- Email: [email protected]
Decreased transcranial Doppler carbon dioxide reactivity is associated with disordered cerebral metabolism in patients with internal carotid artery stenosis
Decreased transcranial Doppler carbon dioxide reactivity is associated with disordered cerebral metabolism in patients with internal carotid artery stenosis G.H. Visser, MD, PhD, J. van der Grond, PhD, A.C. van Huffelen, MD, PhD, G.H. Wieneke, PhD, and B.C. Eikelboom, MD, PhD, Utrecht, The Netherlands Purpose: The hemodynamic effect of stenosis of the internal carotid artery (ICA) can be assessed by measuring, with transcranial Doppler (TCD), the carbon dioxide (CO2) reactivity of the cerebral vessels. The aim of this study was to determine whether a decreased CO2 reactivity is associated with a compromised cerebral metabolism, as evaluated with 1H magnetic resonance spectroscopy (MRS). Methods: Sixty-six patients with unilateral or bilateral stenosis of the ICA, who were scheduled for carotid endarterectomy (CEA) and who had undergone both a TCD CO2 reactivity test and a MRS examination, were included in this study. The ICA stenosis on one side (CEA side) was always more than 70%, and the extent of the stenosis on the contralateral side varied. Results: The CO2 reactivity and the N-acetyl aspartate (NAA)/choline ratio were correlated in both hemispheres (r = .43; P < .001). Patients with an ICA occlusion contralateral to the CEA side are especially at risk for disordered cerebral hemodynamics and metabolism; in the contralateral hemisphere, the mean CO2 reactivity and NAA/choline ratio were abnormal (18% and 1.52, respectively), and lactate was present in 85% of the patients. Changes indicative of disordered hemodynamics were found more often in symptomatic than in asymptomatic patients. Conclusion: A decreased CO2 reactivity appears to be associated with a disordered cerebral metabolism. Patients with severe bilateral ICA stenosis are at risk for disordered cerebral metabolism and hemodynamics. Therefore, the indication for CEA based on the degree of ICA stenosis and clinical grounds might be refined with an additional test, such as the TCD CO2 reactivity test. (J Vasc Surg 1999;30:252-60.)
The hemodynamic effect of stenosis of an internal carotid artery (ICA) depends on the severity of the stenosis and on the potential of the collateral circulation. When the ICA stenosis results in a decreased cerebral perfusion pressure (CPP), cerebral blood flow (CBF) can be kept constant with a decrease in cerebrovascular resistance (CVR). This mechanism is called cerebral autoregulation and is From the Department of Clinical Neurophysiology, University Hospital Utrecht and Rudolf Magnus Institute for Neurosciences (Drs Visser, van Huffelen, and Wieneke), and the Departments of Radiology (Dr van der Grond) and Vascular Surgery (Dr Eikelboom), University Hospital Utrecht. Reprint requests: G.H. Visser, University Hospital Utrecht, Department of Clinical Neurophysiology (F.02.230), P.O. Box 85500, 3508 GA Utrecht, The Netherlands. Copyright © 1999 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/99/$8.00 + 0 24/1/97707
252
mediated by dilatation of arteriolar cerebral vessels. If the collateral blood flow is inadequate, the CPP may decrease so much that the cerebral autoregulation mechanism fails. In this condition, cerebral arterioles are maximally or almost maximally dilated and are resistant to any further vasodilator stimulus, such as hypercapnia. The change in CBF in response to hypercapnia, which is experimentally induced by inhalation of a gas mixture with a raised carbon dioxide (CO2) content, is called CO2 reactivity. This CO2 effect is mediated by vasodilatation of the cerebral arterioles, whereas the proximal cerebral arteries are not affected.1 Because the diameter of the proximal arteries remains constant, changes in blood flow velocity (BFV) in these vessels reflect the changes in blood flow as a result of vasodilatation of the cerebral arterioles. With transcranial Doppler (TCD) sonography, the BFV in the proximal parts of cerebral arteries can be measured, and the CO2 reac-
JOURNAL OF VASCULAR SURGERY Volume 30, Number 2
tivity of the cerebral vessels is expressed as the relative change in BFV during provocation with CO2 inhalation.2 When the cerebral autoregulation mechanism fails, CBF also fails. At first, the rate of oxygen extraction from the blood increases to maintain aerobic glycolysis, but a further decrease in CPP may result in anaerobic metabolism and finally ischemic cerebral damage (low-flow infarction). The extent of anaerobic metabolism can be determined by measuring N-acetyl aspartate (NAA), choline, and lactate levels with 1H magnetic resonance spectroscopy (MRS). It has been shown that the extent of anaerobic metabolism (increased lactate and decreased NAA/choline ratio) corresponds with the severity of carotid lesions3 and is associated with a low blood flow in the middle cerebral artery (MCA).4 The aim of the present study was to investigate whether a decreased CO2 reactivity is associated with a disordered cerebral metabolism, measured with MRS, in patients with unilateral or bilateral stenosis of the ICA. This finding would support the use of the CO2 reactivity test to identify patients at risk for disordered cerebral hemodynamics. SUBJECTS AND METHODS Patients. Sixty-six patients with an ICA stenosis of at least 70% who were scheduled for carotid endarterectomy (CEA) and who had undergone both a TCD CO2 reactivity test and a MRS examination during the period from June 1995 to July 1996 were included in this study. When a bilateral ICA stenosis of more than 70% was found (n = 14), the symptomatic side was scheduled for surgery (n = 8), or, if both sides were asymptomatic (n = 5) or symptomatic (n = 1), the most severely stenosed side was operated on. The extent of ICA stenosis was determined by means of angiography (70% stenosis according to North American Symptomatic Carotid Endarterectomy Trial criteria). The hemisphere of the carotid artery scheduled for CEA was termed the CEA side, and the other hemisphere was termed the contralateral side. The Ethics Committee of the University Hospital Utrecht approved the study, and the patients gave their informed consent. The mean age of the patients was 65 ± 10 years, and 65% of the patients were men (n = 43). The contralateral ICA was occluded in 20% of the subjects (n = 13) and was stenosed by at least 70% in 21% of the subjects (n = 14) and by less than 70% in 59% of the subjects (10 patients with 30% to 69% stenosis, and 29 patients with less than 30% stenosis or no stenosis). Three groups were formed on the
Visser et al
253
basis of these differences in the extent of stenosis of the contralateral ICA. In the year before inclusion, 43 patients (65%) had symptoms only of the CEA-side hemisphere. Of these patients, 14 had a minor stroke, 20 experienced one or more transient ischemic attacks (TIAs), and nine experienced ocular symptoms (ie, seven patients with transient monocular blindness and two patients with impaired vision caused by chronic retinal ischemia). In eight patients, signs and symptoms were restricted to the contralateral side (two patients with a minor stroke and six patients with TIAs). All these eight patients had an ICA occlusion at the contralateral side. Five patients had symptoms on both sides (three patients with bilateral ocular symptoms and two patients with bilateral transient cerebral ischemic symptoms). The three patients with bilateral ocular symptoms had normal vertebrobasilar angiogram findings, and the symptoms could not be attributed to vertebrobasilar insufficiency. Ten patients (15%) were asymptomatic. The distribution of unilateral or bilateral clinical symptoms between the groups based on the contralateral ICA stenosis is shown in Table I. Transcranial Doppler with CO2 reactivity test. The TCD examination was performed with a DWL multidop-X device (Sipplingen, Germany) and two 2MHz pulsed Doppler probes. The method used has been described previously.5 First, a standard TCD examination was performed to locate the vessels. Then, the CO2 reactivity was measured simultaneously in both MCAs, the main cerebral arteries supplied by the ICA. The TCD probes were fitted in a light metal frame, which was firmly fixed to the head with two earpieces and an adjustable nose saddle (manufactured by DWL). The BFV in the MCA was recorded while the patients were lying down with their eyes closed. If an intracranial vessel could not be insonated, this was considered as a missing value and was not included in the statistical analysis. Hypercapnia was induced with inhalation of a gas mixture with 5% CO2 and 95% oxygen (carbogene) through a mouthpiece connected to a respiratory balloon. A nose clip ensured proper inhalation of carbogene. The CO2 content of the breathing gas was continuously monitored with an infrared gas analyzer (Mijnhardt). After a 2-minute baseline period, patients inhaled carbogene for 2 minutes. A spectral TCD recording of 5 seconds duration was made after 1 minute during the baseline period and after 1.5 minutes of carbogene inhalation. At the same time, blood pressure (BP) was measured with an automatic device (Omega 1000; Invivo Research Laboratories, Scott’s Valley, Calif) to
254 Visser et al
