Abstract TP53: NIHSS and Variation of Infarct Volume by Hemisphere

Abstract

Introduction: The National Institute of Health Stroke Scale (NIHSS) is the most widely used measure of neurological deficits in clinical trials. Using the placebo arm of the NINDS t-PA Stroke Trial, it has been demonstrated that the total volume of cerebral infarction in patients with similar NIHSS scores is greater for right compared to left hemisphere strokes. Our objective was to verify this finding in a non-clinical trial, independent data set of acute ischemic strokes. Methods: The Greater Cincinnati/Northern Kentucky Stroke Study (GCNKSS) is a population-based study that tracks the incidence of stroke. A convenient subsample from the 2010 GCNKSS ischemic stroke cohort (N= 368) underwent detailed imaging analysis. Research nurses abstracted patient reords to include baseline retrospective NIHSS score. The 24 hour infarct volume was segmented using manual tracing. NIHSS was compared between left and right brain using Wilcoxon rank sum test. Spearman rank correlation determined the association between the NIHSS score and infarct volume by hemisphere. Patients were stratified by NIHSS (0 to 5, 6 to 20, and greater than 20). Results: Among 368 ischemic stroke subjects with imaging data, excluded were 77 brainstem or cerebellar infarcts, 37 undetermined volume, 1 missing NIHSS, and 3 undetermined laterality. For the remaining 250 patients, 132 were left and 118 were right hemisphere strokes, and 210 had an MRI and 40 had CT. Median time from onset to imaging was 24 hours. Baseline NIHSS was similar by hemisphere and correlated with stroke volume (r=0.38, p<0.01). The infarct volumes of right hemisphere strokes were greater than left hemisphere (p-value=0.02) (Table). Conclusion: The NIHSS score correlates with volume by hemisphere, but has a larger infarct volume for right hemisphere than left hemisphere for similar NIHSS. This likely reflects the different weighting of the NIHSS with regard to language. This finding confirms prior results in an independent dataset.