Current Smoking and Silent Brain infarct are associated with Stroke Occurrence in the migrainers with Right to Left Shunt

초록

Background and Purpose: Right-to-left shunt (RLS) is known to be associated with cryptogenic cerebral infarction and migraine. However, it has not been elucidated, which factors affect the stroke occurrence in the patients with RLS. Here, we sought to determine the prevalence of risk factors in two groups and to find the radiological predictors for stroke. Methods: We performed a retrospective review of the medical records and brain MR image of patients with RLS, who visited our hospital for treatment of migraine or cryptogenic infarction from January, 2008 to June, 2012. Patients with potential risk factors for cardioembolism like aortic atheroma(>4mm) and atrial fibrillation were excluded. RLS was defined as ≥11 microembolic signals(MESs) on the agitated saline test, using Transcranial Doppler sonography. We defined white matter hyperintensities(WMHs), as the Fazekas score≥2, and assessed the prevalence of each risk factors and radiological findings in stroke and migraine group. Results: A total of 489 patients were included in this study(stroke; 55.7±11.0, N=257 vs migraine; 56.0±10.5, N=232). On univariate analysis, the patients with stroke had the higher ratio of men, hypertension, diabetes, dyslipidemia and current smoking than the migrainers. On multivariate binary regression test, the odds ratio of current smoking was highest(OR=10.08, 95% CI; 4.38~23.17). The stroke group had the high frequency of silent brain infarct(28.3% vs 5.2 %, P<0.01) and WMHs(13.1% vs 6.9%), compared to those with migraine. There was no difference in age, intra/extracranial stenosis, microembolic signal number or microbleeds between two groups. Conclusions: Some cardiovascular risk factors, especially silent brain infarct and smoking history were associated with stroke occurrence in the patients with RLS. These results suggest that the evaluation of risk factors and brain images could be helpful for the prediction of stoke in the patients with RLS.