With advanced endoscopic and laparoscopic techniques readily accessible to the treating surgeon, determining the wisest path to the successful treatment of choledocholithiasis and cholangitis has become more challenging. Nevertheless, a large number of options allow one to tailor specific therapy to each individual clinical situation in order to achieve the highest probability of safety and success. In this chapter, we offer the reader a review of the methods available for the diagnosis and treatment of common bile duct (CBD) stones and cholangitis so that treatment plans can be developed that are patient-specific and have the highest chance of success.

Classification and Epidemiology

A common entity in Western societies, gallstones are found in approximately 15% of Americans and result in 700,000 cholecystectomies a year. The annual cost of medical care for gallstones is almost $6.5 billion compared with colorectal cancer ($9.5 billion), viral hepatitis ($3.4 billion), and gastroesophageal reflux disease (GERD) ($12.6 billion).^1,2 CBD (downstream of the confluence of the hepatic ducts) stones have been noted in 10% to 15% of patients with cholelithiasis, and this incidence increases with age to over 80% in those who are over 90 years old.³ Choledocholithiasis in Western countries usually results from stones originating in the gallbladder and migrating through the cystic duct. These secondary bile duct stones are cholesterol stones in 75% and black pigment stones in 25% of patients. Cholesterol stones contain more than 70% cholesterol by weight, and variable amounts of bile salt and calcium. Over 90% of all cholesterol stones are radiolucent. Cholesterol stones are formed in the presence of cholesterol saturation, biliary stasis, and nucleating factors. Behavioral factors associated with cholesterol gallstones include nutrition, obesity, weight loss, and physical activity. Biologic factors linked to gallstones include increasing age, female sex and parity, serum lipid levels, and the Native American, Chilean, and Hispanic races.¹ The formation of black pigment stones is associated with hemolytic disorders, cirrhosis, ileal resection, prolonged fasting, and total parenteral nutrition.³ These conditions lead to supersaturation of unconjugated bilirubin, which results in precipitation of bilirubinate with calcium and other anions in bile. The precipitated salt then becomes a nidus for black stone formation.

Primary bile duct stones, on the other hand, form within the bile ducts and usually are of the brown pigment variety. These tend to be less than 20% cholesterol and higher in bilirubin content as compared with secondary stones. Unlike secondary stones, primary stones are associated with biliary stasis and bacteria.⁴ In fact, in the pathogenesis of brown pigment stones, bacterial enzymes unconjugate bilirubin glucuronide to form free bilirubin, which then precipitates with calcium to become the nidus for stone formation.⁵ Moreover, bacteria have been found in brown pigment stones by electron microscopy but not in black pigment stones.

Primary bile duct stones are more common in Asian populations, and these often are associated with primary intrahepatic stones in this population.¹ These intrahepatic stones usually are calcium bilirubinate and mixed stones and contain more cholesterol and less bilirubin than the extrahepatic bile duct ­pigmented stones. The pathogenesis of these intrahepatic stones appears to involve bile infection, biliary stasis, low-­protein, low-fat diets and malnutrition, and parasitic infections. However, the role of Ascaris lumbricoides and Clonorchis sinensis in the formation of intrahepatic stones is controversial. While these parasites are found in many geographic areas, ­primary intrahepatic stones are found mainly in Southeast Asia. Therefore, in addition to parasitic infections, other factors must play a role in the formation of these stones.¹

Clinical Presentation and Natural History

Asymptomatic bile duct stones may be found incidentally during evaluation of patients with suspected gallstones. In fact, 5% of common duct stones found during surgery may be unsuspected by preoperative findings and discovered only during intraoperative evaluation of the biliary tree. In one autopsy study of 615 patients over age 60, 1% were found to have bile duct stones.³ Patients with choledocholithiasis may present with biliary colic, bile duct obstruction, bilirubinuria (or tea-colored urine), pruritus, acholic stools, and jaundice. However, the biliary obstruction usually is incomplete. There may be nausea and vomiting with intermittent or constant epigastric or right upper quadrant pain.⁶ The clinical course may be complicated by acute gallstone pancreatitis, cholangitis, or rarely, hepatic abscess. Infected patients may present with back pain, fever, hypotension, and mental status changes suggestive of cholangitis and ascending cholangitis. An asymptomatic state is also recognized.

CBD stones are covered by a bacterial biofilm of adherent quiescent bacteria residing in a hermetic environment. When stones cause obstruction of the ducts, cytokines released by epithelial cells activate these bacteria to the planktonic and virulent forms.¹ Therefore, bile duct obstruction secondary to stones often is accompanied by bacterial sepsis resulting from activation of the bacterial biofilm on these stones. Sepsis is much less likely to occur in the context of malignant obstruction without choledocholithiasis.

