OUP user menu

Journal of Crohn's and Colitis: 10 (8)

Editor-in-Chief

Laurence J. Egan, Ireland

Associate Editors

Shomron Ben-Horin, IsraelSilvio Danese, ItalyPeter Lakatos, HungaryMiles Parkes, UKJesús Rivera-Nieves, USABritta Siegmund, GermanyGijs van den Brink, NLSéverine Vermeire, Belgium

6.585
5.586

Published on behalf of

Impact of ileal pouch-anal anastomosis on the surgical outcome of orthotopic liver transplantation for primary sclerosing cholangitis

Emmanuel C. Obusez, Lei Lian, Zhuo Shao, Udayakumar Navaneethan, Robert O'Shea, Ravi P. Kiran, Bo Shen
DOI: http://dx.doi.org/10.1016/j.crohns.2012.06.001 230-238 First published online: 1 April 2013

Abstract

Background: The definitive treatment for patients with primary sclerosing cholangitis (PSC) is orthotopic liver transplantation (OLT), while the surgical treatment of choice for UC is restorative protocolectomy with ileal pouch-anal anastomosis (IPAA). While studies to date show that OLT may impact the outcome of IPAA, the effect of IPAA on the surgical outcome of OLT is not known.

Methods: All eligible patients (those with PSC and OLT) from our prospectively maintained OLT and Pouch Databases were included. Patient and OLT graft survivals along with surgical outcomes were assessed. Univariable and multivariable analyses were performed.

Results: Seventy-nine patients with OLT for PSC were studied, including those with UC (PSC + OLT + UC, n = 27) or without UC (PSC + OLT, n = 30), and with UC and IPAA (PSC + OLT + UC + IPAA, n = 22). In the PSC + OLT + UC group, 23 (85.2%) had UC before OLT and 4 (14.8%) had UC diagnosed after OLT. In the PSC + OLT + UC + IPAA group, 9 (40.9%) had IPAA-then-OLT and 13 (59.1%) had OLT-then-IPAA, while 21 (95.5%) had UC before OLT and 1 (4.5%) had UC diagnosed after OLT. Kaplan–Meier survival curve showed no statistical differences in patient and graft survivals between the 3 groups. In univariable analysis, there was no statistical difference for acute and chronic rejection, hepatic artery thrombosis, stricture, bile leak and acute and chronic renal failure for the 3 groups. In multivariable analysis, no factors were found to be associated with bacteremia or intra-abdominal abscess.

Conclusions: The presence of the IPAA in OLT for PSC patients appears not to have a negative impact on patient and graft survivals and post-operative complications.

Keywords
  • Inflammatory bowel disease
  • Ileal pouch-anal anastomosis
  • Orthotopic liver transplantation
  • Primary sclerosing cholangitis
  • Restorative proctocolectomy
  • Ulcerative colitis

1 Introduction

Primary sclerosing cholangitis (PSC) is an ideopathic chronic disorder characterized by progressive inflammation, fibrosis, and stricturing of intrahepatic and extrahepatic bile ducts. Long-term complications of PSC include cirrhosis, liver failure, and cholangiocarcinoma.1 The definitive treatment for advanced PSC with cirrhosis is orthotopic liver transplantation (OLT).2 On the other hand, PSC represents the most common hepatobiliary manifestation of inflammatory bowel disease (IBD).3 In particular, ulcerative colitis (UC) is strongly associated with PSC, with approximately 75% of PSC patients having concomitant UC and 2–11% of UC patients having concurrent PSC.4,5 Restorative proctocolectomy with ileal pouch-anal anastomosis (IPAA) is the surgical treatment of choice in patients with medically refractory UC or UC with neoplasia.6 While the impact of OLT surgery on the natural course of the ileal pouch has been studied,7,8 the effect of the IPAA procedure on the outcome of OLT has not been investigated.