exclude the possibility that BFV changes were caused by BP changes. The CO2 reactivity after 1.5 minutes of inhalation of carbogene was measured as the relative change in BFV from the mean baseline BFV. The mean of the maximal BFV during the spectral TCD recordings was used in this calculation. Control CO2 reactivity was measured in 30 subjects without cerebrovascular disorders (mean age, 59 ± 10 years; 83% men) who were scheduled for implantation of an internal cardioverter defibrillator. In both hemispheres, the median CO2 reactivity was 52%, and the fifth percentile was 30%. Based on these data, we considered a CO2 reactivity below 30% to be abnormal. Magnetic resonance imaging and MRS. MRI and MRS studies were performed on a 1.5T whole body system (ACS/NT15 model; Philips Medical Systems, Best, The Netherlands). For MRI, the following pulse sequences were used: a sagittal T1weighted spin-echo sequence (repetition time, 545 ms; echo time, 15 ms; one signal acquired; 4-mm section thickness; 0.6-mm intersection gap; 225-mm FOV; 256 × 256 matrix), followed by a transaxial T2weighted spin-echo sequence (repetition time, 2 seconds; echo time, 20 and 100 ms; two signals acquired; 7-mm section thickness; 1.5-mm intersection gap; 225-mm field of view; 256 × 256 matrix). As in a previous study, 1H MRS was performed with a single volume technique, using T2-weighted transaxial MRI images. In each subject, two single volumes of interest (VOI) were selected in the centrum semiovale, one on the CEA side and one on the contralateral side. The anterior-posterior and leftright dimensions of the VOI were chosen so that regions containing subcutaneous lipid were excluded and were typically 70 mm and 35 mm in anteriorposterior and left-right directions. In all subjects, the caudal-cranial dimension of the VOI was 15 mm. All VOI were positioned at least 2 cm away from gray/white matter hyperintensities. In each MR examination, the dimensions of the selected VOI were the same for both hemispheres. Thus, all VOI contained mainly white matter. After selection of a VOI, the 90-degree pulse length was determined. To minimize eddy currents and to maximize the water echo signal, localized spectroscopy was first performed without water suppression for adjustments of the gradients (gradient tuning). This was followed by localized automatic shimming of the VOI, which resulted in a water resonance line-width of 6 Hz (full width at half maximum) or less. Water suppression was performed with selective excitation (60-Hz bandwidth), followed by a spoiler gradient. A double
JOURNAL OF VASCULAR SURGERY August 1999
spin-echo point-resolved spectroscopy (PRESS) sequence was used for VOI localization. Each measurement was performed with a repetition time of 2 seconds and an echo time of 136 ms, and 2048 time domain data points, 4000-Hz spectral width, and 64 or 128 signals were acquired. After zerofilling to 4096 data points, Gaussian multiplication of 5 Hz, exponential multiplication of 4 Hz (line broadening), Fourier transformation and linear baseline correction, total choline, NAA (referenced at 2.01 ppm), and lactate peaks were identified by their chemical shifts. Because this MRS method does not enable us to calculate absolute metabolic concentrations, metabolic data are expressed as the ratio between peak intensities (NAA/choline ratio). To distinguish lactate resonances from lipid resonances at an echo time of 136 ms, lactate was defined as an inverted resonance at 1.33 ppm, with a signal-to-noise ratio larger than 2 and a clearly identifiable 7-Hz J-coupling. Lactate was documented as either absent or present. In 19 healthy volunteers, (mean age, 58 ± 11 years; 74% men), the mean NAA/choline ratio was 1.93 ± 0.17. Therefore, we considered a ratio of less than 1.59 to be decreased (ie, < mean – 2SD). Statistical analysis. For continuous variables, the t test for independent samples was used for the analysis of differences between two groups. For differences among more than two groups, analysis of variance (ANOVA) was used for the first overall analysis. When the ANOVA procedure showed a significant difference, the least significance difference (LSD) method was used for further analysis of differences. All values are given as means ± SD, unless otherwise indicated. The relation between two continuous variables is expressed as the Pearson correlation coefficient. Data for lactate were treated as ordinal variables, because in many cases no lactate peak could be detected in the MRS spectra. For noncontinuous variables, differences between groups in cross-tabs were assessed with Pearson’s chi-square test or Fisher’s exact test. A P value less than .05 was considered statistically significant. RESULTS It was not possible to perform a CO2 reactivity test in eight patients at the CEA side (12%) and in 10 patients at the contralateral side (15%) because of the inability to insonate the MCAs through the temporal bone. The mean CO2 reactivity on the CEA side was 40% ± 19%, and on the contralateral side, it was 41% ± 22%. For those 56 patients in whom a CO2 reactivity test at one or both sides could be performed, the mean BP increased 6% from 113 mm
JOURNAL OF VASCULAR SURGERY Volume 30, Number 2
Visser et al
255
Fig 1A. Transcranial Doppler registration of the blood flow velocity in both middle cerebral arteries during the CO2 reactivity test. The 51-year-old female subject has more than 70% stenosis of the left internal carotid artery (carotid endarterectomy side) and occlusion of the right internal carotid artery (contralateral side). The CO2 reactivity on the carotid endarterectomy side was 31% and on the contralateral side 12% (ET CO2, end-tidal CO2).
Hg during the baseline period to 120 mm Hg during the carbogene inhalation (P < .001). If cerebral autoregulation is disregarded, such a slight increase in BP could only explain an increase in BFV of 6% during the CO2 reactivity test. However, autoregulation would normally ensure a stable CBF (and BFV in the MCA) with such fluctuations in BP. MRS could be performed in all patients. The results for a 51-year-old woman with occlusion of the contralateral ICA are shown in Fig 1. The CO2 reactivity on the CEA side was 31%, and on the contralateral side, it was 12% (Fig 1A). After CEA, the CO2 reactivity increased to 44% on the CEA side and to 40% on the contralateral side. The rectangular box in the MRI shown in Fig 1B indicates the dimensions of the VOI chosen, and the corresponding 1H MR spectra of both sides are shown in Fig 1C. The NAA/choline ratio was 1.82 on the CEA side and 1.64 on the contralateral side. Lactate was present in both hemispheres. There was a modest but statistically significant correlation (r = .43) between the CO2 reactivity and the NAA/choline ratio in both hemispheres (P < .001; Fig 2). In both hemispheres, the CO2 reactivity was significantly lower in patient groups with a decreased NAA/choline ratio in the corresponding hemisphere (Table II). In groups formed on the basis of the presence of lactate, only in the corresponding contralateral hemisphere was the CO2 reactivity significantly decreased, although a similar trend was found at the CEA side (Table III). In nine patients, a bilateral decreased CO2 reactivity was
Fig 1B. Transversal T2-weighted magnetic resonance image showing the centrum semiovale from the same patient represented in Fig 1A. The white box within the brain parenchyma indicates the volume of interest selected for 1H magnetic resonance spectroscopy.
Fig 1C. 1H magnetic resonance spectra from the volumes of interest selected in the left and right hemispheres from the patient shown in Figs 1A and 1B. The chemical-shift axis in parts per million (ppm) is positioned below each spectrum (Cho, choline; Cr, creatine; NAA, N-acetyl aspartate; Lac, lactate). The NAA/choline ratio on the carotid endarterectomy side was 1.82, and on the contralateral side, it was 1.64. Lactate was present in both hemispheres.
found. In this subgroup, the presence of a decreased NAA/choline ratio or lactate was not significantly different, compared with the patients with an unilateral decreased CO2 reactivity at the CEA side.
JOURNAL OF VASCULAR SURGERY August 1999
256 Visser et al
Table I. The distribution of clinical symptomatology between groups based on the degree of contralateral internal carotid artery stenosis Degree of contralateral ICA stenosis Clinical symptoms
Fig 2. Relation between the CO2 reactivity and the N-acetyl aspartate (NAA)/choline ratio on the carotid endarterectomy side •; solid trend line) and on the contralateral side (o; dashed trend line). Trend lines represent an assumed linear correlation between these variables (with r = 0.43 for both lines).
The CO2 reactivity, NAA/choline ratio, and the presence of lactate on the CEA side were not significantly different for the three groups formed on the basis of the contralateral ICA stenosis, but were on the contralateral side (Table IV). The CO2 reactivity and NAA/choline ratio on the contralateral side were significantly lower in the group with an occluded ICA than in the groups of patients with a more than 70% stenosis or less than 70% stenosis. In addition, lactate was present in 85% of the patients in the occluded ICA group, compared with 29% and 13% of the patients in the groups with more than 70% or less than 70% stenosis, respectively. The CO2 reactivity or the presence of lactate on the CEA side was not significantly different between groups formed on the basis of symptoms on this side (Table V). However, the NAA/choline ratio was lower in the groups of patients who had a minor stroke or transient symptoms than in the group of patients with ocular symptoms or the asymptomatic group. If groups were formed on the basis of symptoms on the contralateral side (Table VI), both the CO2 reactivity and the NAA/choline ratio on this contralateral side were lower in the groups of patients who had a minor stroke or transient cerebral symptoms than in the group of patients with ocular symptoms or the asymptomatic group. In addition, lactate was present in 77% of the symptomatic patients (either ocular symptoms, transient symptoms, or minor stroke), which contrasts with the presence of lactate in only 19% of the patients in the asymptomatic group (P < .001 with Fisher’s exact test).
< 70%
CEA side only 33 Ocular 7 TIA 16 Minor stroke 10 Contralateral side only – TIA – Minor stroke – Bilateral symptoms 2 Bilateral ocular 1 Bilateral TIA 1 Asymptomatic 4
≥ 70%
Occlusion
Total
8 2 3 3 – – – 1 1 – 5
2 – 1 1 8 6 2 2 1 1 1
43 9 20 14 8 6 2 5 3 2 10
ICA, Internal carotid artery; CEA, carotid endarterectomy; TIA, transient ischemic attack.