Although a majority of stones will pass spontaneously into the duodenum within hours, prolonged biliary obstruction can lead to biliary cirrhosis and portal hypertension. The average time for choledocholithiasis to lead to biliary cirrhosis is about 5 years, depending on the extent of obstruction.¹ Even with cirrhosis, however, the obstruction should be relieved because some reversal of portal hypertension and secondary biliary cirrhosis may be possible.

Physical examination of patients with choledocholithiasis may be normal or may reveal jaundice, scleral icterus, and abdominal tenderness over the right upper quadrant without peritoneal signs. Early in the course, physical examination may not be very different from that of patients with cholecystitis. Severe tenderness may point to acute gallstone pancreatitis, whereas fever, hypotension, and confusion may suggest cholangitis.⁷

Blood tests may reveal elevation of serum alkaline phosphatase (ALP), gamma-glutamyl transpeptidase (GGT), and bilirubin. Mild elevations of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) can be seen, whereas these are particularly abnormal in the situation of cholangitis. Although bilirubin and aminotransferase levels are high in 70% to 90% of patients at the onset of symptoms, almost all patients have elevation of ALP and GGT.⁷ Elevated amylase and lipase may suggest pancreatitis. Leukocytosis may be seen with cholangitis, pancreatitis, or associated acute cholecystitis. It is worth noting that laboratory evaluation of patients with bile duct stones can be normal repeatedly, and this should not dissuade further evaluation of patients suspected to harbor duct stones.⁸

Evaluation and Management

The evaluation and treatment of choledocholithiasis are best discussed by considering the three clinical circumstances in which patients who may have bile duct stones are seen: prior to cholecystectomy, during cholecystectomy, or some time after cholecystectomy.

PREOPERATIVE

The diagnosis of choledocholithiasis cannot be made on the basis of history, physical examination, and laboratory investigations alone. Moreover, the distinction between the symptoms of bile duct stones and gallbladder stones is difficult. Increasing age, history of fever, cholangitis, and pancreatitis are risk factors for bile duct stones, whereas elevations of serum bilirubin, AST, or ALP are independent positive predictors.^1,9

Biochemical tests can be used as an initial screen to identify patients with high probability of CBD stones. ALP has the highest sensitivity (79.5%) for CBD stones. Total bilirubin (TB) has the highest specificity (87.5%) and accuracy (84.1%) for CBD stones. A normal gamma-glutamyl transferase is excellent for exclusion of CBD stones (odds ratio of 3.2; negative predictive value [NPV] of 97.9%). For a patient with normal GGT, the likelihood of CBD stone is only 2.1%. An elevated GGT, ALP, TB, ALT, and AST has a sensitivity of 87.5%, compare to sensitivity of 96% for endoscopic retrograde cholangiopancreatography (ERCP).¹⁰

Transcutaneous ultrasound has been the traditional method of evaluating patients with biliary disease. It is highly accurate in identifying acute calculous cholecystitis and the presence of gallstones greater than 2 mm. Sensitivities and specificities of 48% to 100% and 64% to 100%, respectively, have been reported.¹¹ However, the ability of transcutaneous ultrasound to establish the diagnosis of choledocholithiasis is only about 50%, varying from 30% to 90%.^7,12 The role of ultrasound as a screening test for bile duct stones was evaluated prospectively by Gross and colleagues.¹³ Patients who were about to undergo ERCP were examined by right upper quadrant sonography to assess the size of the intra and extrahepatic ducts and for the presence or absence of bile duct stones. The findings were compared with ERCP, percutaneous transhepatic cholangiography, or surgical follow-up. Ultrasound was not found to be accurate in the diagnosis (sensitivity of 25%) or the exclusion (73% NPV) of choledocholithiasis.

Costi and colleagues studied the usefulness of the number and size of gallbladder stones for predicting asymptomatic choledocholithiasis.¹⁴ Ultrasound data of 300 consecutive patients undergoing laparoscopic cholecystectomy were analyzed. Patients were divided into two groups: those with multiple small (<5 mm) gallbladder stones or variable (≤5 mm and >5 mm) stones and those with large (>5 mm) stones only. The classification of stone size was confirmed by surgery in 95% of patients. Moreover, the presence of multiple small and variable gallbladder stones represented a risk factor for synchronous asymptomatic bile duct stones (9.5%) as compared with large stones only (2.5%). In another study, ultrasonography was found to have a positive predictive value (PPV) of 69% and an NPV of 78% for choledocholithiasis in patients suspected to have bile duct stones.¹⁵ This compared with serum transaminase tests having predictive values of 68% and 93%, respectively. In comparison to elevated serum transaminases and/or increased amylase levels, ultrasonographic evidence of CBD dilatation (>7 mm) has been described to be the best predictor of choledocholithiasis.¹⁶ Nonetheless, it is worth noting that almost half the patients with CBD stones do not have dilated ducts by ultrasonography, hence a negative study has limited value.¹⁷

