Postoperative Biliary Stricture
The introduction and widespread use of laparoscopic cholecystectomy in the 1990s resulted in a significant increase in the frequency of biliary injuries and associated bile duct strictures. Postoperative bile duct injuries may present early in the postoperative period with biliary leak, or months to years later with jaundice or cholangitis from biliary stricture. Proper management begins with delineation of biliary anatomy followed by repair. Nonoperative balloon dilation via percutaneous transhepatic or endoscopic routes is appropriate in select patients with intact biliary-enteric continuity. Operative repair, however, remains the mainstay of treatment in patients with benign strictures.
Most bile duct injuries and strictures occur in patients following abdominal surgery in the right upper quadrant. Cholecystectomy is performed on over 750,000 patients on an annual basis in the United States and accounts for over 90% of postoperative biliary strictures and injuries. Although the exact incidence of injuries is unknown because many cases go unreported, numerous studies have attempted to define the incidence and mechanisms of bile duct injuries associated with cholecystectomy. An incidence of one to three major bile duct injuries per 1000 cases was consistently reported during the era of open cholecystectomy. Roslyn and colleagues demonstrated a 0.2% incidence of major bile duct injuries from a series of over 42,000 open cholecystectomies.44 A literature review by Strasberg and associates45 of over 25,000 open cholecystectomies performed since 1980 revealed a 0.3% incidence of major bile duct injuries. In contrast, in a review of nearly 125,000 laparoscopic cholecystectomies reported in the literature in the years 1991–1993, Strasberg and associates reported an overall incidence of biliary injuries of 0.85% and an incidence of major injuries of 0.52%. Multiple large surveys from numerous centers have estimated the rate of major bile duct injury with laparoscopic cholecystectomy to be 0.4–1.3%.46–49 Therefore it appears that the incidence of bile duct injury associated with laparoscopic cholecystectomy is two to three times greater than that with open cholecystectomy.
In the early 1990s, many authors ascribed the increased incidence of bile duct injuries with laparoscopic cholecystectomy as a “learning curve” associated with the new technique and projected that the rate of injury associated with laparoscopic cholecystectomy would decline with time. Unfortunately, these projections were not correct, and the rate of bile duct injuries appears now to have stabilized at a level still higher than that of the prelaparoscopic era. A report of over 10,000 cases at US military institutions46 and nationwide reviews in New Zealand50 and Italy48 have demonstrated no significant improvement in the incidence of injury as surgeons have passed through the learning curve. It is likely that the technology and technique associated with laparoscopic cholecystectomy will need fundamental enhancements for the current rate of injury to diminish.
Several factors are associated with increased risk of bile duct injuries at the time of cholecystectomy. Some of these factors may be pathologic, anatomic variations, and/or technical problems that are unique to the laparoscopic approach. Ultimately, the final common pathway of most injuries is either a technical error or misinterpretation of the anatomy. The “classic” biliary injury during laparoscopic cholecystectomy includes misidentification by the surgeon of the common bile duct as the cystic duct or misidentification of an aberrant right sectoral duct as the cystic duct (Fig. 50-7).
Classic laparoscopic bile duct injury. Confusion of the common bile duct with the cystic duct leads to clipping and division of the common bile duct. In many cases the common hepatic duct will not be clipped but will instead be divided by scissors or cautery. (Reproduced, with permission, from Davidoff AM, Pappas TN, Murray EA, et al. Mechanisms of major biliary injury during laparoscopic cholecystectomy. Ann Surg. 1992;215:196.)
A number of patient-related factors have been associated with bile duct injury. Patients with acute cholecystitis may have severe inflammation in the porta hepatis and Calot's triangle, which can make a laparoscopic approach difficult. Patients with complicated gallstone disease also have a higher risk of injury than those with chronic cholecystitis, symptomatic cholecystitis, or biliary colic. Fletcher and colleagues47 reported that complex cases, which included patients with acute cholecystitis, cholangitis, and gallstone pancreatitis, are associated with an increased incidence of bile duct injuries (1.7 vs 0.6%) versus other indications for laparoscopic cholecystectomy. These patients are also associated with a higher rate of conversion to open cholecystectomy (29 vs 8%).
Anatomic variations can also contribute to bile duct injury. A congenitally short cystic duct or a duct that appears shortened by an impacted stone may also lead to misidentification of the common bile duct, resulting in injury or transection. Other high-risk congenital anatomic anomalies include a long common wall between the cystic and common bile duct or the cystic duct inserting into the right hepatic duct. The cystic duct has a very variable pattern ranging from joining the common hepatic duct quite high, almost at the biliary confluence, or running parallel to the common hepatic duct before inserting into the common bile duct almost at the level of the pancreas. The risk of bile duct injury also appears to be increased in patients with obesity, chronic inflammation, excessive fat in the dissection area, inadequate exposure, poor or excessive clip placement, injudicious use of electrocautery, and bleeding into the operative field.
Several technical factors associated with laparoscopic cholecystectomy make it prone to bile duct injury. First, standard laparoscopy gives a limited perspective from its end, viewing a two-dimensional picture of the operative field. The classic laparoscopic injury occurs when the cystic duct and the common bile duct are aligned in the same plane, leading to clipping and dividing the common bile duct. Retraction of the gallbladder infundibulum excessively cephalad aligns the cystic and common bile duct, leading to misidentification and injury. As the operative dissection is carried cephalad, the common hepatic duct may also be transected, often without recognition, resulting in a postoperative bile leak. The right hepatic artery may also be injured, creating excessive bleeding. This classic injury is estimated to occur in over 75% of major bile duct injuries referred to major centers. The classic laparoscopic injury is usually also associated with excision of a segment of bile duct, making the proximal extent of the injury high, usually at or near the hepatic duct bifurcation.