DISCUSSION The hemodynamic effect of ICA obstruction is dependent on the individual potential of the collateral cerebral circulation.6-8 An ipsilaterally lowered CO2 reactivity can be found in patients with severe stenosis of the ICA, especially when the contralateral ICA is also obstructed.9-14 The CO2 reactivity might be superior to the measured degree of ICA stenosis alone as an indicator of the hemodynamic consequences of ICA stenosis, because CO2 reactivity also depends on the quality of the collateral circulation. As illustration of the clinical significance of the CO2 reactivity, especially in patients with low-flow infarctions, patients with chronic retinal ischemia or hypostatic TIAs have a lowered CO2 reactivity.13,15-20 Also, a low or absent cerebrovascular reactivity, either measured with CO2 or acetazolamide as vasodilator stimulus, is a risk factor for subsequent ischemic cerebral complications.21-24 Finally, after CEA, the CO2 reactivity has been shown to increase on both sides, reflecting the improvement of cerebral hemodynamics.5,25-27 Similar results were found with alternative methods to measure cerebrovascular reserve based on changes in regional cerebral blood flow and/or volume, such as with single-photon emission computed tomography, positron emission tomography, or MRI.28-30 Also, an association was found between a decreased acetazolamide response and the presence of high intensity signals on MRI, especially when located in the centrum semiovale.31 With MRS, a decreased NAA and choline concentration and increased lactate concentration can be found in the
JOURNAL OF VASCULAR SURGERY Volume 30, Number 2
Visser et al
257
Table II. Difference in CO2 reactivity and percentage of patients with decreased CO2 reactivity (less than 30%) between groups formed on the basis of the magnetic resonance spectroscopy N-acetyl aspartate (NAA)/choline ratio NAA/choline ratio on CEA side
Statistics
CEA side
Normal (≥ 1.59)
Decreased (< 1.59)
Results
P value
CO2 reactivity (%)
44 ± 19 (n = 39) 18% (n = 7)
31 ± 17 (n = 19) 53% (n = 10)
t (56) = –2.6 (t test)
< .05
Decreased (< 30%)*
NAA/choline ratio on contralateral side Contralateral side CO2 reactivity (%) Decreased (<30%)*
Statistics
Normal (≥1.59)
Decreased (<1.59)
Results
P value
45 ± 19 (n = 44) 25% (n = 11)
28 ± 26 (n = 12) 50% (n = 6)
t (54) = –2.5 (t test)
<.05
*Based on control data of 30 subjects (see methods section). CEA, Carotid endarterectomy.
Table III. Difference in CO2 reactivity and percentage of patients with decreased CO2 reactivity (less than 30%) between groups formed on the basis of the magnetic resonance spectroscopy lactate presence Lactate on CEA side CEA side CO2 reactivity (%) Decreased (< 30%)*
Absent 43 ± 19 (n = 27) 19% (n = 5)
Statistics
Present 36 ± 19 (n = 31) 39% (n = 12)
Results t (56) = 1.4 (t test)
Lactate on contralateral side Contralateral side CO2 reactivity (%) Decreased (< 30%)*
Absent 48 ± 20 (n = 38) 18% (n = 7)
P value NS
Statistics
Present 27 ± 19 (n = 18) 56% (n = 10)
Results t (54) = 3.8 (t test)
P value < .001
*Based on control data of 30 subjects (see methods section). CEA, Carotid endarterectomy.
area of a cerebral infarct. Similar changes, reflecting disordered metabolism, are also found in noninfarcted cerebral regions with a reduced MCA blood flow.4 A statistically significant correlation was found between the CO2 reactivity and the NAA/choline ratio for both hemispheres. However, the correlation was only moderate. A possible explanation is that a short period of compromised cerebral circulation may result in long-term changes in NAA, choline, and lactate concentrations. These changes can be detected for weeks or months, even after restoration of the cerebral circulation. In contrast, CO2 reactivity can be expected to reflect the temporary situation. So, when formerly compromised hemodynamics have been restored as a result of collateral pathway development, or when the disordered metabolism was the result of a temporary decrease in blood pres-
sure, MRS findings may still be abnormal, whereas the CO2 reactivity may be normal. In the present study, hemodynamics were most compromised in patients with an ICA occlusion on the contralateral side. On this contralateral side, both the mean CO2 reactivity and NAA/choline ratio were decreased to values we consider abnormal, and in 85% of the cases, lactate was found. In the group of patients with more than 70% stenosis and the group of patients with maximally 70% stenosis on the contralateral side, CO2 reactivity and the NAA/choline ratio were normal, and lactate was present in considerably fewer patients. The presence of an ICA occlusion on the contralateral side might be expected to influence the variables measured on the CEA side, because collateral flow would be strongly directed towards the side of the ICA occlusion, thus limiting collateral flow to the
JOURNAL OF VASCULAR SURGERY August 1999
258 Visser et al
Table IV. Differences between groups formed on the basis of the extent of internal carotid artery stenosis contralateral to the carotid endarterectomy side and percentage of patients with abnormal test values ICA stenosis on contralateral side ≥ 70%
< 70% CEA side CO2 reactivity (%)
40 ± 17 (n = 33)
45 ± 27 (n = 12)
Decreased (< 30%)† 27% (n = 9) 25% (n = 3) NAA/choline ratio 1.65 ± 0.23 (n = 39) 1.73 ± 0.24 (n = 14) Decreased (< 1.59)‡ Lactate: Absent (n) Present (n) Present (%) Contralateral side CO2 reactivity (%)
Occlusion
Results
P value
33 ± 17 (n = 13)
F (2, 55) = 1.1 (ANOVA)
NS
F (2, 63) = 1.1 (ANOVA)
NS
NS
39% (n = 5) 1.62 ± 0.15 (n = 13)
41% (n= 16)
29% (n= 4)
15% (n= 2)
19 20 51%
5 9 64%
7 6 46%
X2 = 1.0 (cross-tabs)
50 ± 13 (n= 32)
43 ± 27 (n= 12)
18 ± 18 (n= 12)
F (2, 53) = 14.4 (ANOVA)
< .001*
42% (n= 5) 1.87 ± 0.17 (n= 14)
75% (n= 9) 1.52 ± 0.14 (n= 13)
F (2, 63) = 12.3 (ANOVA)
< .001*
13% (n= 5)
0% (n= 0)
54% (n= 7)
34 5 13%
10 4 29%
2 11 85%
Decreased (< 30%)† 9% (n= 3) NAA/choline ratio 1.81 ± 0.23 (n= 39) Decreased (< 1.59)‡ Lactate: Absent (n) Present (n) Present (%)
Statistics
X2 = 23.8 (cross-tabs)
< .001
*Least significance difference tests: Group with an internal carotid artery occlusion was significantly different from the other two groups. †Based on control data of 30 subjects (see methods section). ‡Based on control data of 19 subjects (see methods section). ICA, Internal carotid artery; CEA, carotid endarterectomy; NAA, N-acetyl aspartate; ANOVA, analysis of variance.
Table V. Differences between groups formed on the basis of clinical symptoms on the carotid endarterectomy side and percentage of patients with abnormal test values Symptoms on CEA side Asymptomatic CEA side CO2 reactivity (%) Decreased (< 30%)† NAA/choline ratio Decreased (< 1.59)‡ Lactate: Absent (n) Present (n) Present (%)
Statistics
Ocular
TIA
Minor stroke
Results
P value
42 ± 25 (n = 18) 28% (n = 5) 1.74 ± 0.16 (n = 18) 11% (n = 2)
49 ± 9 (n = 9) 0% (n = 0) 1.83 ± 0.17 (n = 12) 0% (n = 0)
39 ± 15 (n = 19) 26% (n = 5) 1.61 ± 0.19 (n = 22) 46% (n = 10)
30 ± 19 (n = 12) 58% (n = 7) 1.51 ± 0.24 (n = 14) 71% (n =10)
F (3, 54) = 1.9 (ANOVA)
NS
F (3, 62) = 7.3 (ANOVA)
< .001*
11 7 39%
8 4 33%
9 13 59%
3 11 79%
X2 = 7.3 (cross-tabs)
NS
*Least significance difference tests: Asymptomatic and ocular symptoms groups were significantly different from the groups with transient ischemic attack or a minor stroke. †Based on control data of 30 subjects (see methods section). ‡Based on control data of 19 subjects (see methods section). CEA, Carotid endarterectomy; TIA, transient ischemic attack; NAA, N-acetyl aspartate; ANOVA, analysis of variance.
