Department of Radiology, The Second Affiliated Hospital of Nantong University, Nantong, Jiangsu 226001, China.
Department of Radiology, The Second Affiliated Hospital of Nantong University, Nantong, Jiangsu 226001, China.
J Stroke Cerebrovasc Dis. 2022 Sep;31(9):106624. doi: 10.1016/j.jstrokecerebrovasdis.2022.106624. Epub 2022 Aug 4.
In the current study we aim the identification of the culprit plaque characteristics of intracranial arteries using high-resolution magnetic resonance vessel wall imaging (HR-MR-VWI). Moreover, we target the evaluation of the predictive value of culprit plaque characteristics for short-term stroke recurrence combined with ESRS.
We conducted a prospective cohort study on 342 patients diagnosed with symptomatic intracranial atherosclerotic stenosis (sICAS), out of which 243 were men and 99 were women with an average age of 64 ± 12 years. 184 cases of anterior circulation ischemia (ACIS) and 158 cases of posterior circulation ischemia (PCIS) were included in the study. All of them underwent HR-MR-VWI during the period between February 2020 and June 2021 in the Second Affiliated Hospital of Nantong University, Nantong, China. The culprit vessel and culprit plaque characteristics were assessed based on HR-MR-VWI images, and the patients' ESRS were obtained from the electronic medical records of the hospital. Concerning the obtained results from the 6-month follow-up, the patients were divided into the non-recurrence group and the recurrence group, and the differences in the above-mentioned features between the two groups were compared. The univariate Cox regression analysis combined with ESRS was performed to screen out the independent risk factors associated with recurrent stroke with P < 0.1. Receiver operating characteristic curves (ROC curves) were plotted to analyze the predictive performance of the culprit plaque characteristics, ESRS and combined variables for stroke recurrence. We used the area under the curve (AUC) ROC, while the sensitivity and specificity were calculated at the optimal threshold. The Delong test was employed to compare the quality of the AUC of the predictors.
A total of 15.5% (53/342) of patients had a stroke recurrence within six months, with statistically significant differences (P < 0.05) between the two groups regarding the ESRS, medical history of diabetes mellitus, myocardial infarction, data for previous acute ischemic stroke (AIS) or transient ischemic attack(TIA), history of peripheral vascular disease, and serum brain natriuretic peptide level. In the patients with ACIS, the incidence of hyperintensity on the T1-weighted imaging (T1WI) was significantly different between the recurrence and the non-recurrence groups (P < 0.05). In the patients with PCIS, statistically significant differences between the recurrence and the non-recurrence group were detected in the culprit plaque burden, degree of enhancement, and incidence of hyperintensity on T1WI (P < 0.05). The ESRS (hazard ratios [HR], 1.598, 95% confidence interval [CI], 1.193-2.141, P = 0.002) ,degree of enhancement (HR = 1.764, 95% CI 0.985-3.087, P = 0.047) and hyperintensity on T1WI (HR = 2.745, 95% CI 1.373-5.488, P = 0.004) proved to be independent risk factors for stroke recurrence. The ESRS predicted stroke recurrence with AUC = 0.618 (95% CI 0.564-0.670), while the best cut-off value was 2 points. Furthermore, the registered sensitivity and specificity were 60.4% and 58.5%, respectively. Regarding the degree of enhancement in the culprit plaque, the prediction of stroke recurrence was with AUC = 0.628 (95% CI 0.574-0.679) as well as with sensitivities and specificities of 58.5% and 64.4%, respectively. Regarding the hyperintensity on T1WI in culprit plaque, the prediction of stroke recurrence was with AUC = 0.678 (95% CI 0.626-0.727) as well as with sensitivities and specificities of 66.0% and 70.0%, respectively. The ESRS combined with the degree of enhancement predicted stroke recurrence with an AUC = 0.685 (95CI% 0.633-0.734), while the recorded sensitivity and specificity were 56.6% and 73.4%, respectively. The ESRS combined with hyperintensity on the T1WI predicted stroke recurrence with an AUC = 0.745 (95CI% 0.696-0.791). The recorded sensitivity and specificity were 64.2% and 76.8%, respectively. The AUC quality of the ESRS combined with hyperintensity on T1WI was higher than that of other indices (P < 0.05).
