Endoscopy

Endoscopy ´ Reprint

Co-Editors G. Costamagna, Italy J. Devire, Belgium P. Fockens, The Netherlands H. Neuhaus, Germany T. Ponchon, France N. Vakil, USA K. Yasuda, Japan Section Editors J. Baillie, USA (Clinical Case Conference) J. Bergman, The Netherlands (Expert Approach and Innovation Forum Sections) J-F. Rey, France, T. Sauerbruch, Germany (Guidelines)

Assistant Editor H. Pohl, Germany

Managing Editor H. Hamilton-Gibbs, Germany

Chief Copy Editor T. Brady, UK

Editorial Assistants T. Michelberg, Germany F. Heidenreich, Germany Statistical Advisors Principal Advisor: K. Ulm, Germany Advisors: S. Wagenpfeil, Germany R. Hollweck, Germany

M. Classen, Germany M. Cremer, Belgium J. Geenen, USA G. A. Lehman, USA N. Soehendra, Germany H. Suzuki, Japan G. Tytgat, The Netherlands C. Williams, UK

Former Editors L. Demling, Germany M. Classen, Germany

´

T. Rösch, Germany

Advisory Board

ESGE European Society of Gastrointestinal Endoscopy (ESGE) J. Devire, Belgium (President) G. Costamagna, Italy (President Elect) J-F. Rey, France (Past President) S. D. Ladas, Greece (Vice President) P. Fockens, The Netherlands (Secretary General) H. Neuhaus, Germany (Treasurer) L. Aabakken, Norway (Chairman, Education Committee) C. Gheorghe, Romania J. Morris, UK I. Mostafa, Egypt M. Mun Äoz, Spain I. Rµcz, Hungary S. Rejchrt, Czech Republic T. Rösch, Germany (Endoscopy Journal) Address European Society of Gastrointestinal Endoscopy (ESGE) HG Editorial & Management Services Mauerkircher Str. 29 81679 Munich Germany Tel. + 49-89-20 14 856 Fax + 49-89-20 20 64 59 Email: [email protected]

Publishers Georg Thieme Verlag KG Ruedigerstraûe 14 D-70469 Stuttgart P.O. Box 30 11 20 70451 Stuttgart Germany Thieme Medical Publishers, Inc. 333 Seventh Avenue USA, New York, NY 10001 For subscription information please contact: [email protected] Reprint Volume 43 ´ 2011 Georg Thieme Verlag KG Reprint with the permission of the publishers only

´

Editor-In-Chief

Endoscopy

This is a copy of the author's personal reprint

Official Organ of the European Society of Gastrointestinal Endoscopy (ESGE) and affiliated societies

208

Original article

Long-term follow-up after biliary stent placement for postcholecystectomy bile duct strictures: a multicenter study

N. Tuvignon1, C. Liguory2, T. Ponchon3, B. Meduri4, J. Fritsch5, J. Sahel6, J. Boyer7, J. L.Legoux8, J. Escourrou9, C. Boustiere10, J. P. Arpurt11, M. Barthet12, P. Tuvignon13, G. Bommelaer14, B. Ducot15, F. Prat1, 16

Institutions

Institutions are listed at the end of article.

submitted 14 April 2009 accepted after revision 6 October 2010

Background and study aims: Endoscopic stenting is a recognized treatment of postcholecystectomy biliary strictures. Large multicenter reports of its long-term efficacy are lacking. Our aim was to analyze the long-term outcomes after stenting in this patient population, based on a large experience from several centers in France. Methods: Members of the French Society of Digestive Endoscopy were asked to identify patients treated for a common bile duct postcholecystectomy stricture. Patients with successful stenting and follow-up after removal of stent(s) were subsequently included and analyzed. Main outcome measures were long-term success of endoscopic stenting and related predictors for recurrence (after one stenting period) or failure (at the end of follow-up). Results: A total of 96 patients were eligible for inclusion. The mean number of stents inserted at

the same time was 1.9 ± 0.89 (range1 – 4). Stentrelated morbidity was 22.9 % (n = 22). The median duration of stenting was 12 months (range 2 – 96 months). After a mean follow-up of 6.4 ± 3.8 years (range 0 – 20.3 years) the overall success rate was 66.7 % (n = 64) after one period of stenting and 82.3 % (n = 79) after additional treatments. The mean time to recurrence was 19.7 ± 36.6 months. The most significant independent predictor of both recurrence and failure was a pathological cholangiography at the time of stent removal. Conclusion: Endoscopic stenting helps to avoid surgery in more than 80 % of patients bearing postcholecystectomy common bile duct strictures. However, a persistent anomaly on cholangiography at the time of stent removal is a strong predictor of recurrence and may lead to consideration of surgery.

Introduction

grade cholangiopancreatography (ERCP) in the management of bile duct strictures has been established over the past decade by several expert centers reporting success rates of 63 % – 100 %, with mean follow-up rates of 6 months to 10 years [14 – 20]. Endoscopic treatment typically consists of dilation and plastic stent insertion for 1 year, with elective stent exchanges every 3 months to avoid cholangitis caused by stent clogging [14, 19, 21, 22]. Moreover, prior endoscopic treatment does not seem to preclude surgery [17]. Some reliable data on the long-term outcome of endoscopic management of these patients are available, but most are from limited series and/or centers with a specific expertise [14, 19, 21, 22]. The purpose of this study was therefore to analyze the long-term outcome of stent therapy in postcholecystectomy biliary strictures in a large group of patients from various French settings. In order to strengthen the significance of our data, updated information was prospectively collected

Bibliography DOI http://dx.doi.org/ 10.1055/s-0030-1256106 Endoscopy 2011; 43: 208–216 © Georg Thieme Verlag KG Stuttgart · New York ISSN 0013-726X Corresponding author F. Prat, MD, PhD Department of Gastroenterology Endoscopy unit Université Paris-Descartes Pavillon Achard Hôpital Cochin 27 rue du Faubourg St Jacques 75014 Paris France Fax: +33-1-58411965 [email protected]

·

This is a copy of the authorʼs personal reprint

·

Authors

!

