Regulation of Cerebral Blood Flow: Estimation Methods
in Neurosurgery

Semenyutin V.B., Svistov D.V.

Polenov Research Neurosurgical Institute, Saint Petersburg, Russia.
Medicomilitary Academy, Saint Petersburg, Russia.

The main purposes of the system of cerebral circulation consist in minimizing deviations of circulatory and chemical homeostasis of the brain in different functional states. Thus, a process, regulating cerebral blood flow, is assumed to have a complex structural-and-functional organization. Today the interaction of three basic mechanisms of its regulation (myogenic, metabolic and neurogenic) is a generally acknowledged fact.

Autoregulation is one of the fundamental properties of cerebral circulation. Being of extreme importance for adequate cerebral blood supply, it is characterized by an ability of cerebral vessels to preserve a relatively stable volumetric velocity of cerebral blood flow in changes of perfusion pressure (difference between systemic blood and intracranial pressures) within broad ranges (from 50 up to 170 mm Hg). When perfusion pressure oversteps these limits, one can watch autoregulation breakdown and linear dependence of cerebral blood flow on pressure changes. Intensity of cerebral blood flow is a relatively constant value within the above limits of autoregulation and its changes are insignificant (Fig. 1). A pressure range, characterized by stable blood flow, represents a quantitative state of autoregulation, whereas blood flow changes within this range reflect its qualitative parameters.

Fig. 1. Dependence of total cerebral blood flow (TCBF) on perfusion pressure (PP).
1, 2 are the lower and upper limits of cerebral autoregulation respectively.

Perfusion pressure changes, caused by fluctuations of blood and intracranial pressures and watched in a postoperative period, are a frequent phenomenon among neurosurgical patients. These fluctuations lead to brain hypoxia and edema in disorders of cerebral auroregulaton. Its impairment in ischemic and hemorrhagic strokes, craniocerebral trauma is the main link of pathogenesis of cerebral circulation disorders in brain pathology. An analysis of a state of autoregulation mechanisms is of great practical value. Their knowledge is necessary for solving such important problems as support of an optimum level of blood and intracranial pressures during operations and in a postoperative period and providing optimum vasoactive therapy. In case of local disturbances of cerebral autoregulation and focal cerebral lesions, regional blood flow depends on a ratio between values of intravascular and local tissue pressures. A normal value of intracranial pressure is known to be only some mm Hg; thus, even a small increase of tissue pressure results in a sharp decrease of tissue blood flow. Irregularity of tissue pressure in focal lesions conditions irregularity of brain perfusion. It aggravates disorganization of cerebral circulation.

The above peculiarities of structural-and-functional organization of the cerebral circulation system determine informative significance of these or those indices of its functioning in different neurosurgical pathology to a considerable extent. More precise information on functional goals of the system, responsible for regulation of cerebral circulation in a specific type of incoming disturbances, provides a conceptual basis for development and introduction of methods, which would allow to estimate a state of the cerebral circulation system under conditions of clinical practice.

Nowadays cerebrovascular reactivity is considered to be an integral index of adaptation potentialities of the cerebral circulation system, ability of cerebral vessels to respond to changing conditions of functioning and to optimize blood flow in compliance with them.

As for reactivity of cerebral vessels, the essence of this notion lies in possible objectification of the activity of regulatory mechanisms, guiding cerebral circulation and ensuring its functional stability, which is achieved by use of special stimuli. Their intensity and duration should be dosed; they should be induced and abolished quickly, simulate natural disturbances, suffered by the cerebral circulation system under physiologic conditions, and have no cumulative effect.

Estimation of reactivity indices is based on relative parameters (blood flow changes), so it is not necessary to use methods of measuring volumetric characteristics of blood flow for reactivity determination. Physiologic test loads are known to lead to inconsiderable changes of a lumen of basal arteries in comparison with a diameter of smaller arteries. Taking this fact into account, transcranial Doppler (TCD) may be regarded as a quite sufficient method of indirect estimation of cerebral hemodynamics. Thus, dynamics of linear blood flow velocity (LBFV) in basal segments reflects mainly changes of volumetric blood flow in regions of appropriate arteries.

Reactivity of the cerebral circulation system is estimated in clinical practice on the basis of hypercapnic (inhalation of 4-8% carbogen, voluntary breath holding, apneic oxygenation) and hypocapnic (spontaneous/induced hyperventilation) functional tests of a chemical nature.

