Background and Purpose If it could be determined whether cerebral blood flow can be maintained (autoregulated) during transient falls in arterial blood pressure, we might be able to identify patients with carotid stenosis who are at risk of stroke. However, conventional methods of determining autoregulation in such patients are invasive and/or expensive.
Methods We used a new noninvasive method to estimate dynamic cerebral autoregulation in 27 patients with carotid stenosis and 21 age-matched normal controls. After a stepwise fall in arterial blood pressure, we determined the rate of rise of middle cerebral artery blood flow velocity compared with that of arterial blood pressure. We compared the method with a conventional method of determining cerebral hemodynamics, CO2 reactivity.
Results Autoregulatory index (ARI) was significantly reduced in middle cerebral arteries ipsilateral to a stenosed/occluded carotid artery: mean±SD 3.3±2.2 compared with normal controls (6.3±1.1; P<.0001) and nonstenosed carotid arteries in patients (5.9±2.1; P<.002). A subgroup of patients with severe impairment was identified. ARI returned to normal after carotid endarterectomy was performed. In a number of cases, ARI was impaired in the presence of CO2 reactivity.
Conclusions This simple technique allows identification of impaired autoregulation in patients with carotid artery disease. It may allow identification of patients at risk from transient falls of blood pressure as may occur at the onset of antihypertensive therapy and during surgery. It may allow a subgroup of patients with asymptomatic carotid stenosis who are at risk of hemodynamic stroke to be identified.
Recent trials have demonstrated that carotid endarterectomy results in a significant reduction in stroke in patients with symptomatic carotid stenosis >70%.1 2 However, even in this group, 5 to 10 patients need to be operated on to prevent one stroke.3 The benefit of carotid endarterectomy in asymptomatic carotid stenosis is more marginal, with 17 patients needing to be operated on to prevent one stroke over a 5-year period.4 5 The benefits in the real world with less selective patients and perhaps higher operative mortality and morbidity rates may be even less. There is a need for better clarification of high-risk groups.6 A number of markers of high stroke risk have been suggested, including degree of stenosis, plaque ulceration, infarction on brain CT, echogenic plaque on B-mode ultrasound, and asymptomatic circulating emboli detected by transcranial Doppler ultrasound in the ipsilateral middle cerebral artery.6 In addition, it has been suggested that impaired hemodynamics may correlate with increased stroke risk. Critically, ischemic tissue can be identified using positron emission tomography to demonstrate increased oxygen extraction fraction.7 More simply, transcranial Doppler ultrasonography has been used to provide a surrogate measure of cerebral perfusion or “cerebral reactivity.” This utilizes the fact that normal cerebral artery blood flow velocity increases in response to a vasodilator such as CO2.8 This reactivity can be shown to be reduced in a minority of patients with tight carotid stenosis, the presence of impaired reactivity correlating with poor collateral supply. It is assumed that in the presence of hemodynamic compromise the ipsilateral middle cerebral artery territory vessels are already vasodilated and therefore that there can be little further vasodilatation, and increase in blood flow velocity, in response to increased inspired CO2. Impaired CO2 reactivity has been associated with a markedly increased stroke risk in subjects with an occluded internal carotid artery,9 and a small study suggests either CO2 reactivity or acetazolamide reactivity may also be a marker in patients with tight carotid stenosis.10 Although measuring slightly different responses, both CO2 and acetazolamide reactivity correlate well in patients with carotid artery stenosis or occlusion.11 However, there are some potential problems with this technique. First, it has been demonstrated recently that breathing increased levels of inspired CO2 can cause a significant rise in blood pressure.12 This may result in increased cerebral artery blood flow velocity related not to active vasodilatation but to passive autoregulation. The technique might therefore be expected to underestimate the degree of hemodynamic compromise in some patients. Second, the vasodilatory response to CO2 is not necessarily an important physiological response and a more relevant approach would be to determine the ability of the cerebral circulation to maintain cerebral blood flow or to “autoregulate” in response to brief transient reductions in blood pressure that may occur in such patients. In addition, the increase in cerebral blood flow in response to hypercapnia is a distinct response with separate effectors from those involved in dynamic autoregulation, and the two may dissociate in certain circumstances.13 Whether cerebral blood flow can be maintained in response to such changes in blood pressure may also determine the risk of cerebral ischemia in patients with carotid stenosis who suffer hypotension during initiation of antihypertensive therapy or during surgery, particularly during cardiopulmonary bypass.
