摘要目的 评价内镜超声检查术(EUS)在预测乙肝后肝硬化患者食管静脉曲张(EV)进展方面的临床应用价值.方法 采用回顾性队列研究方法,以2014年9月至2015年9月于天津市第二人民医院住院治疗的乙肝后肝硬化合并轻度EV的299例患者为研究对象,通过EUS测量并描述食管周围侧枝静脉(peri-ECV)、食管旁侧枝静脉(para-ECV)的直径和数量,以首次EUS时间为起点,随访期24个月,以EV进展或随访结束为终点,采用多因素Cox回归模型评估乙肝后肝硬化患者EV进展的风险因素,采用ROC曲线分析EUS对乙肝后肝硬化患者EV进展的预测价值.结果 随访至6个月、12个月、18个月、24个月时,EV进展的累积发生率分别为2.3％(7/299)、14.8％(44/297)、33.7％(96/285)和44.0％(120/273).多因素Cox回归分析结果显示,peri-ECV直径(P=0.0112,HR=1.3232,95％CI:1.0656～1.6429)、数量(P=0.0001,HR=1.3666,95％CI:1.1634～1.6052)及para-ECV直径(P=0.0002,HR=1.3641,95％CI:1.1558～1.6100)是EV进展的危险因素,使用核苷类似物抗乙肝病毒治疗(P=0.0020,HR=0.4969,95％CI:0.3186～0.7751)及非选择性β受体阻滞剂降门脉压治疗(P=0.0765,HR=0.5732,95％CI:0.3097～1.0611)是EV进展的保护性因素.ROC曲线分析结果显示,peri-ECV的直径[P<0.001,曲线下面积(AUC)=0.850,95％CI:0.804～0.895]、数量(P<0.001,AUC=0.831,95％CI:0.784～0.878)以及para-ECV的直径(P<0.001,AUC=0.924,95％CI:0.895～0.954)、数量(P<0.001,AUC=0.761,95％CI:0.704～0.817)对EV进展均有较好预测价值,各指标的最佳界值分别为1.85 mm、3.5条、3.35 mm、4.5条,此时预测EV进展的准确率分别为76.60％、75.19％、84.48％、70.29％.结论 EUS可用于预测乙肝后肝硬化的EV进展,peri-ECV直径>1.85 mm、数量>3.5条以及para-ECV直径>3.35 mm、数量>4.5条提示EV进展高风险.对于乙肝后肝硬化合并轻度EV的患者,选择使用核苷类似物抗乙肝病毒及非选择性β受体阻滞剂降低门脉压治疗可起到预防EV进展的作用.

abstractsObjective To assess the clinical value of endoscopic ultrasonography ( EUS ) for predicting esophageal varices ( EV ) progression in patients with hepatitis B virus ( HBV )-related hepatocirrhosis. Methods A retrospective cohort study was performed on 299 HBV-related hepatocirrhosis patients with light EV in Tianjin Second People′s Hospital admitted from September 2014 to September 2015. The diameter and number of peri-esophageal collateral veins ( ECV ) and para-ECV were measured and described by EUS. The first EUS examination time was the starting point, and the follow-up of 24 months or EV progression was the end. Risk factors of EV progression were evaluated by multivariate Cox regression model, and the predictive value of EUS for EV progression was analyzed by receiver operating characteristic ( ROC) curve. Results The cumulative incidence of EV progression was 2. 3％ ( 7/299 ) , 14. 8％( 44/297) , 33. 7％ ( 96/285) and 40. 0％ ( 120/273) at 6 months, 12 months, 18 months and 24 months of follow-up, respectively. The results of multivariate Cox regression analysis showed that the diameter of peri-ECV ( P=0. 0112, HR=1. 3232, 95％CI: 1. 0656-1. 6429 ) , the number of peri-ECV ( P=0. 0001, HR=1. 3666, 95％CI:1. 1634-1. 6052) and para-ECV diameter ( P=0. 0002, HR=1. 3641, 95％CI:1. 1558-1. 6100) were risk factors for EV progression. The use of nucleoside analogues treating HBV (P=0. 0020, HR=0. 4969, 95％CI: 0. 3186-0. 7751) and non-selective β-blockers descending portal venous pressure ( P=0. 0765, HR=0. 5732, 95％CI:0. 3097-1. 0611) were the protective factors for EV progression. The results of ROC curve analysis showed that the diameter of peri-ECV[ P<0. 001, area under the curve (AUC)= 0. 850, 95％CI: 0. 804-0. 895], the number of peri-ECV (P<0. 001, AUC=0. 831, 95％CI: 0. 784-0. 878), the diameter of para-ECV (P<0. 001, AUC=0. 924, 95％CI: 0. 895-0. 954) , and the number of para-ECV ( P<0. 001, AUC=0. 761, 95％CI: 0. 704-0. 817 ) had higher predictive value for EV progression;and the optimum cut-off values of each index were 1. 85 mm, 3. 5, 3. 35 mm, and 4. 5, respectively. The accuracies of prediction for EV progression were 76. 60％, 75. 19％, 84. 48％ and 70. 29％, respectively. Conclusion EUS can be used to predict EV progression in HBV-related hepatocirrhosis patients. Peri-ECV diameter>1. 85 mm, number>3. 5, and para-ECV diameter>3. 35 mm, number>4. 5 suggest a high risk of EV progression. For patients with HBV-related hepatocirrhosis complicated with mild EV, nucleoside analogues to anti-HBV and non-selective β-blockers to reduce portal hypertension can prevent EV progression.