MALDI-TOF MS imaging of controlled release implants.

MALDI-TOF MS (matrix-assisted laser desorption/ionization time-of-flight mass spectrometry) imaging is used to characterize novel lipid implants allowing for controlled drug delivery. Importantly, this innovative technique provides crucial information on the inner structure of the implants before and after exposure to the release medium and does not require the addition of marker substances. Implants were prepared by extrusion at room temperature. Thus, in contrast to hot-melt extruded systems, the risks of drug inactivation and solid state transformations of the lipid matrix former are reduced. Hydrogenated/hardened soybean oil and glyceryl tristearate were studied as lipids and propranolol hydrochloride and theophylline as drugs, exhibiting significantly different solubility in water. The implants were also characterized by optical microscopy, differential scanning calorimetry, water uptake and lipid erosion studies, mathematical modeling as well as in vitro drug release measurements. Importantly, broad spectra of drug release patterns with release periods ranging from a few days up to several months could easily be provided when varying the initial drug content and type of lipid, irrespective of the type of drug. The diameter of the implants can be as small as 1mm, facilitating injection. MALDI-TOF MS imaging revealed homogeneous macroscopic drug distributions within the systems, but steep drug concentration gradients in radial and axial direction at the lower micrometer level, indicating drug- and lipid-rich domains. As the implants do not significantly swell, local irritation upon administration due to mechanical stress can be expected to be limited. Good agreement between experimentally measured and theoretically calculated drug release kinetics revealed that diffusional mass transport plays a major role for the control of drug release from this type of advanced drug delivery systems.

MALDI imaging of formalin-fixed paraffin-embedded tissues: application to model animals of Parkinson disease for biomarker hunting.

A common technique for the long-term storage of tissues in hospitals and clinical laboratories is preservation in formalin-fixed paraffin-embedded (FFPE) blocks. Such tissues stored for more than five years have not been useful for proteomic studies focused on biomarker discovery. Recently, MS-based proteomic analyses of FFPE showed positive results on blocks stored for less than 2 days. However, most samples are stored for more than one year, and thus our objective was to establish a novel strategy using as a model system 6-hydroxydopamine (6-OHDA) treated rat brain tissues stored in FFPE blocks for more than 9 years. We examined MALDI tissue profiling combining the use of automatic spotting of the MALDI matrix with in situ tissue enzymatic digestion. On adjacent sections, the identification of compounds is carried out by tissue digestion followed by nanoLC/MS-MS analysis. The combination of these approaches provides MALDI direct analysis, MALDI/MS imaging, as well as the localization of a large number of proteins. This method is validated since the analyses confirmed that ubiquitin, trans-elongation factor 1, hexokinase, and the Neurofilament M are down-regulated as previously shown in human or Parkinson animal models. In contrast, peroxidoredoxin 6, F1 ATPase, and alpha-enolase are up-regulated. In addition, we uncovered three novel putative biomarkers, the trans-elongation factor 1 (eEF1) and the collapsin response mediator 1 and 2 from protein libraries. Finally, we validate the CRMP-2 protein using immunocytochemistry and MALDI imaging based on the different ions from trypsic digestion of the protein. The access to archived FFPE tissue using MALDI profiling and imaging opens a whole new area in clinical studies and biomarker discovery from hospital biopsy libraries.

[Image guided drainage of pelvic fluid collections: results in 42 patients]. / Drainage radiologique des collections pelviennes: résultats chez 42 malades.

PURPOSE: To review the etiology, location, and morphology of pelvic collection as well as the technique and results of image guided drainage. MATERIALS AND METHODS: From June 1996 to August 2002, we performed image guided drainage of pelvic fluid collections in 21 males and 21 females. In patients where a direct contact between the collection and the endocavitary probe was present, the drainage was performed either by transrectal or transvaginal approach using 10F, 12F, 14F or 16F catheters according to the viscosity of the fluid. When the patients were no longer septic, when drainage had stopped, the drains were removed at day 5. When a fistula was present, the drain was left in place until the fistula healed. RESULTS: The most common location of pelvic collections was the cul-desac (43%). A total of 81% of pelvic abscesses were digestive in origin, either from the colon or appendix. Transrectal or transvaginal drainage was possible in 83% of cases. Mean follow-up was 41 months. No drainage related complication was recorded. In two patients with collections of clear fluid, a simple aspiration was performed without insertion of a drain. In the 40 other patients, a drainage catheter was inserted. Twenty-nine patients were cured after 15 days of drainage. Two patients had recurrent collections. Image guided drainage failed in five patients, and all underwent successful surgical management. CONCLUSION: Image guided drainage of pelvic collections is a safe and effective procedure. Failures were due to initially undiagnosed pathology requiring surgical treatment.

