Acta Neurochir (Wien) (1996) 138:68-76 Acta Neurochirurgica 9Springer-Verlag 1996 Printed in Austria Cerebral Haemodynamic Considerations in Obstructive Carotid Artery Disease A. Sorteberg, W. Sorteberg, K.-F. Lindegaard, and H. Nornes Department of Neurosurgery, Rikshospitalet, The National Hospital, University of Oslo, Oslo, Norway Summary 46 subjects with obstructive carotid artery disease were investigated with transcranial Doppler ultrasonography. Their baseline blood velocities (V) in the middle, anterior and posterior cerebral artery (MCA, ACA and PCA) and in the extracranial internal carotid artery (ICA) were measured and the pulsatility index (PI) calculated for each vessel. Thereafter the vasomotor reserve in both MCAs was tested. Typical patterns of V, PI and vasomotor reactivity are presented. Arterial collaterals were recognized by their relatively increased velocities. We demonstrated a close association of the baseline variables V and PI and the total vasomotor reactivity (hypocapnic plus no, hypercapnic response) by calculating an index of ghemrelated to the cerebrovascular tone. The Uhem index is expressed by: ghe m index = VMcA'PIMcA/VpcA'PIpcA The relationship between Uh~mindex and the total vasomotor reactivity seemed to correspond to a hyperbolic curve. The hyperbolic tangent of Uhem index and total vasomotor reactivity correlated highly significantly, r = 0.8203, p <0.0001, n = 49, the best fit for the regression line was Y = -0.005+Uh~m index 51.3. On the 99% significance level an Uhem index >0.94 indicated normal total cerebral vasomotor reactivity in contrast to an impaired reactivity when ~<0.81. Findings in 20 patients investigated post hoc supported the validity of our concept. Keywords: Transcranial Doppler sonography; vasomotor reserve; arterial collaterals; cerebrovascular impedance. Introduction In obstructive carotid artery disease the cerebral circulatory status can be evaluated by assessing the cerebral v a s o m o t o r reserve [2, 3, 8, 10, 14, 16, 21, 30]. Using non-invasive transcranial Doppler ultrason o g r a p h y (TCD), this is c o m m o n l y done by measuring changes in blood velocity (V) in response to step changes in the carbon dioxide partial pressure (pCO2) [14]. Individuals with reduced cerebral v a s o m o t o r reserve m a y be especially prone to l o w - f l o w cerebral infarction [6, 16] and could benefit from surgery [6, 8, 22, 29]. The functional state o f the cerebral microvascular bed exerts considerable influence both on the magnitude o f the cerebral v a s o m o t o r reserve [7] and on the T C D variables blood velocity and the pulsatility index (PI). Functional changes in the cerebral microcirculation should therefore alter the cerebral vasom o t o r reserve as well as V and PI. Attempts to define this relationship have so far been unsatisfactory [9], although a trend has been shown [13]. The aim of this study was to investigate the association o f cerebral v a s o m o t o r reserve and the T C D variables V and PI in patients with inflow disturbance, n a m e l y obstructive carotid artery disease. Based on these clinical findings we introduce a concept to estimate the total cerebral v a s o m o t o r reserve in individual subjects from their T C D baseline variables. Theoretical Considerations The oscillating velocity of the b l o o d flow in arteries can be regarded as the rheological analogue to an electrical alternating current [15]. The blood velocity V (in cm.s -i) then represents the current (I) and the vascular impedance (Z, in dyn.s.(cm3) -I or 0.1-kg.(m.s) -1) the total opposition to flow. The electromotive force (U) is represented by the contractile state, or tone (in dyn.(cm2)-l), of the heart and vasculature. The relationship between these units is derived f r o m O h m ' s law (I.Z = U), and we then obtain: V(cm.s-1).Z(dyn.s.(cm3) -1) = tone (dyn.