Individual variability of cerebral autoregulation, posterior cerebral circulation and white matter hyperintensity

Abstract

Key points: Cerebral autoregulation (CA) is a key mechanism to protect brain perfusion in the face of changes in arterial blood pressure, but little is known about individual variability of CA and its relationship to the presence of brain white matter hyperintensity (WMH) in older adults, a type of white matter lesion related to cerebral small vessel disease (SVD). This study demonstrated the presence of large individual variability of CA in healthy older adults during vasoactive drug-induced changes in arterial pressure assessed at the internal carotid and vertebral arteries. We also observed, unexpectedly, that it was the 'over-' rather than the 'less-reactive' CA measured at the vertebral artery that was associated with WMH severity. These findings challenge the traditional concept of CA and suggest that the presence of cerebral SVD, manifested as WMH, is associated with posterior brain hypoperfusion during acute increase in arterial pressure.

Abstract: This study measured the individual variability of static cerebral autoregulation (CA) and determined its associations with brain white matter hyperintensity (WMH) in older adults. Twenty-seven healthy older adults (13 females, 66 ± 6 years) underwent assessment of CA during steady-state changes in mean arterial pressure (MAP) induced by intravenous infusion of sodium nitroprusside (SNP) and phenylephrine. Cerebral blood flow (CBF) was measured using colour-coded duplex ultrasonography at the internal carotid (ICA) and vertebral arteries (VA). CA was quantified by a linear regression slope (CA slope) between percentage changes in cerebrovascular resistance (CVR = MAP/CBF) and MAP relative to baseline values. Periventricular and deep WMH volumes were measured with T2-weighted magnetic resonance imaging. MAP was reduced by -11 ± 7% during SNP, and increased by 21 ± 8% during phenylephrine infusion. CA demonstrated large individual variability with the CA slopes ranging from 0.37 to 2.20 at the ICA and from 0.17 to 3.18 at the VA; no differences in CA were found between the ICA and VA. CA slopes measured at the VA had positive correlations with the total and periventricular WMH volume (r = 0.55 and 0.59, P < 0.01). Collectively, these findings demonstrated the presence of large individual variability of CA in older adults, and that, when measured in the posterior cerebral circulation, it is the higher rather than lower CA reactivity that is associated with WMH severity.

Figure 1. Three typical patterns of (normal, inferior, over‐reactive) CA measured at ICA (upper row) and VA (bottom row)

CA, cerebral autoregulation; ICA, internal carotid artery; VA, vertebral artery; CBF, cerebral blood flow; MAP, mean arterial pressure; CVR, cerebrovascular resistance (= MAP/CBF). Open circles represent the percentage change of CBF from baseline (ΔCBF%), and solid circles represent CVR (ΔCVR%); the dashed lines show the trend of changes in ΔCBF%, and the continuous lines show changes in ΔCVR%. The CA slope between ΔCVR% and ΔMAP% is calculated as the regression coefficient from each of the estimated equations.

Figure 2. Individual variability of regional CA slopes and its association with PWMH

A, scatter plots of CA slopes at ICA and VA. The individual variability of CA slopes at both the ICA and the VA has a normal distribution by the Kolmogorov–Smirnov test with no differences between mean values. B, box plots of CA slopes in the ICA and VA between the tertiles of PWMH. The CA slope measured at the VA in the group with high PWMH (> 0.3 %ICV, n = 8) was much greater than that at the ICA and those at the VA or ICA in the groups with low (< 0.15 %ICV, n = 8) and moderate (0.15–0.30 %ICV, n = 11) PWMH. The horizontal dotted and continuous lines within the box represent the mean and median, respectively. CA, cerebral autoregulation; ICA, internal carotid artery; VA, vertebral artery; PWMH, periventricular white matter hyperintensity; %ICV, percentage of intracranial volume.

Figure 3. An increase in arterial pressure induced a paradoxical hypoperfusion, indicating over‐reactivity of CA, measured at the VA in a 78‐year‐old male subject with prominent PWMH

A, representative FLAIR image showing salient PWMH (0.565 %ICV). B, the diameter of VA was measured in a longitudinal view on high‐resolution B‐mode video using automated software. C, the pulsed Doppler sampling volume was placed on the site of diameter measurement in colour mode. D, Doppler measurement of blood flow velocity at the VA when MAP was 92 mmHg at baseline. E, blood flow velocity at the VA was reduced when MAP elevated to 109 mmHg by intravenous infusion of phenylephrine with an estimated CA slope of 3.18. CA, cerebral autoregulation; VA, vertebral artery; PWMH, periventricular white matter hyperintensity; FLAIR, fluid‐attenuated inversion recovery; %ICV, percentage of intracranial volume; PSV, peak systolic velocity; EDV, end‐diastolic velocity; TAPV, time‐averaged peak velocity; TAMV, time‐averaged mean velocity; MAP, mean arterial pressure; CBF, volumetric cerebral blood flow; CVR, cerebrovascular resistance (= MAP/CBF).