A suitable technique for determining dynamic autoregulation with the use of transcranial Doppler ultrasound has been described recently.14 The rate of recovery of MCAV is compared with that of the ABP after a sudden reduction in blood pressure induced by rapid deflation of inflated thigh cuffs.
We have applied this technique to patients with >60% carotid stenosis to determine its reproducibility, the extent of abnormal autoregulation in such patients, and its correlation with cerebrovascular reactivity measured using the CO2 method. We also determined whether impairments of dynamic autoregulation were corrected after carotid endarterectomy or percutaneous carotid angioplasty, which have previously been shown to improve impaired CO2 reactivity.15 16
Subjects and Methods
Twenty-seven subjects with carotid stenosis >60% or carotid occlusion were studied. Degree of stenosis was determined using duplex ultrasound (Acuson XP) with a combination of B-mode and color-flow Doppler imaging; grading of stenosis was based on Doppler velocities17 in combination with B-mode imaging. In 10 of them there was a contralateral stenosis >60% and, therefore, 31 stenoses were studied in all. Twelve were on antihypertensive medication that was withheld for 24 hours before the study. Twenty-one healthy age-matched nonsmoking volunteers with carotid stenosis excluded on duplex ultrasound were also studied as the control group. There was no difference in age between case subjects and controls: mean±SD 63.4±11.4 years versus 67.8±7.4, P=.4. Male/female ratios were 13:7 in controls and 16:5 in patients. Mean±SD ABP was 99.4±11.0 mm Hg in the control group and 113.1±16.8 mm Hg (P=.01) in the study group . In 8 of the carotid stenosis subjects CO2 reactivity and autoregulation were reassessed 1 month after either carotid endarterectomy (7 subjects) or percutaneous transluminal carotid angioplasty (1 subject). Local Hospital Ethics Committee approval was obtained.
Dynamic Autoregulation Testing
During the study the subjects were in a supine position with their heads slightly elevated. MCAV was recorded bilaterally simultaneously via the transtemporal window using 2 MHz transducers (DWL, Langerach). The MCAV was insonated at a mean±SD depth of 50.2±3.5 mm for the control population and 52.6±3.4 mm for the carotid stenosis population. Continuous ABP recording was made via a servo-controlled finger plethysmograph (Finapres 2300, Ohmeda), with the subject’s hand maintained at the same level as the head. Finapres provides a reliable assessment of rapid changes in ABP, but its accuracy for absolute measurement is affected by baseline shifts and unpredictable offsets. For the purposes of the autoregulation model, absolute measures of ABP were not required, and baseline measurement of resting ABP was made by automated arm cuff (Omega 1400 series, In Vivo Laboratories Inc). A sudden stepwise drop in ABP was induced by rapid deflation of bilateral thigh cuffs that had been inflated suprasystolically for 3 minutes. Only a decrease of ABP of more than 10 mm Hg was considered to be a sufficient stimulus. Autoregulatory responses were analyzed off-line using the time-averaged mean velocities of the maximum velocity outlines of the Doppler spectrum and mean ABP. Dynamic autoregulation was assessed as previously described14 18 using the software program supplied by the transcranial Doppler manufacturers. This program compares the rate of return of ABP and MCAV to baseline following the drop in ABP. Starting at the moment of cuff release and based on the actual ABP curve of the 30 seconds that followed, a series of 10 hypothetical MCAV autoregulatory curves were calculated that model between a passive MCAV and ABP relationship (ie, if the rate of rise in MCAV is identical to that of ABP, this resulted in an ARI of 0), while more rapid rates of rise in MCAV resulted in increased ARI (maximum ARI, 9). Each model curve is compared with the actual MCAV recording for the best fit (ie, lowest SEM of the differences between the actual and each hypothetical curve at each of the 30 seconds after the cuff release). Full details of the equations used have been published previously.18 Five cycles of inflation/deflation were performed per subject with a 3-minute rest interval between cycles. A mean ARI was calculated for each subject only from runs in which a sufficient magnitude ABP fall was attained. End-tidal CO2 was recorded between cycles.
