水摄取率在急性小脑梗死不良预后的预测作用

The Predictive Role of Water Uptake Rate in Poor Prognosis of Acute Cerebellar Infarction

摘要: 目的：早期识别具有神经系统恶化风险的急性小脑梗死患者仍然是一个临床难题。每脑体积的净水摄取(Net Water Uptake Rate, NWU)是一种可在CT上测量的缺血性水肿的定量影像学生物标志物，但这种生物标志物目前仅应用于评价前循环卒中的患者。我们假设梗死病灶早期的NWU可以预测急性小脑梗死患者不良预后的发生。本研究的目的是比较急性小脑梗死的NWU与其他常见的预测因素。方法：连续性收集2013年3月~2022年10月在青岛大学附属医院神经内科住院的急性小脑梗死患者，按照纳入及排除标准筛选，回顾性收集基线资料。根据住院期间的最终结局，将患者分为预后良好组和预后不良组。“良好预后”被定义为经内科保守治疗，症状好转。而“不良预后”被定义为符合下列条件之一：1) 住院期间死亡；2) 美国国立卫生院卒中量表(NIHSS评分)第Ia项意识水平下降至1分或更高，且未发现其他继发恶性原因；3) 入院时或住院期间行后颅窝减压术和/或侧脑室外引流术；4) 住院期间出现神经功能恶化，需要手术治疗，因各种原因拒绝手术治疗，且出院时需要长期护理或需临终关怀患者。对所有研究对象进行出院后30天病情随访，以出院后30天内是否死亡为评判标准。使用SPSS 25.0统计软件进行分析，采用多因素Logistic回归分析与神经系统恶化风险相关的独立危险因素。基于发病24小时内颅脑CT，应用标准化程序来量化每早期梗死体积的水摄取率(NWU)，即通过密度测量(Dischemic)患侧及健侧(Dnormal)灰度值，然后，应用公式计算NWU：公式[水摄取率(%) = 1 − 患侧灰度值/健侧灰度值]，计算各层面的平均值，得到最终的水摄取率。分析两组患者的特征并绘制受试者工作特征(ROC)曲线，以比较水摄取率及其他相关危险因素(四脑室受压变形、磁共振弥散加权成像(DWI)上的梗死体积等)的价值，分析NWU识别不良预后的患者的诊断效能及最佳临界值。结果：在373例急性小脑梗死患者中，共有76例患者符合入组标准，其中预后良好组48例(63.2%)，预后不良组28例(36.8%)，其中外科手术干预了22例，住院期间临床死亡5例，出院30天内死亡6例。基于ROC曲线，水摄取率对于急性小脑梗死不良预后的预测效能较好，AUC达到0.920，显著高于其他因素(DWI上梗死面积、入院GCS评分的AUC分别为0.894、0.224)。水摄取率最佳截断值为11.50%，即当水摄取率 ≥ 11.50%时，高度怀疑不良预后的发生，敏感度为0.893，特异度为0.875。DWI上梗死体积最佳截断值为19.88 cm³，即当DWI上的梗死体积为19.88 cm³时，高度怀疑为预后不良的发生，敏感度为0.893，特异度为0.812。入院GCS评分不能预测不良预后的发生。结论：基于颅脑CT的早期梗死病灶水摄取(NWU)定量是急性小脑梗死患者发生预后不良的重要预测指标，除早期梗死体积外，早期病灶NWU的测定有助于识别神经系统恶化风险的患者，以考虑进一步的治疗措施及是否需要外科手术干预。

Abstract: Objective: Early identification of patients with acute cerebellar infarction risking neurological deterioration remains a clinical challenge. Net Water Uptake Rate (NWU) per unit of brain volume is a quantitative imaging biomarker of ischemic edema measurable on CT, but this biomarker is currently only applied to evaluating patients with anterior circulation stroke. We hypothesize that early NWU in the infarction lesion can predict the occurrence of adverse outcomes in patients with acute cerebellar infarction. The purpose of this study is to compare NWU in acute cerebellar infarction with other common predictors. Methods: Acute cerebellar infarction patients admitted to the neurology department of the Affiliated Hospital of Qingdao University from March 2013 to October 2022 were continuously collected, and screened according to inclusion and exclusion criteria, and baseline data were retrospectively collected. Patients were divided into a good prognosis group and a poor prognosis group based on the final outcome during hospitalization. “Good prognosis” was defined as improvement of symptoms after conservative medical treatment. “Poor prognosis” was defined as meeting one of the following conditions: 1) Death during hospitalization; 2) The level of consciousness on the National Institutes of Health Stroke Scale (NIHSS score) item Ia dropped to 1 point or higher, and no other secondary malignant causes were found; 3) Posterior fossa decompression and/or ventricular drainage were performed at admission or during hospitalization; 4) Neurological deterioration occurred during hospitalization, requiring surgical treatment, refusal of surgical treatment due to various reasons, and long-term care or end-of-life care was required at discharge. All subjects were followed up for 30 days after discharge, and whether death occurred within 30 days after discharge was used as the criterion for evaluation. SPSS 25.0 statistical software was used for analysis, and multifactorial Logistic regression analysis was performed to analyze independent risk factors related to the risk of neurological deterioration. Based on the brain CT within 24 hours of onset, a standardized procedure was used to quantify the water uptake rate (NWU) per unit of early infarction volume, i.e., by measuring the gray values of the affected side (Dischemic) and the healthy side (Dnormal) using density measurements. Then, the NWU was calculated using the formula: [Net Water Uptake Rate (%) = 1 − Gray value of the affected side/Gray value of the healthy side]. The average value of each slice was calculated to obtain the final water uptake rate. The characteristics of the two patient groups were analyzed, and a receiver operating characteristic (ROC) curve was plotted to compare the value of water uptake rate and other related risk factors (e.g., compression and deformation of the fourth ventricle, infarction volume on diffusion-weighted imaging (DWI)) in predicting poor prognosis. The diagnostic efficacy and optimal cut-off value of NWU in identifying patients with poor prognosis were analyzed. Results: Among 373 patients with acute cerebellar infarction, 76 patients met the inclusion criteria, including 48 patients (63.2%) in the good prognosis group and 28 patients (36.8%) in the poor prognosis group. Among them, 22 patients received surgical intervention, 5 patients died during hospitalization, and 6 patients died within 30 days after discharge. Based on the ROC curve, the water uptake rate had a good predictive efficacy for poor prognosis in acute cerebellar infarction, with an AUC of 0.920, significantly higher than other factors (AUC of infarction area on DWI and GCS score at admission were 0.894 and 0.224, respectively). The optimal cut-off value of water uptake rate was 11.50%, indicating that when the water uptake rate was ≥ 11.50%, there was a high suspicion of poor prognosis, with a sensitivity of 0.893 and a specificity of 0.875. The optimal cut-off value of infarction volume on DWI was 19.88 cm³, indicating that when the infarction volume on DWI was 19.88 cm³, there was a high suspicion of poor prognosis, with a sensitivity of 0.893 and a specificity of 0.812. The GCS score at admission could not predict the occurrence of poor prognosis. Conclusion: Quantitative measurement of early infarction lesion water uptake (NWU) based on brain CT is an important predictor of poor prognosis in patients with acute cerebellar infarction. In addition to the early infarction volume, the measurement of early lesion NWU helps identify patients with a risk of neurological deterioration, considering further treatment measures and the need for surgical intervention.