[Percutaneous treatment of intrahepatic lithiasis]. / Traitement percutané de la lithiase intra-hépatique.

AIM: To present our experience with percutaneous intracorporeal electrohydrolic lithotripsy in the treatment of intrahepatic lithiasis. SUBJECTS AND METHODS: From January 1989 to November 1998, 53 patients with intrahepatic lithiasis were treated with percutaneous intracorporeal electrohydrolic lithotripsy. Twenty-six patients had primary intrahepatic lithiasis. Intrahepatic stones were associated with intrahepatic duct abnormalities in 11 patients, 9 had strictures and 2 had cystic dilatations. Twenty-seven patients had secondary intrahepatic lithiasis formed a biliodigestive bypass in 20 patients. Intracorporeal electrohydrolic lithotripsy was performed under cholangioscopic guidance in all patients. The endoscope was introduced into the biliary ducts through a cutaneobiliary tract in 51 patients, through a cutaneocholecystic tract in one and through a cutaneojejunal tract in one. These tracts were created and gradually dilated in two sessions three days apart. In twenty-two patients stenosis or sharp angulation prevented adequate positioning of the scope which was only successful after balloon dilation or insertion of a stiff wire. RESULTS: Complete clearance of stones was achieved in 49 patients (92%). Biliary or hepaticojejunostomy strictures were successfully dilated with an angioplasty balloon in all patients. Ten patients (19%) had early complications: four had bilomas treated by percutaneous drainage, three had resolutive onset of cholangitis, two had transient arterial hemobilia, and one had a pneumothorax. The mean duration of follow-up was five years. During this period, 5 patients (9%) had recurrent symptoms of biliary obstruction. Among these patients, three (5.7%) had recurrent symptomatic intrahepatic lithiasis, one had a recurrent biliary stricture and one had secondary sclerosing cholangitis. Treatment of recurrent stones was repeated intracorporeal electrohydrolic lithotripsy in two and left hepatectomy in one; recurrent biliary stricture was treated by hepaticojejunostomy and secondary sclerosing cholangitis by antibiotics. CONCLUSION: Intracorporeal electrohydrolic lithotripsy is effective and safe and should be proposed as the first line treatment of primary or secondary intrahepatic lithiasis.

Percutaneous incision of stenotic uroenteric anastomoses with a cutting balloon catheter: long-term results.

PURPOSE: To describe the technique and results of incision of strictures in anastomotic urinary diversions with a commercially available cutting balloon catheter. MATERIALS AND METHODS: Thirty-seven stenoses were treated in 32 patients. Most (28 [88%]) of the patients had undergone surgery for bladder cancer 17.7 months +/- 17.4 (SD) (range, 3-72 months) before incision. Thirteen patients had undergone ileal conduit diversion, and nineteen had undergone enterocystoplasty. All stenoses were shorter than 3 cm. The presence of adjacent ileal loops and/or iliac vessels was assessed with computed tomography before incision. The cutting wire was oriented anteriorly or anterolaterally, and the balloon was inflated with diluted contrast material during the incision. A Kaplan-Meier survival curve was constructed to illustrate the success rates over time. RESULTS: No major complications occurred. Twelve (32%) stenoses recurred in nine patients 15 months +/- 10 (range, 6-36 months) after stent removal; the failure rate was 53% (eight of 15 stenoses) for ileal conduits and 18% (four of 22 stenoses) for enterocystoplasties. Late failure (>12 months) was observed in four patients. The patency of the other 25 stenoses (23 patients) was checked 25 months +/- 11 after stent removal (range, 5-43 months). The actuarial patency rate was 77% at 1 year, 68% at 2 years, and 62% at 3 years. CONCLUSION: Cutting balloon incision is a safe and simple alternative to surgery, particularly when the urinary diversion is enterocystoplasty.

Placement of metallic stents for treatment of postoperative biliary strictures: long-term outcome in 25 patients.