Estimation of cerebral autoregulation is carried out with the help of tests, having a physical nature (compression of the common cerebral artery, induced non-pharmacological hypotension). Orthostatic and antiorthostatic loads, Valsalva test are used less frequently.

Pharmacological tests with I/V administration of diamox/acetazolamide (1 g), sublingual taking of nitroglycerin (0.25 mg) are spread in clinical practice too.

Hypercapnic tests trigger dilation reserve of resistive vessels, which results in decrease of circulatory resistance in a vascular region and increase of volumetric blood flow, linear blood flow velocity in basal arteries. Hypercapnia is achieved with the help of different methods. Inhalation of 4-8% CO2 was considered a golden standard of reactivity estimation for a long time. An alternative of inhalation load is closed-contour breathing with constant introduction of oxygen at a speed of 1 l/min. The main advantages of such loads are as follows: CO2 is a natural information carrier in the cerebrovascular system; the test is short-term; blood is saturated by gas very quickly; a response of vessels is determined. However, they have several drawbacks: an examined person with hypercapnia has such unpleasant sensations as lack of air, blood rush to a head; there can be unspecific reactions of respiratory and circulatory systems, which can mask a response to a load.

Inhalation loads demand technical support, that is why a test with voluntary breath holding (in normal inspiration, deep inspiration, expiration, after intensive inspiration-expiration) or hypoventilation is often used for estimation of a response of resistive vessels to hypercapnia. A vascular response develops during 20-30 sec of apnea due to accumulation of endogenic carbon dioxide under conditions of temporary ventilation discontinuance. One should use this test with great care in patients with chronic respiratory and cardiovascular diseases and limited reserves of main systems of life support. As for patients with artificial pulmonary ventilation, hypercapnic load can be achieved by switching off an apparatus for some time, decrease of a respiratory rate. Being the safest test, apneic oxygenation is at the same time technically difficult and time-consuming. It consists in switching off an apparatus for artificial pulmonary ventilation under conditions of continuous administration of oxygen at a speed of 30 l/min into the trachea.

The first step in reactivity estimation is recording of LBFV in arteries under examination at rest (a patient is in a recumbent position). After loading repeated recording of maximum values of mean LBFV is carried out. If calculation of a reactivity index demands information on a specific level of saturation with CO2, capnographs are used for estimation of pCO2 in the expired air, as it correlates with pCO2 in blood.

A test with I/V administration of acetazolamide (1 g) is a simple method, which allows to obtain comparable data. Being an inhibitor of erythrocytes carboanhydrase, this drug impairs balance of the CO2 buffer system and leads to accumulation of endogenic carbon dioxide. Its side effects and impact on the cardiovascular system are minimum. It appears, that effective concentration of CO2 is achieved on the 15-20th minute after administration, when one can watch maximum values of cerebral blood flow. An effect of acetazolamide is not inferior to that of carbon dioxide and causes increase of arterial LBFV by 35-42%. The test short-comings are necessity of intravenous injection and a response peak, observed in 15-20 minutes.

The above functional loads are reproduced and compared with each other quite easily. Thus, every researcher can choose an optimum type of load in compliance with his needs.

Hypocapnia due to spontaneous or induced hyperventilation results in narrowing of resistive vessels, increase of vascular resistance, reduction of blood flow and LBFV in basal arteries by 40-55%. Dependence of LBFV on pCO2 is exponential both in hypo- and hypercapnia. It is necessary to use a minimum stationary value of LBFV for calculating a reactivity index. A mean duration of hyperventilation, sufficient for revealing a response, is 25-30 sec; a rate of respiratory movements during this period is 60/min-1.

Reactivity indices, i.e. qualitative characteristics of a state of the system, regulating cerebral circulation, are calculated on the basis of results of functional tests. They include:

- A coefficient of reactivity in hypercapnic load (calculation is based on results of any tests with CO2 inhalation, breath holding, acetazolamide administration):

(Lindegaard K.-F. et al., 1986),

- A reactivity index in hypercapnic load:

(Lindegaard K.-F. et al., 1986),

- A normalized autoregulatory response:

(Widder B. et al., 1986),

- A coefficient of reactivity in hypocapnic load:

(Lindegaard K.-F. et al., 1986),

- A reactivity index in hypocapnic load:

(Lindegaard K.-F. et al., 1986),

- An index of vasomotor reactivity:

(Ringelstein al., 1988),

V is mean maximum LBFV at rest (cm/s),
V + is mean maximum LBFV against a background of hypercapnic load (cm/s),
V - is mean maximum LBFV against a background of hypocapnic load (cm/s),
20 is CO2 tension in the expired air at rest (mm Hg),
2+ is the same parameter against a background of hypercapnic load (mm Hg),
2- is the same parameter against a background of hypocapnic load (mm Hg),
20 is CO2 concentration in the expired air at rest (%),
2+ is the same parameter against a background of hypercapnic load.