In order to predict the presence of bile duct stones more accurately, the combination of clinical, laboratory, and ultrasound risk factors has been used by several investigators.^1,18,19 For patients over 55 years old with a bilirubin greater than 30 μmol/L (1.75 mg/dL) and a CBD more than 6 mm on ultrasound, the probability of a CBD stone is 72%.²⁰ By multivariate logistic regression analysis, the combination of dilated CBD with evidence of stones by ultrasonography, clinical evidence of cholangitis, elevated aspartate transaminase and bilirubin, the likelihood of having stones in the bile duct was 99%.¹⁹ In the absence of all four of these findings, the probability of synchronous choledocholithiasis in patients with cholelithiasis was only 7%.¹⁹ Unfortunately, many patients present with only some of these findings, and the prediction of bile duct stones based on these criteria becomes difficult. Moreover, ultrasound sensitivity is in part operator-dependent and altered by bowel gas, making the findings inconsistent.²¹

In 1968, ERCP was introduced as a diagnostic tool to aid in the management of biliary and pancreatic diseases.²² Five years later, with the development of endoscopic sphincterotomy, ERCP was transformed into a therapeutic modality.²³ In 2009, 228,000 biliary endoscopies were performed in the United States, totaling $900 million in healthcare costs.²⁴ Short of intraoperative examination, ERCP has long been considered the standard reference for the diagnosis of CBD stones.²¹ The specificity and sensitivity of ERCP were reported in 1982 by Frey and colleagues.²⁵ ERCP was compared with findings on common duct exploration or cystic duct cholangiography in 72 patients and was found to have a sensitivity of 90%, specificity of 98%, and a 96% accuracy. Interestingly, the interval between performance of the procedure and operation was particularly important in patients with multiple small stones. Since small stones pass more readily from the gallbladder to the common duct and from the common duct to the duodenum, the longer the interval between ERCP and operation, the greater was the chance of discordant findings.

ERCP has the advantage of being both diagnostic and therapeutic for CBD stones (Figs. 63-1 and 63-2). That is, after stones in the bile duct are identified, endoscopic sphincterotomy and stone extraction can be performed at the same setting.²¹ ERCP stone extraction is successful 80% to 90% of the time using the techniques of sphincterotomy and balloon catheter or Dormia basket stone retrieval.^23,26 The addition of mechanical, electrohydraulic, laser, or extracorporeal shockwave lithotripsy for large stones increases the success rate to over 95%.

Sphincterotomy entails division of the papilla and sphincter muscles to widen the distal end of the CBD using a sphincterotome, a device consisting of a Teflon catheter with exposed cautery wire at the tip. The length of the intraduodenal part of the CBD limits the extent of the cut. Balloon sphincteroplasty is a sphincter-preserving alternative to sphincterotomy that uses a high-pressure hydrostatic balloon of either 6 or 8 mm diameter to dilate the papilla. One drawback of sphincteroplasty is the limited size of the papillary opening created as compared with sphincterotomy. Failure rates of 22% for stone extraction with balloon dilatation and the need for mechanical lithotripsy in 31% have been reported.²⁶ ­Furthermore, sphincteroplasty has been associated with a pancreatitis rate of 19 times greater than the rate associated with sphincterotomy.²⁷ A study evaluating the use of sphincteroplasty, on the other hand, found that severe pancreatitis only occurred in 1 patient out of 63, whereas the successful stone extraction rate was 84%.²⁸

Once the sphincter has been divided, most stones can be removed using a Dormia basket or a balloon catheter. The Dormia basket has better traction than the balloon and consequently is recommended for larger stones (>1 cm). The balloon catheter occludes the bile duct lumen after inflation and therefore is useful for removal of small stones and gravel. The catheter also can be inserted over a guidewire, ­making it useful for intrahepatic duct stones. Three situations that may lead to a difficult extraction are stone size greater than 1.5 cm, stone location proximal to a stricture, and multiple stones that are impacted. Alternative approaches to these situations include mechanical lithotripsy, ­electrohydraulic or laser lithotripsy, and extracorporeal shock wave lithotripsy. Mono-octanoin and methyl tertiary butyl ether (MTBE) have been used in the past to dissolve bile duct stones through nasobiliary drainage catheters or T-tubes. The practice largely has been abandoned because of high complication rates, poor results, and the technical difficulty of performing the dissolution.²⁶

