Endotoxemia after relief of biliary obstruction by internal and external drainage in rats

Systemic and portal endotoxemia were studied in rats with biliary obstruction and after relief of the obstruction by internal and external drainage. Endotoxemia was increased after bile duct ligation (p less than 0.001) compared with control values. The incidence of systemic and portal endotoxemia was significantly reduced after internal drainage (p less than 0.001). A significantly higher incidence of portal (86 percent) and systemic (57 percent) endotoxemia, however, was found after external drainage. The persistence of endotoxemia after external drainage, when serum bilirubin levels returned to normal units, indicates that bile flow is important in controlling endotoxemia during preoperative biliary drainage. These results suggest that the systemic endotoxemia observed after relief of obstruction by external drainage may contribute to the increased mortality, as found in previous rat studies. This observation may contribute to an understanding of why patients with preoperative external drainage of biliary obstruction have a higher incidence of septic complications.     

Characterization of the Kupffer cell response to exogenous endotoxin in a rodent model of obstructive jaundice.

BACKGROUND: Sepsis and endotoxaemia occur frequently in biliary obstruction. Impaired Kupffer cell endocytosis is implicated in these events. Tumour necrosis factor and interleukin 6, secreted by Kupffer cells, are important mediators of sepsis. Kupffer cell clearance of endotoxin and secretion of cytokines in experimental obstructive jaundice were investigated. METHODS: Wistar rats were randomized to bile duct ligation, sham operation or control. Groups (n = 8) were studied 1 and 3 weeks after operation. Kupffer cell function was assessed using in situ hepatic perfusion. RESULTS: Clearance of endotoxin was significantly depressed 1 week (median (interquartile range) 20.3 (10.5-27.1) per cent) and 3 weeks (22.1 (20.2-23.2) per cent) after bile duct ligation compared with that in respective sham animals (35.5 (29.9-41.6) and 40.9 (37.7-47.0) per cent) and controls (39.5 (37.3-46.8) per cent). Secretion of tumour necrosis factor was significantly greater 1 week (1113.7 (706.5-1436. 8) pg/ml) and 3 weeks (1118.2 (775.7-1484.1) pg/ml) following bile duct ligation compared with that in respective sham animals (114.3 (0-178.5) and 107.6 (63.7-166.4) pg/ml) and controls (0 (0-20.7) pg/ml). Interleukin 6 was not secreted by sham or control animals but was present in the perfusate from jaundiced animals at 1 and 3 weeks (52.5 (9.9-89.5) and 66.2 (60.2-193.1) pg/ml). CONCLUSION: These data demonstrate simultaneous impairment of Kupffer cell clearance of endotoxin and increased secretion of proinflammatory cytokines in experimental obstructive jaundice. These diverse responses may contribute to the development of sepsis-related complications in biliary obstruction.     

Absence of intestinal bile promotes bacterial translocation.

Previously, the authors documented that extrahepatic biliary obstruction promotes the systemic translocation of bacteria from the intestine to visceral tissues. The current experiments were performed to determine whether it was the absence of intestinal bile or the presence of biliary obstruction that promoted bacterial translocation. Four groups of rats were studied: 1) nonoperated controls (n = 20), sham common bile duct-ligated (n = 22), common bile duct-ligated (n = 25), and common bile duct-diverted (choledochovesical bypass) (n = 23). The sham-ligated group underwent laparotomy and manipulation of the portal region; whereas the ligated group had their common bile ducts ligated, while the choledochovesical group had a silastic tube placed from the common bile duct to the bladder. Seven days later, at death, the incidence of bacterial translocation was higher in the groups of rats subjected to common bile duct ligation (41%) or diversion (32%) than in the control (3%) or sham-ligated (5%) groups (P less than 0.05). Histologic sections of ileums of ligated and diverted animals both showed subepithelial edema. These findings suggest that it is primarily the absence of bile in the intestine that promotes mucosal injury and bacterial translocation and not biliary obstruction.     

