Can the detection of misery perfusion in chronic cerebrovascular disease be based on reductions in baseline CBF and vasoreactivity?

The aim of this study was to clarify whether decreases in baseline regional cerebral blood flow (rCBF) and in residual cerebral vasoreactivity (CVR), assessed by the acetazolamide (ACZ) challenge, can detect misery perfusion in patients with chronic cerebrovascular disease (CVD). Oxygen extraction fraction (OEF) and other haemodynamic parameters were measured in 115 patients (64±9 years old) with unilateral cerebrovascular steno-occlusive disease (>70% stenosis) using 15O-gas and water PET. A significant elevation of OEF, by greater than the mean+2SD compared with healthy controls, was defined as misery perfusion. CBF, CVR determined by percent change in CBF after ACZ administration, OEF and other haemodynamic parameters in the territories of the bilateral middle cerebral arteries were analysed. Diagnostic accuracy for the detection of misery perfusion using the criteria determined by baseline CBF and CVR was evaluated in all patients and in only those patients with occlusive lesions. Ten of 24 patients with misery perfusion showed a significant reduction in CVR. Using criteria determined by significant decreases in CVR and baseline CBF, misery perfusion was detected with a sensitivity of 42% and a specificity of 95% in all patients. In patients with occlusive lesions (n=50), sensitivity was higher but specificity was slightly lower. The diagnostic accuracy of the threshold determined by baseline CBF alone was similar in all patients and in only those patients with occlusive lesions, and was higher than that achieved using the asymmetry index of OEF. Reductions in CVR and baseline CBF in the ACZ challenge for CVD would detect misery perfusion with high specificity. Reduction in baseline rCBF is more accurate than reduction in CVR alone for the detection of misery perfusion. (orig.)