Orthotopic liver transplantation (OLTx*) is the treatment of choice for end-stage liver disease. The introduction of potent immunosuppressive agents and the improvement in organ preservation and surgical techniques have increased overall allograft and patient survival. However, biliary tract complications continue to be a major cause of morbidity and mortality among liver transplant recipients. The biliary tract anastomosis has been referred to as the Achilles heel of orthotopic liver transplants(1, 2). In large clinical studies, the incidence of biliary tract complications has been reported to range between 10% and 50%, although the true incidence may be underestimated as a result of early posttransplant deaths (3, 4). The percentage of patients who experience biliary tract complications will probably continue to increase as a function of improved patient and allograft survival(5-7).
Biliary tract complications fall into two main categories: leak at the anastomosis and/or stent exit site and stricture, either anastomotic or intrahepatic. End-to-end choledochocholedochostomy (CDCD) using a T tube evolved as the standard method for biliary drainage. When the native and/or allograft ducts are unable to be used for CDCD, Roux-en-Y choledochojejunostomy (CDJ) is the preferred approach (8). Placement of a T tube provides access for performing biliary imaging procedures and monitoring liver allograft function, can be utilized to control and treat an early anastomotic leak, may provide a better flow of bile, and may prevent late anastomotic stricture (9, 10). More recently, the routine use of a T tube has been challenged(11). Rolles et al. (12) and Randall et al. (13) have recently reported that routine T-tube stenting of the OLTx biliary anastomosis is unjustified. The risks associated with routine T-tube placement were identified as added operative time, patient discomfort, the risk of T-tube exit site leak, and the increased cost of imaging procedures. Using a different anastomotic technique, Neuhaus and co-workers (14) reported a remarkably low rate (2.3%) of biliary complications with the novel biliary reconstruction of side-to-side CDCD including placement of a T tube. Based on these reports, we combined these two approaches using side-to-side CDCD without T-tube stenting. The objectives of this retrospective study were to evaluate the incidence of extra-hepatic biliary tract complications using this new method of side-to-side choledochocholedochostomy without a T tube (CDCD SS w/o T tube) and to compare the results of this technique with our historical results using end-to-end CDCD with and without T tubes.
PATIENTS AND METHODS
The study population consisted of 279 consecutive adult OLTx that were performed in 268 patients at the Oregon Health Sciences University and the Portland Veterans Affairs Medical Center, Portland, OR, between September 1991 and June 1996. All transplants were performed with ABO-compatible organs. Two hundred sixty-eight were primary transplants and 11 were first retransplants(4%). The indications for transplantation were hepatitis C combined with alcoholic cirrhosis (n=60), alcoholic cirrhosis (n=58), hepatitis C (n=43), cryptogenic cirrhosis (n=29), primary biliary cirrhosis (n=25), primary sclerosing cholangitis (n=21), autoimmune chronic active hepatitis (n=8), hemochromatosis (n=6), α1-antitrypsin deficiency (n=5), hepatitis B(n=5), Wilson's disease (n=3), drug-in-duced fulminant hepatic failure (n=2), Alagille's syndrome (n=1), and cholangiocarcinoma (n=1). The indications for the 11 re-transplants were recurrent hepatitis C (n=4), primary allograft nonfunction (n=3), chronic rejection (n=2), and hepatic arterial thrombosis(n=2).
Forty-eight transplants were excluded from the analysis. Twenty-seven of these transplants were primary CDJ for their biliary tract reconstruction. The major indications for CDJ were primary sclerosing cholangitis, small caliber recipient bile duct, and retransplantation. Two additional patients were excluded for unconventional anastomoses: side-to-side CDCD with T tube and side-to-end CDCD. Nineteen patients were excluded due to death within 30 days after OLTx (30-day mortality=7%). Three additional transplant recipients were excluded due to allograft primary nonfunction and allograft loss within 30 days after transplantation. The resulting study population for final analysis consisted of 227 OLTx with primary CDCD in 220 recipients with a minimum follow-up of 6 months after transplantation.