Effect of internal biliary drainage on plasma levels of endotoxin,cytokines, and C-reactive protein in patients with obstructive jaundice

Preoperative biliary drainage may improve the cytokine and acute-phase response derangements observed in patients with obstructive jaundice. We conducted a prospective longitudinal, before-after trial in our 600-bed teaching hospital. Twenty-four patients with obstructive jaundice were investigated, 11 with benign obstruction and 13 with malignant disease. Endoscopic internal biliary drainage was performed in all patients (7 by papillotomy and 17 by endoprostheses). Endotoxin, tumor necrosis factor alpha (TNF-a), interleukin-6 (IL-6), nitric oxide production, and C-reactive protein (CRP) were determined at admission and on days 2 and 7 after internal biliary drainage was accomplished. Bile cultures were obtained before and at the time of drainage. Endotoxin, IL-6, TNF-a, and CRP were significantly higher in patients with cancer. After internal drainage, endotoxin (11.4 vs. 2 EU/L; p <0.05), TNF-a (87.5 vs. 48 pg/ml; p = 0.03), and IL-6 (324 vs. 232 pg/ml; p <0.05) plasma levels decreased significantly in the early postdrainage period in patients with cancer. Endotoxin, cytokines, as well as the CRP plasma values, however, increased again on day 7 after drainage. This trend was less marked in patients with benign obstruction. Patients with positive bile cultures after drainage displayed higher levels of CRP (115 vs. 62 mg/L; p = 0.03), IL-6 (598 vs. 330 pg/ml; p = 0.04), and endotoxin (10.6 vs. 4.8 EU/L; p = 0.02) than those with negative bile cultures. Biliary tract obstruction is associated with an increase in endotoxin levels, a positive acute-phase response, and plasma cytokine elevation. After biliary drainage a transitory improvement of these alterations was observed, although values remained high 1 week postdrainage. These findings were associated with positive bile cultures.     

阻塞性黄疸大鼠术前胆道内、外引流对肝细胞再生能力的影响

目的　探讨阻塞性黄疸大鼠肝叶切除术前胆道内、外引流对肝细胞再生能力的影响和机制。方法　将大鼠胆总管结扎 5d后 ,分别行胆道内、外引流 5d ,再行 70 %肝叶切除术。结果　胆道外引流组与内引流组和对照组大鼠相比 ,反映肝细胞再生能力的肝细胞核DNA含量、增殖细胞核抗原 (proliferatingcellnuclearantigen ,PCNA)指数、有丝核分裂指数 (mitoticindexMI)明显减低 (P <0 0 5 )。胆道外引流组肝细胞C met/HGF R基因表达也减低 (P <0 0 1)。结论　阻塞性黄疸大鼠肝叶切除术前胆道外引流对肝细胞再生能力有明显抑制作用 ;胆道外引流组肝细胞C met/HGF R基因表达减弱可能是该组肝细胞再生能力下降的重要因素。     

Endogenous endotoxemia in patients with liver cirrhosis--a quantitative analysis of endotoxin in portal and peripheral blood

In order to investigate the mechanism of endogenous endotoxemia (that is, endotoxemia observed in the absence of infection) in patients with liver cirrhosis, the concentration of endotoxin in the portal (PO-Et) and peripheral blood (PE-Et) from fifty three patients undergoing abdominal surgery was simultaneously measured by a quantitative endotoxin assay. The PE-Et of the patients with liver cirrhosis (19.8 +/- 20.2 pg/ml, n = 23) was significantly elevated, when compared with that of the patients without liver cirrhosis (9.2 +/- 5.1 pg/ml, n = 30), and was close to the normal range of PE-Et obtained from thirty healthy volunteers (7.2 +/- 4.1 pg/ml, n = 30). The PO-Et was also higher in the patients with liver cirrhosis than in the patients without liver cirrhosis. Moreover, PO-Et was significantly higher than PE-Et in all the patients (p less than 0.05). The per cent difference in the endotoxin concentration between the portal and peripheral blood (percentage of delta Et) was significantly decreased in the cirrhotic patients, especially in those with esophageal varices, which was well correlated with the phagocytic activity of the reticuloendothelial system (RES) determined by the clearance of iron colloid. The endogenous endotoxemia is thus likely to be due to the impaired clearance of endogenous endotoxin in portal blood, resulting from both the decreased phagocytic activity of RES in the liver and the coexisting porta-systemic bypass.     

