Treatment for Retrieved Common Bile Duct Stones During Laparoscopic Cholecystectomy: The Rendezvous Technique

Objective  To determine the feasibility and efficacy of the laparoscopic intraoperative rendezvous technique for common bile duct stones (CBDS).

Design  Case series.

Setting  Verona University Hospital, Verona, Italy.

Patients  A total of 110 patients were enrolled in the study; 47 had biliary colic; 39, acute cholecystitis; 19, acute biliary pancreatitis; and 5, acute biliary pancreatitis with associated acute cholecystitis.

Interventions  In all patients, CBDS diagnosis was reached by intraoperative cholangiography. Intraoperative endoscopy with rendezvous performed during laparascopic cholecystectomy for confirmed CBDS; for such a procedure, a transcystic guide wire was positioned into the duodenum. Intraoperative endoscopy with rendezvous was performed for retrieved CBDS during a laparoscopic cholecystectomy.

Main Outcome Measures  Laparoscopic rendezvous feasibility, morbidity, postprocedure pancreatitis, and mortality.

Results  The laparoscopic rendezvous proved to be feasible in 95.5% (105 of 110 patients). The rendezvous failed in 3 cases of successfully performed laparoscopic cholecystectomy, and a conversion of the laparoscopy was needed in 2 cases of successful rendezvous. Two major complications and 2 cases of bleeding were registered after sphincterotomy was successfully performed with rendezvous, and severe acute pancreatitis complicated a traditional sphincterotomy performed after a failed rendezvous.

Conclusions  Rendezvous is a feasible option for treatment of CBDS; it allows one to perform only 1 stage of treatment, even in acute cases such as cholecystitis and pancreatitis. Positioning of the guide wire may allow reduced complications secondary to papilla cannulation but not those of the endoscopic sphincterotomy.

Treatment of common bile duct stones (CBDS) complicating cholecystolithiasis has evolved since endoscopy was introduced for diagnostic and therapeutic purposes.^1,2 The role of the endoscopic approach in the diagnosis of CBDS has been supplanted by new noninvasive imaging techniques such as magnetic resonance imaging and magnetic resonance cholangiopancreatography, as well as by endoscopic ultrasound.^3-5 Increasing expertise in laparoscopy has also led to investigation of a 1-stage treatment through the cystic duct or using choledocotomy,^6-8 but such approaches may still be limited by the biliary anatomy and diameter of stones⁹ and may also be questionable in the setting of acute inflammation such as cholecystitis and/or pancreatitis. The transpapillary approach is, therefore, not yet ready to be definitively abandoned and is still recommended as a treatment option.¹⁰

Intraoperative sphincterotomy has been described,^8,11,12 but complications of endoscopy in this setting may be increased owing to the supine position of the patient.^13,14 Introduction of a guide wire through the cystic duct may reduce the failure rate of papilla cannulation and related complications as a result.

In this prospective, unselected series of CBDS, we report the results of the laparoscopic rendezvous to assess its feasibility and efficacy.

From 2001 to 2008, a total of 780 patients underwent cholecystectomy at our institution, 747 by laparoscopy and 33 by laparotomy. Intraoperative cholangiography was performed in 575 cases and confirmed the diagnosis of CBDS in 120 patients. In 6 patients the CBDS were cleared by flushing with saline solution during laparoscopic cholecystectomy. In 1 patient with a previous gastrectomy, treatment was achieved by laparotomy, and in 3 patients, a first intention open cholecystectomy was planned; these patients were therefore excluded from the study. A total of 110 patients were submitted to first-intention laparoscopic cholecystectomy with intraoperative rendezvous for treatment of associated CBDS.

There were 63 women and 47 men (ratio, 1.34), with a mean (SD) age of 64.2 (17.2) years; 36 patient were younger than 60 years, 51 between 60 and 80 years, and 23 older than 80 years.

