Management of benign biliary strictures by percutaneous interventional radiologic techniques (PIRT)

Abstract

Introduction. Some biliary strictures may be manageable by percutaneous interventional radiologic techniques (PIRT), but long-term efficacy of this approach is scarce. Methods. We reviewed retrospectively all patients with biliary strictures secondary to traumatic bile duct injury or strictured bilioenteric anastomoses. Patients in whom the initial management was by PIRT from 1998 through 2003 were selected. Subjects with sclerosing cholangitis, hepatic transplantation, or malignant strictures were excluded. Data were obtained from medical records and/or direct patient contact. Comparisons were made by Fisher's exact test and Wilcoxon rank-sum test. Results. Twenty-seven patients with biliary strictures were treated by PIRT. Mean age was 54 years (range 11–86). Most frequent etiology was laparoscopic cholecystectomy injury in 11 patients (41%). Eight patients (29%) had undergone biliary resection for malignancy, seven (26%) a pancreatoduodenectomy, and one for presumed ischemic cholangiopathy; no strictures were secondary to neoplastic recurrence. PIRT was successful in 10 of 11 patients (91%) with short, isolated bile duct strictures secondary to laparoscopic cholecystectomy and in seven of 15 patients (41%) with strictured bilioenteric anastomosis, but not in the patient with ischemic cholangiopathy. Twenty patients (74%) were stent-free at follow-up. Anastomotic biliary strictures were more likely to fail PIRT than isolated strictures secondary to laparoscopic cholecystectomy injury (p=0.02). Conclusion. Percutaneous balloon dilatation and stenting can be an effective strategy for patients with bile duct strictures, especially short bile duct strictures after laparoscopic cholecystectomy. Anastomotic strictures are associated with less good results when managed by PIRT but are successful in up to 40% of patients.

Introduction

Percutaneous interventional radiologic techniques (PIRT) have had an increasingly important role in the multidisciplinary diagnosis and management of biliary strictures. Before interventional techniques (radiologic, endoscopic) were developed, operative intervention was the only option for definitive treatment of biliary strictures. While excellent results have been reported with a primary operative approach, some benign biliary strictures may be amenable to a non-operative, interventional approach, thereby avoiding the morbidity and mortality of an operative approach. In recent years, percutaneous transhepatic biliary interventional techniques have been used increasingly to evaluate and treat selected patients with complex biliary lesions; however, little data in the literature address the long-term efficacy of this approach. Indeed, there is no consensus regarding which patients are best treated by this form of interventional treatment.

Methods

We reviewed retrospectively all patients with biliary strictures secondary to traumatic operative bile duct injury or who had an operative bilioenteric anastomosis performed that strictured subsequently. We then selected all those patients managed by PIRT at Mayo Clinic Rochester between 1 January 1998 and 31 December 2003.

Patients with a history of primary sclerosing cholangitis, hepatic transplantation, or malignant strictures were excluded. Demographic, clinical, histopathologic, treatment, survival, and outcome variables were collected and analyzed. Clinical notes, imaging studies, operative reports, and pathology reports were reviewed to determine the etiology and type of stricture whenever possible. Clinical symptoms, pre and post-intervention cholangiograms, results of liver function tests, and need for continued intervention were used to assess outcome. Data were obtained from the hospital record or by direct patient contact via a mailed, written questionnaire. The results were defined arbitrarily as “excellent” when the patient was asymptomatic, had normal liver function tests, and required no further treatment; “good” if mild symptoms were present and liver function tests were mildly abnormal (serum activities of γ-glutamyl-transferase, aminotransferases, or alkaline phosphatase equal to or less than 1.5 times the upper limit of the reference range) but no further intervention was needed; “fair” when the stricture was ultimately managed non-operatively, but the patient still had ongoing symptoms (e.g. intermittent cholangitis) or abnormal liver function tests (serum activities of γ-glutamyl-transferase, aminotransferases, or alkaline phosphatase greater than 1.5 times the upper limit of the reference range); or “poor” if a stricture recurred, was symptomatic, required active interventional treatment, or was managed ultimately by operative intervention.

We decided specifically not to describe the level of the strictures according to available classification systems (i.e. Bismuth, Strasberg) due to the heterogeneity of the lesions and the different mechanisms by which the strictures occurred (primary/iatrogenic vs. anastomotic).