Donor hepatectomy, bile duct anastomosis, recipient operation, and management. All donor organs were obtained from heartbeating donors, perfused in situ, and preserved in University of Wisconsin solution (Via Span, DuPont Corp., Wilmington, DE). Donor age ranged from 8 to 69 years, and cold ischemia times ranged from 2.45 to 30.2 hr (mean, 9.6 hr). The biliary tree in all cases was flushed with University of Wisconsin solution immediately after removal of the organ from the cadaver.
The recipient procedure involved reperfusion of the allograft through the portal vein before reconstruction of the arterial inflow. Veno-venous bypass was rarely used. The biliary anastomosis in the end-to-end procedure was accomplished after trimming the transected ends of the native and allograft common hepatic/common bile ducts until arterial bleeding was encountered. These anastomoses subsequently were performed with interrupted 6-0 polyglyconate absorbable sutures. When a stent was used, a no. 8 or no. 10 T tube with multiple side holes in the T-portion was brought out through a separate opening in the side of the native common bile duct. Additional interrupted 6-0 polyglyconate sutures were used to close the exit site around the tube. The end of the T tube was then brought out of the right upper quadrant of the abdomen through a separate stab wound above the level of the incision (Fig. 1). A routine intraoperative cholangiography was performed to assess for immediate patency of the allograft and native bile ducts as well as to verify the absence of leaks. Repeat cholangiography was performed before T-tube closure on postoperative day 6 or 7. Side-to-side CDCD was performed by incising the opposing anterior and posterior surfaces of the common hepatic/common bile ducts of the native and allograft biliary tree using a fistulectomy excisor. The ends of the ducts were then ligated with single 3-0 silk sutures. Running 5-0 polyglyconate absorbable suture material was then used to create a 1.2- to 1.5-cm anastomosis (Fig. 2).
The posttransplantation immunosuppressive protocol consisted of antilymphocyte induction with either Minnesota antilymphocyte globulin or with antithymocyte immunoglobulin (Atgam, Upjohn, Kalamazoo, MI) at 15 mg/kg/day for 5 days and continued until therapeutic cyclosporine (Sandimmune, Novartis, Hanover, NJ) blood levels were achieved. The dose of cyclosporine was adjusted according to the whole blood cyclosporine concentration trough level on oral therapy to a steady state of 250 μg/L (Incstar radioimmuno-assay, whole blood). The dosage was further adjusted based on clinical evidence of toxicity(nephrotoxicity, neurotoxicity, and/or hypertension) and efficacy(rejection).
Methylprednisolone (Solu-Medrol, Upjohn, Kalamazoo, MI), at 1000 mg, was given intraoperatively, followed by 2 mg/kg/day of methylprednisolone or prednisone from the day of surgery, and tapered at 3-day intervals to 5 mg by the end of the fourth month after transplantation. Azathioprine (Imuran, Glaxo Wellcome, Research Triangle, NC) at 1 mg/kg/day i.v. or 2 mg/kg/day p.o., with dosing reduced or discontinued for leukopenia (defined as white blood cell count <3000 cells/mm [2]), was also administered for maintenance therapy.
Routine liver biopsy was performed at weeks 1, 2, and 52 and when hepatic allograft dysfunction or serum chemistries indicated. Acute cellular allograft rejection was confirmed histologically with allograft biopsy. Rejection episodes were treated with intravenous methylprednisolone and followed with allograft re biopsy. Steroid-resistant rejections were treated with OKT3(Muromonab CD3, Ortho Pharmaceutical, Raritan, NJ) at 5 mg i.v. for 10 days and a repeat prednisone taper schedule. All patients received trimethoprim/sulfamethoxazole and clotrimazole troches as prophy-laxis forPneumocystis carinii and oral candidal infection, respectively. Intravenous ganciclovir followed by oral acyclovir was used for cytomegalovirus prophylaxis.
Data were collected regarding method of biliary reconstruction, type of biliary tract complication, treatment intervention for biliary tract complication, patient sex, age, etiology and indication for transplantation, cold and warm ischemia time, United Network for Organ Sharing (UNOS) status at the time of transplantation, number of HLA mismatches, rejection, and allograft and patient survival from a computerized liver transplant database and a review of hospital and clinic records. The chi-square test was used to compare categorical variables. The SPSS version 6.1 computer system (SPSS, Inc., Chicago, IL) was used to conduct these analyses. The level of statistical significance was set at a P-value of less than 0.05.
