Background: the ASPECTS- collateral score on CT-angiograms was shown to be successful in prognosticating functional outcomes and complications during intravenous thrombolysis in acute ischemic stroke (AIS). We studied predetermined topological information to see if any specific region had more prognostic value.
Methods: consecutive patients from 2010-2014 with intracranial internal carotid artery, M1 or M2 middle carotid artery occlusions treated with intravenous thrombolysis were included. The primary outcome measure was good clinical outcome (3-month modified Rankin Scale score 0-1). We scored each region as 0= no collaterals, 1= poor compared to contralateral and 2= good collaterals. Prognostic value of the 6 cortical ASPECTS-collateral regions in predicting outcomes was determined by multivariable logistic regression.
Results: 310 patients were included (Median age, 66.1±14.5 years; median National Institutes of Health Stroke Scale (NIHSS)- 18 points (range 3-36). Inter-rater reliability for ASPECTS-collaterals was good (κ=0.78). There was no Statistical collinearity among ASPECTS-collateral regions. Using multivariable logistic regression, only the M5 region (odds ratio, 2.72, 95%CI 1.52-4.84, p =0.001), age (OR 0.957 per yr 95%CI 0.936-0.978, p <0.001), Diabetes (OR 0.367, 95%CI 0.193- 0.700, p =0.002) and NIHSSS (OR 0.878 per point, 95%CI 0.836-0.922, p=0.001) were significantly associated with good outcomes. When compared with NIHSS, the receiver operating characteristic curves for NIHSS+M5 (area under the curve, 0.749) correlated well with clinical severity scores. Addition of M5 collateral score showed a statistically significant additive effect to the NIHSS score for predicting good outcomes (Z score: -1.684, p=0.045).
Conclusions: Involvement of the parietal region (M5) regions is a reliable predictor of clinical outcome in anterior circulation large artery occlusion. This simple radiological marker can strengthen the clinical NIHSS score and may be considered during prognostication