Patient and graft survivals of liver transplantation for PSC have been excellent, with improved rejection rate, reduced incidence of hepatic artery thrombosis, low retransplantation and low recurrent PSC due to better immunosuppressive therapy and higher standards of postoperative management.9,10 The majority of PSC patients in the Western world have concomitant UC and approximately 1/4–1/3 of UC patients would eventually need restorative proctocolectomy. Whether restorative proctocolectomy with IPAA has impact on the outcome of OLT is not known. We hypothesized that restorative proctocolectomy with bowel reconstruction and bacterial overload may increase the risk for post-OLT infectious complications and graft survival. The aim of this study was to evaluate the impact of the IPAA on the post-operative outcomes for patients who underwent OLT for PSC and the outcomes of patients with IPAA and OLT.

2 Patients and methods

2.1 Patients

This single-center historical cohort study was approved by the Cleveland Clinic Institutional Review Board. All patients with OLT for PSC who had or had no concurrent UC with or without IPAA were identified from our prospectively maintained electronic OLT Database and Pouchitis Registry. Demographic, clinical, and post-operative variables were reviewed and tabulated. In addition to information from the databases, both hard-copy and electronic medical records were reviewed. All eligible patients were included in the study. The patients were divided and compared the 3 groups based on underlying disease entities: PSC + OLT, PSC + OLT + UC, and PSC + OLT + UC + IPAA.

2.2 Inclusion and exclusion criteria

Inclusion criteria were patients who had OLT for PSC with or without concurrent UC. The UC patients may or may not have had IPAA. Exclusion criteria were (1) patients with OLT for advanced liver diseases other than PSC, including hepatocellular cancer, hepatic vascular diseases and other cholestatic diseases; and (2) IPAA for familial adenomatous polyposis (FAP), Lynch syndrome, or sporadic colorectal cancer (CRC). Patients with UC-associated neoplasia, however, were included in the study.

2.3 Demographic and clinical variables

Demographic and clinical variables were studied, including age, gender, ethnicity, body mass index, smoking and alcohol use, and family history of IBD, PSC, or liver/colon cancer in first-degree relatives, concurrent autoimmune disorders, and significant comorbidities (central nervous system [CNS] disease, heart disease and renal disease). In addition, pertinent clinical data on UC were evaluated, including age of UC diagnosis, and pre- and post-proctocolectomy medicines, pre- or peri-operative diagnosis of UC or indeterminate colitis (IC) (a pre-operative diagnosis of UC or IC was determined by colonoscopy with biopsy and/or peri-operative diagnosis was based on intra-operative/immediate post-operative histopathological evaluation of colectomy specimens), extent of colitis, age at colectomy, indication for colectomy, pouch configuration, stage of ileal pouch and post pouch complication (pouchitis, cuffitis, pouch strictures, pouch leaks, pouch bleeding and pouch abscess). Extra-intestinal manifestations (EIM) of IBD, such as associated diseases of skin and joints and thromboembolic events, were assessed. There were overlaps between EIM of IBD and skin, joint, and thromboemolic manifestations in liver disease, we lumped these manifestations as “EIM of IBD”. The sequence of OLT-then-IPAA or IPAA-then-OLT and intervals between the two procedures were recorded.

Pertinent OLT data were evaluated, including age and date of the procedure, indication for OLT, patient and graft survivals, MELD scores, complications of procedure (biliary leak or stricture, acute and chronic renal failure, bacteremia and intra-abdominal abscess and acute and chronic rejection). Data used for the measurement of complications from the procedures were based on the most recent follow-up at the time of data entry. This was applicable to biliary stricture, acute and chronic rejection. For bacteremia and intra-abdominal abscess a 90-day time frame was used. The use of long-term medical therapy including post-OLT antibiotics, corticosteroids, cyclosporine, mycophenolate, sirolimus and tacrolimus was also evaluated with the time frame from the procedure to last follow-up.

2.4 Clinical practice pattern for patients with IPAA and OLT

All IPAA patients received standard mechanical bowel preparation and intravenous amoxicillin 1 g before IPAA surgery. Postoperatively, all patients undergoing OLT for PSC patients were administered trimethoprim/sulfamethoxazole (160 mg trimethoprim and 800 mg sulfamethoxazole) prophylaxis against Pneumocystis. Routine antiviral ganglicyclovir prophylaxis was only administered in seronegative recipients whose liver allograft was from a sero-positive donor. Standard post-OLT prophylaxis regimens included gancyclovir or valganciclovir for the first 3 months after the surgery and trimethoprim–sulfamethoxazole once a day indefinitely. OLT patients routinely received anti-rejection medications (tacrolimus, cyclosporine, low-dose prednisone, or mycophenolate mofetil). Episodes of rejection were treated with high dose-corticosteroids; and steroid-resistant rejection was treated with OKT3.