RESULTS
The study population for analysis consisted of 227 OLTx with CDCD in 220 recipients. These 227 CDCDs were distributed as follows: 124 end-to-end CDCD with T tube (CDCD EE with T tube) in 118 patients, 44 end-to-end CDCD without T tube (CDCD EE without T tube) in 44 patients, and 59 side-to-side CDCD without T tube (CDCD SS w/o T tube) in 58 patients. One hundred fifty-three patients were male (71%) and 67 patients were female (29%), ranging in age from 16 to 69 years (mean age, 47.9 years). The mean cold and warm ischemia times were 9.8 hr and 0.9 hr, respectively. Fourteen patients were UNOS status 1 (6%), 26 patients were status 2 (12%), and 180 patients were status 3 (82%) at the time of transplantation. The mean number of HLA mismatches between donor and recipient was 4.5.
Demographic data on the 220 patients according to the biliary anastomotic technique used are summarized in Table 1. All three groups were similar in mean age, gender distribution, total ischemia time, UNOS status at the time of transplantation, and number of HLA mismatches among donor and recipient. No statistically significant differences in these demographic characteristics were found among the different CDCD anastomotic groups.
Biliary tract complications. Sixty-nine biliary tract complications were observed in 227 CDCD anastomoses (30%). These results are summarized in Table 2. A total of 43 patients (19%) experienced anastomotic and/or T-tube exit site-leak. Three patients (2%) in the CDCD EE with T tube group experienced anastomotic leaks, all of which were noted within the initial posttransplant hospital stay at the time of routine cholangiography. These three anastomotic leaks were treated aggressively with operative revision to CDJs. Four additional patients were noted to have minor extravasation of contrast at the T-tube exit site at the time of initial postoperative cholangiography. These small leaks were considered to be due in part to overinjection of contrast during performance of the cholangiogram. All four of these patients had continued external bag drainage for an additional week and subsequently underwent repeat cholangiography. All four of these T-tube exit site leaks were noted to have sealed, and their T tubes were closed at that time. These four leaks were not included in the statistical analysis because they did not prolong the posttransplant hospitalization or recovery and were not considered to be morbid events for the patients. One additional patient dislodged his T tube at 6 weeks after transplantation and underwent reoperation to remove the T tube and close the exit site. This early exit site leak is included in the analysis as a reoperation. Thirty-two (26%) additional patients in the CDCD EE with T tube group experienced leaks at the T-tube exit site from the native common bile duct that resulted in bile peritonitis. These additional exit site leaks occurred at the time of T-tube removal at 6 months after transplantation and required rehospitalization and antibiotics. One patient (2%) in the CDCD EE w/o T tube group and six patients(10%) in the CDCD SS w/o T tube group also experienced anastomotic leaks early in their posttransplant course. These seven patients underwent reoperation for anastomotic revision to CDJ (Table 3).
Focusing strictly on the anastomotic leaks, only 10 patients (8%) experienced anastomotic leak: three (2%) in the CDCD EE with T tube group, one(2%) in the CDCD EE without T tube group, and six (10%) in the CDCD SS w/o T tube group. Therefore, the incidence of anastomotic leak was higher in the CDCD SS w/o T tube group than in the CDCD EE w/o T tube and CDCD EE with T tube groups (10% vs. 2%). These anastomotic leaks were all identified early after transplantation.
Anastomotic strictures were found in 26 patients (12%): 7 patients (6%) with CDCD EE with T tube, 10 patients (23%) with CDCD EE w/o T tube, and 9 patients (15%) with CDCD SS w/o T tube. The incidence of anastomotic stricture was statistically lower in the CDCD EE with T tube group compared with the CDCD without T tube groups (both end to end and side to side)(P<0.05).
Treatment intervention. All patients with a biliary anastomotic or stent exit site leak were hospitalized and treated with intravenous antibiotics. When leaks were identified early after transplantation and were found to be clinically significant regardless of the site (anastomotic vs. exit site), they were treated operatively, with all but the one patient undergoing conversion of the CDCD to a CDJ anastomosis. Late exit site leaks following T-tube removal were initially treated noninvasively. Evidence of persistent leak either from protracted ileus and/or additional cholangiography prompted stent placement via either endoscopic retrograde cholangiopancreatography or percutaneous transhepatic cholangiography.