Mechanical lithotripsy is the most commonly used and simplest means of fragmenting large bile duct stones or when a significant discrepancy between the stone size and the diameter of the exit passage exists.²⁹ A large, strong basket is used to trap the stone. The stone then is crushed against a metal sheath by applying tension to the wires by the use of a crank handle. Reimann and colleagues first described the technique in 1982, and since then, many variations in design have become available.^30,31 When stones are extremely large, repeat application of the technique may be needed to further break the stone fragments and thus allow removal. Success rates between 80% and 90% have been reported for clearing the bile duct using the procedure.^32-34 One retrospective study of 162 patients undergoing mechanical lithotripsy found that the probability of bile duct clearance was over 90% for stones less than 1 cm diameter versus 68% for stones greater than 2.8 cm diameter.³⁵ Garg and colleagues presented data on 87 patients with stones greater than 1.5 cm that required mechanical lithotripsy.³⁶ They analyzed ­various predictive factors, including size and number of stones, stone impaction, serum bilirubin, presence of cholangitis, and bile duct diameter, in relation to the success or failure of lithotripsy. Impaction of the stones in the bile duct was found to be the only significant factor that predicted failure of mechanical lithotripsy and subsequent bile duct clearance. The composition of the stone also has been found to affect the success of stone removal. Soft stones, such as those found in Oriental cholangitis, are large but amenable to crushing, sometimes even with the Dormia basket.²⁹ However, calcified stones are hard and resist mechanical crushing.

Large-balloon dilatation of the distal bile duct has been reported as a means of removing difficult bile duct stones after standard extraction has been unsuccessful.³⁷ In a retrospective analysis, 58 patients who failed standard sphincterotomy and standard basket or balloon extraction underwent dilation with a 10- to 20-mm-diameter balloon (esophageal type) followed by standard basket or balloon extraction. The patients were divided into two groups: 18 patients with a tapered distal bile duct (group 1) and 40 patients with square, barrel-shaped, and/or large (>15 mm) stones (group 2). Stone clearance was successful in 89% of group 1 patients and 95% of group 2 patients. In the two patients in each group in whom extraction was not possible after dilatation alone, mechanical lithotripsy allowed for stone removal. Complication rates were high: 33% for group 1 and 7.5% for group 2, including mild pancreatitis (two patients), mild cholangitis (two patients), and bleeding (five patients). No bleeding required surgery. Large-balloon dilatation offers an alternative in managing difficult bile duct stones, and further studies are needed to establish its role as compared with other lithotripsy options.

The management of complicated situations of choledocholithiasis may require several procedures or several sessions of the same procedure for successful clearance of the CBD. In such situations, partial stone impaction may lead to biliary stasis and cholangitis. Along with the administration of broad-spectrum antibiotics to cover gram-negative and gram-positive bacteria, it is important to decompress the biliary tree with either a nasobiliary ­catheter or a biliary stent as a temporizing measure pending more definitive treatment.^26,29 By doing this, serum bilirubin ­levels are allowed to decrease, and the rate of post-procedure cholangitis becomes similar to that after stone clearance. ­Interestingly, up to 30% of patients in whom a stent has been left in place for large stones have spontaneous disappearance of the stones, as noted on subsequent ERCP.²⁹ This may be secondary to the frictional movement of stone against the stent or as a result of improved bile flow with dissolution effects. Furthermore, by adding oral ursodeoxycholic acid to stent placement, 9 of 10 patients have been reported to become stone-free by this combination as compared with 0 of 40 with stent placement only.³⁸ Long-term stent placement is an unconventional management option for patients with large, inextricable stones who are at high risk for surgical intervention, and should be used with caution. In a long-term follow-up study of 58 elderly patients, 40% of patients treated with permanent stents for endoscopically irretrievable stones developed 34 complications in 23 patients, with cholangitis being the most frequent.³⁹ At median follow-up of 36 months, 44 patients had died, 9 as a result of biliary-related causes. Hui and colleagues prospectively evaluated 36 high-risk patients with difficult CBD stones.⁴⁰ Of these, 19 underwent stent placement, and 17 underwent complete stone clearance with electrohydraulic lithotripsy. The actuarial incidence of recurrent acute cholangitis was 8% in the lithotripsy group versus 63% in the stent group. The actuarial mortality also was higher in the stent group compared with the lithotripsy group, 74% and 41%, respectively.

