Gut 1997; 40: 272-276

Endoscopic management of Mirizzi's syndrome

R E England, D F Martin

AbstractBackground-The accepted managementof Mirizzi's syndrome is surgical, butendoscopic and percutaneous manage-ment have been described.Aim-To review our experience of endos-copic intervention for Mirizzi's syndrome.Patients and Methods-ERCP reports ofpatients presenting for endoscopic man-agement of choledocholithiasis betweenMarch 1989 and June 1995 were reviewed.Those with cholangiographic evidence ofMirizzi's syndrome were selected forstudy. Patient records and cholangio-grams were reviewed and follow up wasrecorded from the notes or by telephonecontact with patients, their relatives, ordoctors.Results-Twenty five patients hadMirizzi's syndrome. Sixteen were femaleand their median age was 67 years (range28-91). Ten presented with painlessjaundice, nine with painfil jaundice, fourwith cholangitis, and two had pain as theironly symptom. Twelve were referred forsurgery and 11 of these had preliminaryendoscopic therapy. Thirteen have beentreated solely with endoscopic therapy.Treatment in this group was aimed atrelieving jaundice and removing stones.Stones were completely removed in threepatients. Nine patients have been treatedwith long term stents, and one awaitsextracorporeal shockwave lithotripsy ofthe gall bladder. Complications of treat-ment occurred in four of 25 after ERCP.Conclusions-Endoscopic treatment ofMirizzi's syndrome is effective as atemporising measure before surgery andcan be definitive treatment for unsuitablesurgical candidates.(Gut 1997; 40: 272-276)

Keywords: Mirizzi's syndrome, ERCP, endoscopictherapy.

tectomy. ' 2The established treautent is surgicalbut with the development of endoscopic andpercutaneous biliary techniques non-surgicaltherapy has been described. We describe ourexperience of Mirizzi's syndrome, emphasisingthe role ofERCP in diagnosis and treatment.

MethodsWe searched our database for all patients witha cholangiographic diagnosis of Mirizzi's syn-drome between March 1989 and June 1995and reviewed their notes, and retrograde chol-angiograms. Demographic details, presen-tation, clinical and cholangiographic findings,endoscopic and surgical treatment, and com-plications of treatment were recorded. Patientswere followed up by contacting their primaryphysician or referring doctor. Follow up wasjudged complete if the patient had cholecys-tectomy, cholangiography showed clear ducts,and they were discharged from review, or if thepatient died.

Department ofRadiology, SouthManchester UniversityHospitals NHS TrustWithingtonR E EnglandD F MartinCorrespondence to:Dr D F Martin,Department of Radiology,South Manchester UniversityHospitals, NHS TrustWithington, Nell Lane,M20 2LR.Accepted for publication6 September 1996

Mirizzi's syndrome was described in 1948 andconsists of a stone impacted in the neck of thegall bladder or in the cystic duct, causing in-flammation and extrinsic compression of thecommon duct, leading to obstructive jaundice.Mirizzi's syndrome may be subdivided intotypes I and II: in the first there is extrinsiccompression of the common duct by an in-flamed or enlarged gall bladder (Fig 1) while intype II the stone erodes from the gall bladderor the cystic duct into the common duct result-ing in a cholecystcholedochal fistula (Fig 2).The syndrome is rare occurring in only0-7-110% of patients undergoing cholecys-

Figure 1: Retrograde cholangiogram showing extrinsiccompression of the common duct with dilatation of thebiliary tree upstream - Mirizzi type I.

272

on 16 July 2019 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.40.2.272 on 1 February 1997. D

ownloaded from

Endoscopic management ofMirizzi's syndrome

Figure 2: Mirizzi type II. Cholangiography shows contrastin the cystic duct and stones in the gall bladder. One stone is

eroding into the common hepatic duct resulting in a

cholecystcholedochalfistula.