There is also a growing understanding of surgeon cognitive factors associated with bile duct injury during laparoscopic cholecystectomy. A report examined 252 laparoscopic cholecystectomy bile duct injuries using the human error factor and cognitive science techniques and found that 97% of injuries were due to a visual perceptual illusion or inadequate visualization.51 In a subsequent study from the same group, one of the main explanations for the surgeon's frequent inability to recognize a bile duct injury associated with laparoscopic cholecystectomy appears to be confirmation bias, which is the propensity to seek clues to confirm a belief and to discount clues that might discount that belief.52 While cognitive factors are important for understanding the psychological issues associated with bile duct injuries, surgeons must continue to have appropriate corrective mechanisms in place to minimize the chance of these injuries, including knowledge of anatomy, typical mechanisms of injury, appropriate level of suspicion, and logic.53
The role of intraoperative cholangiography in preventing bile duct injury remains controversial, with mixed results from reported series. A large series in Australia demonstrated a protective effect,47 whereas a review from the Veteran's Administration Hospitals demonstrated that bile duct injury occurred more commonly in patients undergoing cholangiography (0.7 vs 0.2%).45 Clinical information from patients in the Medicare claims database and surgeon data from the American Medical Association Physician Masterfile were recently used to examine the influence of intraoperative cholangiography on the rate of major bile duct injury, finding the rate of injury to be significantly higher when intraoperative cholangiography was not used.54 In this study, surgeons who routinely performed intraoperative cholangiography had a lower rate of injuries than those who did not, and this lower rate disappeared when intraoperative cholangiography was not used by these surgeons. Whether or not intraoperative cholangiography actually prevents bile duct injury, the procedure can often lead to early recognition of the injury, and therefore potentially minimize the injury and its associated morbidity (Fig. 50-8). The best technical approach in preventing and limiting bile duct injuries, regardless of the use of cholangiography, includes methodical dissection with careful exposure and identification of the structures of the triangle of Calot.44
Intraoperative cholangiogram obtained during laparoscopic cholecystectomy. Cholangiogram demonstrates an injury to the common bile duct (which is clipped such that contrast does not fill the proximal biliary tree). Contrast fills the normal distal bile duct and duodenum.
The operative technique for laparoscopic cholecystectomy, which defines the “critical view of safety,” is a corrective mechanism that helps prevent misidentification and injury to the major bile ducts.55 In this method, the triangle of Calot is cleared of fat and fibrous tissue. Only two structures are connected to the lower end of the gallbladder once this is done; the cystic duct and cystic artery and the lowest part of the gallbladder attachment to the liver have been exposed. Once the critical view is attained, the cystic duct and artery may be clipped and divided, as they have been conclusively identified. Failure to achieve the critical view is an indication for conversion or possible cholangiography.
Several physiologic processes have been implicated in the formation of bile duct strictures. Ischemia of the bile duct from excessive periductal dissection may have an important role in the formation of postoperative anastomotic strictures. Studies show that the blood supply to the ducts can be thought of having three elements: afferent arteries, marginal arteries, and the epicholedochal plexus. The afferent arteries are branches of the hepatic arteries or less commonly of the superior mesenteric artery or other upper abdominal arteries. The marginal arteries lie on and run parallel to the long axis of the bile ducts. Anatomically, these are the major arteries of the common bile duct located at the 3- and 9-o'clock positions that can be injured or divided by unnecessary dissection during cholecystectomy, or more commonly the bile duct can be excessively “skeletonized” while performing a bile duct anastomosis.
Fibrosis and scarring can be intense following a bile duct injury. In canine models, bile duct ligation results in an elevation of bile duct pressure that is immediate and sustained and is accompanied by an increased bile duct diameter and formation of high local concentrations of bile salts at the canalicular membrane.56 A month after bile duct ligation, the bile duct wall is thickened, will have reduced mucosal folds, and have loss of surface microvilli with epithelial degeneration. On pathologic staining 2 weeks after ligation, there is evidence of increased synthesis of collagen and proline hydroxylase activation. Recently, an animal model of bile duct injury demonstrated healing in traumatized bile duct tissue to occur in a mode of overhealing, implicating myofibroblasts as the main cause of contracture of scar and stricture of the bile duct.57 Inflammation in the surrounding tissues compounds the problem by encouraging fibrosis, especially when associated with bile leakage.
Injuries and strictures of bile ducts occur less commonly in association with other operative procedures. After cholecystectomy, common bile duct exploration is the next most frequently associated procedure with stricture, typically occurring at the site of choledochotomy or an impacted stone. Procedures requiring biliary-enteric anastomoses may be complicated by postoperative stricture. Typically, these procedures involve choledochoenteric or hepaticoenetric anastomosis in such cases as reconstruction after pancreaticoduodenectomy, bile duct resection for mid–bile duct tumors, and after excision of choledochal cysts. Gastrectomy and hepatic resection are the most common nonbiliary operations associated with postoperative strictures. Injuries associated with gastrectomy typically occur during pyloric and proximal duodenal dissection associated with closure of the duodenal stump or with creating a Billroth I gastroduodenostomy. Injuries during hepatic resection often take place during dissection of the hepatic hilum. Bile duct injury and stricture is also associated with hepatic transplantation, pancreatic procedures, and penetrating or blunt trauma. Finally, the recurrence of stricture after an initial attempt at repair is not uncommon and may occur over a decade following initial repair (Fig. 50-9).58
The cumulative percentage of recurrent strictures is shown with respect to the time interval from the initial repair to the next repair. (Adapted, with permission, from Pitt HA, Miyamoto T, Parapatis SK et al. Factors influencing outcome in patients with postoperative biliary strictures. Am J Surg 1982;144:14–21.)
Strictures and injuries to the bile duct vary widely in their complexity and nature. The ease of management, operative risk, and outcome of biliary injuries vary considerably depending on the location and the type of injury. Injuries associated with laparoscopic cholecystectomy are often complex, located at or near the level of the hepatic duct bifurcation, and potentially include one or more hepatic duct branches. Minor injuries to the bile duct include lacerations of the bile duct, clip placement on an intact bile duct, injury via electrocautery, or avulsion of the cystic duct.
A number of classification systems or major bile duct strictures have been presented, with the traditional classification being that described by Bismuth (Fig. 50-10), which classifies major injuries based on the level of obstruction of the biliary tree regarding the hepatic duct confluence or the involvement of an aberrant right sectoral hepatic duct with or without a concomitant hepatic duct stricture.59 A drawback of the Bismuth classification system is that patients with limited strictures, isolated right hepatic duct strictures, or cystic duct leaks cannot be classified. The Strasberg classification system has been developed to classify all types of injury and is used extensively in describing bile duct injuries associated with laparoscopic cholecystectomy (Table 50-3).55
Bismuth classification system. Classification of bile duct strictures based on the level of the stricture in relation to the confluence of the hepatic ducts. Types III–V are usually considered complex injuries. (Reproduced, with permission, from Bismuth H. Postoperative strictures of the biliary tract. In: Blumgart LH, ed. The Biliary Tract. Clinical Surgery International Series, Vol. 5. Edinburgh, Scotland: Churchill Livingstone; 1983:209–218.)
