Choledocholithiasis

Roles of Liver Biochemistry and Transabdominal Ultrasonography

Initial investigations in patients suspected to have bile duct stones should include liver biochemical tests and TUS. Normal liver function and biochemical tests are useful in excluding bile duct stones. In 1002 patients undergoing laparoscopic cholecystectomy, normal liver biochemistry accurately predicted the absence of bile duct stones. The negative predictive values (NPVs) of gamma glutamyl transferase (GGT), alkaline phosphatase, total bilirubin, alanine, and aspartate aminotransferases were all high and ranged from 94.7% (for total bilirubin) to 97.9% (for GGT). Unfortunately, the positive predictive value of only one abnormal liver biochemical test was just 15%.

TUS has a sensitivity of <50% in diagnosing bile duct stones. However, a bile duct stone seen during TUS is highly specific for stones found at ERCP and surgery. TUS is sensitive in detection of bile duct dilation (>6 mm in diameter), which is associated with the presence of bile duct stones. Mild biliary ductal dilation can be seen, however, in the elderly and in those with prior cholecystectomy. A TUS finding of a normal-sized bile duct has a 95% NPV of finding bile duct stones at ERCP.

No single parameter accurately predicts presence of bile duct stones in patients with gallstones. Most predictive models are based on a combination of clinical, biochemical, and TUS findings. For example, a patient older than 55 years with a serum bilirubin >30 µmol/L (1.8 mg/dL) and a dilated bile duct on TUS has a 72% probability of having bile duct stones found at ERCP. In 2010, the American Society for Gastrointestinal Endoscopy (ASGE) Standards of Practice Committee proposed stratification of symptomatic patients with gallstones into those at low (<10%), intermediate (10% to 50%), and high (>50%) risk of harboring bile duct stones. Very strong predictors include clinical cholangitis, bilirubin >4 mg/dL, and bile duct stone on TUS. Strong clinical predictors are a serum bilirubin of 1.8 to 4 mg/dL or a dilated bile duct on TUS (>6 mm with an intact gallbladder). The presence of any one very strong predictor or both strong predictors categorizes a patient at high risk of having a bile duct stone. Age greater than 55 years, clinical gallstone pancreatitis, and abnormal liver function tests other than a raised serum bilirubin are intermediate predictors of choledocholithiasis. The absence of any of these predictors is considered low risk. Those at high risk of harboring a bile duct stone should undergo preoperative ERCP and stone extraction. Patients at low risk should undergo cholecystectomy without further investigation. Patients at intermediate risk should be offered preoperative imaging, such as magnetic resonance cholangiography (MRC) or EUS to minimize diagnostic ERCP and the associated risk of post-ERCP pancreatitis (PEP).

Multiple groups recently prospectively applied the approach described earlier to patients with suspected bile duct stones and showed that the diagnostic accuracy for choledocholithiasis in the ASGE-designated high-risk group was only 59% to 69%. More refined stratification tools are needed to avoid unnecessary ERCPs. Among patients at intermediate risk of bile duct stones, a pooled analysis of 213 patients in 4 trials comparing an EUS-first versus direct ERCP approach found that initial use of EUS could reduce the need for ERCP in 67% of patients and the risk of PEP (relative risk [RR] 0.21; 95% confidence interval [CI] 0.06 to 0.83).

Roles of Magnetic Resonance Cholangiography and Endoscopic Ultrasonography

A meta-analysis found both high sensitivity (92%; 95% CI 80% to 97%) and high specificity (97%; 95% CI 90% to 99%) of MRC in the detection of bile duct stones. The sensitivity of MRC appears to be related to stone size. In one study, the sensitivity was 100% for stones approximately 1 cm in diameter and decreased to 71% for stones <5 mm in diameter. Another study, in which bile duct stone diameter >4 mm as cutoff on MRC was an indication for ERCP, had the best sensitivity (83%; 95% CI 78% to 88%) and specificity (66%; 95% CI 53% to 77%) for a positive ERC. False positives can also occur and are mostly related to air bubbles or bilioenteric anastomosis, such as a choledochoduodenostomy. MRC has the distinct advantages of being entirely noninvasive, usually not requiring sedation, and being unaffected by altered anatomy; it is contraindicated in patients with pacemakers, defibrillators, or other indwelling metallic objects.

