The present study investigates the pathogenesis of focal cerebral hyperemia, its effect on brain tissue and discusses its pathophysiological and therapeutic importance in the light of interpreting severe hyperemia as a sign of arterial reopening probably due to embolic migration. Cerebral angiography, serial CT-scans and serial TC99 -scans were performed in a consecutive group of 73 patients with completed stroke all admitted to hospital within 3 days after stroke onset. When possible the regional cerebral blood flow (rCBF) was studied with the intracarotid Xe 133 injection method. Twenty-nine patients had evidence of middle cerebral artery (MCA) occlusion; rCBF was investigated in 24. Fourteen patients had either occlusion or severe internal carotid artery (ICA) stenosis; rCBF was not measured in these patients. Thirty patients had no angiographical evidence of MCA occlusion, ICA occlusion or severe ICA stenosis; rCBF was investigated in 24. Focal hyperemia was observed in 21 patients but exclusively in the group with evidence of MCA occlusion. Hence, these 21 patients are typical and representative for the group of patients with evidence of MCA occlusion. Hyperemia was found in infarcted as well as in non-infarcted tissue. Apparently, it is the severity of the initial ischemic episode and not the hyperemia that determines whether or not tissue necrosis develops. Interpreting severe hyperemia as a sign of arterial reopening and embolic migration (evidenced by partial reopening affecting only some MCA branches) reopening had occurred in about 1/3 of the patients with MCA occlusion before they were examined 1 to 4 days after stroke onset. Autopsy studies performed in 8 of the patients with MCA occlusion indicate that arterial reopening also takes place in many patients later on (7 of 8). According to this interpretation, hypothetical as it is, the changing position of the embolus is associated with partial or complete reperfusion leading to hyperemia in the initially ischemic brain tissue. The hemodynamic basis for appropriate therapy therefore may change from one day to the next in the acute state of stroke due to MCA occlusion.

In a group of 48 patients with completed stroke, 8 patients had viable collaterally perfused brain tissue which was accessible for rCBF recordings with a two dimensional technique. All 8 had deep subcortical infarcts on CT-scan, and angiographic occlusion of the arteries normally supplying the infarcted territory. The brain tissue overlying the deep infarcts appeared normal on CT-scan and was supplied by collateral circulation. rCBF was measured in all within 72 hours after the stroke. The intra-carotid Xe-133 injection method and a 254 multidetector camera were used to study rCBF. Relatively ischemic low flow areas were a constant finding in the collaterally perfused tissue. In 6 of the patients, the collaterally perfused part of the brain had low flow values comparable to those of an "ischemic penumbra" (viable, but functionally depressed brain tissue due to inadequate perfusion). Autoregulation was impaired in all of the collaterally perfused areas while the CO2-response always was preserved. Steal phenomena were not seen. In the surrounding brain tissue, autoregulation was normal in 5 patients and impaired in 3 while the CO2-response seemed to be normal. The results confirm the experimental finding of an ischemic penumbra associated with acute cerebral infarcts and suggest that early restoration of the blood flow in acute stroke patients might improve recovery and prognosis in selected patients.