ResultsTwenty five patients had a cholangiographicdiagnosis of Mirizzi's syndrome. Sixteen were

female and the median age was 67 years (range28-91). Ten presented with painless jaundice,nine with pain and jaundice, four with cholan-gitis, and two had pain as their only symptom.None presented with a clinical diagnosis ofacute cholecystitis and in only one was thediagnosis of Mirizzi syndrome suspected on

preliminary sonography. At ERCP 16 hadMirizzi type I, with a stone impacted in theneck of the gall bladder (7) or cystic duct (9).

Stone size (corrected for magnification) was a

median of 1-5 cm (range 0-75-2 25). Ninepatients had Mirizzi type II with a biliary fistulabetween the wall of the gall bladder and thecommon duct (5), or the cystic duct and thecommon duct (4). Stone size was larger,median 2-25 cm (range 1.12-5-25). Therewere coincidental common bile duct stones inonly one patient.Where patients were medically fit they were

referred for definitive surgical treatment (12 of25), and their clinical details are given inTable I. Thirteen of 25 were treated solely withendoscopic therapy, and their details are givenin Table II. Most of this group were elderly or

unfit for surgery, or both.Of the 12 patients referred for surgical ther-

apy 11 had preliminary endoscopic therapy.Ten had Mirizzi type I. One patient (case 8),presenting with pain alone, had Mirizzi type Idiagnosed on retrograde cholangiography withextrinsic compression of the common hepaticduct by a distended gall bladder with milddilatation of the intrahepatic biliary radicals. Asthe patient had not developed jaundice by thetime of the initial ERCP no endoscopic drain-age was performed and she was referreddirectly for surgery. She remained well (with-out jaundice) for the subsequent year before an

elective cholecystectomy. Her inclusion in thisseries was on the basis of her cholangiographicand subsequent surgical findings: she requiredopen conversion of a laparoscopic procedurebecause of two impacted stones in Hartmann'spouch. Her lack of jaundice despite thecholangiographic findings must be explainedby incomplete obstruction of the biliary tree. Intwo patients (cases 9 and 25) endoscopicsphincterotomy was the only treatment: case

25 had previously had sphincterotomy forF common bile duct stones and presented with

pain alone and was referred for surgery. In case

9 attempts to disimpact the stone from thecystic duct failed and he was referred forcholecystectomy during the same admission.The remaining patients had biliary drainage

TABLE I Clinical details of the surgically treated group

Patient Sex Age (y) Presentation Mirizzi type Endoscopic treatment Surgery Comment

4 F 37 Jaundice+pain Type I ES, cystic duct disimpacted, ESWL LC at 15 days Stone had been fragmented. Alive and wellof gall bladder

8 F 65 Pain Type I None LC converted to OC Surgery recommended. Alive and wellat 365 days

9 M 56 Jaundice+pain Type I ES OC+CBD Surgery recommended. Alive and wellexploration at 3 days

11 F 60 Jaundice Type I ES, CBD stent OC at 116 days Invastive carcinoma of gall bladder onhistology and died from this 388 days aftercholecystectomy

12 F 28 Jaundice+pain Type I ES, CBD stent OC+CBD Alive and wellexploration at 1 day

14 M 65 Cholangitis Type II ES, CBD stent, mechanical OC+CBD No response to endoscopic treatment.lithotripsy exploration 7 days Alive and well after surgery

15 M 70 Jaundice Type I CBD stent OC at 120 days Polya gastrectomy. Alive and well17 F 28 Jaundice+pain Type I Cystic duct disimpacted, cystic duct OC at 63 days ES not performed as young and wanted to

stent preserve sphincter. Alive and well20 M 59 Cholangitis+renal Type I ES, CBD stent OC at 165 days Surgery postponed till renal failure

failure resolved. Alive and well22 F 71 Cholangitis Type II ES, CBD stents, mechanical Laparotomy at 11 Coincident pseudoaneurysm of cystic

lithotripsy days. Unable to artery with haemobilia. Hepaticremove gall bladder embolisation performed, bleeding stopped,

but died with multi-organ failure at 18 days23 F 58 Jaundice+pain Type I CBD stent OC at 9 days No improvement with stent, developed

acute cholecystitis. Alive and well aftersurgery

25 M 62 Pain Type I ES done previously for CBD stone. Awaits surgery Alive and wellNo endoscopic treatment at this time

ESWL=extracorporeal shockwave lithotripsy, CBD=common bile duct, OC=open cholecystectomy, LC=laparoscopic cholecystectomy.

