Spontaneous perforation of the bile duct (SPBD) is a rare condition affecting infants and children. However, it has been described as the second most common surgical cause of jaundice in infancy after biliary atresia (1). The first case report was published in 1932 by Dijkstra (2). Since then only 150 cases have been reported.
The aetiology of SPBD is unknown, but proposed theories include congenital mural weakness of the common bile duct (CBD) (3), ischaemia (4), distal biliary obstruction (5), and pancreaticobiliary malunion (PBM) (6).
There are several methods of management reported. Procedures performed are simple peritoneal drainage (1,7–11), repair of the perforation (12–17), T-tube drainage (6,18–21) with or without cholecystectomy, Roux-en-Y intestinal anastomosis (22–25), and endoscopic retrograde cholangiopancreatography (ERCP) (26,27). This wide variation suggests that the appropriate management needs further clarification. This review attempts to suggest the most appropriate management. We have reviewed all of the cases reported in the last 20 years and summarised their demographics and management. In addition, we further present 2 cases.
MATERIALS AND METHODS
A literature search of all of the cases of SPBD was undertaken. Searches were carried out using MEDLINE, Google, and PubMed databases. The key words “spontaneous,” “perforation,” “bile duct,” “biliary tract,” “infancy,” and “childhood” were used. The search was then expanded using the related articles facility in PubMed and by reviewing the references within these articles. All of the articles published between March 1988 and March 2008 were reviewed and data of demographics including age, sex and gestation, and management including the investigations and the procedures performed were extracted. In total, 58 cases were identified from 29 publications.
A 3-month-old boy presented with a history of jaundice, abdominal distension, vomiting, pale stools, and failure to thrive. On examination, the child was jaundiced but not unwell. The abdomen was distended with the presence of massive ascites. Bilateral hydroceles were evident. Serum liver enzymes were within normal limits. Plain abdominal film showed a ground glass appearance. Abdominal ultrasound scan (USS) showed normal liver and gallbladder with no dilatation of the intrahepatic bile ducts. No mass was detected. Radionuclide scanning demonstrated prompt visualisation of the liver and gallbladder, followed by visualisation of the CBD that looked slightly dilated. The tracer was found to be leaking into the peritoneal cavity, and there was no tracer in the intestines even at 24 hours (Fig. 1). Ascitic aspiration revealed yellow-coloured fluid. With the diagnosis of spontaneous biliary perforation in mind, a laparotomy was performed. There was a thin-walled pseudocyst at the porta hepatis. Intraoperative cholangiogram through the gallbladder showed extravasation of contrast from the CBD. A small perforation was found on the anterior wall of the CBD at the junction with the cystic duct. There was no distal obstruction. A cholecystectomy was performed and a size 8F T tube was placed into the perforation and the perforation closed around this. An external drain was left in the area of the pseudocyst. The T tube was removed once it had stopped draining and the external drained removed 2 days later. This patient was alive and well 1 year postoperatively.
A 6-month-old girl presented with a 3-month history of gradually increasing abdominal distension and failure to thrive. There was also a 5-month history of pale stools. On examination, she had a distended abdomen, was not jaundiced, and had an umbilical hernia. Serum liver enzymes were within normal limits. Abdominal plain radiograph showed a ground glass appearance. Ultrasound of the abdomen demonstrated massive ascites with internal septations. Radionuclide scanning showed prompt uptake and excretion of tracer from the liver. However, there was extravasation of the isotope from the extrahepatic biliary tract and accumulation of the isotope in a well-walled-off cyst in the right hypochondrium (Fig. 2). A liver biopsy suggested mild cholestasis. An ascitic tap removed 1 L of bilious fluid. Findings at laparotomy were a necrotic walled pseudocyst containing bile in the right hypochondrium adherent to the colon and inferior surface of the liver. The gallbladder was collapsed and minimally inflamed. On dissecting out the pseudocyst, a tiny perforation was found on the anterior wall of the CBD at the junction with the cystic duct. A T-tube insertion was attempted but not possible because the perforation was so small. Simple closure of the perforation was performed and an external drain placed in the vicinity of the CBD.