Experimental study on the development of endotoxemia in biliary tract infection with special reference to lymphatic pathway at controlled biliary duct pressure

A study was made on the pathway of the endotoxin from the biliary tract into the circulation, using adult mongrel dogs divided into 10 groups primarily with or without thoracic duct drainage. The biliary duct pressure level of 25, 35, and 55 cm saline was maintained by infusing physiological saline into the bile duct. Each level was maintained for a 10-minute period following the intrabiliary infusion of endotoxin 10 mcg/kg. The endotoxin level in arterial and portal blood and in thoracic lymph increased with increment of biliary duct pressure. The endotoxin level in the thoracic lymph was more than 10 times higher than that in blood. At 35 and 55 cm saline of biliary duct pressure, the groups with or without thoracic lymph drainage did not show any significant difference in blood endotoxin level, or moreover, endotoxin shock developed in each group. These findings imply that at high biliary pressure such as 35 or 55 cm saline, endotoxin in the bile duct directly appears in the sinusoid. At low biliary pressure such as 25 cm saline, endotoxin was detected only in thoracic lymph and endotoxemia did not develop.     

Bacterial translocation in obstructive jaundice in rats: role of mucosal lacteals.

OBJECTIVE: To study the ultrastructure of the ileal wall in rats with obstructive jaundice alone and after passive external biliary drainage to see if we could discover the reason for the increased risk of infective complications and multisystem failure in the presence of obstructive jaundice and after external biliary drainage. DESIGN: Histological examination of the wall of the terminal ileum using light microscopy, as well as scanning and transmission electron microscopy (EM). SETTING: Experimental laboratory, Republic of Georgia. ANIMALS: 56 adult male Wistar rats. Interventions: Rats were divided into 7 groups: controls (not operated on, n = 6); sham-operated and studied after 6 and 12 days (n = 6 in each); bile duct ligation alone studied after 6 and 12 days (n = 10 in each); and bile duct ligation followed 6 and 12 days later by one-day of external biliary drainage (n = 9 in each). MAIN OUTCOME MEASURES: Percentage of destroyed villi. RESULTS: The extent of oedema of villi, necrosis of neurons, and disturbances in the secretory capacity of enterocytes correlated well with the duration of cholestasis. After 6 and 12 days ligation alone 7.2% and 17.3%, respectively, of villi were found to be destroyed; their connective tissue framework including lymphatics was in direct contact with the intestinal contents. The changes were not reversed by one day of external biliary drainage. CONCLUSION: The gaps in the ileal mucosa caused by obstructive jaundice (and not relieved by one day of external biliary drainage) may enable gut bacteria and their endotoxin to reach the systemic circulation through the lymphatic-system. This could increase the risk of infective complications.     

Endotoxemia during percutaneous manipulation of the obstructed biliary tree

Fourteen patients undergoing percutaneous biliary manipulation were studied on 21 occasions. Using a chromogenic limulus-based assay technique, peripheral endotoxin concentration prior to the procedure was found to be 6.1 +/- 9 pg/ml, increasing to 30.7 +/- 26 pg/ml after the procedure (p less than 0.001). Fourteen of 21 bile cultures and only 1 of 21 blood cultures drawn during the procedure were positive for bacterial growth. There was a clinically modest but statistically significant decrease in blood pressure (p less than 0.05), an increase in temperature (p less than 0.01), and an increase in creatinine concentration (p less than 0.05) before and after biliary manipulation. The presence of infected bile and the preprocedure bilirubin level predicted development of endotoxemia.     

The effect of preoperative internal and external biliary drainage on mortality of jaundiced rats

Mortality following abdominal infection induced by cecal ligation and puncture was studied in rats with obstructive jaundice and after relief of the obstruction by preoperative internal or external biliary drainage. Four groups of adult Sprague-Dawley rats were used: common bile duct ligation (BDL), BDL followed by internal drainage after two weeks, BDL followed by external drainage after two weeks, and sham operation. The serum bilirubin concentration was significantly increased in the BDL group and returned to normal following internal and external drainage. Mortality in the sham group was 16.5% and increased to 83.3% following BDL, but it decreased significantly (25%) to near-normal values after preoperative internal drainage. However, after external drainage no significant reduction in mortality (63%) was found. These findings confirm the poor results of preoperative external drainage in previous studies and establish the efficacy of internal biliary drainage in a well-controlled experimental model that has to be evaluated in the clinical situation.     

