Perfusion CT Imaging Follows Clinical Severity in Left Hemispheric StrokesFurtado A.D.a · Smith W.S.b · Koroshetz W.d · Dillon W.P.a · Furie K.L.d · Lev M.H.e · Vittinghoff E.c · Schaeffer S.a · Biagini T.b · Hazarika O.b · Wintermark M.a
aNeuroradiology Section, Department of Radiology, bDepartment of Neurology and cDepartment of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, Calif., and dStroke Service and eNeuroradiology Section, Department of Radiology, Massachusetts General Hospital, Boston, Mass., USA
Objective: The purpose of this study was to assess how imaging findings on admission perfusion CT (PCT) and follow-up noncontrast CT (NCT), and their changes over time, correlate with clinical scores of stroke severity measured on admission, at discharge and at 6-month follow-up. Methods: Fifty-two patients with suspected hemispheric acute ischemic stroke underwent a PCT within the first 24 h of symptom onset and a follow-up NCT of the brain between 24 h and 3 months after the initial stroke CT study. NIH Stroke Scale (NIHSS) scores were recorded for each patient at admission, discharge and 6 months; modified Rankin scores were determined at discharge and 6 months. Baseline PCT and follow-up NCT were analyzed quantitatively (volume of ischemic/infarcted tissue) and semiquantitatively (anatomical grading score derived from the Alberta Stroke Program Early CT Score). The correlation between imaging volumes/scores and clinical scores was assessed. Analysis was performed for all patients considered together and separately for those with right and left hemispheric strokes. Results: Significant correlations were found between clinical scores and both quantitative and semiquantitative imaging. The volume of the acute PCT mean transit time lesion showed best correlation with admission NIHSS scores (R2 = 0.61, p < 0.001). This association was significantly better for left hemispheric strokes (R2 = 0.80, p < 0.001) than for right hemispheric strokes (R2 = 0.39, p = 0.131). Correlation between imaging and NIHSS scores was better than correlation between imaging and modified Rankin scores (p = 0.047). The correlation with discharge clinical scores was better than that with 6-month clinical scores (p = 0.012). Conclusions: Baseline PCT and follow-up NCT volumes predict stroke severity at baseline, discharge and, to a lesser extent, 6 months. The correlation is stronger for left-sided infarctions. This finding supports the use of PCT as a surrogate stroke outcome measure.
© 2008 S. Karger AG, Basel