- Email: [email protected]
Endoscopic retrograde cannulation of the gallbladder: direct dissolution of gallstones
()()16-5107/90/3605-0444$02.00 GASTROINTESTINAL ENDOSCOPY Copyright © 1990 by the American Society for Gastrointestinal Endoscopy
Endoscopic retrograde cannulation of the gallbladder: direct dissolution of gallstones Ernst Christoph Foerster, MD, PhD Werner Matek, MD Wolfram Domschke, MD Erlangen, Federal Republic of Germany
Percutaneous transhepatic catheterization of the gallbladder for dissolution of cholesterol stones by instillation of methyl tert-butyl ether (MTBE) is an invasive therapeutic procedure. The only non-invasive alternative available to now, endoscopic retrograde cannulation of the cystic duct, was difficult because of the cystic duct's tortuosity and spiral valves. We therefore developed a catheter system which, using conventional duodenoscopes during a routine endoscopic retrograde cholangiography (ERC) procedure, permits reliable and safe catheterization of the gallbladder without the need for endoscopic sphincterotomy. In 18 of 22 patients (82%) we were able to place a cysto-nasal catheter, and in 14 patients MTBE dissolution therapy was then performed. Eight patients (57%) were completely free of stones after treatment; the other six (43%) had residual debris. In 4 of 22 patients (18%) cannulation attempts failed, in 3 patients due to cystic duct blockage by a calculus. Endoscopic retrograde cannulation of the gallbladder (ERCG) represents a promising alternative to the invasive percutaneous transhepatic catheterization procedure. (Gastrointest Endosc 1990;36:444-450)
After publication of in vitro studies suggesting that the cholesterol solvent methyl tert-butyl ether (MTBE) dissolved the majority of cholesterol stones leaving little or no residue,! Allen et a1. 2 attempted dissolving gallbladder stones with MTBE administered via percutaneous transhepatic gallbladder catheterization. MTBE is liquid at body temperature and dissolves cholesterol stones within a few hours. 3 By 1989, the same Mayo Clinic group had instilled MTBE into the gallbladders of 75 patients; in 72 patients there was complete dissolution of stones or more than 95% dissolution. 4 In one case, however, bile leakage from the catheter track occurred, requiring intervention. Several other patients had transient pain immediately after removal of the catheter, suggesting minor leakage events. As an alternative to this invasive transhepatic ap-
proach, we have developed a new catheter system (U.S. patent no. 4,905,667; Schneider Europe AG, Zurich, Switzerland) providing reliable endoscopic retrograde cannulation of the gallbladder. 5 Another recently reported technique by Ponchon et a1. 6 permits successful catheterization only in exceptional cases where a favorable anatomical situation is present. Using our catheter system, we initially performed studies in autopsy specimens5 and preliminary reports of our early experience with patients have been published. 7,8 The present study describes the efficacy and safety of retrograde cannulation of the gallbladder in 22 patients, and direct chemolitholysis of symptomatic radiolucent gallstones with MTBE administered through a cysto-nasal catheter in 14 patients.
Received February 26, 1990. For revision April 18, 1990. Accepted July 2, 1990. From the Department of Medicine I, University of Erlangen-Nuremberg, Erlangen, Federal Republic of Germany. Reprint requests: E. C. Foerster, MD, Medizinische Klinik mit Poliklinik I der Universitiit Erlangen-Nurnberg, Krankenhausstr. 12, D-8520 Erlangen, Federal Republic of Germany.
Patients
444
PATIENTS AND METHODS
Twenty-two patients, 18 women and 4 men aged 21 to 79 years (mean, 54 years) were included in the study. Solitary stones (16 and 22 mm in diameter, respectively) were present in two patients. Multiple stones ranging widely in size (8 to 35 mm) and number (2 to 10 or more) were found in the GASTROINTESTINAL ENDOSCOPY
remaining 20 patients. No patient was excluded due to the size or number of stones. All patients had suffered one or more episodes of biliary colic or pancreatitis, but none had any signs, symptoms, or laboratory evidence of acute inflammation of the gallbladder, bile ducts, or pancreas at the time of treatment. Exclusion criteria included irreversible coagulopathy, known obstruction of the cystic duct or common bile duct, and pregnancy. All participating patients had previously refused cholecystectomy or were considered at high risk for surgery. Written informed consent was obtained from all patients; approval of the study protocol was granted by the ethics committee of our hospital. Following a fatty meal, cystic duct function was examined in all patients by means of ultrasonography (US). Plain films of the abdomen or radiographs of the kidney, ureter, and bladder were not routinely obtained. Computer tomographic (CT) scans without intravenous or oral radiopaque contrast media were performed in all patients to verify that the stones were not calcified. The scans, consisting of contiguous 4-mm thick slices at 4-mm intervals, were reconstructed with a 512 2 matrix display format. Hematological studies, serum aminotransferases, alkaline phosphatase, bilirubin, creatinine and amylase or lipase levels, prothrombin time, and hemolysis screen (haptoglobin and free hemoglobin) were done before treatment (normal in all cases) and repeated after each day of MTBE treatment.
M
~::::~.:::;:,.~~:::::::;::::::::=-z:;. -:.. ' .. -:-.-.:.,::.
A
-
-
.-G. _
:::::::===============-------_._-_.
B
Endoscopic retrograde insertion of the catheter into the gallbladder
After an overnight fast, patients received intravenous diazepam (2 to 5 mg) and pethidine (50 to 100 mg) prior to endoscopy. Antibiotics were not routinely given before or during the procedure. An intravenous line was maintained to permit rapid administration of medications as needed. Once the TJF 10 duodenoscope (Olympus Optical Co., Hamburg, Federal Republic of Germany) was positioned opposite the papilla of Vater, a "thick," flexible guide wire (length 280 cm, diameter 0.039 inch = 1.00 mm) was first inserted into the common bile duct (CBD) without papillotomy (Fig. 1A), and a modified angiography catheter (Schneider Europe AG, Zurich, Switzerland) (Fig. 1B), henceforth referred to as common bile duct catheter (CBDC) (length, 175 cm; outside diameter, 7 F, i.e., 2.15 mm; inside diameter, 1.2 mm), passed over it and placed in the common bile duct. The intra- and extrahepatic bile duct system was then filled with diluted contrast medium (Conray 60, Byk Gulden, Federal Republic of Germany). The thick, flexible guide wire was then withdrawn, and the tip of the CBDC hooked into the opening of the cystic duct (Fig. 1A). A custom-made thin steerable guide wire (Fig. lC) (0.02 inch = 0.52 mm in diameter, length 400 cm), with a 6-cm long soft distal tip, was placed in a Teflon (stiffening) cystic duct catheter (inside diameter, 0.60 mm; outside diameter, 1.05 mm; length, 205 cm) in such a manner that the tip of the wire extended just to the tip of the Teflon catheter (Fig. ID). The guide wire/Teflon catheter combination was then advanced along the CDBC and manipulated past the spiral valves and along the cystic duct. Cannulation was thus performed after conventional premedication with no need for previous sphincterotomy. Placement of the CBDC was performed in such a manner VOLUME 36, NO.5, 1990
uuum
2
il••••••• ==:==:==:=:
.11111111111111111111_. . . . . .
c ::: ~ ~~:..: .: .: .::::::::::
,
:'.:'.:",:'.:",:".:",::.:'
: , .. ::::-:::<::<:<:::::':::'::-:-:'::j I III'II'II~~==========~~~~"~~,
.
