Management of Benign Biliary Strictures: Biliary Enteric Anastomosis vs Endoscopic Stenting

Hypothesis  Although advances in endoscopic procedures have provided alternative options for relieving biliary obstructions, the overall chance of cure for patients with benign biliary stricture is the same using surgical or endoscopic treatment.

Design  Case-control study.

Setting  Tertiary care university hospital.

Patients  Of 163 patients referred for treatment with diagnoses of benign strictures of the common bile duct between January 1, 1975, and July 1, 1998, we studied 42 patients with postcholecystectomy stricture and a follow-up longer than 60 months. Twenty of these patients were treated with endoscopic stenting and 22 with surgery (hepaticojejunostomy, choledochojejunostomy, or intrahepatic cholangiojejunostomy).

Main Outcome Measures  Postoperative mortality and morbility and long-term outcome. The rate of restenosis was also determined.

Results  Morbidity occurred more frequently in patients treated with endoscopic procedures than with surgical ones (9 vs 2; P = .34). Hospital mortality was 0%. Surgery achieved excellent or good long-term outcome in 17 of 22 patients. Endoscopic biliary stenting was successful in 16 of 20 patients. Overall, excellent or good outcomes were achieved in 34 patients (81%).

Conclusion  The ability to achieve steady, long-term results confirms hepaticojejunostomy as the best procedure in the treatment of benign biliary strictures, even if endoscopic procedures are gaining a new role in the treatment of a greater number of patients.

STANDARD surgical techniques offer a good chance of cure for the majority of patients affected by extrahepatic benign biliary stricture.^1-11 Nevertheless, operative repair has a long-term recurrence rate of stricture in 10% to 30% of patients. Advances in endoscopic procedures have provided alternative options of relieving biliary obstructions, but prolonged length of treatment and rehospitalization have to be considered if endoscopy is performed.^12-15 The purpose of this study was to compare the effectiveness of endoscopic and surgical treatments of benign stenosis of the bile duct.

Between January 1, 1975, and July 1, 1998, 163 patients were referred for treatment to our department with diagnoses of benign strictures of the common bile duct. We considered for further analysis only patients referred to our institution from January 1, 1988, the date that the endoscopy unit for biliary disease started to operate. Only patients with postcholecystectomy bile duct stricture not previously treated were enrolled. Medical records were prospectively analyzed. Between January 1, 1988, and July 1, 1993, 42 patients with major bile duct injuries that occurred during cholecystectomy were treated at the First Department of Surgery, University of Rome "La Sapienza," Rome, Italy.

The level of biliary obstruction was defined according to the Bismuth classification.¹⁶ The degree of common bile duct dilatation was classified as follows: A, width above the stricture is less than 1.5 cm; B, width is 1.5 to 3.0 cm; and C, width is more than 3.0 cm.

Patients were randomly assigned into 2 treatment groups (endoscopy vs surgery) according to random-number tables. Informed consent was obtained from all participating subjects.

In patients treated with endoscopy, the procedure was performed after obtaining visualization of the biliary tree by endoscopic retrograde cholangiography. After a diagnostic endoscopic retrograde cholangiography, a small sphincterotomy was done and initial dilatation was completed if a firm stenosis interfered with the passage of the catheter. One or more 10F- to 12F-gauge straight endoprostheses with a proximal and distal flaps were inserted over the guidewire, with the catheter bridging the stenosis. Stent exchange was performed routinely but was considered only if clinically necessary for cholangitis, dislocation, or clogging or for clinically evident cholestasis.

In patients treated with surgery, the restoration of biliary enteric continuity was achieved with a defunctionalized Roux-en-Y jejunal loop by means of hepaticojejunostomy, choledochojejunostomy, or intrahepatic cholangiojejunostomy. In the last case, the right and left hepatic ducts were exposed by excising the liver tissue as previously described.¹⁷ Transanastomoting splinting was never used in the surgical procedure.¹⁷

Follow-up was accomplished by trimonthly clinical examinations, ultrasonographic liver scans, and biochemical tests for liver function and cholestasis.

