Relationship between Clinical Factors and Hyperintense Acute reperfusion Marker Sign in Patients with Transient Ischemic Attack

다국어 초록 (Multilingual Abstract)

Relationship between Clinical Factors and Hyperintense Acute reperfusion Marker Sign in Patients with Transient Ischemic Attack by Taewoo Kim Doctor of Philosophy in Medicine Graduate School of Kyung Hee University Advised by Dr. Sung Hyuk Heo Background: Transient ischemic attack (TIA) is a transient focal cerebral dysfunction of vascular origin for which there is no definitive diagnostic tool. The hyperintense acute reperfusion marker (HARM) sign is a hyperintense signal observed on postcontrast fluid- attenuated recovery inversion images and is strongly associated with cerebral ischemic insults. The clinical significance of the HARM sign in TIA has rarely been studied, unlike that in stroke. This study investigated the relationship between the HARM sign and various clinical factors in diffusion-weighted imaging (DWI)-negative TIA. Furthermore, we investigated the relationship between the HARM sign with recurrence of TIA and ischemic stroke. Methods: We included 329 consecutive patients with DWI-negative TIA and divided them into two groups according to the HARM sign: 299 patients in the HARM(-) group and 30 patients in the HARM(+) group. Clinical information, brain imaging, and follow-up data were gathered from medical records and phone calls and compared using the HARM sign. Results: The patients with HARM sign were older (70.7 vs. 64.4 years, p = 0.007), had more previous TIA or stroke history within 12 months (26.7% vs. 4.0%, p < 0.001), and had higher systolic blood pressure (154.3 vs. 144.1, p = 0.022). The HARM(+) group also had a shorter symptom duration of <1 hour (63.3% vs. 38.8%, p = 0.009) and more symptomatic stenosis (50–99%) or occlusion (60.0% vs. 14.0%, p < 0.001). Among the transient neurological symptoms, only cortical symptoms were more prevalent in the HARM(+) group (30.0% vs. 8.7%, p = 0.002). The total follow-up duration of both groups was similar, and the Kaplan- Meier analysis showed a higher cumulative incidence of recurrent stroke in the HARM(+) group (log-rank test, p = 0.007). However, in the multivariate Cox analysis, the HARM sign was not independently associated with recurrent stroke, and only symptomatic stenosis or occlusion was independently associated with recurrent stroke in DWI-negative TIA cases (hazard ratio 3.412, 95% confidence interval 1.061–10.972). Conclusion: The HARM sign was related to various clinical factors in patients with DWI- negative TIA. The HARM sign provides insight into the etiology of TIA and may be a useful imaging biomarker for predicting prognosis after TIA. Keywords Transient ischemic attack; hyperintense acute reperfusion marker; recurrent stroke

목차 (Table of Contents)

• Abstract iv
• 1. Introduction · 1
• 2. Materials and Methods · 2
• 1) Patient Selection and Study Design · 2
• 2) Clinical Information 3
• 3) Image Acquisition and Analysis 4
• 4) Follow-up Data 6
• 5) Statistical Analysis 7
• 3. Results · 8
• 1) Baseline Characteristics of Study Participants 8
• 2) Clinical Symptoms in Patients with TIA 9
• 3) Recurrence of TIA and Stroke in Patients with TIA 10
• 4. Discussion · 20
• 5. References · 24
• Table Contents
• Table 1.
• Baseline Characteristics 13
• Table 2.
• Logistic Regression Analysis for HARM sign 15
• Table 3.
• Transient Neurological Symptoms 16
• Table 4.
• Follow-up Data · 17
• Table 5.
• Cox Proportional Model for Risks of Recurrent Stroke 19
• Figure Contents
• Figure 1.
• Flow Chart of Patient Selection · 11
• Figure 2.
• Brain MRI Examples of patients with HARM 12
• Figure 3.
• Kaplan-Meier Analysis · 18