1. Background
Primary sclerosing cholangitis (PSC) is a chronic liver disease characterized by cholestasis, biliary inflammation and duct stricture that will lead to cirrhosis in approximately half of subjects (1-5). There are no curative medical treatments for PSC, and the only definitive treatment for advanced disease is liver transplantation (LT) (1, 6-8). There are two main techniques for biliary reconstruction of the transplanted liver: duct-to-duct (DD) anastomosis, and Roux-en-Y hepaticojejunostomy (RYH). Historically, the success rates of DD anastomosis have been lower than RYH, because of the risk of anastomotic strictures associated with the former technique (9, 10). Given that RYH is associated with several complications, including gastrointestinal bleeding, bacterial overgrowth, cholangitis and biloma formation (11-13), there has therefore existed a trend towards performing DD anastomosis, in suitable PSC patients, to avoid the morbidity of RYH. Furthermore, DD anastomosis has the added advantage of facilitating endoscopic retrograde cholangiography (ERCP) post-LT, if clinically warranted. Given the controversial and conflicting literature surrounding DD anastomosis in PSC, further observational data are necessary to justify this surgical approach.
2. Objectives
The primary objective of this study was to assess the safety and efficacy of DD anastomosis, compared to RYH reconstruction, among adults transplanted for PSC. The secondary objective was to analyze the postoperative resource utilization rates between the two groups.
3. Patients and Methods
From the liver transplant database at the multi organ transplant unit of the University of Western Ontario, London, Canada, all subjects ≥ 18 years old, who underwent primary LT between January 1997 and January 2012 for PSC, were identified and stratified by type of biliary drainage performed at the time of surgery.
All subjects were diagnosed with PSC prior to transplantation and underwent an extensive workup with cross-sectional abdominal imaging and cholangiograms [either ERCP or magnetic resonance cholangiopancreatography (MRCP)]. Subjects who underwent LT had either end-to-end DD anastomosis or RYH reconstruction. The DD anastomosis was performed in select cases, where the recipient extrahepatic duct looked grossly normal or the recipient had previous surgery or extensive bowel edema that precluded the safe creation of a Roux loop. Among patients who underwent DD anastomosis, end-to-end, rather than end-to-side, is the preferred method of anastomosis at our center, as it facilitates therapeutic intervention by ERCP, if required.
3.1. Data Collection
Following approval by the Institutional Review Board at University of Western Ontario, London, Canada, baseline clinical and demographic data were collected for each subject, using hospital records. Recipient and graft survival, postoperative medical and surgical complications, and postoperative health resource utilization rates were compared between the two groups.
3.2. Statistical Analysis
Categorical variables were reported as frequencies and percentages, and continuous variables were reported as means with standard deviations (SDs) or medians. Differences between groups were examined using the t test, or Wilcoxon two-sample test, where appropriate, for continuous variables and by the χ2 test, or Fisher’s exact test, where appropriate, for categorical variables. All statistical tests were two-sided and differences were considered significant when P < 0.05. Statistical analyses were performed using SAS Version 9.1.2 (SAS Inc., Cary, NC, USA).
4. Results
Totally, 73 adult patients underwent primary LT for PSC. Of them, 58 patients (79.4%) had RYH and 15 patients (20.5%) had DD reconstruction. A total of 53 patients were male (73%) and the mean ± SD age at LT was 43.3 ± 14.4 years. Baseline recipient characteristics are summarized in Table 1. The mean length of hospital stay in intensive care unit was shorter in the RYH compared to DD group (4.8 days vs. 9.9 days, P = 0.06). However, the total length of hospital stay was similar between the two groups. There was no statistically significant differences seen in recipient death (17.2% vs. 13.3%, P > 0.999), graft failure (22.4% vs. 20%, P > 0.999), biliary complications (8.6% vs. 13.3%, P = 0.627), need for reoperation (27.6% vs. 33.3%, P = 0.751), readmission (50% vs. 26.7%, P = 0.148) or retransplantation (14% vs. 0%, P = 0.191), between the RYH and the DD groups. More cholangiograms (through ERCP or MRCP) were needed post LT for DD patients, compared to RYH patients (33.3% vs. 8.6%, P = 0.026), although there were otherwise no statistically significant differences between the two groups in postoperative complications or resource utilization after LT. Two patients developed anastomotic biliary stricture in the DD group. In the RYH group, two patients were diagnosed with bile leak, one patient had a non-anastomotic biliary stricture, one patient was diagnosed with biliary sump syndrome, and one patient was diagnosed with bile leak by hepatobiliary iminodiacetic acid scan (Table 2). The most common identifiable cause of death post-LT in RYH group was cancer. Causes of death are listed in Table 3. There was no statistically significant difference in rate or causes of graft failure, between the two groups. The two most common causes of graft failure were PSC recurrence and chronic rejection (Table 4). Eight patients required retransplantation; one of whom underwent LT thrice, after loss of the first two grafts for chronic ductopenic rejection and hepatic artery thrombosis.
