Prognostic value of "tissue-based" definitions of TIA and minor stroke: Population-based study.

OBJECTIVE: Since use of diffusion-weighted imaging (DWI) positivity in the "tissue-based" definition of stroke in patients with a clinical TIA is supported by the high associated 90-day risk of recurrent stroke, we aimed to determine long-term prognostic significance, stratified by etiologic subtype, and whether the same tissue-based distinction is predictive in minor strokes. METHODS: Consecutive eligible patients with TIA or minor stroke (NIH Stroke Scale [NIHSS] ≤3) in the population-based Oxford Vascular Study underwent brain MRI at baseline. Stroke risk on 10-year follow-up was stratified by NIHSS (0/1 vs 2/3) and Trial of Org 10172 in Acute Stroke Treatment classification of the initial event. RESULTS: Among 1,033 patients (633 TIA; 400 minor stroke), 248 (24.0%) had acute lesions on DWI (13.9% of TIAs; 40.0% of minor strokes). A positive DWI was associated with an increased 10-year risk of recurrent ischemic stroke after an index TIA (hazard ratio [HR] 2.66, 95% confidence interval [CI] 1.28-5.54, p = 0.009) or a stroke with NIHSS 0-1 (3.03, 1.29-7.08, p = 0.011), but not after a stroke with NIHSS 2-3 (0.70, 0.24-2.10, p = 0.53). Ischemic stroke risk after DWI-positive TIA was at least equivalent to that after DWI-negative stroke (1.81, 0.82-4.00, p = 0.14). Among all patients, DWI positivity was most predictive of 10-year risk after cryptogenic events (4.68, 1.70-12.92, p = 0.003). CONCLUSION: DWI positivity is associated with an increased long-term risk of recurrent stroke after TIA and minor stroke, supporting a tissue-based definition of minor stroke as well as TIA. Prognostic value is greatest after cryptogenic events.