Table 50-3: Strasberg Classification of Biliary Injury and Stricture ||Download (.pdf)
Table 50-3: Strasberg Classification of Biliary Injury and Stricture
|Class A||Injury to small ducts in continuity with the biliary system, with cystic duct leak|
|Class B||Injury to sectoral duct with consequent obstruction|
|Class C||Injury to sectoral duct with consequent bile leak|
|Class D||Lateral injury to extrahepatic ducts|
|Class E1||Stricture >2 cm distal to bifurcation|
|Class E2||Stricture <2 cm distal to bifurcation|
|Class E3||Stricture at bifurcation|
|Class E4||Stricture involving right and left bile ducts; ducts are not in continuity|
|Class E5||Complete occlusion of all bile ducts|
Most patients with bile duct injuries unfortunately are not recognized at the time of laparoscopic cholecystectomy. After open cholecystectomy, only 10% of injuries are suspected after the first week, but nearly 70% are diagnosed within the first 6 months after operation.58 However, injuries after laparoscopic cholecystectomy are recognized earlier more likely because of heightened awareness and suspicion.
Large series reviews have demonstrated that less than one-third of major bile duct injuries are detected at the time of injury during laparoscopic cholecystectomy.60,61 Possible indications that a bile duct injury had occurred intraoperatively include a persistent and unexpected bile leak, atypical anatomy, or a second bile duct discovered during dissection. Injuries may also be discovered if the removed gallbladder specimen and cystic duct are carefully examined to ensure normal duct anatomy. Intraoperative cholangiography will also diagnose bile duct injuries at the time of cholecystectomy and may minimize injury, allowing early repair (see Fig. 50-8).
The clinical presentation of patients with a bile duct injury in the early postoperative period depends on the type of injury. In most cases the injury is associated with uncontrolled bile leakage into the peritoneal cavity, while in others the duct is completely ligated by clip placement leading to obstructive jaundice usually without cholangitis. Patients with significant bile leaks generally present within the first week after operation with abdominal pain, distention, nausea, vomiting coupled with fever, or other signs of sepsis. Prompt investigation is required if patients have bilious drainage from incision sites or from intraoperatively placed drains. Bile leaks result in either biliary ascites with associated chemical peritonitis if allowed to drain freely into the abdominal cavity or, alternatively, bile can become loculated resulting in biloma (Fig. 50-11) or, if infected, a subhepatic or subdiaphragmatic abscess. In the latter scenario, presentation is more subtle with low-grade fever and localized abdominal pain. Because significant abdominal complaints are uncommon after uncomplicated laparoscopic cholecystectomy, all patients with such symptoms should be appropriately evaluated without delay for possible bile leak to prevent progression to frank sepsis. Failure to recognize a major bile leak or to institute appropriate treatment can result in life-threatening sepsis and the development of multisystem organ failure. In a recent series of 200 major bile duct injuries treated at The Johns Hopkins Hospital, three patients were transferred to this tertiary care center and died of complications of sepsis secondary to delayed or inadequate treatment.62
CT scan demonstrating biloma associated with biliary leak after bile duct injury. (Reprinted, with permission, from Lillemoe KD, Pitt HA, Cameron JL. Postoperative bile duct strictures. Surg Clin North Am. 1990;70:1362.)
Bile duct strictures may also present months to years after the original operation. Patients with a slowly evolving stricture may have nonspecific abdominal complaints, jaundice, pruritus, cholangitis, or derangements in liver function tests. In addition, patients with an isolated right sectoral hepatic duct injury may present with a history of unexplained fevers, pain, or generalized malaise. Episodes of cholangitis are typically mild and respond effectively to antibiotics. Less often, patients can present with painless jaundice, which can be confused with a malignant stricture.
The findings on physical examination are usually not specific. Abdominal distention and pain may be seen in patients with bile peritonitis or focal tenderness if the patient presents with a collection or abscess. If the patient has jaundice, there may be multiple excoriations from pruritus. Hepatomegaly may be present in patients with chronic biliary obstruction or possible splenomegaly if there is any portal hypertension from portal venous injury or severe underlying hepatocellular damage.
Patients presenting with a biliary leak from injury usually present without evidence of biliary obstruction, and bilirubin levels are normal or slightly elevated due to absorption of bile from the peritoneal cavity. Patients with postoperative bile leak or cholangitis will also have an elevated white blood cell count, pyrexia, or occasionally frank sepsis. Patients with postoperative bile duct strictures typically reveal a stereotypical biochemical profile of cholestasis. In particular, liver function tests typically consist of an elevated alkaline phosphatase and normal or slightly elevated liver transaminases (alanine and aspartate aminotransferases). Serum bilirubin levels are usually elevated in the range of 2–6 mg/dL. In rare cases, patients with long-term obstruction will present late in the course of disease with cirrhosis, diminished serum albumin, and abnormal coagulation studies from altered hepatic synthetic function.
Definitive diagnosis for bile duct strictures and injuries requires radiographic imaging. Ultrasound and abdominal CT scan are both helpful in patients who present in the early postoperative period for the detection of bilomas and biliary ascites, as well as bile duct dilation from obstruction. Ultrasound has little value in assessing the extent of a stricture and is unhelpful if the biliary tree is decompressed. Abdominal CT scan is usually the best first-line study often showing a dilated biliary tree or intra-abdominal collections or ascites, which can direct further investigations. The CT should be performed with arterial-phase contrast to evaluate for concomitant vascular injury. Nuclear medicine imaging with technetium-HIDA (hepatobiliary iminodiacetic acid) scanning can demonstrate bile leakage noninvasively but typically does not have the sensitivity to define the specific anatomic site of injury. MRCP has been demonstrated to be an effective noninvasive method for demonstrating biliary leakage or obstruction, as well as precisely defining biliary anatomy and the nature of the injury (Fig. 50-12). Last, sinography, typically performed by injecting water-soluble contrast via operatively placed drains, can define the biliary anatomy and the source of bile leakage.