This study was approved by the Mayo Clinic Institutional Review Board and complied with the HIPPA and Minnesota State regulations.

Clinical management

All patients underwent percutaneous transhepatic cholangiography (PTC) followed by placement of percutaneous biliary drainage catheters (PBD). This procedure was performed in a standard fashion and is described briefly below. All patients were treated with intravenous antibiotics before the procedure. Intravenous conscious sedation was administered by anesthesiology personnel consisting either of diprivan or a combination of midazolam and sublimazime. Using the right midline, transaxillary approach, a 22-gauge Chiba needle (Cook, Inc., Bloomington, IN) was inserted, and iodinated contrast material was injected to opacify the intrahepatic bile ducts and to define the level and nature of the bile duct lesion. Based on the percutaneous cholangiogram, an appropriate intrahepatic duct was selected and punctured, often a right, laterally oriented duct. When the injury was at or near the bifurcation, both right and left hepatic ducts were accessed. Attempts were made to cross the injury at the time of initial tube insertion. At our institution, the initial catheter and guide wire combination used most frequently was a five-Fr glide head catheter (Boston Scientific, Boston, MA) and an angled 0.035^? hydrophilic Glidewire (Boston Scientific). This combination was manipulated intraductally through a small percutaneous sheath. Other catheter and guide-wire combinations were used when this initial approach failed. After crossing the biliary lesion, the stricture was dilated with an appropriately sized (usually 6–8 mm diameter), high-pressure balloon. Finally, a locking-loop biliary drainage catheter (either 8.5 or 10 Fr size) (Cook Inc.) was inserted with its tip coiled beyond the site of obstruction. After the procedure, the catheter was placed initially to gravity drainage. The patient was discharged to home when clinically stable often with the catheter capped and internalized.

At two- to three-month intervals, patients returned to our institution on an outpatient basis for routine biliary catheter exchanges, balloon dilatation of the underlying stricture, and re-evaluation of hepatic function and overall clinical status. The overall duration of biliary intubation was determined by the location of the injury, the patient's clinical course, and follow-up cholangiograms. Prior to removal of the stenting catheters, most patients underwent a clinical trial with the tube capped for two weeks prior to removal.

For the purposes of analysis, the results were grouped into two categories: Group A: patients with a good response to PIRT (excellent and good outcomes) and Group B: patients who failed PIRT (fair and poor outcomes). Data were analyzed by descriptive statistics, and comparisons between groups were performed using Fisher's exact test and Wilcoxon rank-sum test with SPSS 13.0 statistical package (Chicago, IL). Differences were considered clinically important at p≤0.05. Data are presented as x±SD unless specified otherwise.

Results

Patients

A total of 27 patients with biliary strictures met our criteria for inclusion during the time period. There were 15 men (56%) and 12 women (44%) with an age of 54±19 years (range 11–86). The most frequent etiology of biliary stricture (Table I) was ductal injury during laparoscopic cholecystectomy in 11 patients (41%).

Table I. Etiology of biliary strictures.

Etiology	n	%
Laparoscopic bile duct injury	11	40
Pancreatoduodenectomy	7	26
Resection of extrahepatic biliary tree	5	19
Hepatectomy with resection of extrahepatic biliary tree	3	11
Presumed ischemic cholangiopathy	1	4

Eight patients (30%) had undergone biliary resection for cholangiocarcinoma, and seven patients (26%) had a biliary anastomosis during pancreatoduodenectomy as the cause of the stricture, but none of postoperative biliary strictures were secondary to neoplastic recurrence at the stricture site. Three of these patients had undergone radiation therapy prior to developing the stricture.

Treatment

The majority (24 patients) were managed by a combination of balloon dilatation and stenting (Table II). Mean number of balloon dilatations was 2.7 (range 1–7) with just over half of the patients requiring one or two dilatations (Table II). One patient developed transient pancreatitis after initial percutaneous transhepatic cholangiography; there were no treatment-related deaths.

Table II. Interventional radiology management.

