Influential Factors And Predictive Models Of Unfavorable Outcomes In Acute Ischemic Stroke With Successful Recanalization By Mechanical Thrombectomy

BackgroundAlthough MT increases the percentage of successful recanalization in patients with acute ischemic stroke,approximately 50%of patients still have poor function outcomes at 90 days after stroke.Besides,some patients develop malignant courses during the acute phase,including symptomatic hemorrhage transformation(sHT),malignant brain edema,and brain herniation.Therefore,early identification of patients at risk of unfavorable outcomes has great significance for adjusting the treatment plan after MT and communicating with patients and their families.High-density areas(HDAs)on NCCT frequently appear after successful recanalization by MT,but the relationship between HD As and unfavorable outcomes is still controversial,and there is no prediction model for unfavorable outcomes that includes HD As.ObjectiveThis study aimed to explore the relationship between HDAs on NCCT in AIS patients within 24 hours after successful recanalization by MT and unfavorable outcomes,and seek to create predictive models including HDAs and evaluate their accuracy of predicting unfavorable outcomes after successful recanalization.MethodsA retrospective analysis was conducted on 110 patients with anterior circulation stroke who achieved successful recanalization by MT between January 2016 and December 2019.Patients’ clinical characteristics and postoperative image data were collect.According to imaging data,patients were dichotomized into two groups:those with no-minor versus those with moderate-severe HD As.In this study,the unfavorable outcomes included futile recanalization and malignant courses.Futile recanalization(successful recanalization via MT,but mRS>3 at 90 days)was the main outcome,and malignant courses(the occurrence of sHT,malignant brain edema,and brain herniation within 7 days after MT)was the secondary outcome.By univariate and multivariate analysis,we explored the relationship between HD As and unfavorable outcomes and screened out variables that could independently predict unfavorable outcomes,and the predictive value of each variable was determined through the ROC curve.According to the results of multivariate analysis,the futile recanalization predictive model(FRPM)and malignant courses predictive model(MCPM)were established respectively,and evaluated their predictive ability by the area under curve(AUC),and the Delong test was used to explore whether the predictive models can improve the predictive ability of each single variable for unfavorable outcomes.The SPSS(23.0)and MedCalc(19.2.0)were used for statistical analysis.A P-value of<0.05 was considered statistically significant.ResultsIn this study,moderate-severe HD As(OR=4.632,95%CI:1.226-17.495,P=0.024),initial National Institute of Health Stroke Scale(NIHSS)score(OR=1.168,95%CI:1.034-1.320,P=0.012),moderate low-density areas(LDAs)(OR=6.231,95%CI:1.549-25.076,P=0.010),and severe LDAs(OR=12.518,95%CI:3.171-49.416,P<0.001)were independent predictors of futile recanalization.Besides,moderate-severe HDAs(OR=4.058,95%CI:1.257-13.105,P=0.019),moderate LDAs(OR=5.724,95%CI:1.139-27.475,P=0.029),and severe LDAs(OR=27.342,95%CI:6.277-117.233,P<0.001)were significant risk factors for malignant courses.FRPM including HDAs,LDAs,and initial NIHSS score,significantly improved the predictive value of LDAs(AUC:0.878 VS 0.801,P=0.001),HDAs(AUC:0.878 VS 0.719,P<0.001),and NIHSS score(AUC:0.878 VS 0.745,P=0.003)for futile recanalization.FRPM score increased by 1 point,the risk of futile recanalization increased by 4.590 times(per point:OR=4.590,95%CI:2.635-7.993,P<0.001).MCPM including HDAs and LDAs,was not significantly different from LDAs(AUC:0.874 VS 0.847,P=0.174)in predicting malignant courses,but significantly improved the predictive value of HDAs(AUC:0.874 VS 0.719,P<0.001).MCPM score increased by 1 point,the risk of malignant courses increased by 4.894 times(per point:OR=4.894,95%CI:2.591-9,245,P<0.001).ConclusionIn this study,moderate-severe HDAs was an independent predictor of futile recanalization and malignant courses in AIS patients.FRPM score based on HDAs,LDAs,and initial NIHSS score significantly improved the predictive value of a single predictor for futile recanalization.MCPM score based on HDAs,and LDAs significantly improved the predictive value of HDAs in predicting malignant courses,but was not significantly different from LDAs.