ESSEN Score Combined With Magnetic Resonance Imaging Predict The Risk Of Cerebral Infarction After TIA During1-Year

Objective: Transient ischemic attack (TIA) is a early warning signs ofcerebral infarction. TIA repeately attack could progress into cerebral infarction.Cerebral infarction has a high morbidity and mortality. Therefore we have topredict the risk of cerebral infarction in patients with TIA, and give timelyintervention.Many studies respectively confirm that ESSEN score andmagnetic resonance imaging can predict the risk of cerebral infarction afterTIA. We firstly explore the predictive value, which is the ESSEN scorecombined with cranial MRI predicted the risk of cerebral infarction after TIAduring1-year. It will help the risk stratification and optimal treatment inpatients with TIA, and prevent TIA progressing into cerebral infarction.Methods: Consecutive patients with TIA during7days of onset werehospitalized in the department of neurology of second Hospital of HebeiMedical University between June2009and January2011were recruited. Allpatients underwent MRI. The following data were collected: history, clinicalsymptoms, previous history, family history, bad habits and related auxiliaryexamination results. In addition, we calculated the ESSEN score and the newscore of ESSEN score combined with magnetic resonance imaging. Patientswere Periodically followed after TIA during1-year. Cerebral infarction, deathor1-year is the endpoint.All analyses were performed by the SPSS statistical package version13.0.Chi-square for categorized variables. Receiver operator characteristic(ROC)curve was used to analyse the prediction accuracy of the ESSEN score andnew score to the risk of cerebral infarction after TIA during1-year. Logisticregression analyze was for the correlation of factors and cerebral infarctionwith TIA. P＜0.05was considered statistically significant. Results:1A total of332TIA patients were included in the study.219cases were male,113cases were female. There was not significant differences between men andwomen (P＞0.05). Age from17to96years old (mean56.35±12.845SD).There were143cases with hypertension;105cases with diabetes mellitus;16cases with myocardial infarction;128cases with other heart diseases (exceptfor atrial fibrillation);34cases with peripheral vascular disease;117caseswith smoking history;84cases with history of TIA or stroke. MRI showedsignal intensity changes suggestive of cerebral ischemia in138patients. MRAdetected stenoses≥50%in87patients. There were94cases(28.31%) cerebralinfarction with TIA, no death.2ESSEN score was0points in36cases,1points in77cases,2points in92cases,3points in60cases,4points in44cases,5points in16cases,6pointsin7cases. The area under the receiver operating characteristic curve was0.727(95%CI:0.668,0.787, P＜0.05) by the ESSEN score predicted the riskof cerebral infarction after TIA. The sensitivity and specificity was themaximum when the interception points of new score=2.5. According toESSEN score for risk stratification to all patients with TIA, low risk group(0~2),205cases; high-risk group (＞2),127cases. There were35cases(17.1%) cerebral infarction in low risk group,59cases(46.5%) cerebralinfarction in high risk groups. Probability of cerebral infarction in high riskgroup was higher than the low risk group. By χ2test, there were statisticalsignificant between each layer of ESSEN score predicted the risk of cerebralinfarction after TIA (P＜0.05).3The each factor was analyzed by stepwise Logistic regression analysis forcerebral infarction after TIA. the standard for selecting was0.05, for excludingwas0.10, age group (X1), hypertension (X2), diabetes mellitus (X3), durationof symptoms(X4) could enter into the equation, logitP=－2.591+0.985X1+1.319X2+0.524X3+0.336X4，P=0.000； χ2=57.197, P＜0.05, there wasstatistical significance. All the regression coefficient B value was positive,which showed predictive factors being positively interrelated with cerebralinfarction to TIA patients. 4332patients with TIA, there were138cases with ischemic lesions in MRI,85cases of cerebral infarction. The probability of cerebral infarction washigher than the patients without ischemic lesions in MRI, there was statisticalsignificant between no lesion group and lesion group of cerebral infarctionafter TIA (P＜0.05). There were87cases MRA stenosis of more than50%,cerebral infarction in54cases. MRA show stenosis≥50%had a higherprobability of cerebral infarction. There was statistical significant betweenstenosis≥50%group and stenosis＜50%group of cerebral infarction afterTIA (P＜0.05).5The area under the receiver operating characteristic curve was0.828(95%CI:0.781,0.874, P＜0.05) by the new score predicted the risk ofcerebral infarction after TIA. The sensitivity and specificity was the maximumwhen the interception points of new score=3.5. According to new score forrisk stratification to all patients with TIA, low risk group (0~3),218cases;high-risk group (＞3),114cases. There were30cases cerebral infarction inlow risk group,64cases cerebral infarction in high risk group. Probability ofcerebral infarction in high risk group was higher than the low risk group. Byχ2test, there were statistical significant between each layer of new scorepredicted the risk of cerebral infarction after TIA during1-year (P＜0.05).Conclusion: An ESSEN score＞2was associated with an increasedgeneral risk for cerebral infarction in1year follow-up after TIA. Stenosis≥50%in MRA, ischemic changes in MRI respectively was higher risk forcerebral infarction in1year follow-up after TIA. The new score had a biggerprediction value to the risk of cerebral infarction after TIA during1-year. Thenew score＞3was associated with an increased general risk for cerebralinfarction in1year follow-up after TIA.