Diagnostic magnetic resonance cholangiopancreatography (MRCP) demonstrating biliary anatomy associated with a cystic duct leak after laparoscopic cholecystectomy. There is an intact biliary system with extravasation of contrast in the subhepatic space.
Cholangiography currently remains the gold standard for evaluating the biliary tree. Endoscopic retrograde cholangiography (ERC) is performed via a distal approach to the biliary tree and is useful only in patients if the native bile duct is intact, such as with partial injuries or after end-to-end repair. ERC is the procedure of choice for patients suspected of cystic duct leaks (Fig. 50-13A) or leaks from peripheral hepatic radicals (ducts of Luschka). In these cases, the biliary leak may be effectively controlled with the use of an endoprosthesis. Most cases of major bile duct injury, however, are associated with complete duct transection, and the cholangiogram via the retrograde endoscopic route will demonstrate a normal distal bile duct terminating in misapplied clip(s) devices (Fig. 50-13B). Therefore, ERC will neither define the site of bile leakage nor the proximal anatomy necessary for reconstruction. In such cases, PTC is necessary to define the proximal biliary anatomy and the site of injury (Fig. 50-14). In addition to delineating the anatomy, a percutaneous biliary drainage catheter should be placed at the time of PTC to decompress the biliary tree, and treat cholangitis and control the biliary leak. Percutaneous biliary drainage catheters will also be useful at the time of operative repair as a guide for dissection and identification of the transected bile duct, which is often retracted high into the liver hilum. Finally, in those cases in which biliary-enteric continuity exists, percutaneous catheters allow access for balloon dilation.
A. Endoscopic retrograde cholangiopancreatogram demonstrating cystic duct leak. B. Endoscopic retrograde cholangiopancreatogram with multiple clips across the common bile duct without visualization of the proximal biliary tree in a patient with total transection of the common bile duct during laparoscopic cholecystectomy.
Percutaneous transhepatic cholangiogram in a patient with complete transection of the common hepatic bile duct. Note the surgical clips near the cutoff point.
Significant arterial injury associated with major bile duct injury has been increasingly reported in recent years. The “classic” biliary injury during laparoscopic cholecystectomy in which the common bile duct is mistaken for the cystic duct often includes injury to the right hepatic artery as it enters either above or below the hepatic duct. While this injury may cause bleeding at the time of operation, the arterial injury often is unnoticed, usually resulting in arterial occlusion or less commonly a hepatic artery pseudoaneurysm. In a large study by Stewart et al63 on combined right hepatic artery and bile duct injury, there were 7 pseudoaneurysms compared to 77 right hepatic artery occlusions. The incidence of disruption of the right branch of the hepatic artery during major bile duct injury ranges between 12 and 39%.64 However, the presence of an arterial injury does not appear to affect either early or late outcomes.63,65 Because of the recognized association of vascular injuries during laparoscopic bile duct injuries especially if there is a history of excessive bleeding at the time of cholecystectomy, a CT scan with arterial and venous phase contrast or arteriography should be obtained. Some authors believe if arterial injury has occurred, biliary reconstruction should be delayed to decrease the risk of late stricture recurrence.55 In patients presenting in a delayed manner after cholecystectomy, the combination of biliary and vascular injuries often leads to segmental or lobar atrophy, which may suggest a role for hepatic resection rather than reconstruction.
The timing of presentation is often a primary determinant of the preoperative management of a patient with a postoperative bile duct stricture or injury. In the early postoperative period, patients with a bile leak associated with a bile duct injury are often either septic due to intra-abdominal infections or otherwise manifesting a systematic inflammatory response from chemical peritonitis associated with the bile leak. Treatment and control of sepsis may require broad-spectrum parenteral antibiotics, percutaneous biliary drainage, and percutaneous or, rarely, operative drainage of bilomas. Once sepsis is controlled, there is no hurry in proceeding with surgical reconstruction of the bile duct injury. Most biliary fistulae can be controlled with the combination of proximal biliary decompression and external drainage. After early control and clinical improvement, the patient may be discharged home for several weeks to permit return of overall health and for the resolution of inflammation in the periportal region.
It should be stressed that despite the belief of many surgeons that a suspected bile leak warrants urgent reoperation, exploration with an attempt at repair should be avoided early after presentation with a bile leak. In this situation exploration often reveals marked inflammation associated with bile spillage and small, decompressed bile ducts retracted high into the porta hepatis, making recognition of the injury and repair virtually impossible. Instead of proceeding to urgent exploration, a more prudent approach is to define biliary anatomy via preoperative cholangiography and to control the bile leak with percutaneous stents. Early operative intervention to deal with bile collections or ascites is not usually required because the intraperitoneal bile either can be drained percutaneously or is simply absorbed by the peritoneal cavity. Delayed reconstruction, with facilitation by percutaneous biliary catheters, allows for the most favorable operative results especially when concurrent hepatic artery injury is suspected.
Patients who present with a biliary stricture remote from the initial operation usually experience symptoms of cholangitis that necessitate urgent cholangiography and biliary decompression. The choice of technique depends on the nature of any prior repair. If the native bile duct is intact, endoscopic drainage with stent placement can sometimes be achieved. If a prior hepaticojejunostomy has been performed, transhepatic biliary drainage will be necessary for diagnosis. Both parenteral antibiotics and biliary drainage are central to controlling sepsis. Patients who present with jaundice without cholangitis should undergo either ERC or PTC to define the anatomy. As with patients presenting early in the postoperative period, ERC may not completely define the proximal biliary anatomy, making PTC the more favorable procedure. Preoperative biliary decompression in patients presenting with jaundice without cholangitis has not been demonstrated to improve outcome.