Postoperatively, there was a rapid decrease in the serum bilirubin after initial mild increase in levels. The drain was removed on day 10 postoperatively, once it stopped draining. Postoperative radionuclide scan showed prompt uptake and excretion from the liver, and the tracer was visualised in the CBD and intestines. There was no evidence of extrabiliary leak. This patient was alive and well 3 months postoperatively.
A review of the last 20 years' literature has revealed 60 cases of SPBD in infants and children, including the 2 described here. The ages at presentation ranged from 25 weeks' gestation (28) to 15 years (29). It most commonly occurs in the first year of life. The median age at presentation is 4 months. Male-to-female ratio is 2:1. In only 28 cases did the authors state the gestational age at birth. Twenty-two of these were born at full term. The other 6 ranged from 26 to 36 weeks' gestation.
The most common presenting symptom is abdominal distension and/or jaundice. Thirty-five of the 60 cases presented with abdominal distension. Thirty-two of the 60 cases were jaundiced at presentation. Other symptoms include fever, vomiting, abdominal mass, and pale-coloured stools. A hernia or a hydrocele was present in 9 cases, presumably due to ascites and increased intraabdominal pressure. Most cases were relatively well at presentation, with a subacute onset and potentially fluctuating symptoms. However, it may present acutely; 6 patients in this series were unwell at presentation.
Serum bilirubin and liver enzymes are often normal or only mildly elevated. There may be a mild leucocytosis. If an ascitic tap is performed, the concentration of bilirubin in the fluid is higher than that in the serum.
The most common investigation performed was abdominal USS (Table 1). In 1 case, the multicystic mass was thought to be an enteric duplication cyst (21), and in another, a choledochal cyst (23). The demonstration of tracer in the peritoneal cavity on a radionuclide scan with failure of tracer to enter the bowel is diagnostic. The site of perforation was found in 56 of the cases. Table 2 shows the sites. A summary of the management is shown in Table 3.
In total, 46 patients were managed with drainage ± repair of perforation ± cholecystectomy; 8 of these subsequently required biliary reconstruction (BR). This was because of the continued bile leak in 2 (17,24), distal obstruction in 2 (6,30), PBM in 3 (6,31), and for unknown reasons in 1 (30). In the 2 cases of continued leak, the authors did not state for how long this bile leak continued before further laparotomy. In the majority of successfully drained patients, bile leakage ceased by 14 days, and in another case, it continued to leak for 4 weeks (21) but then settled spontaneously. It is therefore possible that in these 2 cases the bile leak may have settled spontaneously in time. Two patients had postoperative distal obstruction; the obstruction was found at 3 weeks and went on to have BR at this time. In the case described by Hirigoyen et al (19), postoperative obstruction took 24 days to resolve spontaneously.
The fact that most perforations occurred close to the junction of the cystic duct supports the theory of a developmental weakness of the wall of the duct that perforates after a certain intraductal pressure is reached. This was first postulated by Peterson in 1955 (32) and later by Johnston in 1961 (33).
Most patients were healthy before the perforation occurred. Associated conditions were found in 3 cases. The associated conditions were Ivemark syndrome (16), HIV with stage IV Hodgkin lymphoma (27), and previous necrotizing enterocolitis in the first week of life (25).
The case reported by Chilukuri et al (28) supports the theory that it may be a congenital abnormality because the perforation in this case occurred before 25 weeks' gestation. This is the only reported case of an antenatal diagnosis of SPBD.
A viral aetiology was suspected by Moore (34) but without any substantive evidence. There were no abnormal viral titres or immunoglobulin levels commented on in these patients, except that cytomegalovirus was isolated from the urine of a very premature infant (23) and the 1 patient who was known to be HIV positive (27).
An ischaemic aetiology has been proposed by Northover and Terblanche (4) due to the posterolateral arterial supply making the anterior wall of the CBD susceptible to ischaemia. They have shown that the CBD receives its blood supply from 2 pairs of marginal arteries; 1 pair arises inferiorly from the superior pancreaticoduodenal artery and the other arises superiorly from the right hepatic artery. These contribute 60% and 40%, respectively, of the vascular supply of the CBD. The junctional area between these 2 vascular territories is a potential ischaemic watershed. This may predispose to focal ischaemia of the CBD wall in the event of a period of hypoperfusion, resulting in perforation. A particular susceptibility to splanchnic ischaemia in neonates may explain localized injury in infants with precarious bile duct blood supply (35).