273

on 16 July 2019 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.40.2.272 on 1 February 1997. D

ownloaded from

England, Martin

TABLE II Clinical details of the patients managed solely by endoscopic therapy

CumulativeAge Coexistent Minzzi Endoscopic No of Duration of inpatient

Patient Sex (y) Presentation Illlness type treatment ERCPs treatment days Outcome

1 F 88 Jaundice+ Cardiac failure, Type II ES, Mechanical 6 4 years 7 13 Long term stenting. Died ofpain Hypertension lithotripsy, CBD stents months CCF, IHD, CVA, non-biliary

causes2 M 39 Jaundice Angina, MI, Type I ES, ESWL GB, MTBE, 2 20 days 20 Ducts clear. Alive with no biliary

epilepsy, CVA, Mechanical lithotripsy problems at 6 years 1 monthPulmonary follow upembolism

3 F 84 Jaundice Ischaemic heart Type I ES, cystic duct 3 5 years 2 33 Long term stenting. Alive anddisease disimpacted, Mechanical months well, no biliary problems

lithotripsy, CBD stents5 F 78 Jaundice Diabetes Type II ES, Mechanical 5 2 years 7 19 Long term stenting. Died with

lithotripsy, ESWL, CBD months secondary biliary cirrhosisstents

6 F 63 Jaundice Type II ES, Mechanical 7 2 years 10 16 Ducts clear. Not anxious forlithotripsy, ESWL, CBD months surgery. Alive and well at 4 yearsstents and 5 months follow up

7 F 81 Cholangitis Cardiac failure Type II ES, CBD stent 1 10 days 10 Died of bronchopneumonia 10CVA, Hx of days, jaundice was resolvingPE+DVT

10 F 91 Jaundice CVA, atrial Type II ES, CBD stent, GB stent 3 2 years 9 6 Long term stenting. Alive andfibrillation months well at 2 years 9 months follow

up13 F 73 Jaundice+ Asthma, chronic Type I CBD stent 1 10 months 2 Died of endometrial carcinoma

pain anaemia, 2 weeks at 10 months. Non-biliary deathendometrial Ca

16 F 69 Jaundice Ischaemic heart Type I ES, CBD stent 1 1 1 months 18 Alive and well. Long termdisease, CVA stenting. Not anxious for surgery

18 F 70 Jaundice Type I ES, cystic duct 2 7 months 4 Alive and well. Long termdisimpaction, CBD stent stenting. Not anxious for surgery

19 M 78 Jaundice+ COAD, Type II ES, CBD stent, URSO 1 4 months 2 Alive and well. Long termpain hypertension stenting

21 M 80 Jaundice+ Paraplegia Type I ES, GB decompression, 2 10 days 10 Discharged with clear ducts at 10pain ESWL days. Died at 10 months

non-biliary causes24 M 69 Jaundice Type II ES, CBD stent, GB stent 1 1 month 2 Alive and well, awaiting ESWL

MI=myocardial infarction, CVA=cardiovascular accident, DVT=deep venous thrombosis, COAD=chronic obstructive airway disease, MTBE=methyl tert butylether, URSO=ursodeoxycholic acid, Hx of PE=pulmonary embolus, CCF=congestive cardiac failure, IHD=ischaemic heart disease.

