Predictive value of brain and vascular imaging including intracranial vessels in transient ischaemic attack patients: external validation of the ABCD3-I score



Background and purpose

Recently, brain and vascular imaging have been added to clinical variables to identify patients with transient ischaemic attack (TIA) with a high risk of stroke recurrence. The aim of our study was to externally validate the ABCD3-I score and the same score taking into account intracranial circulation.


We analyzed data from 1137 patients with TIA from the PROMAPA study who underwent diffusion-weighted magnetic resonance imaging (DWI) within 7 days of symptom onset. Clinical variables and diagnostic work-up were recorded prospectively. The end-points were subsequent stroke at 7 and 90 days follow-up.


A total of 463 (40.7%) subjects fulfilled all inclusion criteria. During follow-up, eight patients (1.7%) had a stroke within 7 days, and 14 (3.1%) had a stroke within 3 months. In the Cox proportional hazard multivariate analyses, the combination of large-artery atherosclerosis and positive DWI remained as independent predictors of stroke recurrence at 7- and 90-day follow-up [HR 8.23, 95% confidence interval (CI) 2.89–23.46, P < 0.001]. The ABCD3-I score was a powerful predictor of subsequent stroke. The area under the receiver operating characteristic curve was 0.83 (95% CI 0.72–0.93) at 7 days and 0.69 (95% CI 0.53–0.85) at 90 days. When we include intracranial vessel disease in the score, the area under the curve increases but the difference observed was non-significant.


The inclusion of vascular and neuroimaging information to clinical scales (ABCD3-I score) provides important prognostic information and also helps management decisions, although it cannot give a complete distinction between high-risk and low-risk groups.