Benign biliary tract diseases

Published on 11/04/2015 by admin

Filed under Surgery

Last modified 22/04/2025

Print this page

This article have been viewed 2170 times

Benign biliary tract diseases

Benjamin N.J. Thomson and O.James Garden

Introduction

Apart from those disorders related to choledocholithiasis, benign diseases of the biliary tree are relatively uncommon (Box 11.1). The most challenging issues are in patients who present with symptoms associated with biliary strictures, which arise more commonly following iatrogenic injury during cholecystectomy. Congenital abnormalities such as choledochal cysts and biliary atresia are usually in the domain of the paediatric surgeon, although later presentation of cysts may occur after missed diagnosis or when revisional surgery is required. Most of the published literature regarding benign non-gallstone biliary disease is retrospective or at best prospectively gathered, non-randomised data, but clear guidelines can be followed based upon observation.

Box 11.1 Benign causes of biliary strictures

Strictures of the extrahepatic biliary tree

Iatrogenic biliary injury

Postcholecystectomy

Mirizzi’s syndrome

Recurrent pyogenic cholangitis

Parasitic infestation

Clonorchis sinensis

Opisthorchis viverrini

Echinococcus

Ascaris

Primary sclerosing cholangitis

Benign strictures imitating malignancy

Pancreatitis

Lymphoplasmacytic pancreatitis

Inflammatory pseudotumour

Idiopathic strictures

HIV cholangiopathy

Congenital anomalies

Biliary atresia

Biliary atresia occurs in approximately 1 per 10 000 live births but its aetiology remains unclear. There is experimental evidence for a primary perinatal infection as well as cellular and humoral autoimmunity. An inflammatory process before birth may result in failure of the biliary lumen to develop in all or part of the extrahepatic biliary tree.

Presentation is usually in the early neonatal period with prolongation of neonatal jaundice. Most patients are treated in specialist neonatal surgical units; however, occasionally patients may be referred to adult units for assessment for liver transplantation following previous unsuccessful treatment. Management in the neonate is by porto-enterostomy (Kasai’s operation), which involves anastomosis of a Roux limb of jejunum to the tissue of the hilum. Restoration of bile flow has been reported in 86% of infants treated before 8 weeks of age, but only 36% in older children.¹ Four-year survival is dependent on the timing of surgery. Of 349 North American children with biliary atresia, 210 (60%) required later liver transplantation, with a 4-year transplantation survival of 82%.² Recent evidence has suggested better outcomes following maternal liver-related liver transplantation, potentially due to tolerance to non-inherited maternal antigens.³

Choledochal cysts

The earliest description of a choledochal cyst was by Douglas in 1952,⁴ who described a 17-year-old girl with jaundice, fever and a painful mass in the right hypochondrium. However, presentation is usually in childhood and around 25% are diagnosed in the first year, although prenatal diagnosis is now possible with improvements in antenatal ultrasonography. Adult centres treat a small proportion of those presenting with delayed diagnosis as well as those with complications from previous cyst surgery.

The incidence of choledochal cysts in Western countries is around 1 in 200 000 live births but it is much higher in Asia. There is frequent association with other hepatobiliary disease such as hepatic fibrosis, as well as an aberrant pancreatico-biliary duct junction.⁵

Magnetic resonance cholangiopancreatography (MRCP) now allows images that are superior to traditional cholangiography (Fig. 11.1), and it should be recommended due to its non-invasive nature.⁶

Figure 11.1 MRCP (a) and macroscopic photograph (b) demonstrating a type I choledochal cyst with a distal cholangiocarcinoma in a 42-year-old Caucasian woman requiring a pancreaticoduodenectomy. Gallbladder (GB), tumour (T), pancreatic duct (single arrow) and aberrant common channel (double arrow) are shown. Courtesy of Professor Prithi S. Bhathal, Pathology Department, University of Melbourne, Australia.

Classification

The modified Todani classification is employed to describe the various forms of choledochal cyst ⁷ (Fig. 11.2). Type I, the most common, represents a solitary cyst characterised by fusiform dilatation of the common bile duct. Type II comprises a diverticulum of the common bile duct, whilst type III cysts are choledochocoeles. Type IV is the second most common, with extension of cysts into the intrahepatic ducts. Lastly, type V involves intrahepatic cystic disease with no choledochal cyst, which merges into the syndrome of Caroli’s disease.

Figure 11.2 Modified Todani classification for choledochal cysts.⁷ Reproduced from Todani T, Watanabe Y, Narusue M et al. Congenital bile duct cysts: classification, operative procedures, and review of thirty-seven cases including cancer arising from choledochal cyst. Am J Surg 1977; 134:263–9. With permission from Elsevier.

Risk of malignancy

In the Western literature, the incidence of cholangiocarcinoma is reported to be approximately 12% (Fig. 11.1),⁸ compared to Todani et al.’s Japanese experience of 16% in 1353 patients.⁹ The incidence of malignancy is reported to be 2% at 20 years, increasing to 43% for those in their sixties.¹⁰ Cyst drainage without cyst excision does not prevent later malignant change, and there is continuing debate regarding the precise ongoing risk following cyst resection. Takeshita et al. reported 180 patients who underwent primary surgery for a choledochal cyst. Synchronous malignancy was found in 36 patients (20%), with only one of the remaining 144 patients developing malignancy during follow-up.¹¹

Management

Surgical resection is required to prevent recurrent episodes of sepsis and pain, to prevent the risk of pancreatitis from passage of debris and calculi, and because of the association with cholangiocarcinoma. Complete cyst excision with preservation of the pancreatic duct is required, with hepatico-jejunostomy for reconstruction. Some authors advocate liver resection for type IV cysts with intrahepatic extension for complete removal of the cyst, although the advantage is debatable. For those patients with Caroli’s disease, resection may be feasible if the biliary involvement is localised to one part of the liver. For other patients, endoscopic or radiological techniques may be required to address biliary sepsis by improving biliary drainage, while others may need to be considered for hepatic replacement if liver failure develops.

For extrahepatic cysts, cyst-enterostomy, or drainage of the cyst into the duodenum, should no longer be performed as the cyst epithelium remains unstable and malignant potential exists. If previous drainage has been performed, symptoms of cholangitis generally persist and conversion to a Roux-en-Y hepatico-jejunostomy is advisable.

Special operative techniques

During operative exposure, intraoperative ultrasound is very useful to identify the biliary confluence, the intrahepatic extension of the cyst, and the relationship to the right hepatic artery above and to the pancreatic duct below (Fig. 11.3). Small aberrant hepatic ducts may enter the cyst below the biliary confluence and these are missed frequently on preoperative imaging.¹² Such aberrant ducts are usually identified once the cyst has been opened. The cyst is normally best excised in its entirety and this is facilitated by opening it along its anterior length. This aids identification of the vessels from which the cyst is freed. Early identification of the biliary confluence aids the surgeon in planning the incorporation of any segmental duct into the eventual hepatico-jejunal Roux-en-Y anastomosis. Dissection into the head of the pancreas is made easier by use of bipolar scissors and the CUSA™ (ultrasonic surgical aspiration system, ValleyLab, Boulder, CO) if the plane of dissection is obscured by fibrosis or inflammation. It may be necessary to leave a small oversewn lower common bile duct stump to avoid compromise to the pancreatic duct lumen; however, recurrent pancreatitis and possible malignant transformation remain possible complications.