with plastic stents (8) and sphincterotomy (6),extracorporeal shockwave lithotripsy of thegall bladder (1), and adjuvant mechanicallithotripsy (2). Sphincterotomy was avoided incase 17 as she was young, and was notperformed for technical reasons in case 15 asthe patient had a Polya gastrectomy. Case 23had stenting without sphincterotomy as theinitial diagnosis was thought to be a malignantstricture; however, on review of her imagingMirizzi's syndrome was diagnosed. Treatmentin this group was aimed at relieving jaundiceand allowing elective planned cholecystec-tomy. Surgery was performed at a median timeof 15 days after ERCP (range 1-365 days).Three patients failed to improve after endos-

copic therapy: case 23 developed acute chole-cystitis, case 14 had persistent cholangitis, andcase 22 was operated on for continuedhaemobilia but surgery failed to identify thesource and cholecystectomy was not possiblebecause of severe inflammation of Calot'striangle. This patient subsequently had trans-catheter embolisation of a pseudoaneurysm ofthe cystic artery but although this controlledher haemorrhage she succumbed a week laterto multiorgan failure. All but one of thepatients who had surgery required an openprocedure.

Thirteen patients had been treated solelywith endoscopic therapy (Table II). Treatmentwas aimed at relieving jaundice and removingstones. Those in whom complete clearance wasnot achieved were treated with long termstenting. The stones were completely removedin three patients (cases 2, 6, and 21) using acombination of sphincterotomy, mechanicallithotripsy, methyl tert butyl ether, extra-

corporeal shockwave lithotripsy, and commonduct bile stents. Nine patients were treatedwith long term stenting and of these five arealive and well at follow up ranging from fourmonths to five years and two months. Fourpatients in this group died, two of non-biliarycauses, another from bronchopneumonia 10days after ERCP, and one of secondary biliarycirrhosis at two years and seven months. Theremaining patient in this group is awaitingextracorporeal shockwave lithotripsy of the gallbladder.

Complications of endoscopic treatmentoccurred in four of 25 patients: (case 7)bronchopneumonia after ERCP, (case 23)acute cholecystitis after stenting, (case 1 1) liverabscess treated conservatively with antibiotics,and (case 5) secondary biliary cirrhosis.The overall mortality of the group was seven

of 25, with a 30 day mortality of two of 25.Four of seven deaths were biliary related: (case5) secondary biliary cirrhosis; (case 7) broncho-pneumonia after ERCP; (case 1 1) carcinoma ofgall bladder; (case 22) multiorgan failuresecondary to prolonged haemobilia from apseudoaneurysm of the cystic artery.

DiscussionMirizzi syndrome, a rare complication ofgallstone disease, is increasingly recognisedwith the widespread availability of cholan-giography. In the past the diagnosis was oftenfirst made at laparotomy, and because of thesevere inflammation in Calot's triangle asso-ciated with a shrunken gall bladder, the riskof bile duct injury was high. The syndrome,as originally described, comprised a stone

274

on 16 July 2019 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.40.2.272 on 1 February 1997. D

ownloaded from

Endoscopic management ofMinizzi's syndrome

impacted in the cystic duct or neck of the gallbladder and a functional disorder of a putativesphincter within the common hepatic duct,without any lesion of the common bile duct.3Inflammation, aberrant vessels or an impactedcystic duct stone were thought to stimulatecontraction of this 'sphincter' resulting in ob-structive jaundice. It is now known that nosuch 'sphincter' exists and that biliary obstruc-tion results from extrinsic compression of thecommon duct by the inflamed gall bladder andthat the inflammatory process may result incommon duct disruption.