Although ERCP has developed over the years as a relatively safe endoscopic diagnostic and therapeutic tool, there are well-defined, potentially severe, and life-threatening complications associated with it. The reported rates of complications vary widely in different studies, and this may be related in part to study design, with retrospective studies being prone to under-reporting. Furthermore, the complication rates may diverge depending on the patient mix in the study and may be influenced in part by the definitions used for these complications.²²

The mortality rate after diagnostic ERCP is about 0.2%, and this rate is more than doubled by therapeutic interventions, to 0.5%.^21,22 Remember, these are essentially the same rates as for laparoscopic cholecystectomy. Cardiopulmonary complications are the leading cause of death and include cardiac arrhythmia, hypoventilation, and aspiration. These may be the result of premorbid conditions or related to medications used during sedation and analgesia. Other ­significant complications include perforations (0.3%-0.6%), ­bleeding related primarily to sphincterotomy (0.8%-2%), cholecystitis (0.2%-0.5%), and cholangitis (1%). In a recent ­meta-­analysis, prophylactic²¹ antibiotics were not found to be beneficial in reducing infectious complications of ERCP. Moreover, another study failed to show a decrease in the rate of cholangitis in patients with distal bile duct stones or biliary strictures receiving antibiotic prophylaxis.²²

Pancreatitis is the most common complication seen after ERCP. The consensus definition for ERCP-induced pancreatitis is new or worsened abdominal pain, serum amylase that is greater than three times the upper limits of normal at 24 hours post procedure, and a requirement of at least 2 days of hospitalization. Although the transient elevation of serum pancreatic enzyme levels is frequent, based on the consensus definition of ERCP pancreatitis, the expected rate of this complication is between 1% and 7%. Risk factors associated with ERCP-induced pancreatitis include a prior history of ERCP pancreatitis, nondilated biliary ducts, normal bilirubin, young age, female gender, and suspected sphincter of Oddi dysfunction. In fact, the risk of pancreatitis in women with normal bilirubin and suspected sphincter of Oddi dysfunction is 18%, compared with 1.1% for the low-risk patient.^22,41 Moreover, one of five episodes of pancreatitis in this setting will be severe, requiring more than a 10-day hospital stay and/or resulting in necrosis, pseudocyst, abscess formation needing surgery or percutaneous drainage, or death. Since the highest rate of complications appears to exist in the group of patients that is least likely to benefit from ERCP, the most effective method of reducing post-ERCP pancreatitis would appear to be to avoidance of unnecessary ERCP.

Pharmacologic methods of pancreatitis prophylaxis have been attempted to reduce this complication after ERCP.²² Meta-analyses have suggested that somatostatin and gabexate are useful in reducing pancreatitis rates, but multicenter randomized, controlled trials have failed to confirm this. The use of nonionic contrast agents has not reduced the rate of pancreatitis. Glyceryl trinitrate (GTN) administered by both sublingual and transdermal routes has been shown to decrease post-ERCP pancreatitis in two placebo-controlled trials, supposedly by decreasing sphincter of Oddi pressure. Its hypotensive effects limit its use. The placement of pancreatic stents has been found to reduce the incidence of postbiliary sphincterotomy pancreatitis in patients suspected of sphincter of Oddi dysfunction.

Based on clinical, laboratory, and ultrasound criteria for CBD stones, up to 70% of patients may be found not to have duct stones at the time of preoperative ERCP.^19,42,43 Given this, a large number of patients may be subjected to an unnecessary ERCP and suffer its risks and costs. Several methods have become available to diagnose the presence of bile duct stones accurately prior to having patients undergo ERCP or operative interventions. The most important of these are magnetic resonance cholangiopancreatography (MRCP), endoscopic ultrasound (EUS), and computed tomography (CT).

Sensitivities of conventional CT for choledocholithiasis in the setting of suspected bile duct stones is 76% to 90%, whereas unenhanced helical CT has been shown to have a sensitivity of 88%, a specificity of 97%, and an accuracy of 94%.²¹ When compared with ERCP as the reference standard, CT without biliary contrast material showed poor concordance with ERCP (sensitivity 65% and specificity 84%) but compared better when oral biliary contrast material was given (sensitivity and specificity greater than 90%).⁴⁴ CT with intravenous (IV) biliary contrast material in other studies has been found to have a sensitivity of 71% to 85% and a specificity of 88% to 95%.⁴⁴ Patel and colleagues reported a comparison between noncontrast-enhanced helical CT and the reference standard of EUS and found that CT had both a sensitivity and a specificity of 83% for the detection of CBD dilatation in the setting of choledocholithiasis. However, when CT was evaluated for identifying duct stones, it had a sensitivity of only 22% and a specificity of 83%.