Strictures, when identified, were initially addressed non-operatively with stent placement, again either by endoscopic retrograde cholangiopancreatography or percutaneous transhepatic cholangiography. Inadequate patency or recurrence of the stricture after stent removal following a 3- to 6-month trial of stenting prompted operative revision of the CDCD to CDJ. Ultimately, 40 patients (18%) required operative or nonoperative biliary tract interventions to address the biliary complications of leak and/or anastomotic stricture (Table 3).
Definitive treatment of the leak and/or stricture was accomplished in 13 patients (6%) with endoscopic or percutaneous stents. Twenty-seven patients(12%) ultimately required surgery. Fourteen (6.3%) of the operated patients failed initial nonoperative intervention due to inadequate control of a leak or inadequate cannulation of the duct. Twelve (5.4%) other patients with strictures initially treated nonoperatively ultimately underwent surgical revision due to recurrence of the anastomotic stricture after removal of the stent. The patients requiring surgery comprised 10 patients (8%) in the CDCD EE with T tube group, 8 (17%) in the CDCD EE w/o T tube group, and 9 (15%) in the CDCD SS w/o T tube group. Therefore, although the total number of biliary anastomotic complications was significantly higher in the CDCD EE with T tube group (P<0.05), the overall number of biliary tract interventions as well as the number of operative revisions was significantly lower when a T tube was used (P<0.05).
Overall hospitalization was increased in the groups not using a T tube due to increased operative interventions. Those patients who underwent CDJ conversion had a median hospital length of stay of 12 days (range, 7-28 days). For patients hospitalized and treated without an operation for bile peritonitis from a leak, the median length of hospitalization was only 5.5 days (range, 2-12 days).
Survival. Patient survival was 83% at 1 year. The overall mortality rate was not significantly different among the different CDCD anastomotic groups. However, the median length of follow-up was substantially longer in the CDCD EE with T tube group, as this historically was our standard technique before implementation of the more recent CDCD techniques (3.5 vs. 2.0 vs. 1.0 years). Mortality within the first 6 months was significantly less in the patients using T tubes (P<0.05) (Table 4).
DISCUSSION
Over the last 30 years, investigators have suggested a variety of methods of biliary reconstruction to minimize anastomotic complications, but few have been accepted for clinical practice(1, 4, 9, 12-16). The most common methods of biliary tract reconstruction in OLTx are end-to-end CDCD with or without a T tube and Roux-en-Y hepaticojejunostomy/CDJ(18). The type of biliary anastomosis and the use of a T tube have recently become the subject of debate. In early clinical studies, biliary tract complications following liver transplantation occurred in up to 50% of all patients, with an associated mortality of 25-30%(19-22). Anastomotic strictures resulting in biliary obstruction, biliary lithiasis and/or anastomotic leak, stent site leaks, intrahepatic strictures, and cholangitis account for the majority of biliary tract complications (19). In addition to surgical technique, other factors, such as organ preservation technique, ischemia time, rejection, hepatic artery thrombosis, disease recurrence, and ABO incompatibility, have been demonstrated to influence the incidence of biliary tract complications (23-25). Biliary anastomotic complications following transplantation are initially managed nonoperatively with endoscopic or transhepatic dilation and/or stents(28-31). Persistent stricture or restenosis of a CDCD often requires surgical revision to a Roux-en Y CDJ.
CDCD using a T tube has become the standard biliary anastomosis for liver transplant recipients with healthy native bile ducts(10, 27). The use of a T tube was thought to decease the risk of anastomotic leak or stricture, as well as provide access for evaluation of allograft dysfunction (8, 27). However, the use of a T tube has repeatedly been demonstrated to increase the risk of bile leak. Leaking of bile around the T tube from where it exits the common bile duct or directly from the duct after inadvertent dislodgment of the T tube can occur early after transplantation. The majority of the leaks, however, are documented at the time of T-tube removal due to disruption of the tract that forms around the T tube at the time the T tube is removed(27-32). Certainly, the longer the T tube remains, the greater the chance that there is a well-formed tract. Nevertheless, if the tract is disrupted at the time that the T tube is removed, a leak with subsequent sequelae of bile peritonitis, etc., can ensue irrespective of the interval between transplantation and T-tube removal.
