A novel magnetic resonance imaging approach to collateral flow imaging in ischemic stroke.

Dedicated magnetic resonance (MR) imaging (MRI) sequences for evaluation of collaterals can be generated using MR perfusion (MRP) source data. We compared a novel collateral flow imaging technique with digital subtraction angiography (DSA) for determining collateral circulation in acute stroke and evaluated the ability of MR-based collateral flow maps to predict outcomes after recanalization therapy. Consecutive patients who were candidates for endovascular treatment were enrolled. A collateral flow map derived from MRP source data was generated by manual or automatic postprocessing. Collateral grading based on the collateral flow map was performed and compared with grading based on DSA. Clinical and radiological outcomes were evaluated according to MR-based collateral grading and early reperfusion (ER) status. There was good correlation between MRI-based and DSA-based collateral grades (weighted κ-coefficient = 0.70). Collateral status and achievement of ER were the 2 main determinants of a favorable functional outcome and neurological improvement, in addition to infarct growth. Regardless of achievement of ER, better collaterals were significantly associated with a lower modified Rankin score at day 90 (p < 0.001 for trend in both ER(-) and ER(+) ). Most symptomatic intracranial hemorrhages occurred in patients with a poor collateral grade and ER(+) , whereas no patient with excellent collaterals suffered symptomatic intracranial hemorrhage or died. MRI techniques to assess collaterals are rapidly being developed, and may provide insight into collateral perfusion. The combination of collateral images derived from pretreatment MRP source data and reperfusion status is a robust predictor of outcomes in acute ischemic stroke.