Since its introduction over a decade ago, the use of MRCP for the diagnosis of CBD stones has increased eightfold, while the use of ERCP as a diagnostic tool has decreased substantially.²⁴ With sensitivities and specificities that approach those of ERCP, MRCP has emerged as a diagnostic alternative to ERCP for the detection and exclusion of choledocholithiasis.²¹ Additionally, economical and clinical cost analysis showed diagnostic MRCP is favorable over diagnostic ERCP in selected patients.⁴⁵ Performed with T₂-weighted sequences, the biliary tract is seen as a bright structure with high-signal intensity without the use of contrast material, instrumentation, or ionizing radiation. Common duct stones are seen as low-signal-intensity filling defects surrounded by high-­intensity bile. Improvements over the past decade have resulted in the ability to image the entire biliary tract in a single breath-hold of 20 seconds with a resolution that allows visualization of fourth-order intrahepatic bile ducts and small stones. Stones as small as 2 mm can be detected even in the absence of biliary dilatation.²¹ In one study of 97 patients, sensitivity of MRCP was 100% for stone diameters of 11 to 27 mm, 89% for stone diameters of 6 to 10 mm, and 71% for stone diameters of 3 to 5 mm.⁴¹ In this study, MRCP had a 91% sensitivity compared with 100% for ERCP, whereas both tests had a specificity of 100%. Studies with state-of-the-art techniques have found sensitivities of 90% to 100% with specificities of 92% to 100%.²¹ In a prospective analysis by Ke and colleagues, 267 patients believed to have CBD stones were evaluated by MRCP and ERCP.⁴⁶ MRCP was found to have a sensitivity of 100%, a specificity of 96%, and an NPV of 100%. Kejriwal and colleagues retrospectively examined patients with cholelithiasis who underwent MRCP for suspected choledocholithiasis.⁴⁷ Patients were considered not to have clinically relevant common duct stones if they had a negative MRCP and did not present for readmission for choledocholithiasis after treatment of their cholelithiasis. MRCP was negative for bile duct stones in 74% of patients (60 of 81) and missed clinically relevant stones in two patients, resulting in a PPV of 95% and an NPV of 97%. With its ability to exclude bile duct stones, MRCP may allow the avoidance of unnecessary diagnostic ERCP. Demertines and colleagues found that even in patients with high and moderate risk of CBD stones based on laboratory findings, the performance of MRCP could have resulted in the avoidance of ERCP in 52% and 80% of patients, respectively.⁴⁸

One of the limitations of MRCP is that its resolution remains less than that of ERCP, and therefore it cannot detect small stones and crystals consistently. Claustrophobia also may influence the use of MRCP, and patients may need sedation or even general anesthesia for its performance. Patient obesity may diminish the quality of images, whereas morbid obesity, pacemakers, and aneurysm clips preclude entry into the scanner.²¹ Conversely, ERCP may be limited by an inability to access and cannulate the papilla and opacify the ductal system. Failed ERCP rates vary greatly among endoscopists, from 5% to 20%.²¹ Moreover, alterations in the gastrointestinal tract anatomy, such as a Billroth II gastrojejunostomy, may preclude access to the ampulla. In summary, MRCP offers a method of evaluating the biliary system for bile duct stones with sensitivities and specificities that approach those of ERCP in a manner that is noninvasive and avoids the risks and limitations of ERCP. Patients with a positive MRCP then may be considered for more invasive therapeutic procedures.

Another sensitive method of evaluating the biliary system for CBD stones is EUS. EUS has been shown to have a diagnostic accuracy of 95% for bile duct stones.⁴⁹ With the high ultrasound frequencies used (7.5 and 12 MHz), EUS has a resolution of less than 1 mm, making it the best imaging technique available for the extrahepatic biliary tract. Several studies have found EUS to be similar to ERCP in sensitivity and specificity for the evaluation of choledocholithiasis, with some showing ERCP to be better and others showing EUS to be better.⁴⁴ Compared with ERCP, EUS is semi-invasive, with almost no procedure-related complications and a negligible failure rate. In fact, several series comprising over 1000 patients have reported no complications.⁴⁹ In a prospective study by Buscarini and colleagues, 485 patients suspected to have choledocholithiasis based on clinical, laboratory, and ultrasound or CT findings underwent EUS.⁴⁹ Positive EUS findings were confirmed by surgery or ERCP with sphincterotomy; negative findings were confirmed by clinical ­follow-up of at least 6 months. EUS findings were verified in 463 patients as follows: 237 true positive, 216 true negative, 2 false positive, and 4 false negative, and in 4 patients EUS was incomplete (sensitivity 98%, specificity 99%, PPV 99%, NPV 98%, accuracy 97%). No complications were noted in the study. EUS offers higher resolution than MRCP and therefore is better able to detect small stones. It is able to identify bile duct stones as well as microlithiasis and is able to detect pathology that is not seen by ERCP. EUS prior to performing invasive diagnostic or therapeutic techniques would lower the rate of procedure-related complications in patients suspected of having bile duct stones. Cost analysis of EUS followed by ERCP versus ERCP alone is also in favor of EUS as a pre-therapeutic procedure.⁴⁹