...
;,;.~~~~ '
.
'.'
..
'.'
,
D 5
2
Figure 1. A, Conventional duodenoscope (4) with the tip positioned at the papilla of Vater; transpapillary cannulation with the common bile duct catheter (1) and positioning of the tip in the opening of the cystic duct (3). ''Thick'' guide wire (2) already advanced along the catheter. B, Configuration of the pre-shaped common bile duct catheter (1) in the resting state; the curves (a) and (b) are stable in the axial direction. C, ''Thin'' guide wire (2) with soft, Teflon-coated flexible tip. 0, The tip of the common bile duct catheter (1) with advanced thin guide wire (2) enclosed within the stiffening Teflon catheter (5).
that the segment between the first and second curve (Fig. lB) became obliquely anchored within the CBD, while the flexible tip came to rest within the orifice of the cystic duct (Fig. lA). The usual distance between the two curves of the CBDC (Fig. 1B) was 4.2 cm. The crucial point was to ensure that the tip of the catheter lay freely within the orifice of the cystic duct and not at right angles to it, which would block off the opening of the catheter. Catheter placement was confirmed and documented by fluoroscopy (Fig. 2). 445
Figure 2. A, Radiograph of a 79-year-old woman. Contrast medium filling extrahepatic bile duct system including cystic duct and gallbladder. B, Common bile duct catheter with tip hooked in the opening of the cystic duct. C, Advancement of the guide wire/ catheter system. D, Guide wire positioned within the gallbladder.
After placing the guide wire in the gallbladder, the Teflon catheter was withdrawn and a specially designed polyethylene MTBE instillation catheter (outside diameter, 5 F, i.e., 1.55 mm; inside diameter, 1.0 mm; 360-cm long) with a 2to 3-cm diameter double pigtail tip containing multiple sideholes was passed. The pigtail tip was looped in the gallbladder fundus so that it encircled as many stones as possible (Fig. 3A). We found that placing one or two loops of the catheter in the gallbladder helped to prevent dislodgment. After withdrawal of the duodenoscope, the irrigation catheter was withdrawn through the nose, reduced to an external length of approximately 50 em, and fitted with a metalluerlock connector. 446
With the catheter in place, the gallbladder contents were completely aspirated, and contrast medium diluted with 0.9% saline was instilled to determine the volume needed to fill the gallbladder without spilling over into the cystic duct. An amount of MTBE equal to half this volume was later used for MTBE treatment. During treatment, the amount of MTBE instilled was increased with progressive dissolution of the stones. Discontinuation of treatment due to untoward effects was not necessary in any patient. Treatment with MTBE
Treatment usually commenced 2 to 3 hours following catheter placement in a treatment room fitted with an GASTROINTESTINAL ENDOSCOPY
cholesterol free. The gallbladder was then thoroughly rinsed and filled with sterile saline, and the catheter removed over the guide wire under fluoroscopic control to avoid knotting it. Before discharging the patients, US of the gallbladder and endoscopy of the upper gastrointestinal tract were performed. In the six patients with subtotal dissolution of gallstones, oral urso- and chenodeoxycholic acid (6 to 8 mg/ kg body wt/day) were administered for 3 months.
RESULTS Efficacy
Figure 3. A, Seventy-nine-year-old woman. Radiolucent stones filling the gallbladder as demonstrated by contrast medium instillation via a retrograde catheter. B, Same gallbladder as demonstrated by contrast medium after 8 hours of MTBE dissolution therapy.
externally vented spark-free exhaust fan. MTBE was instilled manually using 20- to 50-ml glass syringes at initial volumes ranging from 3 to 5 ml, and gradually increased to a maximum eventual volume of 80 ml (mean, 40 ml; range, 8 to 80). During the course of treatment, the MTBE volumes to be instilled were adjusted at 1-hr intervals by catheter cholecystography. Ether was instilled and aspirated in 3- to 5-min cycles. The amount of MTBE was decreased, particularly at the beginning of treatment, if the patient complained of pain. To avoid solvent saturation, the MTBE was replaced by fresh solvent after each instillation/aspiration cycle. The catheter within the cystic duct largely prevented overflow of MTBE into the duodenum, thus avoiding adverse intestinal effects of the solvent. All MTBE volumes instilled were carefully monitored. After establishing a rate and rhythm of MTBE infusion and aspiration which was comfortable for the patient, treatment was continued by trained paramedical staff. Fluoroscopy was performed after every 1 to 2 hours of treatment to assess the progress of dissolution. MTBE therapy was performed until the evaporated aspirates became VOLUME 36, NO.5, 1990
Catheterization of the CBD via the papilla of Vater was easily performed and positioning of the tip of the CBDC in the opening of the cystic duct was possible in each case. In 18 of 22 patients (82%), the cystic duct was cannulated with the catheterization technique described (Fig. 3A). In one case, however, the middle third of the cystic duct followed an acuteangled course and contained a pocket-like sacculation which prevented passing the guide wire. In three patients, cannulation attempts proved to be unsuccessful due to cystic duct blockage by calculus; in one of these patients, the cystic duct was inadvertently perforated and a cholecystectomy was subsequently carried out. Of the 18 patients with successful cannulation, 4 underwent endoscopic inspection of the gallbladder by means of a miniscope inserted through the cannulation catheter and 14 patients were treated by MTBE instillation. The maximum diameter of the lumen of the 22 cystic ducts was 3 mm. The entire cannulation procedure from the papilla of Vater to the gallbladder required a fluoroscopy time between 3 and 25 min (mean, 360 rad·cm 2 ). The overall procedure time for catheter placement was from 8 to 45 min (mean, 30 min). Based on changes in the size of the stones as determined by catheter cholecystography, we estimated that at least 95% ofthe stone mass was dissolved (Fig. 3B) in all but four patients (22%). In these four patients, the estimated reduction in stone mass did not exceed 50%. These relative failures were related to discrete calcifications of the gallstones, expressed in the CT scan by Hounsfield units between 60 and 80. Consequently, only stones with an HU value less than 60 (average 45) were subsequently treated with MTBE. In the majority of patients, the only residual debris detected by US was not greater than 4 mm in diameter. One patient with incomplete litholysis was re-admitted to the hospital 2 weeks post-treatment after developing temperatures to 38°C and right upper quadrant pain suggestive of acute cholecystitis 24 hours previously. The day after admission, the patient underwent an uneventful cholecystectomy. The intra447