Normal parameters of reactivity are given in the Table.

The Table

Reactivity of the Cerebral Circulation System and Its Parameters (M±SD)

An Examined Vessel

Age (years)

Cr+ (%)

Cr- (%)

RI+ (%)

RI- (%)

NAR (%)

IVMR (%)


< 40
> 40

43 ± 4
37 ± 6

55 ± 4
42 ± 2

3.1 ± 0.1
2.6 ±0.2

-3.4 ± 0.1
-2.6 ±0.2

24.6 ± 2.3
17.1 ±1.9

96.8 ± 10.1
79.8 ±11.3


< 40
> 40

36 ± 3
29 ± 6

44 ± 9
42 ± 11

2.7 ± 0.1
2.5 ± 0.1

-3.2 ± 0.1
-2.9 ± 0.2

21.4 ± 3.6
16.6 ± 2.7

80.1 ± 9.4
71.6 ± 10.4


< 40
> 40

51 ± 7
38 ± 12

59 ± 11
50 ± 8

3.2 ± 0.1
2.4 ± 0.2

-3.4 ± 0.1
-2.8 ± 0.1

26.7 ± 4.1
19.0 ± 2.2

90.6 ± 12.4
88.6 ± 11.7

A change of reactivity indices can be both a diagnostic sign and reflection of a functional state of the cerebral circulation system. Acute or chronic cerebral ischemia leads to reduction of reactivity indices in hypercapnic load, which is indicative of a limited functional reserve. As a rule, if they decrease two and more times, one can think of an organic lesion of the brain or its vascular system. Reduction of a response to hypocapnic load is watched in pathologic arteriovenous anastomosis, characteristic of arteriovenous cerebral malformations. When an amplitude of responses to hyper- and hypocapnic loads is approximately equal, it tells of neutral/normal tonicity of resistive vessels. Cerebral angiodystonia is characterized by impaired correlation of response amplitudes, i.e. a response to hypocania is dominant in dystonia of a hypotonic type, whereas a response to hypercapnia dominates in dystonia of a hypertonic type. At the same time the whole range of a response, represented by a value of IVMR, is normal. Narrowing of a homeostatic range (reduction of IVMR up to 50%) is an indication of considerable decrease of reactivity, conditioned, as a rule, by an organic lesion of the brain and its vascular system.

A state of cerebral autoregulation can be estimated by a degree, to which hyperemic changes are marked after more or less durable regional hypotension, induced by temporary occlusion of a major vessel. Compression of the common carotid artery in a neck area causes reduction of perfusion pressure in internal carotid and middle cerebral arteries by 52±11.4% and 32.2±16.3%, respectively. Transient increase of blood flow due to compensatory vasodilation is watched after compression discontinuance. It can be used as an autoregulation indicator. A transient hyperemic response in the MCA, which demonstrates itself in a short-term growth of LBFV, allows to calculate an index, characterizing vasodilation in response to transient reduction of perfusion pressure.

Recording of background LBFV in the MCA is carried out at rest with the purpose of estimating results of a compression test. Then, careful compression of the ipsilateral common carotid artery is performed during 5 cardiac cycles, which results in reduction of LBFV up to 57±11% from the initial value. Compression is discontinued in a diastole of the 5th complex and LBFV is recorded during 3-5 cardiac cycles. In a normal state compression completion is followed by a marked growth of LBFV (overshoot). It appears against a background of stable indices of central hemodynamics. In other words, it is realized by cerebral mechanisms only. Taking into account, that indices of peripheral resistance during overshoot are reliably lower than initial values, one can suppose it to result from reduction of circulatory resistance in the MCA region in response to decrease of perfusion pressure.