In patients for whom ERCP is not available, not possible secondary to anatomic considerations, or not successful, an alternative method of cholangiography and nonsurgical therapy is percutaneous transhepatic cholangiography (PTC) followed by transhepatic methods of stone removal. A needle is introduced into the intrahepatic bile ducts through the skin, and a cholangiogram is performed, followed by wire insertion and then a catheter over the wire for external biliary drainage and access to the biliary system. The method was introduced in Denmark in the 1970s and has been refined over the years with the addition of several therapeutic options.⁵⁰ This technique is particularly useful for evaluating intrahepatic stones or other proximal bile duct disease. After diagnosis of bile duct stones, several therapeutic options are available through the percutaneous route. In 1981, the removal of an 8-mm CBD stone by percutaneous transhepatic technique was reported by Fernstrom and colleagues.⁵¹ In 1990, Stokes and colleagues, from Boston, reported a series of 53 patients in whom surgery was contraindicated and ERCP unsuccessful.⁵² By inserting a modified Dormia basket via a percutaneous transhepatic route, stones were advanced whole or after fragmentation into the duodenum. Mono-octanoin or MTBE were used in 30 patients to reduce stone size or remove debris. Morbidity and mortality were 12% and 4%, with a success rate of 93%. Transhepatic cholangioscopy and lithotripsy can be performed after PTC and dilatation of the intrahepatic channel with success rates of 90% to 100% and 5% to 8% complications.⁵³ In a series of 12 patients with bile duct stones, PTC in combination with laser or electrohydraulic lithotripsy to deliver stone fragments into the duodenum was found to be successful in all the patients.⁵⁴ In another series of 13 patients, laser lithotripsy was used with percutaneous cholangioscopy performed either transhepatically (12 patients) or through the T-tube track.⁵⁵ Stone fragmentation was successful in 92%, and stone clearance was possible in all patients. However, 11 patients required the addition of sphincterotomy (either by ERCP or by antegrade method with fluoroscopic monitoring) or stent insertion. Bleeding in two patients accounted for a 15% severe complication rate. Percutaneous transhepatic papillary balloon dilatation was reported by a Japanese group for the management of choledocholithiasis.⁵⁶ In the five patients in whom the method was used, bile duct stones were able to be pushed into the duodenum in all, with no complications or deaths. Ponchon and colleagues reported percutaneous choledochoscopy for stone extraction in 75 patients, with the transhepatic route used in 48 patients and T-tube tract used in 27patients.⁵⁷ Complete clearance of bile duct stones was accomplished in 69 patients (92%).

Role of Cholecystectomy Following CBD Stone Extraction

After bile duct clearance is achieved by nonoperative methods, cholecystectomy generally is recommended in younger patients to decrease the risk of future cholecystitis and recurrent biliary colic. As many as 24% of patients have been found to require cholecystectomy at follow-up after endoscopic papillotomy at an average of 14 months.⁵⁸ Some authors have argued that sphincterotomy results in gallbladder stasis, bacterial overgrowth, and an increase in bile acids, and these may increase the risk of gallbladder cancer in 10 to 20 years.³ On the other hand, Dhiman and colleagues studied the changes in gallbladder emptying and lithogenicity of bile following endoscopic sphincterotomy in patients with choledocholithiasis and gallbladder in situ.⁵⁹ Sphincterotomy was found to decrease stasis of gallbladder bile, improve gallbladder emptying, and decrease the lithogenicity of bile as measured by prolongation of nucleation time. Meanwhile, there is much evidence to support leaving the gallbladder in situ after bile duct clearance in high-risk or elderly patients.^60-69 In a study of 191 patients (median age 76 years) in whom the gallbladder was left in situ post-ERCP, only 10 patients (5%) required subsequent uneventful cholecystectomy.⁶³ Twenty-six percent (49 patients) died during the review period from nonbiliary pathology. Kwon and colleagues followed 146 patients without elective cholecystectomy after endoscopic CBD stone removal for a period of 3 months or more to see if they could identify factors that predict subsequent gallbladder-related symptoms and need of cholecystectomy.⁶⁵ Fifty-nine patients had cholelithiasis, whereas 87 patients had no gallbladder stones. During a mean follow-up of 24 months, seven patients (5%) underwent cholecystectomy, on average 18 months after ERCP as a result of acute cholecystitis (four patients), biliary pain (two patients), and acute pancreatitis (one patient). Nine patients (6%) died of causes unrelated to biliary disease. Interestingly, Cox regression analysis revealed that the need for subsequent cholecystectomy did not correlate with age, sex, presence of gallbladder stones, number of gallbladder stones, or underlying disease. Kullman and colleagues found that at a median observation time of 42 months, cholecystectomy was needed in 11% (13 patients) of 118 patients with a gallbladder in situ after ERCP bile duct clearance.⁶⁶ Forty-nine patients (42%) died within 2 to 87 months after ERCP during the follow-up period. In another study of 33 elderly patients who were followed for an average of 42 months with gallbladder in situ after successful ERCP for choledocholithiasis, 3% (one patient) required cholecystectomy for acute cholecystitis, and 6% (two patients) had mild right upper quadrant pain, whereas 91% remained asymptomatic.⁶⁷ Over the course of the study, 30% of the patients died from nonbiliary causes.

