Beyond collaterals: brain frailty additionally improves prediction of clinical outcome in acute ischemic stroke

• Published in: European radiology 2022-10, Vol.32 (10), p.6943-6952
• Main Authors: Zhou, Jia-Ying, Shi, Yi-Bin, Xia, Cong, Lu, Chun-Qiang, Tang, Tian-Yu, Lu, Tong, Huang, Shan, Wang, Yuan-Cheng, Han, Xiang-Qian, Ju, Sheng-Hong
• Format: Article
• Language: English
• Subjects: Angiography; Atrophy; Brain; Clinical outcomes; Computed tomography; Diagnostic Radiology; Frailty; Head and Neck; Imaging; Internal Medicine; Interventional Radiology; Ischemia; Leukoaraiosis; Markers; Medical prognosis; Medicine; Medicine & Public Health; Neuroradiology; Patients; Radiology; Risk analysis; Risk factors; Stroke; Ultrasound; Vascular diseases
• ISSN: 1432-1084; 0938-7994; 1432-1084
• DOI: 10.1007/s00330-022-08792-6

Objectives We aimed to investigate the additional significance of cerebral small vessel disease (SVD) beyond collaterals in determining the clinical outcome after acute ischemic stroke (AIS). Methods We retrospectively reviewed large vessel-involved stroke patients who had baseline CTA within 24 h after symptom onset and had an MRI scan 5 days after admission from October 1, 2018, to October 31, 2021. Collaterals and SVD markers (including atrophy, leukoaraiosis, lacunes, and perivascular space) were graded on CT angiography and MR images, respectively. Modified Rankin Scale (mRS) score at 90 days was recorded, and mRS ≤ 2 was regarded as a good clinical outcome. The associations between SVD markers, collaterals, and mRS were analyzed using logistic and causal mediation regression. Results We finally enrolled 119 patients (70 ± 13 years). The multivariable regression showed atrophy (evidence: OR 0.05 [95% CI 0.01–0.31], p = 0.002; severe: OR 0.08 [95% CI 0.01–0.44], p = 0.007) and evidence of lacune (OR 0.30 [95% CI 0.08–0.96], p = 0.049) were associated with poor clinical outcomes after correcting covariables. Collaterals mediated 25.74% of the effect of atrophy on poor clinical outcomes ( p < 0.001), while lacune impacted clinical outcomes without collaterals’ mediation effect ( p = 0.54). The classification model with atrophy and lacune had a significantly higher AUC than without markers to distinguish good and poor outcomes ( p = 0.036). Conclusions Beyond collaterals, brain frailty, specifically assessed by atrophy and lacune, was essential in evaluating stroke patients and could additionally improve the stroke outcome prediction. Key Points • Beyond collaterals, brain frailty, specifically assessed by brain atrophy and lacune, was still an independent risk factor of unfavorable clinical outcomes after AIS. • Adding brain atrophy and lacune into the model has an extra benefit in predicting stroke outcomes. • The effect of atrophy on stroke outcomes was proportionally mediated through collaterals, but about three-quarters of the effect of brain atrophy and the total effect of lacune directly impacted stroke outcomes without a mediation effect of collaterals.