Figure 11.3 Operative ultrasound scan of a type I choledochal cyst. The junction of the undilated proximal biliary tree with the cyst (long dotted line) is demonstrated. The right hepatic artery is posterior (two arrows), as is the right branch of the portal vein (short dotted line).

There is an accepted association between choledochal cyst and cholangiocarcinoma. The cyst should be excised and the biliary tree reconstructed by means of a hepatico-jejunostomy Roux-en-Y.

Iatrogenic biliary injury

The commonest cause of an injury to the extrahepatic biliary tree is as a result of an iatrogenic injury at the time of cholecystectomy. Although it is recognised that injury may also occur during other gastric or pancreatic procedures, this is much less common with the reduction in ulcer surgery and increasing specialisation in pancreatico-biliary surgery. Rarely, the injury may be related to abdominal trauma, injection of scolicidal agents in the management of hydatid cyst, ablation of hepatic tumours or radiotherapy.

The true incidence of biliary injury following laparoscopic cholecystectomy remains obscure. It has been suggested that there was a slight increase in the incidence of injuries following initial introduction of the laparoscopic technique,¹³ with a reported incidence of 0.3–0.7%.^14–16 Open cholecystectomy is said to have a lower incidence of biliary injury, with a rate of 0.13%.¹⁷ Recent variations in technique such as single-incision laparoscopic surgery (SILS) cholecystectomy are not immune to biliary injury. Han et al. recently reported two (1.5%) bile duct injuries in 150 patients having single-port laparoscopic surgery.¹⁸

Aetiology

Previous reports of injury during laparoscopic cholecystectomy suggested that injury was more likely to occur when performed for pancreatitis, cholangitis or acute cholecystitis.¹⁹ However, in a prospective analysis of patients referred following biliary injury, 71% occurred in patients in whom the indication for cholecystectomy was biliary colic alone,²⁰ and thus surgeons should always be vigilant regardless of the indication.

In the majority of patients the problem is misinterpretation of the biliary anatomy, with the common bile duct being confused with the cystic duct. Associated injury to the right hepatic artery often occurs as it is mistaken for the cystic artery. Partial injury may occur to the common bile duct after a diathermy burn or due to rigorous traction on the cystic duct, leading to its avulsion from the bile duct.

Techniques to avoid injury

Many techniques have been described to decrease the risk of injury to the common bile duct during cholecystectomy. The main risk factors are thought to be inexperience, aberrant anatomy and inflammation.¹⁹^,²¹ However, in an analysis of 252 laparoscopic bile duct injuries, the authors suggested that the primary cause of error was a visual perceptual illusion in 97% of cases, whilst faults in technical skill were thought to have been present in only 3% of injuries.²²

Correct identification of the biliary anatomy is essential in avoiding injury to the extrahepatic bile duct. Dissection of Hartmann’s pouch should start at the junction of the gallbladder and cystic duct and continue lateral to the cystic lymph node, thus staying as close as possible to the gallbladder. The biliary tree and hepatic arterial anatomy is highly variable and therefore great care must be taken in identifying all structures within Calot’s triangle before ligation. In Couinaud’s published study of biliary anatomy, 25% had drainage of a right sectoral duct directly into the common hepatic duct.²³ Sometimes this structure may follow a prolonged extrahepatic course, where it can be at greater risk from cholecystectomy. The right hepatic artery may also course through this area. All structures should be traced into the gallbladder to minimise the risk of injury (Fig. 11.4 and 11.5).

Figure 11.4 Aberrant biliary anatomy. The normal biliary anatomy is a trifurcation of the right sectoral and left hepatic ducts forming the common hepatic duct which receives the cystic duct after a variable distance. Operative photograph (a) and a cholangiogram (b) of a short cystic duct (single arrow) draining into the right posterior sectoral duct (double arrow), which has a long extrahepatic course.

Figure 11.5 Operative cholangiography of an aberrant right sectoral duct. The injury was recognised after division of the duct following cholangiography. The cholangiogram catheter was used to obtain a cholangiogram of the aberrant duct. The surgeon obtained advice by telephone and a decision was made to ligate the duct. The patient remains asymptomatic.

Calot’s original description of gallbladder anatomy described a triangle formed by the cystic duct, common hepatic duct and superior border of the cystic artery.²⁴ For satisfactory visualisation of the structures, dissection should also extend above the cystic artery to the liver. Extensive dissection should be avoided in Calot’s triangle as diathermy injury may occur to the lateral wall of the common hepatic duct. Furthermore, arterial bleeding in this area should not be cauterised or clipped blindly. Most bleeding can be controlled with several minutes of direct pressure with a laparoscopic forceps compressing Hartmann’s pouch on to the bleed point. During the era of open cholecystectomy many advocated complete excision of the cystic duct to its insertion into the common bile duct to avoid a cystic duct stump syndrome. However, extensive dissection around the common bile duct with or without diathermy may cause an ischaemic stricture due to damage to the intricate blood supply of the common hepatic duct.

Many authors argue that operative cholangiography is essential to avoid biliary injury.¹⁶^,¹⁹ Fletcher et al. reported an overall twofold reduction in biliary injuries with the use of operative cholangiography, with an eightfold decrease in complex cases.¹⁹ Flum and colleagues analysed retrospectively the Medicare database in the USA and identified 7911 common bile duct injuries following cholecystectomy. After adjusting for patient-level factors and surgeon-level factors the relative risk was 1.49 when intraoperative cholangiography was not used.¹⁶ When the use of intraoperative cholangiography has undergone cost analysis, routine cholangiography has been found to be the most cost-effective during high-risk operations when employed by less experienced surgeons.²⁵

Unfortunately, many operative cholangiograms are interpreted incorrectly and injuries are missed. Although this event should be less frequent with the use of modern C-arm imaging, in reported series of biliary injuries only 6–33% of operative cholangiograms are interpreted correctly.²⁰^,²⁶ For correct anatomical interpretation of the proximal biliary tree, both right sectoral/sectional ducts and the left hepatic duct should be visualised. In the presence of an endoscopic sphincterotomy, contrast will preferentially flow into the duodenum and the patient may need to be placed in a head-down position to fill the intrahepatic ducts. If the anatomy is unclear no proximal clip should be placed on what is presumed to be the cystic duct, to avoid a crush injury to what may be the common hepatic duct.