The impact of gallbladder status on patient outcome after ESWL for complicated CBD stones was studied by a ­German group.⁶⁴ One-hundred twenty patients with an average age of 68 years (range 28-86 years) were followed for 3 to 9 years (mean 4 years). Thirty-seven had their gallbladder in situ, 27 had had a cholecystectomy after ESWL, and 56 had already undergone cholecystectomy prior to diagnosis of choledocholithiasis. During the follow-up period, 30% (36 patients) experienced biliary symptoms. However, there was no significant difference in the incidence of these symptoms between the three groups. Repeat ERCP revealed 28 cases of recurrent bile duct stones. Although not reaching statistical significance (p = .077), the recurrences occurred more often in the cholecystectomy groups. Given the multiple studies supporting leaving the gallbladder in situ after CBD clearance, it seems reasonable to perform cholecystectomies on high-risk or elderly patients as needed rather than prophylactically following nonoperative treatment of bile duct stones.

INTRAOPERATIVE

When presenting to the operating room for cholecystectomy, patients may have CBD stones confirmed by preoperative studies (eg, ERCP, MRCP, or EUS), be suspected to have CBD stones by clinical presentation, laboratory values, or transabdominal ultrasound, or have no suspicion of bile duct stones. At the time of surgery, intraoperative cholangiography (IOC) is the diagnostic method used most often. Mirizzi first introduced IOC to open biliary surgery in the 1930s.⁷⁰ With the universal acceptance of laparoscopic cholecystectomy, laparoscopic IOC has developed into a very useful method to evaluate the biliary tree. IOC and ECRP have similar sensitivity and specificity in CBD stone detection.⁷¹ The technique may be performed by injecting contrast material through a catheter introduced into the cystic duct via a variety of techniques.⁷² Cannulation rates with successful cholangiography range from 75 to 100%, and the use of fluoroscopy has become standard.^72,73 The reported sensitivity, specificity, PPV, NPV, and accuracy for laparoscopic cholangiography are 80% to 90%, 76% to 97%, 67% to 90%, 90% to 98%, and 95%, respectively, and these are comparable with the values for open IOC.⁷⁰ The rate of false positive IOC in a recent large review was found to be 0.8% (34 of 4209 patients).⁷⁴

Although approximately 10% to 15% of patients undergoing laparoscopic cholecystectomy harbor CBD stones, the need for routine IOC is a matter of much debate.⁷⁴ In a large Medline literature review, Metcalfe and colleagues found a 4% rate of CBD stones in eight laparoscopic cholecystectomy trials in which routine IOC was performed on 4209 patients without suspected bile duct stones preoperatively.⁷² This finding was felt to be consistent with previous reviews. On the other hand, in a total of 5179 patients without suspicion for bile duct stones who did not undergo IOC during laparoscopic cholecystectomy, 32 (0.6%) proceeded to develop symptoms from residual bile duct stones. Extrapolating this data, it would seem that of the 4% of patients with silent CBD stones at laparoscopic cholecystectomy, only 15% go on to develop symptoms from retained stones. In other words, 167 IOCs would have to be performed during laparoscopic cholecystectomy in order to detect one CBD stone that would go on to cause symptoms in patients without ­preoperative evidence of duct stones. This would result in eight unnecessary bile duct explorations or ERCPs.⁷² It is possible that stones that are not manifested preoperatively are of the size that can pass spontaneously into the duodenum, never presenting with symptoms.