Accepte

ReprintCouturipbr.aph

0039-60

� 2013

http://d

Treatment of failed Roux-en-Yhepaticojejunostomy afterpost-cholecystectomy bile ductsinjuriesAmine Benkabbou, MD,a Denis Castaing, MD,a,b,c Chady Salloum, MD,a Ren�e Adam, MD, PhD,a,c,d

Daniel Azoulay, MD, PhD,a,c and Eric Vibert, MD, PhD,a,b,c Villejuif, France

Background. Roux-en-Y hepaticojejunostomy (RYHJ) is the most well-accepted treatment for most post-cholecystectomy bile duct injuries (BDI). RYHJ failure is a complex situation that requires expertplanning and the possibility of using a combination of operative, radiologic, and endoscopic techniques.The aim of this study was to report our experience with a multidisciplinary approach to failed RYHJafter post-cholecystectomy BDI.Methods. Between January 1996 and March 2008, 44 consecutive patients were managed for RYHJfailure in our department. They presented with recurrent cholangitis in 40 patients (91%) and/orjaundice in 9 (20%). First-line treatment consisted of primary revisionary surgery in 26 cases (59%;repeat RYHJ in 22 and hepatectomy in 4) and a percutaneous approach in 18 cases (41%; biliaryinterventions in 16 and portal vein embolization in 2).Results. Postoperative mortality was nil. Postoperative morbidity was 11% after repeat RYHJ withouthepatectomy, 80% (bile leaks) after hepatectomy, and 10% (mild cholangitis and hemobilia) after apercutaneous approach. Delayed revisionary surgery with the intent to wait for bile duct dilation failedin all 5 patients. With a mean follow-up of 49 ± 40 months, second- or third-line treatment wasattempted in 7 patients (16%). One patient (2%) died because of suicide. Overall clinical successdefined by the absence of incapacitating biliary symptoms after treatment was achieved in 39 patients(89%).Conclusion. An immediate, multidisciplinary approach including repeat biliary surgery and/or apercutaneous approach in a tertiary hepatobiliary center is required to obtain good, long-term resultswhen treating the failure of RYHJ post-cholecystectomy BDI. (Surgery 2013;153:95-102.)

From the AH-HP,a Hopital Paul Brousse, Centre H�epato-Biliaire, Inserm,b Unite 785, the Universit�eParis-Sud,c and Inserm,d Unite 776, Villejuif, France

ROUX-EN-Y HEPATICOJEJUNOSTOMY (RYHJ) is the stan-dard treatment for most post cholecystectomy bileduct injuries (BDI) with long-term clinical successrates reaching 90%.1 Some patients who undergoRYHJ for BDI will experience incapacitating biliarysymptoms, such as jaundice or recurrent cholangi-tis.2 However, in addition to anastomotic stricture,several other, isolated or associated pathogenic fac-tors for RYHJ failure include intrahepatic calculi,intrahepatic stricture, and improper technicalconstruction of the Roux-en-Y limb. Management

d for publication June 14, 2012.

requests: Eric Vibert, MD, PhD, 12 avenue Paul Vaillanter, 94804 Villejuif Cedex, France. E-mail: eric.vibert@p.fr.

60/$ - see front matter

Mosby, Inc. All rights reserved.

x.doi.org/10.1016/j.surg.2012.06.028

of this complex situation requires careful and ex-pert management and the possibility of having acombination of operative, radiologic, and endo-scopic techniques. Few reports have specificallyanalyzed the results of failed biliary repairs forpost-cholecystectomy BDI.2-6 The aim of our studywas to evaluate the short- and long-term results ofa multidisciplinary approach regarding failedRYHJ after post-cholecystectomy BDI.