Retrograde cholecystectomy has been described previously as a safe technique when inflammation around Calot’s triangle makes identification of the anatomy difficult. Nonetheless, care still needs to be exercised during dissection to avoid injury to the right hepatic artery and common hepatic duct, which may be adherent to an inflamed gallbladder. Eight such patients have recently been described by Strasberg and Gouma.²⁷ If identification remains impossible then the gallbladder can be opened to facilitate identification of the cystic duct. A subtotal cholecystectomy should be considered if a safe plane of dissection cannot be established, thus avoiding injury to the common hepatic or left hepatic ducts. Originally described for open cholecystectomy, these techniques have now also been performed laparoscopically.²⁸

Bile duct injury can be avoided by careful identification of the biliary anatomy, dissection close to the gallbladder and avoidance of diathermy in Calot’s triangle. The use of operative cholangiography and its correct interpretation is associated with a reduced incidence of bile duct injury.

Classification

Injury to the distal biliary tree is less technically demanding to repair than involvement of the biliary confluence. The success of reconstruction depends on the type of injury and the anatomical location.²⁹ Bismuth first described a classification system for biliary strictures reflecting the relationship of the injury to the biliary confluence (Table 11.1).³⁰ Strasberg et al. further proposed a broader classification to include a number of biliary complications, including cystic stump leaks, biliary leaks and partial injuries to the biliary tree (Fig. 11.6).¹⁷

Table 11.1

Bismuth classification of biliary strictures

Bismuth classification	Definition
Bismuth 1	Low common hepatic duct stricture – hepatic duct stump > 2 cm
Bismuth 2	Proximal common hepatic duct stricture – hepatic duct stump < 2 cm
Bismuth 3	Hilar stricture with no residual common hepatic duct – hepatic duct confluence intact
Bismuth 4	Destruction of hepatic duct confluence – right and left hepatic ducts separated
Bismuth 5	Involvement of aberrant right sectoral hepatic duct alone or with concomitant stricture of the common hepatic duct

Figure 11.6 Strasberg classification. Type A injuries include leakage from the cystic duct or subvesical ducts. Type B involves occlusion of part of the biliary tree, most usually an aberrant right hepatic duct. If the former injury involves transection without ligation this is termed a type C injury. A lateral injury to the biliary tree is a type D injury. Type E injuries are those described by Bismuth and subdivided into his classification (Table 11.1). Adapted from Strasberg SM, Hertl M, Soper NJ et al. An analysis of the problem of biliary injury during laparoscopic chole cystectomy. J Am Coll Surg 1995; 180:102–25. With permission from the American College of Surgeons.

Presentation

It is preferable that injuries are recognised at the time of surgery to allow the best chance of repair, but this occurs in less than a third of patients. An unrecognised injury may present early with a postoperative biliary fistula, symptoms of biliary peritonitis or jaundice. Early symptoms or signs may be lacking but ductal injury should be suspected in the patient whose recovery is not immediate or is complicated by symptoms of peritoneal or diaphragmatic irritation and/or associated with deranged liver function tests in the first 24–48 hours of surgery. Signs may range from localised abdominal tenderness through to generalised peritonitis with overwhelming sepsis. Ligation of the bile duct will present early with jaundice; however, later presentation may occur as a result of stricture formation from a partial injury, localised inflammation or ischaemic insult.

Ligation of sectoral ducts may cause subsequent or late atrophy of the drained liver segments, which may become infected secondarily. Occasionally liver resection or transplantation may be required for fulminant hepatic failure secondary to combined biliary and vascular injuries.³¹^,³² More commonly, injuries may present late with secondary biliary cirrhosis, which may require liver transplantation when liver failure results.³¹^,³²

In many patients there is a delay until referral, despite evidence of a biliary injury. In a report by Mirza et al., the median interval until referral was 26 days.³³ This delay is not inconsequential as the opportunity for an early repair is lost and results in the liver sustaining further damage.

Management

Intraoperative recognition

In a review by Carroll et al., only 27% of patients underwent a successful repair by the primary surgeon responsible for the injury, whilst 79% of repairs performed following referral had a successful outcome.³⁴ If experienced help is not at hand, no attempt should be made to remedy the situation since this may compromise subsequent successful management. A T-tube or similar drain should be placed to the biliary injury and drains left in the subhepatic space, followed by referral to a specialist centre. No attempt should be made to repair a transection or excision of the bile duct.

A partial injury to the bile duct may sometimes be managed by direct closure with placement of a T-tube through a separate choledochotomy. Primary repair with or without a T-tube for complete transection of the common bile duct is nearly always unsuccessful. This may result from unappreciated loss of common duct or an associated arterial injury, or result from local diathermy injury or devascularisation of the duct from overzealous dissection of the common bile duct ³⁵ (Fig. 11.7a,b). Succesful endoscopic treatment is possible for failed primary repair; however, as many as 32% will require subsequent hepatico-jejunostomy.³⁶

Figure 11.7 (a) Failure of primary repair with T-tube. Primary repair was performed for an injury to the common bile duct presenting with biliary peritonitis. A T-tube was inserted through the anastomosis and this was removed at 4 weeks. An anastomotic stricture developed and the patient required a hepatico-jejunostomy 2 months later. (b) Failure of primary repair for ligation of the common bile duct. A complete transection of the common bile duct identified at postoperative endoscopic retrograde cholangiopancreatography (ERCP). Immediate repair was performed with a direct duct-to-duct repair. (c) A tight anastomotic stricture is demonstrated at a later ERCP.

If an injury to the biliary tree is suspected during cholecystectomy, help must be sought from an experienced hepatobiliary surgeon. A successful repair by the surgeon who has caused the injury is far less likely than one performed by a surgeon experienced in performing a hepatico-jejunostomy.

Postoperative recognition: biliary fistula

Any patient who is not fit for discharge at 24 hours due to ongoing abdominal pain, vomiting, fever or bile in an abdominal drain should be considered to have a biliary leak. The lack of bile in an abdominal drain does not exclude the possibility of a biliary leak, particularly if there is liver function test derangement. Symptoms and signs vary widely, and widespread soiling of the abdominal cavity may be present with few signs.

Initial investigation should include full blood examination and determination of serum levels of urea, electrolytes, creatinine and liver function tests. Ultrasound is usually the initial investigation but it cannot readily differentiate bile and blood from a residual fluid collection following uneventful cholecystectomy. It may provide important information about the presence of intra-abdominal or pelvic fluid, biliary dilatation or retained stones within the bile duct.

If there is evidence of significant peritoneal irritation from widespread biliary peritonitis, laparoscopy allows confirmation of this and provides an opportunity for abdominal lavage. The porta hepatis can be inspected to determine the cause of the bile leak. Whilst dislodged clips from the cystic duct can be managed by application of further clips or suture, any other form of bile leak should lead to specialist referral. Drains can be placed to the subhepatic space as well as the subdiaphragmatic space and pelvis if required. No attempt should be made to repair an injury laparoscopically. If laparotomy is required, this should be considered in conjunction with specialist assistance if bile duct injury is suspected.