Nearly 80 % of all cholecystectomies are currently completed laparoscopically, in most developed countries [1 – 3], with up to 80 000 cholecystectomies being performed each year in France [4]. The incidence of biliary strictures following laparoscopic cholecystectomy ranges from 0.2 % to 0.5 % [5 – 9], as a consequence of either direct bile duct injury with subsequent fibrosis, or ischemic phenomena. Whereas early symptoms are often associated with biliary leaks, late symptoms include cholestasis, recurrent cholangitis, ductal stones, or secondary biliary cirrhosis. Strictures develop either early on bile ducts primarily repaired on a T-tube, or more commonly later on unrecognized injuries. Management was traditionally surgical, the favorite option being Rouxen-Y choledochojejunostomy, with a successful outcome in 76 % – 90 % of cases, but significant associated morbidity (18 % – 51 %) and mortality (4 % – 13 %) [10 – 13]. The role of endoscopic retro-

Tuvignon N et al. Long-term follow-up after biliary stent placement … Endoscopy 2011; 43: 208 – 216

Original article

209

Fig. 1 Study population. Flow diagram of outcomes of endoscopic management in 204 patients with postoperative bile duct injury. BDI, biliary duct injury; ERCP, endoscopic retrograde cholangiopancreatography; FU, follow-up.

Postsurgical BDI explored by ERCP (n = 204)

Excluded patients (n = 80)

Postcholecystectomy bile duct stenosis (n = 124)

Unsuccessful stenting (n = 28)

Successful endoscopic stenting (n = 96)

Primary Success (after 1 endoscopic treatment) (n = 64)

Recurrence (n = 32)

Failure: Metallic stent, Continous stenting, Symptoms (n = 7)

in a subset of patients who were followed up for more than 3 years after stent removal.

Patients and methods !

A retrospective multicenter database on endoscopic stenting for postcholecystectomy stricture including 14 French surgical endoscopic centers was conducted under the auspices of the Société Française d’Endoscopie Digestive (SFED). The study was based on voluntary participation. SFED members claiming to practice ERCP (approximately 120 gastroenterologists from academic, public, and private practices) were asked by mail whether they had followed patients with postsurgical bile duct injuries and were willing to include them in a retrospective study. Those who answered positively were asked about the number of such patients they had treated by ERCP, whether or not they used stents and for how long, and whether or not they were able to identify patients and collect relevant retrospective follow-up data. Endoscopists providing negative responses or no response to these endpoints or with a personal experience of less than three cases were not considered for participation in the study. A total of 20 gastroenterologists from 14 centers (eight academic, four public general hospitals, and two private clinics) fulfilled the criteria for participation. All participants were required to identify all patients treated consecutively in order to avoid drop-outs of patients with failed ERCP or stenting, who were ex" Fig. 1). cluded subsequently (see flow-chart, ● The inclusion period ran from 1984 through to September 2003 (date of last stent removal) in order to achieve a follow-up period

Surgery (n = 1)

of at least 3 years after stent removal. Updated follow-up data collection was started in 2006 and was completed by January 2007. Inclusion criteria were: postcholecystectomy (open or laparoscopic) benign primary biliary duct stricture; successful endoscopic stenting; use of one or more plastic stents; all procedures for an individual patient performed in the same endoscopy center by the same team; follow-up after stent removal. Exclusion criteria were as follows: diagnosis of malignancy; postsphincterectomy stricture; biliary leakage without stenosis; Mirizzi syndrome; stenoses on biliodigestive anastomoses; extrinsic compression secondary to chronic pancreatitis; primary sclerosing cholangitis; stenosis on bile duct stone; post-liver transplantation and post-hepatectomy strictures; stricture treated by dilation without stent placement.

Treatment process Before inclusion in the study, it was verified by questionnaire that each center had followed the guidelines described below and had forwarded all relevant data. The initial endoscopy procedure consisted of the insertion of polyethylene stents bridging the stenosis. There was no standardized protocol for all centers regarding the diameter or the number of stents but each center followed the same guidelines. An endoscopic sphincterotomy was optional, but was performed in most cases to facilitate stent placement during the first procedure. If the stenosis was too tight to allow stenting, endoscopic bougienage or hydrostatic balloon dilation was performed when necessary. Only polyethylene stents were introduced. The diameter and number of stents introduced during each ERCP were tailored to the size of the bile duct, but there was no systematic in-

·

Success after 2 Endoscopic treatments (n = 15)

·

Endoscopic retreatment (n = 23)


Failures, n = 17

Final success (n = 79)

Surgery (n = 9)


210

Original article

tent to maximize the number of stents inserted. The rule followed by all endoscopists was that stents were electively exchanged every 3 months for a 1-year period or more if necessary, based on ERCP features. However, in some cases stents were removed earlier and not exchanged if the stenosis was considered to be adequately dilated based on the following criteria: (1) dilation of the stenosis on cholangiography was complete; (2) drainage of contrast from the biliary tree was quick and complete; and (3) an inflated extraction balloon could be passed through the stenosis without significant resistance. Additional endoscopic treatments (after a first period of stenting) could be based on hydrostatic balloon dilations or a second stenting period. Administration of prophylactic intravenous antibiotics was not systematic.