PATIENTS AND METHODS

Between January 1996 and March 2008, 44consecutive patients were treated in our depart-ment (Centre H�epato-Biliaire, Paul Brousse Hos-pital, Assistance Publique des Hopitaux de Paris,Villejuif, France) for the failure of RYHJ per-formed because of post-cholecystectomy BDI.Our group of patients comprised 13 males (30%)and 31 females (70%) with a mean (± SD) age of

SURGERY 95

Table I. Serum biochemistry findings at referral

Normal range Median Min Max

PT (%) >70 94 68 100Bilirubin (mmol/L) <17 13 5 134AP (UI/L) <120 183 123 1,128GGT (UI/L) <50 210 52 2,074AST (UI/L) <35 48 15 491ALT (UI/L) <43 62 9 776Creatinine (mmol/L) 18–106 64 47 140Protein (g/L) 60–80 71 48 80Albumin (g/L) >38 41 29 50Leukocytes (N./mL) 4,800–10,800 6,150 3,240 14,800Hemoglobin (g/dL) 12–16 12.9 8 15.8Platelets (N.103/mL) 150–400 256 85 658

ALT, Alanine aminotransferase; AP, alkaline phosphatase; AST, aspartate aminotransferase; GGT, gamma glutamyl transferase; PT, prothrombin time.

SurgeryJanuary 2013

96 Benkabbou et al

51 ± 14 years (range, 17–78). All BDI weresustained during cholecystectomy performedfor cholecystolithiasis. The approach for the cho-lecystectomy was laparoscopic in 35 patients (80%)and open in 9 patients (20%). The level of BDI wasassessed according to Bismuth’s classification (7):Type 1 in 2 patients (5%), type 2 in 18 (41%), type3 in 12 (27%), type 4 in 5 (11%), and type 5 in 7(16%). The time elapsing between BDI and initialrepair (RYHJ) was a median of 6 days (range, 0–703). The initial repair was performed very early(<48 hours) in 9 patients (20%), early (<45 days)in 22 (50%), and delayed (>45 days) in 13 (30%).The BDI was incurred and the failed RYHJ wasperformed in the same hospital in 34 patients(72%). Before referral to our department, aninitial revisionary operation for the failed RYHJwas performed in 10 patients (23%), and involveda hepaticojejunostomy repair in 3 patients, Roux-en-Y limb repair in 2 patients, and percutaneousdilatation of a stricture in 5 patients. These 10patients were referred to our department becauseof persistent biliary symptoms despite this revision-ary operation.

The patients were admitted to our departmentfor recurrent cholangitis in 40 patients (91%)and/or jaundice in 9 (20%). Recurrent cholangitiswas defined as fever >388C or episodic right upperquadrant pain with no identifiable source outsidethe hepatobiliary system occurring a minimum of 3times in the preceding year. Continuous or inter-mittent biliary symptoms had developed within amedian period 4 months (range, 0–204) since thepre-referral procedure. An external biliary drainwas present in 7 patients (16%), and no patienthad an active bile leak.

Cholestasis had been present in all patients. Nomajor coagulation and renal function

abnormalities were present (Table I). Leukocytosisand thrombocytopenia were present in 3 patients(7%) and 1 patient (2%), respectively.

Vascular and liver parenchymal assessments(Table II) were performed using routine abdominalultrasonography and computed tomography of theliver with intravenous contrast. These imaging mo-dalities revealed liver atrophy in 7 patients (16%),and evidence of injury to the main (2 patients[4%]) or right branch (6 patients [14%]) of the he-patic artery injury in 8 patients (18%). Liver atrophyand vascular injury were both present in 2 patients.

Biliary assessments were performed using per-cutaneous cholangiography in 34 patients (77%)and/or magnetic resonance cholangiography in 23(52%). These procedures revealed intrahepaticcalculi in 18 patients (41%) and bile duct dilationin 11 (25%). The level of obstruction was suspectedto be hilar or suprahilar in 39 patients (89%).