Indications for cholecystectomy were uncomplicated symptomatic gallbladder stones in 47 patients, acute gallstones cholecystitis in 39 patients, acute biliary pancreatitis in 19 patients, and acute biliary pancreatitis with associated acute cholecystitis in 5 patients. In 10 patients (9.1%) there were no preoperative predictive signs of CBDS because no jaundice or pancreatitis was observed at admission or reported in the medical history; results of liver function tests were in the reference range, and preoperative ultrasound did not show dilation of intrahepatic and extrahepatic bile ducts. In 54 patients, preoperative suspicion of CBDS was high based on abnormal results of liver function testing, with CBD dilation on ultrasound or CBDS evidence on magnetic resonance cholangiopancreatography, while in 46 patients, at least 1 abnormal result on the liver function test or ultrasound was considered an intermediate risk of having CBDS.

For the rendezvous technique, A 150-cm double-channel balloon catheter 6F in diameter and a 450-cm Teflon guidewire of 0.035 in was used. Laparoscopic cholecystectomy was performed with a traditional 4-trocar technique. After dissection of the Calot triangle and section of the cystic artery, a small incision in the upper part of the cystic duct allowed introduction of a 5F catheter for intraoperative cholangiography. Once CBDS was confirmed and did not resolve by flushing, the endoscopic team was alerted. Before endoscopy, the cholangiography catheter was removed and the double lumen balloon catheter was introduced through the cystic duct. The balloon was filled with 2 mL of air to make it more visible using radioscopy. Positioning the tip of the catheter under radioscopic guidance at the level of the middle to distal part of the common bile duct facilitated passage of the guide wire over the duct stones, and then into the duodenum through the papilla. At this stage, the tip of the guide wire was viewed using endoscopy (Pentax ED-3440T endoscope; Pentax, Golden, Colorado) with a polypectomy snare and pulled out of the patient's mouth by combined maneuvers. A traditional sphincterotome was introduced along the guide wire that allowed direct cannulation of the papilla and sphincterotomy. Stone extraction was also performed using the guidance of the wire with a second balloon catheter, as well as Dormia basket when necessary. Contrast was used through the second lumen of the transcystic catheter each time it was required. At the end of the procedure, a nasobiliary catheter was left in place after the cystic stump was clipped. Cholecystectomy was then completed after aspiration of the stomach.

A cholangiography was systematically performed through the nasobiliary catheter on the first or second day after surgery before removing it.

Because patients were not randomized into treatment arms, approval to collect and analyze the data in clinical records was received from the Department of Surgery, University of Verona, according to the rules of our institutional review board. The institutional bodies that granted permission waived the need to obtain patients' informed consent for inclusion in the study. All patients gave informed consent for the surgical procedures and were treated according to the ethical and clinical standards of the University Hospital of Verona.

Laparoscopic treatment of CBDS was successfully completed with rendezvous in 105 of 110 patients (95.5%). In 3 patients, it was not possible to pass the guide wire through the papilla; in these cases, laparoscopic cholecystectomy was successfully completed, and CBDS treatment was performed with a traditional intraoperative sphincterotomy. Conversion of the laparoscopic approach was required to achieve the rendezvous by laparotomy in a patient with a previous gastrectomy. In this case, the guide wire allowed the endoscope to reach the duodenal stump, but cannulation of the papilla could not be performed safely owing to difficulty positioning the endoscope; after conversion to open surgery, handling the endoscope made this maneuver feasible, and endoscopic sphincterotomy was achieved. Another conversion of the laparoscopic approach was required owing to duodenal perforation after sphincterotomy was completed.

Intraoperatively, difficult dissection owing to severe inflammatory changes involving the infundibulum and the hepatico-duodenal ligament was described in 41 patients (37.3%); 27 patients had cholecystitis; 7, gallbladder empyema; 4, acute biliary pancreatitis; and 3, acute pancreatitis with associated cholecystitis. No other intraoperative complication was observed other than the duodenal perforation. The mean (SD) operative time was 155 (57) minutes. Intraoperative cholangiography showed a single stone in 37 patients and multiple stones in 73 patients.

Overall postoperative morbidity was 16.4% (18 of 110). Complications were secondary to the endoscopic intervention in 4 patients. Bleeding from the sphincterotomy, treated by a second endoscopy, was observed in 2 patients, and a delayed retroperitoneal phlegmon requiring surgical drainage was observed in a patients who were successfully treated with the rendezvous technique, while acute pancreatitis was observed after traditional sphincterotomy was performed intraoperatively after a failed rendezvous attempt. In the latter case, acute pancreatitis had a severe course, and the patient died after 3 months. We observed no other deaths, and the overall mortality rate was 0.9%.

