Research on a prediction model for futile recanalization after mechanical thrombectomy for acute anterior circulation large vessel occlusion based on the fusion of multimodal imaging features
10.3969/j.issn.1672-5921.2025.11.003
- VernacularTitle:基于多模态影像学特征的急性前循环大血管闭塞机械取栓后无效再通预测模型研究
- Author:
Zifeng LI
1
;
Youmeng WANG
1
;
Guofang WANG
1
;
Xinping BAI
1
;
Mingren YAO
1
Author Information
1. 236010 安徽省阜阳市人民医院神经内科
- Publication Type:Journal Article
- Keywords:
Ischemic stroke;
Perfusion imaging;
Collateral circulation;
Mechanical thrombectomy;
Futile recanalization
- From:
Chinese Journal of Cerebrovascular Diseases
2025;22(11):755-762
- CountryChina
- Language:Chinese
-
Abstract:
Objective To establish a prediction model for futile recanalization after mechanical thrombectomy(MT)in acute anterior circulation large vessel occlusion(ACLVO)stroke patients based on multimodal imaging features,and to evaluate its predictive performance.Methods Retrospectively enrolled consecutive ACLVO patients who underwent MT with successful recanalization(modified thrombolysis in cerebral infarction[mTICI]grade≥2b)at the Department of Neurology of Fuyang People's Hospital between June 2023 and December 2024.Demographic and clinical data were collected,including age,gender,hypertension,diabetes mellitus,atrial fibrillation,smoking history,alcohol consumption history,National Institutes of Health stroke scale(NIHSS)score upon admission,intravenous thrombolysis,wake-up stroke,onset-to-puncture time(OPT),puncture-to-recanalization time(PRT),occlusion vessel(internal carotid artery,middle cerebral artery),treatment method(suction thrombectomy,stent thrombectomy,suction+stent thrombectomy).All patients underwent pre-procedural CT perfusion(CTP)+CT angiography(CTA)of the head and neck.Imaging parameters included hypoperfusion(defined as time to peak＞6 s)volume(HPV),core infarct(defined as cerebral blood flow＜30%)volume(CIV),mismatch ratio(MMR;HPV/CIV),and Tan collateral score(poor collaterals:0-1 score,good collaterals:2-3 score).Patients were followed up at 90 days post-procedure via outpatient clinic or re-admission.Patients with a modified Rankin scale(mRS)score≤2 were classified into the effective recanalization group,while those with mRS score≥3 were classified into the futile recanalization group.Imaging variables with statistically significant differences between the futile recanalization and effective recanalization groups were included in multivariate Logistic regression analysis to identify independent predictors of futile recanalization and construct a nomogram model.The predictive value of the model was assessed using the receiver operating characteristic(ROC)curve.Model calibration was evaluated using the Hosmer-Lemeshow test(goodness-of-fit defined as P＞0.50).Results(1)A total of 105 ACLVO patients with successful MT recanalization were included(65 males,40 females,mean age[66±11]years,ranged 31～87 years).There were 60 patients in the effective recanalization group and 45 in the futile recanalization group.Compared to the effective recanalization group,the futile recanalization group had significantly higher age([69±11]years vs.[63±11]years,P=0.012),higher proportion of diabetes mellitus(33.33%[15/45]vs.16.67%[10/60],P=0.047),higher pre-treatment NIHSS score([15.51±2.73]vs.[13.25±2.71],P＜0.01),longer OPT([516.40±192.48]min vs.[322.98±171.22]min,P＜0.01)and PRT([94.96±17.37]min vs.[87.58±15.99]min,P=0.026),larger CIV([74.00±12.76]ml vs.[24.28±14.72]ml,P＜0.01)and HPV([121.43±22.21]ml vs.[91.62±11.34]ml,P＜0.01),smaller MMR([1.65±0.15]vs.[9.42±1.91],P＜0.01),higher 90-day mRS score([3.60±0.54]score vs.[1.22±0.83]score,P＜0.01),and a significantly different distribution of Tan collateral scores(P＜0.01).(2)Multivariate Logistic regression analysis was performed with futile recanalization as the dependent variable,identified the following independent predictors of futile recanalization:HPV(OR,2.042,95%CI 1.296-3.218,P=0.002),CIV(OR,2.373,95%CI 1.315-4.280,P=0.004),MMR(OR,1.758,95%CI 1.135-2.721,P=0.011),and Tan collateral score(OR,5.166,95%CI 2.100-12.651,P＜0.01).(3)A nomogram prediction model for futile recanalization after MT in ACLVO stroke was constructed based on the four imaging parameters as aforementioned.ROC curve analysis demonstrated that the area under the curve for this model in predicting futile recanalization after MT was 0.846(95%CI 0.739-0.912),with a sensitivity of 0.844 and a specificity of 0.817.The calibration curve and the Hosmer-Lemeshow test indicated the goodness-of-fit was high(P=0.617),and the overall stability of the model was good.Conclusion The predictive model for futile recanalization after MT for acute ACLVO constructed base on HPV,CIV,MMR and Tan collateral score facilitates the identification patients with high-risk of futile recanalization.