vesicular debris was analyzed, and consisted entirely of minute flakes of cholesterol-free pigment which crumbled on touch. At pre-operative ultrasonography, the pigment flakes had presented as echodense, mobile, non-shadowing reflexes. Because we were able to instill up to 80 ml of MTBE into the gallbladder via the cystic duct that was largely occluded by the irrigation catheter, without any significant spillage into the common bile duct, virtually all calculi were successfully washed by the MTBE without having to change catheter position. All MTBE volumes instilled were carefully monitored (1000 ml mean total volume; range, 200 to 2600); loss ofMTBE, probably due to absorption by the gallbladder mucosa, did not exceed 1.8%. In this manner, dissolution treatment continued for a mean period of 5.6 hours (range, 2.5 to 16). The maximum individual yield of dissolved cholesterol was 106 g (range, 1 to 106; mean, 17.8) in this series. Ordinarily, dissolution therapy was done in one session on 1 day, while only one patient needed daily treatment sessions on 3 consecutive days. The patients were monitored by means of US at 3, 6, and 12 months post-treatment. In eight patients, US obtained immediately after stone dissolution revealed no evidence of residual debris. In the six patients with residual debris (four patients with more than 50%, two patients with more than 95% reduction in stone mass), during the subsequent adjuvant treatment with ursodeoxycholic and chenodeoxycholic acid over a period of 3 months, US never revealed any detectable increase in stone mass. One patient became stone free after 3 months with this adjuvant treatment. To date, we have carried out follow-up examinations for up to 12 months; patients remain continuously well despite minor residual debris. Side effects
During cystic duct and gallbladder cannulation, patients sedated in a standard manner for endoscopic retrograde cholangiography experienced no or only minor discomfort. During MTBE litholysis, 10 mg of diazepam needed to be administered intravenously in only 1 of 14 patients. Two patients experienced transient nausea with or without mild emesis, which responded to prochlorperazine suppositories. The smell of MTBE was often detected on the patients' breath. Depending on the volume of MTBE absorbed, mild transient sedation was also occasionally apparent. Upon termination of treatment, sedation and expired ether usually resolved within 30 min. There was no biochemical evidence of intravascular hemolysis in any patient. No changes occurred in serum haptoglobin concentration or free plasma hemoglobin; there were no significant changes in serum 448
hemoglobin concentration (13.6 g/dl before and 13.5 after MTBE treatment). In two patients, the aminotransferase activity was elevated (mean maximal level, 54 units/liter; normal 6 to 28). Serum bilirubin concentration rose transiently to above 26 JLmol/liter (normal <17) in two other patients. None of the 14 patients showed more than two-fold increase in serum alkaline phosphatase. As occasionally observed in ERCP procedures, an increase in serum amylase to a mean of 160 JLmol/ml (range, 140 to 340, normal <100) was found in seven patients. In six patients, total leukocyte count increased transiently (mean maximum, 14,000; range, 12,000 to 19,000). All abnormal laboratory values returned to normal within 2 days post-treatment. No parenteral analgesics were required to control abdominal pain after catheter removal; none of the patients developed fever. Subsequent endoscopy with biopsy showed no macroscopic or histological signs of duodenitis. None of the patients developed bile duct obstruction, cholangitis, or pancreatitis. Mean duration of hospitalization was 3 days (range, 2 to 5). DISCUSSION
The introduction of MTBE as a potent solvent of cholesterol stones opened up new possibilities in the treatment of gallbladder stones. Extensive pharmacological and toxicological studies, together with direct percutaneous transhepatic dissolution therapy of gallbladder stones in more than 400 patients throughout the world, have shown that MTBE has no demonstrable local toxic effects on the mucosa of the gallbladder. The percutaneous transhepatic approach to the gallbladder, however, is associated with a number of basic problems. 4 Although a 5 French polyethylene catheter can be introduced under radiological control into the gallbladder transhepatically in 100% of patients, contrast medium instillation via the catheter has been associated with extravasation from the site of the puncture. An additional problem associated with this invasive approach is the fact that spillage of the MTBE via the cystic duct into the duodenum must be carefully avoided. In the duodenum, in contrast to the gallbladder, MTBE is well absorbed, which can produce severe adverse reactions, such as intravascular hemolysis. Endoscopic inspection occasionally reveals acute duodenitis following treatment with MTBE. Ponchon et al. 9 have recently published a report of a patient who developed acute renal failure requiring dialysis following intravascular hemolysis arising from massive extravasation of MTBE during percutaneous transhepatic dissolution therapy of gallbladder stones. Marked changes in the duodenal mucosa following the absorption of MTBE have also been described in GASTROINTESTINAL ENDOSCOPY
MTBE treatment of common bile duct stones. 10 Attempts have been made to reduce the undesirable spillage of MTBE into the duodenum by minimizing the amount of MTBE instilled and placing the pigtail catheter as close to the stone(s) as possible. A further problem inherent in the invasive procedure of percutaneous transhepatic puncture of the gallbladder is bile leakage from the catheter track into the abdominal cavity upon removal of the catheter. Large leakages of this type necessitate corrective gallbladder puncture or surgical intervention. Patients' complaint of pain immediately on removal of the irrigation catheter is suggestive of minor leakage episodes. Attempts have been Made to plug biliary fistulas remaining after removal of the catheter with gelatin sponge, but results have not yet been published. All of these problems associated with the percutaneous transhepatic approach to the gallbladder prompted us to develop an improvement to the endoscopic retrograde approach. Selective catheterization of the cystic duct and the gallbladder during endoscopic retrograde cholangiography has to date been successful in less than 10%, or only in the presence of a favorable anatomical situation, or after post-inflammatory destruction of the spiral valves. Frimberger et al. 11 were the first to report on routine attempts to cannulate the gallbladder with the aid of a special short-channel duodenoscope; with this relatively complicated system, however, reliable cannulation of the gallbladder has so far not proven possible. Ponchon et al.