(cm2) -1) (1) With the heart generating a continuous train o f pressure waves, the vascular impedance determines the shape o f the pulsatile blood velocity w a v e f o r m in A. Sorteberg et al.: Cerebral Haemodynamic Considerations 69 one cardiac cycle. Exact description o f this w a v e f o r m envelope requires c o m p l e x analysis, in clinical routine we might therefore use a m o r e simplified procedure. The pulsatility index, PI (Systolic minus diastolic blood velocity divided by the time mean, [5]), serves to describe the blood velocity w a v e f o r m and will be an approximation of the vascular impedance. Replacing Z with the PI in (1), incorporates this approximation into the equation, thus the product V.P1 will, instead of representing the vascular tone, express a unit closely related to it (Uhem): V - PI = Uhem (2) Absolute values of V and PI in h u m a n cerebral arteries m a y show large inter-individual differences. However, in the normal situation there is a relatively constant relationship between blood velocities and the PI values observed in one brain hemisphere [27]. Thus, normalized values, the so-called hemispheric indices, allow for the detection of h a e m o d y n a m i c alterations that might be obscured when considering absolute values only [12, 27]. In the present study on patients with obstructive carotid artery disease we chose to emphasize the hemispheric middle cerebral artery/posterior cerebral artery ( M C A / P C A ) indices because: the M C A is the end artery to the obstructed carotid artery and the P C A is the only collateral channel which usually has its source in a separate arterial system (the vertebrobasilar) and can develop Willisian collaterals (through the posterior c o m m u n i c a t i n g artery) as well as supra-Willisian collaterals (through small pial arteries and arterioles). In accordance with (2) we obtain: VMcA'PIMcA/VpcA'PIpcA = Uhem index (3) C o m p l e x m e c h a n i s m s regulate the tone o f the vascular bed. The regulatory upper and lower v a s o m o t o r endpoints are m a x i m a l constriction and complete dilation, respectively. In a given situation the vascular tone should therefore determine the amount o f vasom o t i o n (further constriction and/or dilation) to a vasoactive stimulus. Materials and Methods Patients The study comprised 46 individuals, 8 females and 38 males with a median age of 61 years (range 41-76 years). The diagnosis of obstructive carotid artery disease was established by angiography and confirmed by means of Doppler ultrasonography. All subjects first underwent an examination at rest (baseline examination) followed by investigation of their cerebral vasomotor reserve. Lying supine in a comfortable position they initially adapted to the situation. Blood Velocity Measurements The blood velocity measurements were performed on 2 MHz range-gated Doppler instruments, either an EMETrans-scan (Ueberlingen, Germany)(first 38 patients) or a Multi Dop X (DWL, Sipplingen, Germany) (last 8 patients). Baseline examination: Using a hand-held probe, the MCA, anterior cerebral artery (ACA), PCA and proximal and distal portion of the extracranial internal carotid artery (ICA) were identified in sequence and their highest blood velocity reading scored by the procedure previously described [1, 12]. The PI was calculated for each artery; the percentage area stenosis estimated from the carotid artery blood velocity readings [11]. Vasomotor reserve testing: The subjects now breathed through a mouthpiece with a one-way valve. After achieving a stable normocapnic state, changes in pCOz were obtained by breathing 5% CO2 in air directly followed by voluntary hyperventilation. The end-expiratory pCO2 was recorded continuously on an infrared CO2 analyser (CD-102, Normocap, Datex, Finland). Using the fixed probe technique, the blood velocity was monitored at the same time in both MCAs [19]. Under off-line data processing, samples from stable normo-, hyper- and hypocapnic conditions were extracted. The vasomotor reserve is expressed as percentage change in V per unit change in pCO2 in kPa (1 kPa equalling 1 Vol%) referred to each individual's normocapnic