Prediction of Subsequent Stroke Risk in TIA Patients with Brain Infarction

Hakan Ay, Stroke Service, Dept of Neurology and AA Martinos Ctr for Biomedical Imaging, Dept of Radiology, Massachusetts General Hosp, Boston, MA; Ethem Murat Arsava, AA Martinos Ctr for Biomedical Imaging, Dept of Radiology, Massachusetts General Hosp, Boston, MA; Karen L Furie, Lee H Schwamm, Stroke Service, Dept of Neurology, Massachusetts General Hosp, Boston, MA; AGregory Sorensen, AA Martinos Ctr for Biomedical Imaging, Dept of Radiology, Massachusetts General Hosp, Boston, MA

Introduction: A new tissue-based definition of stroke proposes to classify neurological events associated with brain infarction as "stroke" whether symptoms are transient or permanent (Stroke 2009;40:2276). Because risk of stroke shortly after "TIA with infarction" is approximately 10 times higher than the risk after ischemic stroke, combining "TIA with infarction" and "stroke" in the same category may result in a loss of valuable prognostic information conveyed by the term "TIA". A prognostic tool to accurately stratify the risk of early recurrent stroke according to the new definition would be of value. The Recurrence Risk Estimator-90 (RRE-90) is a web-based (http://www.nmr.mgh.harvard.edu/RRE-90/), 7-point prognostic score that integrates baseline clinical and imaging information to quantify the risk of recurrent stroke at both 14-days and 90-days. The RRE-90 score was originally derived and validated in patients with ischemic stroke (excluding "TIA with infarction"). In this study, we evaluated the utility of the RRE-90 score for predicting 7-day risk of recurrent stroke in patients with "TIA with infarction".

Methods: We retrospectively identified consecutive patients with "TIA with infarction" during a 6-year period, calculated their RRE-90 and ABCD2 scores, and assessed the predictive ability of these scores by computing receiver operator characteristic (ROC) curves. We defined subsequent stroke as a clinical deterioration associated with new infarction spatially distinct from the index lesion.

Results: There were 255 patients with "TIA with infarction" diagnosed by DWI within 24 hours of symptom onset. Subsequent stroke developed in 15 patients (5.9%) within 7-days. The area under the ROC curve (AUC) was 0.83 (95%CI, 0.76-0.91) for the RRE-90 score. The sensitivity and specificity of RRE-90 score >2 for predicting 7-day stroke risk were 87% and 63% respectively. The AUC for the ABCD2 score was 0.56 (95%CI 0.44-0.68). The p value for comparison between the RRE-90 and ABCD2 scores was <0.01.

Conclusion: The RRE-90 score performs well at predicting 7-day risk of subsequent stroke in TIA patients with infarction on DWI. This suggests that RRE-90 score can be reliably applied to the general population of stroke per the new tissue-based definition to identify those with imminent risk of developing recurrent stroke. Accurate identification of high-risk patients may facilitate early intervention with targeted stroke prevention strategies.