Fluid-attenuated inversion recovery vascular hyperintensity combined with clinical indicators for predicting intravenous thrombolysis outcomes in acute cerebral infarction
10.3969/j.issn.1672-5921.2025.05.004
- VernacularTitle:液体衰减反转恢复血管高信号征联合临床指标对急性脑梗死患者静脉溶栓预后的预测价值研究
- Author:
Xiaoqing YU
1
;
Peilan ZHANG
1
Author Information
1. 300041 天津医科大学附属环湖医院神经内科
- Publication Type:Journal Article
- Keywords:
Ischemic stroke;
Fluid-attenuated inversion recovery;
Vascular hyperintensity;
National Institutes of Health stroke scale score;
Neutrophil
- From:
Chinese Journal of Cerebrovascular Diseases
2025;22(5):317-327
- CountryChina
- Language:Chinese
-
Abstract:
Objective To investigate the predictive value of fluid-attenuated inversion recovery(FLAIR)vascular hyperintensity(FVH)combined with clinical indicators for the prognosis of intravenous thrombolysis in patients with acute cerebral infarction caused by middle cerebral artery(MCA)stenosis or occlusion.Methods This study retrospectively and consecutively enrolled patients with acute cerebral infarction caused by MCA stenosis or occlusion who underwent intravenous thrombolysis within 4.5 h of symptom onset at the Department of Neurology,Tianjin Huanhu Hospital,between January 2022 and October 2024.Baseline clinical data were collected,including sex,age,stroke risk factors(hypertension,diabetes,coronary heart disease,atrial fibrillation,prior stroke history,smoking,alcohol use),pre-thrombolysis blood pressure,National Institutes of Health stroke scale(NIHSS)score,serological markers(white blood cells,neutrophils,lymphocytes,blood glucose,uric acid,creatinine),completeness of the Willis circle,and thrombolysis time window(onset-to-treatment time:＜3.0h vs.3.0-4.5 h).FVH was assessed before thrombolysis using the Lee scoring system(0,1,or 2 points)based on FLAIR imaging.Inter-rater reliability for FVH scores between two radiologists was evaluated using Cohen's Kappa coefficient,with agreement levels defined as follows:almost perfect(Kappa value 0.81-1.00),substantial(Kappa value 0.61-0.80),moderate(Kappa value 0.41-0.60),fair(Kappa value 0.21-0.40),or slight/none(Kappa≤0.20).At 90 d post-thrombolysis,telephone follow-up was conducted to assess functional outcomes using the modified Rankin scale(mRS).Patients were dichotomized into favorable outcome group(mRS score 0-2)and poor outcome group(mRS score 3-6).Variables with statistical significance in univariate analysis underwent collinearity testing via variance inflation factor(VIF).Predictors with high collinearity(VIF＞10)were excluded,and the remaining variables were incorporated into a multivariate Logistic regression model to identify independent risk factors for thrombolysis prognosis.Individual clinical indicators and a combined predictive model were constructed,and their performances were evaluated using receiver operating characteristic curve analysis.The optimal cutoff values for individual and combined predictors were determined by maximizing the Youden index.Comparative analysis of individual clinical indicator and combined models were performed to assess predictive efficacy.Results Among 404 acute cerebral infarction patients treated with intravenous thrombolysis,282 achieved favorable outcomes while 122 had poor outcomes.FVH positivity was observed in 202 cases.(1)Inter-rater reliability:radiologists demonstrated substantial agreement in FVH scoring(Kappa=0.757,95%CI 0.702-0.812,P＜0.01).(2)Univariate analysis showed that significant intergroup differences were found in pre-thrombolysis laboratory values:white blood cell(9.21[7.37,10.85]×109/L vs.7.30[5.83,8.62]×109/L,Z=-7.064,P＜0.01),neutrophils(7.25[5.27,9.02]×109/L vs.4.93[3.81,6.20]×109/L,Z=-8.173,P＜0.01),lymphocytes(1.40[0.92,1.74]×109/L vs.1.57[1.16,1.99]×109/L,Z=-3.380,P=0.001),glucose(7.00[6.30,8.70]mmol/L vs.6.60[5.80,7.80]mmol/L,Z=-2.913,P=0.004),uric acid(318.50[257.50,392.00]μmol/L vs.340.50[282.00,416.25]μmol/L,Z=-2.168,P=0.030),and NIHSS score(8[5,10]vs.3[1,5],Z=-9.286,P＜0.01),FVH score(1[0,2]vs.0[0,1],Z=-4.036,P＜0.01).(3)Collinearity testing excluded white blood cell(VIF=20.561,margin of tolerance=0.049)due to high correlation with neutrophils(VIF=20.303,margin of tolerance=0.049).The rest variables with significant statistic difference in the univariate Logistic analysis were incorporated into the multivariable Logistic regression analysis,high neutrophils(OR,1.489,95%CI 1.314-1.688,P＜0.01),NIHSS score(OR,1.306,95%CI 1.213-1.407,P＜0.01),FVH score(OR,1.976,95%CI 1.398-2.794,P＜0.01)pre-thrombolysis were the independent risk factors for the poor outcome of acute cerebral infarction patients with intravenous thrombolysis treatment.(4)Predictive performance:the optimal cutoff values of FVH score,NIHSS score and neutrophils pre-thrombolysis was 1.000,5.000 and 6.586×109/L respectively,and their area under the curve(AUC)were 0.616(95%CI 0.566-0.663;55.7%sensitivity,63.1%specificity),0.789(95%CI 0.746-0.828;77.7%sensitivity,71.3%specificity),0.756(95%CI 0.711-0.797;80.9%sensitivity,59.8%specificity),respectively;the cutoff value of combined model was 0.786(AUC=0.863;95%CI 0.825-0.895,71.6%sensitivity,88.5%specificity).Conclusion The high FVH score,NIHSS score and neutrophil count pre-thrombolysis were independent predictors of poor prognosis of intravenous thrombolysis in patients with acute cerebral infarction,and the model combining the aforementioned predictors has enhanced predictive value.