state. Hyperplus hypocapnic response is denoted as total vasomotor reactivity (VMRtotal). Arterial Blood Pressure (ABP) During the vasomotor reactivity testing, the ABP was repeatedly measured using a conventional manual blood pressure cuff. Mean ABP was calculated as the diastolic pressure plus one third of the systolic-diastolic difference. The values obtained were averaged separately for the three pCO2 levels for each indiviual. Patients Examined Post Hoc After completion of the study, the next 20 subjects with obstructive carotid artery disease, 3 females and 17 males with a median age of 61 years (range 38-74 years), were examined in order to validate our concept for estimating the VMRtotat from V and PI. Using the Multi Dop X, they were investigated according to the protocol described above. Results In the following, the hemisphere above by the (more) obstructed I C A is termed side A and the contralateral hemisphere side B. We identified 92/92 M C A s , 72/92 A C A s , 49/92 P C A s , and 83/92 I C A s in the 46 subjects (9 I C A s were occluded). T h e y were divided into five groups according to our findings: Patients with unilateral stenoses and antegrade A C A flow (group 1, n = 15) or retrograde A C A flow (group 2, n = 8), those with bilateral stenoses and antegrade (group 3, n = 10) or retrograde (group 4, n = 4 ) A C A flow and those A. Sorteberg et al.: Cerebral Haemodynamic Considerations 70 Table 1. Percentage Area Stenosis of the Carotid Artery (Median Values with Ranges)Estimated fi'om the Carotid Artery Blood Velocity Readings Group No. Hemisphere A Hemisphere B 1 2 3 4 5 15 8 l0 4 9 76(42-92) 80(72-90) 81(53-89) 93(75-94) 100 59 (25-73) 47 (30-62) 70(10-87) For all 46 subjects combined, the VMCA was lower on side A than on side B (61 vs 75 cm.s -1, p <0.001); especially when ACA flow was reversed (groups 2, 4, and 5, Fig. 1). The PINTA o n side A was also dampened (0.71 vs 0.85, p <0.01, Fig. 2). Similarly, in the distal extracranial ICA the V was lower and the PI was dampened on side A compared with side B (VIcA: 31 vs 44 cm.s -1, p <0.001 and PIIcA: 0.83 vs 0.94, p <0.001). With antegrade ACA flow (groups 1 and 3), the VACA o n side A was normal to low (53 c m . s -1, h e m i s p h e r i c VAcA/VMcA = 0.76). When ACA flow was reversed (groups 2, 4, and 5), the with unilateral stenoses and contralateral occlusion (group 5, n = 9). Table 1 gives the percentage carotid artery area stenosis in the five groups. TCD Baseline Variables In this section we used the Wilcoxon test for paired observations for statistical purpose; the results are presented as mean values. .... HUA [CA .... HL:-A 1 3 2 4 VACA e x c e e d e d the i p s i l a t e r a l VMCA (VAcA 60 c m . s -1, hemispheric VAcAfVMca = 1.35).On side B, the VACA was highly increased (92 cm.s -1, hemispheric VAcA/VMcA = 1.30).'No clear tendency was found for the PIAcA (0.76 VS 0.77, n.s.). The VpcA on side A was higher than on side B (70 vs 59 cm-s -1, p <0.05). No definite relation between the PIpcA and ICA obstruction was detected. ICA GROUP I IV '-'*-' PC_A ICA .... ~ ICA ..... t-t.~ ICA VELOCITIES ( c m . s - 1 ) Fig. 1. Blood velocities in the middle, anterior and posterior cerebral arteries (MCA, ACA, PCA) and the extracranial internal carotid arteries (ICA) in obstructive carotid artery disease and in the normal situation. Groups 1 and 2: unilateral stenosis with antegrade and retrograde ACA flow; groups 3 and 4: bilateral stenosis with antegrade and retrograde ACA flow; group 5: unilateral stenosis with contralateral occlusion. A hemisphere above the (more) obstructed ICA. B contralateral hemisphere. + from Sorteberg et al. [27] A. Sorteberg et aL: CerebraI Haemodynamic Considerations 71 0Z , JB 3 A IVCA f ~ A K ' ~ ICA "-" I Group t-~A ICA 3 NORMAL -t" PULSATILITY INDICES ! .2 :.:.:.:.:. :.:.:.:.:. :.:.:.:.:. :.:.:.:.:. :.:.:.:.:. :.:.:.:.:. :.:.:.:.:. :.:.:.:,:. :.:.:.:.:. :.:.:.:,:. :.:.:.:.:. :.:.:,:.:. :.:.:.:.:, :.:.:.:.:. :,:.:.:.:. :.:.