梗阻性黄疸-选择性胆管外引流大鼠模型的建立

目的建立梗阻性黄疸-选择性胆管外引流（约占30%肝脏体积）动物模型,了解其对胆汁分泌及肝功能改善的影响,为肝门部胆管癌减黄实验及临床研究提供参考。方法采用结扎切断SD大鼠左中叶肝胆管及胆总管预制外引流管10d的方法,制作大鼠选择性胆管外引流模型,在引流的0、1、4、7、10、14d收集胆汁及检测肝功。结果成功建立了大鼠完全梗阻性黄疸部分肝脏外引流的模型并观测到其对胆汁分泌及肝功能改善的影响。结论与梗阻性黄疸全肝外引流模型相比,梗阻性黄疸约30%肝脏外引流能够改善肝功能,而且可以增加预保留肝的功能代偿。     

Impaired specific cell-mediated immunity in experimental biliary obstruction and its reversibility by internal biliary drainage

Little is known of the effect of cholestasis on host immunity. This study evaluates lymphocytic responsiveness to PHA and LPS mitogen and to allogeneic F344 antigen in Sprague-Dawley rats 21 days following bile duct ligation and 31 days following relief of jaundice by internal biliary drainage. Serum bilirubin level was significantly elevated in the bile duct ligated animals at Day 21 (P less than 0.001) and thereafter returned to preoperative levels following internal biliary drainage. Results demonstrate depressed responsiveness to PHA (P less than 0.001) and allogeneic F344 antigen in vivo (P less than 0.04) and in vitro (P less than 0.02) in bile duct ligated animals as compared to sham, sham pair-fed, and normal control rats. The observed deficiency in responsiveness to T-cell-dependent mitogen and antigen cannot be explained on the basis of complicating nutritional, renal, or infective factors. Subsequent internal biliary drainage results in some improvement in T-cell responsiveness in the bile duct ligated group although recovery is not complete. B-Lymphocytic response to LPS mitogen is not affected by bile duct ligation. We conclude that cholestasis subsequent to extrahepatic biliary obstruction per se results in impairment of cell-mediated immunity in vivo. This impairment is partly reversible by internal biliary drainage. In vitro B-cell function does not appear to be affected in this model. Further study of impaired cell-mediated immunity in extrahepatic biliary obstruction will improve our understanding of the immunological status of patients with obstructive jaundice and cholestatic liver diseases.     

Development and reversibility of T lymphocyte dysfunction in experimental obstructive jaundice

This study evaluates the effect of experimental biliary obstruction by bile duct ligation (BDL) and biliary drainage on cell-mediated immunity in Wistar rats. Immune status has been assessed by a mitogen stimulation test of T lymphocytes with phytohaemagglutinin. Animals were followed for up to 35 days after BDL. Regression analysis showed a significant negative correlation between lymphocyte function and the period of jaundice (correlation coefficient -0.57, P less than 0.001). Following BDL for 21 days, groups of animals had internal biliary drainage for 7, 14 and 28 days, and external drainage for 14 days. Compared with obstructed animals, 14 days internal drainage was required to improve lymphocyte function (P less than 0.05). Animals which had 14 days of external drainage had significantly lower lymphocyte stimulation than internal drainage animals (P less than 0.05). The results demonstrate that obstructive jaundice produces a progressive reduction of T lymphocyte function. This can be reversed by biliary drainage, internal drainage being more effective than external drainage.     

经皮经肝胆管引流术治疗肝移植相关性胆瘘

目的 探讨经皮经肝胆管引流术(PTBD)治疗肝移植(OLT)相关性胆瘘的方法 及疗效.方法 回顾性分析经PTBD治疗的7例OLT相关性胆瘘,OLT术后受体胆瘘6例,活体肝移植(LDLT)供体胆瘘1例.吻合口瘘6例(1例合并吻合口狭窄),其中2例合并肝动脉狭窄;肝切缘小胆管残端瘘1例.引流方式采用外引流,引流管前端跨过瘘口部位,成畔固定于胆总管内.瘘口闭合、腹腔引流管无胆汁引出为治愈.结果 7例胆瘘病人PTBD引流量180～450 ml/d,经引流21～87 d(中位时间62 d)后,复查造影瘘口闭合,胆瘘治愈率为7/7.随访时间93～675 d(中位时间124 d),4例无症状生存,2例因多器官功能不全放弃治疗,1例肝癌复发病死.1例在引流期间出现胆道感染.结论 PTBD是治疗0LT相关性胆瘘的一种安全、简便、有效的方法 .     

