ERCP Following Laparoscopic Cholecystectomy: A Safe and Effective way to Manage CBD Stones and Complications

The efficacy of ERCP in detecting and treating post-laparoscopic cholecystectomy problems was examined in a series of consecutive patients undergoing directed examination of the biliary tree over a two-year period. Three major diagnostic groups were identified: leaks and bile duct injuries (n = 9), retained common bile duct stones (n = 18), and post-cholecystectomy pain (n = 13). These diagnostic groups differed in degree of abnormal bilirubin (p = .004) and time between surgery and ERCP (p = .0005). Diagnosis of a post-operative complication was successful in 92% of attempted cases. Therapy was successful in 92% of attempted cases. Three patients developed mild pancreatitis as a result of ERCP. This series underscores the efficacy of a multi-disciplinary approach to problems which occur after laparoscopic cholecystectomy.

With the rapid endorsement of LC, the role of the biliary endoscopist in diagnosis and management of two major post-operative problems is increasing. The most important ofthese is bile duct injury. Regional experiences with LC have demonstrated an increased rate of bile duct injury compared with OC. This rate appears disproportionately high in the early experience of the surgeon5-7. The second dilemma is retained common bile duct stones which require removal. Patients with these post-operative problems may be referred to the biliary endoscopist for diagnosis or therapy. biliary tree. Demographic data included age, gender, and referral area. Operative data included results of intraoperative cholangiogram (ifperformed), and whether conversion to OC was required. Pre-ERCP data included number of days from surgery until referral, type of symptoms (e.g., fever, abdominal pain, fistulous drainage), requirement for percutaneous drainage of a biloma, and elevation of liver chemistries (upper limit of normal: AST 45 u/l, ALT 45 u/l, AP 110 u/l, TB 1.2 mg/dl). ERCP data included type of therapeutic procedure(s) performed, number of procedures required to complete therapy, duration of stenting, and type and outcome ofother invasive therapy required (percutaneous or surgical). Success of therapeutic ERCP was defined as (1) avoidance of surgical correction, and (2) resolution of the specific problem leading to the procedure (e.g., resolution of leak or stricture, removal of retained common bile duct (CBD) stones). Complications were defined according to Cotton et al. Statistical tests were performed using nonparametric methods (Wilcoxon R.ank Sum test and Kruskal-Wallis test for multiple groups) as data were not assumed to have normal distribution.

RESULTS
During this 2 year period, 58 patients were referred to the DUMC biliary endoscopy service. Six patients were excluded for examination not directed at the bile duct (n 5) and routine follow-up of primary sclerosing cholangitis (n 1). Fifty-two patients were included for study. There were 19 males and 33 females. Median age was 43.5 years (range, 16-93 years). LC had been performed outside DUMC in 36 cases (69.2%). For the purposes of this review, patients were categorized into three major diagnostic groups: Group leaks and bile duct injuries, Group 2 retained CBD stones, and Group 3 postcholecystectomy pain (post-LC pain) ( Table 1). Patients most commonly presented with abnormal liver chemistries and abdominal pain ( Table 2). Liver chemistries were elevated in 81% of patients with median (range), AST 43 u/1 (5-400), ALT 51 u/I (1-440), AP u/1 162 (52-678), TB   Table 3). The median delay between surgery and ERCP was 21 days(range, 1-354) ( Table 4).
Endoscopic therapy was initiated in 12 patients (Table  6). Six cases of bile duct injury or leak were treated with endoscopic prosthesis (EP). Leaks from the CD stump (n 5) and RHD (n 1) were stented for a median duration of 37 days. Three patients with strictures ofthe CBD (n 2) and CHD (n 1) underwent a trial of EP for a median duration of96 days. EP was successful in 5 of6 leaks (83%) and 2 of 3 strictures (67%). Three of 3 leaks (100%) were successfully treated by ES alone. Two patients underwent endoscopic sphincterotomy (ES) for concomitant stone extraction. ES was not required for stent placement.
Surgical correction was performed as primary therapy in 6 patients: hepato-jejunostomy was performed for complete CBD obstruction (n 2), CHD stricture (n 2), and disrupted CBD (n 1); patient with a leak of the CD stump underwent surgical repair.
Diagnostic ERCP failed in one patient who presented with a biloma; a RHD leak was indentified by percutaneous cholangiogram and the patient underwent hepato-jejunostomy. Access for therapeutic ERCP failed in one patient who presented with a dislodged tube which had been placed in the CBD at surgery. At that time, complete transection of the CBD was recognized and immediate choledochoduodenostomy was performed. The patient was successfully treated with a percutaneously placed stent. EP failed in the patient with the RHD leak who underwent subsequent Longmire procedure. EP (total duration, 300 days) failed in one CBD stricture which occurred after primary repair of a transected duct. This patient underwent successful hepato-jejunostomy.
Overall, definitive therapy was provided by EP in 7 of9 patients (78%) and ES in 3 of 3 patients (100%). Overall Fourteen patients were evaluated for asymptomatic elevations of liver chemistries (n= 1) or post-LC pain (n= 13). Abnormal liver chemistries were present in 4 (28.5%): median (range) AST 37 u/1 (8-81), ALT 32 u/1 (11-112), AP 111 u/1 (55-391), TB 0.55 mg/dl (0.1-1.9). Diagnostic ERCP failed in patient who could not tolerate the procedure. Two patients had evidence for papillary stenosis with delayed drainage of contrast from the bile duct; no therapy was undertaken in one while one patient underwent surgical sphincteroplasty. One patient had evidence of pancreatitis and normal bile ducts. The remaining 10 patients had normal bile ducts. For this entire series, diagnostic procedures alone were attempted in 25 patients; therapeutic procedures were attempted in 27 cases. Diagnostic ERCP failed in 2 of 25 cases (8%). Access for therapy failed in of 27 cases (3.7%). Three complications occurred in this series. Two patients evaluated for post-LC pain developed  mild pancreatitis. One patient who underwent ES and stone extraction with interim stent placement developed mild pancreatitis (Table 7).
Differences in liver chemistry abnormalities were statistically significant between the 3 major diagnostic groups. (Table 3 and Figure 1) The difference in delay time between surgery and ERCP for the 3 major diagnostic groups was very highly statistically significant, p .0005 ( Figure 2).