Outcome measures

·


·

Outcome parameters included: procedure-related complications; number of stents introduced during the first procedure; maximal number of stents introduced; stenting period duration; cholangiographic features on final ERCP; incidence and delay of relapse after one stenting period; incidence of failure; number of patients referred for surgical biliary repair, with particular attention to surgical difficulties and outcome. The term “adequate calibration” was used to define a bile duct with complete or nearly complete recovery in diameter and shape after stent removal. “Success” was defined as an uneventful follow-up after stent removal with no or minimal change on final ERCP, no clinical complaint and, when available, no or minimal biological anomaly on follow-up blood tests (i. e. gamma-glutamyl transferase 55 years, n (%)

57 (59.4)

20 (62.5)

14 (82.4)

Initial surgery, n (%) Laparoscopy Open surgery Laparoscopy with conversion to laparotomy Surgical repair or choledocotomy

50 (52.1) 29 (30.2) 17 (17.7) 47 (49.0)

13 (40.6) 11 (34.4) 8 (25.0) 17 (53.1)

8 (47.1) 3 (17.6) 6 (35.3) 9 (52.9)

Timespan between surgery and symptoms Median (range), days Mean (SD), days

13.1 (0 – 1924) 121.0 (327.5)

20.0 (0 – 988) 119.2 (256.9)

7 (41.2)

25.0 (0 – 972) 122.5 (259.3)

Presenting signs and symptoms, n (%) Cholestasis Jaundice Cholangitis External bile leak

16 (16.7) 22 (22.9) 50 (52.1) 8 (8.3)

5 (15.6) 7 (21.9) 17 (53.1) 3 (9.4)

4 (23.5) 4 (23.5) 9 (52.9) 0

ERCP diagnosis, n (%) Bismuth stricture type I II III IV Associated fistula

6 (6.3) 66 (68.8) 20 (20.8) 4 (4.2) 15 (15.6)

25 (78.1)

15 (88.2)

7 (21.9)

2 (11.8)

5 (15.6)

2 (11.8)

·

Recurrence* (n = 32)

*After the first stenting period. **At the end of follow-up.

" Table 3 and 4) Follow-up and outcomes (●

At the end of the study, the median follow-up after stent removal was 6.1 years (range 0.08 – 20.3 years). The outcome was deemed successful in 66.7 % (n = 64) of the patients after one stenting period (primary success). A total of 28 of the 32 stricture relapses occurred within the first 3 years, as well as nine of 10 surgical re" Fig. 2). The 32 patients with recurrent strictures repairs (● lapsed within a median of 5.5 months (mean 19.7 ± 36.6, range 0 – 180). Relapse was revealed by cholangitis in 75 % of cases, as were all seven cases of “late” relapse (i. e. more than 18 months after stent removal). After relapse, nine patients were referred for surgery and 23 for a second endoscopic treatment. The second endoscopic treatment consisted of simple dilation for nine patients and a new stenting period for 14 patients. The median number of stents inserted was two (range 1 – 3), and the median duration of the second stenting period was 10 months (range 0 – 96 months). By the end of the study, eight failures of endoscopic " Fig. 1): five patients were still on plastic therapy were noted (● stenting, one had a metallic stent in place, one was referred for surgery, and one finally became symptomatic at the end of the

study 24 months after stenosis dilation. Of the nine patients retreated with only balloon dilation, generally performed 6 – 12 months after stent removal, only two developed a second recurrence which was treated endoscopically 6 and 17 years, respectively, after balloon dilation. The final success rate including patients with additional treatment was therefore 82.3 % (n = 79). Updated data for patients with a follow-up period longer than 3 years were obtained from 67.7 % of the patients (n = 65), including eight of 10 patients with surgical repair and 57 of the 86 patients undergoing endoscopy. The other 31 patients (follow-up shorter than 3 years or lost at follow-up) were all symptom-free at the end of follow-up (median follow-up: 2.33 years, range 0.08 – 9.5 years). After a median follow-up of 7.0 years (range 3.2 – 17 years), 48 of the 57 nonoperated patients with a followup longer than 3 years (84.2 %) were symptom-free. Biological data were obtained for 43 patients (75.4 %) and radiological data for 35 patients (61.4 %). Normal results were obtained for liver function tests in 38 patients (88.3 %) and morphological data in 34 patients (97.1 %). Among surgical repair patients, one died during follow-up, one was symptomatic because of anastomotic stenosis, and six were symptom-free with normal biology and imaging. Surgical repair reports from nine out of 10 patients were analyzed: no significant surgical difficulty, in particular with regard to the dissection of the hepatic pedicle or construction of the bilio-enteric anastomosis was reported. There was no report of significant postoperative morbidity in these patients. Fourteen patients died during follow-up, 53.6 ± 36.2 months after stent removal; none of the deaths was due to biliary disease.