PBM is a congenital anomaly in which a union of the pancreatic and biliary ducts is located outside the duodenal wall containing the sphincter of Oddi (6). Pancreatic juice refluxing into the bile duct is activated by mixing with the bile to become potentially destructive. PBM has been postulated to contribute to the aetiology of SPBD. Ohkawa et al (36) states that the dilated bile duct caused by the PBM may lead to acute inflammation and microabscess formation in the duct wall from bile stasis, eventually resulting in perforation. Ando et al (37) stressed that SPBD in association with PBM is related to an abrupt increase in intraluminal pressure from impaction of a protein plug, which is likely to be associated with PBM and congenital dilatation of the bile duct. Ng et al (38) report that virtually all Asian patients with SPBD, who are fully investigated, had in common a radiographically demonstrable long common pancreaticobiliary channel. Two patients who underwent CBD excision and hepaticojejunostomy remained completely symptom free for up to 20 years, whereas a girl who had been managed with T-tube drainage, suffered from recurrent pancreatitis in the ensuing 8 years. Yano and Matsumoto (39) reported 3 patients who were initially treated by drainage alone and who developed choledochal cysts during 2 to 8 years of follow-up. According to Ng, all of the patients will ultimately require excision of CBD and bilioenteric anastomosis, although this may not be appropriate in the acute situation. Another reason for this is the association of PBM with malignant degeneration of the biliary tract in the long term. The incidence of biliary carcinoma developing from PBM with bile duct dilatation was estimated to be 17.5%. The incidence without bile duct dilatation was reported to be as high as 53.6% (40,41). It seems reasonable, therefore, to suggest that any patient with demonstrable PBM should have excision of the CBD with bilioenteric anastomosis.
There is evidence to suggest that SPBD and choledochal cysts are interrelated entities with a common pathogenic factor (39). PBM, intrinsic bile duct weakness, and distal bile duct stenoses have been implicated in both of these conditions.
Obstruction distal to the perforation has been implicated in a number of cases in which stones or actual ductal stenosis has been found. However, in some cases obstruction is considered as a consequence of perforation, slow bile transit, and biliary stasis rather than the cause of the perforation, and the obstruction usually resolves with adequate drainage. From these 60 cases, 20 were known to have obstruction, 9 of these were managed with primary BR and 11 with drainage ± repair of perforation ± cholecystectomy. Of these 11, 2 had PBM and therefore required BR; in 6 patients, the obstruction resolved spontaneously.
Davenport et al (23) reported a case with acquired biliary atresia; this contrasts with those with congenital biliary atresia in the relatively late presentation and the rapid return of normal liver function after the surgery. The prognosis in congenital biliary atresia treated at 24 weeks is extremely poor and liver function tests would not return to normal even if biliary drainage were achieved. The acquired occlusion was thought to be a result of the sclerosing effect of the bile leak from the perforation.
Previously, almost all of the reports of nonoperative management have ended in the death of the child (20). However, in the last 20 years 1 patient was managed nonoperatively (24) and another had a drain sited under local anaesthetic because he was too unstable for surgery (10). These 2 patients did well and were alive and well at 9 years and 9 months, respectively.
Management should depend on the condition of the child. Intraoperative cholangiography should ideally be performed in all of the cases to detect ductal abnormalities. Distal obstruction demonstrated on cholangiography resolves with decompression in the majority of cases, because this is thought to be secondary to the perforation rather than being the primary causative factor. However, Spigland et al (31) dispute this widely held belief and recommend prompt biliary intestinal bypass. We believe exploration of the porta hepatis may be hazardous at this time because of the inflammatory nature of the bile leak and therefore recommend simple peritoneal drainage with T-tube drainage even if there is a distal obstruction. This entails less morbidity and has a good chance of curing the condition or at least stabilising the patient for second-look definitive surgery should the obstruction not resolve. In the case of PBM detected on cholangiography, biliary intestinal anastomosis is necessary to prevent biliary cirrhosis, portal hypertension, recurrent pancreatitis, and ultimately biliary carcinoma (40,41). However, this can be done at a second laparotomy when inflammation has settled. Repair of the perforation is unnecessary and may be hazardous; there is also the risk of postoperative stricture.