These stent-related complications have recently led to a re-evaluation of the purported benefits of the T tube. Rolles and his colleagues(12) observed a lower incidence of biliary leak when T tubes were not placed. Only 10 patients (11%) experienced anastomotic strictures that were managed with either endoscopic dilation (1 patient), percutaneous dilation (1 patient), or Roux loop reconstruction (8 patients). They concluded that abandoning routine T-tube placement was not detrimental to overall outcome (12).
In a similar study, Randall and his co-workers (13) compared the results of 51 patients who had received CDCD EE w/o T tubes to 59 patients who had received CDCD EE with T tubes (13). No significant difference between the two groups was noted in patient survival, allograft survival, rate of conversion to CDJ (6.7% vs. 7.8%), or other biliary tract complications (13.7 vs. 22%). However, this report documented the earlier posttransplant occurrence of biliary tract complications and earlier requirement for intervention when a T tube was not used compared with the group using a T tube. Despite the earlier requirement for intervention to treat these early complications, their patients in the CDCD EE w/o T tube group received far fewer radiologic studies. The authors concluded that routine use of a T tube is unnecessary in most patients.
A different anastomotic approach was developed by Neuhaus and his co-workers (14). They reported on anastomotic complications in 300 patients using a side-to-side CDCD with routine use of a T tube. Only one anastomotic leak in 300 patients (0.3%) was observed after liver transplantation. Late biliary tract complications (>30 days) were seen in only two patients (0.6%). The side-to-side anastomosis was theorized to protect the biliary tract from mechanical trauma by avoiding the traction on the sutures in an end-to-end anastomosis, thereby decreasing resultant necrosis and leakage (33). Side-to-side CDCD with use of a T tube was recommended as the ideal method for reconstruction of the biliary tract after liver transplantation.
Based on these results and findings from previous investigators, we adopted the new technique of side-to-side CDCD, but without the use of a T tube. The aim of this analysis was to compare our historical results with this new technique. We found that our incidence of biliary tract complications after liver transplantation is still significant despite these refinements in surgical technique. Biliary tract complications were identified in 68 patients(30%): 34% in the CDCD EE with T tube group, 25% in the CDCD EE w/o T tube group, and 25% in the CDCD SS w/o T tube group. The overall incidence of anastomotic biliary complications in our series is greater than that of recent reports in the literature (Table 5) and confirmed that T-tube placement is associated with a significant incidence of T-tube exit site leak. All of the patients with exit site leak were hospitalized for bile peritonitis, and all but one of these patients were promptly discharged(median length of stay, 5.5 days). Therefore, most of these leaks were minor and were treated and resolved without surgery. However, one patient had an unrecognized biliary anastomotic dehiscence after T-tube removal and died at an outlying facility from this complication. Despite our increased incidence of leaks, our rate of interventions when we used a T tube was low, similar to recent reports (12, 13, 34).
We found a significantly higher incidence of biliary stricture and surgical revision to CDJ in the groups without a T tube. Our results are consistent with the more recent report from Nuño and co-workers(34). These investigators reported an increased incidence of biliary leak and stenosis (33% vs. 10%) and a higher incidence of reoperation (31% vs. 6%) after end-to-end CDCD when they did not use a T tube.
In our hands, the side-to-side CDCD did not have a beneficial effect on decreasing biliary anastomotic complications compared with end-to-end CDCD. The major technical difference between the report from Neuhaus et al.(14) and our series is the use of a T tube. Our findings suggest that the biliary anastomosis may benefit from T-tube stenting to prevent leaks and anastomotic strictures along with the attendant morbidity of CDJ conversion. In our series, four patients died directly as the result of biliary tract complications and associated multiorgan failure (1.8%). The mortality was not significantly different among the three biliary anastomotic technique groups. However, we believe the increased number of early biliary anastomotic complications observed in the groups without T tubes contributed to the statistically significant increased mortality within 6 months after transplantation (9% vs. 2%, P<0.05). This may be due to the compromised condition of the patient who has recently undergone liver transplantation. These patients are more susceptible to infections and heal more slowly due to their immunocompromised state. Biliary complications late after transplantation routinely are better tolerated than early complications.