·

stricture calibration was adequate (bile duct with complete or nearly complete recovery in diameter and shape after stent removal) in 79 patients (82.3 %). Stent-related complications occurred in 22 of 348 procedures (6.3 %) and involved 21 patients (21.9 %). Ten cases of acute pancreatitis (10.4 %), two of them severe, and two duodenal perforations (2.1 %) were observed. In one case the duodenal perforation was revealed by cervical emphysema during endoscope reintroduction after stent removal. A new stent was introduced without dilation. The second perforation was related to a needle-knife precut-sphincterotomy (no dilation done). Both were treated conservatively with digestive aspiration and antibiotics and were discharged within a few days. Seven cases of symptomatic stent clogging (7.3 %) with four cholangitis, two stent migrations (2.1 %), and one nonsymptomatic biliary fissure (1.0 %) after balloon dilation were noted.


Patients (n = 96)

Multivariate analysis for recurrence and failure Factors independently associated with recurrence and failure, as " Table 5 and 6. Indepenpreviously defined, are summarized in ● dent predictors for recurrence identified using the multivariate Cox proportional hazard model were a pathological final cholangiography (hazard ratio [HR] 16.3; 95 %CI 4.37 – 60.7; P < 0.001)


212

Original article

Table 2 Endoscopic stenting. Duration of treatment and short-term complications. Patients

Recurrence*

Failure**

(n = 96)

(n = 32)

(n = 17)

73 (76.0) 22 (22.9) 1 (1.0)

24 (75.0) 8 (25.0)

14 (82.4) 3 (17.6)

38 (39.6) 32 (33.3) 22 (22.9) 4 (4.2)

13 (40.6) 12 (37.5) 7 (21.9)

9 (52.9) 3 (17.6) 5 (29.4)

Complementary percutaneous drainage, n (%)

8 (8.3)

2 (6.3)

0

No. of stent changes, median (range)

4.0 (1 – 7)

4 (1 – 6)

4.0 (1 – 6)

No. of stents inserted during 1st procedure, n (%) 1 2 3

2.0 (1 – 4)

Maximum no. of stents inserted, median (range) Maximum no. of stents inserted, n (%) 1, n (%) 2, n (%) 3, n (%) 4, n (%)

12.0 (2 – 96) 11.9 (10.3) 52 (54.2)

Cumulated duration of stents in situ (1st and subsequent stentings) Median (range), months Mean (SD), months > 12 months, n (%)

12.0 (2 – 113) 14.62 (16.2) 57 (59.4)

Final cholangiography after 1st stenting period, n (%) Stent removal without cholangiography Normal Mild Stricture Tight Stricture

4 (4.2) 29 (30.2) 50 (52.1) 13 (13.5)

Procedure-related complication (348 ERCPs), n (%) Acute pancreatitis Cholangitis Migration Others

22 (6.3) 10 (45.4) 4 (18.2) 2 (9.1) 6 (27.3)

11.0 (3 – 96) 13.7 (16.1) 14 (43.8)

12 (3 – 24) 12.0 (6.25) 9 (52.9) 12 (3 – 113) 23.82 (28.39) 9 (52.9)

3 (9.4) 13 (40.6) 16 (50.0)

1 (5.9) 6 (35.3) 10 (58.8)

*After the first stenting period. **At the end of follow-up.

Table 3 Long-term outcomes of endoscopic stenting, overall. Patients (n = 96) Primary success, n (%)

64 (66.7)

Recurrence (after 1st stenting period), n (%)

32 (33.3)

Median time to stricture recurrence (range), months

5.5 (0 – 180)

No. of patients referred for 2nd endoscopic treatment, n (%)

23 (24)

Success after retreatment , n (%)

15 (15.6)

Failure after retreatment, n (%)

8 (8.3)

No. of patients referred for surgery, n (%)

10 (10.4)

Overall failure rate (at the end of follow-up), n (%)

17 (17.7)

Final success, n (%)

79 (82.3)

Median follow-up after stent removal (range), years

6.1 (0.08 – 20.3)

No. of patients with updated data, n (%)

23 (24.0)

Deaths during follow-up / biliary or treatmentrelated, n (%)

14 (14.6) / 0

·


·

Duration of stents in situ during the first stenting period Median (range), months Mean (SD), months 1st stenting > 12 months, n (%)

smaller the number of stents inserted, the more pathological the final cholangiography. After stratification of the number of stents on stenting duration this inverse relationship was even stronger. This explains why these factors, although nonsignificant in the univariate analysis, became significant in the multivariate model. A pathological final cholangiography was also the main predictive factor of failure (HR 68.8; 95 %CI 6.8 – 695.8; P = 0.003). Two other independent outcome predictors were identified: age older than 55 years was predictive of failure (HR 7.5; 95 %CI 1.7 – 34.3; P = 0.0089) whereas female sex was a protective factor (HR 0.2; 95 %CI 0.06 – 0.8; P = 0.018). Although nonsignificant initially, sex became predictive of failure also in the univariate analysis after stratification on final cholangiography (HR 0.3; 95 %CI 0.1 – 0.9; P = 0.032), which explains why this factor was independently predictive in the multivariate analysis.

Discussion

and a maximum number of two stents (HR 2.5; 95 %CI 1.1 – 5.8; " Table 5 that, paradoxically, P = 0.036). It can be noted from ● two stents were associated with more recurrences than one stent, but a substantial fraction of these patients had more severe strictures and required more aggressive management. Conversely, a stenting period of at least 1 year appeared to be a protective factor against recurrence (HR 0.4; 95 %CI 0.2 – 0.9; P = 0.022). A complementary analysis revealed that the maximum number of stents and final cholangiography were inversely correlated: the

!