Operative repair for postoperative bile duct strictures is aimed at reestablishing a reliable, long-term conduit for bile flow from the biliary tree to the gastrointestinal tract. Complications of an unsuccessful operative procedure include bile leak resulting in fluid collection or abscess, recurrent stricture with stones or sludge and potentially cholangitis, or biliary cirrhosis. To this end, the ideal technical procedure results in a tension-free, mucosa-to-mucosa repair to a segment of uninjured bile duct. Ideally, surgeons should also seek to maintain ductal length by not sacrificing tissue. Options for operative repair may include end-to-end repair, Roux-en-Y hepaticojejunostomy, or choledochoduodenostomy. The optimal operative procedure is contingent upon the timing of presentation, overall clinical status of the patient, and level and type of injury.
Injury Recognized at Initial Operation.
If injury to the bile duct is recognized at the time of initial cholecystectomy, the surgeon should consider his or her ability to technically perform immediate reconstruction and should consider seeking the counsel and assistance of a more experienced surgeon. Studies show that immediate open repair by an experienced surgeon is associated with reduced morbidity, shorter duration of illness, and lower cost.62 Each failed attempt at repair is associated with loss of bile duct length and exacerbation of a difficult situation. If the surgeon is unable to repair the injury and competent help is unavailable, drains should be placed to control any bile leak and the patient referred immediately to a tertiary specialty center.
When the surgeon suspects an injury or variant anatomy, biliary anatomy must be clearly defined using intraoperative cholangiography and/or careful dissection, being cautious to avoid additional injury or devascularizing the bile duct. Conversion from laparoscopic to open cholecystectomy is often necessary to properly identify anatomy and the injury. Segmental or accessory duct injuries where the diameter of the bile duct is less than 3 mm and where the bile duct does not communicate with the major duct system or drain a large segment of hepatic parenchyma on cholangiography may be ligated. Bile ducts that are 4 mm or larger in diameter or when the cholangiogram shows sectoral or lobar drainage, then the ducts must be operatively repaired, as they likely drain multiple hepatic segments or an entire liver lobe.
Immediate intraoperative repair is indicated in most cases for a major injury to the common hepatic or common bile duct. The nature of that repair is determined by the length of separation between opposed residual, viable ends of the injured duct. Partial common duct transections, involving less than 180-degree circumference of the biliary tree, may be closed primarily over a T-tube using interrupted absorbable sutures (Fig. 50-15). Transection of the common duct involving more than 180-degree circumference or complete transactions with an injury less than 1 cm in length can usually be repaired with an end-to-end anastomosis with a T-tube that exits either above or below the anastomosis via a separate choledochotomy. Primary reconstruction of the bile duct, however, should be used very selectively and be avoided when the injury is near the bifurcation or when duct approximation cannot be accomplished without tension. A generous Kocher maneuver should be done to mobilize the duodenum out of the retroperitoneum and should be used to alleviate tension at the repair. In at least one series, a 100% restricture rate following primary end-to-end repair has been reported.66 Other series have shown better results and suggest an advantage that, if a stricture occurs, endoscopic access for balloon dilation remains an option.67
Primary end-to-end repair of the biliary tree over a T-tube. In general, this technique is used for partial transections of the bile duct, when there has been no associated loss of duct length. Note that the T-tube does not exit at the site of injury.
Transections of the bile duct high in the biliary tree or with significant loss of bile duct length cannot be repaired with a primary biliary anastomosis that remains tension-free. These injuries require reconstruction using a biliary-enteric anastomosis typically using Roux-en-Y hepaticojejunostomy to ensure a tension-free repair. In this situation, the distal bile duct should be oversewn, the injured tissue in the proximal end debrided, and then a biliary-enteric end-to-side anastomosis to the Roux-en-Y jejunal limb. Transhepatic Silastic biliary stents should be placed to control potential anastomotic leaks and for postoperative cholangiography. A perianastomotic drain should also be placed in all cases so that any potential postoperative leak is well-controlled.
Injury Recognized in the Immediate Postoperative Period.
Biliary injuries that are not appreciated in the intraoperative period may present in the first few days. The presentation may include bile drainage from the wound, bile peritonitis, or progressive jaundice. The initial management of a patient who presents in the delayed fashion following laparoscopic cholecystectomy depends on the nature of the injury and the mode and timing of presentation. Any elective repair should generally occur only after preoperative clinical optimization of the patient, and exact anatomy of the biliary system has been identified. Those presenting with biliary leak should have the bile leak and sepsis controlled prior to having definitive repair. In this situation, the result of reconstruction is almost always better if the definitive repair is made well after the leak and the consequent intra-abdominal inflammation and sepsis are controlled with percutaneous biliary drainage. Biliary spillage and marked inflammation can obscure fields and can make identification of ducts difficult making urgent early laparotomy prior to biliary decompression problematic. Finally, the patient should be clinically stabilized prior to elective repair to correct fluid and electrolyte balances, anemia, and malnutrition. The repair is ideally performed 6–8 weeks after adequate control of the leak has been attained.
In patients who present with biliary stricture weeks to months after cholecystectomy, identification of the biliary system is also essential. Patients with a stricture and symptoms of cholangitis should be treated with broad-spectrum antibiotics until sepsis is controlled, followed by biliary decompression with transhepatic percutaneous catheter placement.
Definitive Management of Bile Duct Stricture.
The goal of operative management of a bile duct stricture is the establishment of bile flow into the proximal gastrointestinal tract in a manner that prevents sludge, stone formation, cholangitis, restructure, and cirrhosis. The type of repair should be determined by several factors: previous history of attempted repair, location of stricture or injury, surgeon experience, and surgeon preference. Intraoperatively, biliary anatomy must be carefully defined followed by exposure of healthy proximal bile ducts. Care must be taken to avoid excessive dissection and devascularization of tissue. A biliary-enteric anastomosis is performed using a mucosa-to-mucosa technique in a tension-free manner.
The preferred technique, with few exceptions, is a hepatico- or choledochojejunostomy to a Roux-en-Y limb of jejunum. End-to-end anastomosis after excision of the stricture or area of injury is not prudent because of the loss of bile duct length and associated fibrosis. Significant loss of bile duct length is also a strict contraindication to performing choledochoduodenostomy, which is unlikely to be performed in a tension-free fashion and is also associated with duodenal fistula if leak occurs.
