Crossed cerebellar diaschisis and brain recovery after stroke.

BACKGROUND AND PURPOSE Although crossed cerebellar diaschisis is well recognized after stroke, there is controversy concerning its clinical correlations and serial changes, and little is known about its prognostic value. METHODS We studied crossed cerebellar diaschisis and cerebral hypoperfusion in 47 patients with acute middle cerebral cortical infarction using 99mTc-hexamethylpropyleneamine oxime and single-photon emission computed tomography within 72 hours of stroke onset. Thirty-one of these patients had outcome studies at 3 months; 15 of the 31 underwent an additional scan after acetazolamide injection. Tissue loss was determined with computed tomography, performed at outcome in 28 patients. Clinical stroke severity was assessed with the Canadian Neurological Scale and Barthel Index. Cerebellar blood flow asymmetry was studied in 22 healthy, age-matched control subjects. RESULTS Cerebellar blood flow asymmetry was significant in patients (mean +/- SE, 9.76 +/- 0.78%; P < .001) but not in control subjects (-0.22 +/- 0.56%). Crossed cerebellar diaschisis was strongly associated with infarct hypoperfusion volume at both acute (regression coefficient +/- SEb, b = 6.76 +/- 0.65; P < .001) and outcome stages (b = 6.13 +/- 0.63; P < .001). Cross-sectionally over the first 72 hours, infarct hypoperfusion volume decreased by 2% for each hour from onset (P < .05), while crossed cerebellar diaschisis remained unchanged. Canadian Neurological Scale score at the acute stage was negatively associated with acute crossed cerebellar diaschisis (b = -0.10 +/- 0.05; P < .05) after allowing for infarct hypoperfusion volume. Crossed cerebellar diaschisis did not change between acute-stage, outcome, and postacetazolamide scans. Acute-stage crossed cerebellar diaschisis predicted outcome Barthel Index score (b = -0.28 +/- 0.14; P = .05) and tissue loss (b = 3.81 +/- 0.96; P < .001) but was no longer an independent prognostic factor after allowing for acute-stage infarct hypoperfusion volume. CONCLUSIONS This study shows that crossed cerebellar diaschisis is a functional phenomenon that correlates with both stroke severity and infarct hypoperfusion volume and persists despite neurological recovery. Although acute-stage crossed cerebellar diaschisis has no prognostic value independent of acute-stage hypoperfusion volume, it might indicate the proportion of nutritional to nonnutritional perfusion at the infarct site and hence be useful in the evaluation of reperfusion therapies in the acute stage.