Four major, interdependent mechanisms are involved in the control of cerebral blood flow: metabolic coupling: neural control, involving both extrinsic and intrinsic neural pathways: Pco2 and autoregulation. Although this division may be somewhat artificial and these control mechanisms probably operate in concert, it is useful to consider each separately.

Metabolic Control

Local cerebral blood flow (CBF) is regionally heterogeneous. The varied pattern of CBF is neither random nor related to the anatomic organization of the cerebral vasculature or to known differences in the innervation patterns of the cerebral vessels. Neuronal activity is the principal energy-consuming process in the brain. Local cerebral blood flow adjusts to the level of energy generation; therefore, it is the activity in the neuronal circuits that is the major determinant of variations and regional patterns of cerebral blood flow. Normally there is exquisite coupling between the regional cerebral metabolic demand for oxygen and glucose generated by local neuronal activity and the volume of blood flowing through that tissue. This coupling, termed metabolic regulation, was first demonstrated by simultaneous measurements of regional glucose metabolism and local blood flow in 1975, although indirect evidence supporting this mechanism had existed for many years. From their classic experiments in 1890, Roy and Sherrington noted that, the chemical products of cerebral metabolism contained in the lymph which bathes the wall of the arterioles of the brain can cause variations of the calibre of the cerebral vessels. In this reaction the brain possesses an intrinsic mechanism by which its vascular supply can be varied locally in correspondence with local variations of functional activity. While this doctrine was undisputed for nearly a century. the precise mechanisms responsible for this coupling have remained elusive. Alterations in the concentrations of local metabolites may lead to changes in regional CBF. Several chemical species capable of altering local vascular tone generated during periods of enhanced neuronal or glial activity have been considered as mediators of the coupling between flow and metabolism. These include extracellular pH, Pco2 adenosine, glycolytic intermediates, and extracellular potassium (indirectly, through its role in neuronal and smooth muscle cell membrane function, even though it is not directly involved in energy metabolism).

Silver has shown that local blood flow increases as soon as 1 s after neuronal excitation and that the zone of increase is limited to 250 µm around the site of the increased activity. These results indicate that flow can be adjusted very rapidly at the microvascular level according to metabolic demands in discrete functional subunits. There are a number of examples of increases in flow that are disproportionate compared to metabolism during activation. Such overcompensation has been demonstrated at both cellular and macroscopic levels.

A review of the information indicates that (1) the metabolic hypothesis of Roy and Sherrington does not fully explain the phenomenon of metabolic coupling. since CBF may increase out of proportion to metabolic demands: (2) to date, none of the candidates suggested as mediating this coupling demonstrates the necessary temporal profile between its accumulation in the perivascular space and the flow increase: and, finally, (3) these metabolites may be involved in the maintenance of flow and metabolism levels after this relationship has been set at a different level by a yet undetermined rapid initiator. such as a neurogenic stimulus.

Neurogenic Control

The association of metabolism and flow does not prove that metabolism determines flow. The two variables may be governed by a common third factor. The perivascular innervation is a candidate for playing such a role. It is important to consider not only the extrinsic nerve supply from the cranial ganglia to the cerebral arteries. arterioles, and veins, but also the role of intracerebral neurons serving the intracerebral vasculature. A dense plexus of nerve fibers in the walls of cerebral vessels. forming a "minibrain" or regulatory center. Given this arrangement, neurons could form the coupling mechanism between metabolism and flow. Although systemic administration of various neurotransmitter agonists and antagonists may not produce dramatic effects on cerebral blood flow, in experiments where the blood-brain barrier has been circumvented, marked changes are seen, again suggesting a more prominent role for neurotransmitter action from nerves that synapse directly on the cerebral vasculature. Extrinsic nerves, intrinsic nerves, and intrinsic brain regions all may bring their influence to bear on the cerebral vessels.

