Stroke recurrence is associated with unfavorable intracranial venous outflow in patients with symptomatic intracranial atherosclerotic large vessel severe stenosis or occlusion

Abstract

Objective: The purpose of this study was to investigate the predictive value of intracranial venous outflow for recurrent cerebral ischemic events (RCIE) in patients with symptomatic intracranial atherosclerotic large-vessel severe stenosis or occlusion (sICAS-S/O).

Methods: This retrospective study included sICAS-S/O patients with anterior circulation who underwent dynamic computed tomography angiography (dCTA) and computed tomography perfusion (CTP). Arterial collaterals were evaluated using the pial arterial filling score for dCTA data, tissue-level collaterals (TLC) were assessed using the high-perfusion intensity ratio (HIR, Tmax >10 s/Tmax >6 s), and cortical veins were evaluated using the multi-phase venous score (MVS) for the vein of Labbé (VOL), sphenoparietal sinus (SPS), and superficial cerebral middle vein (SCMV). The relationships between multi-phase venous outflow (mVO), TLC, and 1-year RCIE were analyzed.

Results: Ninety-nine patients were included, 37 of whom had unfavorable mVO (mVO-) and 62 of whom had favorable mVO (mVO+). Compared with the mVO+ patients, mVO- patients had a higher admission National Institutes of Health Stroke Scale (NIHSS) score (median, 4 [interquartile range (IQR), 0-9] vs. 1 [IQR, 0-4]; p = 0.048), larger ischemic volume (median, 74.3 [IQR, 10.1-177.9] vs. 20.9 [IQR, 5-86.4] mL; p = 0.042), and worse tissue perfusion (median, 0.04 [IQR, 0-0.17] vs. 0 [IQR, 0-0.03]; p = 0.007). Multivariate regression analysis showed that mVO- was an independent predictor of 1-year RCIE.

Conclusion: For patients with sICAS-S/O of the anterior circulation, unfavorable intracranial venous outflow is a potential imaging indicator for predicting higher 1-year RCIE risk.

Keywords: collaterals; cortical vein; intracranial atherosclerotic disease; perfusion; stenosis.

Figure 1

Flow Chart. ICA, Internal Carotid Artery; dCTA, dynamic Computed Tomography Angiography; NCCT, noncontract Cranial Computed Tomography; CTP, Computed Tomography Perfusion; mVO, Multi-phase Venous Flow; TLC, Tissue-Level Collaterals and MCA, Middle Cerebral Artery.

Figure 2

Comparison of ROC Curve. (A) The comparison of ROC Curve demonstrates the result when assessing MVS, collateral score and HIR against the recurrent ischemic events. Among these, area under the curve (AUC) of MVS is 0.920. Optimal cut-off for MVS was ≥9.5, (i.e., ≥10), which scored the highest on Youden’s index with a sensitivity of 89.7% and a specificity of 84.3%. (B) Table showed the data of ROC curve analysis.

Figure 3

Association between multi-phase venous outflow (mVO), recurrent cerebral ischemic events (RCIE) and tissue-level collateral (TLC) in symptomatic intracranial atherosclerotic large vessel severe stenosis or occlusion (sICAS-S/O). A patient with sICAD due to occlusion of the M1 segment of the middle cerebral artery (MCA; A), with recurrent ischemic stroke (C) after the original recurrent stroke (B) for 6 months, with an unfavorable mVO profile (E–M) and unfavorable TLC (N,O), and a good quality image (P). Red arrows indicate poor venous contrast opacification and green arrows indicate cortical veins with moderate or good contrast filling. In this patient, unfavorable mVO (multi-phase venous outflow, multi-phase venous score = 6; E–M) was found to be associated with the recurrence of ischemic events. Poor venous outflow showed contrast opacification of the vein of Labbé (VOL; E,H,K), superficial middle cerebral vein (SMCV; F,I,L), and sphenoparietal sinus (SPS; G,J,M) in all three phases of mCTA on the corresponding infarcted right hemisphere.