~ have recently published a report on successful retrograde placement of a nasobiliary 6.5 French Teflon drain in two of five patients. However, all patients, including those in whom the attempt was unsuccessful, experienced marked pain; all had a prior papillotomy. Because ofthe spiral valves, negotiation ofthe cystic duct is technically the most difficult part of retrograde gallbladder cannulation. Although Kozarek 12 succeeded in entering the opening of the cystic duct in 70% of a series of 50 patients, cannulation of the gallbladder itself was only rarely possible. Using our newly developed catheter system, we succeeded in safely, and for the most part painlessly, cannulating the gallbladder in 18 of22 patients (82%). In extensive preliminary studies on autopsy specimens, we were able to tailor our catheter system to the particular anatomical situation. Morphological studies performed by Nuboer l3 showed that the cystic duct has an average diameter of 2 to 3 mm; a diameter of less than 1 mm has been observed only rarely. In the present study, the spiral valve was responsible for only one of the four (18%) unsuccessful attempts at cannulation. In the other three cases, an incarcerated calculus in the cystic duct was responsible for failure; inadequate contraction of the gallbladder following a VOLUME 36, NO.5, 1990
fatty meal observed at US had already raised the suspicion of a cystic duct calculus, but at least partial patency of the duct had been expected. The single case of perforation occurring during retrograde cannulation of the cystic duct resulted from accidentally advancing the common bile duct catheter beyond the opening of the duct and impinging on an area of wall weakened by inflammation in the region of the stone. A suspected cystic duct calculus should therefore represent an exclusion criterion for retrograde chemolitholysis. Because the gallbladder catheter we employ partially occludes the cystic duct, compared with the percutaneous transhepatic system, a considerably greater irrigation volume (up to 80 ml vs. 7 to 15 ml) can be instilled without any major spillage of MTBE into the duodenum. Accordingly, the duration of dissolution treatment with the retrograde catheter seems to be appreciably shorter (5.6 hours) than with the percutaneous transhepatic approach (12.5 hours). Clearly, a more definite comparison between the endoscopic retrograde and the percutaneous transhepatic access routes cannot be made without a carefully matched randomized study. Problems were neither expected nor observed during withdrawal of the retrograde catheter from the gallbladder. Furthermore, it must be emphasized that none of the patients required a papillotomy for cannulation. Although most gallbladder stones are radiolucent,14 we performed a plain CT scan of the gallbladder to exclude calcifications. We found that at <60 HU, coarse calcifications could virtually be excluded. While US revealed complete freedom from stones in 8 of 14 patients, 6 showed a residual stone mass up to 4 mm in diameter. In vitro studies have shown that after removal of cholesterol, this residual material is composed mainly of calcium bilirubinate, calcium carbonate, and a glycoprotein matrix. l5 .16 Provided it is fully functional, the gallbladder should be able to extrude this material. Nevertheless, we recommended adjuvant ursodeoxycholic/chenodeoxycholic acid treatment in these patients in order to prevent the residual material from acting as a nidus for crystallization of new stones. Endoscopic retrograde litholytic after-treatment of stone fragments remaining after extracorporeal shockwave lithotripsyl? might replace the previously recommended oral long-term dissolution therapy with chenodeoxycholic and ursodeoxycholic acids. 18 Postlithotripsy stone fragments should be even more susceptible to chemolitholysis than intact stones. Endoscopic retrograde cannulation of the gallbladder plus direct chemolitholysis might eventually prove to be an economical alternative to extracorporeal shockwave lithotripsy. Due to the short follow-up «12 months), no defin449
itive information on the recurrence rate associated with our form of treatment is yet available. The stonefree patients' post-treatment (8 of 14) are being carefully monitored with US. Should US reveal recurrent stones, oral chemolitholysis will be instituted. Because papillotomy is not required, and the size and number of stones in the gallbladder do not restrict its use, the relatively non-invasive retrograde catheterization procedure described in this article promises to be a well-tolerated form of treatment with wide application in symptomatic cholecystolithiasis.
REFERENCES 1. Nelson PE, Moyer TP, Thistle JL. Gallstone dissolution with methyl tert-butyl ether (MTBE): radiologic selection criteria. Gastroenterology 1990;98:1280. 2. Allen MJ, Borody TJ, Bugliosi TF, May GR, La Russo NF, Thistle JL. Rapid dissolution of gallstones by methyl tert-butyl ether. Preliminary observations. N Engl J Med 1985;312:21720. 3. Allen MJ, Borody TJ, Thistle JL. In vitro dissolution of cholesterol gallstones: a study of factors influencing rate and a comparison of solvents. Gastroenterology 1985;89:1097-103. 4. Thistle JL, May GR, Bender CE, et a1. Dissolution of cholesterol gallbladder stones by methyl tert-butyl ether administered by percutaneous transhepatic catheter. N Engl J Med 1989;320:633-9. 5. Foerster ECh, Auth J, Runge D, Ell C, Kerzel W, Domschke W. ERCG: Endoscopic retrograde catheterization of the gall-
450
bladder. Endoscopy 1988;20:30-3. 6. Ponchon T, Baroud J, Mestas JL, Chayvialle JA. Gallbladder lithotripsy: retrograde dissolution of fragments. Gastrointest Endosc 1988;34:468-9. 7. Foerster EC, Biihler H, Domschke W. Direct dissolution of gallbladder stones. Lancet 1989;1:954. 8. Foerster EC, Schneider MD, Matek W, Domschke W. Endoscopic retrograde catheterization of the gallbladder (ERCG). Gastrointest Endose- 1989;35:164. 9. Ponchon T, Baroud J, Pujol B, Vallette PJ, Perrot D. Renal failure during dissolution of gallstones by methyl-tert-butylether. Lancet 1988;2:276-7. 10. Allen MJ, Borody TJ, Bugliosi TF, May GR, LaRusso NF, Thistle JL. Cholelitholysis using methyl tertiary-butyl ether. Gastroenterology 1985;88:122-5. 11. Frimberger E, Kiihner W, Stiilzle L, Ottenjann R. Short-channel duodenoscope: innovations in selective biliography. Endoscopy 1983;15:359-63. 12. Kozarek RA. Selective cannulation of the cystic duct at the time of ERCP. J Clin Gastroenterol1984;6:37-41. 13. Nuboer JF. Studien iiber das extrahepatische Gallenwegssystem. Frankfurt Z Pathol 1931;41:198-246. 14. Soloway RD, Trotman BW, Ostrow JD. Pigment gallstones. Gastroenterology 1977;72:167-82. 15. Smith BJ, LaMont JT. Identification of gallbladder mucinbilirubin complex in human cholesterol gallstone matrix. J Clin Invest 1985;76:439-45. 16. Sutor DJ, Wooley SF. The organic matrix of gallstones. Gut 1974;15:487-91. 17. Brendel W, Enders G. Shock waves for gallstones: animal studies. Lancet 1983;1:105. 18. Sauerbruch T, Delius M, Paumgartner G, et a1. Fragmentation of gallstones by extracorporeal shock waves. N Engl J Med 1986;314:818-22.