Further assessment depends on the clinical situation. The majority of biliary fistulas are due to leaks from the cystic duct stump or subvesicle ducts, and endoscopic retrograde cholangiopancreatography (ERCP) allows anatomical definition, endoscopic sphincterotomy or stent placement. As complete transection of the bile duct precludes ERCP, computed tomography intravenous cholangiography (CT-IVC) or MRCP can determine continuity of the biliary tree prior to endoscopy. Occasionally, persistent bile drainage is associated with choledocholithiasis requiring sphincterotomy and stone extraction. Most simple cystic duct stump leaks can be resolved by endoscopic stenting if cannulation is possible at ERCP ³⁷ and occasionally side injury to the biliary tree can be controlled with endoscopic stent placement.³⁸

If ERCP is unsuccessful or the bile duct is ligated or occluded by clips, percutaneous transhepatic cholangiography may facilitate biliary decompression but it is less frequently employed for diagnosis or delineation of the biliary anatomy. Occasionally, both sides of the liver may need to be externally drained to gain control of a biliary fistula, especially with E4 injuries to the biliary confluence. However, injury to the biliary tree detected in this way may allow surgical repair to be considered within the first week of injury in the stable non-septic patient, and again such further investigation or management decisions should only be considered following specialist referral.

Where the diagnosis of bile duct injury has been delayed, the aim should be to control the biliary fistula with external drainage using surgical or radiologically placed drains. Further control may be required with endoscopic stenting or external biliary drainage. Delayed repair can be considered subsequently once sepsis and intra-abdominal soiling have resolved, as a planned elective procedure in a specialist unit usually 2–3 months following injury. Such an initial conservative approach renders a potentially difficult operation into a repair that will be considerably easier.

Diagnosis of a bile duct injury in the postoperative period should lead to immediate referral to a specialist centre since inappropriate attempts to manage this outwith a specialist centre will compromise the outcome.

Postoperative recognition: biliary obstruction

Ligation or inadvertent clipping of the biliary tree presents early in the postoperative period with jaundice. Later, stricture formation may occur as a result of direct trauma during dissection, clips placed inadvertently on the cystic duct but compromising the bile duct, or from damage to the intricate vascular supply of the bile duct by extensive mobilisation or diathermy. Initial investigation should include haematology, assessment of coagulation by estimation of prothrombin time, and liver function tests. Ultrasound may indicate the level of obstruction or exclude the presence of a correctable cause of obstructive jaundice, such as a retained stone in the common bile duct.

ERCP will identify a stricture or complete transection of the bile duct; however, identification of complete transection with MRCP will avoid the risks of an unnecessary ERCP. Overzealous instillation of contrast should be avoided, and placement of an endoscopic stent should only be considered after consultation with a specialist unit since this may introduce sepsis into the biliary tree and compromise further management. Furthermore, an undrained biliary tree may allow proximal biliary dilatation, thereby facilitating later reconstruction. Although some have reported satisfactory resolution of biliary strictures with endoscopic stenting alone, the follow-up has usually been short and almost all patients require later surgery in our experience. Partial occlusion of the duct by a clip may be remedied by balloon dilatation with or without placement of a stent; however, delay in diagnosis may result in subsequent recurrent stricture formation. Nonetheless, de Reuver et al. reported 110 patients with bile duct strictures following cholecystectomy that were treated with endoscopic stenting, 48 (44%) of which had already undergone attempted surgical repair. At a mean follow-up of 7.6 years, 74% of patients had a successful outcome.³⁷ The development of removable endoscopic expandable metal stents has recently been described, although long-term results and large series are not yet available. Furthermore stent migration can complicate treatment.

If the diagnosis of ductal obstruction is made early within the first week postsurgery, the bilirubin level is only moderately elevated and there is no coexisting coagulopathy or sepsis, immediate repair offers the best chance of a successful outcome.

If repair needs to be delayed, stent placement may still be avoidable and a decision will generally be made based on the individual patient circumstances. Suspicion or evidence of arterial injury may influence the management decision.

For strictures that declare late, appropriate indications for stent placement are the presence of sepsis, severe itch resistant to medical therapy, or significant hepatic dysfunction.

The timing of repair

Early repair: When an injury is recognised in the early postoperative period and there is minimal peritoneal contamination or sepsis, a definitive repair by an experienced surgeon can be successful (Fig. 11.8). In our series of 123 patients referred with injury to the biliary tree, 22 patients underwent primary biliary repair in the first 2 weeks following injury and three had revision of a failed biliary repair. Between 2 weeks and 6 months, a further 22 injuries were repaired selectively. Successful repair was possible in 22 of 25 early repairs compared with 20 of 22 delayed repairs.³⁹

Figure 11.8 Operative picture of an early repair of an E4 injury. A right-angle forceps is placed in the opening of the left hepatic duct whilst the open right hepatic duct is visible below. The portal vein is skeletonised with ligation and excision of both the extrahepatic biliary tree and right hepatic artery (held by forceps).

Delayed repair: Many injuries continue to be unrecognised or referral delayed. In a prospective audit of major bile duct injuries from Australia, the median delay before referral was 9 days (2–28 days), and this included five patients with generalised peritonitis.²⁰

Controlling the biliary injury and associated sepsis is the first treatment aim, which may require endoscopic or percutaneous biliary decompression, allowing jaundice to settle or biliary sepsis to be drained. Intra-abdominal collections may be drained percutaneously, or in the early postoperative period this may be better achieved by laparoscopic means. It is accepted, however, that bile collections are frequently loculated and difficult to eradicate in patients with intra-abdominal sepsis or widespread biliary contamination or peritonitis. The most effective treatment may be laparotomy with extensive lavage and the placement of large intra-abdominal drains. Definitive repair should not be contemplated if there is severe peritoneal soiling since injudicious attempts to repair the injury may aggravate the injury and result in a poor outcome.

Once these objectives have been met, the patient should be allowed to recover from the combined insult of surgery and sepsis. A period of rehabilitation at home is generally required before repair is contemplated in these compromised patients. Abdominal and biliary drainage can be managed on an outpatient basis with community nursing support. Nutritional supplementation may be required, particularly in those who have required a prolonged admission to the intensive care unit and hospital. Attention should be paid to the consequences of prolonged external biliary drainage and consideration given to recycling of bile.

Associated vascular injury: Abdominal CT is required to ensure resolution of intra-abdominal collections and before repair to exclude the presence of liver atrophy. Atrophy can occur from prolonged obstruction to the segmental, sectional or hepatic ducts, but is generally associated with the presence of a vascular injury, most usually of the right hepatic artery. Liver resection may occasionally be needed at the time of definitive repair to remove a source of ongoing sepsis, or if satisfactory reconstruction to the left or right duct is not possible.