EUS is an alternate modality to evaluate the bile duct among intermediate-risk patients. With the patient in the left lateral decubitus position and the transducer at the apex of the duodenal bulb, the second portion of the duodenum, and the papilla, the bile duct can be well visualized from the biliary hilum down to the ampullary region. Both radial and linear EUS have high sensitivity (~90%) and specificity (>93%) for the diagnosis of bile duct stones, though direct comparisons have not been performed ( Fig. 46.1 ). Importantly, sensitivity does not seem to be affected by stone size or bile duct diameter, though EUS is subject to operator variability. If bile duct stones are detected on EUS, ERCP can be performed during the same endoscopy session, reducing the need for an additional procedure and sedation and minimizing the risk of interval cholangitis.

A 2016 systematic review of eight prospective blinded studies that compared MRC and EUS against a gold standard of ERCP and/or intraoperative cholangiography identified 538 patients with intermediate probability of choledocholithiasis. The sensitivity of EUS (94%) was numerically superior to MRCP (84%). MRCP had a numerically higher specificity at 92% compared with EUS (89%), though statistical significance was not reached. For small stones and biliary sludge, EUS is likely more sensitive. Ultimately, the choice between MRC and EUS is often determined by resource availability and patient preference.

Patient Preparation

Patients with an indication for ERCP and endoscopic sphincterotomy (ES) for stone extraction should be evaluated for fitness for the procedure. Patient age is one of the first considerations, but procedural technical success and adverse events, including cardiopulmonary complications and PEP, were comparable between octogenarians and a younger patient group in a Korean multicenter review of over 600 patients. With the wider use of novel oral anticoagulant drugs and antiplatelets, endoscopists must become familiar with their indications (see Chapter 10 ). An updated ASGE guideline on the management of antithrombotic agents for patients undergoing gastrointestinal endoscopy outlined an approach weighing clinical urgency, bleeding risk of ERCP, and cardiovascular risk of withholding antithrombotics. In patients with sepsis mandating urgent ERCP, the use of anticoagulation should not defer the procedure, as waiting for ERCP >48 hours after hospitalization was associated with a threefold risk of persistent organ failure on a retrospective review of 203 patients with cholangitis. A short-length plastic biliary stent can be safely inserted for drainage without the need for ES. Otherwise, in nonurgent situations, ES is considered to be a procedure at high risk for bleeding, whereas endoscopic papillary balloon dilation (EPBD) without ES is categorized as low bleeding risk, and antithrombotics can be maintained. In patients at low cardiovascular risk with planned ES, anticoagulation can likely be held, but anticoagulation bridging should be considered in higher-cardiovascular-risk patients. The value of bridging therapy has, however, been questioned, as a recent multicenter randomized controlled trial of atrial fibrillation patients with a mean CHADS ₂ score of 2.3 showed that low-molecular-weight heparin bridging resulted in more major and minor bleeding but similar incidence of thromboembolism compared with no bridging. As 90% of the procedures in this trial were low risk for bleeding, the increased bleeding observed in the bridging group is likely even greater for high-bleeding-risk procedures like ES. Patients on dual-antiplatelet therapy at low and high cardiovascular risk anticipating ES should continue aspirin/nonsteroidal anti-inflammatory drugs, while holding thienopyridines for at least 5 days before endoscopy. On the use of preprocedural antibiotic prophylaxis (see Chapter 10 ), ASGE guidelines from 2015 recommend it in liver transplant patients and patients with biliary obstruction in whom incomplete biliary drainage is anticipated (those with multiple stones or with complex strictures). In those with biliary obstruction in the absence of cholangitis and in whom complete biliary drainage is likely after ERCP, it has not been shown conclusively that preprocedural antibiotics decrease post-ERCP cholangitis. We recommend routine use of antibiotics in immunocompromised patients. Recently a multicenter randomized controlled trial of over 2500 Chinese patients undergoing ERCP mostly for bile duct stones showed routine use of 100 mg rectal indomethacin 30 minutes before ERCP in an unselect group of patients was superior to post-ERCP rectal indomethacin given selectively to high-risk patients for reducing PEP (RR 0.47; 95% CI 0.34 to 0.66; p < 0.0001). Adverse events, specifically bleeding and perforation, were similar in both groups. Patients with sepsis who have unstable hemodynamics or potential airway problems should be resuscitated and endotracheally intubated for the procedure, which can be performed in the supine or left lateral decubitus position. We prefer to perform ERCP with the use of anesthesia-administered propofol.

