Objective:To explore the effect of antireflux valve on the prognosis of infants with biliary atresia (BA) during Kasai procedure to guide later surgical procedure.Methods:Randomized and non-randomized controlled studies were retrieved from the databases of PubMed, EMBASE, Cochrane Library, Web of Science, VIP, WanFang and CNKI until July 2021.Revman5.3 software was utilized for comparing the incidence of postoperative cholangitis, jaundice clearance rate and 2-year survival rate of autologous liver in BA infants during Kasai procedure with and without antireflux valve.Stata12.0 Begg's test was employed for examining the publication bias.Results:A total of 6 literatures were retrieved, including 1 rectangular valve (n=60) and 5 intussusception valves (n=2924). Among 2984 BA infants, antireflux valve was applied (n=670) and non-applied (n=2314). Meta-analysis indicated that inter-group statistical difference existed in the incidence of postoperative cholangitis ( OR=0.71, 95% CI: 0.59-0.86, P<0.01); no inter-group statistical difference existed in postoperative jaundice clearance rate ( OR=1.02, 95% CI: 0.85-1.22, P=0.85) or 2-year autogenous liver survival rate ( OR=1.12, 95% CI: 0.50-2.51, P=0.77). And Begg's test revealed no significant publication bias ( P=0.028-0.310). Conclusions:During Kasai procedure for BA, antireflux valve has no effect on postoperative jaundice clearance rate or 2-year survival rate of autologous liver.However, it may effectively lower the incidence of postoperative cholangitis.Thus antireflux valve is recommended during Kasai procedure.

目的 探讨胆道闭锁(biliary atresia,BA)Kasai手术后发生胆管炎的危险因素.方法 回顾性分析西安交通大学附属儿童医院2018年1月至2020年9月收治的109例因BA行Kasai手术患儿的临床资料.收集并记录性别、手术时日龄[＜60 d、60～90 d(含60 d)、≥90 d]、术前体重、心血管畸形、EB病毒(Epstein-Barr virus,EBV)感染、巨细胞病毒(cytomegalovirus,CMV)感染、术前肝脏弹性模量值、术前肝功能指标[总胆红素(total bilirubin,TBIL)、直接胆红素(direct bilirubin,DBIL)、谷草转氨酶(aspartate aminotransferase,AST)、谷丙转氨酶(alanine aminotransferase,ALT)、碱性磷酸酶(alkaline phos-phatase,ALP)、谷氨酰转肽酶(glutamyl transpeptidase,GGT)、总胆汁酸(total bile acid,TBA)、白蛋白(al-bumin,ALB)、总胆固醇(total cholesterol,TC)]、手术时长、术后禁食时长、术后住院时长、术后激素使用情况、术后6个月内黄疸消退情况以及术后胆管炎发生情况;根据术后是否出现胆管炎分为胆管炎组(n=71)和非胆管炎组(a=38);采用卡方检验或非参数检验进行单因素分析,采用Logistic回归进行多因素分析.结果 109例患儿中,7例(7/109,6.4％)术后1个月黄疸完全消退,41例(41/109,37.6％)术后3个月黄疸完全消退,53例(53/109,48.6％)术后6个月黄疸完全消退;随访期内共发生胆管炎71例(71/109,65.1％),术后3个月内发生胆管炎51例(51/109,46.8％),术后6个月内发生胆管炎60例(60/109,55.0％);早期胆管炎34例(34/71,47.9％),复发性胆管炎40例(34/71,56.3％).两组术前肝功能(TBIL、DBIL、ALT、AST、ALP、GGT、TBA、TC、ALB)比较,差异均无统计学意义(P＞0.05);两组男性占比[64.7％(33/51)比35.3％(18/51)]、术前体重[4.54(4.03,5.00)kg 比4.50(3.90,4.54)kg]、合并心血管畸形比例[63.6％(42/66)比36.4％(24/66)]、术前彩超检查肝右叶弹性模量值[12.80(10.50,15.60)kPa比 10.70(8.70,14.70)kPa]、术前 EB 病毒感染率[62.7％(47/75)比 37.3％(28/75)]、手术时长[(3.42(2.92,3.70)h比3.24(2.90,3.75)h]、术后禁食时长[3.00(3.00,3.29)d 比3.00(3.00,4.00)d]、术后激素使用率[54.8％(17/31)比 45.2％(14/31)]、住院时长[23.00(18.00,27.00)d 比 24.00(20.00,28.00)d]比较,差异均无统计学意义(P＞0.05);术前CMV感染患儿胆管炎的发生率明显低于术前无CMV感染的患儿[50％(13/26)比74.1％(40/54),x2=4.549,P=0.033],差异有统计学意义;术后6个月内黄疸消退患儿胆管炎的发生率明显低于黄疸未消退患儿[50.9％(27/53)比78.6％(44/56),x2=9.153,P=0.002],差异有统计学意义;多因素Logistic回归分析结果显示,手术时日龄(OR=2.764,95％CI:1.017～7.515,P＜0.05)及术后6 个月内黄疸是否消退(0R=0.283,95％CI:0.123～0.653,P＜0.05)是影响 BA术后胆管炎发生的危险因素.结论 手术时日龄60～90d及术后6个月内黄疸未消退是BA患儿Kasai手术后发生胆管炎的危险因素;选择合适的手术时机以及早期退黄能降低胆管炎的发生率.