In 1982 McSherry4 classified Mirizzi's syn-drome into type I and type II based oncholangiographic findings at ERCP. Type I ispresent when there is extrinsic compression ofthe common duct by a stone in the gall bladderor in the cystic duct (Fig 1). Type II is presentwhen a cholecystcholedochal fistula exists(Fig 2). Type II has been further subclassifiedby Csendes et al5 into three types depending onthe size of the defect in the common duct -type II: erosion of less than one third thecircumference of the bile duct; type III: erosionof up to two thirds of the common duct; typeIV: complete destruction of the entire wall ofthe common duct. The size of the fistula inCsendes' series was determined at peroperativecholangiography. In our experience exactquantification of fistula size at pre-operativecholangiography is impossible, and we there-fore use McSherry's classification. The import-ance of recognising type II Mirizzi is stressedas the operative approach differs in thepresence of a cholecystcholedochal fistula. Anopen surgical approach is considered moreappropriate than a laparoscopic approach if afistula is present as dissection may be difficultand a direct common duct repair or bilioentericanastomosis is necessary.5 6 Indeed an openprocedure is often necessary for Mirizzi's syn-drome irrespective oftype due to severe inflam-mation and this is reflected in our series whereall but one of our patients who had surgeryrequired an open procedure.There are a number of published reports of

non-surgical management of Mirizzi's syn-drome. Percutaneous management is usuallyreserved for patients who have failed endos-copic access and who are unsuitable surgicalcandidates. Cairns7 and Oxtoby8 describepercutaneous cholecystodocholithotomy andelectrohydraulic lithotripsy, performed underlight general anaesthesia, in patients withMirrizzi type I. Adam9 used a percutaneousmetal stent to decompress the biliary tree in anelderly patient with Mirizzi syndrome, and leftthe gall stones in situ. This patient was well atfollow up at 2-5 years.Endoscopic management of Mirizzi syn-

drome comprises biliary drainage of thecommon duct with or without gall bladderdrainage, and attempted gall stone removalwith either basket or balloon, mechanical,electrohydraulic or extracorporeal lithotripsyor dissolution therapy. Endoscopic sphinctero-tomy is followed by either placement of anasobiliary catheter or more commonlyinsertion of an indwelling prosthesis into the

common bile duct, cystic duct or both.' '2After resolution of jaundice or cholangitisattempts are made to clear the stones whensurgery is not deemed appropriate. The largestpublished series of solely endoscopic manage-ment of Mirizzi syndrome is by Binmoelleret al. 3 In this series electrohydraulic lithotripsywas administered at ERCP under cholangio-scopic guidance to 14 patients with Mirizzi'ssyndrome. In all patients the stone wasimpacted distally in the cystic duct and stonesize ranged from 1-5 cm to greater than 3 cm.Stones were cleared in all 14 patients with onlyone complication: there was leakage of contrastfrom the duct system into the peritoneal cavityafter removal of a large 3 5 cm stone. This leakwas caused by pressure necrosis from the largestone rather than injury to the duct by theelectrohydraulic lithotripsy probe and resolvedspontaneously with conservative management.The authors conclude that this is a safe andeffective alternative to surgery. Limitations ofthis technique include the need for con-siderable expertise at ERCP, cost of equipmentincluding mother and baby scope system, andelectrohydraulic lithoprobe per patient treat-ment session and the procedure is time con-suming. In addition stones must be accessibleto the electrohydraulic lithoprobe and henceMirizzi's syndrome patients with a stoneimpacted in Hartmann's pouch are not suitablefor this treatment (12 of 25 of our patients).

In our series almost half had definitivesurgery after endoscopic therapy. As can beseen from Table I those who were referred forsurgery were generally fit patients or patientswho did not initially respond to endoscopicdrainage (cases 14, 22, and 23). The import-ance of preliminary cholangiography should beemphasised: the diagnosis was only suspectedin one of 25 patients after ultrasound. Chol-angiography permits diagnosis, classification oftype, and provides a roadmap for subsequentsurgery. If ERCP is performed attempts canalso be made to dislodge or remove theobstructing calculus at the initial diagnosticprocedure. Interestingly in our series only two(case 14 and 22) of the surgically treated grouphad type II Mirizzi syndrome. Case 14 had aninitial ERCP with sphincterotomy and mech-anical lithotripsy and partial removal of theobstructing calculus. A stent was left for tem-porary drainage; however, the patient had per-sistent cholangitis and was referred for chole-cystectomy and common duct exploration.Surgery was difficult as anatomical landmarkswere very distorted and there was a large fistulabetween the shrunken gall bladder and thecommon duct with no identifiable cystic duct.During dissection a hole was made accidentallyin the right main duct. The gall bladder wasopened, stone debris was removed from theduct with the endoscopic stent. The gallbladder was resected and the fistula wasdirectly repaired over a T tube. A second Ttube was placed across the right hepatic ductdefect. The T tubes were removed after asubsequent cholangiogram, which confirmedclear ducts and no biliary leak. Case 22 was apatient who at ERCP for cholangitis was found