Biliary Cannulation, Cholangiography, and Sphincterotomy (see Chapters 14 and 17 )

ERCP with ES and stone extraction is a time-honored technique with a pooled success rate of ~90%, according to a 2013 meta-analysis of ERCP intraprocedural performance and quality. We recommend wire-guided biliary cannulation because injection of contrast can increase hydrostatic pressure and cause mechanical trauma to the pancreatic duct. In a pooled analysis of controlled trials that compared contrast versus wire-guided techniques of biliary cannulation, primary cannulation success was significantly higher and PEP was significantly lower with wire-guided cannulation (RR 0.19; 95% CI 0.06 to 0.58). We use a pull-type sphincterotome, typically with a 25-mm cutting wire preloaded with a 0.025-inch or 0.035-inch guidewire that has a hydrophilic terminal portion. Flexing of the sphincterotome provides additional angulation for cannulation. Upon deep cannulation of the bile duct, the sphincterotome is advanced above the cystic duct junction. Bile should first be aspirated, followed by injection of half-strength contrast for better visualization of stones. In patients with cholangitis, frank pus may be aspirated, and overdistension of the bile duct with contrast should be avoided as an increase in biliary pressure can induce bacteremia. Prompt biliary drainage can be accomplished by insertion of a 7-Fr nasobiliary drain or a short plastic stent to prevent calculous impaction. Although several randomized controlled trials showed no difference in biliary drainage and adverse events, we prefer to use a short plastic stent, as nasobiliary drains can kink at the back of the oropharynx, can be a source of discomfort, and are prone to accidental dislodgment, particularly in delirious or elderly patients.

When stone extraction is permitted by the patient's clinical status, an optimal cholangiogram should be performed to highlight stone characteristics such as size, shape, location, number, and presence of impaction, and duct characteristics, including distal bile duct diameter and the presence of strictures, as they will influence the approach to stone extraction. An occlusion cholangiogram may be needed to prevent contrast drainage if prior sphincterotomy was performed. However, the sensitivity of cholangiography in diagnosing bile duct stones is imperfect, ranging from 67% to 94%. In particular, small stones can be missed in a spacious and dilated bile duct. To visualize the bile duct behind the duodenoscope, the endoscope should be advanced into a semilong position. When a cholangiogram is obtained but no apparent stones are identified, the decision to perform an empiric ES is influenced by the pretest probability of finding a bile duct stone. When the clinical suspicion is high (e.g., stone seen on TUS or for a patient with clinical cholangitis), we advocate a more liberal policy of performing empiric ES. This is consistent with results of a randomized study showing that in patients with clinical cholangitis and cholelithiasis but no bile duct stones on ERCP, empiric ES led to a reduction in recurrent stones and sepsis (hazard ratio [HR] 0.305; 95% CI 0.095 to 0.975; p = 0.045) at a mean follow-up of 22 months compared with no ES. In most circumstances, the risk of missing a bile duct stone outweighs that of empiric ES. When expertise is available, ancillary techniques like intraductal ultrasonography or EUS may aid in resolving the dilemma.

ES should be performed in a stepwise manner with the distal portion of the cutting wire in the duct and with minimal wire tension (see Chapter 17 ). Excessive wire tension during ES can otherwise result in a zipper cut with coagulated tissue being forced open, resulting in perforation and bleeding. We prefer to use a blended current in a pulsed mode, such as ENDO CUT (Erbe Elektromedizin GmbH, Tuebingen, Germany). A meta-analysis of 4 randomized studies with over 800 patients showed that ES with pure-cut mode resulted in more intraprocedural bleeding, with no difference in PEP, compared with mixed current.

The diameter that a sphincterotomy can be created depends on the shape of the papilla and diameter and configuration of the distal bile duct. In the presence of a prominent papilla and a dilated bile duct with a flat and square distal end, a full ES to the top of the overlying transverse fold is possible. In a patient with a small papilla and a narrow, tapering distal bile duct, only a limited ES is possible. As muscle fibers to the biliary sphincter are severed, one should see free bile flow. Another sign of an adequate sphincterotomy is free passage of a fully bowed sphincterotome with a 25-mm cutting wire through the sphincterotomy orifice.