Objective:To explore the diagnostic value and standard of serum gammaglutamyl transferase (GGT) and to improve the diagnostic accuracy in biliary atresia (BA) patients at different ages.Methods:Preoperative general data and liver function were reviewed for 350 children (180 boys and 170 girls) operated for cholestasis from April 2012 to July 2019. According to the findings of intraoperative cholangiography, they were divided into two groups of BA (n=244) and non-BA (n= 106). In BA group, there were 106 boys and 138 girls. The race was Han (n=239) and other ethnic groups (n=5). They were full-term (n=238) and premature (n=6). Body weight was (4.6±0.8) kg; The last preoperative detection value of gamma-glutamyl transferase (GGT) was (622.6±505.8) U/L. In non-BA group, there were 74 boys and 32 girls. The nationality was Han (n=105) and other ethnic group (n=1) . They were full-term (n=63) and premature (n=43) . Body weight was (3.7±1.1) kg. The last GGT detection value preoperatively was (181.5±136.5) U/L. According to the last preoperative GGT detection value and age, two groups were divided into 4 ages: ≤30 d, 31-60 d, 61-90 d and >90 d. In BA group, there were 6 children ≤30 d, 130 31-60 d, 86 61-90 d and 22 >90 d. In non-BA group, there were 3 children ≤30 d, 58 31-60 d, 34 61-90 d and 11 >90 d. GGT values of children of the same age in two groups were compared. According to the Jorden index, cut-off value of GGT was utilized for diagnosing BA in all ages. Sensitivity (Se) , specificity (Sp) and accuracy were compared.Results:In terms of general status, compared with non-BA group, statistically significant inter-group differences existed in gender, full-term birth and body weight at the ages of 31-60 d and 61-90 d ( P< 0.05) . In children aged >90 d, inter-group difference was statistically significant ( P<0.05 ) . GGT values were higher in BA group than those in non-BA group at different ages and regardless of age. And the differences were statistically significant ( P<0.05) . In terms of receiver operator characteristic curve (ROC) of GGT values at ≤30 d , 31-60 d, 61-90 d and >90 d, area under the curve (AUC) was 1.000, 0.874, 0.839 and 0.872. The best truncation values of GGT for diagnosing BA were 305, 236, 319 and 331 U/L respectively. AUC of all ages was 0.851 with an optimal cut-off value of 305 U/L. In terms of discriminance of using GGT value in diagnosing BA at different ages, Se, Sp and accuracy of using GGT in diagnosing BA at ≤30 d were 100% ; Se , Sp and accuracy 71.5%, 87.9% and 76.6% from 31 to 60 days ; Se , Sp and accuracy 70.0% , 85.3% and 74.2% from 61 to 90 days. Se , Sp and accuracy of >90 d 77.3%, 100% and 81.8%; Se, Sp and accuracy 68.3%, 87.7% and 74.2% respectively. Conclusions:GGT value has a certain value in the diagnosis of BA. However, great differences exist in BA children at different ages. The value of GGT value in the diagnosis of BA at 31 to 60 days is relatively high.

1临床资料 病例1,患儿,男,1岁11月,因"发现左侧阴囊包块1年"于2019年2月23日就诊于西安市儿童医院.母孕期体健,G1P1,足月顺产,出生体重3.1 kg,生长发育无特殊.查体:体温36.6℃,呼吸20次/min,脉搏94次/min,血压90/58 mmHg,体重14.5 kg;腹平坦,左下腹可触及一包块,质软,按压后见左阴囊包块增大,无压痛及反跳痛;左侧阴囊明显肿大,触质软,未触及左侧睾丸,透光试验阳性,右侧睾丸位于阴囊,大小正常.辅助检查:血常规:白细胞5.42× 109/L,中性粒细胞2.32×109/L,淋巴细胞3.01×109/L,血红蛋白123 g/L,血小板254×109/L.心电图、胸片正常,腹部CT示(图1):左下腹-左侧阴囊囊性包块,考虑淋巴管瘤.彩超示(图2):腹腔淋巴管瘤部分疝入阴囊.入院诊断:左侧腹阴囊鞘膜积液.

Objective:To establish a personalized prediction model for biliary atresia (BA) risk in cholestatic children.Methods:Retrospective analysis was conducted for clinical data of 329 cholestatic children diagnosed intraoperatively by cholangiography. There were BA (n=227) and non-BA (n=102). The parameters of gender, age, weight, ethnicity, term, white stool and liver function were collected. There were 173 boys and 156 girls. SPSS software was utilized for randomly splitting 67.2%(221/329) and 32.8%(108/329) into modeling and validating groups. Single-factor analysis was employed for screening out risk factors of two groups. Multi-factor Logistic regression model was used for analyzing the risk factors of BA. R software was utilized for establishing a nomogram model for predicting infantile BA. The effect of the model was tested by area under curve (AUC) value and fitness.Results:Single-factor analysis modeling group indicated that gender, weight, term, white stool, total bilirubin, direct bilirubin, albumin, alanine aminotransferase, aspartate aminotransferase and glutamyl transpeptidase had inter-group statistical differences ( P<0.05). Finally five factors including gender, weight, white stool, direct bilirubin and glutamyl transpeptidase were modeled by stepwise regression. The AUC of nematic diagram modeling group was 0.927 and AUC of validating group 0.951 and the discrimination was excellent. Ultimately the accuracy of the model was 86.4% in modeling group and 88.9% in validating group. Conclusions:This study has successfully established a nomogram prediction model based upon five factors of gender, weight, white stool, direct bilirubin and glutamyl transpeptidase. With a high level of predicating accuracy of BA in cholestatic infants, this model is worth promoting.