Combining Diffusion-Weighted Imaging Patterns with ABCD2 Score Predicts Stroke Risk after Transient Ischemic Attack.

Abstract

Recent studies indicated that multiple acute infarctions on Diffusion- Weighted Imaging (DWI) were associated with a higher risk of stroke. The study aims to estimate the association of different infarction patterns and ABCD2 score with the prognosis of Transient Ischemic Attack (TIA). We prospectively analyzed the data from TIA database of the First Affiliated Hospital of Zhengzhou University. The predictive outcome was a 90-day ischemic stroke. Cox proportional hazards model was used to evaluate the predictive value of risk factors associated with stroke. The receiver-operating characteristics curves were plotted, and the predictive value was assessed by computing the Area Under the Curve (AUC). A total of 1376 eligible patients were enrolled. DWI patterns were significant predictors for stroke (single acute infarction: hazard ratio [HR] =2.942, p <0.001; multiple acute infarctions: HR =5.552, p <0.001, in comparison with no acute infarction). Patients with both multiple infarctions and ABCD2 ≥4 were associated with approximately 15.5-fold risk of stroke at 90 days (28.8% vs. 2.1%, HR =15.455, 95% confidence interval [CI], 7.946-30.057, p <0.001), compared with those with no infarction or ABCD2 <4. The ABCD2+ DWI patterns showed a better discrimination with an AUC of 0.765 (95% CI, 0.741-0.787) than the ABCD2 score (AUC =0.651; 95% CI, 0.625-0.676; Z =4.777; p <0.0001) and ABCD3-I score (AUC =0.724; 95% CI, 0.700- 0.748; Z =2.697; p =0.007). Combining infarction patterns with ABCD2 score could enhance the predictive value for early stroke risk in TIA.