The postoperative course may be complicated with cholangitis, which requires prolonged antibiotic treatment (18). Although bilious ascites is usually sterile, infection occurred in only 2 cases; a broad-spectrum antibiotic is recommended (9). Portal venous thrombosis occurred in 4 cases (24,30). Four patients in the Charcot series had postoperative portal hypertension and 2 of these were attributed to portal vein thrombosis. The aetiology is assumed to be as simple as irritative bile or a biloma abutting onto the portal vein (30). This is thought to occur more commonly with posterior perforations.
The aetiology of SPBD is unclear, but we believe that there is probably more than 1 cause. The diagnosis should be suspected in any infant with an insidious onset of jaundice and abdominal distension with alcoholic stools. An abdominal USS and biliary radionuclide scan can be used to make a rapid diagnosis.
Minimal surgery in the form of simple peritoneal drainage and T-tube drainage is curative in the majority of cases unless there is PBM, which will require BR, although this can be done at a second laparotomy once the child is more stable and the inflammation has settled. Even in the cases of known distal obstruction, this will resolve spontaneously in the majority of cases.
With recent advances in laparoscopic surgery, diagnosis and percutaneous drainage is a definite alternative (9). In centres with appropriate equipment and expertise, ERCP is an alternative primary therapy, avoiding the need for surgery (26).
1. Stringel G, Mercer S. Idiopathic perforation of the biliary tract in infancy. J Pediatr Surg 1983; 18:546–550.
2. Dijkstra CH. Graluistorting in de buikholte bij een zuigeling, Maandschr. Kindergeneeskd 1932; 1:409–414.
3. Lilly JR, Weintraub WH, Altman RP. Spontaneous perforation of the extrahepatic bile ducts and bile peritonitis in infancy. Surgery 1974; 75:664–673.
4. Northover JMA, Terblanche J. A new look at the arterial supply of the bile duct in man and its surgical implications. Br J Surg 1979; 66:379–384.
5. Colver HD. Perforation of the biliary tract due to gallstones in infancy: an established clinical entity. Ann Surg 1964; 2:226–231.
6. Hasegawa T, Udatsu Y, Kamiyama M, et al
. Does pancreatico-biliary maljunction play a role in spontaneous perforation of the bile duct in children. Pediatr Surg Int 2000; 16:550–553.
7. Hammoundi SM, Alauddin A. Idiopathic perforation of the biliary tract in infancy and childhood. J Pediatr Surg 1988; 23:185–187.
8. Haller JO, Condon VR, Berdon WE, et al
. Spontaneous perforation of common bile duct in children. Radiology 1989; 172:621–624.
9. Banani SA, Bahador A, Nezakatgoo N. Idiopathic perforation of the extrahepatic bile duct in infancy: pathogenesis, diagnosis and management. J Pediatr Surg 1993; 28:950–995.
10. Kasat LS, Borwankar SS, Jain M, et al
. Spontaneous perforation of the extrahepatic bile duct in an infant. Pediatr Surg Int 2001; 17:463–464.
11. Sharma SB, Sharma SC, Gupta V. Spontaneous biliary perforation: a rare entity in late infancy and childhood. Indian J Pediatr 2003; 70:829–831.
12. Enell H, Cavell B, Malmfors G. Spontaneous perforation of the common bile duct. Acta Paediatr Scand 1979; 68:625–626.
13. Megison SM, Votteler TP. Management of common bile duct obstruction associated with spontaneous perforation. Surgery 1992; 111:237–239.
14. Kumar R, Sriram M, Bhatnagar V, et al
. Spontaneous perforation of the common bile duct in infancy: role of Tc-99m mebrofenin hepatobiliary imaging. Clin Nucl Med 1999; 24:847–848.
15. Bingöl-Koloğlu M, Karnak I, Ocal T, et al
. Idiopathic perforation of the bile duct in an infant. J Pediatr Gastroent Nutr 2000; 31:83–85.
16. Prabakaran S, Kumaran N, Regunanthan SR, et al
. Spontaneous biliary perforation in a child with features of Ivemark syndrome. Pediatr Surg Int 2000; 16:109–110.
17. Smethurst FA, Carty H. Case report: spontaneous perforation of the common bile duct in infancy. Br J Radiol 1993; 66:556–557.