Treatment strategy was defined at a multidisci-plinary staff meeting including surgeons, radiolo-gists, and hepatogastroenterologists during acase-by-case analysis in our tertiary center that offersdifferent multidisciplinary approaches to hepato-biliary disorders (operative, endoscopy and inter-ventional radiology). The treatment strategycomprised 2 types of treatments: Revisionary sur-gery, including a revision of hepaticojejunostomyand/or hepatectomy, or a percutaneous approach,including biliary maneuvers and/or portal veinembolization. These treatments were performedalone or in combination and subsequently definedthe different lines of treatment in the same patient.

Revisionary surgery was considered in patientsin good general condition without uncontrolledbiliary sepsis and was designed to perform an end-to-side, wide, healthy, mucosa–mucosa hepaticoje-junostomy without tension and with a 70-cm long

Table II. Radiologic findings at referral

Findings n

Calculi 18 (41%)Bile duct dilation 18 (41%)

Bilateral 11Left liver 4Right liver 1Right sector 2

Vascular injury 8 (18%)Right branch of hepatic artery 6Hepatic artery 2

Parenchymal liver atrophy 7 (16%)Right liver 5Left lateral lobe 1Segment 4 1

Level of biliary obstructionInfrahilar 5 (11%)Hilar 25 (57%)Suprahilar 14 (32%)

SurgeryVolume 153, Number 1

Benkabbou et al 97

retrocolic Roux-en-Y limb. In patients with a lon-gitudinal stricture extending into the intrahepaticbile ducts, associated with liver atrophy, a hepatec-tomy was performed with or without RYHJ revi-sion. A percutaneous approach was considered inpatients with a (1) marked worsening in theirgeneral condition or hepatic function, or severesepsis, all of which contraindicated an operativeprocedure, (2) a local contraindication at the levelof the bile duct owing to a cavernous transforma-tion of a thrombosed portal vein or after numer-ous previous biliary interventions, or (3) anisolated short intra-hepatic biliary stricture.7 Portalvein embolization was performed in patients withan isolated longitudinal intrahepatic biliary stric-ture in attempt to induce parenchymal atrophyin the distribution of the diseased bile ducts.

Techniques of revisionary surgery. An end-to-side, wide, healthy, mucosa–mucosa hepaticojeju-nostomy without tension and with a 70-cm longretrocolic Roux-en-Y limb was the goal in eachpatient. Operative exploration consisted of 3 stages:Verification of the erroneous construction of theRoux-en-Y limb, exposure of the anastomotic areawith collection of a sample of bile, and assessmentof biliary anatomy and/or abnormalities (calculi,stricture) using intraoperative cholangiography.When feasible in patients with intra-hepatic bileduct dilation, a preoperative transhepatic cholan-giography followed by transhepatic biliary drainagewas performed. Intraoperatively, this drainage wasvery useful in localizing the bile duct after removalof the RYHJ and dissecting the hilar plate to exposethe primary biliary confluence and notably the left

bile duct. A hepatotomy between segments 5 and 4through of the bed of the gallbladder was used toaccess the secondary right biliary confluence.Visual magnifying aids were used routinely tooptimize biliary dissection, the recognition ofhealthy mucosa, and the anastomoses. These anas-tomoses were performed using 5/0 or 6/0 inter-rupted, nonabsorbable, monofilament sutures withthe knots tied on the external surface of theanastomosis. An ultrasonic dissector and bipolarcoagulation forceps were used routinely during anyhepatectomy.

Percutaneous approaches. All percutaneousprocedures were carried out in the operatingroom under full aseptic conditions as applicableto any operative procedure.7 The operating suitewas equipped with a Doppler Ultrasound (AlokaSSD 680, Aloka, Tokyo, Japan) and a light ampli-fier (Diasonics 3800; Diasonics, Milpitas, CA).These procedures were performed under eitherlocal anesthesia, neuroleptic analgesia with pre-medication, or general anesthesia with intubationif the duration of the procedure was expected tobe of a greater duration. Biliary maneuvers con-sisted of 3 successive stages: Establishing adequatetranshepatic and/or transjejunal8 access to thebiliary tract if not present, performing the re-quired intervention, and obtaining a contraststudy to demonstrate if the procedure was success-ful. When necessary, endoscopic control of theprocedure was ensured using a pediatric broncho-scope (diameter, 4 mm). Strictures were treatedby balloon dilatation and calculi by extractionand/or lithotripsy (Lithotron EL27, Walz Elektro-nik GMBH, Germany). Portal vein embolizationwas performed via a transhepatic approach.