The exact details of the reconstruction depend on the particular anatomic features of the stricture. For strictures where there is more than 2 cm of healthy common hepatic duct present (Bismuth I), a simple end-to-side biliary-enteric anastomosis will suffice. For strictures in where there is less than 2 cm of healthy common hepatic duct (Bismuth II) or the stricture involves the bifurcation of the hepatic duct but the left and right still communicate (Bismuth III), it may be necessary to lower the hilar plate and extend the dichotomy along a short length of the left hepatic duct to allow a common biliary-enteric anastomosis. Strictures that completely separate the right and left biliary system (Bismuth IV and V) require separate right and left biliary-enteric anastomosis. When duct length cannot be found outside of the hepatic parenchyma, an intraoperative ultrasound is essential to locate the segments II and III ducts. Often, a wedge of liver may need to be resected until an adequate duct can be found to do a biliary-enteric anastomosis.
The use of percutaneous biliary stents with elective reconstruction of the biliary tree remains a topic of debate for hepatobiliary surgeons. Preoperatively placed stents act as intraoperative aids for defining anatomy, especially if the stricture is located proximally. Stents left in place after reconstruction also allow postoperative cholangiography and control early anastomotic leaks in the immediate postoperative period. Many surgeons also advocate extended postoperative transanastomotic stenting, with the purpose of minimizing fibrosis and risk of late anastomotic stricture. In this setting, follow-up cholangiography will reveal early evidence of anastomotic stricture and provide access for balloon dilation if necessary.
Biliary reconstruction with the technique of hepaticojejunostomy with a Roux-en-Y limb with transhepatic biliary stents is depicted in Fig. 50-16. Dissection of the porta hepatis is performed to clear any adhesions between the duodenum or colonic hepatic flexure to the gallbladder fossa, subhepatic space, or Glisson's capsule. Preoperatively placed percutaneous stents are essential in assisting in dissection and bile duct identification in patients with a high bile duct transection. In patients with an intact but strictured bile duct, the duct is divided at the most distal portion of the stricture, and a segment of the strictured duct should be resected and sent to pathology for frozen section. The distal end of the stricture is then oversewn. The proximal extent of the duct should be debrided for a length not to exceed 5 mm to obtain healthy bile duct circumferentially for use in the anastomosis. Careful limited dissection is important to avoid vascular compromise to the bile duct. Preoperatively placed percutaneous transhepatic catheters, which now protrude from the proximal end, are usually exchanged for soft Silastic stents. Silastic stents range from 12 to 22F in size, with multiple side holes that are generally interspersed along 40% of the length of the catheter. A radiologic guidewire is placed through the percutaneous transhepatic catheter; using the Seldinger technique, a series of progressively larger coudé catheters are passed over the guidewire in order to dilate the system for Silastic stent placement. The Silastic stent is arranged with the side holes extending beyond the anastomosis distally and within the liver parenchyma proximally. The end of the Silastic stent without holes is brought through the hepatic parenchyma and out through the upper anterior abdominal wall. A Roux-en-Y jejunal limb is then created by mobilizing a suitable segment of intestine of approximately 60 cm in length. The anastomosis is then constructed with a standard end-to-side Roux-en-Y hepatico- or choledochojejunostomy, typically using a single layer of 4-0 or 5-0 absorbable sutures.
Roux-en-Y hepaticojejunostomy reconstruction of biliary tree. A. Repair of common hepatic duct stricture with transhepatic ring catheter exiting at the bifurcation. The stricture has been resected, and the distal biliary tree is oversewn. The hepaticojejunal anastomosis can then be performed over the ring catheter, or the ring catheter can be exchanged for a Silastic transhepatic stent. B. The Silastic transhepatic stent is shown exiting the biliary tree, with the Roux-en-Y jejunal limb prepared for the hepaticojejunostomy. C. Completed repair showing the Silastic biliary stent traversing the liver and the hepaticojejunostomy. The Roux-en-Y jejunal limb has been brought to the hepatic hilum in retrocolic position. (Reproduced, with permission, from Cameron JL, ed. Atlas of Surgery, Vol. I. Hamilton, Ontario, Canada: BC Decker; 1990:43, 53, 57.)
In the postoperative period, Silastic stents are left to external gravity drainage. A cholangiogram is then performed on postoperative day 4 or 5 (Fig. 50-17). If the biliary tree is adequately decompressed and no leakage is seen, the stents can be internalized and the perianastomotic drain is removed.
Postoperative cholangiography after hepaticojejunostomy via percutaneous Silastic biliary stents; the image shows no evidence of anastomotic leak.
The length of postoperative transanastomotic stenting is dependent on the individual patient, the clinical setting, and surgeon preference. Long-term stenting involves fluoroscopic exchange of stents at regular 2- to 3-month intervals. Timing of stent removal can be aided by biliary manometric flow studies that give objective data about the adequacy of the anastomosis, or by passing a clinical trial with the stent placed above the anastomosis.68
An alternative described approach of doing a hepaticojejunostomy involves an anterior longitudinal opening created in the bile duct and a long side-to-side anastomosis performed. Often, this is done to the extrahepatic portion of the left hepatic duct after it is lowered by dividing the hepatic plate (Hepp-Couinaud approach). This approach is particularly suitable for injuries at or just below the bifurcation. Right ducts do not lend themselves to this approach as well, because they have a short extrahepatic length. Sometimes the end of the right duct is used. However, dissection of the left duct provides a guide to the coronal plane in which the intrahepatic right hepatic ducts will be found and may further be exposed by removing liver tissue. During these procedures, exposure can be improved by dividing the bridge of tissue between segments 3 and 4 and opening the gallbladder fossa. Finally, if still more exposure is needed, resecting part of segments 4b and 5 will open the upper porta hepatis. The technique can avoid the need for postoperative stenting.
Nonoperative interventional radiology and endoscopic techniques have also been developed for the management of select patients with bile duct strictures and injuries. The most common nonoperative technique in these patients is interventional radiologic percutaneous stenting and balloon dilation, which may be possible in patients with intact biliary-enteric continuity. With the administration of conscious sedation, the proximal biliary tree is accessed so that the stricture can be traversed using a guidewire under fluoroscopic guidance (Fig. 50-18). Angioplasty-type balloon catheters are used to perform dilation of the stricture to a goal diameter based on the stricture location and the normal bile duct diameter. Following dilation, a transhepatic biliary stent is left in place across the stricture. The stent allows for future cholangiography, repeat dilation, and maintenance of the lumen while the bile duct heals. Complications of balloon dilation occur in up to 20% of patients and include cholangitis, hemobilia, and bile leaks. Percutaneous management will usually require numerous dilations.