Although the number of hospitalizations for biliary anastomotic complications was higher in the group with a T tube, the cumulative number of hospital days with the associated costs was greater when no T tube was used. Despite other reports of decreased radiologic studies and the associated cost savings when no T tube was used, our own data using a T tube along with the reports from Neuhaus et al. (14) and Nuño et al.(34) suggest that the use of a T tube may be superior in a comprehensive cost consequence model(12-14, 34, 35).
In summary, the objective of this retrospective study was to evaluate the incidence of biliary tract complications using the new technique of side-to-side CDCD without T tube and to compare these results with end-to-end CDCD with and without T tube. The highest incidence of biliary tract complications was observed in CDCD EE with T tube. This included a low incidence of early anastomotic leak but a high incidence of leak from the T-tube exit site at the time of T-tube removal. Patients who experienced the T-tube exit site leaks were readily treated with a short hospitalization. The lowest anastomotic stricture and subsequent intervention rate was also seen with CDCD EE with T tube. Our data demonstrate a significantly higher incidence of biliary stricture and operative repair (CDJ conversion) as well as a higher 6-month mortality rate in CDCD patients without T tubes. In addition, the placement of T tubes was found to decrease the total length of hospitalization and likely the costs associated with transplantation as a function of a decreased need for operative and nonoperative intervention. CDCD SS w/o T tube did not offer significant advantages over conventional biliary anastomotic techniques.
Footnotes
Presented at the 23rd Annual Meeting of the American Society of Transplant Surgeons, May 14-16, 1997, Chicago, IL.
REFERENCES
1. Calne RY, McMaster P, Portmann B, et al. Observations on preservation, bile drainage and rejection in 64 human orthotopic liver allografts. Ann Surg 1977; 186: 282.
2. Starzl TE, Putnam CW, Hansbrough JF, et al. Biliary complications after liver transplantation: with special reference to the biliary cast syndrome and techniques of secondary duct repair. Surgery 1977; 81: 212.
3. Lerut J, Gordon RD, Iwatsuki S, et al. Biliary tract complications in human orthotopic liver transplantation. Transplantation 1987; 43: 47.
4. Calne RY. A new technique for biliary drainage in orthotopic liver transplantation utilizing the gallbladder as a pedicle graft conduit between the donor and recipient common bile ducts. Ann Surg 1976; 184: 605.
5. Klein AS, Savader S, Burdick JF, et al. Reduction of morbidity and mortality from biliary complications after liver transplantation. Hepatology 1991; 14: 818.
6. Lopez RR, Benner KG, Ivancev K, Keeffe EB, Deveney CW, Pinson CW. Management of biliary complications after liver transplantation. Am J Surg 1992; 163: 519.
7. Greif F, Bronsther OL, Van Thiel DH, et al. The incidence, timing, and management of biliary tract complications after orthotopic liver transplantation. Ann Surg 1994; 219: 40.
8. Langnas AN, Marujo W, Stratta RJ, Wood RP, Shaw BW Jr. Vascular complications after orthotopic liver transplantation. Am J Surg 1991; 161: 76.
9. O'Connor TP, Lewis WD, Jenkins RL. Biliary tract complications after liver transplantation. Arch Surg 1995; 130: 312.
10. Hiatt JR, Quinones-Baldrich WJ, Ramming KP, Brems J, Busuttil RW. Operations upon the biliary tract during transplantation of the liver. Surg Gynecol Obstet 1987; 165: 89.
11. Rouch DA, Emond JC, Thistlethwaite JR Jr, Mayes JT, Broelsch CE. Choledochocholedochostomy without a T tube or internal stent in transplantation of the liver. Surg Gynecol Obstet 1990; 170: 239.
12. Rolles K, Dawson K, Novell R, Hayter B, Davidson B, Burroughs A. Biliary anastomosis after liver transplantation does not benefit from T tube splintage. Transplantation 1994; 57: 402.
13. Randall HB, Wachs ME, Somberg KA, et al. The use of the T tube after orthotopic liver transplantation. Transplantation 1996; 61: 258.
14. Neuhaus P, Blumhardt G, Bechstein WO, Steffen R, Platz KP, Keck H. Technique and results of biliary reconstruction using side-to-side choledochocholedochostomy in 300 orthotopic liver transplants. Ann Surg 1994; 219: 426.