The value of bile replacement during external biliary drainage: an analysis of intestinal permeability, integrity, and microflora

OBJECTIVE: To investigate the effect of bile replacement following percutaneous transhepatic biliary drainage, ie, external drainage, on intestinal permeability, integrity, and microflora in a clinical setting. SUMMARY BACKGROUND DATA: Several authors have reported that internal biliary drainage is superior to external drainage. However, it is unclear whether bile replacement following external drainage is beneficial. METHODS: Twenty-five patients with biliary cancer underwent percutaneous transhepatic biliary drainage (PTBD) as a part of presurgical management. All externally drained bile was replaced either per os or by administration through a nasoduodenal tube. The interval between PTBD and the beginning of bile replacement was 21.3 +/- 19.7 days, and the length of bile replacement was 20.7 +/- 9.6 days. The lactulose-mannitol test, measurement of serum diamine oxidase (DAO) activity, and analyses of fecal microflora and organic acids were performed before and after bile replacement. RESULTS: The volume of externally drained bile varied widely from patient to patient, ranging from 220 +/- 106 mL/d to 1616 +/- 394 mL/d (mean, 714 +/- 346 mL/d). Biliary concentrations of bile acids, cholesterol, and phospholipids increased significantly after bile replacement. The lactulose-mannitol (L/M) ratio decreased from 0.063 +/- 0.060 before bile replacement to 0.038 +/- 0.032 after bile replacement (P < 0.05). Serum DAO activity increased from 3.9 +/- 1.4 U/L before bile replacement to 5.1 +/- 1.6 U/L after bile replacement (P < 0.005), and the magnitude of change in serum DAO activity correlated with the length of bile replacement (r = 0.483, P < 0.05). Neither the L/M ratios nor serum DAO activities before bile replacement correlated with the interval between PTBD and the beginning of bile replacement. Fecal microflora and organic acids were unchanged. CONCLUSION: Impaired intestinal barrier function does not recover by PTBD without bile replacement. Bile replacement during external biliary drainage can restore the intestinal barrier function in patients with biliary obstruction, primarily due to repair of physical damage to the intestinal mucosa. Our results support the hypothesis that bile replacement during external drainage is beneficial.     

十字型胰胆引流管的应用研究

胰瘘和胆瘘是胰十二指肠切除术后最常见和最严重的并发症，为预防其发生，除吻合方法的不断改进外，胰液及胆汁的引流是预防胰、胆瘘的关键。为了合理引流胰液及胆汁，作者研制了一种十字型胰胆引流管，经临床应用１６例，效果满意，术后未发生胰瘘和胆瘘。作者认为，该管集胰液内、外引流管和胆总管Ｔ型引流管于一体，能起到合理引流胰液和胆汁，预防胰、胆瘘的作用，是一种简便、实用、有推广价值的新方法     

Role of pancreatic duct occlusion with prolamine (Ethibloc) in necrotizing pancreatitis

There is evidence that the pancreatic duct plays an important role in the evolution of necrotizing pancreatitis. We hypothesized that occlusion of the pancreatic duct and its smaller ductules with prolamine (Ethibloc) in opossums at risk of severe necrotizing pancreatitis would have a beneficial effect on the progression of the disease. Sixteen opossums underwent bile duct ligation below the entrance of the pancreatic duct. They were divided into four groups at 6 days. Group I (control, n = 5) opossums were killed for histologic observation of the degree of necrosis of the pancreas; group II (n = 5) underwent external drainage of the pancreatic biliary duct; group III (n = 4) had external biliary drainage and ligation of the pancreatic duct; group IV (n = 7) was treated with external biliary drainage and ligation of the main pancreatic duct after instillation of prolamine. Serum amylase, lipase, and calcium values were determined. The pancreas was examined by inspection and histologically at the time of death, and the severity of the disease was determined by quantitation of pancreatic tissue necrosis. All animals in groups II and III died 8 to 14 days after bile duct ligation, and all had severe necrotizing pancreatitis. All animals in group IV survived and were killed at 2 to 10 weeks after prolamine (Ethibloc) injection into the pancreatic duct. A mild edematous pancreatitis was observed in all seven animals. Prolamine (Ethibloc) provided dramatic protection from progressive necrosis. This study does not provide an explanation, but it allows for speculation that ductal injection interrupted the deleterious effect of proteolytic enzymes and their leakage into the interstitial space of the pancreas.     

