Quantitative contrast enhancement volume on immediate post-thrombectomy CT predicts symptomatic intracranial hemorrhage and functional outcomes in acute large vessel occlusion stroke

Abstract

Objectives: Symptomatic intracranial hemorrhage (sICH) following endovascular thrombectomy (EVT) for acute ischemic stroke due to anterior circulation large vessel occlusion (AIS-LVO) significantly impacts clinical outcomes. Contrast enhancement (CE) on immediate post-EVT non-contrast CT (NCCT) may reflect blood-brain barrier disruption, but its volumetric correlation with sICH and functional independence remains underexplored.

Methods: We performed a retrospective screening on consecutive AIS-LVO patients who had CE on NCCT immediately within 2 h after EVT. The quantitative volume of CE was calculated by using 3D Slicer software. Multivariable logistic regression was performed to achieve the risk factors of sICH and functional independence. The discrimination and calibration of the multivariable models were assessed using the area under the receiver operator characteristic curve, fivefold cross-validation, calibration curve, and decision curve analysis.

Results: In this study, 111 patients were enrolled in the final analysis. According to the restricted cubic spline, 10.6 mL was the optimal threshold of CE volume dichotomization for patients with AIS-LVO. In multivariate regression analysis, the CE₊ group (CE volume beyond 10.6 mL) was significantly associated with sICH (aOR: 5.24, 95% CI: 1.45-18.99, p = 0.012) and functional independence (10.9% vs. 51.8%; aOR 0.05, 95% CI: 0.01-0.28, p < 0.001). The multivariable models demonstrated good discrimination and calibration in this cohort, as well as the fivefold cross-validation.

Conclusion: Volumetric quantification of CE on immediate post-EVT NCCT serves as a novel biomarker for early sICH risk stratification and functional prognosis in AIS-LVO. Incorporating CE volume into predictive models enhances clinical utility, enabling timely diagnosis and intervention.

Keywords: acute ischemic stroke; contrast enhancement; endovascular treatment; functional independence; symptomatic intracranial hemorrhage.

Figure 1

An 82-year-old patient presented with a right-sided cerebral ischemic stroke. (A,B) Non-contrast CT immediately after EVT shows contrast enhancement (CE). (C) The three-dimensional structure of CE in the brain is reconstructed with a volume of 24.9 mL.

Figure 2

Restricted cubic spline for contrast enhancement (CE) volume and the risk of functional independence (90-day mRS 0–2). A nonlinear relationship between the contrast enhancement volume and risk of functional outcomes was observed and CE produced a significant risk for functional independence as the volume of CE exceeded 10.6 mL.

Figure 3

Receiver operating characteristic curve for the prediction of symptomatic intracranial hemorrhage (A) and functional independence (B). sICH model 2, consisting of conventional variables, is displayed with AUC of 0.771 (95% CI 0.673 to 0.868; P<0.05); sICH model 1, including conventional variables plus CE volume, shows an AUC of 0.870 (95% CI 0.784 to 0.957; p<0.05), significantly higher than sICH model 2 (the difference value of AUC 0.100, p = 0.01); Functional independence model 2, consisting of conventional variables, is displayed with AUC of 0.865 (95% CI 0.795 to 0.936; p<0.05); Functional independence model 1, including conventional variables plus CE volume, shows an AUC of 0.927 (95% CI 0.878 to 0.977; p<0.05), significantly higher than Functional independence model 2 (the difference value of AUC 0.062, p = 0.02). The conventional variables include age, pre-stroke mRS, baseline NIHSS score, baseline ASPECTS, ASITN/SIR score, eTICI 2b-3, and occlusion site. sICH, symptomatic intracranial hemorrhage; AUC, area under the curve; CI, confidence interval.

Figure 4

Receiver operating characteristic curve for the prediction and symptomatic intracranial hemorrhage (A) and functional independence (B). sICH model 1, including conventional variables plus CE volume, shows an AUC of 0.870 (95% CI 0.784 to 0.957; p<0.05); sICH model 1 based on fivefold cross validation with AUC of 0.758 (95% CI 0.526 to 0.896; p < 0.05). Functional independence model 1, including conventional variables plus CE volume, shows an AUC of 0.927 (95% CI 0.878 to 0.977; p<0.05); Functional independence model 1 based on fivefold cross validation with AUC of 0.862 (95% CI 0.830 to 0.933; p < 0.05). The conventional variables include age, pre-stroke mRS, baseline NIHSS score, baseline ASPECTS, ASITN/SIR score, eTICI 2b-3, and occlusion site. sICH, symptomatic intracranial hemorrhage; AUC, area under the curve; CI, confidence interval.