Gut 1994; 35: 122-126

Management ofcommon bile duct stricture causedby chronic pancreatitis with metal mesh selfexpandable stents

J Deviere, M Cremer, M Baize, J Love, B Sugai, A Vandermeeren

AbstractTwenty patients with chronic pancreatitis andsigns of biliary obstruction were treated byendoscopic placement of self expandablemetal mesh stents, and followed up prospect-ively. Eleven had been treated previously withplastic endoprostheses. All had persistentcholestasis, seven patients had jaundice, andthree overt cholangitis. Endoscopic stentplacement was successful in all cases. No earlyclinical complication was seen and cholestasis,jaundice or cholangitis rapidly resolved in allpatients. Mean follow up was 33 months (range24 to 42) and consisted of clinical evaluation,ultrasonography, and endoscopic retrogradecholangiopancreatography (ERCP). In 18patients, successive ERCPs and cholangio-scopies have shown that the metal meshinitially embeds in the bile duct wall and israpidly covered by a continuous tissue bythree months. The stent lumen remainedpatent and functional throughout the follow upperiod except in two patients who developedepithelial hyperplasia within the stent resultingin recurrent biliary obstruction, three and sixmonths after placement. They were treatedendoscopically with standard plastic stentswith one of these patients ultimately requiringsurgical drainage. No patient free of clinical orradiological signs of epithelial hyperplasiaafter six months developed obstruction later.This new treatment could become an effectivealternative to surgical biliary diversion iffurther controlled follow up studies confirmthe initial impression that self expandablemetal mesh stents offer a low morbidityalternative for longterm biliary drainage inchronic pancreatitis without the inconveni-ence associated with plastic stents.(Gut 1994; 35: 122-126)

MedicosurgicalDepartment ofGastroenterology,H6pital Erasme,Universite Libre deBruxelles, Brussels,BelgiumJ DeviereM CremerM BaizeJ LoveB SugaiA VandermeerenCorrespondence to:Dr J Deviere, ULB H6pitalErasme, Route de Lennik 808,B - 1070 Brussels, Belgium.Accepted for publication8 April 1993

About 20% of patients with chronic pancreatitis(CP) develop common bile duct (CBD) stenosisand subsequent cholestasis or jaundice.1`S Asbiliary obstruction can lead to secondary biliarycirrhosis or recurrent episodes of cholangitis,surgical treatment has been recommended forstricture associated with persistent cholestasis.3-5

Endoscopic biliary drainage using large plasticstents, is an alternative to surgery in the manage-ment of these patients, and can rapidly resolvecholestasis, jaundice, and cholangitis. Longtermresults of biliary stenting, however, have beendisappointing because stent clogging or migra-tion are responsible for significant morbidity.2For this reason, close monitoring of thesepatients is required, with stent replacement

when necessary. Moreover, a true recalibrationof the biliary stenosis after longterm plasticstenting is rarely achieved even after insertion ofmultiple 10 French plastic stents.

Also, surgical morbidity related to biliarydrainage remains high in these alcoholic andoften debilitated patients.3" Accordingly, it isreasonable to continue the search for alternativetreatments that provide equivalent or betterresults with lower morbidity.

Self expandable steel braided endoprostheseswere first used for the treatment of vascular andurological benign strictures, with good clinicalresults.78 Since then these stents have been usedin bile ducts for malignant obstructive jaundiceand are placed either percutaneously or endo-scopically.?12 Their use in malignant stricturesprovides immediate drainage, avoiding the earlycomplications encountered with plastic stents.The longterm results are impaired, however,by tumoral overgrowth or more frequentlyby ingrowth through the metallic mesh intothe stent, resulting in recurrent jaundice orcholangitis. 'I2As ingrowth and subsequent reocclusion

seems to be tumour related, we studied this newtreatment method in patients with severe CP andbenign symptomatic biliary stricture.