15. Bismuth H, Castaing D, Gugenheim J, Traynor O, Ciardullo M. Roux-en-Y hepatojejunostomy: a safe procedure for biliary anastomosis in liver transplantation. Transplant Proc 1987; 19: 2413.
16. Anselmi M, Sherlock D, Buist L, et al. Gallbladder conduit vs end-to-end anastomosis of the common bile duct in orthotopic liver transplantation. Transplant Proc 1990; 22: 2295.
17. Evans RA, Raby ND, O'Grady JG, et al. Biliary complications following orthotopic liver transplantation. Clin Radiol 1990; 41: 190.
18. Porayko MK, Kondo M, Steers JL. Liver transplantation: late complications of the biliary tract and their management. Semin Liver Dis 1995; 15: 139.
19. Iwatsuki S, Shaw BW Jr, Starzl TE. Biliary tract complications in liver transplantation under cyclosporin-steroid therapy. Transplant Proc 1983; 15: 1288.
20. Sherman S, Jamidar P, Shaked A, Kendall BJ, Goldstein LI, Busuttil RW. Biliary tract complications after orthotopic liver transplantation: endoscopic approach to diagnosis and therapy. Transplantation 1995; 60: 467.
21. Stratta RJ, Wood RP, Langnas AN, et al. Diagnosis and treatment of biliary tract complications after orthotopic liver transplantation. Surgery 1989; 4: 675.
22. Ringe B, Oldhafer K, Bunzendahl H, Bechstein WO, Kotzerke J, Pichlmayr R. Analysis of biliary complications following orthotopic liver transplantation. Transplant Proc 1989; 21: 2472.
23. Sanchez-Urdazpal L, Batts KP, Gores GJ, et al. Increased bile duct complications in liver transplantation across the ABO barrier. Ann Surg 1993; 218: 152.
24. Dousset B, Conti F, Houssin D, Calmus Y. Acute vanishing bile duct syndrome after interferon therapy for recurrent HCV infection in liver-transplant recipients [Editorial]. N Engl J Med 1994; 330: 1160.
25. Valente JF, Alonso MH, Weber FL, Hanto DW. Late hepatic artery thrombosis in liver allograft recipients is associated with intrahepatic biliary necrosis. Transplantation 1996; 61: 61.
26. Tzakis AG, Gordon RD, Shaw BW Jr, Iwatsuki S, Starzl TE. Clinical presentation of hepatic artery thrombosis after liver transplantation in the cyclosporine era. Transplantation 1985; 40: 667.
27. Krom RAF, Kingma LM, Wesenhagen H, Slooff MJH, Haagsma EB, Gips CH. Choledochocholedochostomy is successful in orthotopic liver transplantation. Transplant Proc 1984; 16: 1228.
28. Ward EM, Wiesner RH, Hughes RW, Krom RAF. Persistent bile leak after liver transplantation biloma drainage and endoscopic retrograde cholangiopancreatographic sphincterotomy. Radiology 1991; 179: 719.
29. Donovan J. Nonsurgical management of biliary tract disease after liver transplantation. Gastroenterol Clin N Am 1993; 22: 317.
30. Zajko AB, Campbell WL, Bron KM, et al. Cholangiography and interventional biliary radiology in adult liver transplantation. AJR 1985; 144: 127.
31. Osorio RW, Freise CE, Stock PG, et al. Nonoperative management of biliary leaks after orthotopic liver transplantation. Transplantation 1993; 55: 1074.
32. Ramirez P, Parrilla P, Bueno FS, et al. Reoperation for biliary tract complications following orthotopic liver transplantation. Br J Surg 1993; 80: 1426.
33. Neuhaus P, Brölsch C, Ringe B, Lauchart W, Pichlmayr R. Results of biliary reconstruction after liver transplantation. Transplant Proc 1984; 16: 1225.
34. Nuño J, Vicente E, TurriĂ³n VS, et al. Biliary tract reconstruction after liver transplantation: with or without T-tube? Transplant Proc 1997; 29: 564.
35. Stergachis A. Overview of cost-consequence modeling in outcome research. Pharmacotherapy 1995; 15: 405.