Interventions	n
Balloon dilatation and stenting	24
Balloon dilatation only	2
Balloon dilatation + expandable metallic stent	1
Number of dilatations
1	8
2	7
3	6
5	3
6	2
7	1
Mean	2.7

Follow-up

Adequate follow-up was available for 26 of the 27 patients. The mean duration of follow-up was 42±25 months (range 2–81). Group A patients (treatment success group) were followed a mean of 46 months (range 5–81). The only patient lost to follow-up had a capped external drainage catheter, and although the patient was asymptomatic and had normal liver function parameters at last follow-up, we considered him a treatment failure. Eight patients died during follow-up from several different causes not related to the biliary stricture management, including three in Group A and five in Group B.

Outcomes

Overall, treatment was considered a success in 63% of patients. Interventional radiologic management was successful in 10 of 11 patients (91%) with a bile duct stricture related to a laparoscopic cholecystectomy and in seven of 15 patients (41%) with anastomotic strictures.

When comparing Group A to Group B, we found no difference in the mean number of balloon dilatations (2.7 vs. 2.7; p=NS) or in the duration of endobiliary stenting (eight vs. 10 months; p=NS), respectively. We did find an association between a diagnosis of cancer and outcome; patients with cancer were more likely to have a treatment failure than patients without a history of cancer (p=0.02). The duration of endobiliary stenting was less than four months in 13 patients (48%), four to nine months in four (15%), and greater than nine months in nine (33%). The duration of time of endobiliary stenting did not correlate with outcomes.

Of the 10 treatment failures (37%), one patient underwent successful operative reconstruction, and one was lost to follow-up. The remaining eight patients were not considered good candidates for operative management. Three were managed by active, radiologic intervention, three had stents placed (one endoscopic, two percutaneous), and one each was managed by continuous external–internal biliary drainage or intermittent courses of antibiotics for recurrent cholangitis. Twenty patients (74%) were stent-free at last follow-up, including the 16 in Group A and four in Group B. Variables potentially related to treatment outcomes are shown in Table III.

Table III. Interventional radiology management.

Etiology	Success/Total	%	p
Laparoscopic bile duct injury*	10/11	91
Pancreatoduodenectomy	2/7	29
Resection of extrahepatic biliary tree	3/5	60
Hepatectomy	2/2
Ischemic cholangiopathy	0/1
Intervention
Balloon dilatation and stenting	15/24	63
Balloon dilatation only	1/2
Balloon dilatation + expandable metallic stent	1/1
Duration of stenting†
< 4 months	10/13	77	0.63
Four to nine months	3/4	75
> 9 months	4/9	44
Cancer diagnosis
Yes	2/8	25	0.02
No	15/19	79

*Strictures were selected that were isolated and short in length <2 cm.

†One patient was lost to follow-up with a catheter stenting the stricture site and the duration of stenting is not known.

Discussion

The number of benign biliary strictures has increased in the recent years, in part as a result of the widespread use of laparoscopic cholecystectomy and the increase in the volume of hepatobiliary surgery. Iatrogenic bile duct injuries after upper gastrointestinal tract surgery (especially cholecystectomy) are the most common cause of postoperative biliary strictures and account for over 80% of the patients in some series ¹. Management of biliary strictures continues to be a challenging problem and requires a multidisciplinary approach.

Repair of bile duct injuries secondary to cholecystectomy increases the cost of an uncomplicated procedure markedly and carries substantial morbidity and mortality. It has been estimated that a bile duct injury adds over $50,000 to the cost of care ². These figures only take into account the cost of medical care and fail to include the loss in productivity and other additional expenses.

Operative approaches for repair of biliary injuries are a time-tested management option and achieve good long-term results with minimal morbidity and mortality in most patients ³,⁴,⁵,⁶. Our study of patients who underwent operative reconstruction of laparoscopic cholecystectomy-induced bile duct injuries at Mayo Clinic, Rochester found that the quality of life assessed a minimum of five years after reconstruction was as good as that of a control population after uneventful laparoscopic cholecystectomy ⁷.

Endoscopic management with dilatation, followed by insertion of plastic, biliary stents has also been used with good outcomes in other series ⁸,⁹,¹⁰,¹¹. As with PIRT, this technique requires biliary-enteric continuity to be present, but it also requires the stricture zone to be accessible endoscopically accessible via the gastrointestinal tract, a difficult procedure in those patients with Roux-en-Y anastomoses or more complex postoperative anatomy. Selected biliary strictures may be managed by PIRT to avoid the morbidity of an operation and allow the patient to return to an active life sooner. Other series have reported the results of PIRT for laparoscopy-associated bile duct injuries, but data on the long-term efficacy is scarce ¹²,¹³,¹⁴,¹⁵. Post-transplant biliary strictures also have been managed successfully by PIRT ¹⁶,¹⁷.