Patients and methods

PATIENTSBetween June 1989 and January 1991, 20patients with CP (mean age: 45 (27-61) years,four women, 16 men), and with signs of persist-ent biliary obstruction associated with CBDstricture were treated by endoscopic placementof self expandable metallic stents (Wallstent,Schneider SA, Biilach, Switzerland). All but onewere alcoholics. Eleven ofthem had been treatedpreviously with size 10 French plastic stents for amean duration of 15 (2-36) months. Plasticstents were replaced by metallic stents only incase of clogging or dislodgement. Formal con-sent for this new type of treatment was obtainedfrom each patient and the study conformed toour institution's guidelines concerning medicalethics.

Stenosis discovered during ERCP withoutpersistent cholestasis, was not considered anindication for stenting. The initial indicationwas, in all cases, CBD stricture associated with atleast persistent cholestasis (alkaline phosphataseconcentrations greater than twice the normalvalues for more than three months and notdecreasing for more than two weeks). Sevenpatients also had jaundice (mean bilirubin values

122

on March 3, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.35.1.122 on 1 January 1994. D

ownloaded from

Management ofcommon bile duct stricture caused by chronic pancreatitis with metal mesh selfexpandable stents

46 (20-142) mg/l) and three had overt cholangitisat the time of stent placement.The mean duration of the disease from first

diagnosis to the time of first biliary stent place-ment was 65 months (range 5-180). Fourpatients had cholestasis and stricture at the timeof the first diagnostic procedure. Five patientshad had previous surgery of the pancreas(3 lateral pancreaticojejunostomies, 2 tail pan-createctomies and 1 cyst-jejunostomy).

Figure 1: Patient no 5:47year old womanwith previouspancreaticojejunostomy.(A) appearance ofthe biliarystenosis after removal ofaplastic stent left in placefornine months; (B) the stentcarrying double membranecatheter is introduced in thecommon bile duct and thestent is endoscopically andfluoroscopically adjusted;(C) control cholangiographyafter stent expansion; (D)control endoscopic retrogradecholangiopancreatography24 months later showing thatthe stent is covered with a2 mm thick epithelium. Thelumen remains largely open.

R _| ,,

TECHNIQUE (Fig 1)The technique of endoscopic placement ofbiliary metal mesh stents has been described."Briefly, after diagnostic ERCP, endoscopicsphincterotomy is carried out with a long nosesphincterotome. A guide wire is impacted intotho.intrahepatic ducts, through the stenosis. Thestent carrying double membrane catheter ispassed over this guide wire. The proximal end ofthe stent is adjusted to be level with the papillaryorifice. The insertion catheter is then pressurisedwith contrast medium to 75 mm Hg to separatethe two membranes. The outer layer is then

progressively withdrawn to permit stent expan-sion. Proximal realignment remains possible aslong as the stent is not completely opened. Thedelivery catheter is then removed and a finalopacification of the CBD is performed to assessthe quality of drainage.A 34 mm long stent was used in all cases. The

internal diameter becomes 10 mm after fullexpansion. This length was adapted to the intra-pancreatic choledocus leaving the proximalcommon bile duct free. The stent covered theentire stricture in each case and, as only the distalintrapancreatic portion of the CBD was stentedwith this unremovable material, it should notinterfere with a possible future hepaticojejuno-stomy or choledocoduodenostomy. No dilata-tion was performed before or after insertion ofthe metal mesh stent whose complete expansionis spontaneously achieved after five days. Thiswas seen in the first five patients treated, by serialcholangiographies through indwelling naso-biliary catheters.

FOLLOW UPPatients were prospectively evaluated one monthafter stenting, by clinical examination, labora-tory tests, ultrasonography, and ERCP. There-after, ultrasonography and laboratory tests weredone every three months and ERCP every sixmonths, during the two first years of follow up.These examinations were done on an ambulatorybasis and patients were readmitted only in case ofrecurrent cholestasis or cholangitis. All thepatients were seen in consultation and laboratorytests performed at the end of the presentlyreported follow up.

