Accuracy of the Alberta Stroke Program Early CT Score during the first 3 hours of middle cerebral artery stroke: comparison of noncontrast CT, CT angiography source images, and CT perfusion

BACKGROUND AND PURPOSE: The Alberta Stroke Program Early CT Score (ASPECTS) is a reliable method of delineating the extent of middle cerebral artery (MCA) stroke. Our aim was to retrospectively compare the accuracy of ASPECTS on noncontrast CT, CT angiography (CTA) source images, and CT perfusion maps of cerebral blood volume (CBV) during the first 3 hours of middle cerebral artery (MCA) stroke.MATERIALS AND METHODS: First-time patients with MCA stroke who presented <3 hours from symptom onset and were evaluated by noncontrast CT/CTA/CT perfusion, had confirmed acute nonlacunar MCA infarct on diffusion-weighted MR imaging (DWI) within 7 days, and had follow-up angiography were included. Patients were excluded for persistent MCA occlusion or stenosis. Two raters through consensus assigned an ASPECTS on the noncontrast CT, CTA source images, and the section-selective (2 × 12 mm coverage) CT perfusion CBV maps. ASPECTS on follow-up DWI served as the reference standard. For each CT technique, the detection rates of regional infarction, the mean ASPECTS, and the linear correlation to final ASPECTS were determined and compared. P values <.05 were considered significant.RESULTS: Twenty-eight patients satisfied the criteria with DWI performed at a mean of 50.3 hours (range, 22–125 hours) post-CT imaging. Of 280 ASPECTS regions, 100 were infarcted on DWI. The accuracy of noncontrast CT, CTA source images, and CT perfusion CBV for detecting regional infarct was 80.0%, 84.3%, and 96.8%, respectively (P < .0001). The mean ASPECTSs of noncontrast CT, CTA source images, CT perfusion CBV, and DWI were 8.4 ± 1.8, 8.0 ± 1.8, 6.8 ± 1.9, and 6.5 ± 1.8, respectively. The mean noncontrast CT and CTA source image ASPECTS was different from that of DWI (P < .05). Correlation of noncontrast CT, CTA source images, and CT perfusion CBV ASPECTS with final ASPECTS was r² = 0.34, r² = 0.42, and r² = 0.91, respectively.CONCLUSION: In a retrospective cohort of MCA infarcts imaged <3 hours from stroke onset, ASPECTS was most accurately determined on CT perfusion CBV maps.

The Alberta Stroke Program Early CT Score (ASPECTS) is a semiquantitative method of defining infarct extent in the middle cerebral artery (MCA) territory that is more reliable than the “1/3 MCA” rule^1,2 and is highly predictive of outcome.^3-6 ASPECTS was designed for conventional noncontrast CT, the ubiquitously performed imaging technique for triaging acute stroke. However, several recent studies have reported application of ASPECTS to CT angiography (CTA) source images^7-9 and CT perfusion parametric color maps,^10-12 with data to suggest that the acute ASPECTS is more accurately determined on these advanced CT techniques. Our aim was to retrospectively compare the accuracy and strength of the correlation of the acute ASPECTSs obtained on noncontrast CT, CTA source images, and CT perfusion mapping of cerebral blood volume (CBV) for nonlacunar MCA territory strokes imaged during the initial 3 hours of symptom onset. We hypothesized that with arterial recanalization, the CT perfusion CBV map is the best predictor of the final ASPECTS.