自发性颅内出血患者的CT血管造影或数字减影血管造影：回顾性病例系列研究

CT angiography or digital subtraction angiography in patients with spontaneous intracranial hemorrhage： a retrospective case series study

目的探讨脑血管造影，包括cT血管造影（CT angiography，CTA）或数字减影血管造影（digital subtraction angiography, DSA）在自发性颅内出血患者中的意义。方法纳入自发性颅内出血患者，分析其人口统计学和影像学资料。根据血管造影结果分为阳性和阴性组，进一步分析自发性颅内出血的危险因素及其脑血管造影的意义。结果共纳入343例自发性颅内出血患者，年龄6～95岁，平均（56．3±14．2）岁，78例（22．7％）≤45岁，265例（77．3％）〉45岁；男性208例（60．6％），女性135例（39．4％）。在343例自发性颅内出血患者中，血管造影阳性117例（CTA或DSA任意一项阳性），其中脑动脉瘤75例（64．1％），脑血管畸形（包括动静脉畸形和动静脉瘘）22例（18．8％），烟雾病20例（17．1％）；血管造影阴性（CTA和DSA均阴性）226例。血管造影阳性组平均年龄显著性小于血管造影阴性组[（52．45±14．33）岁对（58．25±13．58）岁；t=-3．679，P=0．000]。血管造影阳性组≤45岁（46．15％对53．85％；χ^2=6．515，P=0．011）、男性（28．36％对71．63％；χ^2=7．762，P=0．005）、高血压（62．40％对83．20％；χ^2=18．321，P=0．000）、吸烟（15．40％对27．90％；χ^2=7．831，P=0．005）、饮酒（10．30％对21．20％χ^2=6．443，P=0．011）的患者构成比显著性低于血管造影阴性组。两组间出血部位存在显著性差异（χ^2=80．087，P=0．000），血管造影阳性组以蛛网膜下腔出血多见（49．60％），而血管造影阴性组以基底节出血（36．30％）多见。两组糖尿病（χ^2=0．391，P=0．532）和高脂血症（χ^2=0．387，P=0．534）患者的构成比无显著性差异。多变量logistic回归分析显示，年龄[优势比（oddsratio，OR）1．035，95％可信区间（confidence interval，CI）1．015～1．056；P=0．001]和出血部位（OR 0．690，95％CI 0．614～0．776；P=0．000）与血管造影阳性独立相关。对出血部位的分析显示，蛛网膜下腔出血的病因主要为脑动脉瘤（82．8％），其次为脑血管畸形（12．1％）和烟雾病（5．2％）；脑室出血的病因主要为烟雾病（58．3％），其次为脑动脉瘤（33．3％）。脑叶出血的病因主要为脑动脉瘤（50．0％）和脑血管畸形（45．0％）。小脑出血的病因主要为脑血管畸形（60．0％）和脑动脉瘤（40．0％）。蛛网膜下腔和脑室出血患者的血管造影阳性率分别为49．60％和10．30％，脑叶和小脑出血的血管造影阳性率分别为17．10％和4．30％，基底节和丘脑出血的血管造影阳性率分别为12．00％和5．10％，而脑干和多部位出血的血管造影阳性率极低（均为0．90％）。在343例接受CTA检查的自发性颅内出血患者中，111例经DSA或手术证实，其中CTA误诊6例，漏诊6例。CTA诊断的准确率为89．2％，敏感性为93．6％，特异性为64．7％，阳性预测值为93．6％，阴性预测值为64．7％。结论根据年龄和出血部位推测病因具有一定参考意义。在条件允许的情况下，所有自发性颅内出血患者均应行CTA检查。