Percutaneous balloon dilation of postoperative bile duct stricture using an angioplasty-type balloon catheter. Cholangiogram showing mid–bile duct stricture.
Results for the treatment of bile duct strictures using percutaneous balloon dilation are limited. In a retrospective comparison, percutaneous balloon dilation was compared to surgical repair in 43 patients with postoperative bile duct strictures treated between 1979 and 1987.69 Twenty-five patients underwent surgical repair with postoperative stenting and 20 patients had percutaneous balloon dilation with transhepatic stenting (mean: four dilations). Three patients underwent both surgical management and balloon dilation. Successful outcome was achieved in 89 and 52% of surgical and balloon dilation patients, respectively. These results would appear comparable to other series in the prelaparoscopic cholecystectomy era.70–73 However, the follow-up in most studies was less than 3 years, which is insufficient to make a definitive comment regarding long-term efficacy.
A series of 51 patients undergoing percutaneous balloon dilation therapy for bile duct strictures following laparoscopic cholecystectomy was reported by Misra and associates.74 At a median 76-month follow-up, overall success with balloon dilation, defined as stent-free without the need for further intervention, was 58%. With additional stenting and balloon dilation for two patients and surgical reconstruction for the remaining patients, all but one patient (98%) had a successful long-term outcome. These results suggest that in highly selected patients, percutaneous balloon dilation can provide long-term successful results.
Endoscopic balloon dilation has a more limited application, because it is technically possible only in patients with primary bile duct stricture repair or with choledochoduodenal anastomosis. ERC is performed, followed by endoscopic sphincterotomy. Sequential balloon dilation is performed after the stricture is traversed by a guide wire, often with one or more endoprostheses left in place after dilation. Complications associated with stent placement include cholangitis, pancreatitis, stent occlusion, migration, dislodgment, and ductal perforation, and have a reported incidence between 9 and 70%.75–77
Repeat cholangiography, often with repeat dilations, may be performed at regular intervals of every 3–6 months. While most endoscopists advocate regular follow-up and reevaluation of the stricture, the risks of stent occlusion and replacement need to be weighed against the risks and costs of the repeat procedures, and there is still some debate about timing of stent change to avoid occlusion. Bergman and associates demonstrated a 70% reobstruction rate with resultant jaundice or cholangitis when stents were not exchanged at 3-month intervals.77 In contrast, De Masi and colleagues describe and advocate leaving the stents in place until patients were symptomatic.78
In addition, the rate of stent occlusion appears to vary with the type of stent used. While metallic stents provide a longer period of patency than plastic stents for patients with malignant obstruction, the indications for their use in patients with benign strictures are limited. Metallic stents cannot be routinely exchanged or removed, and several studies have demonstrated high reocclusion rates at long-term follow-up.79–81 Newer covered metallic stents may provide a suitable alternative as they can be changed or removed after completion of treatment.82
While there have been no determinative studies for the length of time that stents should remain in place, most studies having excellent results have used larger-bore stents (≥10F) left in place for 2–6 months.76,77,83 Long-term studies reporting the endoscopic treatment of benign bile duct strictures are few. One of the few studies that directly compare endoscopic therapy to surgical reconstruction was done by Davids and colleagues from the Netherlands.75 In 66 patients, endoscopic therapy consisted of dilation and placement of an endoprosthesis, which was exchanged every 3 months. Surgical repair in 35 patients consisted of Roux-en-Y hepaticojejunostomy. Surgery was associated with excellent or good results in 29 patients (82%), with 6 patients (17%) developing a recurrent stricture at mean 40 months from initial surgery. In contrast, endoscopic stenting resulted in 81% with excellent or good outcome and 18% developing restricture at a mean of 3 months after stent removal. Recurrent strictures after stent removal in several other series have been reported to occur at a rate varying from 0 to 20% at median follow-up of 29–108 months.76,77,83
Biliary injury and stricture repair are associated with significant morbidity and mortality. With improved medical technology and experience, the incidence of operative mortality has decreased markedly. A recent series of 200 consecutive patients repaired at The Johns Hopkins Hospital reported a perioperative mortality of only 1.7%.62 Advanced age, comorbid disease, and a history of major biliary tract infection are factors associated with operative mortality. Underlying liver disease is the most important correlated factor for operative mortality and morbidity, with advanced biliary cirrhosis and portal hypertension having mortality rates approaching 30%. Fortunately, in the modern era such advanced disease is uncommon.
A recent analysis of Medicare claims, patients' examined mortality associated with major bile duct injuries over an 8-year period in 791 elderly patients demonstrated a perioperative mortality of 2.7% associated with repair.84 In addition, the study demonstrated an adjusted hazard ratio for death during the follow-up period was significantly higher for patients with a bile duct injury than in patients without a bile duct injury. The hazard increased with advancing age and comorbidities and decreased with experience of the repairing surgeon. The adjusted hazard for death during follow-up was 11% greater if the repairing surgeon was the same as the injuring surgeon. This study gives supportive evidence for improved survival in patients with major bile duct injuries treated by experienced hepatobiliary surgeons at tertiary referral centers.
In most series, postoperative morbidity rates are in the range of 20–40%. Morbidity nonspecific to biliary surgery includes hemorrhage, infection, and risks associated with general anesthesia. Complications specific to biliary repairs include anastomotic leak, cholangitis, and hepatic insufficiency associated with preexisting liver disease. Anastomotic leaks can typically be managed via nonoperative means, especially when transanastomotic stenting has been utilized. Percutaneous transhepatic stenting may also have specific morbidity, including bile leaks from hepatotomy sites, hemobilia, and cholangitis from stent occlusion.