=.:. :.:.:.:.:. :.:.:.:.:. :.:.:.:.:. :.:.:.:.:. :.:.:q.:. :.:.:.:.:. I~ IW-A/~C-A FCA EA " - " H.YA ICA 2 A [vr~ 'z~'A ~ I:C_.A 4 ICA 5 Fig. 2.The Pulsatility Index, PI (systolic minus diastolic blood velocity divided by the time mean) in the MCA, ACA, PCA, and tCA in groups 1-5 of obstructive carotid artery disease and in the normal situation. A hemisphere above the (more) obstructed ICA. B contralateral hemisphere. + from Sorteberg et al. [27] VASOMOTOR REACTIVITY {V%.kPa-1) 29 30 23 / 6O 50, " --. 20 ~ - ' - ~.~.. ~. 49 , 18 -, 18 . 51 44 43 ". 43 40 10 i I - I 43 I 5% CO2 0 4O 1 30 30 2 3 22 20 4 5 31 30 2O 22 26, 20 10 !21 TOTAL t6 10 0 1 2 3 4 5 ItV 0 3 4 5 Fig. 3. Total, hypercapnic (5% CO2) and hypocapnic (HI/) vasomotor reactivity in the cerebral hemisphere above the (more) obstructed ICA (straight line, A) and the contralateral hemisphere (dotted line, B) in groups 1-5 of obstructive carotid artery disease A. Sorteberg et al.: Cerebral Haemodynamic Considerations 72 (ANOVA, F = 1.06, n.s.). There were considerable individual variations within each group. V%kPa-1 60 O n 50 9 " Z'"Y O 9D DO 30 20 ~o TCD Baseline Variables and Vasomotor Reactivity * o " 40 9 ":~ o ~176 4, 9 go ~ 9 Y 10 0 I I I I I I 0.5 1 1.5 2 2.5 3 Uhem index a V%kPa-1 60 50 40 De ip ,1~ 30 Oo D ~3 20 D 10 0 I 1 I 1 I 0.2 0.4 0,6 0.8 1 tanh (ghe m index) b Fig. 4. (a) The relationship between the index of Uhem,expressed as VMcA'PIMcA/VpcA'PIpcA and the total cerebral vasomotor reactivity (VMRtotal) in 46 patients with obstructive carotid artery disease in this study (black squares) and 20 patients examined post hoc (white squares). The graph suggests a hyperbolic curve. (b) Correlation of the hyperbolic tangent (tanh = e2•215 of Uhem index and VMRtotal, r = 0.8203 (p <0.0001, n = 49). The best fit for the regression line w a s Y = - 0 . 0 0 5 + U h e m index.51.3 (black line). Study series (black squares) and patients examined post hoc (white squares) For all 46 patients combined, there was no significant correlation of VMCA and PIMcA (r = 0.0237, n.s.) or of VMCA and VMRtotai (r = 0.2689, n.s.). The PIMCA and VMRtotai correlated with r = 0.459 (p <0.0001, n = 92). The relationship between Uhem index and the VMRtotat are illustrated in Fig. 4 a (black squares). The graph suggests a hyperbolic curve. The hyperbolic tangent (tanh = eZ• 2x+1) of Uhem index and VMRtotal correlated highly significantly (Fig. 4 b, black squares) with r = 0.8203 (p <0.0001, n = 49) and the best fit for the regression line was Y = --0.005+Uhem index.51.3. The VMRtotal was also estimated according to the best fit for the regression line. The relationship between measured and estimated VMRtotai is shown in Fig. 5 (black squares). When calculating on the precision of the estimated VMRtota], we defined a "critical" Uh~m index for the 99% confidence interval. An Uhem index />0.94 then indicates a normal VMRmtai ( > / 3 6 V%.kPa -~) in contrast to a n Uhem index ~<0.81 when the VMRtotal is impaired. Substituting contralateral ACA values for the ipsilateral P C A p a r a m e t e r s in (3), Uh~m index and VMRtotal led to r = 0.2524, p <0.05. V%kPa-1 60 measured n 9~ 50 9 30 [] 9 * Vasomotor Reactivity For all 46 patients combined, the VMRtotal was lower on s i d e A than on s i d e B (38+12 vs 46+8 V%-kPa <, p <0.00t, t = 5.587, t-test for paired samples). Figure 3 (left) gives the VMR~ot~Iin groups 1-5, showing no significant difference between groups and hemispheres (ANOVA, F = 2.85, n.s.). While the hypercapnic response (Fig. 3, upper right) was significantly lower on side A ( A N O V A , F = 5.42, p <0.001), the hypocapnic response (Fig. 3, lower right) was similar between groups and hemispheres 20 oD, , ~ 0 0 = 10 = 20 I 30 = I 40 50 estimated I V%kPa-1 60 Fig. 5. The relationship between measured and estimated VMetotal in patients with obstructive carotid artery disease in the study (black squares) and those examined post hoc (white squares) A. Sorteberg et al.: Cerebral Haemodynamic Considerations Arterial Blood Pressure The mean ABP increased 9% (range 1-29%)in hypercapnia and decreased 2% (range -22-14%) in hypocapnia. In 25 normotensive patients the cerebral vasomotor reactivity was independent of the ABP (side A: r = - 0 . 