This may or may not be related to the increased frequency among African populations of the sickle cell trait, which is a known cause of stroke due to the obstruction of small blood vessels by abnormal red blood cells. Even in monogenic disorders such as sickle cell disease, multiple genes may interact to modify risk. In a study of 1398 individuals with sickle cell anemia, 12 genes were found to interact with the mutated hemoglobin and modulate the risk of stroke [49].

Monogenic disorders — A number of monogenic syndromes are associated with an increased risk of ischemic stroke, including the following [50]:

●Marfan syndrome and Ehlers-Danlos syndrome, which predispose to cervical artery dissection (see "Cerebral and cervical artery dissection: Clinical features and diagnosis", section on 'Associated conditions and risk factors' and "Genetics, clinical features, and diagnosis of Marfan syndrome and related disorders")

●Familial moyamoya disease (see "Moyamoya disease and moyamoya syndrome: Etiology, clinical features, and diagnosis", section on 'Etiology and Pathogenesis')

●Fabry disease (see "Fabry disease: Neurologic manifestations")

●Pseudoxanthoma elasticum (see "Pseudoxanthoma elasticum")

●Homocystinuria (see "Ischemic stroke in children and young adults: Epidemiology, etiology, and risk factors", section on 'Metabolic disorders')

●Menkes disease (see "Ischemic stroke in children and young adults: Epidemiology, etiology, and risk factors", section on 'Metabolic disorders')

●Cerebral autosomal dominant arteriopathy with subcortical infarctions and leukoencephalopathy (CADASIL) (see "Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL)")

●Cerebral autosomal recessive arteriopathy with subcortical infarctions and leukoencephalopathy (CARASIL) [51-53]

●Hereditary endotheliopathy with retinopathy, nephropathy, and stroke (HERNS) [54,55]

●Sickle cell disease (see "Acute stroke (ischemic and hemorrhagic) in children and adults with sickle cell disease")

●Mitochondrial encephalomyopathy with lactic acidosis and stroke-like episodes (MELAS) (see "Mitochondrial myopathies: Clinical features and diagnosis", section on 'MELAS')

It is important to note that all of these conditions together account for only a small percentage of ischemic strokes [56].

SUMMARY

●Under normal conditions, the rate of cerebral blood flow is primarily determined by the amount of resistance within cerebral blood vessels. Dilation of vessels leads to an increased volume of blood in the brain and increased cerebral blood flow, whereas constriction of vessels has the opposite effect. Cerebral blood flow is also determined by variation in the cerebral perfusion pressure. (See 'Cerebral autoregulation' above.)

●The brain is exquisitely sensitive to even short durations of ischemia. Multiple mechanisms are involved in tissue damage that results from brain ischemia. (See 'Consequences of reduction in blood flow during stroke' above.)

●Brain ischemia initiates a cascade of events that eventually lead to cell death, including depletion of adenosine triphosphate (ATP); changes in ionic concentrations of sodium, potassium, and calcium; increased lactate; acidosis; accumulation of oxygen free radicals; intracellular accumulation of water; and activation of proteolytic processes. (See 'Mechanisms of ischemic cell injury and death' above.).

●Cell death following cerebral ischemia or stroke can occur by either necrosis or by apoptosis. Low levels of ATP within the core infarct are insufficient to support apoptosis, and cell death occurs by necrosis. In the ischemic penumbra, ATP levels are sufficiently high that cell death by apoptosis can occur. As the duration of ischemia increases, however, ATP levels are eventually depleted and the proportion of cells that undergo necrosis is increased. (See 'Necrosis and apoptosis' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Daniel B Zemke, PhD, who contributed to earlier versions of this topic review.