CEA side. Although in a previous study a statistically significant decreased CO2 reactivity was found on the CEA side in these patients,5 in the present study, only a trend towards a lower CO2 reactivity and
NAA/choline ratio was found for the CEA side. However, not all patients with a contralateral ICA occlusion show abnormal CO2 reactivity and MRS values, and abnormal values are also found in patients
JOURNAL OF VASCULAR SURGERY Volume 30, Number 2
Visser et al
259
Table VI. Differences between groups formed on the basis of clinical symptoms on the contralateral side and percentage of patients with abnormal test values Symptoms on contralateral side Asymptomatic CEA side CO2 reactivity (%)
46 ± 19 (n = 44) Decreased (< 30%)† 23% (n = 10) NAA/choline ratio 1.81 ± 0.22 (n = 53) Decreased (< 1.59)‡ 11% (n = 6) Lactate: Absent (n) 43 Present (n) 10 Present (%) 19%
Statistics
Ocular
TIA
Minor stroke
Results
51 ± 24 (n = 3) 33% (n = 1) 1.84 ± 0.25 (n = 3) 0% (n = 0)
20 ± 21 (n = 7) 57% (n = 4) 1.52 ± 0.17 (n = 8) 50% (n = 4)
8 ± 15 (n = 2) 100% (n = 2) 1.46 ± 0.04 (n =2) 100% (n = 2)
F (3, 52) = 5.8 (ANOVA)
2 1 33%
0 8 100%
1 1 50%
§
F (3, 62) = 5.8 (ANOVA)
P value
< .01*
< .01*
§
*Least significance difference tests: Asymptomatic and ocular symptoms groups were significantly different from the groups with transient ischemic attack or a minor stroke. †Based on control data of 30 subjects (see methods section). ‡Based on control data of 19 subjects (see methods section). §Fisher’s exact test of asymptomatic versus symptomatic (ocular, transient ischemic attack, or minor stroke); P < .001. TIA, Transient ischemic attack; NAA, N-acetyl aspartate; ANOVA, analysis of variance.
with a less severe degree of contralateral stenosis. The variance in these findings may be explained by different collateral contributions from, for example, the basilar system. Therefore, a contralateral ICA occlusion can only be regarded as a risk factor for disordered cerebral hemodynamics and metabolism. Additional tests, such as the CO2 reactivity test, are required to obtain a stratified classification of patients at risk for cerebral ischemic complications. The groups with cerebral symptoms (transient or minor stroke) had a lower mean NAA/choline ratio than the asymptomatic group (including only ocular symptoms). Lactate was more often present in the groups with cerebral symptoms, although differences were only statistically significant on the contralateral side. The CO2 reactivity of the groups with cerebral symptoms was also decreased only on the contralateral side. Thus, in general, changes indicative of disordered hemodynamics were found more often in symptomatic patients than in asymptomatic patients. This suggests that the hemodynamic effect of an ICA stenosis contributes to the transition from asymptomatic to symptomatic stenosis. In conclusion, a decreased CO2 reactivity is indicative of not only inadequate collateral blood supply, but also of a disordered cerebral metabolism, as indicated by the abnormal MRS findings. This is in accordance with the finding that a decreased CO2 reactivity in patients with an ICA stenosis is a risk factor for ischemic cerebral events.6,32,33 Our results show that patients with a contralateral ICA occlu-
sion are especially at risk for disordered cerebral hemodynamics and metabolism. However, not all of these patients with severe bilateral ICA stenosis have disordered hemodynamics, and decisions about the indication for CEA based on clinical grounds alone might be open to refinement. Our results support the use of the CO2 reactivity test to identify patients with disordered hemodynamics at one or both sides, and this test might have the potential to refine the indication for CEA. REFERENCES 1. Kleiser B, Scholl D, Widder B. Doppler CO2 and Diamox test: Decreased reliability by changes of vessel diameter? Cerebrovasc Dis 1995;5:397-402. 2. Bishop CCR, Powell S, Rutt D, Browse NL. Transcranial Doppler measurement of middle cerebral artery blood flow velocity: A validation study. Stroke 1986;17:913-5. 3. Van der Grond J, Balm R, Kappelle LJ, Eikelboom BC, Mali WPThM. Cerebral metabolism of patients with stenosis or occlusion of the internal carotid artery. A 1H-MR spectroscopic imaging study. Stroke 1995;26:822-8. 4. Van der Grond J, Eikelboom BC, Mali WPThM. Flow-related anaerobic metabolic changes in patients with severe stenosis of the internal carotid artery. Stroke 1996;27:2026-32. 5. Visser GH, Van Huffelen AC, Wieneke GH, Eikelboom BC. Bilateral increase in CO2 reactivity after unilateral carotid endarterectomy. Stroke 1997;28:899-905. 6. Powers WJ. Cerebral hemodynamics in ischemic cerebrovascular disease. Ann Neurol 1991;29:231-40. 7. Powers WJ, Press GA, Grubb RL, Gado M, Raichle ME. The effect of hemodynamically significant carotid artery disease on the hemodynamic status of the cerebral circulation. Ann Intern Med 1987;106:27-35.
260 Visser et al
8. Fürst H, Hartl WH, Janssen I, Fink B, Piepgras A, Lauterjung L, et al. Pathophysiology of unilateral high-grade carotid artery stenosis: Evaluation of intracranial haemodynamics by analysis of velocity waveforms from the middle cerebral artery. Clin Sci 1992;83:357-66. 9. Widder B, Kleiser B, Krapf H. Course of cerebrovascular reactivity in patients with carotid artery occlusions. Stroke 1994;25:1963-7. 10. Markus HS, Harrison MJ. Estimation of cerebrovascular reactivity using transcranial Doppler, including the use of breathholding as the vasodilatory stimulus. Stroke 1992;23:668-73. 11. Miller JD, Smith RR, Holaday HR. Carbon dioxide reactivity in the evaluation of cerebral ischemia. Neurosurgery 1992;30:518-21. 12. Maeda H, Matsumoto M, Handa N, Hougaku H, Ogawa S, Itoh T, et al. Reactivity of cerebral blood flow to carbon dioxide in various types of ischemic cerebrovascular disease: Evaluation by the transcranial Doppler method. Stroke 1993;24:670-5. 13. Ringelstein 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;19:963-9. 14. Quilitz B, Rimpel J, Mehdorn M, Rohm N, Lehmann HJ. Cerebrovascular reserve capacity in carotid artery disease: A routine test in selection for surgical treatment? Thorac Cardiovasc Surg 1988;36:217-20. 15. Bullock R, Mendelow AD, Bone I, Patterson J, Macleod WN, Allardice G. Cerebral blood flow and CO2 responsiveness as an indicator of collateral reserve capacity in patients with carotid arterial disease. Br J Surg 1985;72:348-51. 16. Kleiser B, Krapf H, Widder B. Carbon dioxide reactivity and patterns of cerebral infarction in patients with carotid artery occlusion. J Neurol 1991;238:392-4. 17. Ringelstein EB, Weiller C, Weckesser M, Weckesser S. Cerebral vasomotor reactivity is significantly reduced in lowflow as compared to thromboembolic infarctions: The key role of the circle of Willis. J Neurol Sci 1994;121:103-9. 18. Kleiser B, Widder B, Hackspacher J, Schmid P. Comparison of Doppler CO2 test, patterns of infarction in CTT, and clinical symptoms in carotid artery occlusions. Neurosurg Rev 1991;14:267-9. 19. Baumgartner RW, Regard M. Role of impaired CO2 reactivity in the diagnosis of cerebral low flow infarcts. J Neurol Neurosurg Psychiatry 1994;57:814-7. 20. Weiller C, Ringelstein EB, Reiche W, Buell U. Clinical and hemodynamic aspects of low-flow infarcts. Stroke 1991;22: 1117-23.
JOURNAL OF VASCULAR SURGERY August 1999
21. Kleiser B, Widder B. Course of carotid artery occlusions with impaired cerebrovascular reactivity. Stroke 1992;23:171-4. 22. Yonas H, Smith HA, Durham SR, Pentheny SL, Johnson DW. Increased stroke risk predicted by compromised cerebral blood flow reactivity. J Neurosurg 1993;79:483-9. 23. Webster MW, Makaroun MS, Steed DL, Smith HA, Johnson DW, Yonas H. Compromised cerebral blood flow reactivity is a predictor of stroke in patients with symptomatic carotid artery occlusive disease. J Vasc Surg 1995;21:338-45. 24. Gur AY, Bova I, Bornstein NM. Is impaired cerebral vasomotor reactivity a predictive factor of stroke in asymptomatic patients? Stroke 1996;27:2188-90. 25. Hartl WH, Janssen I, Fürst H. Effect of carotid endarterectomy on patterns of cerebrovascular reactivity in patients with unilateral carotid artery stenosis. Stroke 1994;25:1952-7. 26. Silvestrini M, Troisi E, Matteis M, Cupini LM, Caltagirone C. Transcranial Doppler assessment of cerebrovascular reactivity in symptomatic and asymptomatic severe carotid stenosis. Stroke 1996;27:1970-3. 27. Bornstein NM, Gur AY, Shifrin EG, Morag BA. Does carotid endarterectomy modify cerebral vasomotor reactivity? Cerebrovasc Dis 1997;7:201-4. 28. Russell D, Dybevold S, Kjartansson O, Nyberg-Hansen R, Rootwelt K, Wiberg J. Cerebral vasoreactivity and blood flow before and 3 months after carotid endarterectomy. Stroke 1990;21:1029-32. 29. Naylor AR, Merrick MV, Gillespie I, Sandercock PAG, Warlow CP, Cull RE, et al. Prevalence of impaired cerebrovascular reserve in patients with symptomatic carotid artery disease. Br J Surg 1994;81:45-8. 30. Gückel FJ, Brix G, Schmiedek P, Piepgras A, Becker G, Köpke J, et al. Cerebrovascular reserve capacity in patients with occlusive cerebrovascular disease: Assessment with dynamic susceptibility contrast-enhanced MR imaging and the acetazolamide stimulation test. Radiology 1996;201:405-12. 31. Isaka Y, Nagano K, Narita M, Ashida K, Imaizumi M. High signal intensity on T2-weighted magnetic resonance imaging and cerebral hemodynamic reserve in carotid occlusive disease. Stroke 1997;28:354-7. 32. Yonas H, Pindzola R. Physiological determination of cerebrovascular reserves and its use in clinical management. Cerebrovasc Brain Metab Rev 1994;6:325-40. 33. Czosnyka M, Harris NG, Pickard JD, Piechnik S. CO2 cerebrovascular reactivity as a function of perfusion pressure—A modeling study. Acta Neurochir (Wien) 1993;121:159-65.