Objective To investigate the significance of cerebrovascular angiography, including CT angiography （CTA） or digital subtraction angiography （DSA） in patients with spontaneous intracranial hemorrhage. Methods Three hundred forty-three patients with spontaneous intracranial hemorrhage were enrolled in the study. Their demographic and imaging data were analyzed. According to the results of angiography, the patients were divided into either a positive group or a negative group. The risk factors for spontaneous intracranial hemorrhage and theh＂ significance of cerebral vascular angiography were further analyzed. Results A total of 343 patients with spontaneous intracranial hemorrhage were enrolled. Their ages were 6 to 95 years old （mean 56. 3 ± 14.2 years）, 78 patients （22.7%） were ≤ 45 years, 265 （77.3%） were 〉 45 years, 208 （60. 6%） were males, and 135 （39.4%） were females. Of the 343 patients with spontaneous intracranlal hemorrhage, 117 were angiography positive （CTA or DSA）, in which 75 had cerebral aneurysms （64. 1% ）, 22 （18.8%） had cerebral vascular malformations （including arteriovenous malformations and arteriovenous ftstula）, and 20 （17. 1% ） had moyamoya disease. The angiography of 226 patients was negative （Both CTA and DSA were negative）. The mean age of the angiography positive group was significantly lower than that of the angiography negative group （52.45 ±14. 33 years vs. 58.25 ±13.58 years; t = -3. 679, P= 0. 000）. The proportions of the patients of ≤45 years （46. 15% vs. 53.85% ; χ^2 = 6. 515, P = 0. 011 ）, males （28.36% vs. 71.63% ; χ^2 = 7. 762, P = 0. 005）, hypertension （62.40% vs. 83.20% ;χ^2 = 18.321, P =0. 000）, smoking （15.40% vs. 27.90% ;χ^2 =7. 831, P =0. 005）, alcohol consumption （10.30% vs. 21.20% ;χ^2 =6. 443, P =0. 011） in the angiography positive group were significantly lower than those in the angiography negative group. There was significant difference in bleeding sites between the two groups （χ^2= 80. 087, P= 0. 000）. Subarachnoid hemorrhage （49.60%） were common in the angiography positive group, while basal ganglia hemorrhage （36. 30% ） were common in the negative angiography group. There were no significant difference in the proportions of patients with diabetes （χ^2 =0. 391, P = 0. 532） and hyperlipidemia （χ^2 = 0. 387, P = 0. 534） between the two groups. Multivariate logistic regression analysis showed that the age （odds ratio [ OR] 1. 035,95% confidence interval [ CI] 1.015 -1.056; P=0.001） and the bleeding site （ OR 0.690, 95% CI 0.614-0.776; P=0. 000） were independently correlated with the positive angiography. The analysis of the bleeding site showed that the cause of subarachnoid hemorrhage was mainly cerebral aneurysms （82.8%）, followed by cerebral vascular malformations （12, 1%） and moyamoya disease （5.2%）. The cause of intraventricular hemorrhage was mainly moyamoya disease （58.3%）, followed by cerebral aneurysms （33.3%）. The cause of lobar hemorrhage was mainly cerebral aneurysms （50. 0% ） and cerebral vascular malformations （45.0%）. The cause of cerebellar hemorrhage was mainly cerebral vascular malformations （60.0%） and cerebral aneurysms （40. 0%）. The positive rates of angiography in patients with subarachnoid hemorrhage and intraventricular hemorrhage were 49.60% and 10. 30%, respectively, the positive rates of angiography of lobe and cerebellar hemorrhage were 17.10% and 4. 30%, respectively, and the positive rates of angiography of basal ganglia and thalamic hemorrhage were 12.00% and 5.10%, respectively, while the positive rates of angiography of brainstem and multi-site bleeding were very low （both 0.90% ）. Of the 343 patients with spontaneous intracranial hemorrhage who received CTA examination, 111 were confn-med by DSA or surgery, in which 6 were misdiagnosed by CTA and 6 were missed diagnosis. The accuracy rate of CTA diagnosis was 89.2%, sensitivity was 93.6%, specificity was 64.7%, positive predictive value was 93.6%, and negative predictive value was 64. 7%. Conclusions According to the age and the bleeding site, speculating about the etiology has a certain reference value. Under conditions permit, all patients with spontaneous intracranial hemorrhage should undergo CTA examination.