Theories of neurogenic control of the cerebral vasculature have focused on the role of efferent nerves that follow large arteries to innervate the cerebral vessels. Three types of extrinsic nerve systems, with distinct origins and neurotransmitters. have been identified. One consists of sympathetic neurons arising principally from the superior cervical ganglion. These neurons contain norepinephrine (NE) and neuropeptide-Y (NPY), which are both vasoconstrictors. A second system consists of parasympathetic neurons in the sphenopalatine and otic ganglia. which contain acetylcholine (ACh) and often coexpress vasoactive intestinal peptide (VIP). The third consists of sensory fibers originating in the trigeminal ganglion. These contain substance P (SP) and calcitonin gene­related peptide (CGRP), both of which are vasodilators.

Most of the neuron fibers investing the cerebral vasculature are sympathetic. They appear to function by reducing CBF under conditions where it has been increased by metabolic demand. and they may raise the threshold for the breakthrough of autoregulation that occurs with arterial hypertension. Attempts to manipulate CBF by either stimulation or ablation of sympathetic innervation have been largely unsuccessful. The parasympathetic nerves do not appear to play an important role in the tonic control of flow, but they may have some effect in pain-mediated vasodilatory responses.

Trigeminal nerves appear to become important only under special circumstances, such as hypertension and seizures, when their stimulation can effect a substantial increase in CBF. Despite the abundance of these nerve fibers, CBF appears to be primarily regulated by local metabolism with only minor modulation by extrinsic nerves. It is unclear how these peripheral neurons may contribute to the moment-by-moment governance of the cerebral circulation during normal activity.

The possibility that the brain could regulate its own blood flow was recognized by several early researchers in the field, though for many years most investigators did not consider this to be a significant regulatory mechanism. With the advent of both autoradiographic methods for the determination of regional CBF and positron emission tomography (PET) studies of cerebral circulation and metabolism in humans. as well as an explosion of techniques and knowledge in biochemical neuroanatomy, a growing body of evidence supports the concept that the brain can regulate its own blood flow through intrinsic neural networks. Cerebrovascular control may in part be regulated by intrinsic neural systems within the medulla oblongata and may not (as traditionally believed). depend entirely on responses by vessel walls and/or endothelium.

 

Carbon Dioxide (C02)

 

It has been well established that alterations in Pa_co2 result in marked vasodilation. There is an exponential relationship between Pa_co2, and CBF within a Pa_co2 range of 25 to 60 mmHg, with a CBF change of approximately of 4 percent per millimetre of mercury.

Flow changes induced by alterations in Pa_co2 occur within 2 min and reach a new plateau within 12 min. This regulatory mechanism has been shown to be a function of changes in the perivascular pH in the vicinity of the vascular smooth muscle cells, rather than a direct effect of CO2 per se. In addition to the direct effects of hydrogen ions on the vascular smooth muscle, local changes in pH can modulate the vasomotor responses to other agents that affect vessel calibre, such as norepinephrine. Since changes in systemic Pa_co2 are detected by carotid artery chemoreceptors, this regulatory mechanism can be effected by reflex pathways. In support of this. it has been observed that a lesion of the tegmental reticular formation diminishes the cerebrovascular response to alterations in Pa_co2. Investigations has suggested that the powerful effects of CO2 on the cerebral circulation are mediated by endothelium-derived relaxing factor.

Prolonged alterations in Pa_co2 result in chronic adaptation, and after approximately 36 h the blood flow changes tend to return to prealteration levels. At Pa_co2 levels of 70 mmHg, maximal vasodilation has occurred and CBF does not increase as Pa_co2 increases further. Similarly, Pa_co2 levels less than 20 mmHg cause no further decrease in CBF. These low Pa_co2 levels should be avoided in the clinical setting, since the ensuing blood flow reductions can lead to tissue ischemia.

Autoregulation

Autoregulation is defined as the physiologic maintenance of a constant flow over a moderate range of perfusion pressures. This restricted use of the term autoregulation, in contrast to the broader meaning of the term - the capacity of an organ to regulate its blood supply in accordance with its underlying functional or metabolic needs - avoids confusion with other mechanisms involved in cerebrovascular regulation, such as metabolic coupling.