1 6 3 5 , n.s. and side B: r = - 0 . 0 5 6 8 , n.s.). In 21 subjects where the systolic ABP exceeded 170 mmHg during the vasoreactivity testing (all 21 had chronic hypertension) the vasomotor reactivity was also independent of the ABP in hemispheres above obstructed carotid arteries (side A: r = 0.0506, n.s. and side B: r = -0.1327, n = 9, n.s.). There was, however, an inverse relation between mean ABP and the hypercapnic response for the 12 hemispheres above nonobstructed arteries in the hypertensive patients (Side B in groups 1 and 2, r---0.6444, p <0.01). Patients Examined post hoc The 20 patients examined post hoc belonged to the different groups as follows: group 1, n = 6; group 2, n = 6; group 3, n = 1; group 4, n = 3; group 5, n = 4. Uhom index and its hyperbolic tangent were calculated from the V and PI in 35/40 hemispheres (five PCAs were not identified). These values and the measured VMRtotal are presented in Fig. 4 a and b, respectively (white squares). Using the best fit for the regression line, we also estimated the VMRtotal in the 35 hemispheres. The relationship between measured and estimated VMRtotal is shown in Fig. 5 (white squares). We obtained a mean of 38.8 V%.kPa -1 for the measured - versus 39.6 V%-kPa -~ for the estimated VMRtotal; the mean difference was 0.8 V%.kPa -1 and the standard error of the mean was 0.8 V%.kPa -1. The nine hemispheres where the measured VMRtotal was ~<36V%, were also estimated to be impaired. Correspondingly, the 26 hemispheres with measured VMRtot~l >~36V% were also recognized as being normal by estimation. Discussion TCD Baseline Variables The TCD baseline variables, blood velocity and pulsatility are resultants of complex factors interacting, including the size and quality of the input signal, rheological/chemical properties of the blood and organ related factors [ 13]. These factors have equivalent influence on the different brain regions [23]. 73 Superimposed on this, local factors such as intracranial space occupying lesions or disease altering the cerebral haemodynamics exert regional influences on the V and PI [17]. In obstructive carotid artery disease, the cerebral circulation depends on the degree of carotid artery lumen narrowing as well as on the extent of the collateral blood supply [29]. Using TCD, functioning collaterals can be recognized by analysing the individual's cerebral haemodynamic pattern: In the normal situation, the blood velocities in the two hemispheres show no significant side differences (Fig. 1, upper right). Moreover, within each hemisphere they are subject to characteristic proportions forming a "stairway pattern" (Fig. 1, upper right). Alterations in this pattern suggest a haemodynamic situation related to pathology. Figure 1 thus illustrates how increasing ICA pathology distorts this stairway pattern more and more. Communication through the circle of Willis also makes it necessary to consider the blood velocities in a bilateral context, even when the obstruction is unilateral. Typically, collateral flow patterns are recognized from the relatively increased blood velocity in the collateral arteries, an increase reflecting enlargement of the respective arteries' perfusion territories [26]. In obstructive carotid artery disease, the main collaterals are the contralateral ACA and the ipsilateral PCA. Potential collaterals may be revealed by compressing the common carotid artery [24]; however, we do not use this as a routine procedure. The stenosis itself is recognized by the increased blood velocity at the point of obstruction and the decreased velocity and pulsatility distally. The present findings of decreased ipsilateral VMCA and PIMcA concur with previous studies [13, 21, 25]. Not only velocity, but also the pulsatility contains information about the capacity of collateral channels. Collaterals showing more distinct pulsatile patterns have thus