Cerebral Hemodynamics

Brain blood flow can be modeled from a physical standpoint as flow in a tube with the assumptions that flow is steady, laminar, and uniform through thinned-walled (the wall is <10% of the lumen) non-distensible tubes [87]. These assumptions do not apply to large arteries that have thick walls or in the microcirculation in which flow is non-Newtonian [161]. Ohm’s law states that flow is proportional to the difference in inflow and outflow pressure (ΔP) divided by the resistance to flow (R): flow = ΔP/R. In the brain, ΔP is cerebral perfusion pressure (CPP), the difference between intra-arterial pressure and the pressure in veins. Venous pressure is normally low (2–5 mmHg) and is influenced directly by intracranial pressure (ICP). Therefore, ΔP is calculated as the difference in CPP and either venous pressure or ICP, whichever is greater. Blood flow is also estimated by Poiseuielle’s law that states that flow is directly related to ΔP, blood viscosity, and the length of the vessel (assumed to be constant) and inversely related to radius to the fourth power: flow = (8 × η × L)/r⁴ [136]. Thus, radius is the most powerful determinant of blood flow and even small changes in lumen diameter have significant effects on cerebral blood flow, and it is by this mechanism that vascular resistance can change rapidly to alter regional and global cerebral blood flow [137].

Autoregulation of Cerebral Blood Flow

Autoregulation of cerebral blood flow is the ability of the brain to maintain relatively constant blood flow despite changes in perfusion pressure [137]. Autoregulation is present in many vascular beds, but is particularly well-developed in the brain, likely due to the need for a constant blood supply and water homeostasis. In normotensive adults, cerebral blood flow is maintained at ~50 mL per 100 g of brain tissue per minute, provided CPP is in the range of ~60 to 160 mmHg [138]. Above and below this limit, autoregulation is lost and cerebral blood flow becomes dependent on mean arterial pressure in a linear fashion [71,72,139]. When CPP falls below the lower limit of autoregulation, cerebral ischemia ensues [27,140]. The reduction in cerebral blood flow is compensated for by an increase in oxygen extraction from the blood [141]. Clinical signs or symptoms of ischemia are not seen until the decrease in perfusion exceeds the ability of increased oxygen extraction to meet metabolic needs. At this point, clinical signs of hypoperfusion occur, including dizziness, altered mental status, and eventually irreversible tissue damage (infarction) [140,141].

The mechanisms of autoregulation in the brain are not completely understood and likely differ with increases vs. decreases in pressure. Although a role for neuronal involvement in autoregulation is appealing, studies have shown that cerebral blood flow autoregulation is preserved in sympathetically and parasympathetically denervated animals, indicating that a major contribution of extrinsic neurogenic factors to autoregulation of cerebral blood flow is unlikely [70] (see Perivascular Innervation). Recently, a role for neuronal nitric oxide in modulating cerebral blood flow autoregulation has been shown, suggesting that although extrinsic innervation may not be involved, intrinsic innervation may have a role [62]. Biproducts of metabolism have also been proposed to have a role in autoregulation [142]. Reductions in cerebral blood flow stimulate release of vasoactive substances from the brain that cause arterial dilatation. Candidates for these vasoactive substances include H⁺, K⁺, O₂, adenosine, and others. Autoregulation of cerebral blood flow when pressure fluctuates at the high end of the autoregulatory curve is most likely due to the myogenic behavior of the cerebral smooth muscle that constrict in response to elevated pressure and dilate in response to decreased pressure [68,69–71]. The important contribution of myogenic activity to autoregulation is demonstrated in vitro in isolated and pressurized cerebral arteries that constrict in a response to increased pressure and dilate in response to decreased pressure [71,105] (see Myogenic Response). Autoregulation at pressures below the myogenic pressure range likely involves hypoxia and release of metabolic factors [68].

The importance of autoregulation in normal brain function is highlighted by the fact that significant brain injury occurs when autoregulatory mechanisms are lost. For example, during acute hypertension at pressures above the autoregulatory limit, the myogenic constriction of vascular smooth muscle is overcome by the excessive intravascular pressure and forced dilatation of cerebral vessels occurs [143–146]. The loss of myogenic tone during forced dilatation decreases cerebrovascular resistance, a result that can produce a large increase in cerebral blood flow (300–400%), known as autoregulatory breakthrough [143–146] (Figure 16). In addition, decreased cerebrovascular resistance increases hydrostatic pressure on the cerebral endothelium, causing edema formation [143–145], the underlying cause of conditions such as hypertensive encephalopathy, posterior reversible encephalopathy syndrome (PRES), and eclampsia [143,147] (see Vasogenic Edema Formation).

FIGURE 16

Tracing of CBF (in laser Doppler units) and ABP (in mmHg) in response to increasing doses of PE. In this experiment, CBF increased four times greater than baseline as ABP increased from 140 to 210 mmHg, demonstrating autoregulatory breakthrough. Used (more...)

Although uncommon since the advent of effective antihypertensive therapy, hypertensive encephalopathy occurs as a result of a sudden, sustained rise in blood pressure sufficient to exceed the upper limit of cerebral blood flow autoregulation (>160 mmHg) [148–150]. Early studies on the reaction of cerebral vessels to high blood pressure produced the concept of hypertensive vasospasm. Acute hypertensive encephalopathy was thought the result of spasm—;defined as an uncontrolled vasoconstriction—;of the cerebral arteries, causing brain tissue ischemia [151,152]. This concept originated from the observations of Byrom [151] who produced experimental renal hypertension and found ~90% of hypertensive rats with neurologic manifestations showed multiple cortical spots of trypan blue extravasation, whereas rats without cerebral symptoms appeared to have normal cerebrovascular permeability. He also noted what he called an alternating vasoconstriction/vasodilation in the pial vessels, a phenomenon known as a “sausage-string” appearance. This observation led him to the conclusion that cerebral vasospasm caused ischemia and edema formation in response to acute hypertension. Byrom later modified his view and referred to a finding in the mesenteric circulation that vessels with this “sausage-string” appearance had protein leakage in the dilated parts of the vessels only [153,154]. Since then, it has been established that high blood pressure results in increased cerebral blood flow and “breakthrough of autoregulation” [155]. Further experiments confirmed that loss of myogenic vasoconstriction during forced dilatation rather than spasm is the critical event in hypertensive encephalopathy [156].