According to the general equation of flow, CBF can be described by the relationship between cerebral perfusion pressure (CPP) and cerebrovascular resistance (CVR):

CBF = CPP/CVR

Cerebral perfusion pressure is equal to mean arterial blood pressure (MABP) [where MABP = 1/3(systolic pressure - diastolic pressure) + diastolic pressure] minus intracranial pressure and sagittal sinus pressure. In the absence of pathologic conditions, intracranial pressure and sagittal sinus pressure are negligible compared to systemic arterial pressure, and CPP is roughly equivalent to MABP. According to the previous equation, autoregulation must be mediated by changes in CVR. The Hagen-Poiseuille equation, which describes the flow of Newtonian fluids in rigid tubes, offers an approximation of the factors that govern CVR and suggests that resistance is inversely proportional to blood viscosity and proportional to the fourth power of eheradius of the vessel. Thus, changes in the radius of cerebral blood vessels can produce marked alterations of CVR. A decrease in CPP produces dilation of the precapillary resistance vessels, whereas an increase produces constriction. Largely by variation in the degree of constriction of the cerebral resistance vessels, average hemispheric CBF is maintained at a fairly constant level, near 50 ml/100 g per minute in the adult human brain at rest.

Although myogenic, neurogenic, and metabolic mechanisms have been postulated, the precise control of the autoregulatory response remains unknown. Pioneering work done by Bayliss on the myogenic basis of autoregulation showed that reflex changes in the tone of arteriolar smooth muscle are elicited by changes in transmural pressure. According to this hypothesis, an increase in the transluminal pressure leads to stretching of smooth muscle within the vessel wall. Reflex contraction of radial fibers then results in constriction of vessel diameter, and an opposite effect is seen with a decrease in transluminal pressure.

 

A growing body of evidence suggests that endothelium-dependent mechanisms function as the primary mediating factor of vascular tone, and they are now considered as a facet of the myogenic hypothesis of autoregulation. The endothelium acts as a transducer of hemodynamic forces that lead to the release of vasoactive substances. Synthesis of the endothelium-derived relaxing factor, either nitric oxide (EDRF/NO) or a closely related molecule derived from the amino acid L-arginine, appears to affect vascular tone, both under basal conditions and in response to the application of specific agonists. The proposed mechanism for this effect is that stimulation of soluble guanylate cyclase by EDRF/ NO raises the level of cyclic guanosine monophosphate (cGMP) in vascular smooth muscle and results in vascular relaxation. Dilation of large cerebral arteries and pial arterioles in response to the application of acetylcholine in vivo is dependent on the formation of NO by nitric oxide synthase and can be blocked by a competitive antagonist of that enzyme, N-monomethyl-L-arginine (L-NMMA). Intravenous administration of a similar NO synthase antagonist in rats caused a 40 percent increase in MABP and a 60 percent reduction in the lumen diameter of pial arteries.  This L-arginine/NO/cGMP pathway appears to be critical in the control of vascular tone and is increasingly accepted as the dominant mediator of the autoregulatory response.

 

In the brain, autoregulation is manifest as the lack of major fluctuation in CBF despite changes in mean arterial blood pressure

between 60 and 150 mmHg. Cerebral autoregulation may be thought of as a homeostatic mechanism that is superimposed on the baroceptive reflexes. It is important to stress that both the upper and the lower limits of autoregulation can be affected by many factors, including sympathetic nerve activity, Pa_co2 and pharmacologic agents. The most important factor that can affect autoregulation is chronic arterial hypertension. As a result of thickening of the cerebral arteries, the upper and lower limits of autoregulation are both displaced to higher levels in patients with chronic hypertension. The consequence of these alterations is that symptoms of cerebral hypoperfusion can occur at higher values of mean arterial pressure in patients with chronic hypertension than in normotensive individuals.

 

A knowledge of the cerebrovascular status of the patient with respect to hypertension may also be a consideration in the planning of temporary vessel occlusion during cerebrovascular surgery. Some experimental evidence suggests that intermittent temporary occlusion is less damaging than a single sustained episode. Other experimental work directed specifically at determining the response in both normotensive and hypertensive animals demonstrates that in the latter group, intermittent occlusion was associated with a greater degree of ischemic injury.

