Objective:
Previous studies have shown that the prolonged duration of TIA symptoms or ABCD2 score are associated with DWI abnormality, and the presence of DWI abnormality is associated with an increased early risk of stroke. However, there are few reports that show the relation between TIA clinical etiology and DWI abnormality. Our aim of this study is to clarify the prevalence of positive DWI in relation to characteristics of patients and TIA.
Methods:
The subjects were enrolled from patients who were admitted to our stroke unit within 7 days after symptom onset from January 2006 to July 2013. The diagnosis of TIA was done by NINDS criteria, and we classified TIA etiology by TOAST classification based on clinical symptoms, ECG monitoring, carotid ultrasound, MR angiography and transesophageal echocardiography. All patients underwent DWI-MRI within 7 days after symptom onset. We examined an association between TIA etiology, symptom, duration of symptoms and DWI abnormality.
Results:
A total of 141 patients (mean 64 years; 63% men) were admitted with TIA during this period. Those included lacuna TIA (n=17, 12.1%), atherothrombotic TIA (n=32, 22.7%), cardioembolic TIA (n=23, 16.3%), TIA due to other causes (n=35, 24.8%), and TIA with unknown etiology (n=34, 24.1%). Prevalence of positive DWI findings were 47.1% in lacunar TIA, 43.7% in atherothrombotic TIA, 52.1% in cardioembloic TIA, 42.8% in TIA due to other causes, and 23.5% in TIA with unknown etiology. DWI abnormality was the most frequent in cardiogenic TIA. In relation to symptom duration, the prevalence of DWI positive findings were 45.2% in less than 1 hour (N=53), 36.6% in 1-3 hour (N=41), 25.0% in 3-6 hour (N=12) and 42.9% in 6-24 hours (N=35). In relation to motor symptoms, there was no difference in prevalence of DWI abnormality between patient with motor symptoms (39.8%, N=113) and without (42.8%, N=28). There was no relation between DWI abnormalities and age, a history of stroke/TIA episode or vascular risk factors.
Conclusion:
Prevalence of DWI positive findings was high in cardiogenic TIA, and low in TIA with unclassified etiology. There were no relation between DWI abnormality, duration of symptom, and motor symptom.