been found to have the highest index of collateral capacity, while those with a low collateral capacity have a throttled pulsatile flow [18]. The observation of an undampened pulsatility was considered indicative of low impedance in the collateral vessel [18]. In our subjects investigated there was a!so a tendency towards higher PIs in the collaterals with the relatively most increased velocities (ipsilateral PCA in groups 3, 4, and 5 and reversed ipsilateral ACA in groups 2 and 5, Figs. 1 and 2). The PI is often associated with the cerebrovascular 74 resistance, which is mainly confined to the cerebral microvasculature. However, the bloodstream in the cerebral arteries is oscillating and subjected to the more complicated cerebrovascular impedance, including the vascular closing pressure and the total system compliance distal to the point of observation. The term resistance is used in steady flow dynamics [15]. Nornes [19] described how mean ICA flow and flow pulsatility decreased similarly during graded ICA clamping for saccular aneurysm, while a corresponding reduction of ICA flow caused by hyperventilation increased the ICA flow pulsatility. He explained this with a nearly similar "resistance" to flow in the two situations, but with different impedances. Ffirst et al. [4] used the Purcelot Index to express vascular impedance. However, this index' only takes into consideration the systolic and diastolic extremes. The PI seems physiologically more relevant for such evaluations because it takes account of the waveform envelope [13]. T C D Baseline Variables and Vasomotor Reactivity The correlation of single TCD baseline variables and the VMRtotal was lower than when using calculations of the Uhem index in (3). The PIMcA and the VMRtotal correlated with r = 0.459, which is close to the results of Ley-Pozo et al. [9] and Lindegaard [13]. We observed that the more the ipsilateral VPCA increased, the more the VMRtotal was impaired. This might seem paradoxical since a good collateral function should in fact prevent the vasomotor reserve from becoming exhausted. However, one must bear in mind that our series probably were selected in the sense that they were healthy enough to be considered for surgery. To this end, only those with the "fittest" collateralization system survive [21]. The maximum exploitation of dilatatory ability goes hand in hand with the maximum use of collateral capacity; a highly increased VecA therefore reflects a marginal ipsilateral hemispheric perfusion. The collateral function on the contralateral ACA acts via the anterior communicating artery (ACoA). Being shorter (average length: 2.6 vs 14 mm) and most often wider (average diameter: 2.3 vs 1.2 mm) than the posterior communicating artery [28], the resistance to flow through the ACoA will usually only be a fraction of that in the posterior communicating artery. Dependence of a hemisphere on ipsilateral PCA collateral flow therefore may reflect a situation A. Sorteberget al.: CerebralHaemodynamicConsiderations where the capacity of the anterior circle of Willis is inadequate. This also explains why our contralateral ACA values showed a poor correlation in the estimation of the VMRtotal. Widder et al. [30] found the normal vasomotor reactivity to be 23_+5 V%.kPa q, and considered it p/~thological when <18 V%.kPaq. Based on their findings, we presently set the lower limit for the normal VMRtotal at 36 V%.kPa-1. The vasomotor tone defines the microvascular capacity to further dilatation and/or constriction. It is therefore appropriate to link the g h e m index in the VMRtot~. This relationship suggests a hyperbolic curve with a breakpoint close to 0.94; which is the lower limit for the compatibility with a normal VMRtot~. Above this threshold, the vasomotor reactivity has a physiologically broader variation than the Uhe m index. Given adequate MCA and PCA blood velocity signals, we could presently estimate the VMRtotal from the baseline variables V and PI. Moreover, findings in the 20 patients examined post