Laparoscopic cholecystectomy has been the treatment of choice for symptomatic cholelithiasis patients for more than 15 years, yet it still induces biliary injuries in 0.4 % to 0.86 % [24 – 26]. Although surgical repair of bile duct injuries has long been the mainstay of treatment, the majority of these injuries and especially bile duct strictures, with or without leakage, can be treated successfully in 70 % – 95 % of the patients by means of endoscopic or percutaneous stenting [14, 19 – 22, 27 – 29]. Nowadays, most patients are initially treated by endoscopy because of comparable efficacy and lower rates of morbidity and mortality [30, 31]. A


Original article

213

Table 4 Follow-up and long-term outcomes of endoscopic stenting in patients with updated data. All patients (n = 96)

Endoscopic stenting only (n = 86)

6.1 (0.08 – 20.3)

Global follow-up, median (range), years Patients with incomplete data: Follow-up < 3 years (n = 17) or nonupdated data (n = 14), n (%)

31 (32.3) 2.33 (0.08 – 9.5)

Follow-up, median (range), years Patients with updated data and follow-up > 3 years, n (%)

65 (67.7) 7.3 (3.2 – 20.3)

Follow-up, median in (range), years Clinical, n (%)

65 (100)

Asymptomatic

29 (33.7) 2.3 (0.08 – 8.0) 57 (66.3) 7.0 (3.2 – 17) –

scopic failures (n = 10) 9.2 (1.3 – 20.3) 2 (20.0) 5.41 (1.33 – 9.5) 8 (80.0) 9.41 (6.75 – 20.3) –

54 (83.1)

48 (84.2)

Symptomatic

3 (4.6)

2 (3.5)

1 (12.5)

Death (none of which related to procedure or biliary disease)

8 (12.3)

7 (12.3)

1 (12.5)

50 (76.9)

43 (75.4)

7 (87.5)

43 (86.0)

38 (88.3)

5 (71.4)

7 (14)

5 (11.6)

2 (28.6)

40 (61.5)

35 (61.4)

5 (62.5)

39 (97.5)

34 (97.1)

1 (2.5)

1 (2.9)

Biological, n (%) Normal hepatic biology Pathological hepatic biology Radiological, n (%) Normal biliary duct Pathological biliary duct

6 (75.0)

5 (100) 0

bile duct stricture with long-term follow-up to date. Only one other study included more patients for this indication but it was not a multicenter study [22]. As a nonprospective study with no pre-defined treatment protocol, the results might have been biased by the selection of participating centers. However, the majority of French endoscopy centers with an experience in the management of biliary injury agreed to be involved, and no center willing to participate was rejected for applying different management strategies to those defined in the methods section. This is therefore probably a fairly reliable report of the actual outcomes of endoscopic therapy of postcholecystectomy biliary strictures with plastic stents. The success rate after one stenting period and the final success of endoscopic treatment in our patients are similar to previously " Table 7. However, this study is published results, as shown in ● different in that it was a multicenter one, enrolling not only tertiary referral, academic centers, but also local hospitals from all parts of the country. It is noteworthy that 39 % of the endoscopic re-treatments (9/23) consisted of only dilation performed during a systematic control ERCP 6 – 12 months after stent(s) removal and that only one patient had more than two additional endoscopic treatments (two dilations during systematic ERCP). Costamagna et al. reported a success rate of 89 % in 45 patients after a follow-up of 4 years [19]. Bergman et al. followed 44 patients over 9 years and found a 20 % rate of stricture recurrence [14]. Interestingly, they observed that all recurrences occurred within 2 years of stent removal. De Reuver et al. evaluated the recurrence rate at 26.4 % (n = 29) [22]. We observed a 33.3 % stricture recurrence rate after one endoscopic treatment. In accordance with Bergmann et al., 50 % of the recurrences occurred during the first 6 months after stent removal but some cases were also noted after 2 years (6/32, 18.8 %). All relapses observed after more than 18 months of follow-up were symptomatic and usually revealed by cholangitis, so that “late” nonsymptomatic recurrences that could lead to secondary biliary cirrhosis are very unlikely. It is therefore reasonable and advisable to discontinue the follow-up after 2 years of stent removal, leaving only patients with biliary symptoms to be explored [22]. Before this delay, clinical and biological follow-up should be recommended on a 6-monthly basis, with abdominal ultrasonography and a set of the usual liver blood tests being sufficient. Even the use of an imaging study is

50

25

0 0 n = 96

24

48

72

61

50

20

96

120

8 8 Time (months) Patients at risk (n)

144

168

5

2

192

Fig. 2 Kaplan-Meier relapse-free survival. Time (0) corresponds to stent removal at the end of the first endoscopic treatment. The continuous line shows the proportion of patients with no evidence of relapse among 96 patients with postcholecystectomy biliary stenosis after one stenting period. The dotted line shows the proportion of patients at risk of relapse after one endoscopic treatment.

second reason is the rate of symptomatic relapse (5 % – 20 %) after surgical repair because of recurrent stricture formation at the anastomotic site. Moreover, surgery remains available when endoscopy fails [17, 32 – 34], but the results of surgical repair after endoscopic treatment remain poorly described. Despite the limitations due to the retrospective design of this study, most patients were followed up for more than 3 years after stent removal and updated data were obtained for a large majority of the patients included. Although this design precluded the study of short-term results of endoscopic treatment, it increased the power of the study and allowed for a reliable analysis of the long-term outcomes of stenting in real practice. This is the largest multicenter cohort of patients treated for post-cholecystectomy

