Early spontaneous hyperperfusion after stroke

To clarify the relationships between early hyperperfusion (i.e. the hallmark of early, efficient recanalization in animal stroke models) and ultimate infarction, we have compared acute-stage perfusion PET images and chronic-stage CT scans in patients with middle cerebral artery (MCA) stroke. We used PET and the oxygen-15 (15O) equilibrium method to obtain cerebral blood flow (CBF), cerebral blood volume (CBV), oxygen extraction fraction (OEF) and cerebral metabolic rate of oxygen consumption (CMRO2) parametric images in 30 consecutive, still symptomatic, first-ever MCA territory stroke patients without sign of haemorrhage at admission CT scan. Each subject was studied twice, first within 5-18 h of stroke onset, and, in survivors, approximately 1 month later; a plain CT scan (co-registered with PET) was performed approximately 1 month after onset. Following initial screening based on acute-stage perfusion images, 10 survivors with focal hyperperfusion in the appropriate MCA territory confirmed by computer were declared eligible. In each patient, the topography and volume of both hyperperfusion and infarction (delineated on late CT scan) were recorded, and all PET parameters were obtained for both areas and both times. The hyperperfused areas affected the cortical MCA territory, often widely so and in a patchy fashion; they were topographically distinct from, and consistently larger than (P < 0.01, Wilcoxon sign test) the final infarcts, which were small and generally deep-seated. In none of the nine patients in whom it was successfully performed did transcranial Doppler reveal MCA stem occlusion. In the hyperperfused regions, the acute-stage perfusion, blood volume and oxygen consumption were significantly increased, and the OEF significantly reduced, while all these variables had significantly returned toward normality in the chronic-stage PET study. The ultimately infarcted area did not exhibit significant hyperperfusion in the acute stage. The areas with acute-stage hyperperfusion exhibited haemodynamic and metabolic abnormalities consistent with post-recanalization hyperperfusion, i.e. vasodilatation and "luxury perfusion'. Increased oxidative metabolism, previously reported only in animals, presumably reflects an overshoot of protein synthesis. The fact that the areas with hyperperfusion, though extensive, were topographically distinct from the infarcted region, suggests that spontaneous non-haemorrhagic hyperperfusion, when documented 5-18 h after onset, is a harmless and even perhaps beneficial phenomenon. These results have implications for clinical trials.