Fig. 2. A spectrogram of flow in the MCA during a compression test.
V1 mean LBFV at rest, V2 mean LBFV of the first spike after compression discontinuance.

An overshoot coefficient (OC) is calculated according to the following formula:

A normal value of OC under conditions of normocapnia is 1.39±0.11. A range of normal indices of OC, indicative of preserved autoregulation, varies from 1.23 up to 1.54. OC, exceeding 1.5, reflects increased tonicity of resistive vessels and is typical of arterial hypertension. If OC is less than 1.2, tonicity of resistive vessels is reduced. It is characteristic of many pathologic states (brain ischemia, intracranial hypertension, vascular spasm, arteriovenous shunting). OC values, close to 1.0, are an indication of impaired autoregulation, being an unfavorable prognostic sign. In severe cerebral lesions, large malformations and marked circulatory insufficiency a level of LBFV can be lower than initial values after compression discontinuation (an inverted response). It is a manifestation of absence/failure of cerebral autoregulation in this vascular region. Such response is indicative of exhausted vasodilation reserve due to additional reduction of perfusion pressure. It corresponds to a high risk of repeated episodes of a hemodynamic ischemic lesion and reflects a phenomenon of intracranial steal of ischemic areas at the expense of segments of a vascular bed with a normal reaction. Mean values of OC in organic lesions of the cerebrovascular system (arteriovenous malformation, vascular spasm, occluding lesions) vary from 1.01 up to 1.10; it is considerably lower than normal parameters.

A test load with nitroglycerin possesses a bidirectional effect on the system, regulating cerebral circulation. On the one hand, it is conditioned by an effect on central hemodynamics (reduced venous return to the heart, reduced cardiac output) and on the other by a direct endothelium-independent dilating effect on middle-sized arteries. As for the system of cerebral circulation, it results in growth of volumetric cerebral blood flow against a background of reduction of LBFV in major arteries due to their dilation. In a normal state decrease of mean LBFV by 10-12%, growth of volumetric blood flow by 70±35 ml/min, enlargement of the MCA diameter by 37±12% are watched 3 minutes after taking the drug (according to data of transcranial sonogaphy with color coding of Doppler spectrum energy). It is considered to be an adequate response (Lelyuk V.G., Lelyuk S.E., 1999).

A combination of compression and nitroglycerin tests allows to estimate tonicity of major cerebral arteries and those of a peripheral bed.

In 1987 R. Aaslid et al. proposed a quantitative method of estimation of cerebral autoregulation, known as a cuff test. Its essence lies in comparative analysis of changes of systemic blood pressure and LBFV in cerebral vessels in response to acute reduction of blood pressure (by 20-25%). The authors achieved it by postischemic hyperemia in lower extremities, developing after thigh compression with pneumatic cuffs. As for this method, monitoring of blood pressure and LBFV helps to reveal their relative changes. Difference in a speed of restoration of blood pressure and LBFV demonstrates a degree of autoregulaton. In its absence a trend of LBFV is coherent to a trend of blood pressure; in its presence restoration of LBFV is much quicker.

A speed of cerebral autoregulation (RoR) is calculated according to the following formula:


DCVR is a relative change of cerebrovascular resistance after pressure decrease in cuffs,
D50 is a term, necessary for LBFV restoration up to 50% of its initial value after its maximum reduction during a cuff test,
DCPP is a relative change of perfusion pressure after reduction of pressure in cuffs.

Fig. 3. Changes of systemic blood pressure (BP), LBFV
and intracranial pressure (ICP) in a cuff test.

A normal calculated value of RoR is 20±3% sec-1. In a normal state RoR is dependent on CO2 tension. It decreases in hypercapnia (11±2), and inreases in hypocapnia (38±4) (Fig. 3). The lower RoR, the poorer autoregulation. Gross disorders of autoregulation are watched in severe craniocerbral trauma, intracranial hemorrhages, vascular spasm (Fig. 4). The shortcoming of this test is necessity to use special systems for monitoring TCD indices, blood and intracranial pressures, as well as special software.


Fig. 4. A - Results of a cuff test in a patient, whose state is compensated.
Indices of cerebral autoregulation rate correspond to normal values.

B - Results of a cuff test in a patient with spasm of the left MCA in an acute period of subarachnoid hemorrhage.
Indices of cerebral autoregulation rate are much lower on the spasm side (impaired auioregulation).