·

75


·

100

Survival (%)

5.9 (0.08 – 17)

Surgery for endo-


214

Original article

Table 5 Univariate and multivariate analysis of predictive factors of recurrence after first stenting period. Total n = 96

Univariate analysis*

Multivariate analysis P value

Hazard ratio [95 %CI] 39 57

1 1.2 [0.6 – 2.5]

0.60

Sex Male Female

39 57

1.0 [0.5 – 2.0]

0.99

Initial surgery Laparoscopy Initial open surgery Laparoscopy with conversion

50 29 17

1 1.5 [0.7 – 3.4] 2.5 [1.04 – 6.1]

0.33 0.04

Location of strictures Bismuth I – II Bismuth III – IV

72 24

1 0.8 [0.4 – 2.0]

Final ERCP Normal Mild defect Tight stricture or N/A

29 50 17

1 2.7 [0.8 – 9.5] 13.7 [3.9 – 47.6]

0.12 < 0.001

1 2.5 [0.7 – 9.0] 16.3 [4.37 – 60.7]

Maximum number of stents inserted (1st treatment) 1 stent 2 stents ≥ 3 stents

38 32 26

1 1.3 [0.6 – 2.9] 0.7 [0.3 – 1.8]

0.51 0.45

1 2.5 [1.1 – 5.8] 1.1 [0.42 – 3.1]

Stenting duration (first treatment) < 12 months > 12 months

44 52

1 0.6 [0.3 – 1.2]

·

Age < 55 years > 55 years

0.72

0.12

P value

1 0.9 [0.4 – 2.2] 2.3 [0.9 – 5.8]

0.89 0.091

1 1.2 [0.5 – 3.1]

0.064

1 0.4 [0.2 – 0.9]

0.14 < 0.001

0.036 0.80 0.022

ERCP, endoscopic retrograde cholangiopancreatography; N/A, not available.

Table 6 Univariate and multivariate analysis of predictive factors of failure (at the end of follow-up). Total n = 96

Univariate analysis*

Multivariate analysis

Hazard ratio [95 %CI]

P value


P value

39 57

1 5.0 [1.1 – 21.9]

0.03

1 7.5 [1.7 – 34.3]

0.0089

Sex Male Female

39 57

1 0.5 [0.2 – 1.3]

0.16

1 0.2 [0.06 – 0.8]

0.018

Initial surgery Laparoscopy Initial open surgery Laparoscopy with conversion

50 29 17

1 0.4 [0.1 – 2.1] 2.8 [0.97 – 8.1]

Location of strictures Bismuth I – II Bismuth III – IV

72 24

1 2.3 [0.5 – 10.4]

Final ERCP Normal Mild defect Tight stricture or N/A

29 50 17

1 3.5 [0.4 – 28.8] 19.9 [2.5 – 157.5]

Maximum number of stents inserted (1st treatment) 1 stent 2 stents ≥3 stents

38 32 22

1 0.4 [0.1 – 61.4] 0.7 [0.2 – 2.2]

Stenting duration (first treatment) < 12 months > 12 months

40 56

1 0.7 [0.3 – 1.8]

Age < 55 years > 55 years

·



ERCP, endoscopic retrograde cholangiopancreatography; N/A, not available.


1 0.2 [0.05 – 1.3] 3.3 [0.9 – 12.0]

0.093 0.064

0.25 0.005

1 4.7 [0.5 – 40.9] 68.8 [6.8 – 695.8]

0.16 0.003

0.14 0.57

1 0.4 [0.08 – 1.7] 0.96 [0.3 – 3.5]

0.2 0.96

0.29 0.058 0.25

0.45

1 0.4 [0.1 – 1.1]

0.08

Original article

215

Table 7 Results of endoscopic treatment of cholecystectomy-related biliary strictures. Study

n

Stenting

Complica-

BDI-related

Length of follow-up,

Final

Stricture

duration,

tions, %

mortality,

years

success, %

recur-

Median: 9

80

20

months

%

rence, %

Bergman et al. 2001 [14]

44

40

de Reuver et al. 2007 [22]

110

11

33

1.8

Mean: 7.6 ± 2.9

74

29

Costamagna et al. 2001 [19]

45

12

10

2

Mean: 4.06

89

10

Davids et al. 1993 [17]

35

12

35

Mean 3.5

80

17

Vitale et al. 2008 [29]

48

12

8

0

Mean: 2.5 ± 2.0

91

9

Current study

96

12

23

0

Median: 6.1

82

23

BDI, bile duct injury.

·

(type B and C) for any stricture location. Surgery should be reserved for endoscopic treatment failures and could be discussed at the end of a well-conducted 1-year endoscopic stenting period only if the final ERCP does not demonstrate adequate calibration. It is also very interesting to observe that in none of the nine patients with surgical repair and for whom operative and postoperative reports were obtained, did the surgeon describe any specific difficulty nor were postoperative complications reported. Based on patient outcomes in this study, several points can be raised, while keeping in mind the limitations of a retrospective study. " Stenting should be maintained for 1 year and protracted if necessary until adequate calibration on ERCP. " At least two stents should be introduced, and more when possible. " Stents should be exchanged during the stenting period, with a time span between stent exchanges that may be extended to 4 months. " Patients with suboptimal ERCP at the end of the initial stenting period and subsequent symptomatic stricture relapse, particularly men over 55 years, should be referred to the surgeon. " Follow-up can probably be discontinued after 2 years of stent removal in nonsymptomatic patients with normal biology. Indeed, all these suggestions apply to conventional polyethylene stents and may have to be reconsidered in the near future if fully covered expansive stents appear to fulfill their promises in the treatment of benign biliary strictures [36].