Follow-up data were obtained by means ofreview of hospital and outpatient records. Allpatients were seen 1 month after hospital dis-charge and underwent computed tomographyand a complete biochemical assessment. Thereaf-ter, they were followed with liver ultrasonographyevery 4 months during the first year and every 6months for 2 years, and the yearly thereafter.Postoperative morbidity was assessed according tothe Clavien-Dindo classification.9 Clinical outcomewas determined according to the Terblanche classi-fication10: grade I, no biliary symptoms; grade II,transitory symptoms and no current symptoms;grade III, biliary symptoms requiring medicaltherapy; and grade IV, recurrent biliary symptomsrequiring correction or related to death. Ter-blanche class IV constituted a poor result. Ter-blanche I, II, and III constituted a clinical successwith excellent, good, and fair results, respectively.

SurgeryJanuary 2013

98 Benkabbou et al

RESULTS

The mean (± SD) follow-up period was 49 ± 40months (range, 2–153). One patient (2%) died asa result of suicide 44 months after the initialhepatectomy. In 7 patients (16%), $2 treatmentswere required, with a mean follow-up of 33 ± 36months (range, 2–85). At the time of last follow-up, clinical success had been achieved in 39patients (89%): 34 patients (77%) were asymptom-atic (Terblanche I–II), and 5 patients (11%) hadexperienced an improvement in their symptoms(Terblanche III). The overall result was poor(Terblanche IV) in 5 patients (11%). An overviewof the results is shown in Fig 1. First-line treatment(Fig 2) consisted of primary revision surgery (first-line revisionary surgery) in 26 patients (59%) anda percutaneous approach (first-line percutaneousapproach) in 18 patients (41%).

First-line revisionary surgery (n = 26 [59%];Table III). The Roux-en-Y limb was found to beshort (<70 cm) in 12 patients (44%) and was re-modeled to a length of 70 cm. Revisionary surgeryincluded repeat hepaticojejunostomy in 23 patients(89%), which involved more than a single duct in13 (54%; (range, 1–5). In 3 patients (11%), revi-sionary surgery was suboptimal because an anasto-mosis of the isolated right sector duct (1 case) orsegment 4 duct (2 cases) was impossible. Drainswere placed through the hepaticojejunostomy in 4patients (17%). An access limb of jejunum wasplaced under the abdominal wall to enable subse-quent percutaneous access in 3 patients (13%).Hepatectomy was performed in 4 patients (15%) in-cluding 2 right hepatectomies, 1 left hepatectomy,and 1 left lateral sectionectomy).

First-line percutaneous approach (n = 18 [41%];Table IV). Access to the biliary tract was establishedby catheterization of the jejunal limb (the ‘‘cul-de-sac’’) in 9 patients (54%), transhepatic catheteriza-tion in 4 patients (23%), or combined techniquesin 4 (23%). The procedures were performed un-der biliary endoscopic control in 12 patients(27%). Balloon dilatation of a stricture, extractionof calculi and/or biliary cast, and lithotripsy wereperformed in 12 (70%), 8 (47%), and 4 patients(23%), respectively. In 2 patients with calculi, su-tures exposed in the bile duct lumen were re-moved percutaneously under endoscopic controlvia access of the jejunal limb. Transhepatic portalvein embolization was performed in 2 patients(right posterior sectoral portal branch and rightportal branch).

Short-term results. In 5 patients (20%), first-linerevisionary surgery was delayed at referral in at-tempt to wait for bile duct dilation to develop and

thus optimize revision. None of these patientsdeveloped bile duct dilation after a mean ofwaiting time of 20 ± 17 months. During this period,4 patients developed recurrent cholangitis. Liverabscess and pylephlebitis of the right portal branchoccurred in 1 patient with a previous injury of theright branch of the hepatic artery.