Submitted Sep 22, 1998; accepted Jan 11, 1999.
The hemodynamic effect of stenosis of an internal carotid artery (ICA) depends on the severity of the stenosis and on the potential of the collateral circulation. When the ICA stenosis results in a decreased cerebral perfusion pressure (CPP), cerebral blood flow (CBF) can be kept constant with a decrease in cerebrovascular resistance (CVR). This mechanism is called cerebral autoregulation and is From the Department of Clinical Neurophysiology, University Hospital Utrecht and Rudolf Magnus Institute for Neurosciences (Drs Visser, van Huffelen, and Wieneke), and the Departments of Radiology (Dr van der Grond) and Vascular Surgery (Dr Eikelboom), University Hospital Utrecht. Reprint requests: G.H. Visser, University Hospital Utrecht, Department of Clinical Neurophysiology (F.02.230), P.O. Box 85500, 3508 GA Utrecht, The Netherlands. Copyright © 1999 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/99/$8.00 + 0 24/1/97707
252
mediated by dilatation of arteriolar cerebral vessels. If the collateral blood flow is inadequate, the CPP may decrease so much that the cerebral autoregulation mechanism fails. In this condition, cerebral arterioles are maximally or almost maximally dilated and are resistant to any further vasodilator stimulus, such as hypercapnia. The change in CBF in response to hypercapnia, which is experimentally induced by inhalation of a gas mixture with a raised carbon dioxide (CO2) content, is called CO2 reactivity. This CO2 effect is mediated by vasodilatation of the cerebral arterioles, whereas the proximal cerebral arteries are not affected.1 Because the diameter of the proximal arteries remains constant, changes in blood flow velocity (BFV) in these vessels reflect the changes in blood flow as a result of vasodilatation of the cerebral arterioles. With transcranial Doppler (TCD) sonography, the BFV in the proximal parts of cerebral arteries can be measured, and the CO2 reac-
JOURNAL OF VASCULAR SURGERY Volume 30, Number 2
tivity of the cerebral vessels is expressed as the relative change in BFV during provocation with CO2 inhalation.2 When the cerebral autoregulation mechanism fails, CBF also fails. At first, the rate of oxygen extraction from the blood increases to maintain aerobic glycolysis, but a further decrease in CPP may result in anaerobic metabolism and finally ischemic cerebral damage (low-flow infarction). The extent of anaerobic metabolism can be determined by measuring N-acetyl aspartate (NAA), choline, and lactate levels with 1H magnetic resonance spectroscopy (MRS). It has been shown that the extent of anaerobic metabolism (increased lactate and decreased NAA/choline ratio) corresponds with the severity of carotid lesions3 and is associated with a low blood flow in the middle cerebral artery (MCA).4 The aim of the present study was to investigate whether a decreased CO2 reactivity is associated with a disordered cerebral metabolism, measured with MRS, in patients with unilateral or bilateral stenosis of the ICA. This finding would support the use of the CO2 reactivity test to identify patients at risk for disordered cerebral hemodynamics. SUBJECTS AND METHODS Patients. Sixty-six patients with an ICA stenosis of at least 70% who were scheduled for carotid endarterectomy (CEA) and who had undergone both a TCD CO2 reactivity test and a MRS examination during the period from June 1995 to July 1996 were included in this study. When a bilateral ICA stenosis of more than 70% was found (n = 14), the symptomatic side was scheduled for surgery (n = 8), or, if both sides were asymptomatic (n = 5) or symptomatic (n = 1), the most severely stenosed side was operated on. The extent of ICA stenosis was determined by means of angiography (70% stenosis according to North American Symptomatic Carotid Endarterectomy Trial criteria). The hemisphere of the carotid artery scheduled for CEA was termed the CEA side, and the other hemisphere was termed the contralateral side. The Ethics Committee of the University Hospital Utrecht approved the study, and the patients gave their informed consent. The mean age of the patients was 65 ± 10 years, and 65% of the patients were men (n = 43). The contralateral ICA was occluded in 20% of the subjects (n = 13) and was stenosed by at least 70% in 21% of the subjects (n = 14) and by less than 70% in 59% of the subjects (10 patients with 30% to 69% stenosis, and 29 patients with less than 30% stenosis or no stenosis). Three groups were formed on the
Visser et al
253
basis of these differences in the extent of stenosis of the contralateral ICA. In the year before inclusion, 43 patients (65%) had symptoms only of the CEA-side hemisphere. Of these patients, 14 had a minor stroke, 20 experienced one or more transient ischemic attacks (TIAs), and nine experienced ocular symptoms (ie, seven patients with transient monocular blindness and two patients with impaired vision caused by chronic retinal ischemia). In eight patients, signs and symptoms were restricted to the contralateral side (two patients with a minor stroke and six patients with TIAs). All these eight patients had an ICA occlusion at the contralateral side. Five patients had symptoms on both sides (three patients with bilateral ocular symptoms and two patients with bilateral transient cerebral ischemic symptoms). The three patients with bilateral ocular symptoms had normal vertebrobasilar angiogram findings, and the symptoms could not be attributed to vertebrobasilar insufficiency. Ten patients (15%) were asymptomatic. The distribution of unilateral or bilateral clinical symptoms between the groups based on the contralateral ICA stenosis is shown in Table I. Transcranial Doppler with CO2 reactivity test. The TCD examination was performed with a DWL multidop-X device (Sipplingen, Germany) and two 2MHz pulsed Doppler probes. The method used has been described previously.5 First, a standard TCD examination was performed to locate the vessels. Then, the CO2 reactivity was measured simultaneously in both MCAs, the main cerebral arteries supplied by the ICA. The TCD probes were fitted in a light metal frame, which was firmly fixed to the head with two earpieces and an adjustable nose saddle (manufactured by DWL). The BFV in the MCA was recorded while the patients were lying down with their eyes closed. If an intracranial vessel could not be insonated, this was considered as a missing value and was not included in the statistical analysis. Hypercapnia was induced with inhalation of a gas mixture with 5% CO2 and 95% oxygen (carbogene) through a mouthpiece connected to a respiratory balloon. A nose clip ensured proper inhalation of carbogene. The CO2 content of the breathing gas was continuously monitored with an infrared gas analyzer (Mijnhardt). After a 2-minute baseline period, patients inhaled carbogene for 2 minutes. A spectral TCD recording of 5 seconds duration was made after 1 minute during the baseline period and after 1.5 minutes of carbogene inhalation. At the same time, blood pressure (BP) was measured with an automatic device (Omega 1000; Invivo Research Laboratories, Scott’s Valley, Calif) to
254 Visser et al
exclude the possibility that BFV changes were caused by BP changes. The CO2 reactivity after 1.5 minutes of inhalation of carbogene was measured as the relative change in BFV from the mean baseline BFV. The mean of the maximal BFV during the spectral TCD recordings was used in this calculation. Control CO2 reactivity was measured in 30 subjects without cerebrovascular disorders (mean age, 59 ± 10 years; 83% men) who were scheduled for implantation of an internal cardioverter defibrillator. In both hemispheres, the median CO2 reactivity was 52%, and the fifth percentile was 30%. Based on these data, we considered a CO2 reactivity below 30% to be abnormal. Magnetic resonance imaging and MRS. MRI and MRS studies were performed on a 1.5T whole body system (ACS/NT15 model; Philips Medical Systems, Best, The Netherlands). For MRI, the following pulse sequences were used: a sagittal T1weighted spin-echo sequence (repetition time, 545 ms; echo time, 15 ms; one signal acquired; 4-mm section thickness; 0.6-mm intersection gap; 225-mm FOV; 256 × 256 matrix), followed by a transaxial T2weighted spin-echo sequence (repetition time, 2 seconds; echo time, 20 and 100 ms; two signals acquired; 7-mm section thickness; 1.5-mm intersection gap; 225-mm field of view; 256 × 256 matrix). As in a previous study, 1H MRS was performed with a single volume technique, using T2-weighted transaxial MRI images. In each subject, two single volumes of interest (VOI) were selected in the centrum semiovale, one on the CEA side and one on the contralateral side. The anterior-posterior and leftright dimensions of the VOI were chosen so that regions containing subcutaneous lipid were excluded and were typically 70 mm and 35 mm in anteriorposterior and left-right directions. In all subjects, the caudal-cranial dimension of the VOI was 15 mm. All VOI were positioned at least 2 cm away from gray/white matter hyperintensities. In each MR examination, the dimensions of the selected VOI were the same for both hemispheres. Thus, all VOI contained mainly white matter. After selection of a VOI, the 90-degree pulse length was determined. To minimize eddy currents and to maximize the water echo signal, localized spectroscopy was first performed without water suppression for adjustments of the gradients (gradient tuning). This was followed by localized automatic shimming of the VOI, which resulted in a water resonance line-width of 6 Hz (full width at half maximum) or less. Water suppression was performed with selective excitation (60-Hz bandwidth), followed by a spoiler gradient. A double