Buell et al. identified associated vascular injury as an independent predictor of mortality, with 38% of patients dying compared to 3% (P < 0.001) where no arterial injury was present.⁴⁰ Some authors advocate arteriography before repair to identify such associated vascular injury as a repair is less likely to be successful,⁴¹ or for consideration of hepatic arterial reconstruction at the time of hepatico-jejunostomy.⁴¹^,⁴² However, a recent paper described 55 patients with postcholecystectomy strictures who underwent surgical reconstruction with a left duct approach and preoperative coeliac axis and superior mesenteric artery angiography.⁴³ Twenty-six patients (47%) had an associated vascular injury, of which 20 (36%) were of the right hepatic artery. In this series only one patient in each group (vascular injury vs. no injury) developed a recurrent stricture after repair.⁴³ A proximal anastomosis may offer a better blood supply, minimising the risk of anastomotic stricturing (Fig. 11.9). In support of this theory, Mercado et al. demonstrated that an anastomosis fashioned below the biliary confluence was more likely to require revisional surgery (16%) compared to an anastomosis performed at the biliary confluence (0%; P < 0.05).⁴⁴ Recent improvements in magnetic resonance imaging (MRI) and spiral CT are producing impressive arterial and venous anatomical reconstructions, which may negate the need for invasive arteriography.

Figure 11.9 Anastomotic stricture following repair of biliary injury. Percutaneous transjejunal cholangiogram (PTJC) of a Bismuth 1 injury repaired by hepatico-jejunostomy at the level of the transection of the common bile duct (not to the left hepatic duct). Three months later the patient required reconstruction of the anastomotic stricture.

Injury to the hepatic arterial supply (usually the right hepatic artery) may present with haemobilia or intra-abdominal haemorrhage from a false aneurysm, usually associated with ongoing subhepatic sepsis. If suspected, urgent angiography is required (Fig. 11.10). Haemorrhage may be controlled by embolisation of the feeding vessel, although re-bleeding can occur and necessitate further embolisation. However, in our experience, further bleeding in the presence of ongoing sepsis usually requires laparotomy for control of bleeding and drainage of any subhepatic collection.

Figure 11.10 Digital subtraction angiogram demonstrating a false aneurysm of the common hepatic artery. Embolisation was required for control. The patient has undergone a primary repair for a complete transection of the common bile duct. FA, false aneurysm; GDA, gastroduodenal artery; HA, common hepatic artery; LHA, left hepatic artery; RHA, right hepatic artery.

Rarely, combined injury to the hepatic artery and portal vein can occur with resultant infarction of the affected hepatic parenchyma, usually the right liver. Such injuries may require urgent hepatic resection or transplantation.³¹^,³²^,⁴⁵

Further imaging: For patients with injury to the biliary confluence (E3 and E4), preoperative imaging will help in the planning of future repair. In the presence of a biliary stricture, invasive cholangiography by ERCP or percutaneous transhepatic cholangiogram (PTC) risks introducing sepsis. However, if PTC is required for external biliary drainage, an adequate cholangiogram may be obtained at this time. The quality of MRCP continues to improve, and detailed biliary anatomical reconstructions can be produced, thereby negating the need for more invasive imaging.

Operative techniques

Biliary reconstruction should be performed under optimal circumstances at the time of injury or soon thereafter. Once this opportunity has been lost, repair should only be considered when the patient has been optimised, in the absence of intra-abdominal sepsis, and when sufficient time has elapsed to allow for maturation of adhesions and the tissues at the porta hepatis.

A right subcostal incision is used for access, which can be extended across the midline if required. Retraction is provided with Doyen’s blades and the Omni-tract® (Omni-tract surgical, St Paul, MN) mechanical retractor. Laparotomy is undertaken to assess the liver and to allow adhesiolysis, thereby freeing the small bowel for reconstruction. Frequently the omentum, hepatic flexure, duodenum and hepatoduodenal ligament are involved in a dense inflammatory mass, and occasionally an unsuspected fistula between bile duct and duodenum or colon is identified. Dissection is often easier if commenced laterally and then directed towards the biliary structures. The common bile duct can be difficult to identify, particularly in the presence of extensive fibrosis, and intraoperative ultrasound is a useful tool in allowing its location and relationship to vessels to be determined.

For injuries that involve the biliary confluence, lowering of the hilar plate allows easier identification of the left and right hepatic ducts. This may be aided by the use of an ultrasonic dissector (CUSA), which is also employed to break down the contracted fibrotic tissue in the gallbladder bed and facilitate the division of any bridge of liver tissue between segments 3 and 4. Opening these two planes on the right and left sides facilitates identification of and access to the biliary confluence.

Since the blood supply to the bile duct is often damaged at the time of injury, the common hepatic duct should be opened as proximally as possible, although frequently there has been retraction of the fibrotic remnant superiorly. Extension of the incision into the left hepatic duct allows a wide anastomosis to be fashioned with adequate views of the left- and right-sided ducts. Care should be taken since there may be a small superficial arterial branch crossing the left duct anteriorly and running above to segment 4. For injuries with separation of the confluence, the right and left hepatic ducts can be anastomosed together before formation of a hepatico-jejunostomy, allowing a single biliary anastomosis. If possible, injuries to an isolated right sectoral duct are best repaired or drained into a Roux limb of bowel (Fig. 11.5). Simple ligation will lead to atrophy of the drained segments, which may become a nidus for sepsis. However, enteric drainage of a small sectoral duct may also lead to sepsis if an anastomotic stricture occurs.

Repair should be effected by a hepatico-jejunostomy with a 70-cm Roux limb of jejunum, thereby minimising the risk of enteric reflux and chronic damage to the biliary tree. Moraca et al. advocate hepatico-duodenostomy for biliary injury on the basis that it is more physiological, quicker to perform, and allows later ERCP for imaging and intervention.⁴⁶ They found no difference in outcome following hepatico-duodenostomy when compared with hepatico-jejunostomy, although median follow-up was only 54 months. Hepatico-duodenostomy has largely been abandoned in the treatment of other benign biliary disease due to ongoing enteric reflux. There have been anecdotal reports of the late development of cholangiocarcinoma,⁴⁷ as well as the need to undertake liver transplantation in patients so managed when secondary biliary cirrhosis due to enteric reflux has resulted. Our own view is that hepatico-duodenostomy has no role in the management of bile duct injury.

Fine absorbable interrupted sutures of 4/0 or 5/0 polydioxanone sulphate (PDS II) should be used to fashion an end-to-side hepatico-jejunostomy, with care being taken to produce good mucosal apposition. Some authors advocate the use of an access limb, particularly for E3 and E4 injuries, to allow subsequent radiological intervention for dilatation of recurrent strictures.⁴⁸ However, others believe that advances in percutaneous transhepatic techniques have made this unnecessary and have achieved satisfactory results without using this surgical approach.

Rarely, there may be no recognisable bile ducts visible in the porta hepatis. In such cases a variation of porto-enterostomy (Kasai procedure) can be considered with the Roux limb sutured to the fibrous structure of the hilar plate (S.W. Banting, personal communication).⁴⁹

Partial injury to the biliary tree can be repaired with fine interrupted sutures, although when resulting from diathermy dissection, formal hepatico-jejunostomy may be necessary as conduction of the thermal injury may cause later stricture formation. If a T-tube is placed to protect a primary duct repair, this should be placed through a separate choledochotomy.