Table 1. Baseline Characteristics for Recipients a
| Characteristics | Roux-en-Y (n = 58) | DD (n = 15) | P Value |
|---|---|---|---|
| Mean age, y b | 59.0 (47 - 81) | 58.0 (30 - 76) | 0.973 |
| Male Gender c | 39 (67.2) | 14 (93.3) | 0.054 |
| MELD score b | 18.0 (8-39) | 17.0 (10 - 29) | 0.904 |
| Inflammatory bowel disease c | |||
| Ulcerative colitis | 26 (44.8) | 11 (73.3) | > 0.999 |
| Crohn’s disease | 10 (17.2) | 2 (13.3) | 0.081 |
| Comorbidities c | |||
| Cholangiocarcinoma | 2 (3.5) | 0 (0.0) | > 0.999 |
| Dialysis | 4 (7.6) | 0 (0.0) | 0.572 |
| Renal Insufficiency | 1 (1.7) | 0 (0.0) | > 0.999 |
| Hypertension | 2 (3.5) | 0 (0.0) | > 0.999 |
| CAD | 1 (1.7) | 0 (0.0) | > 0.999 |
| Diabetes | 3 (5.2) | 1 (6.7) | > 0.999 |
aAbbreviation: CAD, Coronary Artery Disease; DD, Duct-to-duct; MELD, Model for End-Stage Liver Disease.
b Data are presented as median (min-max)
c Data are presented as No. (%).
Table 2. Outcomes Post-Liver Transplantation for Primary Sclerosing Cholangitis by Biliary Reconstruction a
| Post-LT Outcome | Overall (n = 73) | Roux-en-Y (n = 58) | Duct-to-Duct (n = 15) | P Value |
|---|---|---|---|---|
| Recipient Death b | 12 (16.4) | 10 (17.2) | 2 (13.3) | > 0.999 |
| Graft Failure b | 16 (21.9) | 13 (22.4) | 3 (20.0) | > 0.999 |
| Biliary leak or stricture b | 7 (9.6) | 5 (8.6) | 2 (13.3) | 0.627 |
| Reoperation b | 21 (28.8) | 16 (27.6) | 5 (33.3) | 0.751 |
| Readmission b | 33 (45.2) | 29 (50.0) | 4 (26.7) | 0.148 |
| Re-transplantation b | 8 (11.1) | 8 (14.0) | 0 (0.0) | 0.191 |
| Need for Cholangiogram b | 10 (13.7) | 5 (8.6) | 5 (33.3) | 0.026 |
| LOS, d, Median | 14.0 | 15.0 | 13.5 | 0.670 |
| ICU LOS, days, Median | 3.0 | 2.5 | 6.0 | 0.070 |
| Follow-up, mon, Median | 133.0 | 145.5 | 122.0 | 0.128 |
a Abbreviations: Post-LT, Post-Liver Transplantation.
b Data are presented as No. (%).
Table 4. Causes of Graft Failure a
| Roux-en-Y | DD | |
|---|---|---|
| Graft failure | 13 (22.4) | 3 (20.0) |
| PSC recurrence | 8 (13.8) | 1 (6.7) |
| Chronic ductopenic rejection | 2 (3.5) | 1 (6.7) |
| Hepatic artery thrombosis | 1 (1.7) | 0 (0.0) |
| Primary non-function | 1 (1.7) | 0 (0.0) |
| Unknown | 1 (1.7) | 1 (6.7) |
a Abbreviations: DD, Duct-to duct; PSC, primary sclerosing cholangitis.
The donor characteristics and incidence of vascular complications post-LT were not significantly different in the DD group, compared to the RYH group (Tables 5 and 6). Only one patient developed intrahepatic disease in the RYH group, secondary to hepatic artery stenosis and prolonged cold ischemic time. None of the patients had cytomegalovirus CMV disease or blood group incompatibility from liver donors.