Long-term outcome was classified as excellent if no symptoms were referred; good, if symptoms were transitory; fair, if medical therapy was requested; and poor, if recurrent stricture was established. The rate of restenosis was also determined.

Univariate comparisons were conducted using Fisher exact tests, χ² tests, t tests, and analysis of variance, where appropriate. Multiple regression analysis was then performed to determine predictive factors.

Demographic and clinical characteristics of patients are listed in Table 1. In all patients a preoperative ultrasonographic scan of the liver and abdomen was performed. In 39 patients (93%) an endoscopic retrograde cholangiopancreatography was performed, and 3 patients (17%) underwent intravenous cholangiography.

Endoprostheses placement was successful after a mean of 1.75 procedures (range, 1-7). Thirteen patients received one 10F to 12F endoprosthesis, 5 received 2 endoprostheses, and 2 received 3 endoprostheses. Four patients developed 9 early complications. In patients who underwent endoscopic biliary stenting, 3 experienced hemorrhage, 1 had dislocation and 4 clogging of the endoprotheses, and 1 patient developed pneumonia. Overall, stent exchange was needed in 8 instances due to endoprostheses clogging (n = 4), endoprostheses dislocation (n = 1), or cholangitis (n = 3).

End-to-side hepaticojejunostomy was performed in 16 patients. In the remaining 6 patients, the types of biliary reconstruction were end-to-side choledochojejunostomy (5 patients)and intrahepatic cholangiojejunostomy (1 patient). Complications included bile leakage in 1 patient who had hepaticojejunostomy, and hemorrhage in 1 patient who had choledochojejunostomy that required surgical reintervention.

Mean ± SD follow-up time of patients treated with endoscopic procedures was 89.7 ± 17.6 months in the endoscopy group and 92.4 ± 18.7 months in the surgery group. Initial hospitalization was longer for surgical patients than for those undergoing endoscopic treatment (10.7 days vs 3.9 days; P<.001). Even if rehospitalization for complications or clinical evaluation was considered, total hospital stays differed between the groups (10.8 days vs 5.8 days; P<.001). Only procedure-related major complications occurred. Morbidity occurred more frequently in patients treated with endoscopic procedures than with surgery (9 vs 2; P = .34). Hospital mortality was 0%.

Surgery achieved excellent or good results in 17 patients (77%), and fair or poor results in 5 patients (23%). Endoscopic biliary stenting was successful (excellent or good) in 16 patients (80%) and fair or poor in 4 patients (20%). Overall, excellent or good results were achieved in 34 patients (81%). Eight patients (19%) experienced fair or poor results. Four of these 8 patients underwent only medical treatment. The other 4 patients were submitted to surgical therapy for anastomotic stricture (1 patient) or restenosis after endoprostheses removal (3 patients).

No statistically significant differences were demonstrated between the treatment groups, except for length of hospital stay and for American Society of Anesthesiologists score, which reached borderline significance. In univariate analysis, factors influencing the outcome were the degree of dilatation, postoperative complications, and surgical treatment (Table 2). Multivariate analysis confirmed the degree of dilatation as the independent factor affecting the long-term outcome (β = −0.46 ± 0.21 [SE], B = −0.62 ± 0.28, t = −2.1; P = .04).

Benign strictures of the bile ducts represent a significant clinical problem, most often occurring in young patients. The treatment goal for these patients is long-term absence of symptoms and need for further hospitalization.^11,17-19

Previous reports have suggested hepaticojejunostomy as the best treatment for benign biliary stenosis.^{9,10,17,20,21} The vascular patterns of the bile duct and the pathogenesis of the stricture advocate performing the biliary enteric anastomosis as high as possible on the biliary tree.^22-25 However, the outcome of repair of bile duct injuries is difficult to evaluate because reports in the literature do not discern morbidity and mortality directly associated with primary or secondary stricture repair. Moreover, scant information is given on the early outcome of uncomplicated cases vs those with chronic liver impairment or portal hypertension as a consequence of progression of the biliary obstructive damage.