Management of complications related to repair

Revisional surgery: Many patients with biliary injury continue to suffer from complications despite reconstruction. Factors such as the experience of the initial surgeon, the level of injury, the associated sepsis and liver atrophy all increase the chance of an unsuccessful repair. Following primary repair of a ductal tear or laceration, further stricture formation may result if there has been extensive dissection around the common hepatic duct. In such instances, surgical revision with the formation of a Roux-en-Y hepatico-jejunostomy is indicated.

The majority of patients requiring revisional surgery will have undergone a previous biliary enteric drainage procedure. Anastomotic stricturing will require revision of the anastomosis, with extension of the choledochotomy into the left hepatic duct (Fig. 11.9).

Liver resection and transplantation: In the acute setting of bile duct injury, long-term damage to the hepatic parenchyma is difficult to predict. Major vascular injury or unrecognised segmental biliary obstruction may lead to atrophy of the liver, chronic intrahepatic infection, abscess formation or secondary biliary cirrhosis. In such patients, careful operative assessment is required; CT should be performed to identify areas of associated liver atrophy and to exclude portal vein thrombosis.

In our experience, the majority of patients requiring liver resection are those with ongoing sepsis in an obstructed segment or those where drainage of the extrahepatic biliary tree is not possible due to sectoral duct damage or fibrosis.³¹ A recent review identified 99 patients (5.6%) requiring hepatectomy among 1756 postcholecystectomy bile duct injury patients,⁵⁰ with combined arterial and Strasberg E4 and E5 injuries more likely to require hepatic resection.⁵⁰ Occasionally, early hepatic resection is required for combined arterial, portal venous and biliary injury, although results are poor.⁴⁵ Very rarely, resection may be needed to gain access to the biliary tree, especially when the injury involves the biliary confluence (E4), although some authors routinely advocate resection of segments IVb and V for access to the right hepatic ducts.⁵¹ The right lobe is most commonly affected by sepsis and atrophy as the right-sided sectoral ducts and arterial supply are more likely to be damaged during cholecystectomy, although both left- and right-sided hepatic resections have been reported in patients with severe biliary injury. Resection of the right liver can be performed, for example at the time of delayed reconstruction if there is any doubt regarding the integrity of the anastomosis to the right sectoral or hepatic duct and when a satisfactory anastomosis can be achieved to the long extrahepatic left duct.

Failed reconstruction and persistent cholangitis may lead to end-stage liver failure within a few years and this may require liver transplantation ³¹ (Fig. 11.11). A long interval between injury and referral is known to be associated with end-stage liver disease.⁵² Rarely, liver transplantation may be needed when the combined biliary and vascular injury is so severe as to preclude attempted reconstruction, although the results are universally poor.³¹

Figure 11.11 Contrast-enhanced CT image of the liver after unsuccessful revisional hepatico-jejunostomy. The surgeon who performed the laparoscopic cholecystectomy performed the hepatico-jejunostomy for an E4 injury. A revisional hepatico-jejunostomy was performed before referral, which was complicated by an anastomotic stricture and portal vein thrombosis. The CT scan shows evidence of right lobe atrophy and splenomegaly as well as a percutaneous biliary drain.

Prognosis

Success of repair: Successful repair has been well described and can be achieved in 90% of patients in a specialised unit.²⁰^,²⁶ As for laparoscopic cholecystectomy, a learning curve for biliary repair has also been described for tertiary centres managing bile duct injury. Over a 20-year period, Mercado et al. reported an improvement with experience and a reduction of post-repair strictures from 13% to 5%.⁵³ As well as anastomotic strictures, liver atrophy and cirrhosis may also occur many years following repair. Predictors of a poor outcome include involvement of the biliary confluence,²⁹^,⁵² repair by the injuring surgeon,³⁴^,⁵⁴^,⁵⁵ three or more previous attempted repairs²⁹ and recent active inflammation.⁵⁵

Survival: Mortality following injury to the biliary tree is significant. Death may follow the acute injury itself, following the biliary repair, or occur later as a result of biliary sepsis or cirrhosis. In a recent report of a nationwide analysis of survival following biliary injury after cholecystectomy, Flum et al. identified 7911 (0.5%) injuries from 1 570 361 cholecystectomies.⁵⁴ Within the first year after cholecystectomy the mortality rate was 6.6% in the uninjured group and 26.1% in those with injury to the common bile duct. The adjusted hazard ratio for death during follow-up was higher for those with an injury (2.79; 95% confidence interval 2.71–2.88). The risk of death increased significantly with advancing age and comorbidities. If the initial repair was performed by the injuring surgeon then the adjusted hazard of death increased by 11%.

Quality of life: Boerma et al. first undertook an assessment of quality of life in patients who had sustained biliary injury or leak that required additional intervention.⁵⁶ Five years after injury, quality of life in the physical and mental domains was significantly worse than controls, despite a successful outcome in 84% of treated patients and regardless of the type of treatment or severity of injury. However, the length of treatment was an independent predictor of a poor mental quality of life. Melton et al. report that quality of life in 89 patients who had undergone biliary repair following laparoscopic cholecystectomy showed no difference in the physical or social domains when compared to controls.⁵⁷ However, in the psychological domain, patients were significantly worse, particularly in the 31% of patients who sought legal recourse for their injury.

Associated malignancy: A small number of reports exist about the development of cholangiocarcinoma at the site of anastomosis 20–30 years following repair.⁴⁷ It is possible that enteric reflux into the biliary tree with sepsis and the production of mutagenic secondary bile salts may be responsible. Furthermore, hepatocellular carcinoma may develop due to secondary biliary cirrhosis (Fig. 11.12).

Figure 11.12 Hepatocellular carcinoma as a consequence of biliary injury. This patient required a liver transplant for secondary biliary cirrhosis, which developed following hepato-duodenostomy for a biliary injury (a). At pathological examination a hepatocellular carcinoma was detected in the explanted liver (b).

Benign biliary strictures

Mirizzi’s syndrome

Mirizzi first described the syndrome of extrahepatic biliary stricture in association with cholelithiasis in 1948,⁵⁸ a condition that occurs in fewer than 0.5% of cholecystectomies.⁵⁹ Obstruction of the common hepatic duct may occur for two reasons: (i) a stone impacted in the cystic duct may cause direct pressure or oedema (Mirizzi type I; Fig. 11.13), or (ii) occasionally the stone may erode through the wall of the gallbladder or cystic duct and into the common hepatic duct (Mirizzi type II).

Figure 11.13 Mirizzi syndrome (type I). Magnetic resonance cholangiopancreatography (MRCP) demonstrating compression of the common bile duct (single arrow) secondary to acute cholecystitis from a stone impacted in Hartmann’s pouch.

Presentation

Diagnosis can be difficult as symptoms may be the same as for acute cholecystitis, but all patients with the disease process will have abnormal liver function tests and some may present with jaundice. Occasionally the diagnosis is made at the time of laparoscopic cholecystectomy.