Simultaneous segmental obstruction of bile duct and portal vein markedly changes a population of biliary and hepatic cells in human liver

BACKGROUND AND AIMS: No studies have investigated histologic changes caused by simultaneous segmental obstruction of the bile duct and portal vein in human liver. PATIENTS/METHODS: Liver tissues with simultaneous obstruction of the segmental bile duct and portal vein (O(+/+) liver), with segmental bile duct obstruction alone (O(+/-) liver), and without obstruction (O(-/-) liver) were obtained from patients who underwent hepatectomy, and studied morphologically and immunohistochemically. RESULTS: In O(+/+) liver, the proportional area consisting of hepatocytes was significantly less (31.0+/-25.8%) than in O(+/-) liver (78.4+/-18.9%) or O(-/-) liver (86.5+/-9.2%). In contrast, the proportional area consisting of biliary epithelial cells was significantly higher in O(+/+) liver (9.1+/-6.1%) than in O(+/-) liver (1.6+/-1.5%) or O(-/-) liver (0.7+/-0.6%). The proportional area consisting of fibrous tissue also was significantly higher in O(+/+) liver than in the other two groups. In O(+/+) liver, some cells located at the periphery of hepatocyte areas were immunoreactive for both hepatocyte and biliary epithelial cell markers. CONCLUSION: Simultaneous segmental obstruction of the bile duct and portal vein induces a marked ductular increase, periportal fibrosis, and a reduction in the number of hepatocytes in human liver tissue.     

胆汁转流性外引流术对恶性梗阻性黄疸血内毒素水平的影响

目的 :探讨外引流术体外转流胆汁对恶性梗阻性黄疸病人血内毒素水平的影响。方法 :对 14例肿瘤手术不能切除的恶性梗阻性黄疸病人行胆汁转流性外引流术 ,与同期施行的 15例内引流术、2 0例外引流术病人进行手术前后外周血内毒素水平比较。结果 :术前 3组内毒素水平差别无显著性意义 (P>0 .0 5 )。单纯外引流组手术后内毒素水平略高于术前 (P >0 .0 5 ) ;内引流组术后第 2天内毒素水平反而高于术前 (P <0 .0 5 ) ,第 7天、第 14天显著降低 (P <0 .0 5 ,P <0 .0 1) ;体外转流组术后内毒素水平逐渐降低 ,与内引流术组变化基本相同。结论 :胆汁转流性外引流术可降低恶性梗阻性黄疸病人外周血内毒素水平。     

External biliary fistulas selectively managed by endoscopic retrograde cholangiography with sphincterotomy and/or stent placement

External bile duct fistulas are inherent postoperative complications that usually appear after biliary tract surgery, traumatic bile duct injuries and liver surgery for hepatic hydatid disease or liver transplant. The management is highly individualized, while the success and long-term results of endoscopic and surgical techniques are conflicting. The study included 32 cases with external bile duct fistulas managed by endoscopic retrograde cholangiography (ERC) with sphincterotomy and/or stent placement, including "rendez-vous" procedures in 2 cases. The causes of the external fistula were represented by cholecystectomy with/without retained common bile duct stones or strictures (22 cases), cholecystectomy and drainage of a subphrenic abscess caused by severe acute pancreatitis (1 case) and surgical interventions for hepatic hydatid disease (9 cases). Due to the prospective protocol of the study we were able to apply an individualized endoscopic treatment: sphincterotomy with proper relief of the bile duct obstruction (stone extraction) or sphincterotomy with large-size (10 Fr) stent placement for large-sized bile duct defects. The results consisted in closure of the fistula in 3.5 +/- 1.7 days for the subgroup of patients with sphincterotomy alone. Among the patients with stent insertion, fistulas healed slower in 14 +/- 3.5 days. There were no complications after endoscopic treatment; however the stent could not be passed in one patient that required subsequent surgery. In conclusion, endoscopic intervention is the treatment of choice for small external biliary fistulas complicating biliary tract surgery or liver surgery for hepatic hydatid disease. When the fistula is large, the placement of a 10 Fr endoprosthesis becomes necessary, while failure of endoscopic treatment leads to surgery with hepatico-jejunal anastomosis.