Is there an apparent diffusion coefficient threshold in predicting tissue viability in hyperacute stroke?

BACKGROUND AND PURPOSE Rapid and precise identification of the penumbra is important for decision-making in acute stroke. We sought to determine whether an early and moderate decrease in the apparent diffusion coefficient (ADC) may help to identify, within the diffusion/perfusion (DWI/PWI) mismatch, those areas that will eventually evolve toward infarction. METHODS We reviewed 48 patients not treated by thrombolytics who had a DWI/PWI within 6 hours after onset, with infarct evolution documented by follow-up magnetic resonance on days 2 to 4. We calculated absolute values for ADC and the ADC ratio (ADCr) in (1) the initial DWI hypersignal; (2) the final volume of the infarct, ie, the follow-up fluid-attenuated inversion recovery abnormalities; (3) the infarct growth (IGR) area; and (4) the oligemic area (OLI) that remained viable despite initial hemodynamic disturbance. We tested the value of the ADC to predict tissue outcome by using discriminant analysis. RESULTS ADC values were marginally but significantly decreased in the IGR area (ADC 782+/-82x10(-6) mm(2)/s, ADCr 0.94+/-0.08) compared with mirror values (P=0.01) and with OLI (ADC 823+/-41x10(-6) mm(2)/s, ADCr 0.99+/-0.07; P=0.001). Of all quantitative DWI and PWI parameters, the ADCr best discriminated between IGR and OLI (F(1,50)=13.6, cutoff=0.97, 64% sensitivity, 92% specificity) and between the final volume of infarct and OLI (F(1,83)=219, cutoff=0.91, 91% sensitivity, 100% specificity). CONCLUSIONS A simple approach based on ADC alone may allow the identification of tissue at risk of infarction in acute-stroke patients.