hoc (Figs. 4 a, b and 5, white squares) support the validity of this concept. Such an estimation offers an alternative to reactivity testing and is of special interest in subjects in whom such a test is not feasible. Our concept also broadens the view on cerebral haemodynamics in obstructive carotid artery disease because it includes both the vasomotor and collateral capacity. Arterial B l o o d Pressure Hypercapnia presently elevated the mean ABP by about 10%, which is of the same magnitude as reported previously [3, 14, 30]. We also confirm the findings of Widder et al. [30] that the cerebral vasomotor reactivity is independent of ABP at systolic pressures <170 mm Hg. For values >170 mm Hg, however, we obtained an inverse relation between hypercapnic reactivity and the ABP in hemispheres above nonobstructed ICAs. This can be explained by constriction of the cerebral microvasculature to the increased ABP counteracting the vasodilatatory response to hypercapnia. In hemispheres severely affected by obstructed ICAs, an elevated ABP may protect the brain by helping to maintain an adequate cerebral perfusion pressure. An increase in ABP might then not be counteracted by microvascular contraction to the same extent as in the normal situation. Correspondingly, a reduction of the ABP in patients with haemodynamically significant ICA obstruction could in fact provoke low-flow cerebral infarction. A. Sorteberg et al.: Cerebral Haemodynamic Considerations Acknowledgement The authors wish to thank Robert Innes, MD, for reviewing the English manuscript. References 1. Aaslid R (1986) Transcranial Doppler examination techniques. In: Aaslid R (ed) Transcranial Doppler sonography. Springer, Wien New York, pp 39-59 2. Bishop CCR, Powell S, Insall M, Rutt D, Browse NL (1986) Effect of internal carotid artery occlusion on middle cerebral artery blood flow at rest and in response to hypercapnia. Lancet: 710-712 3. Brown MM, Wade JPH, Bishop CCR, Ross Russel RW (1986) Reactivity of the cerebral circulation in patients with carotid occlusion. J Neurol Neurosurg Psychiatry 49:899-904 4. Ftirst H, Hartl W, Jansen I, Fink B, Piepgras A, Lauterjung L, Sehildberg F (1992) Pathophysiology of unilateral high-grade carotid artery stenosis: evaluation of intracranial haemodynamics by analysis of velocity waveforms from the middle cerebral artey. Clin Science 83:357-366 5. Gosling RG, King DI-I (1974) Arterial assessment by Dopplershift ultrasound. Proc R Soc Med 67:447-449 6. Halsey JH, Morawetz RB, Blauenstein UW (1982)The haemodynamic effect of STA-MCA bypass. Stroke 13:163-167 7. Harper AM, Glass HI (1965) Effect of alterations in the arterial carbon dioxide tension on the blood flow through the cerebral cortex at normal and low arterial pressures. J Neurol Neurosurg Psychiatry 28:449~;52 8. Karnik R, Valentin A, Ammerer H-P, Donath P, Slaw J (1992) Evaluation of vasomotor reactivity by transcranial Doppler and acetazolamide test before and after extracranial-intracranial bypass in patients with internal carotid artery occlusion. Stroke 23:812-817 9. Ley-Pozo J, Willmes K, Ringelstein EB (1990) Relationship between pusatility indices of Doppler flow signals and CO2reactivity within the middle cerebral artery in extracranial occlusive disease. Ultrasound Med Biol 16:763-772 10. Levine RL, Dobkin JA, Rozental JM, Satter MR, Nickles RJ (1991) Blood flow reactivity to hypercapnia in strictly unilateral carotid disease: preliminary results. J Neurol Neurosurg Psychiatry 54:204-209 11. Lindegaard KF, Bakke SJ, Grip A, Nornes H (1984)Pulsed Doppler techniques for measuring instantaneous maximum and mean flow velocities in carotid arteries. Ultrasound Med BioI 10:419-426 12. Lindegaard KF, Bakke SJ, Grolimund P, Aaslid R, Huber P, Nornes H (1985) Assessment of intracranial hemodynamics in carotid artery disease by transcranial Doppler ultrasound. J Neurosurg 63:890-898 13. Lindegaard KF (1992) Indices of pulsatility. In: Newell DW, Aaslid R (eds) Transcranial Doppler. Raven, New York, pp 67-82 14. Markwalder