OBJECTIVE: This study was undertaken to evaluate the results of our 7-year experience with Gianturco-Rosch metallic stents, used for the management of postoperative biliary strictures. SUBJECTS AND METHODS: From January 1989 to April 1995, self-expanding Gianturco-Rosch metallic stents were placed in 25 patients with postoperative bile duct stenosis. All patients had a history of bile duct injury during cholecystectomy. Twenty-four patients had a conventional open cholecystectomy and one patient had a laparoscopic cholecystectomy. Eight patients had stenosis at the level of the common bile duct. The other 17 patients, who had undergone surgical repair of the bile duct, had a stricture at the level of the hepaticojejunostomy. These anastomotic strictures recurred after simple cholangioplasty. Patients were monitored for 9-84 months (mean, 55 months). Treatment was considered successful if the initial stenosis did not recur. Treatment was considered a failure if the initial stenosis recurred within the stent. RESULTS: Two patients had early complications: one had bile pleural effusion, treated with percutaneous drainage, and the other had arterial hemobilia, treated with embolization. Eighteen (72%) of 25 patients had no recurrence of the initial strictures. Among these patients, 11 had no further symptoms of biliary obstruction and seven, all with strictured hepaticojejunostomies, had recurrent episodes of cholangitis caused by secondary sclerosing cholangitis or intrahepatic stone formation. Seven (28%) of 25 patients had recurrence of the initial stenoses, causing repeated episodes of cholangitis. Among these seven patients, six had common bile duct stenoses and one had an anastomotic stricture. Recurrent biliary obstruction was treated surgically or with percutaneous methods, despite the presence of the metallic stent. CONCLUSION: Gianturco-Rosch stent placement should be considered in patients with postoperative bile duct stenoses in whom another operation is not indicated and cholangioplasty has failed. The results are better in patients who have hepaticojejunostomy strictures rather than common bile duct strictures. Overall, a long-term recurrence rate of cholangitis of more than 50% of patients was seen because of recurrence of the original stenosis or intrahepatic bile duct obstruction.

[Percutaneous treatment of malignant stenoses of the hilum]. / Traitement percutané des sténoses malignes du hile.

PURPOSE: The endoscopic treatment of malignant hilar obstruction is followed in 70% of the case by infection of undrained biliary sectors. We report the influence of complete biliary drainage on post procedural cholangitis. METHOD: From January 1990 to January 1993 we treated 120 consecutive patients presenting with a malignant hilar obstruction. There were 61 women and 59 men, mean age 65 +/- 7.5 years. The level of stenosis was type II in 45 patients (37%), type III in 18 patients (13%) and above type III in 57 patients (48%). Complete biliary drainage with multiple biliary access was attempted in all patients. Long term internal drainage was achieved by metallic autoexpansive endoprosthesis. RESULTS: Complete drainage was achieved in all patients with type II or type III biliary stenosis. Drainage was incomplete in all patients with biliary stenosis above type III. Early complications were observed in 35% of the patients. Persistent cholangitis, the most frequent complication (22%) was only observed in patients with above type III biliary stenosis. Mortality at 30 days was 17%. Recurrent biliary obstruction was observed in 22% of the patients after an average of 187 days. Median survival was 95 days. CONCLUSION: Complete biliary drainage prevents persistent cholangitis in patients with type II or III biliary stenosis without increasing other complications related to biliary drainage.

[Large calculi of the common bile duct. Value of lithotripsy]. / Calculs volumineux de la voie biliaire principale. Intérêt de la lithotritie.

Certain stones need to be fragmented before being extracted via endoscopic sphincterotomy (ES). From April 1988 to December 1991, extracorporeal lithotripsy was used in this indication in 28 patients (22 females, 6 males) with a mean age of 77 +/- 20 years, using an ultrasound-guided electrohydraulic lithotriptor. Stone detection was performed after perfusion of the nasogastric tube and was easy in 20 cases (71%), difficult in 6 cases (22%) and impossible in 2 cases (7%), which could not be treated by this method. The patients had an average of 1.4 +/- 0.9 stones measuring 19.6 +/- 8 mm and received an average of 2.480 +/- 580 shock waves in a single session for 24 patients and in two sessions for 2 patients. Radiologically obvious fragmentation was achieved in 11 out of 26 cases (42%) and was found to be effective at a further extraction attempt in 4 other cases. Complete clearance of the common bile duct was achieved in 15 cases (57.7%). The size of the stones (> or < 20 mm) and the solitary or multiple nature of the stones did not significantly influence the fragmentation results. No complication related to the technique was observed apart from the constant development of cutaneous petechiae. An improvement in the power of the generator and the use of fluoroscopic rather than ultrasonographic guidance should allow an improvement of these results in the future. Because of its safety, this lithotripsy method can be proposed following failure of mechanical lithotripsy.