Revisionary surgery without hepatectomy (n = 26 pro-cedures in 25 patients): Revisionary surgery withouthepatectomy was performed as first-line treatmentin 22 patients, as second-line treatment in 3patients (after a percutaneous approach), and asthird-line treatment in 1 patient (after revisionarysurgery followed by a second-line percutaneousapproach). Morbidity occurred after 3 of 26 pro-cedures (11%): Abdominal hematoma managedwith transfusion in 1 patient (Clavien-Dindo II),cholangitis managed with antibiotics in 1 (Clavien-Dindo II), and acute pancreatitis that requiredexploratory laparotomy for suspected biliary peri-tonitis in 1 (Clavien-Dindo IIIb).

Revisionary surgery with hepatectomy (n = 5 proce-dures in 4 patients): Hepatectomy was performedas first-line treatment in 4 patients and as second-line treatment in one who underwent operationtwice. Bile leaks occurred after 4 procedures(80%). Morbidity occurred as Clavien-Dindo IIafter 3 procedures and Clavien-Dindo IIIb after1 first-line procedure complicated by a bilio-pleural fistula managed with prolonged drainagethat progressed to a chronic external fistula re-quiring repeat hepatectomy.

Percutaneous approach (n = 120 procedures in 21 pa-tients): A percutaneous approach was adopted asfirst-line treatment in 18 patients and as addi-tional treatment in 3 (after initial revisionarysurgery). In 1 patient, additional treatment con-sisted of a combination of biliary maneuvers andright sectoral portal vein embolization. There wasno mortality. No morbidity was observed afterportal vein embolization. Hemobilia that did notrequire a blood transfusion (Clavien-Dindo I)and/or cholangitis managed with IV antibiotics(Clavien-Dindo I) occurred after 10% of thebiliary interventions.

Long-term results. Revisionary surgery withouthepatectomy (n = 22 patients): Satisfactory primaryresults were achieved in 18 patients (82%): 17patients (94%) became asymptomatic (TerblancheI–II) and 1 patient (6%) improved, although withsome symptoms (Terblanche III). In 4 patients(9%), the symptoms did not improve (TerblancheIV). One of these patients experienced generallyfair results (Terblanche III) after an additionalpercutaneous approach.

Fig 1. Overview of the results as a function of the Terblanche classification of 44 patients, achieved with 1, 2 or 3 step(s).HJ, Hepaticojejunostomy; PVE, portal vein embolization; T, Terblanche classification10; TI, no biliary symptoms; TII,transitory biliary symptoms, no current symptoms; TIII, biliary symptoms requiring medical therapy; TIV, recurrent bil-iary symptoms requiring correction or related to death.

Fig 2. Overall results of the revisionary approach. Fullline: Overall results including additional revision. Dottedline: Results after first-line revision.

SurgeryVolume 153, Number 1

Benkabbou et al 99

Revision surgery with hepatectomy (n = 4 patients):Good primary results were achieved in 3/4 patientswho were asymptomatic (Terblanche I–II), but thesymptoms did not improve in the other patient.This patient (Terblanche IV) underwent addi-tional repeat hepatectomy for a bile leak from anexcluded segment 4 with histologic evidence ofsecondary biliary cirrhosis, but this procedurefailed. She died from suicide.

Biliary interventions (n = 16 patients): Good pri-mary results were achieved in 14 patients (87%):10 (71%) were asymptomatic (Terblanche I–II)

and 4 (29%) had improvement in their symptoms(Terblanche III). In 2 patients (6%), the symptomswere not primarily improved (Terblanche IV), butboth became asymptomatic (Terblanche I–II) afterfurther revisionary surgery.