The series reporting the outcomes in 200 patients undergoing surgical reconstruction demonstrated a 43% overall postoperative complication rate.62 The most common complications were wound infection (8%), cholangitis (6%), minor stent-related complications (6%), and intra-abdominal abscess/biloma (3%). Postoperative cholangiography revealed an anastomotic leak in 4.6% of patients and extravasation at the liver dome–stent exit site in 10.3%. These complications were all managed conservatively with either new biliary stent placement or biliary stent exchanges required in 2.3%. Postoperative percutaneous abscess/biloma drainage was required in nine patients (5.1%). No patients required reoperation in the postoperative period. Despite the relatively high morbidity rate, median length of stay was similar to that in other reports (8 ± 4.6 days).
There are mixed results regarding perioperative complications when vascular injury has occurred in association with a bile duct injury.65,85,86 A report from Schmidt and associates87 reported that a repair in the presence of uncontrolled infection, a concurrent hepatic artery injury, and injury level (at or above the bifurcation) were independent predictors of the development of major biliary complications.
The ultimate goal of the repair of a bile duct stricture is a successful repair with no further symptoms, including jaundice, cholangitis, and preserved liver function. Excellent long-term results following operative repair of postoperative bile duct injuries after open cholecystectomy have been reported with approximately 80–90% having a successful outcome (Table 50-4).69,75,88–94 Early reports and observations from the laparoscopic era were less favorable than those previously reported with open cholecystectomy repairs. Stewart and Way66 reviewed 85 patients who had undergone 112 biliary repairs and defined four factors that influenced success or failure of operative repairs after laparoscopic cholecystectomy bile duct injury: (1) performance of preoperative cholangiography, (2) choice of surgical repair, (3) details of surgical repair, and (4) experience of the repairing surgeon. Procedures without preoperative cholangiography were unsuccessful 96% of the time, and those with incomplete cholangiography data had a success rate of only 31%. With complete cholangiography data, the success rate was 84%. All patients with complete transection of the bile duct who underwent primary end-to-end repair over a T-tube had a failed result. In contrast, 63% of Roux-en-Y hepaticojejunostomy repairs were successful. Initial repair by the original laparoscopic surgeon was successful in only 17% of cases. Repeat attempts at repair by the same surgeon were never successful. Finally, those patients whose first repair was by a tertiary care biliary surgeon achieved a 94% success rate.
Table 50-4: Results of Surgical Repair of Postoperative Bile Duct Strictures ||Download (.pdf)
Table 50-4: Results of Surgical Repair of Postoperative Bile Duct Strictures
|Reference||Year||No. of Patients||Success Rate (%)||Follow-Up (mo)|
|Walsh et al95||2007||144||89||67|
|Lillemoe et al94||2000||156||91||58|
|Tocchi et al88||1996||84||83||108|
|McDonald et al89||1995||72||87||<60|
|Chapman et al90||1995||104||76||86|
|Davids et al75||1993||35||83||50|
|Pitt et al69||1989||25||88||57|
|Innes et al91||1988||22||95||72|
|Genest et al92||1986||105||82||60|
|Pellegrini et al93||1984||60||78||102|
A series providing long-term results after repair of bile duct injuries and strictures in the 1990s was reported by Lillemoe and associates.94 A total of 156 consecutive patients underwent surgical reconstruction with a mean follow-up period of 57.5 months (range 11–119 months; median 54.7 months). The original operation consisted of laparoscopic cholecystectomy in 118 patients (76%), open cholecystectomy in 27 patients (17%), open cholecystectomy with bile duct exploration in 4 patients (3%), or other abdominal surgery or trauma in 7 patients (4%). Sixty patients (41%) had a previous attempt at repair prior to referral with eight patients (5.5%) having more than one attempt at repair prior to referral. Of the 156 operatively repaired patients, 142 patients had completed treatment at the time of final evaluation with an overall success rate of 91%. Even though they were more likely to have had repair prior to referral and higher and more complex injuries, patients with repair of a stricture or injury associated with laparoscopic cholecystectomy had a better success rate than repair after other operations (94 vs 80%; p < .05). There were 13 failures following surgical reconstruction. Ten had successful results following either surgical revision (one patient) or percutaneous balloon dilation (nine patients), resulting in an overall success rate of 98% including secondary intervention. Only three patients continued to require long-term biliary stents to prevent biliary obstruction symptoms and/or cholangitis. Comparable results have been reported from other high-volume hepatobiliary centers with similar volume of patients in the series.96–99 Outcomes after surgical repair for laparoscopic cholecystectomy injury from other series are outlined in Table 50-5.94,97–100
Table 50-5: Surgical Repair of Laparoscopic Cholecystectomy Bile Duct Injuries ||Download (.pdf)
Table 50-5: Surgical Repair of Laparoscopic Cholecystectomy Bile Duct Injuries
|Reference||No. of Patients||Bismuth Level 3–5 (%)||Success Rate (%)|
|Lillemoe et al94||118||63||94|
|Walsh et al97||34||80||91|
|Bauer et al98||32||24||83|
|Mirza et al99||52||53||92|
|Nealon and Urrutia100||23||26||100|
Effect of Surgical Repair in Quality of Life.
Despite the overall high level of success in the surgical management of laparoscopic cholecystectomy bile duct injuries, there is an impression that patients may have an impaired quality of life even after a successful repair of their bile duct injury. Quality of life after laparoscopic cholecystectomy bile duct injury has been addressed in several recent reports, with differing results.101–104 Two studies using the Short Form 36 (SF-36) Health Survey quality-of-life instrument in patients with laparoscopic cholecystectomy injury found both the physical and mental quality-of-life aspects to be reduced compared to controls at approximately 5-year follow-up.101,104 A study with SF-36 found that patients with laparoscopic bile duct injury and subsequent biliary reconstruction had quality of life similar to matched controls and national norms in all eight quality-of-life areas.103 Melton and associates102 assessed quality of life in 54 patients having undergone successful surgical repair of laparoscopic cholecystectomy bile duct injuries and compared these results to quality-of-life measures in patients after uncomplicated laparoscopic cholecystectomy and in healthy controls using a standard quality-of-life instrument, which was used to assess the physical, psychological, and social domains of health-related quality of life. Patients after surgical repair had overall quality-of-life scores comparable to those of controls. Only in the psychological dimension were patients post–bile duct injury repair found to have significantly worse scores compared to controls. Patients who reported pursuing a lawsuit following their injury (31%) had significantly worse quality-of-life scores in all domains when compared to those who did not entertain legal action (p < .01).