[Palliative treatment by endoprosthesis of icterus caused by distal biliary tumoral obstruction]. / Traitement palliatif par endoprothèse des ictères par obstacle tumoral biliaire sous-hilaire.

Between March 1982 and December 1987, 466 patients (256 women, 210 men, mean age 73 years) with tumor obstruction of the common bile duct were referred to our department. The causes of obstruction were carcinoma of the pancreas (298 patients), carcinoma of the ampulla of Vater (32 patients) and carcinoma of the common bile duct (136 patients). Endoscopical insertion of a biliary prosthesis was initially possible in 377 patients (81%). In case of failure, patients were referred to the radiologist for percutaneous drainage. Successful drainage was obtained in 58 patients with an overall success rate of 93% (435 patients). Endoscopic replacement was necessary in 170 cases for 114 patients and was successful in 155 (91%). Pruritus was relieved in 89% of the patients. Serum bilirubin levels decreased more than 75% after initial endoscopic endoprosthesis, repeated endoscopic endoprosthesis and percutaneous prosthesis insertion in 80%, 79%, and 62% of the patients, respectively. Short term complications of endoscopic endoprosthesis occurred in 28% of patients with a mortality rate of 8%. Percutaneous prosthesis complications occurred in 33% of patients with a mortality rate of 11%. In the long term, cholangitis was the main complication and occurred in 27% of patients with a delay of 103 +/- 105 days. Intestinal obstruction was observed in 7% of patients. The average life expectancy of endoscopic endoprosthesis and percutaneous prosthesis was 109 +/- 157 and 92 +/- 101 days, respectively. The average life expectancy of patients was 163 +/- 224 days.(ABSTRACT TRUNCATED AT 250 WORDS)

[Color doppler echography in the exploration of vasculogenic impotence]. / Echographie-doppler couleur dans l'exploration de l'impuissance d'origine vasculaire.

102 patients with suspected vasculogenic impotence were evaluated with color doppler sonography. Measurement of normal systolic and diastolic velocities were obtained from the cavernosal arteries of patients responding by a full erection after intra-cavernosal injection of 20 mg of Papaverine. A correlation with cavernosometry was obtained in 61 patients and with selective internal pudendal arteriography in 11. The 10 patients with abnormal arteriograms had a systolic velocity < 25 cm/sec. 13 out of the 15 patients with an end diastolic velocity > 5 cm/sec had a venous leak defined by a maintenance flow rate of erection during cavernosometry > 25 ml/mn. End diastolic velocity is an excellent index of the function of the veno-occlusive system, provided the systolic velocity remains at a normal value. In case of arterial insufficiency, a diastolic flow < 5 cm/sec is of no value and cavernometry is mandatory to detect a mixed arterio-venous impotence. The addition of color doppler sonography permitted a more rapid detection of vessels and an easily reproducible measurement of velocities which makes color doppler sonography an excellent screening test for examining patients with potential vasculogenic impotence.

Management of malignant hilar biliary obstruction by endoscopy. Results and prognostic factors.

Between January 1983 and December 1987, 103 patients who had hilar biliary obstruction (59 men, 44 women, median age 73 years) were referred to our institution. The causes of hilar biliary obstruction were carcinoma of the bile ducts (55), hepatic metastases or hepatocellular carcinoma (30), and carcinoma of the gallbladder (18). When endoscopic retrograde cholangiography was performed, the stricture was classified as type I in 28%, type II in 41%, and type III in 31% of the patients. In 92 patients, we tried to insert endoscopically a 10, 11, or 12 F Amsterdam type prosthesis; it proved possible in 66 (74%), and the prosthesis proved functional without further procedure in 49 cases (53%); no combined percutaneous and endoscopic method was used. At death or discharge, 45 patients (49%) had a successful drainage. Cholangitis was the main procedure-related complication and occurred in 25 patients. The 30-day mortality was 43%. Results varied according to type of stenosis: successful drainage was performed in 15% of the patients with type III stenosis, compared with 86% when the stenosis was of type I. Under a multivariate analysis the independent prognostic factors of 30-day mortality were: (1) development of infectious complications after endoscopic attempt at drainage (P less than 0.0001), and (2) absence of successful drainage (P less than 0.0001). In conclusion, endoscopic endoprosthesis placement allows a sufficient drainage in 53% of the cases. In type III stenosis, the high rate of 30-day mortality leads us the conclusion that endoscopic drainage must be avoided.