Acknowledgment !

This study was conducted under the auspices of the Société Française d’Endoscopie Digestive (SFED) Competing interests: None Institutions 2 3 4 5 6 7 8 9 10 11 12 13

Gastroenterology Department, Cochin Hospital, Paris, France Gastroenterology Department, “Clinique de l’Alma”, Paris, France Gastroenterology Department, Edouard Herriot Hospital, Lyon, France Gastroenterology Department, “Clinique Bachaumont, Paris, France Gastroenterology Department, Kremlin-Bicêtre Hospital, Paris, France Gastroenterology Department, Conception Hospital, Marseilles, France Gastroenterology Department, St Jean Hospital, Angers, France Gastroenterology Department, Source Hospital, Orleans, France Gastroenterology Department, Rangueil Hospital, Toulouse, France Gastroenterology Department, St Joseph Hospital, Marseilles, France Gastroenterology Department, St Marthe Hospital, Avignon, France Gastroenterology Department, North Hospital, Marseilles, France Gastroenterology Department, Albi Hospital, Albi, France

·

1


questionable, as the only patient in our study with pathological imaging also had pathological biological results as well as clinical symptoms. In the present series, short-term procedure-related complications occurred in 6.3 % of ERCPs and 21.9 % of the patients. This is coherent with the average found in the literature [14, 19, 22, 29, 31, 35] even though a few low-volume centers were involved in our study. Our main objective when undertaking this work was to identify predictive factors for recurrence or failure. The likelihood of failure was lower in women than in men and higher in older patients, but much more significant was the relationship of failure with the final cholangiography (i. e. obtained immediately after the last stent removal). This strongly suggests that when the final cholangiography does not show an adequate calibration, especially if the patient is a man over 55 years, surgery may be the next treatment without waiting for symptomatic relapse. Contrary to previous studies, the duration of the first stenting or the cumulated duration of stentings did not appear to be significantly related to failure, but a short period of stenting was a predictive factor of recurrence [14, 19, 22, 27]. This result can be explained by the fact that 40 % of recurrences were treated by dilation rather than by stenting. Similarly, a small number of stents inserted during the first procedure did not appear to be predictive of failure [19, 22, 29]. This finding should not make one reconsider the idea of multiple stent insertion, because not all strictures bear the same resilience to endoscopic therapy and those with the poorest prognosis may require more stents. The optimal duration of stent implantation is still controversial. The expert opinion is that it is approximately 1 year but Costamagna et al. suggested the key point was the final cholangiography [19]. Our results clearly support this assumption. The final cholangiography and the achievement of adequate calibration can be seen as the result of the combined and often conflicting effects of the severity of the stricture and the endeavor to optimize stenting by, for instance, increasing the number of stents placed. We therefore suggest that patients should be treated until an adequate calibration, as previously defined, is obtained at control cholangiography, and for no less than 1 year. Stent exchanges are often scheduled every 3 months to prevent complications. With regard to the low rate of per-stenting complications (three cases), one could assume that increasing this time span to 4 months (which means four ERCPs over a 1-year period instead of five) should not be detrimental. Contrary to previous studies [14, 19, 22] we did not identify stricture location as a predictive factor for failure, possibly because we excluded patients with unsuccessful initial ERCP. This study confirms that endoscopic stenting is the primary treatment of choice for postcholecystectomy biliary strictures with or without fistula


216

Original article 14

15 16

Gastroenterology Department, “Hotel Dieu” Hospital, Clermont Ferrand, France Inserm U569, Kremlin-Bicêtre Hospital, Paris, France Université Paris-Descartes, Paris, France