TM, Grolimund P, Seller RW, Roth F, Aaslid R (1984) Dependency of blood flow velocity in the middle cerebral artery on end-tidal carbon dioxide partial pressure - a transcranial ultrasound Doppler study. J Cereb Blood Flow Metab 4:368-372 15. McDonald DA (1974) Blood flow in arteries. Edward Arnold, London, p 124 75 16. Miller JD, Smith RR, Holaday HR (1992) Carbon dioxide reactivity in the evaluation of cerebral ischemia. Neurosurgery 30:518-521 17. Newell DW, Seiler RW, Aaslid R (1992) Head injury and cerebral circulatory arrest. In: Newell DW, Aaslid R (eds) Transcranial Doppler. Raven, New York, pp 109-t21 18. Nornes H (1972) Hemodynamic aspects in the management of carotid-cavernous fistula. J Neurosurg 37:687-694 19. Nornes H (1975) Pulsatile blood flow pattern in cerebral circulation. In: Langfitt TW, Mc Henry LC Jr, Reivich M e~ al (eds) Cerebral circulation and metabolism. Springer, Berlin Heidelberg New York Tokyo, pp 152-154 20. Nornes I-I, Sorteberg W, Nakstad P, Bakke SJ, Aaslid R, Lindegaard KF (1990)Haemodynamic aspects of clinical cerebral angiography. Concurrent two vessel monitoring using transcranial Doppler ultrasound. Acta Neurochir (Wien) 105:89-97 21. Norris JW, Krajewski A, Bornstein NM (1990)The clinical role of the cerebral collateral circulation in carotid occlusion. J Vasc Surg 12:113-118 22. Norrving B, Nilsson B, Risberg J (1982) rCBF in patients with carotid occlusion. Resting and hypercapnic flow related to collateral pattern. Stroke 13:155-162 23. Ogawa S, Handa N, Matsumoto M, Etani H, Yoneda S, Kimura K, Kamada T (1988) Carbondioxide reactivity of the blood flow in human basilar artery estimated by the transcranial Doppler method in normal men: a comparison with that of the middle cerebral artery. Ultrasound Med Biol 14:479-483 24. Schneider PA, Ringelstein B, Rossman ME, Dilley RB, Sobel DF, Otis SM, Bernstein EF (1988) Importance of cerebral coilateral pathways during carotid endarterectomy. Stroke 19: 1328-I334 25. Schneider PA, Rossman ME, Bernstein EF, Torem S, Ringelstein B, Otis S (1988) Effect of internal carotid artery occlusion on intracranial hemodynamics. Stroke 19:589-593 26. Sorteberg W, Lindgaard KF, Rootwelt K, Dahl A, Russel D, Nyberg-Hansen R, Nornes H(1989) Blood velocity and regional blood flow in defined cerebral artery systems. Acta Neurochir (Wien) 97:47-52 27. Sorteberg W, Langmoen IA, Lindegaard KF, Nomes H (t990) Side-to-side differences and day-to-day variations of transcranial Doppler parameters in normal subjects. J Ultrasound Med 9:403-409 28. Von Reutern GM, Btidingen HJ (eds)(1989)Ultraschalldiagnostik der himversorgenden Arterien. Thieme, Stuttgart, p 39 29. Wiberg J, Nomes H (1983) Effects of carotid endarterectomy on blood flow in the internal carotid artery. Acta Neurochir (Wien) 68:217-226 30. Widder B, Paulat K, Hackspacher J, Mayr E (1986) Transcranial Doppler CO2 test for the detection of hemodynamically critical carotid artery stenoses and occlusions. Eur Arch Psychiatr Neurol Sci 236:162-168 Correspondence: Angetika Sorteberg, M.D., Department of Neurosurgery, Rikshospitalet, The National Hospital, University of Oslo, N-0027 Oslo 1, Norway. Comments In this paper the group of the Rikshospitalet in Oslo presents a very original and well clone study on cerebral haemodynamics in obstructive carotid artery disease. 46 patients were investigated 76 using transcranial Doppler ultrasonography and the velocities and pulsatility-index of the basal cerebral arteries and of the extracranial internal carotid artery were measured and correlated with the vasomotor reserve in both MCAs. The authors demonstrate a close association of the velocity, the pulsatility-index and the vasomotor reactivity. By the non-invasive Doppler technique, relevant hae- A. Sorteberg et al.: Cerebral Haemodynamic Considerations modynamic information can be gained as demonstrated in this paper. The method used is correct and the results are significant. The manuscript deserves the attention of the readers of Acta Neurochirurgica. R. Seiler