ResultsCBD strictures corresponded to the type I ofCaroli and Nora in 11 cases and type III in ninecases. The morphological appearance of thepancreatitis at ERCP consisted, in all cases, of adistal stricture of the main pancreatic duct withupstream dilatation with or without pancreaticcalculi. We identify this morphology as types IVand V pancreatitis in our classification,'3 thesebeing nearly always encountered in cases ofCBDstenosis in our experience. Two patients hadpancreas divisum. It should be noted that 18 ofthese patients also had therapeutic endoscopy ofthe pancreas with the aim at restoring thepatency of the pancreatic duct (pancreaticsphincterotomy, extracorporeal shock wavelithotripsy, with or without stent placement inthe main pancreatic duct).

IMMEDIATEERCP was successful in all patients and in eachcase the stent was inserted in the CBD by theendoscopic route, during the same session; therewere no sphincterotomy related complications.We encountered two technical complications:one stent was badly positioned and was left tooproximally because of a technical problemduring insertion. It was immediately removedwith a snare and replaced with a new one.Another stent seemed, after six months, to be

B

123

on March 3, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.35.1.122 on 1 January 1994. D

ownloaded from

Deviere, Cremer, Baize, Love, Sugai, Vandermeeren

^ 1200

D1000 -

800

X 600 \0

0 400

200

0 L1 _ ,0 1 2 3 10

Days after stenting

impacted in the CBD at the site of a sharpangulation. Clinically no problem was seen butwe preferred to place a second stent, moreproximally, overlapping the first one, to correctthis angulation.There were no problems of early clogging or

dislodgement. Cholestasis, cholangitis, andjaundice resolved quickly (Fig 2) in all but onepatient who continued despite adequate drainageto have slight and persistent cholestasis. Liverbiopsy showed that this patient already hadsecondary biliary cirrhosis.

LATEThe mean follow up after common bile ductstenting is now 33 months, ranging from 24 to 42months and is more than two years in all patients(Table, Fig 3). Eighteen patients have now beenfollowed up for 33 (24-42) months without anybiliary problems. Follow up ERCPs (in allpatients) and cholangioscopies (in five patients)have shown the metal mesh to be embedded inthe bile duct wall with thickened mucosa grow-ing between the struts of the stent. At sixmonths, the struts were usually buried by themucosa, giving the impression of a continuous'membrane' lining the stent. Its thickness isabout 2 mm and the CBD lumen remains largelypatent inside the stent (Fig 1). Two patients(10%) developed epithelial hyperplasia withinthe stent resulting in recurrent cholestasis forone and jaundice for the other, three and sixmonths after stent placement, respectively (Fig4). A second self expanding stent was insertedinto the first stent in each case without longtermsuccess, as obstruction occurred again after threemonths in both. We then placed two plastic 10French stents through the metallic stent. One ofthese patients (Table; patient 15) required a later

Patients' data: sex, alcoholism, previous surgery ofthe pancreas, age at the first diagnosis ofchronic pancreatitis, duration ofbiliary drainage with plastic IO F stents before metal stent placement, age at the time ofmetal mesh stent placement, date ofimplantation, evolution, and duration offollow up with metal stent in situ

DurationAge at Plastic Age at offollow

Previous diagnosis stent metal stent upPatient Sex Alcoholism surgery (y) (months) (y) Implantation Evolution (months)

1 M + - 23 16 34 6/89 Stent patent 422 F + 34 - 36 9/89 Stent patent 393 M + + 52 36 61 9/89 Stent patent 394 M + - 27 18 34 9/89 Stent patent 395 F + + 36 9 47 9/89 Stent patent 396 M + - 53 13 58 11/89 Stent patent 377 M + - 35 2 43 1/90 Stent patent 358 M + - 36 23 39 1/90* Stent patent 359 M + - 29 - 33 2/90 8/90 Jaundice 3410 M + - 60 - 60 2/90 Stent patent 3411 M + + 37 17 39 3/90 Stent patent 3312 M + 49 - 49 3/90 Stent patent 3313 M + - 39 20 46 4/90 Stent patent 3214 M + - 53 - 54 5/90 Stent patent 3115 F - + 45 - 55 9/90 12/90 Cholestasis 2716 F + 27 - 27 9/90 Stent patent 2717 M + 38 - 45 10/90 Stent patent 2618 M + + 33 10 49 1/91 Stent patent 2419 M + - 43 4 46 1/91t Stent patent 2420 M + 46 - 47 1/91 Stent patent 24