Management

The investigations may be aimed at excluding a diagnosis of bile duct or gallbladder cancer. Ultrasound assessment will identify dilatation of the biliary tree proximal to the stricture and may even have features suggestive of a Mirizzi syndrome. An ultrasound finding of a decompressed gallbladder with stones involving the common hepatic duct would be more suggestive of a Mirizzi syndrome. An associated mass or lymphadenopathy would be more in keeping with biliary malignancy, but associated sepsis or an empyema of the gallbladder may lead the operator to diagnose Mirizzi’s syndrome as cancer. Both conditions may coexist. ERCP allows endoscopic stent placement to relieve jaundice and may demonstrate a fistula between the gallbladder and common hepatic duct (type II). A smooth, tapered stricture is more typical of a benign rather than malignant cause of jaundice. Endoscopic stenting provides resolution of jaundice, an anatomical roadmap and also may help with identification of the common hepatic duct at operation. Occasionally a stone impacted in the cystic duct may produce a distal biliary stricture. This occurs when there is a low insertion of the cystic duct with stone impaction at the cystic duct/common bile duct junction. It can be difficult to visualise or extract the stone at ERCP and therefore MRCP will be required to confirm the diagnosis.

When Mirizzi’s syndrome is difficult to distinguish from a malignant stricture, CT may aid in the diagnosis. Contrast-enhanced ultrasound may allow better delineation of a benign or malignant cause; however, availability of contrast is limited at present. Occasionally laparoscopic ultrasound may be necessary to further delineate the stricture and exclude tumour dissemination. Although the two conditions may appear similar, vascular invasion may be seen and targeted biopsy may confirm a malignant diagnosis.⁶⁰

Successful completion of cholecystectomy by laparoscopic means has been reported for apparent type I Mirizzi’s syndrome,⁶¹ although this would be inappropriate where there was a clear fistulous communication between gallbladder and common hepatic duct. One of these reported patients has subsequently developed a biliary stricture (O.J. Garden, personal communication). The conventional approach for type I strictures is to perform an open cholecystectomy or to convert from a laparoscopic procedure to allow adequate assessment of the associated biliary stricture. Operative cholangiography should be performed, and in those with persistent strictures a hepatico-jejunostomy should be performed. For type II, where a defect results from the removal of the gallbladder and stent, the common bile duct should be explored. The majority of patients will require a hepatico-jejunostomy, although apparently successful reconstructions using grafts of Hartmann’s pouch have been described.⁶² Long-term results of this innovative approach are awaited.

In a recent report, Schafer et al. identified Mirizzi’s syndrome in 39 of 13 023 patients (0.3%) undergoing a laparoscopic cholecystectomy.⁶³ Thirty-four (87%) patients had a type I Mirizzi syndrome. Of these, 23 patients underwent cholecystectomy alone, 10 patients required bile duct exploration and T-tube insertion, and one patient had a Roux-en-Y reconstruction. Twenty-four of the 34 patients (74%) required open conversion. Of the remaining five patients with a type II Mirizzi syndrome, three underwent a hepatico-jejunostomy and two had simple closure with T-tube drainage. All had required open conversion and, interestingly, four patients (10%) were found to have a gallbladder carcinoma on histopathology.

Mirizzi’s syndrome is an uncommon cause of obstructive jaundice secondary to gallstone disease. This condition will normally necessitate conversion to open surgery if a laparoscopic cholecystectomy is attempted, and a hepatico-jejunostomy will normally be required for a type II lesion.

Hepatolithiasis

Intrahepatic gallstone disease (hepatolithiasis) is also known as oriental cholangiohepatitis or recurrent pyogenic cholangitis. It is most common in Taiwan, south-east Asia and Hong Kong. Symptoms include abdominal pain and jaundice as well as cholangitis.

Recently there has been a decline in incidence, possibly related to improved economic conditions and changes in diet.⁶⁴ The cause is unknown, although Clonorchis sinensis, ascariasis and nutritional insufficiency have been suggested as associated factors in its causation.

Pathologically there is gross irregular dilatation and intrahepatic stricture formation of the biliary tree, which frequently contains stones, debris and pus. Bile duct proliferation, portal and periportal inflammation and fibrosis are seen, and occasionally there is liver abscess formation. Stone formation is frequently associated with bacterial superinfection and the bile is infected in 96% of patients with hepatolithiasis, most usually with Escherichia coli.⁶⁵ There is a strong association between hepatolithiasis and cholangiocarcinoma, particularly in the presence of liver atrophy.⁶⁶ Diagnosis is based on history and demographic features, and investigation includes liver biochemistry and abdominal ultrasonography. Ultrasound is usually diagnostic, although abdominal CT may add further information regarding associated liver atrophy or abscess formation. ERCP will provide important anatomical detail and will allow endoscopic stenting if required. If stricture formation or stones prevent filling of the intrahepatic ducts, MRCP should provide further information, avoiding the risk of PTC.

Primary sclerosing cholangitis is a rare condition, and although the precise cause has yet to be determined there is increasing evidence of an immunological basis as well as an overall increase in the incidence.⁶⁹ Around 70% of patients will have ulcerative colitis, or more rarely Crohn’s disease. There are other reports of association with Riedel’s thyroiditis, retroperitoneal fibrosis, lymphoplasmacytic sclerosing pancreatitis⁷⁰ as well as a strong association with a number of human leucocyte antigens.⁷¹

Presentation

Primary sclerosing cholangitis is a progressive obliterative fibrosis of the intrahepatic and extrahepatic biliary tree with a wide clinical spectrum and frequent remissions and relapses. In the early stages of disease most patients are asymptomatic but later in the disease process patients may have pruritus, ill-defined pain, fever, jaundice and weight loss. Many asymptomatic patients are diagnosed by detection of abnormal liver function tests during the investigation of inflammatory bowel disease. Although some patients may present at an advanced stage, signs of liver failure develop over a period of time. Sudden deterioration may suggest the development of cholangiocarcinoma, with which there is a strong association.

Investigation

Liver biochemistry demonstrates a cholestatic picture. Although antineutrophil cytoplasmic antibodies are present in the majority of patients, testing for autoantibodies is usually performed to exclude primary biliary cirrhosis, a condition from which it can be difficult to differentiate.

The mainstay of investigation is cholangiography, which usually demonstrates a diffuse picture of stricturing and attenuated intrahepatic bile ducts. As well as providing anatomical details of the biliary tree, ERCP enables endoscopic therapy and the opportunity for brush cytology if malignancy is suspected. MRCP is highly sensitive, with a diagnostic accuracy comparable to ERCP, and is now preferred as a means of both diagnosing and assessing the extent of disease to avoid the introduction of bacteria, possibly causing severe biliary sepsis. CT intravenous cholangiography also provides a good alternative when MRCP is contraindicated.