Although excellent results are achieved by surgery, advances in endoscopic technology have suggested wider use of nonoperative options, which in the past were mostly restricted to the treatment of biliary fistulas and to patients with associated medical disease.^13,14,26,27

Attempts to compare these new endoscopic strategies and surgical results have been difficult because of the variable criteria in selecting the patients, different length of follow-up, and inconsistency in defining a successful procedure.^1,2,28,29 In our study, all surgical and endoscopic treatments were performed at one institution by one group of endoscopists and surgeons; consequently, surgical and stenting techniques such as the types of stents, the alignment of stenting, and the assessment of results were all standardized.

The immediate success of nonoperative management should be considered with caution because of late complications influencing the outcome of these patients.^12,15 In 35% of our patients, clogging and cholangitis were the main pitfalls of endoscopic procedures.

Bacterial colonization, bile viscosity, and stent characteristics have been reported to play a causal role in transient or definitive obstruction of endoprostheses and in the onset of cholangitis.³⁰ Periodic oral administration of antibiotics and use of large-caliber Teflon stents without side holes have been shown to improve patency and to reduce the impact of cholangitis.^15,31,32 Balloon dilatation of the stricture has been reported to improve the results of biliary stenting.⁹ However, in cases of strictures unyielding to pneumatic distension, endoprosthesis alone has proved of value in preserving luminal dimensions due to favoring the maturation of the scar tissue and allowing biliary duct patency with periprosthetic flux.^9,14 Stent migration is usually promoted by small endoprostheses diameter. Therefore, in our series only 10F to 12F stents were used, which achieved good results.

The short-term success rate of operative endoscopic retrograde cholangiography found in the present series compares favorably with that of surgery, although a prevalence of procedure-related morbidity in endoscopy-treated patients was detected, possibly because of the initial "learning curve." The low rate of endoscopic procedures found to be necessary in the present trial mostly reflects the policy adopted of changing the stent only at the first clinical appearance of complications. The most widely adopted policy of a routine change while not reducing the number and the risks of early complications, would have resulted in a much higher number of endoscopic procedures, related risks, hospital accesses, and consequent costs.^2,12,13,15

Because a recurrent stricture, mostly occurring in the first 5 years, can arise at any time after primary treatment, 8 to 10 years of follow-up is needed to determine whether repair has been successful. Our analysis was focused on patients with a length of follow-up longer than 60 months. Long-term results following both endoscopic procedures and surgical treatment were compared. The long-term outcome in our series is as good as that of other reports, which confirm that more than 80% of patients can be treated successfully with endoscopic or surgical procedures.

Univariate analysis showed the degree of dilatation and type of treatment as factors statistically correlated to long-term positive outcomes and highlighted surgery as the best therapeutic option in the treatment of biliary tract strictures. Although confirming the former correlation, multivariate analysis did not show any correlation between the kind of procedure adopted and the outcome of the treatment. A high degree of dilatation was the only factor affecting the long-term outcome. Dilatation provides not only a wide anastomosis, but also an easy and effectual long-term stenting of the biliary tree.

By proving the safety and the effectiveness of endoscopic procedures, our findings confirm that this conservative treatment should not be restricted to a select group of patients. Indeed endoscopic stenting should be considered greatly useful not only as a bridge between the time of diagnosis and that of more aggressive procedures or for patients with biliary fistula, but as a definitive treatment in patients with associated medical diseases. However, when considered as a definitive treatment, endoscopic stenting, even if performed by an experienced team, could be hindered by the major disadvantage of multiple procedures that must be repeated to accomplish success.^{2,12,13,15,33} However, as was the case for 3 of our patients, endoscopy is not an either-or decision; when it is no longer successful, surgery is still feasible.¹⁴

In conclusion, our results confirm hepaticojejunostomy as the best surgical procedure, with evidence that endoscopic stenting is an authoritative alternative tool in the treatment of biliary benign stricture.

Reprints: Adriano Tocchi, MD, First Department of Surgery, University of Rome "La Sapienza," via Bruno Bruni, 94, 00189 Rome, Italy (e-mail: gluca.costa@iol.it).