Portal vein embolization (n = 2 patients): Poorresults were obtained in these 2 patients. Onepatient who underwent portal vein embolizationalone had overall good results (Terblanche II)after additional revisionary surgery, whereas thesecond with initial combined portal vein emboli-zation and biliary interventions was listed for livertransplantation because of development of second-ary biliary cirrhosis.

DISCUSSION

This study shows that with an experiencedmultidisciplinary approach (according to the strat-egy summarized in Fig 3), patients with a failedRYHJ after post-cholecystectomy BDI can achievegood long-term clinical success in 89%. These re-sults required more than the first-line revision in16% of patients. Although we showed that waitingfor bile duct dilation before revisionary surgery wasnot successful; moreover, we were unable to iden-tify any prognostic predictive factor at referral inour small and heterogeneous population.

Concordant data suggest that both repair ofBDI repair by an expert hepatobiliary surgeon11

and a multidisciplinary approach involving

Table III. Revisionary surgery procedures (n = 31)

First line(n = 26)

Additional(n = 5*)

Hepaticojejunostomy revision 23 (88%) 4 (80%)Number of ducts/anastomosis

1 13 22 5 13 6 —4 1 15 1 —

Endobiliary extraction 18 (69%) 2 (40%)Calculi and/or biliary cast 15 2Clips 2 —Alimentary 1 1

R-en-Y revision with 12 (46%) 0Hepatectomy 4 (15%) 1 (20%)

Right liver 2 —Left liver 2 —Left lateral lobe 1 —

Segment 4 (excluded bile leak) — 1

*Second-line in 4 patients and third-line in 1 patient.

Table IV. Percutaneous approach procedures(n = 21)

First line(n = 18)

Additional(n = 3)

Portal vein embolization 2 (11%) 1Right branch 1 —Right posterior sector branch 1 1

Biliary maneuvers 17 (94%) 3 (100%)Catheterization approachTranshepatic 9 1Transjejunal 4 —Combined 4 2

Endoscopic control 12 (67%) 2ProceduresStricture dilatation 12 (70%) 3Calculi and/or biliary cast

extraction8 (47%) 1

Lithotripsy 4 (23%) 1Suture extraction 2 (12%) —

Median number of procedures(range)

5 (2–23) 10 (2–24)

SurgeryJanuary 2013

100 Benkabbou et al

gastroenterologists, radiologists, and surgeons12

favorably affects outcomes. RYHJ has been success-fully used in such cases since the 1970s13 and is cur-rently the standard treatment, with success ratesof up to 91% and very long-term (>10 years)stricture-free survival.14,15 Nevertheless, $10% ofthese patients will suffer from a failure of theRYHJ.2

This failure involves isolated or associatedpathogenic factors responsible for recurrent epi-sodes of cholangitis in 90% of patients and/orjaundice in 20%.16 These symptoms of biliary ob-struction may occur without anastomotic strictureand are possibly caused by the passage of calculior by enterobiliary reflux induced by inappropri-ate construction of the Roux-en-Y limb.2 Between80% and 90% of patients with failure of the biliaryrepair develop symptoms within 5–7 years.3,17 Thisdelay can vary from a few days to several years18 (amaximum of 17 years in our series), which empha-sizes the need for prolonged follow-up. By con-trast, some patients who are clinically ‘‘normal’’after repair can continue to experience mildincreases in serum bilirubin and/or gamma glu-tamyl transferase activity during long-term fol-low-up.19

The management of RYHJ failure is hamperedby the fact that biliary strictures are found at ahigher level than before the first attempt at BDIrepair. Indeed, with each failed attempt, the levelof the scarred biliary stricture recedes higher intothe hepatic hilum. In our experience, 89% ofreferred patients had hilar or suprahilar strictures,

although at the time of initial biliary repair only38% of them had an injury or stricture at that level.Hence, a precise assessment of bile duct anatomyand the level of the stricture are critical to thesuccess of any revisionary strategy.4,16 In this set-ting, percutaneous cholangiography can be consid-ered as the ‘‘gold standard’’ because it providescritical information on the biliary anatomy andductal communication. Magnetic resonance chol-angiography has been claimed to be as reliable asthe percutaneous approach in defining biliarytree anatomy20 and may be the procedure ofchoice in selected patients.