Cerebral Blood Flow and Ischemic Thresholds

The unique metabolic requirements of the brain form the central basis for understanding the relationship between blood flow and ischemic tolerance. Although the brain represents only 2 percent of the total body weight, it receives 18 percent of the cardiac output and uses 20 percent of the oxygen supply. Its high metabolic demand and lack of appreciable energy reserves render the central nervous system uniquely susceptible to alterations of blood supply. The use of PET has greatly enhanced our understanding of the pathophysiologic alterations that occur in focal cerebral ischemia in humans. The simultaneous measurement of regional CBF, oxygen metabolism (CMRO2), oxygen extraction fraction (OEF, or amount of oxygen extracted from the blood as it travels from artery to vein), and cerebral blood volume (CBV, or the volume of blood in the cerebral parenchyma) has permitted the identification of three successive stages of severity in an ischemic injury to the brain. As CPP initially falls. autoregulation occurs and vasodilation of precapillary resistance vessels causes an increase in CBV while maintaining both CBF and CMRo2 At the lower limit of autoregulation, maximal compensatory dilation of cerebral resistance vessels has occurred. Further reductions in CPP lead to a fall in CBF. The oxygen extraction fraction (OEF) then increases to maintain CMRo2. If CBF reduction is modest, the increased extraction of oxygen and glucose by the brain from remaining blood flow can maintain normal brain metabolism and function. A leftward shift of the oxygen-haemoglobin affinity curve, which is produced by the decreased local pH. results in an increased transfer of oxygen from blood to tissue and is largely responsible for the increased OEF. When the OEF reaches its maximum (approximately 90 percent). no further compensation can occur. As CBF falls further, the metabolic demands of the brain can no longer be satisfied, and CMRo2 decreases.

The alterations of cerebral function that occur as these compensatory systems become overwhelmed in the face of increasingly severe ischemia may be best understood in the context of ischemic thresholds and the ischemic cascade.

Flow Thresholds

Electrophysiologic techniques and accurate cerebral blood flow determinations have refined our understanding of the relationship between neuronal function, tissue viability and critical levels of regional cerebral blood flow.

Experimental studies of middle cerebral artery occlusion in various species have demonstrated a blood flow gradient from normal flow in areas outside the affected territory, to modest decreases in the adjacent perifocal region, to a profound drop in the ischemic core. The slope of this gradient depends on the extent and functional capacity of collateral blood supply. While some authors have attempted to make a distinction between the events occurring in the ischemic core and those that occur in tissue affected by a global ischemic insult, on a biochemical basis it is difficult to clearly dissociate the events occurring in these two environments. The tissue in the border zone between normal perfusion and the ischemic core-the perifocal region-may be subjected to unique challenges to its homeostatic mechanisms, however.

Different cellular functions, which require specific minimum levels of blood flow, are affected in these regions depending on the level of blood flow reduction. Certain functional perturbations occur once blood flow decreases below these thresholds. Critical values for loss of synaptic transmission, corresponding to loss of neuronal function, are between 15 and 18 ml/100 g per minute. The threshold for membrane pump failure. and thus for loss of cellular integrity, is approximately 10 ml/100 g per minute. The level of blood flow reduction for ion pump failure appears to be similar to that for energy failure. The presence of these two distinct thresholds implies that some regions in the perifocal area contain cells that are electrophysiologically quiescent but nonetheless viable. These regions constitute the ischemic penumbra, defined as areas with EEG quiescence and low extracellular K⁺. These thresholds were determined in experimental models using both primates and other higher vertebrates. Similar values have been reported in humans. While absolute values may vary somewhat depending on the species and anaesthetic factors, the percent reduction from normal flow to these thresholds appears to be uniform and constant.

Flow reduction is one component that determines the severity of an ischemic insult, but the duration of flow reduction is also of paramount importance. The threshold for infarction in monkeys is approximately 12 ml/100 g per minute, but that the duration as well as the degree of blood flow reduction was important, since infarction developed only if blood flow was reduced to below 12 ml/100 g per minute for periods lasting 2 h or longer. Since the time course for irreversible damage in complete global ischemia models is much shorter-approximately 10 min-it is reasonable to suspect that areas with more profound blood flow reduction in focal ischemia have a shorter tolerance than areas with higher levels of blood flow.