References

·


·

1 Livingston EH, Rege RV. A nationwide study of conversion from laparoscopic to open cholecystectomy. Am J Surg 2004; 188: 205 – 211 2 McMahon AJ, Fischbacher CM, Frame SH, MacLeod MC. Impact of laparoscopic cholecystectomy: a population-based study. Lancet 2000; 356: 1632 – 1637 3 Nilsson E, Fored CM, Granath F, Blomqvist P. Cholecystectomy in Sweden 1987 – 99: a nationwide study of mortality and preoperative admissions. Scand J Gastroenterol 2005; 40: 1478 – 1485 4 Caroli-Bosc FX, Deveau C, Harris A et al. Prevalence of cholelithiasis: results of an epidemiologic investigation in Vidauban, southeast France. General Practitioner’s Group of Vidauban. Dig Dis Sci 1999; 44: 1322 – 1329 5 Tesdal IK, Roeren T, Weiss C et al. Metallic stents for treatment of benign biliary obstruction: a long-term study comparing different stents. J Vasc Interv Radiol 2005; 16: 1479 – 1487 6 Waage A, Nilsson M. Iatrogenic bile duct injury: a population-based study of 152 776 cholecystectomies in the Swedish Inpatient Registry. Arch Surg 2006; 141: 1207 – 1213 7 Aggarwal R, Darzi A. Training in the operating theatre: is it safe? Thorax 2006; 61: 278 – 279 8 Oleynikov D, Rentschler ME, Dumpert J et al. In vivo robotic laparoscopy. Surg Innov 2005; 12: 177 – 181 9 Sikora SS, Pottakkat B, Srikanth G et al. Postcholecystectomy benign biliary strictures – long-term results. Dig Surg 2006; 23: 304 – 312 10 Archer SB, Brown DW, Smith CD et al. Bile duct injury during laparoscopic cholecystectomy: results of a national survey. Ann Surg 2001; 234: 549 – 558; discussion 558 – 559 11 Way LW, Stewart L, Gantert W et al. Causes and prevention of laparoscopic bile duct injuries: analysis of 252 cases from a human factors and cognitive psychology perspective. Ann Surg 2003; 237: 460 – 469 12 Gadacz TR. Update on laparoscopic cholecystectomy, including a clinical pathway. Surg Clin North Am 2000; 80: 1127 – 1149 13 Strasberg SM, Hertl M, Soper NJ. An analysis of the problem of biliary injury during laparoscopic cholecystectomy. J Am Coll Surg 1995; 180: 101 – 125 14 Bergman JJ, Burgemeister L, Bruno MJ et al. Long-term follow-up after biliary stent placement for postoperative bile duct stenosis. Gastrointest Endosc 2001; 54: 154 – 161 15 Prat F, Pelletier G, Ponchon T et al. What role can endoscopy play in the management of biliary complications after laparoscopic cholecystectomy? Endoscopy 1997; 29: 341 – 348 16 Hartle RJ, McGarrity TJ, Conter RL. Treatment of a giant biloma and bile leak by ERCP stent placement. Am J Gastroenterol 1993; 88: 2117 – 2118 17 Davids PH, Tanka AK, Rauws EA et al. Benign biliary strictures. Surgery or endoscopy? Ann Surg 1993; 217: 237 – 243

18 Smith MT, Sherman S, Lehman GA. Endoscopic management of benign strictures of the biliary tree. Endoscopy 1995; 27: 253 – 266 19 Costamagna G, Pandolfi M, Mutignani M et al. Long-term results of endoscopic management of postoperative bile duct strictures with increasing numbers of stents. Gastrointest Endosc 2001; 54: 162 – 168 20 Draganov P, Hoffman B, Marsh W et al. Long-term outcome in patients with benign biliary strictures treated endoscopically with multiple stents. Gastrointest Endosc 2002; 55: 680 – 686 21 Dumonceau JM, Deviere J, Delhaye M et al. Plastic and metal stents for postoperative benign bile duct strictures: the best and the worst. Gastrointest Endosc 1998; 47: 8 – 17 22 de Reuver PR, Rauws EA, Vermeulen M et al. Endoscopic treatment of post-surgical bile duct injuries: long term outcome and predictors of success. Gut 2007; 56: 1599 – 1605 23 Bismuth H, Majno PE. Biliary strictures: classification based on the principles of surgical treatment. World J Surg 2001; 25: 1241 – 1244 24 Russell JC, Walsh SJ, Mattie AS, Lynch JT. Bile duct injuries, 1989 – 1993. A statewide experience. Connecticut Laparoscopic Cholecystectomy Registry. Arch Surg 1996; 131: 382 – 388 25 The Southern Surgeons Club. A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 1991; 324: 1073 – 1078 26 Roslyn JJ, Binns GS, Hughes EF et al. Open cholecystectomy. A contemporary analysis of 42,474 patients. Ann Surg 1993; 218: 129 – 137 27 Kassab C, Prat F, Liguory C et al. Endoscopic management of post-laparoscopic cholecystectomy biliary strictures. Long-term outcome in a multicenter study. Gastroenterol Clin Biol 2006; 30: 124 – 129 28 Cantwell CP, Pena CS, Gervais DA et al. Thirty years’ experience with balloon dilation of benign postoperative biliary strictures: long-term outcomes. Radiology 2008; 249: 1050 – 1057 29 Vitale GC, Tran TC, Davis BR et al. Endoscopic management of postcholecystectomy bile duct strictures. J Am Coll Surg 2008; 206: 918 – 923; discussion 924 – 925 30 De Palma GD, Persico G, Sottile R et al. Surgery or endoscopy for treatment of postcholecystectomy bile duct strictures? Am J Surg 2003; 185: 532 – 535 31 Tocchi A, Mazzoni G, Liotta G et al. Management of benign biliary strictures: biliary enteric anastomosis vs endoscopic stenting. Arch Surg 2000; 135: 153 – 157 32 Pitt HA, Kaufman SL, Coleman J et al. Benign postoperative biliary strictures. Operate or dilate? Ann Surg 1989; 210: 417 – 425; discussion 426 – 427 33 Lillemoe KD, Melton GB, Cameron JL et al. Postoperative bile duct strictures: management and outcome in the 1990s. Ann Surg 2000; 232: 430 – 441 34 Walsh RM, Vogt DP, Ponsky JL et al. Management of failed biliary repairs for major bile duct injuries after laparoscopic cholecystectomy. J Am Coll Surg 2004; 199: 192 – 197 35 Davids PH, Rauws EA, Coene PP et al. Endoscopic stenting for post-operative biliary strictures. Gastrointest Endosc 1992; 38: 12 – 18 36 Kahaleh M. “Cross-wired” biliary metal stents for malignant hilar strictures: a new window of opportunity? Gastrointest Endosc 2009; 69: 1361 – 1362