Segmental Vascular Resistance

In peripheral circulations, small arterioles (<100 μm diameter) are typically the major site of vascular resistance (157). However, in the brain, both large arteries and small arterioles contribute significantly to vascular resistance. Direct measure of the pressure gradient across different segments of the cerebral circulation found that the large extracranial vessels (internal carotid and vertebral) and intracranial pial vessels contribute ~50% of cerebral vascular resistance [58,158]. Large artery resistance in the brain is likely important to provide constant blood flow under conditions that change blood flow locally, e.g., metabolism. Large artery resistance also attenuates changes in downstream microvascular pressure during increases in systemic arterial pressure. Thus, segmental vascular resistance in the brain is a protective mechanism that helps provide constant blood flow in an organ with high metabolic demand without pathologically increasing hydrostatic pressure that can cause vasogenic edema.

Neural–Astrocyte Regulation

Unlike pial arteries and arterioles, parenchymal arterioles are in close association with astrocytes and, to a lesser extent, neurons. Both these cell types may have a role in controlling local blood flow [2,12,22,32]. Subcortical microvessels are innervated from within the brain parenchyma and are unique in that the majority of vericosities adjoin astrocytic end-feet surrounding arterioles and thus does not have conventional neurovascular junctions [135]. Neurons whose cell bodies are from within the subcortical brain regions (e.g., nucleus basalis, locus ceruleus, raphe nucleus) project to cortical microvessels to control local blood flow by release of neurotransmitter (e.g., ACH, norepinephrine, 5HT) [22] (Figure 17). Release of neurotransmitter stimulates receptors on smooth muscle, endothelium, or astrocytes to cause constriction or dilation, thereby regulating local blood flow in concert with neuronal demand [22,98,134]. It has been known for some time that astrocytes can release vasoactive factors [159]. Evidence for the involvement of astrocytes in local control of blood flow in vivo has recently emerged. Their close apposition to microvessels, encasing almost the entire parenchymal arterioles and capillaries with little neuronal contact, makes astrocytic involvement likely at this level [21,22,98,134]. Studies in brain slices, in which the entire neurovascular unit is intact, showed that direct electrical stimulation of neuronal processes raises calcium in astrocytic end-feet and causes dilation of nearby arterioles [160]. Stimulation of astrocytes also raises calcium in end-feet and has a similar vasoactive effect on parenchymal arterioles; however, whether dilation or constriction occurs seems to depend on the level of calcium and, not surprisingly, resting tone [161]. It has been proposed that an elevation in astrocyte calcium releases vasoactive factors, including K⁺, 20-HETE, and PGE₂ [160–162]. However, a weakness of the brain slice preparation is that it does not allow for arterioles to be pressurized or have flow. Thus, the role of the myogenic response, which may significantly modify any astrocytic-derived signals in vivo is not known.

FIGURE 17

Summary of the regulation of cortical microvessels from cells located in subcortical areas and within the cerebral cortex. The possibility that interneurons also induce the release of vasoactive molecules from astrocytes is not included for clarity purposes. (more...)

Effect of Oxygen

The brain has a very high metabolic demand for oxygen compared to other organs, and thus, it is not surprising that acute hypoxia is a potent dilator in the cerebral circulation that produces marked increases in cerebral blood flow [163]. In general, blood flow does not change in the brain until tissue PO₂ falls below ~50 mmHg, below which cerebral blood flow increases substantially [163]. As hypoxia decreases PO₂ further, cerebral blood flow can rise up to 400% of resting levels [164]. Increases in cerebral blood flow do not change metabolism, but hemoglobin saturation falls from ~100% at PO₂ >70 mmHg to ~50% at PO₂ <50 mmHg [164]. Acute hypoxia causes an increase in cerebral blood flow via direct effects on vascular cells of cerebral arteries and arterioles. Hypoxia-induced drop in ATP levels opens K_ATP channels on smooth muscle, causing hyperpolarization and vasodilation [165]. In addition, hypoxia rapidly increases nitric oxide and adenosine production locally, also promoting vasodilation [166]. Chronic hypoxia increases cerebral blood flow through an effect on capillary density [16–19] (see Microcirculation and Neurovascular Unit).

Effect of Carbon Dioxide

Carbon dioxide (CO₂) has a profound and reversible effect on cerebral blood flow, such that hypercapnia causes marked dilation of cerebral arteries and arterioles and increased blood flow, whereas hypocapnia causes constriction and decreased blood flow [167,168]. The potent vasodilator effect of CO₂ is demonstrated by the finding that in humans 5% CO₂ inhalation causes an increase in cerebral blood flow by 50% and 7% CO₂ inhalation causes a 100% increase in cerebral blood flow [168]. Although several mechanisms involved in hypercapnic vasodilation have been proposed, the major mechanism appears to be related to a direct effect of extracellular H⁺ on vascular smooth muscle [169]. This is supported by findings that neither bicarbonate ion nor changes in PCO₂ alone affect cerebral artery diameter [170]. Other proposed mechanisms involved in the response to changes in PCO₂ include vasodilator prostanoids and nitric oxide; however, the involvement of these mediators appears to be species-specific [171,172].