Reduced Vasomotor Reactivity in Cerebral Microangiopathy
A Study With Near-Infrared Spectroscopy and Transcranial Doppler Sonography
Background and Purpose—Reduction of cerebral blood flow and vasomotor reactivity (VMR) are thought to play an important role in the pathogenesis of cerebral microangiopathy. The aim of our study was to determine whether near-infrared spectroscopy (NIRS) can detect a reduced VMR in patients with microangiopathy, whether NIRS reactivities correlate with VMR assessed by transcranial Doppler sonography (TCD), and whether the differing extents of patients’ microangiopathy demonstrated on MRI or CT can be distinguished by both noninvasive techniques.
Methods—We compared the VMR of 46 patients with cerebral microangiopathy with 13 age-matched control subjects. Patients were classified with the Erkinjuntti scale. We monitored cerebral blood flow velocity (CBFV) in both middle cerebral arteries by TCD, changes in concentration of oxyhemoglobin (Hbo2), deoxyhemoglobin (Hb) and blood volume (HbT) by NIRS, mean arterial blood pressure, and end-tidal CO2 (Etco2) during normocapnia and hypercapnia. VMRs were calculated as percent change of CBFV (NCR) and as absolute change in concentration of HbO2, Hb, and HbT per 1% increase in Etco2 (CR-HbO2, CR-Hb, CR-HbT).
Results—NCR and NIRS reactivities were significantly reduced in patients with cerebral microangiopathy. CR-HbO2 and CR-Hb showed a close correlation with NCR, and NCR and NIRS reactivities were related to the severity of cerebral microangiopathy according to the Erkinjuntti scale. Validity of NCR and NIRS reactivities were similar.
Conclusions—VMR is reduced in patients with cerebral microangiopathy and can be noninvasively assessed in basal arteries (with TCD) and brain parenchyma (with NIRS). Reduction of CO2-induced VMR, as measured by NIRS and TCD, may indicate the severity of microangiopathy.
- cerebral blood flow
- spectroscopy, near-infrared
- ultrasonography, Doppler, transcranial
- vasomotor reactivity
Cerebral microangiopathy is associated with age, arterial hypertension, and other vascular risk factors, and coincides with lacunar infarction and demyelination of the deep white matter.1 2 3 4 5 6 Cerebral microangiopathy is characterized anatomically by lipohyalinosis of cerebral arteries and arterioles and hemodynamically by decreased cerebral blood flow and loss of autoregulation.7 8 9 10 Impaired VMR is thought to cause further ischemic infarctions and progression of Binswanger’s disease.5 11
NIRS measures changes in concentration of the chromophores Hbo2 and Hb, and enables the assessment of cerebral hemodynamics noninvasively (see References 12 14 ). NIRS has been applied to test CO2-induced VMR, and correlations with cerebral blood flow and cerebral blood volume measurements were found in preterm neonates15 and newborn infants.16 In adults with various neurological diseases, CO2 reactivity measured by NIRS was similar to changes in jugular venous oxygen saturation.17 Simultaneous assessment of CO2-induced VMR by TCD and NIRS revealed correlating results in adult volunteers and patients with different degrees of carotid artery stenosis.18 19
TCD measurements of VMR reflect autoregulatory changes of the whole downstream microcirculatory bed. In contrast to TCD, NIRS semiquantitatively measures changes of the regional cerebral blood volume (rCBV) in cerebral tissue and therefore assesses VMR from a different location.
The aim of our study was therefore to evaluate whether cerebral microangiopathy is associated with a reduced CO2-induced VMR, whether VMR measured by NIRS correlates with NCR, and whether there is a correlation between impairment of autoregulation and the pattern and extent of white matter changes on CT or MRI.
Subjects and Methods
We compared CO2-induced VMR in 46 patients with cerebral microangiopathy who were treated in our Department of Neurology (38 male, 8 female; mean age 68.8 years, SD 9.6 years) with a group of 13 age-matched volunteers from our staff or our department without previous or present diseases of the central nervous system (9 male, 4 female; mean age 66.8 years, SD 7.6 years). Patients fulfilled the following criteria: ≥1 ischemic events consistent with transient ischemic attack or lacunar infarction (n=33), dementia (n=13), or both (n=3), and white matter changes on CT or MRI in concordance with cerebral microangiopathy. Patients or volunteers with insufficient temporal bone window, significant carotid artery stenosis, or territorial brain infarction were excluded. All gave informed consent prior to the reactivity test.