The second major problem encountered in themanagement of RYHJ failure is an absence of intra-hepatic bile duct dilation that complicates thebiliary repair. This situation was observed in 59%of our patients. It should be noted that revisionarysurgery was delayed in 5 patients specifically to waitfor bile duct dilation, but this strategy failed in all 5patients and was associated with severe morbidityduring the waiting period. We, therefore, considerthat when surgical revision criteria are fulfilled, anelective procedure assisted by the routine use ofoptical magnification should be scheduled withoutwaiting for bile duct dilation.

From a technical point of view, if the biliaryconfluence is intact, a wide stoma of healthy ductcan be achieved by extending the opening in thebile duct to the extrahepatic portion of the lefthepatic duct.21 If the biliary confluence is obliter-ated, and the left and right hepatic ducts are

Fig 3. Initial treatment algorithm in failed RYHJ after post-cholecystectomy BDI.

SurgeryVolume 153, Number 1

Benkabbou et al 101

isolated, hilar plate dissection is necessary up tothe level at which a healthy duct can be found.When the results of revision of the hepaticojeju-nostomy are expected to be problematic becauseof a suprahilar longitudinal stricture and/or liveratrophy, hepatectomy must be considered. The ra-tionale for partial liver resection in patients withcomplex RYHJ failure is that hepatectomy removesirreversible fibrotic parenchyma and prevents theprogressive liver damage caused by permanentbile stasis and/or recurrent cholangitis.1,22 Hepa-tectomy with biliary reconstruction in the settingof complex BDI produces excellent long-term re-sults despite a high rate of severe postoperativecomplications.22 Total hepatectomy followed byliver transplantation has been considered when re-visionary surgery and percutaneous approacheshave failed or were not technically feasible in thepresence of diffuse, secondary biliary cirrhosis.23,24

In difficult cases, and especially in patients whohave undergone $2 previous operative repairsand/or in whom portal hypertension is present,4

a percutaneous biliary approach is very useful to as-sess the precise level of the stricture and the healthof the mucosa. In this setting, endoluminal dilata-tion associated with complete extraction of all in-traductal debris may represent either a chancefor long-term remission or a step toward revision-ary surgery.7

In conclusion, good long-term results can beachieved after a failed RYHJ failure after post-cholecystectomy BDI by means of a case-by-case

analysis and an immediate multidisciplinary ap-proach in tertiary hepatobiliary centers.

REFERENCES

1. Murr MM, Gigot JF, Nagorney DM, Harmsen WS, IlstrupDM, Farnell MB. Long-term results of biliary reconstructionafter laparoscopic bile duct injuries. Arch Surg 1999;134:604-9.

2. Kozicki I, Bielecki K, Kawalski A, Krolicki L. Repeated re-construction for recurrent benign bile duct stricture. Br JSurg 1994;81:677-9.

3. Pellegrini CA, Thomas MJ, Way LW. Recurrent biliary stric-ture. Patterns of recurrence and outcome of surgical ther-apy. Am J Surg 1984;147:175-80.

4. Chaudhary A, Chandra A, Negi SS, Sachdev A. Reoperativesurgery for postcholecystectomy bile duct injuries. Dig Surg2002;19:22-7.

5. Walsh RM, Vogt DP, Ponsky JL, Brown N, Mascha E, Hen-derson JM. Management of failed biliary repairs for majorbile duct injuries after laparoscopic cholecystectomy. J AmColl Surg 2004;199:192-7.

6. Pottakkat B, Sikora SS, Kumar A, Saxena R, Kapoor VK. Re-current bile duct stricture: causes and long-term results ofsurgical management. J Hepatobiliary Pancreat Surg 2007;14:171-6.