GASTROINTESTINAL ENDOSCOPY
Endoscopic retrograde cannulation of the gallbladder: direct dissolution of gallstones Ernst Christoph Foerster, MD, PhD Werner Matek, MD Wolfram Domschke, MD Erlangen, Federal Republic of Germany
Percutaneous transhepatic catheterization of the gallbladder for dissolution of cholesterol stones by instillation of methyl tert-butyl ether (MTBE) is an invasive therapeutic procedure. The only non-invasive alternative available to now, endoscopic retrograde cannulation of the cystic duct, was difficult because of the cystic duct's tortuosity and spiral valves. We therefore developed a catheter system which, using conventional duodenoscopes during a routine endoscopic retrograde cholangiography (ERC) procedure, permits reliable and safe catheterization of the gallbladder without the need for endoscopic sphincterotomy. In 18 of 22 patients (82%) we were able to place a cysto-nasal catheter, and in 14 patients MTBE dissolution therapy was then performed. Eight patients (57%) were completely free of stones after treatment; the other six (43%) had residual debris. In 4 of 22 patients (18%) cannulation attempts failed, in 3 patients due to cystic duct blockage by a calculus. Endoscopic retrograde cannulation of the gallbladder (ERCG) represents a promising alternative to the invasive percutaneous transhepatic catheterization procedure. (Gastrointest Endosc 1990;36:444-450)
After publication of in vitro studies suggesting that the cholesterol solvent methyl tert-butyl ether (MTBE) dissolved the majority of cholesterol stones leaving little or no residue,! Allen et a1. 2 attempted dissolving gallbladder stones with MTBE administered via percutaneous transhepatic gallbladder catheterization. MTBE is liquid at body temperature and dissolves cholesterol stones within a few hours. 3 By 1989, the same Mayo Clinic group had instilled MTBE into the gallbladders of 75 patients; in 72 patients there was complete dissolution of stones or more than 95% dissolution. 4 In one case, however, bile leakage from the catheter track occurred, requiring intervention. Several other patients had transient pain immediately after removal of the catheter, suggesting minor leakage events. As an alternative to this invasive transhepatic ap-
proach, we have developed a new catheter system (U.S. patent no. 4,905,667; Schneider Europe AG, Zurich, Switzerland) providing reliable endoscopic retrograde cannulation of the gallbladder. 5 Another recently reported technique by Ponchon et a1. 6 permits successful catheterization only in exceptional cases where a favorable anatomical situation is present. Using our catheter system, we initially performed studies in autopsy specimens5 and preliminary reports of our early experience with patients have been published. 7,8 The present study describes the efficacy and safety of retrograde cannulation of the gallbladder in 22 patients, and direct chemolitholysis of symptomatic radiolucent gallstones with MTBE administered through a cysto-nasal catheter in 14 patients.
Received February 26, 1990. For revision April 18, 1990. Accepted July 2, 1990. From the Department of Medicine I, University of Erlangen-Nuremberg, Erlangen, Federal Republic of Germany. Reprint requests: E. C. Foerster, MD, Medizinische Klinik mit Poliklinik I der Universitiit Erlangen-Nurnberg, Krankenhausstr. 12, D-8520 Erlangen, Federal Republic of Germany.
Patients
444
PATIENTS AND METHODS
Twenty-two patients, 18 women and 4 men aged 21 to 79 years (mean, 54 years) were included in the study. Solitary stones (16 and 22 mm in diameter, respectively) were present in two patients. Multiple stones ranging widely in size (8 to 35 mm) and number (2 to 10 or more) were found in the GASTROINTESTINAL ENDOSCOPY
remaining 20 patients. No patient was excluded due to the size or number of stones. All patients had suffered one or more episodes of biliary colic or pancreatitis, but none had any signs, symptoms, or laboratory evidence of acute inflammation of the gallbladder, bile ducts, or pancreas at the time of treatment. Exclusion criteria included irreversible coagulopathy, known obstruction of the cystic duct or common bile duct, and pregnancy. All participating patients had previously refused cholecystectomy or were considered at high risk for surgery. Written informed consent was obtained from all patients; approval of the study protocol was granted by the ethics committee of our hospital. Following a fatty meal, cystic duct function was examined in all patients by means of ultrasonography (US). Plain films of the abdomen or radiographs of the kidney, ureter, and bladder were not routinely obtained. Computer tomographic (CT) scans without intravenous or oral radiopaque contrast media were performed in all patients to verify that the stones were not calcified. The scans, consisting of contiguous 4-mm thick slices at 4-mm intervals, were reconstructed with a 512 2 matrix display format. Hematological studies, serum aminotransferases, alkaline phosphatase, bilirubin, creatinine and amylase or lipase levels, prothrombin time, and hemolysis screen (haptoglobin and free hemoglobin) were done before treatment (normal in all cases) and repeated after each day of MTBE treatment.
M
~::::~.:::;:,.~~:::::::;::::::::=-z:;. -:.. ' .. -:-.-.:.,::.
A
-
-
.-G. _
:::::::===============-------_._-_.
B
Endoscopic retrograde insertion of the catheter into the gallbladder
After an overnight fast, patients received intravenous diazepam (2 to 5 mg) and pethidine (50 to 100 mg) prior to endoscopy. Antibiotics were not routinely given before or during the procedure. An intravenous line was maintained to permit rapid administration of medications as needed. Once the TJF 10 duodenoscope (Olympus Optical Co., Hamburg, Federal Republic of Germany) was positioned opposite the papilla of Vater, a "thick," flexible guide wire (length 280 cm, diameter 0.039 inch = 1.00 mm) was first inserted into the common bile duct (CBD) without papillotomy (Fig. 1A), and a modified angiography catheter (Schneider Europe AG, Zurich, Switzerland) (Fig. 1B), henceforth referred to as common bile duct catheter (CBDC) (length, 175 cm; outside diameter, 7 F, i.e., 2.15 mm; inside diameter, 1.2 mm), passed over it and placed in the common bile duct. The intra- and extrahepatic bile duct system was then filled with diluted contrast medium (Conray 60, Byk Gulden, Federal Republic of Germany). The thick, flexible guide wire was then withdrawn, and the tip of the CBDC hooked into the opening of the cystic duct (Fig. 1A). A custom-made thin steerable guide wire (Fig. lC) (0.02 inch = 0.52 mm in diameter, length 400 cm), with a 6-cm long soft distal tip, was placed in a Teflon (stiffening) cystic duct catheter (inside diameter, 0.60 mm; outside diameter, 1.05 mm; length, 205 cm) in such a manner that the tip of the wire extended just to the tip of the Teflon catheter (Fig. ID). The guide wire/Teflon catheter combination was then advanced along the CDBC and manipulated past the spiral valves and along the cystic duct. Cannulation was thus performed after conventional premedication with no need for previous sphincterotomy. Placement of the CBDC was performed in such a manner VOLUME 36, NO.5, 1990
uuum
2
il••••••• ==:==:==:=:
.11111111111111111111_. . . . . .
c ::: ~ ~~:..: .: .: .::::::::::
,
:'.:'.:",:'.:",:".:",::.:'
: , .. ::::-:::<::<:<:::::':::'::-:-:'::j I III'II'II~~==========~~~~"~~,
.