Patients were classified in a blinded manner according to their MRI scans with the Erkinjuntti scale20 by one of the authors as follows. For periventricular hyperintensities: 0 indicates lesions absent; 1, caps; 2, pencil-thin lining; 3, smooth halo; and 4, irregular hyperintensities extending to the deep white matter. For other white matter hyperintensities: 0 indicates absent; 1, <5 small focal and/or <2 large focal lesions; 2, 5 to 12 small and/or 2 to 4 large focal lesions; 3, >12 small focal and/or >4 large focal or some confluent lesions; and 4, predominantly confluent lesions.
The above groups and all patients were divided into mild and severe white matter changes. In 14 patients with a CT scan only, the Erkinjunti scale was applied in an analogous manner.
We recorded simultaneously the CBFV in both middle cerebral arteries by transcranial Doppler sonography (X4, DWL), changes in concentration of Hbo2, Hb, and HbT by near-infrared spectroscopy (NIRO 500, Hamamatsu Photonics), and end-tidal CO2 (Kapnograph, Datex-Engström) during normocapnia and hypercapnia. Mean arterial blood pressure was measured continuously by a noninvasive, beat-to-beat finger blood-pressure monitor (Portapres, TNO Biomedical Instrumentation). The transmitting probe of the NIRS was placed on the left side of the forehead 2 cm beside the midline and 3 to 4 cm above the supraorbital ridge, and the receiving probe (photomultiplier) was fixed laterally at a distance of 5 cm with an elastic bandage.
For baseline conditions, subjects lay in a supine position, breathing room air through a breathing mask. After registration of a stable signal for at least 5 minutes, a CO2 test was performed by their breathing carbogene gas (5% CO2 and 95% O2), until a 1% increase in Etco2 was achieved. Figure 1⇓ shows a typical recording.
Data Collection and Processing
For offline analysis, data were collected and digitized with use of an AD converter with a sampling rate of 50 Hz, averaged, and stored on a laptop personal computer by means of a data collecting software (Dasylab, Synotech). CO2-induced VMRs were calculated as percentage change of the CBFV in TCD (NCR) and as absolute change in concentration (μmol/L) of the chromophores HbO2, Hb, and HbT per 1% increase in end-tidal CO2 (CR-HbO2, CR-Hb, CR-HBT). To increase signal-to-noise ratio, we calculated Hbdiff as the difference between HbO2 and Hb, and CR-Hbdiff as an additional reactivity index.
analogous calculations for CR-Hb, CR-HbT, and CR-Hbdiff where CR-Hbo2 is the CO2 reactivity of Hbo2, NCR is normalized CO2 reactivity, and CBFVmax and CBFVnormo are cerebral blood flow velocity during hypercapnia and normocapnia, respectively.
For statistical analysis we used the nonparametric Mann-Whitney U test, because most values did not show normality. Differences in VMR between control subjects and patients with mild and severe white matter hyperintensities according to the Erkinjuntti scale were tested by a Kruskal-Wallis 1-way ANOVA. Post hoc tests (2×2) were performed by the use of the Scheffé test.
The correlation coefficients between reactivity indices and between VMRs and the extent of microangiopathy according to the Erkinjuntti scale were tested with the nonparametric Pearson and Spearman formulas. The validity of all reactivity indices was assessed by sensitivity, specificity, and receiver operating characteristic (ROC) analysis.
The classification of the patients according to the Erkinjuntti scale is summarized in Table 1⇓. Patients with cerebral microangiopathy had a significantly reduced CO2-induced VMR demonstrated on TCD and NIRS: NCR, CR-HbO2, CR-Hb, CR-HbT, and CR-Hbdiff were diminished compared with values from the age-matched control group (Table 2⇓). NIRS reactivities showed a higher level of significance compared with NCR, in particular CR-HbO2 and CR-Hbdiff (NCR left, P=0.0215; CR-HbO2, P=0.0001; CR-Hbdiff, P=0.0006).
Blood pressure usually rose during CO2 tests (mean 6.8 mm Hg, SD 7.1 mm Hg). Changes in blood pressure during reactivity tests did not correlate with NCR but did correlate with CR-HbO2, CR-Hb, and CR-Hbdiff (P<0.05).
The reduction of CO2-induced VMR was highest in patients with white matter changes qualified as “severe” (Erkinjuntti scale 5 to 8) and less in patients with “mild” abnormalities (Erkinjuntti scale 1 to 4) compared with the controls. The global difference between the groups was significant (NCR, P<0.05; CR-HbO2, P=0.0001; CR-Hb, P<0.003; CR-Hbdiff, P<0.002; Figures 3 through 7⇓⇓⇓⇓⇓). A 2×2 post hoc test (Scheffé) showed a difference between controls and patients with severe microangiopathy for NCR right (P<0.05) and a trend toward a reduced NCR left (P=0.055). Concerning the NIRS reactivities, differences between controls and patients with mild microangiopathy and between controls and patients with severe microangiopathy were significant for the parameters CR-HbO2 (P=0.001 and P=0.000, respectively), CR-Hb (P=0.017 and P=0.001, respectively), and CR-Hbdiff (P=0.003 and P=0.000, respectively). The homogeneity of variances of these parameters could be verified. We did not find significant differences between patients with mild and those with severe microangiopathy (Scheffé test) in either TCD or NIRS.