18. Kanojia RP, Sinha SK, Rawat J, et al
. Spontaneous biliary perforation in infancy and childhood: clues to diagnosis. Indian J Pediatr 2007; 74:509–510.
19. Hirigoyen MB, Geoghegan JG, Owen ER, et al
. Spontaneous perforation of the bile duct in infancy. Eur J Pediatr Surg 1995; 5:375–376.
20. Sahnoun L, Belghith M, Jouini R, et al
. Spontaneous perforation of the extrahepatic bile duct in infancy: report of two cases and literature review. Eur J Pediatr Surg 2007; 17:132–135.
21. Goldberg D, Rosenfeld D, Underberg-Davis S. Spontaneous biliary perforation: biloma resembling a small bowel duplication cyst. J Pediatr Gastroent Nutr 2000; 31:201–203.
22. Coffey R, Oweida L. Unusual problems in biliary tract surgery. Am J Gastroenterol 1955; 24:485–498.
23. Davenport M, Heaton ND, Howard ER. Spontaneous perforation of the bile duct in infants. Br J Surg 1991; 78:1068–1070.
24. Chardot C, Iskandarani F, Dreuzy OD, et al
. Spontaneous perforation of the biliary tract in infancy: a series of 11 cases. Eur J Pediatr Surg 1996; 6:341–346.
25. Pradas DVI, Vila JJ, Fernandez MS, et al
. Spontaneous biliary perforation and necrotizing enterocolitits. Pediatr Surg Int 1999; 15:401–402.
26. Barnes BH, Narkewicz R, Sokol RJ. Spontaneous perforation of the bile duct in a toddler: the role of endoscopic retrograde cholangiopancreatography in diagnosis and therapy. J Pediatr Gastroent and Nutr 2006; 43:695–697.
27. Makin E, Davenport M. Idiopathic biliary perforation in a 10 year old boy. Pediatric Surg Int 2006; 22:465–467.
28. Chilukuri S, Bonet V, Cobb M. Antenatal spontaneous perforation of the extrahepatic biliary tree. Am J Obstet Gyne 1990; 163:1201–1202.
29. Shah SH, Relekar RG, Naik A, et al
. Spontaneous perforation of the left hepatic duct with acalculous cholecystitis. Ind J Gastroenterol 1990; 9:317.
30. Livesey E, Davenport M. Spontaneous perforation of the biliary tract and portal vein thrombosis in infancy. Pediatr Surg Int 2008; 24:357–359.
31. Spigland N, Greco R, Rosenfeld D. Spontaneous biliary perforation: does external drainage constitute adequate therapy? J Pediatr Surg 1996; 31:782–784.
32. Peterson G. Spontaneous perforation of the common bile duct in infants. Acta Chir Scan 1955; 60:192–201.
33. Johnston JH. Spontaneous perforation of the common bile-ducts in infancy. Br J Surg 1961; 48:532–533.
34. Moore TC. Massive bile peritonitis in infancy due to spontaneous bile duct perforation with portal vein occlusion. J Pediatr Surg 1975; 10:537–538.
35. Lloyd JR. The etiology of gastro-intestinal perforation in the newborn. J Pediatr Surg 1969; 4:77–84.
36. Ohkawa H, Takahashi H, Maie M. A malforation of the pancreatico-biliary system as a cause of perforation of the biliary tract in childhood. J Pediatr Surg 1977; 12:541–546.
37. Ando H, Ito T, Watanabe Y, et al
. Spontaneous perforation of choledochal cyst. J Am Coll Surg 1995; 181:125–128.
38. Ng WT, Cheung CH, Chan S. Is spontaneous perforation of the bile duct in children due solely to pancreatico-biliary maljunction? Pediatr Surg Int 2002; 18:565–566.
39. Yano H, Matsumoto H. Choledochal cyst following operation for idiopathic perforation of the biliary tract in childhood. Jpn J Surg 1983; 13:441–445.
40. Murakami Y, Kodama T, Takesue Y, et al
. Anamalous pancreaticobiliary ductal junction without dilatation of the choledochus. Surg Today 1992; 22:276–279.
41. Todani T, Watanabe Y, Fujii T, et al
. Carcinoma arising from the bile duct in choledochal cyst and anomalous arrangement of the pancreaticobiliary ductal union (in Japanese). Tan Sui 1985; 6:525–535.