7. Castaing D, Vibert E, Bhangui P, Salloum C, Smail A, AdamR, et al. Results of percutaneous manoeuvres in biliary dis-ease: the Paul Brousse experience. Surg Endosc 2011;25:1858-65.

8. Castaing D, Azoulay D, Bismuth H. [Percutaneous catheter-ization of the intestinal loop of hepatico-jejunostomy: a newpossibility in the treatment of complex biliary diseases].Gastroenterol Clin Biol 1999;23:882-6.

9. Clavien PA, Barkun J, de Oliveira ML, Vauthey JN, Dindo D,Schulick RD, et al. The Clavien-Dindo classification of surgi-cal complications: five-year experience. Ann Surg 2009;250:187-96.

SurgeryJanuary 2013

102 Benkabbou et al

10. Terblanche J, Worthley CS, Spence RA, Krige JE. High orlow hepaticojejunostomy for bile duct strictures? Surgery1990;108:828-34.

11. Stewart L, Way LW. Bile duct injuries during laparoscopiccholecystectomy. Factors that influence the results of treat-ment. Arch Surg 1995;130:1123-8.

12. de Reuver PR, Rauws EA, Bruno MJ, Lameris JS, Busch OR,van Gulik TM, et al. Survival in bile duct injury patients afterlaparoscopic cholecystectomy: a multidisciplinary approachof gastroenterologists, radiologists, and surgeons. Surgery2007;142:1-9.

13. Bismuth H, Franco D, Corlette MB, Hepp J. Long termresults of Roux-en-Y hepaticojejunostomy. Surg GynecolObstet 1978;146:161-7.

14. Chapman WC, Halevy A, Blumgart LH, Benjamin IS. Post-cholecystectomy bile duct strictures. Management and out-come in 130 patients. Arch Surg 1995;130:597-602.

15. Gigot JF. Bile duct injury during laparoscopic cholecystec-tomy: risk factors, mechanisms, type, severity and immedi-ate detection. Acta Chir Belg 2003;103:154-60.

16. Lillemoe KD,Melton GB, Cameron JL, Pitt HA, Campbell KA,Talamini MA, et al. Postoperative bile duct strictures: manage-ment and outcome in the 1990s. Ann Surg 2000;232:430-41.

17. Pitt HA, Miyamoto T, Parapatis SK, Tompkins RK, LongmireWP Jr. Factors influencing outcome in patients with postop-erative biliary strictures. Am J Surg 1982;144:14-21.

18. Tocchi A, Costa G, Lepre L, Liotta G, Mazzoni G, Sita A.The long-term outcome of hepaticojejunostomy in thetreatment of benign bile duct strictures. Ann Surg 1996;224:162-7.

19. Fialkowski EA, Winslow ER, Scott MG, Hawkins WG, Line-han DC, Strasberg SM. Establishing "normal" values for liverfunction tests after reconstruction of biliary injuries. J AmColl Surg 2008;207:705-9.

20. Chaudhary A, Negi SS, Puri SK, Narang P. Comparison ofmagnetic resonance cholangiography and percutaneoustranshepatic cholangiography in the evaluation of bileduct strictures after cholecystectomy. Br J Surg 2002;89:433-6.

21. Hepp J. Hepaticojejunostomy using the left biliary trunk foriatrogenic biliary lesions: the French connection. World JSurg 1985;9:507-11.

22. Laurent A, Sauvanet A, Farges O, Watrin T, Rivkine E, Bel-ghiti J. Major hepatectomy for the treatment of complexbile duct injury. Ann Surg 2008;248:77-83.

23. Nordin A, Makisalo H, Isoniemi H, Halme L, Lindgren L,Hockerstedt K. Iatrogenic lesion at cholecystectomy re-sulting in liver transplantation. Transplant Proc 2001;33:2499-500.

24. Thomson BN, Parks RW, Madhavan KK, Garden OJ. Liverresection and transplantation in the management of iatro-genic biliary injury. World J Surg 2007;31:2363-9.