Furthermore, we found a negative correlation between VMR and the extent of microangiopathy according to the Erkinjuntti scale (NCR, P=0.024; CR-HbO2, P=0.000; CR-Hb, P=0.000; CR-Hbdiff, P=0.000; and CR-HbT, P=0.008).
Sensitivity and specificity of NCR left were 72.7% and 64.4% (P=0.021); NCR right, 69.2% and 60.0% (P=0.009); CR-HbO2, 92.3% and 65.2% (P=0.000); and CR-Hbdiff, 76.9% and 63% (P=0.001; Table 4⇓). ROC analysis of all reactivity indices showed the largest areas under the curve for the indices CR-HbO2, CR-Hb, and CR-Hbdiff (Table 4⇓, Figures 8 through 10⇓⇓⇓).
Several studies indicate hemodynamic alterations in cerebral microangiopathy.6 8 9 21 Impaired cerebral autoregulation and altered pulsatility characteristics measured by TCD were found in patients with vascular dementia and lacunes22 23 24 25 ; however, most of these studies did not distinguish between cerebral microangiopathy and lacunar and territorial brain infarction.
In our patients with cerebral microangiopathy (defined as focal neurological signs and/or dementia in the presence of white matter abnormalities on CT or MRI), we found a significantly reduced CO2 reactivity in both noninvasive techniques, TCD and NIRS. Compared with the control subjects, VMR was more reduced in patients with severe than with mild microangiopathy (Figures 3 through 7⇑⇑⇑⇑⇑), and differences (global comparison) between the 3 groups were significant. The reduction of NCR and NIRS reactivities correlated with the extent of periventricular and deep white matter hyperintensities on the Erkinjuntti scale.
Our results are in line with recent hemodynamic studies: reduction of rCBF and rCBF change to acetazolamide measured by xenon-enhanced CT was described in the cerebral white matter and cortex of patients with single and multiple lacunes. Patients with multiple lacunar infarctions showed a significantly lower VMR in the cerebral cortex than did patients with single lacunes.7 In patients with leukoaraiosis and lacunar infarctions, Oishi et al21 found a more reduced rCBF and acetazolamide reactivity than in patients with leukoaraiosis alone. A negative correlation between vasodilatory capacity to acetazolamide in the cortex and the severity of periventricular hyperintensities on MRI of the brain was also shown in asymptomatic subjects with periventricular hyperintensities (xenon clearance).26 In patients with vascular dementia of the Binswanger type, a decreased CO2 reactivity was demonstrated in cerebral cortex and in the deep white matter (by PET),27 and in lacunar dementia De Reuck et al28 found a decreased blood flow to oxygen metabolism with increased oxygen extraction rate. These studies indicate that cerebral microangiopathy goes along with a reduction of rCBF and, as noninvasively shown in our study, an impaired cerebral autoregulation.
Findings of impaired VMR fit into the concept that the disturbed autoregulation in cerebral microangiopathy results in temporary critical hypoperfusion during episodes of hypotension.1 5 28 29 Recurrent decreases in cerebral perfusion (eg, during blood pressure dysregulation or cardiac arrhythmia) may lead to ischemia in the deep white matter.5 11 30
Functional studies31 with NIRS and PET have shown that the sample volume of NIRS is located in the superficial part of the brain. We have hypothesized that measurements with NIRS might therefore be more sensitive and reliable than those with TCD, because NIRS measures the hemodynamic response to hypercapnia closer to the presumed site of the disturbed autoregulation. In our study, sensitivity and specificity of the reactivity indices CR-HbO2 and CR-Hbdiff were higher compared with NCR, and ROC analysis revealed the largest areas under the curve for CR-HbO2, CR-Hb, and CR-Hbdiff (Table 4⇑, Figures 8 through 10⇑⇑⇑). Validity of VMR measured by NIRS might therefore be similar (or even better) compared with TCD in patients with cerebral microangiopathy.
Interestingly, Sabri et al6 32 found that in microangiopathy dementia and neuropsychological impairment correlate with the reduction of regional cerebral blood flow and glucose utilization and not with MRI changes. Whether a normal neurological and neuropsychological status correlates with a preserved autoregulation in patients with cerebral microangiopathy has yet to be determined.