JOURNAL OF VASCULAR SURGERY August 1999
spin-echo point-resolved spectroscopy (PRESS) sequence was used for VOI localization. Each measurement was performed with a repetition time of 2 seconds and an echo time of 136 ms, and 2048 time domain data points, 4000-Hz spectral width, and 64 or 128 signals were acquired. After zerofilling to 4096 data points, Gaussian multiplication of 5 Hz, exponential multiplication of 4 Hz (line broadening), Fourier transformation and linear baseline correction, total choline, NAA (referenced at 2.01 ppm), and lactate peaks were identified by their chemical shifts. Because this MRS method does not enable us to calculate absolute metabolic concentrations, metabolic data are expressed as the ratio between peak intensities (NAA/choline ratio). To distinguish lactate resonances from lipid resonances at an echo time of 136 ms, lactate was defined as an inverted resonance at 1.33 ppm, with a signal-to-noise ratio larger than 2 and a clearly identifiable 7-Hz J-coupling. Lactate was documented as either absent or present. In 19 healthy volunteers, (mean age, 58 ± 11 years; 74% men), the mean NAA/choline ratio was 1.93 ± 0.17. Therefore, we considered a ratio of less than 1.59 to be decreased (ie, < mean – 2SD). Statistical analysis. For continuous variables, the t test for independent samples was used for the analysis of differences between two groups. For differences among more than two groups, analysis of variance (ANOVA) was used for the first overall analysis. When the ANOVA procedure showed a significant difference, the least significance difference (LSD) method was used for further analysis of differences. All values are given as means ± SD, unless otherwise indicated. The relation between two continuous variables is expressed as the Pearson correlation coefficient. Data for lactate were treated as ordinal variables, because in many cases no lactate peak could be detected in the MRS spectra. For noncontinuous variables, differences between groups in cross-tabs were assessed with Pearson’s chi-square test or Fisher’s exact test. A P value less than .05 was considered statistically significant. RESULTS It was not possible to perform a CO2 reactivity test in eight patients at the CEA side (12%) and in 10 patients at the contralateral side (15%) because of the inability to insonate the MCAs through the temporal bone. The mean CO2 reactivity on the CEA side was 40% ± 19%, and on the contralateral side, it was 41% ± 22%. For those 56 patients in whom a CO2 reactivity test at one or both sides could be performed, the mean BP increased 6% from 113 mm
JOURNAL OF VASCULAR SURGERY Volume 30, Number 2
Visser et al
255
Fig 1A. Transcranial Doppler registration of the blood flow velocity in both middle cerebral arteries during the CO2 reactivity test. The 51-year-old female subject has more than 70% stenosis of the left internal carotid artery (carotid endarterectomy side) and occlusion of the right internal carotid artery (contralateral side). The CO2 reactivity on the carotid endarterectomy side was 31% and on the contralateral side 12% (ET CO2, end-tidal CO2).
Hg during the baseline period to 120 mm Hg during the carbogene inhalation (P < .001). If cerebral autoregulation is disregarded, such a slight increase in BP could only explain an increase in BFV of 6% during the CO2 reactivity test. However, autoregulation would normally ensure a stable CBF (and BFV in the MCA) with such fluctuations in BP. MRS could be performed in all patients. The results for a 51-year-old woman with occlusion of the contralateral ICA are shown in Fig 1. The CO2 reactivity on the CEA side was 31%, and on the contralateral side, it was 12% (Fig 1A). After CEA, the CO2 reactivity increased to 44% on the CEA side and to 40% on the contralateral side. The rectangular box in the MRI shown in Fig 1B indicates the dimensions of the VOI chosen, and the corresponding 1H MR spectra of both sides are shown in Fig 1C. The NAA/choline ratio was 1.82 on the CEA side and 1.64 on the contralateral side. Lactate was present in both hemispheres. There was a modest but statistically significant correlation (r = .43) between the CO2 reactivity and the NAA/choline ratio in both hemispheres (P < .001; Fig 2). In both hemispheres, the CO2 reactivity was significantly lower in patient groups with a decreased NAA/choline ratio in the corresponding hemisphere (Table II). In groups formed on the basis of the presence of lactate, only in the corresponding contralateral hemisphere was the CO2 reactivity significantly decreased, although a similar trend was found at the CEA side (Table III). In nine patients, a bilateral decreased CO2 reactivity was
Fig 1B. Transversal T2-weighted magnetic resonance image showing the centrum semiovale from the same patient represented in Fig 1A. The white box within the brain parenchyma indicates the volume of interest selected for 1H magnetic resonance spectroscopy.
Fig 1C. 1H magnetic resonance spectra from the volumes of interest selected in the left and right hemispheres from the patient shown in Figs 1A and 1B. The chemical-shift axis in parts per million (ppm) is positioned below each spectrum (Cho, choline; Cr, creatine; NAA, N-acetyl aspartate; Lac, lactate). The NAA/choline ratio on the carotid endarterectomy side was 1.82, and on the contralateral side, it was 1.64. Lactate was present in both hemispheres.
found. In this subgroup, the presence of a decreased NAA/choline ratio or lactate was not significantly different, compared with the patients with an unilateral decreased CO2 reactivity at the CEA side.
JOURNAL OF VASCULAR SURGERY August 1999
256 Visser et al
Table I. The distribution of clinical symptomatology between groups based on the degree of contralateral internal carotid artery stenosis Degree of contralateral ICA stenosis Clinical symptoms
Fig 2. Relation between the CO2 reactivity and the N-acetyl aspartate (NAA)/choline ratio on the carotid endarterectomy side •; solid trend line) and on the contralateral side (o; dashed trend line). Trend lines represent an assumed linear correlation between these variables (with r = 0.43 for both lines).
The CO2 reactivity, NAA/choline ratio, and the presence of lactate on the CEA side were not significantly different for the three groups formed on the basis of the contralateral ICA stenosis, but were on the contralateral side (Table IV). The CO2 reactivity and NAA/choline ratio on the contralateral side were significantly lower in the group with an occluded ICA than in the groups of patients with a more than 70% stenosis or less than 70% stenosis. In addition, lactate was present in 85% of the patients in the occluded ICA group, compared with 29% and 13% of the patients in the groups with more than 70% or less than 70% stenosis, respectively. The CO2 reactivity or the presence of lactate on the CEA side was not significantly different between groups formed on the basis of symptoms on this side (Table V). However, the NAA/choline ratio was lower in the groups of patients who had a minor stroke or transient symptoms than in the group of patients with ocular symptoms or the asymptomatic group. If groups were formed on the basis of symptoms on the contralateral side (Table VI), both the CO2 reactivity and the NAA/choline ratio on this contralateral side were lower in the groups of patients who had a minor stroke or transient cerebral symptoms than in the group of patients with ocular symptoms or the asymptomatic group. In addition, lactate was present in 77% of the symptomatic patients (either ocular symptoms, transient symptoms, or minor stroke), which contrasts with the presence of lactate in only 19% of the patients in the asymptomatic group (P < .001 with Fisher’s exact test).
< 70%
CEA side only 33 Ocular 7 TIA 16 Minor stroke 10 Contralateral side only – TIA – Minor stroke – Bilateral symptoms 2 Bilateral ocular 1 Bilateral TIA 1 Asymptomatic 4
≥ 70%
Occlusion
Total
8 2 3 3 – – – 1 1 – 5
2 – 1 1 8 6 2 2 1 1 1
43 9 20 14 8 6 2 5 3 2 10
ICA, Internal carotid artery; CEA, carotid endarterectomy; TIA, transient ischemic attack.