*Second stent placed to correct a sharp biliary angulation.tMisplacement, stent removed and immediately replaced by another one.

H

HPatients followed u

20 20 20 20 20 20 20 20 20nnnM n E n n n ni1614

hepaticojejunostomy, which was performedwithout any problem related to the stent in place.

All patients (90%) who were free of obstruc-tion from hyperplasia six months after implanta-tion, remained asymptomatic subsequently. Thetwo patients who developed stent occlusion as aresult of hyperplasia, already had abnormalhyperplasia one month after stenting. No differ-ence in clinical or morphological features, orboth were found in these two patients, comparedwith the others.

Biopsies performed within the stents showedhyperplastic papillary mucosa covered by biliaryor intestinal epithelium. The microscopicappearance was similar in the two patients withabnormal hyperplasia.

.(..

DiscussionPersistent CBD stricture requires treatment

Figure 2: Alkalinephosphatase concentrations(mean (SEM), n-250IU)after stenting. Broken line=normal values.

100::-

C 80ca)a,0C 60a,CD

° 40CD

20

0o 3 6 9 12 15 18 21 24 27 30 33 36 39 42

Months in situFigure 3: Probability of stent patency in chronic pancreatitis associated biliary obstruction.

124

on March 3, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.35.1.122 on 1 January 1994. D

ownloaded from

Management ofcommon bile duct stricture caused by chronic pancreatitis with metal mesh selfexpandable stents

when accompanied by a prolonged rise in serumalkaline phosphatase concentration." This is anextremely valuable marker of significant biliaryobstruction,6 which may lead to the developmentof hepatic damage and secondary biliary cirrho-sis. In our experience, CBD stricture is seen atERCP in 30% of patients with severe CP and isassociated with persistent cholestasis, jaundice,or cholangitis in 9%.? In published reports,biliary tract obstruction is the primary indicationfor surgery in about 10% of patients with CP. Inaddition, some patients having an operationaimed primarily at relieving pancreatic ductobstruction or resecting a diseased pancreas arealso found to have an element of biliary tractobstruction requiring surgical drainage.3 '4Our initial experience ofCBD drainage in CP

with plastic stents was somewhat disappointing.Placement of large bore plastic biliary stents wasa good initial treatment for cholangitis or jaun-dice when present. Longterm results, however,were unsatisfactory because of clogging or dis-lodgement, and required frequent replacement,which is not acceptable to these often youngpatients.2 Our opinion at that time was thatpermanent endoscopic biliary drainage in CPshould be reserved for patients whose poorclinical condition made an operation unsuitable.

Self expandable biliary stents have been usedfor palliative treatment of malignant biliarystrictures. All the published results available atthis time show that this new material is easy toplace and avoids most of the early stent relatedcomplications - that is, early clogging or dis-lodgement."' '5 The internal diameter of thesestents is 10 mm, which is three times thediameter of the largest plastic stents. To ourknowledge, early clogging or dislodgement hasnot been reported. The initial enthusiasm con-cerning longterm patency of these stents inmalignancies, however, has diminished becauseof frequent obstruction because of tumouringrowth through the metallic struts, or tumourovergrowth.91' As obstruction in these cases isbecause of the tumoral process itself, we postu-lated that their longterm patency in benignbiliary strictures was likely to be better. This wassuggested by a report of the preliminary results

