Predictive Utility of Cerebral Blood Flow Transients in Experimental Stroke

A gap in developing novel preclinical treatment strategies for ischemic stroke is predicting long-term outcome in experimental stroke models early during ischemia to reduce heterogeneity and sample size. Besides saving costs through improved risk stratification, reducing the number of animals is a requirement of the 3Rs principle. In this secondary analysis, we analyzed 28 Sprague-Dawley rats of a prospective data base that underwent 90-minute filament-occlusion of the middle cerebral artery (MCAO) to assess the predictive power of early variables at 30, 60, and 90 min after occlusion. The animals were sacrificed after 72 h. Infarct sizes were determined by hematoxylin staining. In a minimally invasive fashion, we recorded cerebral blood flow (CBF) with laser-Doppler flowmetry and direct current (DC)/alternating current-electrocorticography (ECoG) with epidural Ag/AgCl electrodes. Both CBF and ECoG markers correlated with the cortical infarct volumes. Multiclass receiver operating characteristic analysis identified the best predictors of three risk classes, and Spearman's rank correlation was used to explore relationships between ECoG and CBF. The slope of the CBF transients in response to spreading depolarization (SD) was the best biomarker at all time points, while the DC integral was the best epidural biomarker. Both correlated negatively at all time points (ρ < -0.68). In summary, we have found that early risk stratification during MCAO in rats is possible using minimally invasive biomarkers. This would enable, in particular, the early sorting out of animals with a low risk of cortical infarction in neuroprotection studies, where these animals typically distort the statistical results.