...
;,;.~~~~ '
.
'.'
..
'.'
,
D 5
2
Figure 1. A, Conventional duodenoscope (4) with the tip positioned at the papilla of Vater; transpapillary cannulation with the common bile duct catheter (1) and positioning of the tip in the opening of the cystic duct (3). ''Thick'' guide wire (2) already advanced along the catheter. B, Configuration of the pre-shaped common bile duct catheter (1) in the resting state; the curves (a) and (b) are stable in the axial direction. C, ''Thin'' guide wire (2) with soft, Teflon-coated flexible tip. 0, The tip of the common bile duct catheter (1) with advanced thin guide wire (2) enclosed within the stiffening Teflon catheter (5).
that the segment between the first and second curve (Fig. lB) became obliquely anchored within the CBD, while the flexible tip came to rest within the orifice of the cystic duct (Fig. lA). The usual distance between the two curves of the CBDC (Fig. 1B) was 4.2 cm. The crucial point was to ensure that the tip of the catheter lay freely within the orifice of the cystic duct and not at right angles to it, which would block off the opening of the catheter. Catheter placement was confirmed and documented by fluoroscopy (Fig. 2). 445
Figure 2. A, Radiograph of a 79-year-old woman. Contrast medium filling extrahepatic bile duct system including cystic duct and gallbladder. B, Common bile duct catheter with tip hooked in the opening of the cystic duct. C, Advancement of the guide wire/ catheter system. D, Guide wire positioned within the gallbladder.
After placing the guide wire in the gallbladder, the Teflon catheter was withdrawn and a specially designed polyethylene MTBE instillation catheter (outside diameter, 5 F, i.e., 1.55 mm; inside diameter, 1.0 mm; 360-cm long) with a 2to 3-cm diameter double pigtail tip containing multiple sideholes was passed. The pigtail tip was looped in the gallbladder fundus so that it encircled as many stones as possible (Fig. 3A). We found that placing one or two loops of the catheter in the gallbladder helped to prevent dislodgment. After withdrawal of the duodenoscope, the irrigation catheter was withdrawn through the nose, reduced to an external length of approximately 50 em, and fitted with a metalluerlock connector. 446
With the catheter in place, the gallbladder contents were completely aspirated, and contrast medium diluted with 0.9% saline was instilled to determine the volume needed to fill the gallbladder without spilling over into the cystic duct. An amount of MTBE equal to half this volume was later used for MTBE treatment. During treatment, the amount of MTBE instilled was increased with progressive dissolution of the stones. Discontinuation of treatment due to untoward effects was not necessary in any patient. Treatment with MTBE
Treatment usually commenced 2 to 3 hours following catheter placement in a treatment room fitted with an GASTROINTESTINAL ENDOSCOPY
cholesterol free. The gallbladder was then thoroughly rinsed and filled with sterile saline, and the catheter removed over the guide wire under fluoroscopic control to avoid knotting it. Before discharging the patients, US of the gallbladder and endoscopy of the upper gastrointestinal tract were performed. In the six patients with subtotal dissolution of gallstones, oral urso- and chenodeoxycholic acid (6 to 8 mg/ kg body wt/day) were administered for 3 months.
RESULTS Efficacy
Figure 3. A, Seventy-nine-year-old woman. Radiolucent stones filling the gallbladder as demonstrated by contrast medium instillation via a retrograde catheter. B, Same gallbladder as demonstrated by contrast medium after 8 hours of MTBE dissolution therapy.
externally vented spark-free exhaust fan. MTBE was instilled manually using 20- to 50-ml glass syringes at initial volumes ranging from 3 to 5 ml, and gradually increased to a maximum eventual volume of 80 ml (mean, 40 ml; range, 8 to 80). During the course of treatment, the MTBE volumes to be instilled were adjusted at 1-hr intervals by catheter cholecystography. Ether was instilled and aspirated in 3- to 5-min cycles. The amount of MTBE was decreased, particularly at the beginning of treatment, if the patient complained of pain. To avoid solvent saturation, the MTBE was replaced by fresh solvent after each instillation/aspiration cycle. The catheter within the cystic duct largely prevented overflow of MTBE into the duodenum, thus avoiding adverse intestinal effects of the solvent. All MTBE volumes instilled were carefully monitored. After establishing a rate and rhythm of MTBE infusion and aspiration which was comfortable for the patient, treatment was continued by trained paramedical staff. Fluoroscopy was performed after every 1 to 2 hours of treatment to assess the progress of dissolution. MTBE therapy was performed until the evaporated aspirates became VOLUME 36, NO.5, 1990
Catheterization of the CBD via the papilla of Vater was easily performed and positioning of the tip of the CBDC in the opening of the cystic duct was possible in each case. In 18 of 22 patients (82%), the cystic duct was cannulated with the catheterization technique described (Fig. 3A). In one case, however, the middle third of the cystic duct followed an acuteangled course and contained a pocket-like sacculation which prevented passing the guide wire. In three patients, cannulation attempts proved to be unsuccessful due to cystic duct blockage by calculus; in one of these patients, the cystic duct was inadvertently perforated and a cholecystectomy was subsequently carried out. Of the 18 patients with successful cannulation, 4 underwent endoscopic inspection of the gallbladder by means of a miniscope inserted through the cannulation catheter and 14 patients were treated by MTBE instillation. The maximum diameter of the lumen of the 22 cystic ducts was 3 mm. The entire cannulation procedure from the papilla of Vater to the gallbladder required a fluoroscopy time between 3 and 25 min (mean, 360 rad·cm 2 ). The overall procedure time for catheter placement was from 8 to 45 min (mean, 30 min). Based on changes in the size of the stones as determined by catheter cholecystography, we estimated that at least 95% ofthe stone mass was dissolved (Fig. 3B) in all but four patients (22%). In these four patients, the estimated reduction in stone mass did not exceed 50%. These relative failures were related to discrete calcifications of the gallstones, expressed in the CT scan by Hounsfield units between 60 and 80. Consequently, only stones with an HU value less than 60 (average 45) were subsequently treated with MTBE. In the majority of patients, the only residual debris detected by US was not greater than 4 mm in diameter. One patient with incomplete litholysis was re-admitted to the hospital 2 weeks post-treatment after developing temperatures to 38°C and right upper quadrant pain suggestive of acute cholecystitis 24 hours previously. The day after admission, the patient underwent an uneventful cholecystectomy. The intra447