DISCUSSION The hemodynamic effect of ICA obstruction is dependent on the individual potential of the collateral cerebral circulation.6-8 An ipsilaterally lowered CO2 reactivity can be found in patients with severe stenosis of the ICA, especially when the contralateral ICA is also obstructed.9-14 The CO2 reactivity might be superior to the measured degree of ICA stenosis alone as an indicator of the hemodynamic consequences of ICA stenosis, because CO2 reactivity also depends on the quality of the collateral circulation. As illustration of the clinical significance of the CO2 reactivity, especially in patients with low-flow infarctions, patients with chronic retinal ischemia or hypostatic TIAs have a lowered CO2 reactivity.13,15-20 Also, a low or absent cerebrovascular reactivity, either measured with CO2 or acetazolamide as vasodilator stimulus, is a risk factor for subsequent ischemic cerebral complications.21-24 Finally, after CEA, the CO2 reactivity has been shown to increase on both sides, reflecting the improvement of cerebral hemodynamics.5,25-27 Similar results were found with alternative methods to measure cerebrovascular reserve based on changes in regional cerebral blood flow and/or volume, such as with single-photon emission computed tomography, positron emission tomography, or MRI.28-30 Also, an association was found between a decreased acetazolamide response and the presence of high intensity signals on MRI, especially when located in the centrum semiovale.31 With MRS, a decreased NAA and choline concentration and increased lactate concentration can be found in the
JOURNAL OF VASCULAR SURGERY Volume 30, Number 2
Visser et al
257
Table II. Difference in CO2 reactivity and percentage of patients with decreased CO2 reactivity (less than 30%) between groups formed on the basis of the magnetic resonance spectroscopy N-acetyl aspartate (NAA)/choline ratio NAA/choline ratio on CEA side
Statistics
CEA side
Normal (≥ 1.59)
Decreased (< 1.59)
Results
P value
CO2 reactivity (%)
44 ± 19 (n = 39) 18% (n = 7)
31 ± 17 (n = 19) 53% (n = 10)
t (56) = –2.6 (t test)
< .05
Decreased (< 30%)*
NAA/choline ratio on contralateral side Contralateral side CO2 reactivity (%) Decreased (<30%)*
Statistics
Normal (≥1.59)
Decreased (<1.59)
Results
P value
45 ± 19 (n = 44) 25% (n = 11)
28 ± 26 (n = 12) 50% (n = 6)
t (54) = –2.5 (t test)
<.05
*Based on control data of 30 subjects (see methods section). CEA, Carotid endarterectomy.
Table III. Difference in CO2 reactivity and percentage of patients with decreased CO2 reactivity (less than 30%) between groups formed on the basis of the magnetic resonance spectroscopy lactate presence Lactate on CEA side CEA side CO2 reactivity (%) Decreased (< 30%)*
Absent 43 ± 19 (n = 27) 19% (n = 5)
Statistics
Present 36 ± 19 (n = 31) 39% (n = 12)
Results t (56) = 1.4 (t test)
Lactate on contralateral side Contralateral side CO2 reactivity (%) Decreased (< 30%)*
Absent 48 ± 20 (n = 38) 18% (n = 7)
P value NS
Statistics
Present 27 ± 19 (n = 18) 56% (n = 10)
Results t (54) = 3.8 (t test)
P value < .001
*Based on control data of 30 subjects (see methods section). CEA, Carotid endarterectomy.
area of a cerebral infarct. Similar changes, reflecting disordered metabolism, are also found in noninfarcted cerebral regions with a reduced MCA blood flow.4 A statistically significant correlation was found between the CO2 reactivity and the NAA/choline ratio for both hemispheres. However, the correlation was only moderate. A possible explanation is that a short period of compromised cerebral circulation may result in long-term changes in NAA, choline, and lactate concentrations. These changes can be detected for weeks or months, even after restoration of the cerebral circulation. In contrast, CO2 reactivity can be expected to reflect the temporary situation. So, when formerly compromised hemodynamics have been restored as a result of collateral pathway development, or when the disordered metabolism was the result of a temporary decrease in blood pres-
sure, MRS findings may still be abnormal, whereas the CO2 reactivity may be normal. In the present study, hemodynamics were most compromised in patients with an ICA occlusion on the contralateral side. On this contralateral side, both the mean CO2 reactivity and NAA/choline ratio were decreased to values we consider abnormal, and in 85% of the cases, lactate was found. In the group of patients with more than 70% stenosis and the group of patients with maximally 70% stenosis on the contralateral side, CO2 reactivity and the NAA/choline ratio were normal, and lactate was present in considerably fewer patients. The presence of an ICA occlusion on the contralateral side might be expected to influence the variables measured on the CEA side, because collateral flow would be strongly directed towards the side of the ICA occlusion, thus limiting collateral flow to the
JOURNAL OF VASCULAR SURGERY August 1999
258 Visser et al
Table IV. Differences between groups formed on the basis of the extent of internal carotid artery stenosis contralateral to the carotid endarterectomy side and percentage of patients with abnormal test values ICA stenosis on contralateral side ≥ 70%
< 70% CEA side CO2 reactivity (%)
40 ± 17 (n = 33)
45 ± 27 (n = 12)
Decreased (< 30%)† 27% (n = 9) 25% (n = 3) NAA/choline ratio 1.65 ± 0.23 (n = 39) 1.73 ± 0.24 (n = 14) Decreased (< 1.59)‡ Lactate: Absent (n) Present (n) Present (%) Contralateral side CO2 reactivity (%)
Occlusion
Results
P value
33 ± 17 (n = 13)
F (2, 55) = 1.1 (ANOVA)
NS
F (2, 63) = 1.1 (ANOVA)
NS
NS
39% (n = 5) 1.62 ± 0.15 (n = 13)
41% (n= 16)
29% (n= 4)
15% (n= 2)
19 20 51%
5 9 64%
7 6 46%
X2 = 1.0 (cross-tabs)
50 ± 13 (n= 32)
43 ± 27 (n= 12)
18 ± 18 (n= 12)
F (2, 53) = 14.4 (ANOVA)
< .001*
42% (n= 5) 1.87 ± 0.17 (n= 14)
75% (n= 9) 1.52 ± 0.14 (n= 13)
F (2, 63) = 12.3 (ANOVA)
< .001*
13% (n= 5)
0% (n= 0)
54% (n= 7)
34 5 13%
10 4 29%
2 11 85%
Decreased (< 30%)† 9% (n= 3) NAA/choline ratio 1.81 ± 0.23 (n= 39) Decreased (< 1.59)‡ Lactate: Absent (n) Present (n) Present (%)
Statistics
X2 = 23.8 (cross-tabs)
< .001
*Least significance difference tests: Group with an internal carotid artery occlusion was significantly different from the other two groups. †Based on control data of 30 subjects (see methods section). ‡Based on control data of 19 subjects (see methods section). ICA, Internal carotid artery; CEA, carotid endarterectomy; NAA, N-acetyl aspartate; ANOVA, analysis of variance.
Table V. Differences between groups formed on the basis of clinical symptoms on the carotid endarterectomy side and percentage of patients with abnormal test values Symptoms on CEA side Asymptomatic CEA side CO2 reactivity (%) Decreased (< 30%)† NAA/choline ratio Decreased (< 1.59)‡ Lactate: Absent (n) Present (n) Present (%)
Statistics
Ocular
TIA
Minor stroke
Results
P value
42 ± 25 (n = 18) 28% (n = 5) 1.74 ± 0.16 (n = 18) 11% (n = 2)
49 ± 9 (n = 9) 0% (n = 0) 1.83 ± 0.17 (n = 12) 0% (n = 0)
39 ± 15 (n = 19) 26% (n = 5) 1.61 ± 0.19 (n = 22) 46% (n = 10)
30 ± 19 (n = 12) 58% (n = 7) 1.51 ± 0.24 (n = 14) 71% (n =10)
F (3, 54) = 1.9 (ANOVA)
NS
F (3, 62) = 7.3 (ANOVA)
< .001*
11 7 39%
8 4 33%
9 13 59%
3 11 79%
X2 = 7.3 (cross-tabs)
NS
*Least significance difference tests: Asymptomatic and ocular symptoms groups were significantly different from the groups with transient ischemic attack or a minor stroke. †Based on control data of 30 subjects (see methods section). ‡Based on control data of 19 subjects (see methods section). CEA, Carotid endarterectomy; TIA, transient ischemic attack; NAA, N-acetyl aspartate; ANOVA, analysis of variance.
CEA side. Although in a previous study a statistically significant decreased CO2 reactivity was found on the CEA side in these patients,5 in the present study, only a trend towards a lower CO2 reactivity and
NAA/choline ratio was found for the CEA side. However, not all patients with a contralateral ICA occlusion show abnormal CO2 reactivity and MRS values, and abnormal values are also found in patients
JOURNAL OF VASCULAR SURGERY Volume 30, Number 2
Visser et al
259
Table VI. Differences between groups formed on the basis of clinical symptoms on the contralateral side and percentage of patients with abnormal test values Symptoms on contralateral side Asymptomatic CEA side CO2 reactivity (%)
46 ± 19 (n = 44) Decreased (< 30%)† 23% (n = 10) NAA/choline ratio 1.81 ± 0.22 (n = 53) Decreased (< 1.59)‡ 11% (n = 6) Lactate: Absent (n) 43 Present (n) 10 Present (%) 19%
Statistics
Ocular
TIA
Minor stroke
Results
51 ± 24 (n = 3) 33% (n = 1) 1.84 ± 0.25 (n = 3) 0% (n = 0)
20 ± 21 (n = 7) 57% (n = 4) 1.52 ± 0.17 (n = 8) 50% (n = 4)
8 ± 15 (n = 2) 100% (n = 2) 1.46 ± 0.04 (n =2) 100% (n = 2)
F (3, 52) = 5.8 (ANOVA)
2 1 33%
0 8 100%
1 1 50%
§
F (3, 62) = 5.8 (ANOVA)
P value
< .01*
< .01*
§
*Least significance difference tests: Asymptomatic and ocular symptoms groups were significantly different from the groups with transient ischemic attack or a minor stroke. †Based on control data of 30 subjects (see methods section). ‡Based on control data of 19 subjects (see methods section). §Fisher’s exact test of asymptomatic versus symptomatic (ocular, transient ischemic attack, or minor stroke); P < .001. TIA, Transient ischemic attack; NAA, N-acetyl aspartate; ANOVA, analysis of variance.