275

on 16 July 2019 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.40.2.272 on 1 February 1997. D

ownloaded from

England, Martin276

to have haemobilia. Sphincterotomy was per-

formed and attempts to remove the large 3 cmcalculus with mechanical lithotripsy failed.Two large bore stents were placed for tempor-ary drainage; however, a laparotomy was per-

formed for continued haemorrhage. Surgeryproved difficult because of anatomical dis-tortion and bleeding and a cholecystectomyand stone removal were not possible. A T tubewas placed in the common duct for drainage.Subsequent selective hepatic artery embol-isation was performed of a pseudoaneurysm ofthe cystic artery and although this controlledher haemorrhage she succumbed to multi-organ failure. Surgery for Mirizzi type II isdifficult and may be hazardous and endoscopicmanagement may be preferable in this group

as one can avoid the difficult surgical repair ofthe cholecystcholedochal fistula. Seven of ninepatients in this series with Mirizzi type II weretreated endoscopically.The timing between preliminary endoscopic

treatment and surgery was very variable in our

series (Table I) and reflected patient response

to initial biliary drainage as well as surgicalwaiting lists for elective cholecystectomy. Pro-vided the patient's symptoms are relieved withinitial endoscopic treatment definitive surgicaltreatment can be postponed until the patient isfit for elective surgery. In patients unsuitable or

unwilling to undergo surgery further endos-copic attempts are made to disimpact andremove the stones, using a combination ofmechanical lithotripsy, dissolution agents or

extracorporeal shock wave lithotripsy. We were

successful in doing this in three patients, cases

2 and 21 who were medically unfit and case 6who refused surgery.There were nine patients who had long term

stenting. The median age of this group was 78years, range 69-91, and all but one (case 18)had coexistent medical problems making themunsuitable surgical candidates (Table II).Bergman et al'4 recently reported the outcomeof long term stenting in 58 of 117 patients forirretrievable common duct stones and showed a

40% complication rate, most commonly cholan-gitis, and nine of 44 patients had a biliary relateddeath. The authors conclude that permanentbiliary stenting (that is, stent exchange only forrecurrent symptoms) should be restricted to

patients unfit for surgery or further endoscopictherapy and those with a short life expectancyand we agree with this policy. In our unit we

change long term stents at yearly intervals butwe review patients if they have recurrentsymptoms between stent exchanges. Of the ninepatients with long term stenting five are aliveand well. Of the four who died there was one

biliary death (case 5) from secondary biliarycirrhosis. This patient was initially referred forERCP with suspected primary biliary cirrhosisand Mirizzi's syndrome was diagnosed. She was

a diabetic and not anxious for surgery. She hadfive ERCPs with mechanical lithotripsy andextracorporeal shock wave lithotripsy, howeverher stone remained. She was asymptomatic withstenting until finally presenting at two years and

seven months with ascites, and oesophagealvarices, but without jaundice. She succumbedafter gastrointestinal haemorrhage. In retrospectthis patient may have had a better outcome hadshe agreed to surgery. Complications of treat-ment occurred in a further three patients:bronchopneumonia resulting in the demise ofone patient at 10 days; acute cholecystitis afterstenting treated surgically; and a liver abscessthat was treated conservatively. The latterpatient (case 11) had an open cholecystectomyat 1 16 days and histological examination of hergall bladder revealed an unsuspected gallbladder carcinoma. Coexistence of gall bladdercarcinoma and Mirizzi's syndrome has beenrecognised previously6 15 and is not surprising asboth conditions are associated with long termgall stone disease.