ofmesh stent implantation in postsurgical biliarystenosis. With an average follow up of eightmonths, no cases of reobstruction were seen inseven patients.14 These results are, however,probably too optimistic as a few cases of reocclu-sion by hyperplasia through the mesh have beenreported with metallic stents placed either per-cutaneously9'1 or endoscopically (C Liguory,personal communication) for postsurgical CBDstrictures.We chose to treat patients with CP associated

biliary strictures with these devices for tworeasons. The first is that we have a large numberof patients with a good follow up in our institu-tion and we are reluctant to propose surgery onlyfor biliary diversion when there is not yet anindication for pancreatic drainage, as repeatedsurgery is associated with higher morbidity. Thesecond is that the CBD stricture is always distalallowing for the use of comparatively short (34mm) mesh stents.

Indeed, as these stents are not removable andbecause we are treating patients with benigndisease, we did not want to compromise apotential surgical biliary diversion. As the stentis left in the distal part of the CBD, a futurehepaticojejunostomy or choledochoduodeno-stomy still remains possible as was shown in oneof our patients.The potential for malignant transformation at

the site of implantation as a result of chronicirritation from the metal remains a theoreticalproblem worthy of consideration. Clearly, thisstudy cannot provide a definitive answer.Surgeons, however, have used metallic implantssuch as clips and automatic sutures for a longtime without any such complication reported.'6It seems therefore that implanted metal despite apotential to induce fibrosis or inflammation, hasnot been responsible for malignancy. Alsoreassuring are animal studies in which thesestents have shown excellent biological toleranceand no dysplastic changes have been noted. 17

Ninety per cent of stents remain patent atthree years and when abnormal hyperplasiaoccurs, resulting in stent occlusion, it occursduring the first six months. Two of our 20patients required further biliary drainage.Epithelial hyperplasia is a constant finding inthe presence of these stents, probably resultingfrom mechanical irritation. The reason why itbecomes more pronounced in some patientsremains unanswered. In the two cases ofreobstruction, the pattern mimicked the pre-vious stricture, suggesting that the radial forceexerted by the mesh on the wall may have led todeeper embedding of the mesh where thestricture was tight or may have provoked moreimportant hyperplasia, or both. This maypartially explain the better results seen herecompared with those obtained with metal stentsin postsurgical strictures.'5 In these cases, thestricture is short and tight and the radial forceexerted by the individual wires at the level of thestricture is therefore higher.The lower end of the stent must be level with

the papilla to avoid the presence of metallic meshwithin the duodenum and must extend over theentire stricture. In the event ofmisplacement, aswe saw in one patient, immediate removal is

Figure 4: (A) patient no 9.Endoscopic retrogradecholangiopancreatography(ERCP) six months aftermetal stent placement whenthe patient developedjaundice because ofepithelial hyperplasia intothe stent. A plastic stent is inthe pancreatic duct. Bycomparison with (B), patientno 3, control ERCP 24months after metal stentimplantation.

125

on March 3, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.35.1.122 on 1 January 1994. D

ownloaded from

126 Deviere, Cremer, Baize, Love, Sugai, Vandermeeren

possible. Where there is sharp angulation of theCBD, the upper part of the stent should be 1 cmabove or 1 cm below the angulation, but not atthe point of angulation itself, to avoid impactionof the metallic mesh in the CBD wall andpossibly further impaired drainage.Within a month of placement, the metallic

mesh embeds into the bile duct wall and, afterthree months, the struts are buried by themucosa, giving the impression of a continuousmembrane covering the inner stent. This findingis comparable with animal studies,'7 whereimplantation of metal stents is followed bymucosal hyperplasia through the space betweenthe struts, which finally are buried by themucosa. This finding is suggestive of destructionof the original mucosal layer during expansionwith subsequent healing and proliferation.'Clogging' ofthese stents by material comparablewith the one seen in plastic stents has not beenseen, probably because of both a larger diameterof the lumen and the nature of the stent's innerlining after re-epithelialisation.