The existence of two distinct thresholds suggests that some areas in the perifocal region contain cells that are electrically silent but nonetheless viable. These cells are the likely targets for prevention of ischemic injury, since they should be the most susceptible to therapeutic rescue. The ability to maintain a low extracellular potassium concentration in the perifocal region implies that sufficient energy stores remain to maintain near-normal electrochemical gradients. but the neuronal paralysis and reduced blood flow suggest that the penumbra is clearly at risk for further damage. SiesjÖ has applied a pragmatic definition to this region by defining a reperfusion penumbra and a pharmacologic penumbra, which represent the tissue that would inevitably become infarcted without the timely institution of either reperfusion or pharmacologic intervention.

Therapeutic Manipulation of Residual Flow and Alteration of Flow Thresholds

The Role of Blood Viscosity

The importance of blood viscosity in the routine regulation of CBF remains controversial. Several studies suggest that in normal brain the effect of blood viscosity on cerebral perfusion is nonexistent or minimal. Under ischemic conditions. however. even small alterations in the rheologic properties of blood may have significant functional relevance.

This selective contribution of viscosity to the regulation of CBF under impaired flow conditions can be explained by the inconstant nature of blood viscosity, which results from erythrocyte deformability and aggregation. Therefore, the relationship between blood flow and viscosity is imprecisely described by the Hagen­Poiseuille equation. especially at low flow rates. In low-flow states, perfusion pressure is reduced and compensatory vasodilation of the microcirculation occurs. Under these conditions, blood flow is further reduced by an increase in viscosity. Thus, under ischemic conditions an intricate relationship exists between vasomotor compensatory mechanisms and blood viscosity, and even small alterations in the rheologic properties of blood have significant functional relevance.

Blood viscosity is determined by several factors. hematocrit being the most important, especially when shear rates are low. The steep portion of the hematocrit-viscosity curve falls in the physiologic range of hematocrit. Reductions in hematocrit within the physiologic range significantly reduce blood viscosity and this effect is most marked at the low shear rates seen in focal cerebral ischemia.

Experimental studies have shown augmentation of diminished CBF following the acute reduction of hematocrit. Although reduction of hematocrit (and thus haemoglobin content) reduces the oxygen content of blood. relative oxygen transport capacity has been calculated to increase owing to improved CBF with hematocrit reductions to approximately 30 percent. Below this hematocrit level, the decrease in blood oxygen content outweighs the beneficial effect of decreased viscosity on CBF in the microcirculation. This finding is consistent with reports that indicate that a hematocrit range of 30 to 32 percent is optimal for tissue oxygen delivery.

 

The Role of Reperfusion

 

The results of experimental work with reperfusion suggest that there is a definite time limit after which such restoration of flow will not be beneficial. Whether such treatment is in fact harmful is a matter of some debate. That reperfusion can lead to aggravation of edema and hemorrhagic transformation has been clearly documented. On a cellular and pathophysiologic level, it has also been shown that "reperfusion injury" occurs in some organs. This type of injury is speculated to occur in the brain. It has nevertheless been shown that reperfusion achieved by the endovascular route in the early evolution of ischemic damage in the clinical setting does improve neurological outcome and offers a promise of revolutionizing the treatment of stroke.

 

The Role of Calcium Homeostasis

 

While there are many promising areas of investigation into the pharmacologic control of the cerebral circulation, including adrenergic mechanisms and the roles of dopamine, serotonin, acetylcholine, histamine, prostaglandins, neuropeptides, and glutamate, few have generated as much interest among neurosurgeons as the possible uses of calcium antagonists.

The role of calcium in the control of both cerebrovascular smooth muscle function and intracellular homeostasis offers opportunities for the treatment of ischemia and ischemia-producing disorders. The pharmacologic effects exerted by calcium entry blockers on vascular reactivity through an effect on excitation­contraction coupling involve three main processes: the influx of extracellular calcium or release of calcium from intracellular stores. the transport of calcium out of the cells and  uptake into various cell organelles and the processes that are regulated by the intracellular concentration of free calcium and that affect the activity of contractile proteins themselves or other cellular processes.