vesicular debris was analyzed, and consisted entirely of minute flakes of cholesterol-free pigment which crumbled on touch. At pre-operative ultrasonography, the pigment flakes had presented as echodense, mobile, non-shadowing reflexes. Because we were able to instill up to 80 ml of MTBE into the gallbladder via the cystic duct that was largely occluded by the irrigation catheter, without any significant spillage into the common bile duct, virtually all calculi were successfully washed by the MTBE without having to change catheter position. All MTBE volumes instilled were carefully monitored (1000 ml mean total volume; range, 200 to 2600); loss ofMTBE, probably due to absorption by the gallbladder mucosa, did not exceed 1.8%. In this manner, dissolution treatment continued for a mean period of 5.6 hours (range, 2.5 to 16). The maximum individual yield of dissolved cholesterol was 106 g (range, 1 to 106; mean, 17.8) in this series. Ordinarily, dissolution therapy was done in one session on 1 day, while only one patient needed daily treatment sessions on 3 consecutive days. The patients were monitored by means of US at 3, 6, and 12 months post-treatment. In eight patients, US obtained immediately after stone dissolution revealed no evidence of residual debris. In the six patients with residual debris (four patients with more than 50%, two patients with more than 95% reduction in stone mass), during the subsequent adjuvant treatment with ursodeoxycholic and chenodeoxycholic acid over a period of 3 months, US never revealed any detectable increase in stone mass. One patient became stone free after 3 months with this adjuvant treatment. To date, we have carried out follow-up examinations for up to 12 months; patients remain continuously well despite minor residual debris. Side effects
During cystic duct and gallbladder cannulation, patients sedated in a standard manner for endoscopic retrograde cholangiography experienced no or only minor discomfort. During MTBE litholysis, 10 mg of diazepam needed to be administered intravenously in only 1 of 14 patients. Two patients experienced transient nausea with or without mild emesis, which responded to prochlorperazine suppositories. The smell of MTBE was often detected on the patients' breath. Depending on the volume of MTBE absorbed, mild transient sedation was also occasionally apparent. Upon termination of treatment, sedation and expired ether usually resolved within 30 min. There was no biochemical evidence of intravascular hemolysis in any patient. No changes occurred in serum haptoglobin concentration or free plasma hemoglobin; there were no significant changes in serum 448
hemoglobin concentration (13.6 g/dl before and 13.5 after MTBE treatment). In two patients, the aminotransferase activity was elevated (mean maximal level, 54 units/liter; normal 6 to 28). Serum bilirubin concentration rose transiently to above 26 JLmol/liter (normal <17) in two other patients. None of the 14 patients showed more than two-fold increase in serum alkaline phosphatase. As occasionally observed in ERCP procedures, an increase in serum amylase to a mean of 160 JLmol/ml (range, 140 to 340, normal <100) was found in seven patients. In six patients, total leukocyte count increased transiently (mean maximum, 14,000; range, 12,000 to 19,000). All abnormal laboratory values returned to normal within 2 days post-treatment. No parenteral analgesics were required to control abdominal pain after catheter removal; none of the patients developed fever. Subsequent endoscopy with biopsy showed no macroscopic or histological signs of duodenitis. None of the patients developed bile duct obstruction, cholangitis, or pancreatitis. Mean duration of hospitalization was 3 days (range, 2 to 5). DISCUSSION
The introduction of MTBE as a potent solvent of cholesterol stones opened up new possibilities in the treatment of gallbladder stones. Extensive pharmacological and toxicological studies, together with direct percutaneous transhepatic dissolution therapy of gallbladder stones in more than 400 patients throughout the world, have shown that MTBE has no demonstrable local toxic effects on the mucosa of the gallbladder. The percutaneous transhepatic approach to the gallbladder, however, is associated with a number of basic problems. 4 Although a 5 French polyethylene catheter can be introduced under radiological control into the gallbladder transhepatically in 100% of patients, contrast medium instillation via the catheter has been associated with extravasation from the site of the puncture. An additional problem associated with this invasive approach is the fact that spillage of the MTBE via the cystic duct into the duodenum must be carefully avoided. In the duodenum, in contrast to the gallbladder, MTBE is well absorbed, which can produce severe adverse reactions, such as intravascular hemolysis. Endoscopic inspection occasionally reveals acute duodenitis following treatment with MTBE. Ponchon et al. 9 have recently published a report of a patient who developed acute renal failure requiring dialysis following intravascular hemolysis arising from massive extravasation of MTBE during percutaneous transhepatic dissolution therapy of gallbladder stones. Marked changes in the duodenal mucosa following the absorption of MTBE have also been described in GASTROINTESTINAL ENDOSCOPY
MTBE treatment of common bile duct stones. 10 Attempts have been made to reduce the undesirable spillage of MTBE into the duodenum by minimizing the amount of MTBE instilled and placing the pigtail catheter as close to the stone(s) as possible. A further problem inherent in the invasive procedure of percutaneous transhepatic puncture of the gallbladder is bile leakage from the catheter track into the abdominal cavity upon removal of the catheter. Large leakages of this type necessitate corrective gallbladder puncture or surgical intervention. Patients' complaint of pain immediately on removal of the irrigation catheter is suggestive of minor leakage episodes. Attempts have been Made to plug biliary fistulas remaining after removal of the catheter with gelatin sponge, but results have not yet been published. All of these problems associated with the percutaneous transhepatic approach to the gallbladder prompted us to develop an improvement to the endoscopic retrograde approach. Selective catheterization of the cystic duct and the gallbladder during endoscopic retrograde cholangiography has to date been successful in less than 10%, or only in the presence of a favorable anatomical situation, or after post-inflammatory destruction of the spiral valves. Frimberger et al. 11 were the first to report on routine attempts to cannulate the gallbladder with the aid of a special short-channel duodenoscope; with this relatively complicated system, however, reliable cannulation of the gallbladder has so far not proven possible. Ponchon et al.