Penumbral probability thresholds of cortical flumazenil binding and blood flow predicting tissue outcome in patients with cerebral ischaemia.

Active treatment of acute ischaemic stroke can only be successful as long as tissue in the area of ischaemic compromise is still viable. Therefore, the identification of the area of irreversible damage, and its distinction from the penumbral zone, may improve the estimation of the potential efficacy of various therapeutic strategies. Ten patients (seven male, three female, aged 52-75 years) with acute ischaemic stroke, in whom MRI delineated an infarct involving the cortex 3 weeks after the attack, were studied by [(11)C]flumazenil (FMZ) PET to assess their neuronal integrity, and regional cerebral blood flow (CBF) was measured by H(2)(15)O PET 2-12 h (median interval 6 h) after onset of symptoms. Cortical volumes of interest (3 mm radius) were placed on co-registered CBF, FMZ and on late MRI scans. Using initial CBF and FMZ binding data from volumes of interest finally located within or outside the cortical infarct, cumulative probability curves were computed to predict eventual infarction or non-infarction. Positive (at least 95% chance of infarct) and negative (at least 95% chance of non-infarct) prediction limits for CBF (4.8 and 14.1 ml/100 g/min, respectively) and for FMZ binding (3.4 and 5.5 times the mean of normal white matter, respectively) were determined to define the penumbral range. Using the lower FMZ binding threshold of 3.4 for irreversible tissue damage and the upper CBF value of 14.1 ml/ 100 g/min for the threshold of critical perfusion at or above which tissue will likely be preserved, various cortical subcompartments were identified: of the final cortical infarct (median size 25.7 cm(3)) a major portion comprising, on average, 55.1% showed FMZ binding critically decreased, thus predicting necrosis. In 20.5% of the final infarct, on average, CBF was in the penumbral range (<14.1 ml/100 g/min) and FMZ binding was above the critical threshold of irreversible damage. Only 12.9% of the final infarct exhibited neuronal integrity and CBF values above the penumbral range. Therefore, most of the final infarct is irreversibly damaged already at the time of the initial evaluation, when studied several hours after stroke onset. A much smaller portion is still viable but suffers from insufficient blood supply: this tissue may be salvaged by effective reperfusion. Only an even smaller compartment is viable and sufficiently perfused, but eventually becomes necrotic, mainly owing to delayed mechanisms, and may benefit from neuroprotective or other measures targeted at secondary damage. Therefore, early reperfusion is crucial in acute ischaemic stroke.