欢迎访问《中国循证儿科杂志》官方网站，今天是
English Version
中国循证儿科杂志

中国循证儿科杂志 ›› 2020, Vol. 15 ›› Issue (3): 166-176.

• 论著 • 上一篇    下一篇

基于超声、肝胆核素显像和磁共振胆胰管成像影像学检查诊断胆道闭锁准确性研究的系统评价和Meta分析

姜璟瑩1，汤悦1，朱叶1，王瑞2，孙颖华3，沈桢1，郑珊1，陈功1，张崇凡2   

  1. 复旦大学附属儿科医院 上海，201102； 1 外科，2 临床指南制作和评价中心，3 B超室
  • 收稿日期:2020-07-31 修回日期:2020-07-31 出版日期:2020-06-25 发布日期:2020-06-25
  • 通讯作者: 陈功，张崇凡

Accuracy of ultrasound，hepatobiliary scintigraphy and MRCP in the diagnosis of biliary atresia：A systematic review and meta-analysis

JIANG Jing-ying1, TANG Yue1, ZHU Ye1, WANG Rui2,  SUN Ying-hua3, SHEN Zhen1, ZHENG Shan1, CHEN Gong1, ZHANG Chong-fan2   

  1. Children's Hospital of Fudan University, Shanghai 201102, China; 1 Department of Pediatric Surgery, 2 Center for Clinical Practice Guideline Development and Evaluation, 3 Department of Ultrasound
  • Received:2020-07-31 Revised:2020-07-31 Online:2020-06-25 Published:2020-06-25
  • Contact: CHEN Gong, ZHANG Chong-fan

PDF

859

摘要： 目的明确影像学检查［超声、肝胆核素显像和磁共振胆胰管成像（MRCP）］在胆道闭锁诊断中的价值。方法根据“biliary atresia，bile duct atresia”和“胆道闭锁，胆管闭锁”建立的检索式检索英文Pubmed、Embase、Medline数据库（起于建库）和中文CBM数据库（起于2000年1月1日），止于2020年4月26日，按胆道闭锁临床实践指南的纳入、排除标准筛选文献，并使用QUADAS-2量表对纳入文献偏倚风险及临床适用性进行评价。提取文献数据，计算诊断参数。结果30篇超声文献进入Meta分析：肝门三角征（24篇文献），敏感度79%（95%CI：68%~86%），特异度97%（95%CI：94%~98%），I2＞97%，诊断比值比（DOR）=103（95%CI：52~203），SROC曲线提示曲线下面积（AUC）为0.97（95%CI：0.95~098），Deeks检测P=0.28，发表偏倚的可能较小；胆囊形态、胆囊大小、胆总管有无和肝动脉直径各有9、12、5和6篇文献，敏感度分别为73%、78%、92%和83%，特异度分别为94%、76%、76%和78%，I2均＞70%，DOR分别为42、11、12和17，SROC曲线提示AUC分别为0.94、0.81、0.92和0.87。32篇肝胆核素显像文献进入Meta分析，敏感度98%(95%CI：95%~99%)，特异度75%（95%CI：69%~81%），I2＞75%，DOR=140（95%CI：50~392），SROC曲线提示AUC为0.93（95%CI：0.91~0.95），Deeks检测P=0.05，发表偏倚的可能较小；按检查前口服肝酶诱导剂（苯巴比妥）、使用99m锝造影剂、肠道内显影和肠道或胆囊显影行亚组分析，各有18、28、18和11篇文献进入Meta分析，敏感度97%~99%，特异度75%~78%，I2均＞75%，DOR分别为239、251、246和109。7篇MRCP文献进入Meta分析，敏感度92%(95%CI：79%~97%)，特异度82%（95%CI：63%~92%），I2为75%和83%，DOR=52（95%CI：7~382），SROC曲线提示AUC为0.94（95%CI：0.91~0.96）。结论超声探测到肝门三角征较其他超声特征（胆囊形态、肝动脉直径、胆总管有无和胆囊大小）对诊断胆道闭锁有明确的优势；无论是检查前口服肝酶诱导剂、采用99m锝造影剂、肠道内显影或胆囊显影, 肝胆核素显像对胆道闭锁误诊率仍很高；MRCP对诊断胆道闭锁的结果不稳健。

Abstract: ObjectiveTo clarify the value of ultrasound, hepatobiliary scintigraphy and magnetic resonance cholangiopancreatography(MRCP) in the diagnosis of biliary atresia. MethodsAccording to search terms of biliary atresia and bile duct atresia, English literature was searched in Pubmed，Embase，Medline from the establishment of databases to April 26, 2020 and Chinese literature was searched in CBM from January 1, 2000 to April 26, 2020. Retrieved articles were screened and selected according to the inclusion and exclusion criteria. The QUADAS-2 was used to evaluate the risk of bias and applicability concerns. Data were extracted and diagnostic parameters were calculated for the combined data. ResultsThirty studies on ultrasound were included. The sensitivity and specificity of triangular cord sign (24 articles involved) were 79% (95%CI: 68%-86%) and 97% (95%CI: 94%-98%) respectively with I2>97%, and the diagnostic odds ratio (DOR) was 103 (95%CI: 52-203). The SROC curve suggested that the AUC was 0.97 (95%CI: 0.95-0.98), and the funnel plot indicated little publication bias (P=0.28). The sensitivity of abnormal gallbladder morphology (9 articles involved), gallbladder size (12 articles involved), absence of common bile duct (5 articles involved) and hepatic artery diameter (6 articles involved) was 73%, 78%, 92% and 83%, respectively and the specificity was 94%, 76%, 76% and 78% respectively with I2>70%, and DOR was 42, 11,12, and 17, respectively. The SROC curve suggested that AUC was 0.94, 0.81, 0.92 and 0.87, respectively. Thirty-two studies on hepatobiliary scintigraphy were included. The sensitivity and specificity were 98%(95%CI: 95%-99%) and 75%（95%CI: 69%-81%）respectively with I2＞75% and DOR=140（95%CI: 50-392）. The SROC curve suggested that the AUC was 0.93 (95%CI: 0.91-0.95), and the funnel plot indicated little publication bias (P=0.05). In subgroup analysis of liver enzyme inducer before examination (18 articles involved), technetium contrast agent (28 articles involved), detection of contract in intestine only (18 articles involved), and detection of contract in intestine or bile duct (11 articles involved), the sensitivity ranged from 97% to 99%, and the specificity ranged from 75% to 78% with I2>75%, and DOR was 239, 251, 246 and 109, respectively. Seven studies on MRCP were included. The sensitivity and specificity were 92% (95%CI: 79%-97%) and 82% (95%CI: 63%-92%) respectively with I2 of 75% and 83%, and the DOR was 52 (95%CI: 7-382). The SROC curve suggested that the AUC was 0.94 (95%CI: 0.91-0.96). ConclusionCompared with other ultrasonic features (abnormal gallbladder morphology, gallbladder size, absence of common bile duct and hepatic artery diameter), the triangular cord sign detected by ultrasound is more accurate in the diagnosis of biliary atresia. The misdiagnosis rate of hepatobiliary scintigraphy is still high by liver enzyme inducer before examination, technetium contrast agent, detection of contract in intestine or bile duct. MRCP is not stable in terms of diagnosis accuracy of biliary atresia.