This study was supported by a grant from the Thuringia Research Council to Dr Röther (TMWFK B 301-96052). We gratefully acknowledge statistical support by Dr B. Schack (Institute of Medical Statistics, Computer Science and Documentation, University of Jena, Germany).
- Received August 9, 1999.
- Revision received January 27, 2000.
- Accepted January 27, 2000.
- Copyright © 2000 by American Heart Association
Pellissier J-F, Poncet M. Binswanger’s encephalopathy. In: Toole JF, ed. Handbook of Clinical Neurology. Amsterdam, Netherlands: Elsevier Science Publishers BV; 1989;10:221–233.
Roman GC. From UBOs to Binswanger’s disease: impact of magnetic resonance imaging on vascular dementia research. Stroke.. 1996;27:1269–1273.
Pantoni L, Garcia JH. Pathogenesis of leukoaraiosis: a review. Stroke.. 1997;28:652–659.
Sabri O, Ringelstein EB, Hellwig D, Schneider R, Schreckenberger M, Kaiser HJ, Mull M, Buell U. Neuropsychological impairment correlates with hypoperfusion and hypometabolism but not with severity of white matter lesions on MRI in patients with cerebral microangiopathy. Stroke.. 1999;30:556–566.
Mochizuki Y, Oishi M, Takasu T. Cerebral blood flow in single and multiple lacunar infarctions. Stroke.. 1997;28:1458–1460.
De Reuck J, Van Aken J, Decoo D, Strijckmans K, Lehmahieu I. Cerebral blood flow and oxygen metabolism in leuko-araiosis. Cerebrovasc Dis.. 1991;1:25–30.
Yao H, Sadoshima S, Kuwabara Y, Ichiya Y, Fujishima M. Cerebral blood flow and oxygen metabolism in patients with vascular dementia of the Binswanger type. Stroke.. 1990;21:1694–1699.
Erkinjuntti T, Hachinski VC. Rethinking vascular dementia. Cerebrovasc Dis.. 1993;3:3–23.
Jobsis FF. Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science.. 1977;198:1264–1267.
Tateichi A. Maekawa T, Soejima Y, Sadamitsu D, Yamamoto M, Matsushita M, Nakashima K. Qualitative comparison of carbon dioxide-induced change in cerebral near-infrared spectroscopy versus jugular venous oxygen saturation in adults with acute brain disease. Crit Care Med.. 1995;23:1734–1738.
Smielewski P, Kirkpatrick P, Minhas P, Pickard JD, Czosnyka M. Can cerebrovascular reactivity be measured with near-infrared spectroscopy? Stroke. 1995;26:2285–2292.
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.
Mantyla R, Erkinjuntti T, Salonen O, Aronen HJ, Peltonen T, Pohjasvaara T, Standertskjold-Nordenstam CG. Variable agreement between visual rating scales for white matter hyperintensities on MRI: comparison of 13 rating scales in a poststroke cohort. Stroke.. 1997;28:1614–1623.
Maeda H, Matsumoto M, Handa N, Hougaku H, Ogawa S, Itoh T, Tsukamoto Y, Kamada T. Reactivity of cerebral blood flow to carbon dioxide in various types of ischemic cerebrovascular disease: evaluation by the transcranial Doppler method. Stroke.. 1993;24:670–675.
Ries F, Horn R, Hillekamp J, Honisch C, Konig M, Solymosi L. Differentiation of multi-infarct and Alzheimer dementia by intracranial hemodynamic parameters. Stroke.. 1993;24:228–235.
Marcos A, Egido JA, Barquero M, Fernandez C, de Seijas EV. Full range of vasodilatation tested by transcranial Doppler in the differential diagnosis of vascular and Alzheimer types of dementia. Cerebrovasc Dis.. 1997;7:14–18.
Isaka Y, Okamoto M, Ashida K, Imaizumi M. Decreased cerebrovascular dilatory capacity in subjects with asymptomatic periventricular hyperintensities. Stroke.. 1994;25:375–381.
Hagendorff A, Dettmers C, Block A, Pizzulli L, Omran H, Hartmann A, Manz M, Luderitz B. Reduction of cerebral blood flow with induced tachycardia in rats and in patients with coronary artery disease and premature ventricular contractions. Eur Heart J.. 1994;15:1477–1481.
Sabri O, Hellwig D, Schreckenberger M, Cremerius U, Schneider R, Kaiser HJ, Doherty C, Mull M, Ringelstein EB, Buell U. Correlation of neuropsychological, morphological and functional (regional cerebral blood flow and glucose utilization) findings in cerebral microangiopathy. J Nucl Med.. 1998;39:147–154.