DWI-Based Algorithm to Predict Disability in Patients Treated with Thrombectomy for Acute Stroke
Abstract
BACKGROUND AND PURPOSE The reasons for poor clinical outcome after thrombectomy for acute stroke, concerning around half of all patients, are misunderstood. We developed a hierarchic algorithm based on DWI to better identify patients at high risk of disability. MATERIALS AND METHODS Our single-center, retrospective study included consecutive patients with acute ischemic stroke who underwent thrombectomy for large anterior artery occlusion and underwent pretreatment DWI. The primary outcome was the mRS at 3 months after stroke onset. Multivariable regression was used to identify independent clinical and imaging predictors of poor prognosis (mRS > 2) at 3 months, and a hierarchic algorithm predictive of disability was developed. RESULTS A total of 149 patients were analyzed. In decreasing importance, DWI lesion volume of >80 mL, baseline NIHSS score of >14, age older than 75 years, and time from stroke onset to groin puncture of >4 hours were independent predictors of poor prognosis. The predictive hierarchic algorithm developed from the multivariate analysis predicted the risk of disability at 3 months for up to 100% of patients with a high predictive value. The area under the receiver operating characteristic curve was 0.87. CONCLUSIONS The DWI-based hierarchic algorithm we developed is highly predictive of disability at 3 months after thrombectomy and is easy to use in routine practice. © 2020 by American Journal of Neuroradiology.