CO2 was administered as 8% CO2 in air from a Douglas bag reservoir through a mask with inspiratory and expiratory limbs protected by one-way valves. End-tidal CO2 was monitored by continuous sampling from the expiratory limb using an automated capnograph (Normacap 200, Datex Instrumentation). CO2 was administered until MCAV recordings had plateaued. CO2 reactivity was calculated off-line as the percentage increase in MCAV during 8% CO2 inspiration, compared with baseline MCAV while breathing room air.8
A lower limit of the normal range of ARI and CO2 reactivity was calculated from 2 SD below the mean. Differences between groups were determined by t test or ANOVA, with Scheffé’s test for post-hoc analysis as appropriate. Correlations were determined by Pearson’s test.
Dynamic Autoregulation Testing
In 1 control subject no transtemporal acoustic window was obtained and therefore results are presented for 20 normal control subjects (40 middle cerebral arteries). Measurement of ARI was possible in 21 of the 27 carotid stenosis patients (42 middle cerebral arteries). In 2 subjects there was no transtemporal window. In 4 subjects with symptomatic peripheral vascular disease, ARI could not be measured because of poor Finapres signals in 2 and an inability to induce a stepped blood pressure reduction in 2. In 10 of the 21 subjects whose ARI could be measured, there was a contralateral stenosis >60% and therefore 31 carotid stenoses were studied in all.
Five autoregulatory runs were attempted in each subject and this was a well-tolerated procedure. In 3 subjects in the normal population it was not possible to induce a sufficient magnitude drop in MAP for each run, whereas in 12 subjects in the carotid stenosis population a sufficient MAP fall was not attained for each cycle (1 cycle in 4 subjects, 2 cycles in 6 subjects, and 3 cycles in 2 subjects). The mean ABP drop achieved in satisfactory autoregulation runs was 13.8±5.3 mm Hg. There was no significant change in resting CO2 between dynamic autoregulation runs.
The SD of the measurement error19 for the first two successive recordings was 0.87 in all the recordings, 0.83 in the patient recordings, and 0.92 in the control recordings.
Mean±SD (range) ARI in the 40 middle cerebral arteries in 20 controls was 6.3±1.1 (4.2 to 8.2). From this we determined a normal range of >4.1. Severe impairment was defined as an ARI >2.0. No control subjects had an ARI outside the normal range.
ARI was significantly reduced in middle cerebral arteries ipsilateral to a stenosed/occluded carotid artery: mean±SD 3.3±2.2, compared with 5.9±2.1 ipsilateral to a nonstenosed artery in patients (P<.002) and 6.3±1.1 for the normal controls (P<.0001). As shown in Fig 1⇓, mean ARI was progressively reduced with increasing degree of carotid stenosis (ANOVA P=.0016, Scheffé’s test P<.05 for 80% to 95% and >95% compared with <60%). However, even with severe (>80%) carotid stenosis in some cases ARI was normal (5/23), and in only a minority of cases was it severely impaired (7/23). After carotid endarterectomy or angioplasty there was a significant improvement in ARI in 8 subjects, which increased from 4.0±2.3 to 5.9±2.1 (P<.02). In all subjects in whom ARI was outside the normal range preoperatively it returned to normal postoperatively (Fig 2⇓).