In conclusion ERCP plays an important partin diagnosis and provides not only a roadmapfor surgery but permits therapeutic decom-pression of the biliary tree as a temporisingmeasure before surgery. In some patientsendoscopic clearance of stones is possible, andlong term stenting for those in whom clearanceis not achieved is an alternative, effectiveoption to surgery - with the caveat that thesepatients should be followed up at regularintervals to prevent complications. Finallythose patients with Mirizzi type II may be moresuited to endoscopic therapy as this can avoidwhat is often very difficult open surgery.

This work has previously been presented in abstract form at theBritish Society of Gastroenterology Spring meeting 20-22March 1996 (Gut 1996; 38: A55).

1 Blumgart LH. Surgery of the liver and biliary tract.Edinburgh: Churchill Livingstone, 1988.

2 Bower TC, Nagomey DM. Minzzi syndrome. HPB Surg1988; 1: 67-76.

3 Mirizzi PL. Syndrome del conducto hepatico. J7ournalInternational du Chirurgie 1948; 8: 731-77.

4 McSherry CK, Ferstenberg H, Virshup M. The Mirizzisyndrome: suggested classification and surgical therapy.Surg Gastroenterol 1982; 1: 219-25.

5 Csendes A, Carlos Diaz J, Burdiles P, Maluenda F, Nava 0.Mirizzi syndrome and cholecystobiliary fistula: a unifyingclassification. BrJ Surg 1989; 76: 1139-43.

6 Baer HU, Matthews JB, Schweizer WP, Gertsch P,Blumgart LH. Management of the Mirizzi syndrome andthe surgical implications of cholecystcholedochal fistula.BrJ'Surg 1990; 77: 743-5.

7 Cairns SR, Watson GN, Lees WR, Salmon PR. Percu-taneous lithotripsy and endoprosthesis: a new treatmentfor obstructive jaundice in Minzzi's syndrome. BMJ1987; 295: 1448.

8 Oxtoby JW, Yeong CC, West DJ. Mirizzi syndrome treatedby percutaneous stone removal. Cardiovasc InterventRadiol 1994; 17: 207-9.

9 Adam A, Roddie ME, Benjamin IS. Case Report: Mirizzisyndrome - treatment with metallic endoprosthesis. ClinRadiol 1993; 48: 198-201.

10 Vakil N, Sawyer R. Endoscopic drainage of the gallbladderin a septic variant of the Mirizzi syndrome. GastrointestEndosc 1994; 40: 236-8.

11 Delcenserie R, Joly JP, Dupas JL. Endoscopic diagnosis andtreatment of Mirizzi's syndrome. J Clin Gastroenterol1992; 15: 343-6.

12 Martin DF, Tweedle DEF, Rao PN. Endoscopic gallbladdercatheterisation and extracorporeal shockwave lithotripsyin the management of Mirizzi's syndrome. Endoscopy1988; 20: 321-2.

13 Binmoeller KF, Thonke F, Soehendra N. Endoscopic treat-ment of Mirizzi's syndrome. Gastrointest Endosc 1993; 39:532-6.

14 Bergman JJGHM, Rauws EAJ, Tijssen JGP, Tytgat GNJ,Huibregtse K. Biliary endoprostheses in elderly patientswith endoscopically irretrievable common bile ductstones: report on 117 patients. Gastrointest Endosc 1995;42: 195-201.

15 De Bakshi S, Das S, Sengupta A. Mirizzi syndrome in apatient with carcinoma of the gallbladder. Br3 Surg 1992;79: 371-2.

on 16 July 2019 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.40.2.272 on 1 February 1997. D

ownloaded from