Table 5.Vascular Complications a
| Roux-en-Y | DD | P Value | |
|---|---|---|---|
| Portal vein thrombosis | 2 (3.5) | 1 (6.7) | 0.504 |
| Hepatic artery thrombosis | 1 (1.7) | 0 (0.0) | > 0.999 |
| Hepatic artery stenosis | 3 (5.2) | 0 (0.0) | > 0.999 |
a Abbreviation: DD, duct-to-duct.
Table 6.Donor Characteristics
| Characteristics | Roux-en-Y (n = 58) | DD (n = 15) | P Value |
|---|---|---|---|
| Mean age, y a | 39.0 (3 - 69) | 58.0 (21 - 76) | < 0.001 |
| Male Gender, No. (%) | 35 (65.3) | 9 (60.0) | > 0.999 |
| Type of donor, No. (%) | > 0.999 | ||
| Donation after cardiac death | 51 (87.9) | 14 (21.5) | |
| Donation after brain death | 2 (3.5) | 0 (0.0) | |
| Living donor | 5 (8.6) | 1 (6.7) | |
| Cold ischemic time, h a | 6.7 (1 - 13) | 6.7 (2 - 13) | 0.838 |
| Rewarm ischemic time, min a | 53.0 (2 - 45) | 45.0 (12 - 70) | 0.173 |
a Data are presented as median (min-max)
5. Discussion
The RYH has historically been the method of choice for hepatic transplantation for PSC. The theorized basis for this surgical preference is that the distal common bile duct of the recipient may be diseased from chronic PSC, and as such a DD anastomosis may increase the risk of biliary sequels and even cholangiocarcinoma following LT (14).
In this study, the incidence of biliary leak or biliary stricture were not statistically different in the RYH group, compared to the DD group (8.6% versus 13.3%, P = 0.95), therefore challenging the historical convention of RYH as the default anastomotic technique in PSC recipients undergoing LT. This finding is corroborated in other, albeit small, single-center experiences. For instance, in a retrospective study of 53 PSC patients who underwent LT, Esfeh et al. also found that the biliary complication rate was not significantly higher in the DD group, compared to RYH group (11% vs. 4%, P = 0.32) (14). Likewise, in a study by Distante et al. the incidences of biliary stricture and biliary leak were not significantly higher in DD patients compared to RYH patients (19% vs. 10% and 6% vs. 20%, P value non-significant) (15). In another publication by Heffron et al. involving 60 PSC patients, who underwent LT, 22 of whom had DD reconstruction, there was also no significant difference in the rates of anastomotic stricture or biliary leak (16). In a UK liver transplant registry study of 264 PSC patients, who underwent LT, including 264 subjects with RYH and 98 with DD anastomosis, the incidence of biliary leak was not significantly higher in the DD group, compared to the RYH group (7% vs. 4%, P = 0.26), although biliary strictures occurred more often in the DD group (8% vs. 2%, P = 0.05) (9).
Graft failure and recipient death were not significantly different in the DD group, compared to the RYH group, in the present study. This finding conflicts with the UK liver transplant registry database study of PSC recipients, where the mean graft survival was 85 months in the RYH group, compared to the 74 months of the DD group (P = 0.034) (9). However, the causes of death and a detailed analysis of survival differences between the groups were not clear in the UK study, preventing any meaningful conclusions to be drawn (9). Other publications, such as that by Damrah et al. have determined no survival differences when patients were stratified by biliary anastomotic technique (17).
There are several limitations to our study. First, the present study was a retrospective cohort analysis, with a small sample size, raising the possibility of a type 1 error. That being said, the sample size of our study is larger than other published case series, and our study also has the strength of long-term follow-up across different eras for transplantation. Reliable long-term follow-up data is particularly lacking in registry-based studies. While our study may also be limited in external validity due to data derived from a single-center, our findings appear to coincide with multiple published series (10, 18). It should be noted in our study that DD anastomosis was performed in select cases, where the recipient extrahepatic duct looked grossly normal or the recipient had previous surgery or extensive bowel edema that precluded safe creation of a Roux loop and this would be an interesting area for future research.
In conclusion, DD anastomosis represents an efficacious and safe method in selected PSC patients who undergo LT, compared to RYH, and the overall clinical outcomes appear comparable between the two techniques.