In contrast to the significant relationship between ARI and degree of stenosis, there was no relationship between basal MCAV and degree of stenosis: mean±SD MCAV <60% stenosis 58.2±12.7 cm/s, 60% to 70% 57.2±22.0 cm/s, 80% to 99% 62.4±13.2 cm/s, 100% 53.5±14.2 cm/s; P=.6 via ANOVA. Consistent with this there was no relationship between ARI and basal MCAV within the carotid stenosis population (r=.17, P=.28).
CO2 reactivity testing was performed in 18 subjects in the control population, but 2 subjects were unable to tolerate the face mask. Mean±SD (range) in controls was 91.2±29.3% (48.9 to 143%), giving a normal range of >32.6%. No controls had a reactivity outside this normal range.
CO2 reactivity was significantly reduced in middle cerebral arteries ipsilateral to a stenosed/occluded carotid artery: mean±SD 50.6±40.0%, compared with 84.2±31.9% ipsilateral to a nonstenosed artery in patients (P<.01) and 91.2±29.3 for the controls (P<.0001). In the carotid stenosis population there was a correlation between ARI and CO2 reactivity (Pearson’s r=.45, P=.003). However, in a significant number of cases, CO2 reactivity was normal in the presence of an ARI below the normal range (Fig 3⇓). In all subjects end-tidal CO2 increased to greater than 7 kPa. Mean (range) end-tidal CO2 on air was 4.68 (3.9 to 5.4) kPa and on 8% CO2 was 7.94 (2.1 to 4.2) kPa.
Our results demonstrate that in a subgroup of patients with carotid artery stenosis, dynamic autoregulation is impaired, as determined by the ability of the cerebral circulation to maintain middle cerebral artery blood flow in response to a rapid reduction of ABP. In a number of cases this technique identified impaired autoregulation in patients in whom CO2 reactivity was in the normal range. This increased sensitivity may make the technique more useful in clinical practice. It is likely that at least in some patients normal CO2 reactivity but impaired dynamic autoregulation may reflect passive autoregulation due to a rise in ABP associated with inspiration of increased levels of inspired CO2.11 Dynamic autoregulation may be a more relevant marker of impaired hemodynamics. It may represent a marker of increased stroke risk, although this needs to be determined in prospective studies. In addition, patients with markedly impaired dynamic autoregulation would be expected to be at high risk from sudden reductions in blood pressure precipitated by antihypertensive medication. This technique may allow identification of such patients in whom blood pressure should be maintained at relatively higher levels.
We found this method of dynamic cerebral autoregulation testing to be a well-tolerated procedure with no side effects in either the patient group or control group, with a combined number of more than 400 individual autoregulatory runs performed in this study. Provided that an adequate stepped reduction in ABP was achieved, it was a reproducible technique. In some patients with carotid stenosis this was more difficult to achieve, perhaps reflecting the diffuse nature of the atheromatous process and slower development of reactive hyperemia after thigh cuff deflation. In 2 subjects who had symptomatic peripheral vascular disease, no ABP reduction could be achieved. Poor peripheral circulation also prevented Finapres monitoring of ABP in another 2 subjects with carotid stenosis. Monitoring of ABP continuously with the use of an oscillating upper arm cuff13 may help in such cases, although in pilot work we found this device less well tolerated than the Finapres.