~ have recently published a report on successful retrograde placement of a nasobiliary 6.5 French Teflon drain in two of five patients. However, all patients, including those in whom the attempt was unsuccessful, experienced marked pain; all had a prior papillotomy. Because ofthe spiral valves, negotiation ofthe cystic duct is technically the most difficult part of retrograde gallbladder cannulation. Although Kozarek 12 succeeded in entering the opening of the cystic duct in 70% of a series of 50 patients, cannulation of the gallbladder itself was only rarely possible. Using our newly developed catheter system, we succeeded in safely, and for the most part painlessly, cannulating the gallbladder in 18 of22 patients (82%). In extensive preliminary studies on autopsy specimens, we were able to tailor our catheter system to the particular anatomical situation. Morphological studies performed by Nuboer l3 showed that the cystic duct has an average diameter of 2 to 3 mm; a diameter of less than 1 mm has been observed only rarely. In the present study, the spiral valve was responsible for only one of the four (18%) unsuccessful attempts at cannulation. In the other three cases, an incarcerated calculus in the cystic duct was responsible for failure; inadequate contraction of the gallbladder following a VOLUME 36, NO.5, 1990
fatty meal observed at US had already raised the suspicion of a cystic duct calculus, but at least partial patency of the duct had been expected. The single case of perforation occurring during retrograde cannulation of the cystic duct resulted from accidentally advancing the common bile duct catheter beyond the opening of the duct and impinging on an area of wall weakened by inflammation in the region of the stone. A suspected cystic duct calculus should therefore represent an exclusion criterion for retrograde chemolitholysis. Because the gallbladder catheter we employ partially occludes the cystic duct, compared with the percutaneous transhepatic system, a considerably greater irrigation volume (up to 80 ml vs. 7 to 15 ml) can be instilled without any major spillage of MTBE into the duodenum. Accordingly, the duration of dissolution treatment with the retrograde catheter seems to be appreciably shorter (5.6 hours) than with the percutaneous transhepatic approach (12.5 hours). Clearly, a more definite comparison between the endoscopic retrograde and the percutaneous transhepatic access routes cannot be made without a carefully matched randomized study. Problems were neither expected nor observed during withdrawal of the retrograde catheter from the gallbladder. Furthermore, it must be emphasized that none of the patients required a papillotomy for cannulation. Although most gallbladder stones are radiolucent,14 we performed a plain CT scan of the gallbladder to exclude calcifications. We found that at <60 HU, coarse calcifications could virtually be excluded. While US revealed complete freedom from stones in 8 of 14 patients, 6 showed a residual stone mass up to 4 mm in diameter. In vitro studies have shown that after removal of cholesterol, this residual material is composed mainly of calcium bilirubinate, calcium carbonate, and a glycoprotein matrix. l5 .16 Provided it is fully functional, the gallbladder should be able to extrude this material. Nevertheless, we recommended adjuvant ursodeoxycholic/chenodeoxycholic acid treatment in these patients in order to prevent the residual material from acting as a nidus for crystallization of new stones. Endoscopic retrograde litholytic after-treatment of stone fragments remaining after extracorporeal shockwave lithotripsyl? might replace the previously recommended oral long-term dissolution therapy with chenodeoxycholic and ursodeoxycholic acids. 18 Postlithotripsy stone fragments should be even more susceptible to chemolitholysis than intact stones. Endoscopic retrograde cannulation of the gallbladder plus direct chemolitholysis might eventually prove to be an economical alternative to extracorporeal shockwave lithotripsy. Due to the short follow-up «12 months), no defin449
itive information on the recurrence rate associated with our form of treatment is yet available. The stonefree patients' post-treatment (8 of 14) are being carefully monitored with US. Should US reveal recurrent stones, oral chemolitholysis will be instituted. Because papillotomy is not required, and the size and number of stones in the gallbladder do not restrict its use, the relatively non-invasive retrograde catheterization procedure described in this article promises to be a well-tolerated form of treatment with wide application in symptomatic cholecystolithiasis.
REFERENCES 1. Nelson PE, Moyer TP, Thistle JL. Gallstone dissolution with methyl tert-butyl ether (MTBE): radiologic selection criteria. Gastroenterology 1990;98:1280. 2. Allen MJ, Borody TJ, Bugliosi TF, May GR, La Russo NF, Thistle JL. Rapid dissolution of gallstones by methyl tert-butyl ether. Preliminary observations. N Engl J Med 1985;312:21720. 3. Allen MJ, Borody TJ, Thistle JL. In vitro dissolution of cholesterol gallstones: a study of factors influencing rate and a comparison of solvents. Gastroenterology 1985;89:1097-103. 4. Thistle JL, May GR, Bender CE, et a1. Dissolution of cholesterol gallbladder stones by methyl tert-butyl ether administered by percutaneous transhepatic catheter. N Engl J Med 1989;320:633-9. 5. Foerster ECh, Auth J, Runge D, Ell C, Kerzel W, Domschke W. ERCG: Endoscopic retrograde catheterization of the gall-
450
bladder. Endoscopy 1988;20:30-3. 6. Ponchon T, Baroud J, Mestas JL, Chayvialle JA. Gallbladder lithotripsy: retrograde dissolution of fragments. Gastrointest Endosc 1988;34:468-9. 7. Foerster EC, Biihler H, Domschke W. Direct dissolution of gallbladder stones. Lancet 1989;1:954. 8. Foerster EC, Schneider MD, Matek W, Domschke W. Endoscopic retrograde catheterization of the gallbladder (ERCG). Gastrointest Endose- 1989;35:164. 9. Ponchon T, Baroud J, Pujol B, Vallette PJ, Perrot D. Renal failure during dissolution of gallstones by methyl-tert-butylether. Lancet 1988;2:276-7. 10. Allen MJ, Borody TJ, Bugliosi TF, May GR, LaRusso NF, Thistle JL. Cholelitholysis using methyl tertiary-butyl ether. Gastroenterology 1985;88:122-5. 11. Frimberger E, Kiihner W, Stiilzle L, Ottenjann R. Short-channel duodenoscope: innovations in selective biliography. Endoscopy 1983;15:359-63. 12. Kozarek RA. Selective cannulation of the cystic duct at the time of ERCP. J Clin Gastroenterol1984;6:37-41. 13. Nuboer JF. Studien iiber das extrahepatische Gallenwegssystem. Frankfurt Z Pathol 1931;41:198-246. 14. Soloway RD, Trotman BW, Ostrow JD. Pigment gallstones. Gastroenterology 1977;72:167-82. 15. Smith BJ, LaMont JT. Identification of gallbladder mucinbilirubin complex in human cholesterol gallstone matrix. J Clin Invest 1985;76:439-45. 16. Sutor DJ, Wooley SF. The organic matrix of gallstones. Gut 1974;15:487-91. 17. Brendel W, Enders G. Shock waves for gallstones: animal studies. Lancet 1983;1:105. 18. Sauerbruch T, Delius M, Paumgartner G, et a1. Fragmentation of gallstones by extracorporeal shock waves. N Engl J Med 1986;314:818-22.
GASTROINTESTINAL ENDOSCOPY