Nasobiliary cholangiography performed postoperatively showed residual stones in the common bile duct in 15 patients (13.6%). A second endoscopy was necessary in 5 patients to complete common bile duct treatment, while in 10 patients with small stones (less than 3 mm in diameter), a 24-hour wait-and-see policy allowed spontaneous clearance to be confirmed by further nasobiliary cholangiography.

The mean postoperative stay was 6.7 (10) days, and the mean overall hospitalization was 10.3 (11.6) days. Sensitivity and specificity of preoperative diagnosis of acute inflammation were 70.7% and 50.7%, respectively, with a positive predictive value of 46% and a negative predictive value 74.5%. Acute inflammation was observed in 12 patients who were admitted with biliary colic, and 34 patients with a preoperative diagnosis of acute cholecystitis and/or pancreatitis did not show any sign of inflammation, either intraoperatively or during pathological examination of the gallbladder.

Table 1 illustrates the results based on intraoperative and/or pathological confirmation of acute inflammation and in Table 2 on preoperative suspicion of CBDS. The results do not show any significant role of the presence of acute inflammation because no statistical significance were observed when compared with nonacute cases. The only significant difference has been a shorter operative time when CBDS was suspected preoperatively.

Outpatient monitoring was scheduled for 1 month and 6 months after surgery; no clinical or radiological signs of residual CBDS were observed.

This study describes the results of a 1-stage treatment of CBDS with gallbladder in situ in a consecutive series that included patients with acute cholecystitis and acute biliary pancreatitis.

One- and 2-stage treatments are equally recommended in cases of CBDS, depending on local expertise¹⁰; however, 1-stage treatment allows a unique procedure that avoids a second dose of anesthesia, and decisions are based on intraoperative cholangiography findings, the reference method for CBDS diagnosis, without the diagnostic uncertainty of preoperative evaluation.

Minimally invasive 1-stage treatment of CBDS includes different options. The less invasive option is the transcystic approach, but it is limited by the cystic duct diameter and that of the stones that need to be removed through the cystic duct. In a recent series of 505 patients who were treated with laparoscopy, the transcystic approach was attempted in only 50% of cases, with a 75% success rate.⁶

The second approach is transcholedocal by way of a laparoscopic choledochotomy with primary closure or T tube closure of the main bile duct. Although this procedure has been favorably compared with the transpapillary approach, as described in a review⁷ and confirmed in a more recent randomized study,⁸ it is still technically demanding and has limitations related to main bile duct diameter; an increased risk of postoperative bile leakage and of late stenosis has also been described.⁹

Moreover, this approach in the setting of acute complications of the CBDS like acute cholecystitis and pancreatitis may be questionable because the inflammatory process involving the hepatico-duodenal ligament and the gallbladder may increase the risk of bile leakage, but no specific data for such cases have been reported in the literature.

The third option that has been described is intraoperative endoscopic sphincterotomy.^8,11,12 This transpapillary approach may be more difficult owing to the supine position of the patients and may increase the risk of complications because difficult cannulation, injection of contrast into the pancreatic duct, and precut sphincterotomy have been found to correlate with post–endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis.^13,14

With the rendezvous technique, cannulation of the papilla using the guidance of a wire may allow one to avoid these complications, even when patients are in the supine position. Studies reporting results of such a technique are mainly observational,^15-21 while patients were randomized in 3 studies.^22-24

One nonrandomized comparative study reported greater feasibility of the rendezvous technique and a significant reduction in postprocedure pancreatitis compared with the conventional 2-stage endoscopic sphincterotomy,²¹ while results of the randomized clinical studies appear controversial. In the randomized controlled trial by Rabago et al,²² the success rate of common bile duct clearance was not different between the two approaches, while Morino reported a significant difference favoring the rendezvous approach (80% vs 95.6%; P = .06).²³ Post-ERCP pancreatitis incidence was higher in the 2-stage treatment group (12.7% vs 1.7%; P = .03) in one randomized controlled trial,²² while such a difference was not observed in the other.²³ A third randomized study show a significant reduction in post-ERCP pancreatitis in a selected group of patients with related risk factors for postprocedure pancreatitis.²⁴ Finally, the mean hospital stay and mean hospital costs were significantly higher in the 2-stage treatment than in the rendezvous approach in 2 randomized controlled trials.^22,23

This study confirms the feasibility of the laparoscopic rendezvous technique in 95% of cases. The procedure failed because of the difficulty to pass the guide wire through the papilla, but minimally invasive treatment was still achieved with traditional sphincterotomy in such cases.

