Crossed Cerebellar Diaschisis: Diagnostic and Prognostic Value of MRI Derived Imaging—A BOLD-Cerebrovascular Reactivity, MRI Perfusion, and H²O-PET Comparison Study

Introduction: Crossed cerebellar diaschisis (CCD) refers to a reduction in cerebellar blood flow and metabolism contralateral to a supratentorial lesion and corresponds to poorer clinical outcome. It is thought to be caused by an interruption of the cortico-ponto-cerebellar or the cerebello-thalamico-cortical fiber tracts. Positron emission tomography (PET) imaging is typically used to detect CCD, but advancing noninvasive magnetic resonance imaging (MRI) techniques such as dynamic susceptibility contract (DSC), MRI-derived parameters, or blood oxygenation-level dependent (BOLD)-derived cerebrovascular reactivity (CVR) may prove be more optimal tools to characterize this disease. Hence, we aimed to study the novel diagnostic and prognostic potential of DSC-MRI and BOLD-CVR to detect CCD.

Methods: Nineteen subjects with symptomatic unilateral cerebrovascular steno-occlusive disease underwent a BOLD-CVR as well as a 15(O)-H²O PET study. Mean transit time, time to peak, relative cerebral blood flow, and cerebral blood volume maps were derived from DSC-MRI. CCD as well as a cerebellar asymmetry index (CAI) was determined from PET. CAI was then also calculated for BOLD-CVR and all the DSC-MRI derived maps separately and compared with PET CAI. Clinical status at admission and outcome after 3 months were determined with National Institute of Health Stroke Score (NIHSS) and modified Rankin Scale (mRS) scores.

Results: CAI in CCD(+) subjects was significantly different for both BOLD-CVR and PET (CCD(+) vs CCD(-) for BOLD-CVR: 14.76 ± 10.71 vs 1.40 ± 5.43, p < 0.001 and for PET: 7.04 ± 1.75 vs 1.79 ± 3.14, p < 0.001). No difference in CAI was seen for any of the DSC-MRI-derived maps. The area under the curve (AUC) for BOLD-CVR was 0.88 (confidence interval 0.73–1.0). None of the DCS derived maps could accurately detect CCD with cerebral blood volume showing the highest AUC of 0.6. Using a cutoff value of 6.2 CAI for BOLD-CVR, CCD could be detected with a sensitivity of 0.87 and a specificity of 0.77. CCD(+) patients were in poorer clinical condition at baseline (CCD(+) vs CCD(-): NIHSS: 6 vs 1, p = 0.004; mRS: 3 vs 1, p < 0.01) and after 3 months follow-up (NIHSS: 2 vs 0, p = 0.03; mRS: 1 vs 0, p = 0.04).

Conclusions: In comparison to PET imaging, BOLD-CVR can accurately detect CCD, whereas none of the DCS-MRI parameters proved to be able to characterize CCD. Furthermore, CCD(+) subjects had a significantly poorer initial clinical status and outcome after 3 months with both mRS and NIHSS scores.