Management

The prognosis of primary sclerosing cholangitis is poor, with a median survival of only 9.6 years from diagnosis to death or liver transplantation.⁷² Survival may improve with earlier diagnosis and liver transplantation; however, subsequent development of cholangiocarcinoma and colorectal cancer has now become the leading cause of death.⁷³ The use of ursodeoxycholic acid in the treatment of pruritus has been associated with improvements in biochemical function and histological appearance.⁷⁴ Episodes of cholangitis can be treated with antibiotics covering biliary pathogens. There is no evidence that colectomy for inflammatory bowel disease alters disease progression.

Endoscopic or transhepatic dilatation of short dominant strictures with or without endoscopic stenting has been described as effective, safe and well tolerated, although no randomised trials have been performed. Dilatation achieves palliation at 1 and 3 years in 80% and 60% of patients, respectively.⁷⁵ In those patients without cirrhosis but with jaundice secondary to a dominant stricture, surgical drainage with an access limb has been described.

Liver transplantation is necessary to treat end-stage liver disease. However, it is now more usual for patients to be considered if there is persistent jaundice, intractable pruritus, recurrent cholangitis, malnutrition or fatigue. Many patients are now transplanted before liver failure, with survival rates of greater than 80% at 5 years.⁷⁶

Exclusion of associated malignant stricture

Cholangiocarcinoma and gallbladder cancer complicate 10–36% of patients with primary sclerosing cholangitis,⁷² and need to be excluded before liver transplantation.

In the majority of patients, concern regarding occult cholangiocarcinoma is small, and liver transplantation is undertaken in the absence of a dominant stricture. Serum carbohydrate antigen (CA) 19-9 has been used in an attempt to identify cases with an occult biliary malignancy. Patients with a sudden rapid deterioration in their clinical state or with a dominant stricture must be considered to have a cholangiocarcinoma and be investigated extensively. Brush cytology at ERCP may provide the diagnosis if a malignant smear is obtained. CT or MRI may demonstrate a mass lesion in association with the biliary tree, although the usual appearance is of a stricture indistinguishable from a benign disease. Positron emission tomography (PET) scanning is superior to conventional radiological investigations to differentiate primary sclerosing cholangitis and cholangiocarcinoma.⁷⁷

Laparoscopy identifies the majority of patients with unresectable biliary tract cancer,⁷⁸ and may be of use in assessing those considered for transplantation in whom a cholangiocarcinoma is suspected since tumour dissemination often occurs early. Laparoscopic ultrasound may further aid assessment, and occasionally laparotomy may be required if there is diagnostic doubt regarding cholangiocarcinoma.

Biliary strictures imitating malignancy

It is not unusual for benign biliary pathology to be found in resected specimens of the pancreatic head that had been thought to be malignant. Around 10% of Whipple resections for malignancy will be found to have benign pathology. Most commonly the pathology is chronic pancreatitis related to alcohol or gallstone disease. However, other confounding pathologies include lymphoplasmacytic sclerosing pancreatitis, primary sclerosing cholangitis, choledocholithiasis and inflammatory pseudotumours.

Up to 14% of patients undergoing surgery for presumed malignant hilar obstruction are found to have a benign fibrotic stricture of the bile duct.⁷⁹ It is usually impossible to differentiate them from malignancy preoperatively and thus resection is often attempted and nearly always feasible. Successful treatment has been described with the use of steroids.

Lymphoplasmacytic sclerosing pancreatitis

Lymphoplasmacytic sclerosing pancreatitis is also known as autoimmune pancreatitis, sclerosing pancreatitis or primary inflammatory pancreatitis. The disease is associated with other autoimmune diseases such as Sjögren’s syndrome, Riedel’s thyroiditis, retroperitoneal fibrosis, ulcerative colitis and primary sclerosing cholangitis.⁸⁰ Although only accounting for about 2.4% of pancreatic resections, the condition is important since a proportion of patients will develop either biliary anastomotic strictures or intrahepatic strictures following resection. In a series of 31 patients, eight (28%) went on to develop recurrent jaundice after resection.⁸⁰

Increased levels of IgG4 have been described in association with the disease, and successful treatment with steroids has been described, although disease recurrence is reported.⁸¹ As yet the value of steroid treatment in the prevention and treatment of recurrent strictures is not well described.

Functional biliary disorders

Most patients who present for investigation of sphincter of Oddi dysfunction have already undergone cholecystectomy for presumed gallbladder pain. However, up to 39–90% of patients with idiopathic recurrent pancreatitis may also have sphincter of Oddi dysfunction.⁸² In those patients with postcholecystectomy pain, the presentation and investigation identifies three types:⁸³

• Type 1 – Abdominal pain, obstructive liver function tests, biliary dilatation and delayed emptying of contrast at ERCP.

• Type 2 – Pain with only one or two of the above-mentioned criteria.

• Type 3 – Recurrent biliary pain only.

Between 65% and 95% of group 1 patients will be found on biliary manometry to have sphincter of Oddi dysfunction compared to only 12–28% of type 3.⁸² Diagnosis is usually by exclusion of other causes of abdominal pain such as peptic ulcer disease and irritable bowel syndrome. Liver function tests, abdominal ultrasonography, CT, endoscopy and MRCP have often been already performed. Morphine–neostigmine and secretin provocation MRCP studies may also be of diagnostic value.

At ERCP, biliary manometry is not required if there is delayed drainage of contrast in type 1 or 2 patients. This investigation should be reserved for those patients in whom the diagnosis remains unclear.

Medical therapy with calcium channel blockers, nitrates and botulinum toxin is available but long-term results are unknown. Avoidance of opiate analgesia, particularly over-the-counter preparations containing codeine, may prevent the onset of pain in the majority. Endoscopic sphincterotomy is a potential treatment; however, 5–16% of patients will develop postprocedural pancreatitis ⁸⁴ and good or excellent responses are only reported in 69% of patients at long-term follow-up.⁸⁵ Surgical sphincterotomy is now indicated rarely due to the lower cost and lower morbidity of endoscopic sphincterotomy but it may be required if the endoscopic approach has been unsuccessful.

Key points

• Choledochal cysts should be treated with complete cyst excision and hepatico-jejunostomy due to the risk of malignancy in the remaining biliary epithelium.

• Identification of the biliary anatomy and minimisation of diathermy near the common bile duct are essential during laparoscopic cholecystectomy to avoid biliary injury.

• Operative cholangiography is useful for delineating the biliary anatomy during cholecystectomy; however, many cholangiograms are not interpreted correctly at the time of biliary injury.

• Following laparoscopic cholecystectomy, any patient who is not fit for discharge at 24 hours due to ongoing abdominal pain, vomiting, fever or bile in an abdominal drain should be considered to have a biliary leak.

• Diagnosis of a bile duct injury in the postoperative period should lead to immediate referral to a specialist centre since inappropriate attempts to manage this outwith a specialist centre will compromise the outcome.

• In the absence of sepsis, repair of injuries to the biliary tree can be performed successfully within the first week.