Fluoroscopically guided percutaneous transrenal electroincision of ureterointestinal anastomotic strictures.

A new technique for electroincision of a strictured ureterointestinal anastomosis is described that uses a sphincterotome and high frequency current. After placement of a percutaneous nephrostomy tube a 7F "wire guided" sphincterotome was placed into the stenosis. The cutting wire was then deflected while cutting current was applied intermittently. Injection of contrast medium through the papillotome probe assessed the depth of the incision. A 10 mm. angioplasty balloon was inflated at low pressure to verify that the anastomosis had been incised to a depth of 1 cm. The anastomosis was then stented for 8 weeks with an 18F stent. The operative time did not exceed 45 minutes. A total of 9 stenoses was treated in 7 patients: 4 were ileal conduit diversions and 5 were enterocystoplasties. No immediate complication was observed. In 1 case a small urinoma was surgically drained at removal of the stent. Six stenoses are patent with 2, 3, 4, 4, 10 and 13 months of followup after removal of the stent. One patient died of bladder tumor metastases during the stenting period and 1 with bilateral incision still has a stent. The technique can be performed without major complication (bleeding or digestive fistula). Long-term results remain to be assessed.

Endoscopic treatment of postoperative biliary fistulae.

Postoperative biliary fistulae are difficult to manage, particularly in the face of obstruction or malignancy. We used endoscopic sphincterotomy or endoprosthesis placement to aide fistula closure in 52 patients with postoperative biliary fistulae. Thirty-seven patients with a fistula were treated with endoscopic sphincterotomy alone. Twenty-four of these 37 patients had a history of lithiasis; 21 patients were treated successfully by endoscopic sphincterotomy alone. The fistula closed in 2.4 +/- 1.6 days. Among the other 13 patients without history of stone disease, the fistula closed in seven cases (54%), 8.4 +/- 2 days after endoscopic treatment. Three patients ultimately required surgical intervention. In 15 patients an attempt was made to pass a 10F endoprosthesis above the fistula. Among the eight patients with successful prosthesis insertion, the fistula healed in six patients (75%). In the seven patients in whom a prosthesis could not be passed endoscopically, the percutaneous transhepatic approach was used. Surgical treatment (hepaticojejunal anastomosis) was ultimately required in two of these seven patients. Sphincterotomy alone is the preferred treatment for biliary fistulae-complicating surgery for gallstone disease. Alternatively, when a fistula is large, endoscopic placement of a prosthesis can be proposed as the first treatment. In cases of endoscopic failure, placement of a prosthesis through the percutaneous transhepatic approach is a useful alternative, particularly when the fistula source is located in the intrahepatic biliary tract.

Common bile duct and intrahepatic stones: results of transhepatic electrohydraulic lithotripsy in 50 patients.

Percutaneous, transhepatic, intracorporeal, electrohydraulic shock wave lithotripsy was performed in 50 patients after failure of endoscopic treatment (n = 43) or directly in patients with a strictured hepaticojejunostomy (n = 7). Twenty-seven patients had common bile duct stones; 23, intrahepatic stones. Three steps were used: A transhepatic bilicutaneous fistula was created, a wide communication between the bile duct and the gut was established, and contact shock wave lithotripsy was performed under endoscopic guidance. Afterward, 46 patients were free of stones. In four patients with diffuse intrahepatic lithiasis, only 75% of stones could be cleared. Severe complications, seen in 11 patients (hemobilia necessitating transfusion [n = 6], bile duct perforation resulting in cholangitis [n = 3], acute pulmonary edema [n = 1], and hemothorax [n = 1]), were fatal in four patients; all occurred early in the study. The authors modified their technique by dilating the biliary tract in two sessions 3 days apart, waiting 6 days for the tract to mature, and then introducing the cholangioscope directly through the skin, significantly reducing complications and mortality (P less than .005).

Impassable ureteral strictures: management with percutaneous ureteroneocystostomy.

Sixteen of 227 patients referred for percutaneous placement of a ureteral stent had impassable stenoses. Stenoses were benign (n = 8) or attributed to malignant retroperitoneal neoplasm (n = 8). Electrocautery was used to create a neotract between the stenosed ureter and the bladder or ileal loop. A double-J stent was placed after dilation of the tract by use of angioplasty. Neotracts were established and stents were placed in all patients. Complications (digestive tract fistulas) developed in two patients. This technique is safe if the electrode is placed close to the bladder or ileal loop. The procedure can be used as an alternative to surgery or permanent nephrostomy or in initial treatment of benign anastomotic stenosis.