The use of middle cerebral artery blood flow velocity, rather than blood flow, to determine cerebral hemodynamics makes the assumption that the diameter of the middle cerebral artery does not change during the procedure. In operative patients it has been demonstrated that the changes in MCAV induced by such a drop in ABP correlate very closely with those in internal carotid artery flow measured using a flowmeter.13 In addition, dynamic autoregulation measured by this method correlates well with static measures of autoregulation.18
This method of determining dynamic autoregulation is a simple reproducible technique that allows the hemodynamic effect of a carotid stenosis on the intracerebral circulation to be determined. It is more physiologically appropriate and may be a more sensitive measure of hemodynamic impairment than CO2 reactivity. However, it is possible that it detects an overlapping but different subgroup of compromised patients than those detected by CO2 reactivity testing and that the two tests may be synergistic. The increase in cerebral blood flow in response to hypercapnia is a distinct response with separate effectors from those involved in dynamic autoregulation, and the two may dissociate in certain circumstances, such as after severe ischemia.13 However, the very severe impairment of reactivity or autoregulation found in a subgroup of patients with carotid disease is likely to result from impaired perfusion pressure that would affect both autoregulation and CO2 reactivity. It is much greater in magnitude than the alterations we have seen in either CO2 reactivity or ARI in patients with long-standing hypertension and presumed abnormal intracerebral vasculature in the absence of large artery stenosis (authors’ unpublished data).
Dynamic autoregulatory testing may allow identification of a subgroup of patients with tight carotid stenosis who are at risk from subsequent stroke. This hypothesis needs to be examined in large prospective studies, and further comparison needs to be made of this method of measuring dynamic autoregulation with other methods of identifying impaired cerebral hemodynamics. It may also allow identification of subjects at risk of the hyperperfusion syndrome after endarterectomy and those subjects at risk of hypoperfusion and stroke in the face of reductions in ABP, such as antihypertensive therapy.
Selected Abbreviations and Acronyms
|ABP||=||arterial blood pressure|
|MAP||=||mean airway pressure|
|MCAV||=||middle cerebral artery flow velocity|
This work is supported by a Medical Research Council project grant.
- Received February 25, 1997.
- Revision received April 14, 1997.
- Accepted April 30, 1997.
- Copyright © 1997 by American Heart Association
Nicolaides A. Asymptomatic carotid stenosis. Int Angiol. 1995;14:1-4.
Ringlestein EB, Sievers C, Ecker S, Schneider PA, Otis SM. Noninvasive assessment of CO2-induced cerebral vasomotor response in normal individuals and patients with internal carotid artery occlusions. Stroke. 1988;19:963-969.
Kleiser B, Widder B. Course of carotid artery occlusions with impaired carbon dioxide reactivity. Stroke. 1992;23:171-174.
Kleiser B, Scholl D, Widder B. Assessment of cerebrovascular reactivity by Doppler CO2 and diamox testing—which is the appropriate method? Cerebrovasc Dis. 1994;4:134-138.
Smielewski P, Czosnyka M, Pickard JD, Kirkpatrick P. Clinical evaluation of near-infrared spectroscopy for testing cerebrovascular reactivity in patients with carotid artery disease. Stroke. 1997;28:331-338.
Newell DW, Aaslid R, Lam AM, Mayberg TS, Winn R. Comparison of flow and velocity during dynamic autoregulation testing in humans. Stroke. 1994;25:793-797.
Hartl WH, Janssen I, Furst H. Effect of carotid endarterectomy on patterns of cerebrovascular reactivity in patients with unilateral carotid artery stenosis. Stroke. 1994;25:1952-1957.
Markus HS, Clifton A, Buckenham T, Taylor R, Brown MM. Improvement in cerebral hemodynamics after carotid angioplasty. Stroke. 1996;27:612-616.
Roederer GO, Langlois YE, Chan ATW, Primozich JF, Lawrence RJ, Chikos PM, Strandness DE Jr. Ultrasonic duplex scanning of the extracranial carotid arteries: improved accuracy using new features from the common carotid artery. J Cardiovasc Ultrasonogr. 1982;1:373-380.
Tiecks FP, Lam AM, Aaslid R, Newell DW. Comparison of static and dynamic cerebral autoregulation measurements. Stroke. 1995;26:1014-1019.
Bland M. An introduction to medical statistics. Oxford, UK: Oxford University Press; 1987.