Longer follow up and possibly controlledclinical trials are needed to determine if metalstents have the potential to become the firstchoice treatment for persistent, symptomatic CPassociated biliary stricture. This study suggests,however, that there may be a satisfactory lowmorbidity alternative to biliary surgery pro-viding efficient prolonged biliary drainage,without the usual complications and discomfortassociated with plastic stents.

The authors thank G Ghattas, MD, and C Matthys for their helpin proof reading and typing the manuscript.

1 Scott J, Summerfield JA, Elias E, Dick R, Sherlock S. Chronicpancreatitis: a cause of cholestasis. Gut 1977; 18: 196-201.

2 Deviere J, Devaere S, Baize M, Cremer M. Endoscopic biliarydrainage in chronic pancreatitis. Gastrointest Endosc 1990;36:96-100.

3 Littenberg G, Afroudakis A, Kaplowitz N. Common bileduct stenosis from pancreatitis: a clinical and pathologicspectrum. Medicine 1979; 58: 385-411.

4 Stabile BE, Calabria R, Wilson SE, Passaro E. Stricture of thecommon bile duct from chronic pancreatitis. Surg GynecolObstet 1987; 165: 121-6.

5 Warshaw AL, Shapiro RH, Ferrucci JJ, Galdabani JJ.Persistent obstructive jaundice, cholangitis and biliarycirrhosis due to common bile duct stenosis in chronicpancreatitis. Gastroenterology 1976; 70: 562-7.

6 Stahl TJ, O'Connor Allen M, Ansel HJ, Vennes JA. Partialbiliary obstruction caused by chronic pancreatitis. Anapproval of indications for surgical biliary drainage.Ann Surg 1988; 207: 26-32.

7 Sigwart U, Puel J, Mirkovitch V, Joffre F, Kappenberger L.Intravascular stents to prevent reclusion and restenosis aftertransluminal angioplasty. NEnglJMed 1987; 316: 701-6.

8 Milroy E, Chapple C, El Din A, Wallsten H. A new stent forthe treatment of urethral strictures. Preliminary report.BrJ Urol 1989; 63: 392-6.

9 Gillams A, Dick R, Dooley J, Walisten H. Self expandablestainless steel braided endoprosthesis for biliary strictures.Radiology 1990; 174: 137-40.

10 Neuhaus H, Hagenmuller F, Griebel M, Classen M. Selfexpanding metal stents versus conventional plastic endo-prostheses for malignant biliary obstruction [Abstract].GastrointestEndosc 1991; 37: 253.

11 Cremer M, Deviere J, Sugai B, Baize M. Expandable biliarymetal stents for malignancies: endoscopic insertion anddiathermic cleaning for tumor ingrowth. Gastrointest Endosc1990; 36:451-7.

12 Huibregtse K, Cheng J, Coene PPL, Tytgat GNJ. Endoscopicplacement of expandable metal stents for biliary strictures.A preliminary report on experience with 33 patients.Endoscopy 1989; 21: 280-5.

13 Delhaye M, Vandermeeren A, Baize M, Cremer M. Extra-corporeal shock wave lithotripsy of pancreatic calculi.Gastroenterology 1992; 102: 610-20.

14 Foerster EC, Hoepffner N, Domschke W. Bridging of benigncholedocal stenoses by endoscopic retrograde implantationofmesh stents. Endoscopy 1991; 23: 133-5.

15 Maccioni F, Rossi M, Salvatori FP, Ricci P, Bezzi M, Rossi P.Metallic stents in benign biliary strictures: three year follow-up. Cardiovasc InterventRadiol 1992; 15: 360-8.

16 Fekete F, Gayet B, Place S, Biagrin J. Four hundredoesophageal anastomoses with a stapler. Int Surg 1989; 74:69-72.

17 Alvarado R, Palmaz JC, Garcia OJ, Tio FO, Rees CR.Evaluation of polymer coated balloon expandable stents inbile ducts. Radiology 1989; 170: 975-8.

on March 3, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.35.1.122 on 1 January 1994. D

ownloaded from