Diaschisis Profiles in the Cerebellar Response to Hemodynamic Stimuli: Insights From Dynamic Measurement of Cerebrovascular Reactivity to Identify Occult and Transient Maxima

Background: Crossed cerebellar diaschisis (CCD) refers to depressions in perfusion and metabolism within the cerebellar hemisphere contralateral to supratentorial disease. Prior investigation into CCD in cerebrovascular reactivity (CVR) has been limited to terminal CVR estimations (CVR_end ). We recently have demonstrated the presence of unsustained CVR maxima (CVR_max ) using dynamic CVR analysis, offering a fully dynamic characterization of CVR to hemodynamic stimuli.

Purpose: To investigate CCD in CVR_max from dynamic blood oxygen level-dependent (BOLD) MRI, by comparison with conventional CVR_end estimation.

Study type: Retrospective.

Population: A total of 23 patients (median age: 51 years, 10 females) with unilateral chronic steno-occlusive cerebrovascular disease, without prior knowledge of CCD status.

Field strength/sequence: A 3-T, T1-weighted magnetization-prepared rapid gradient-echo (MPRAGE) and acetazolamide-augmented BOLD imaging performed with a gradient-echo echo-planar imaging (EPI) sequence.

Assessment: A custom denoising pipeline was used to generate BOLD-CVR time signals. CVR_end was established using the last minute of the BOLD response relative to the first-minute baseline. Following classification of healthy versus diseased cerebral hemispheres, CVR_max and CVR_end were calculated for bilateral cerebral and cerebellar hemispheres. Three independent observers evaluated all data for the presence of CCD.

Statistical tests: Pearson correlations for comparing CVR across hemispheres, two-proportion Z-tests for comparing CCD prevalence, and Wilcoxon signed-rank tests for comparing median CVR. The level of statistical significance was set at P ≤ 0.05.

Results: CCD-related changes were observed on both CVR_end and CVR_max maps, with all CCD+ cases identifiable by inspection of either map. Diseased cerebral and contralateral cerebellar hemispheric CVR correlations in CCD+ patients were stronger when using CVR_end (r = 0.728) as compared to CVR_max (r = 0.676). CVR correlations between healthy cerebral hemispheres and contralateral cerebellar hemispheres were stronger for CVR_max (r = 0.739) than for CVR_end (r = 0.705).

Data conclusion: CCD-related alterations could be observed in CVR examinations. Conventional CVR_end may underestimate CVR and could exaggerate CCD.

Evidence level: 4.

Technical efficacy: Stage 3.