Patients selected for analysis in our study were those with biliary strictures secondary to operative traumatic bile duct injury or who had an operative bilioenteric anastomosis performed that strictured subsequently and that either were not amenable to endoscopic access or were not good candidates for operative or endoscopic management. Patients with a history of primary sclerosing cholangitis, hepatic transplantation, or malignant strictures were not included in this analysis, because they represent different diseases mechanisms. In this series, 27 patients with biliary strictures treated with PIRT were followed for a mean 42 of months and had an overall successful outcome in 63% of the patients with no treatment-related mortality. In the subgroup of patients that had a laparoscopic cholecystectomy as the cause of the stricture and were deemed candidates for PIRT based on the type and location of the bile duct stricture, PIRT was successful in 91% of the patients (10 of 11 patients). In those 12 patients with strictured bilioenteric anastomoses, the long-term success rate was less (5; 41%). The reasons for the difference in outcomes may be explained by the different mechanisms by which the bile duct strictures occur. Two series from the Johns Hopkins Hospital have reported success rates of 55 and 59%, respectively for percutaneous management of biliary strictures most of which were secondary to bile duct injury during cholecystectomy ¹²,¹³. These results are comparable to the overall success rate of our series but are inferior to our results for primary strictures (secondary to laparoscopic cholecystectomy). These difference may be explained by several factors, such as different selection criteria of patients and different durations of follow-up between the studies, because greater durations of follow-up may find more stricture recurrences.

We found an association between a history of cancer and outcomes; patients with strictures in association with operative procedures for cancer were more likely to fail PIRT than patients without a history of cancer (p=0.02). These results suggest that malignancy may have had a role in the genesis of the anastomotic stricture. Nevertheless, in none of the patients in our series was recurrent neoplasm ever demonstrated as the cause of the stricture. An alternative explanation for this difference may be the use of adjuvant abdominal radiation therapy in some of those patients.

Duration of endobiliary stenting did not correlate with outcomes. This finding contrasts with results reported from other series in which PIRT treatment was more likely to fail in patients stented for shorter durations of time ¹³.

One patient in our series was managed successfully with balloon dilatation and the implantation of a self-expandable metallic stent. Controversy exists regarding the use of “uncovered” metallic stents in patients with benign biliary conditions, because these stents are difficult, if not impossible to remove. Rossi et al. ¹⁸ first reported the use of uncovered metal stents in patients with postoperative biliary strictures after failed attempts at operative repair and endoscopic balloon dilatation; of the 17 patients managed in this fashion, 14 had relief from jaundice and cholangitis at a mean follow-up of eight months.

A recent series addressing removal of metallic stents placed to relieve biliary strictures found that removal of uncovered stents was difficult, while removal of covered stents with a snare was a relatively simple and safe procedure and therefore recommended that a covered stent be considered as initial therapy when malignant biliary obstruction was suspected but unconfirmed ¹⁹. Development of self-expandable, covered stents specifically designed to be removed at a later date may change current management strategies by promoting their use in benign strictures.

In summary, our data show that PIRT can be used with good or excellent outcomes in selected patients as a primary intervention or as an adjunct to operative management. PIRT is less invasive than operative intervention and may be safer in poor operative candidates and in those patients whose anatomy makes endoscopic instrumentation technically difficult; indeed, PIRT may represent their only non-operative management option.

We acknowledge the limitations of this retrospective chart review in which the patient groups reported are heterogeneous and small, potentially leading to a type II error in some cases. The results of this study are also affected by the active management decisions made by the involved clinical physicians during the years of this review, whereby young healthy patients with bile duct injuries ideal for operative repair were often not considered for a trial of PIRT, but instead underwent immediate operative reconstruction. These considerations, together with the fact that several strictures were reported more than 10 years after a biliary injury, bilioenteric anastomosis or attempt at operative repair underscores the need for prospective studies with long-term follow-up before considering any procedure a good option for management.