The hyperintensity on T1WI of the culprit plaque in intracranial arteries combined with ESRS demonstrated better predictive ability for short-term stroke recurrence. We consider this of high importance for clinical application since it provides an easier way of obtaining data for precise diagnosis.
在目前的研究中，我们旨在使用高分辨率磁共振血管壁成像（HR-MR-VWI）识别颅内动脉的罪犯斑块特征。此外，我们的目标是评估罪犯斑块特征与 ESRS 相结合对短期卒中复发的预测价值。
我们对 342 例症状性颅内动脉粥样硬化性狭窄（sICAS）患者进行了前瞻性队列研究，其中 243 例为男性，99 例为女性，平均年龄为 64±12 岁。研究包括 184 例前循环缺血（ACIS）和 158 例后循环缺血（PCIS）。所有患者均于 2020 年 2 月至 2021 年 6 月在南通大学第二附属医院接受 HR-MR-VWI 检查。根据 HR-MR-VWI 图像评估罪犯血管和罪犯斑块特征，并从医院的电子病历中获取患者的 ESRS。根据 6 个月的随访结果，将患者分为无复发组和复发组，比较两组之间上述特征的差异。采用单变量 Cox 回归分析结合 ESRS 筛选与卒中复发相关的独立风险因素，P<0.1。绘制受试者工作特征曲线（ROC 曲线），分析罪犯斑块特征、ESRS 和联合变量对卒中复发的预测性能。我们使用曲线下面积（AUC）ROC，同时在最佳阈值处计算灵敏度和特异性。采用 Delong 检验比较预测指标的 AUC 质量。
342 例患者中有 15.5%（53/342）在 6 个月内复发卒中，两组之间的 ESRS、糖尿病病史、心肌梗死病史、既往急性缺血性卒中（AIS）或短暂性脑缺血发作（TIA）病史、外周血管疾病病史和血清脑利钠肽水平差异有统计学意义（P<0.05）。在 ACIS 患者中，复发组和无复发组 T1 加权成像（T1WI）上的高信号发生率差异有统计学意义（P<0.05）。在 PCIS 患者中，复发组和无复发组之间的罪犯斑块负荷、强化程度和 T1WI 上高信号发生率差异有统计学意义（P<0.05）。ESRS（危险比 [HR]，1.598，95%置信区间 [CI]，1.193-2.141，P=0.002）、强化程度（HR=1.764，95%CI 0.985-3.087，P=0.047）和 T1WI 上的高信号（HR=2.745，95%CI 1.373-5.488，P=0.004）被证明是卒中复发的独立危险因素。ESRS 预测卒中复发的 AUC 为 0.618（95%CI 0.564-0.670），最佳截断值为 2 分。此外，登记的灵敏度和特异性分别为 60.4%和 58.5%。关于罪犯斑块的强化程度，卒中复发的预测 AUC 为 0.628（95%CI 0.574-0.679），灵敏度和特异性分别为 58.5%和 64.4%。关于罪犯斑块 T1WI 上的高信号，卒中复发的预测 AUC 为 0.678（95%CI 0.626-0.727），灵敏度和特异性分别为 66.0%和 70.0%。ESRS 联合强化程度预测卒中复发的 AUC 为 0.685（95%CI 0.633-0.734），记录的灵敏度和特异性分别为 56.6%和 73.4%。ESRS 联合 T1WI 上的高信号预测卒中复发的 AUC 为 0.745（95%CI 0.696-0.791）。记录的灵敏度和特异性分别为 64.2%和 76.8%。ESRS 联合 T1WI 上的高信号的 AUC 质量优于其他指标（P<0.05）。
颅内动脉罪犯斑块的 T1WI 高信号与 ESRS 联合显示出对短期卒中复发的更好预测能力。我们认为这对于临床应用非常重要，因为它提供了一种更简单的方法来获取数据进行精确诊断。