with a less severe degree of contralateral stenosis. The variance in these findings may be explained by different collateral contributions from, for example, the basilar system. Therefore, a contralateral ICA occlusion can only be regarded as a risk factor for disordered cerebral hemodynamics and metabolism. Additional tests, such as the CO2 reactivity test, are required to obtain a stratified classification of patients at risk for cerebral ischemic complications. The groups with cerebral symptoms (transient or minor stroke) had a lower mean NAA/choline ratio than the asymptomatic group (including only ocular symptoms). Lactate was more often present in the groups with cerebral symptoms, although differences were only statistically significant on the contralateral side. The CO2 reactivity of the groups with cerebral symptoms was also decreased only on the contralateral side. Thus, in general, changes indicative of disordered hemodynamics were found more often in symptomatic patients than in asymptomatic patients. This suggests that the hemodynamic effect of an ICA stenosis contributes to the transition from asymptomatic to symptomatic stenosis. In conclusion, a decreased CO2 reactivity is indicative of not only inadequate collateral blood supply, but also of a disordered cerebral metabolism, as indicated by the abnormal MRS findings. This is in accordance with the finding that a decreased CO2 reactivity in patients with an ICA stenosis is a risk factor for ischemic cerebral events.6,32,33 Our results show that patients with a contralateral ICA occlu-
sion are especially at risk for disordered cerebral hemodynamics and metabolism. However, not all of these patients with severe bilateral ICA stenosis have disordered hemodynamics, and decisions about the indication for CEA based on clinical grounds alone might be open to refinement. Our results support the use of the CO2 reactivity test to identify patients with disordered hemodynamics at one or both sides, and this test might have the potential to refine the indication for CEA. REFERENCES 1. Kleiser B, Scholl D, Widder B. Doppler CO2 and Diamox test: Decreased reliability by changes of vessel diameter? Cerebrovasc Dis 1995;5:397-402. 2. Bishop CCR, Powell S, Rutt D, Browse NL. Transcranial Doppler measurement of middle cerebral artery blood flow velocity: A validation study. Stroke 1986;17:913-5. 3. Van der Grond J, Balm R, Kappelle LJ, Eikelboom BC, Mali WPThM. Cerebral metabolism of patients with stenosis or occlusion of the internal carotid artery. A 1H-MR spectroscopic imaging study. Stroke 1995;26:822-8. 4. Van der Grond J, Eikelboom BC, Mali WPThM. Flow-related anaerobic metabolic changes in patients with severe stenosis of the internal carotid artery. Stroke 1996;27:2026-32. 5. Visser GH, Van Huffelen AC, Wieneke GH, Eikelboom BC. Bilateral increase in CO2 reactivity after unilateral carotid endarterectomy. Stroke 1997;28:899-905. 6. Powers WJ. Cerebral hemodynamics in ischemic cerebrovascular disease. Ann Neurol 1991;29:231-40. 7. Powers WJ, Press GA, Grubb RL, Gado M, Raichle ME. The effect of hemodynamically significant carotid artery disease on the hemodynamic status of the cerebral circulation. Ann Intern Med 1987;106:27-35.
260 Visser et al
8. Fürst H, Hartl WH, Janssen I, Fink B, Piepgras A, Lauterjung L, et al. Pathophysiology of unilateral high-grade carotid artery stenosis: Evaluation of intracranial haemodynamics by analysis of velocity waveforms from the middle cerebral artery. Clin Sci 1992;83:357-66. 9. Widder B, Kleiser B, Krapf H. Course of cerebrovascular reactivity in patients with carotid artery occlusions. Stroke 1994;25:1963-7. 10. Markus HS, Harrison MJ. Estimation of cerebrovascular reactivity using transcranial Doppler, including the use of breathholding as the vasodilatory stimulus. Stroke 1992;23:668-73. 11. Miller JD, Smith RR, Holaday HR. Carbon dioxide reactivity in the evaluation of cerebral ischemia. Neurosurgery 1992;30:518-21. 12. Maeda H, Matsumoto M, Handa N, Hougaku H, Ogawa S, Itoh T, et al. Reactivity of cerebral blood flow to carbon dioxide in various types of ischemic cerebrovascular disease: Evaluation by the transcranial Doppler method. Stroke 1993;24:670-5. 13. Ringelstein 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;19:963-9. 14. Quilitz B, Rimpel J, Mehdorn M, Rohm N, Lehmann HJ. Cerebrovascular reserve capacity in carotid artery disease: A routine test in selection for surgical treatment? Thorac Cardiovasc Surg 1988;36:217-20. 15. Bullock R, Mendelow AD, Bone I, Patterson J, Macleod WN, Allardice G. Cerebral blood flow and CO2 responsiveness as an indicator of collateral reserve capacity in patients with carotid arterial disease. Br J Surg 1985;72:348-51. 16. Kleiser B, Krapf H, Widder B. Carbon dioxide reactivity and patterns of cerebral infarction in patients with carotid artery occlusion. J Neurol 1991;238:392-4. 17. Ringelstein EB, Weiller C, Weckesser M, Weckesser S. Cerebral vasomotor reactivity is significantly reduced in lowflow as compared to thromboembolic infarctions: The key role of the circle of Willis. J Neurol Sci 1994;121:103-9. 18. Kleiser B, Widder B, Hackspacher J, Schmid P. Comparison of Doppler CO2 test, patterns of infarction in CTT, and clinical symptoms in carotid artery occlusions. Neurosurg Rev 1991;14:267-9. 19. Baumgartner RW, Regard M. Role of impaired CO2 reactivity in the diagnosis of cerebral low flow infarcts. J Neurol Neurosurg Psychiatry 1994;57:814-7. 20. Weiller C, Ringelstein EB, Reiche W, Buell U. Clinical and hemodynamic aspects of low-flow infarcts. Stroke 1991;22: 1117-23.
JOURNAL OF VASCULAR SURGERY August 1999
21. Kleiser B, Widder B. Course of carotid artery occlusions with impaired cerebrovascular reactivity. Stroke 1992;23:171-4. 22. Yonas H, Smith HA, Durham SR, Pentheny SL, Johnson DW. Increased stroke risk predicted by compromised cerebral blood flow reactivity. J Neurosurg 1993;79:483-9. 23. Webster MW, Makaroun MS, Steed DL, Smith HA, Johnson DW, Yonas H. Compromised cerebral blood flow reactivity is a predictor of stroke in patients with symptomatic carotid artery occlusive disease. J Vasc Surg 1995;21:338-45. 24. Gur AY, Bova I, Bornstein NM. Is impaired cerebral vasomotor reactivity a predictive factor of stroke in asymptomatic patients? Stroke 1996;27:2188-90. 25. Hartl WH, Janssen I, Fürst H. Effect of carotid endarterectomy on patterns of cerebrovascular reactivity in patients with unilateral carotid artery stenosis. Stroke 1994;25:1952-7. 26. Silvestrini M, Troisi E, Matteis M, Cupini LM, Caltagirone C. Transcranial Doppler assessment of cerebrovascular reactivity in symptomatic and asymptomatic severe carotid stenosis. Stroke 1996;27:1970-3. 27. Bornstein NM, Gur AY, Shifrin EG, Morag BA. Does carotid endarterectomy modify cerebral vasomotor reactivity? Cerebrovasc Dis 1997;7:201-4. 28. Russell D, Dybevold S, Kjartansson O, Nyberg-Hansen R, Rootwelt K, Wiberg J. Cerebral vasoreactivity and blood flow before and 3 months after carotid endarterectomy. Stroke 1990;21:1029-32. 29. Naylor AR, Merrick MV, Gillespie I, Sandercock PAG, Warlow CP, Cull RE, et al. Prevalence of impaired cerebrovascular reserve in patients with symptomatic carotid artery disease. Br J Surg 1994;81:45-8. 30. Gückel FJ, Brix G, Schmiedek P, Piepgras A, Becker G, Köpke J, et al. Cerebrovascular reserve capacity in patients with occlusive cerebrovascular disease: Assessment with dynamic susceptibility contrast-enhanced MR imaging and the acetazolamide stimulation test. Radiology 1996;201:405-12. 31. Isaka Y, Nagano K, Narita M, Ashida K, Imaizumi M. High signal intensity on T2-weighted magnetic resonance imaging and cerebral hemodynamic reserve in carotid occlusive disease. Stroke 1997;28:354-7. 32. Yonas H, Pindzola R. Physiological determination of cerebrovascular reserves and its use in clinical management. Cerebrovasc Brain Metab Rev 1994;6:325-40. 33. Czosnyka M, Harris NG, Pickard JD, Piechnik S. CO2 cerebrovascular reactivity as a function of perfusion pressure—A modeling study. Acta Neurochir (Wien) 1993;121:159-65.
Submitted Sep 22, 1998; accepted Jan 11, 1999.