[Use of papillotome for antegrade treatment of ureteral stenoses and uretero-ileal anastomotic strictures under radioscopic control. Experimental study in dogs and clinical application]. / Utilisation du sphinctérotome par voie antégrade et sous contrôle radioscopique dans le traitement des sténoses urétérales et des anastomoses urétéro-iléales sténosées. Etude expérimentale chez le chien et application clinique.

A new technique of electro-incision of ureteral stenoses and strictured uretero-enteric anastomoses is presented. Incision is performed with a papillotome, routinely used to achieve endoscopic retrograde sphincterotomy of the duodenal papilla. Group I: 7 ureteral stenoses were performed, on a dog model after surgical ligation of the lumbar ureter. Ten days later, through a percutaneous approach, the papillotome was placed through the stenosis, deflected, and cutting current was applied to incise the stenosis. The IVP performed one month later showed disappearance of the stenosis in 4 cases, a residual stenosis without obstacle in 2 cases and a residual stenosis with obstacle in one case. Group II: 7 strictured uretero-enteric anastomoses on 6 patients. After placement of a percutaneous nephrostomy, a wire guided papillotome was placed into the stenosis. Cutting current was then applied to cut the stenosis. A 18 F ureteral stent was subsequently placed for 8 weeks. Two patients have a patent anastomosis, 7 and 10 months after removal of the stent. One patient died from metastases of bladder tumor. The three remaining patients are still stented. An urinoma occurred in one patient the day following removal of the stent, and was surgically drained. This new technique which combines electro-incision and stenting with a large caliber stent may be proposed as an alternative to surgery or balloon dilatation for the treatment of strictured uretero-digestive anastomoses.

[Endoscopic retrograde cholangiopancreatography and biliary prosthesis]. / Cholangio-pancréatographie rétrograde endoscopique et prothèses biliaires.

Endoscopic retrograde cholangio-pancreatography (ERCP) confirms the diagnosis of obstructive jaundice and sometimes provides the histological proof that the stenosis is due to cancer. Palliative treatment of biliary and pancreatic cancers is indicated when the extension of the tumour is such that it precludes any oncologically satisfactory excision, in patients at high operative risk, when jaundice recurs after surgery and in cases of biliary metastases from distant cancers. Cancers located below the hilum are usually easily treated by endoscopic insertion of a biliary stent, whereas hilar cancers extending to the bifurcation often require combined endoscopic and percutaneous techniques to drain all liver segments. The most frequent of early complications of biliary stents is cholangitis, notably in hilar cancers. Late complications, notably obstruction of the stent, can be reduced by using expandable metal stents.

[Endoscopic and transhepatic intubation of malignant bile duct strictures for postoperative jaundice]. / Intubation des sténoses biliaires malignes par voie endoscopique et transhépatique pour ictère après chirurgie.

Out of 625 patients referred for stenting for a malignant stricture of the biliary tract, 97 (15.8%) had undergone previous surgery. Resection had been performed in 43 cases, by-pass in 15, surgical stenting in 11, laparotomy in 28. The stricture was located in the porta hepatis in 48 patients (49.5%) in the middle common bile duct (CBD) in 47 (48.5%) and juxtapapillary in 2 (2%). Endoscopic retrograde transhepatic stenting was successful in 51 patients (52.5%). Percutaneous transhepatic stenting was successful in 41 cases out of 46 (85%) and in 5 cases, only external drainage was possible. A 75% reduction in serum bilirubin was observed in 78 patients (81.5%) and normalization was observed in 66 (90%) who survived more than one month. The complication rate was 31.3% in the endoscopic group and 47.7% in the percutaneous transhepatic group, with a mortality related to early complications of 9.8% and 19.6% respectively. The higher complication rate of transhepatic stenting is at least partially related to an unfavourable selection of patients in this group: failures of endoscopic stenting, high frequency of hilar strictures. The mortality at D 30 was 24%, significantly higher in hilar strictures than in middle CBD strictures (p less than 0.02). A late obstruction of the stent occurred in 43 patients (58%) after an interval of 103 +/- 52 days, and endoscopic retreatment was possible in 65% of cases. The median survival was 153 days in subhilar strictures and 104 days in hilar strictures. These results justify considering the possibility of palliative stenting after failure of a surgical treatment especially in peri-ampullary and middle CBD strictures.