Successful single-stage treatment of acute cases is a major advantage of the procedure. In the other series, the rate of acute cholecystitis is not always clearly reported, while in our series, inflammatory changes in the hepatico-duodenal ligament have been observed in at least one-third of patients. This may explain the longer postoperative stay in our series compared with other observational studies^19,20; postoperative stays in acute cases were longer than in nonacute cases, although the observed difference did not reach a statistically significant level. Furthermore, systematic postoperative monitoring of CBD clearance also contributed to longer postoperative stays in acute and nonacute cases. However, neither morbidity nor conversion of the laparoscopic approach were increased in patients with local acute inflammation, allowing us to consider the procedure as safe as it may be for nonacute cases.

The procedure required the collaboration of 2 teams, the rendezvous between surgeons and endoscopist; radiologists may also be involved in the cholangiogram interpretation before the start of the procedure. In cases with a very high or a confirmed preoperative diagnosis of CBDS, operative time was significantly reduced because the presence of the endoscopic team was required in the operating room from the beginning of the intervention. Otherwise, endoscopists were called once diagnosis was confirmed, resulting in a longer operative time.

The complication rate of the rendezvous technique was 3.7% and was due to the sphincterotomy; 2 bleeding incidents and 2 perforations were observed with the rendezvous technique, while a death secondary to post-ERCP pancreatitis was observed in a case of failed rendezvous technique. Results of rendezvous complications reported by other studies^15-21 confirm that the rendezvous technique may reduce complications of papilla cannulation but it does not change the risk of the sphincterotomy.

No early investigation of the common bile duct after the endoscopic approach to CBDS treatment have been published in the literature. In the present series, nasobiliary drainage allowed such investigation and revealed apparently high incidence of residual CBDS. After intraoperative sphincterotomy, air bubbles enter the CBD and make diagnosis much less accurate because of false imaging and migration of small stones proximally in the hepatic ducts. However, only one-third of patients have been treated with a second endoscopy because, in the other cases, spontaneous passage through the sphincterotomy was demonstrated. It may be estimated that the incidence of potentially problematic residual CBDS was less than 5% of cases according to results of traditional ERCP.²⁵

A limit of the present study is that no data regarding late complications of the transpapillary approach have been registered because the follow-up period was too short.

However, the rendezvous technique seems a feasible option for the treatment of CBDS; it allows 1-stage treatment, especially when unexpected stones are retrieved during intraoperative cholangiography or when an acute complication of biliary lithiasis is present such as acute cholecystitis and/or acute pancreatitis. The guide wire reduces complications of the papilla cannulation but not those of the sphincterotomy. The rendezvous technique has, therefore, to be taken into account when deciding to treat CBDS, but the choice of procedure in the different clinical situations needs to be assessed with further randomized controlled trials.

Correspondence: Giuseppe Borzellino, MD, I Divisione Clinicizzata di Chirurgia Generale, Verona University Hospital Borgo Trento, Piazzale A. Stefani 1, 37126 Verona, Italy (gborz@libero.it).

Accepted for Publication: September 9, 2009.

Author Contributions:Study concept and design: Borzellino, Rodella, Politi, and Cordiano. Acquisition of data: Borzellino, Saladino, Catalano, and Minicozzi. Analysis and interpretation of data: Borzellino. Drafting of the manuscript: Borzellino, Saladino, Catalano, Politi, Minicozzi, and Cordiano. Critical revision of the manuscript for important intellectual content: Borzellino, Rodella, and Cordiano. Statistical analysis: Borzellino. Study supervision: Rodella, Politi, and Cordiano.

Financial Disclosure: None reported.