Return to search

Magnetic Resonance Mapping of Cerebrovascular Reserve: Steal Phenomena in Normal and Abnormal Brain

Blood oxygen level-dependent (BOLD) magnetic resonance (MR) imaging enables non-invasive spatial mapping of changes in cerebral blood flow (CBF). By applying a vasodilatory stimulus (such as inhaled CO2) during BOLD MR imaging, one can measure cerebral vasodilatory capacity. "Cerebrovascular reactivity" (CVR) is defined as the change in CBF per unit of vasodilatory stimulus. Vasodilatory capacity is clinically important as vasodilatation is a mechanism by which the brain maintains constant CBF despite reductions in cerebral perfusion pressure.ii
Patients with arterial narrowing commonly demonstrate a paradoxical response: vasodilatory stimulus-induced reduction of BOLD MR signal. BOLD MR depends on CBF but on other factors too. Does a reduction of BOLD MR signal indicate a decrease in flow? Does BOLD MR CVR correlate with CVR measured using arterial spin labeling (ASL) MR? I studied thirty-eight patients with stenosis of brain-supplying arteries and found that the BOLD CVR and ASL CVR results correlate strongly (R=0.83, P<0.0001 for cerebral hemispheric gray matter). The second study aimed to determine whether preoperative CVR predicts the hemodynamic effect of extracranial-intracranial bypass surgery. Whereas prior studies relied on right-left interhemispheric CVR asymmetry indices, this study used “absolute” CVR from each hemisphere. I studied twenty-five patients with intracranial arterial stenosis. I found that the group with normal pre-operative CVR showed no change in CVR following bypass surgery (0.22% ± 0.05% to 0.22% ± 0.01% (mean ± SD)(P=0.881)), the group with reduced pre-operative CVR demonstrated an improvement (0.08% ± 0.05% to 0.21 ± 0.08% (mean ± SD)(P<0.001)), and the group with paradoxical pre-operative CVR demonstrated the greatest improvement (-0.04% ± 0.03% to 0.27% ± 0.03% (P=0.028)). ii
Patients with arterial narrowing commonly demonstrate a paradoxical response: vasodilatory stimulus-induced reduction of BOLD MR signal. BOLD MR depends on CBF but on other factors too. Does a reduction of BOLD MR signal indicate a decrease in flow? Does BOLD MR CVR correlate with CVR measured using arterial spin labeling (ASL) MR? I studied thirty-eight patients with stenosis of brain-supplying arteries and found that the BOLD CVR and ASL CVR results correlate strongly (R=0.83, P<0.0001 for cerebral hemispheric gray matter).
The second study aimed to determine whether preoperative CVR predicts the hemodynamic effect of extracranial-intracranial bypass surgery. Whereas prior studies relied on right-left interhemispheric CVR asymmetry indices, this study used “absolute” CVR from each hemisphere. I studied twenty-five patients with intracranial arterial stenosis. I found that the group with normal pre-operative CVR showed no change in CVR following bypass surgery (0.22% ± 0.05% to 0.22% ± 0.01% (mean ± SD)(P=0.881)), the group with reduced pre-operative CVR demonstrated an improvement (0.08% ± 0.05% to 0.21 ± 0.08% (mean ± SD)(P<0.001)), and the group with paradoxical pre-operative CVR demonstrated the greatest improvement (-0.04% ± 0.03% to 0.27% ± 0.03% (P=0.028)).
The third study arose from an unexpected observation: paradoxical reactivity in the white matter of young healthy subjects. I evaluated healthy subjects using BOLD CVR and ASL CVR, transformed all CVR maps into a common brain space, and generated composite maps of CVR. Composite maps confirmed regions of significant paradoxical
iii
reactivity in the white matter. These regions may represent the physiological correlate of previously anatomically defined border-zones (watershed zones). The regions match the locations where elderly patients develop white matter rarefaction, so-called leukoaraiosis.

Identiferoai:union.ndltd.org:TORONTO/oai:tspace.library.utoronto.ca:1807/43658
Date13 January 2014
CreatorsMandell, Daniel M.
ContributorsMikulis, David J.
Source SetsUniversity of Toronto
Languageen_ca
Detected LanguageEnglish
TypeThesis

Page generated in 0.0023 seconds