In any case, the majority of tests implies an impact (safe as it may be) on the body of a person under examination. It causes some inevitable distortion of responses, which can take a bit different course under natural conditions. Due to this fact, there appeared an idea to estimate autoregulation on the basis of durable monitoring of parameters of cerenral blood flow, systemic blood pressure and CO2 tension without any external distubances. Such monitoring helps to reveal the so-called slow oscillatons of physiologic parameters; besides similarity of oscillatons of blood flow, pressure and saturation of blood with CO2 and its degree can be determined on the basis of calculation of coherence coefficients. Being really non-invasive, such method permits to estimate regulatory profiles in situ, which makes it of great value for clinical practice.

In 1991 C. Giller analyzed spontaneous oscillations of blood pressure and LBFV in patients with ruptured intracranial aneurysms. It demonstrated dependence of a coherence index (Fouriers cross-spectral analysis) between spontaneous oscillations of blood pressure and LBFV on a state of cerebral autoregulation. A normal coherence index is 0.21±0.05. It increased up to 0.45±0.1 in patients with ruptued intracranial aneurysms (p<0.05).

In 2001 E. Lang et al. revealed dependence of a phase displacement angle between spontaneous oscillations of blood pressure and LBFV on a state of cerebral autoregulation. They determined a normal value of this index. It was 69.4±12.10o. Fig. 5 demonstrates spontaneous oscillations of blood pressue and LBFV in patents with a different term of subarachnoid hemorrhage.

Fig. 5. A phase angle between spontanous oscillations of blood pressure and LBFV.

A phase angle is considerably smaller in patients with intracranial aneurysms and dependent on a term of subarachnoid hemorrhage. Its further reducton is observed on the 7-13th day after SAH.

Taking into account anatomic-and-functional integrity of the cerebrovascular system, a test, allowing to estimate a response not only of one isolated vascular region, but at least two symmetrical regions of healthy and damaged hemispheres, should be considered to be optimum. This condition can be realized with bilateral recording of LBFV against a load background. Sensors are fixed with the help of a special helmet for monitoring, which provides their fastening in an arbitrary manner and in projection of an ultrasonic window.

Fig. 6. Normal trends of mean LBFV in both MCA: at rest, in hyper- and hypocapnic loads.
The X-axis mean LBFV (cm/s), the Y-axis time (min).

Normal trends of blood flow velocity in conjugate arteries are characterized by a high degree of similarity and repeat each other (Fig. 6).

In case of unilateral insufficiency, a response of LBFV to a load differs greatly on a lesion side (Fig. 7).

Bilateral TCD-monitoring allows to calculate not only standard indices of reactivity, but also such parameters, characterizing a degree of circulatory insufficiency on a lesion side, as a coefficient of flows correlation (it runs to 1.0 in a normal state), a ratio of regression coeffficients (it runs to 1.0 in a normal state). Fig. 8 represents trends of linear regression of LBFV in conjugate vascular regions, watched in a normal state and ICA occlusion.

Fig. 7. Trends of mean LBFV in both MCA in unilateral occlusion of the internal carotid artery:
A - on a healthy side, B - on an occlusion side.
The Y-axis mean LBFV (cm/s), the X-axis time (min).

Results of a reaction to functional loads allowed R. Aaslid et al. to give the following classification:
1) A unidirectional positive reaction a symmetrical adequate response to a load;
2) A multidirectional reaction a positive response on one side and a reduced or paradoxical response on another.
3) A unilateral negative reaction a bilateral reduced or inverted response.

Fig. 8. Diagrams of linear regression of LBFV in both MCA in a normal state and occlusion.
Y- and X-axes mean LBFV (cms).

Today TCD is the main method of studying regulation of cerebral blood flow in clinical practice. Use of functional loads permits to estimate a state of the cerebral circulation system, which is often the most important factor in determining indications for an operation in cerebrovascular diseases (interventions, aimed at creating microvascular anastomosis, operations for arteriovenous malformations), therapy of acute intracranial hypertension, prescription of vasoactive drugs. As for everyday practice, the most convenient tests are those, allowing to assess a response of vessels to CO2 (tests with breath holding and hyperventilation), autoregulation (compression and cuff tests) and cross-spectral analysis of spontaneous oscillations of systemic and cerebral hemodynamics. One cannot speak of full-value research, if it carried out without estimation of a functional state of the cerebral circulation system.