DISCUSSION
In the absence of prospective, randomized trials, the enthusiasm for LC has been tempered by careful scrutiny of associated complications. Appeals for circumspection in the use of the procedure and credentialing of trainees are now familiar9-1. Attention is turning to the long-term outcome and management ofcomplications after LC. The therapeutic endoscopist can serve an important role in the Bile duct injury has emerged as the most serious complication of LC and appears related to early experience 2, 6. The classic pattern of injury occurs when the common bile duct is mistaken for the cystic duct and a portion of the misidentified duct is then removed. The major post-operative manifestations are bile leak and biliary obstruction The management strategy for injuries which are suspected post-operatively will depend on the expertise available. Adequate visualization of the leak site, percutaneous drainage of fluid collections, and relief of distal obstruction are rules for successful management .R adiologic and endoscopic management have the advantage of avoiding further surgical morbidity.
Percutaneous transhepatic drainage provides definitive management for biliary fistulae in 50-80/,, ofcases 4 5. This method may also temporize by diverting bile flow away from the bile duct defect until the patient's clinical condition improves for surgical repair. However, in addition to the risk of bleeding due to transhepatic puncture, percutaneous drainage may be mor.e difficult in this group of patients since the bile ducts are effectively decom-pressed. For these reasons, we reserve this approach for patients who have failed endoscopic access.
Success rates exceeding 75% for ES and EP have been demonstrated for management of post-operative biliary fistulae 3. ,6. The endoscopic management of similar sequelae after LC appears to be increasing. Wootton reported successful treatment of 2 of 3 cystic duct leaks treated with EP'7. Kozarek et al., have reported the most complete series to date. Three CD leaks, 6 CBD leaks, and 2 CBD transections were identified at ERCP. Bile leaks sealed in the 9 patients treated endoscopically. The discrete referral base and consecutive nature ofthe patient sample allowed better characterization of the range of injuries and man-agement6. In our case series, we have demonstrated successful treatment in 5 of 6 patients (83%) with EP and 3 of 3 patients (100%) with ES. The similar success rates for ES and EP suggest an ana!egous mechanism: relief of relative obstruction at the sphincter of Oddi. 3 The retrospective nature and small number of cases limit valid comparisons of the two approaches. ES has been recommended in this setting because ofinfection and blockage associated with stents '8. We prefer EP over ES to avoid the increased risk of bleeding, perforation, and pancreatitis associated with ES8. The limited duration of stent placement (median, 37 days) obviates the need for routine stent exchange since these complications are unlikely to occur. We do not routinely perform ES for placement of the stent.
High success rates for endoscopic management of post-operative biliary strictures has been similarly demonstrated. Good to excellent results were shown by Davids, et al. in 83% of patients treated with EP for one year, however, restricturing occurred in 17% , 9. Accounts of this modality for managing post-LC strictures are limited to two reports. Weber treated a stricture related to a CD clip with EP2. Wootton's series included 2 CBD strictures; of these was successfully treated with EP'7. We attempted therapy in 3 patients with strictures after LC and achieved resolution in 2. Only long-term follow-up will define the success of this therapy. The reasons for limited endoscopic experience in this clinical setting are unclear. We have not attempted to describe our referral base, therefore the denominator and rate for this particular injury is undefined in our series. The referral pattern for complications of this relatively new surgery may be confined to the surgical community 2. Complete or complicated obstruction will require expert surgical correction.
However, our results suggest that a trial of endoscopic therapy is warranted in the appropriate clinical setting, as for any benign bile duct stricture.
Are retained bile duct stones "complications"? This is a matter of semantics. The endoscopic removal of retained stones is conventional; our success rate after LC is high (100%). Currently, attention is appropriately focused on pre-operative identification and management of patients with suspected choledocholithiasis2-23.
Davidoff et al., emphasized that prolonged or severe abdominal pain post-LC should heighten suspicion for a complication, since one of the benefits of LC is relative lack of post-operative discomfort. We observed a distinct clinical presentation for the 3 major diagnostic groups based on timing of ERCP and degree of abnormal liver chemistries, Patients with findings at ERCP of serious consequence (bile leaks, strictures and retained stones) underwent study within 1-2 weeks of surgery and liver chemistries were more strikingly abnormal. In contrast, patients in the group with post-LC pain had relatively unremarkable elevations in liver chemistries and presented 3.9 months after surgery. A low threshold for study is reflected in earlier referral for patients with objective evidence of significant disease (e.g., presence of biloma, abnormal cholangiogram, abnormal liver chemistries). Our review was retrospective and the sample was drawn from patients referred to a tertiary center for biliary endoscopy. We conclude that a low-threshold for diagnostic evaluation of potential post-operative complications should be maintained, especially in patients with abdominal pain or abnormal liver chemistries. This case series demonstrates that ERCP is ideally suited as an adjunct to laparoscopic surgery to address both diagnosis and therapy ofcomplications.
Outcomes research defining the nature and tong-term significance of post-LC complications is required. The prospect is both exciting and formidable. This series demonstrates an excellent example of multidisciplinary collaboration.