2.5 Outcome measurement

The primary outcome was the comparison of graft and patient survivals between the study groups. The secondary outcome was the assessment of post OLT infectious and surgical complications, renal failure, acute and chronic rejection, and re-transplantation survival in the study groups.

The 90-day post-OLT complications (infectious or surgical) were recorded in patients with IPAA-then-OLT. In patients with OLT-then-IPAA, post OLT complications were recorded 90 days after the IPAA procedure.

Definitions for infectious complications were adopted from the Center for Disease Control (CDC) criteria.11 Bacteremia was defined as the presence of Staphylococcus aureus or, Candida species or gram negative rods found in at least one blood culture. Pathogens other than the above mentioned were considered positive when isolated from 2 separate blood cultures from areas considered as the infection site. We defined primary bacteremia as bacteremia without clinical, radiological or laboratory evidence of infection source. Catheter-sourced bacteremia was defined when more than 15 colony units were cultured from the catheter tip regardless of whether the same microbe was isolated from blood cultures. Intra-abdominal abscess was defined as intra-abdominal fluid collection on computed tomography imaging or ultrasound with the presence of organisms on aspiration and the presence of fever, abdominal pain, tenderness or elevation of liver function indices consistent with local or systemic infection without source. Bacterial infection was defined as the isolation of bacterial species in cultures from specimens collected from sterile sites or the presence of 105 colony forming bacterial units/ml in urine or sputum bronchoaveolar lavage with one of the following: fever (> 38 °C), chills or hypotension or leukocytosis.

2.6 Statistical analysis

Student's t test was used to determine significance of the difference between the means. Chi-square test was used for comparison between the different groups and Fisher's exact test for smaller numbers. Patient survival was calculated using Kaplan–Meier survival and log rank test to compare survival between the different groups. Multivariable analysis was performed. A P-value less than 0.05 was considered as statistical significance.

3 Results

3.1 Demographic and clinical variables for the cohort

A total of 79 patients were included in the study, consisting of 30 with PSC + OLT, 27 with PSC + OLT + UC, and 22 patients with PSC + OLT + UC + IPAA. The demographic and clinical data of the 79 patients with PSC for OLT are listed in Table 1 . The mean follow-up time for the PSC + OLT + UC + IPAA, PSC + OLT + UC and PSC + OLT groups were 118 ± 61 months, 85 ± 57 months and 118 ± 67 months respectively.

View this table:
Table 1

Univariable comparison of demographic and clinical data.

FactorPSC + OLT (N = 30)PSC + OLT + UC (N = 27)PSC + OLT + UC + IPAA (N = 22)P value
Mean current age, ± SD54.9 ± 14.551.6 ± 13.556.2 ± 10.60.45
Mean age at IBD diagnosis, ± SD, yearsNA20.5 ± 14.527.9 ± 12.10.07
Mean age at OLT, ± SD, years44.9 ± 13.546.4 ± 9.944.2 ± 13.10.82
Mean BMI, ± SD28.4 ± 5.924.5 ± 3.427.2 ± 6.10.04
Female gender11(36.7%)12(44.4%)7(31.8%)0.65
EthnicityBlack5(16.7%)5(18.5%)2(9.1%)0.55
Other1(3.3%)0(0.0%)0(0.0%)
White24(80.0%)22(81.5%)20(91.0%)
Family history of UC4(13.3%)3(11.1%)9(40.9%)0.02
Concurrent autoimmune disorders1(3.3%)04(18.2%)0.01
CNS diseases1(3.3%)1(3.7%)00.01
Heart diseases003 (13.6%)0.01
Heart and CNS disease01(3.7%)00.01
Renal disease1(3.3%)3(11.1%)00.01
Extent of colitisExtensive colitis26(96.3%)14(63.7%)< 0.001
Proctitis/left-sided colitis1(3.7%)8(36.7%)
Extraintestinal manifestationsSkin1(3.3%)7(25.9%)7(31.8%)0.008
Joints1(3.3%)4(14.8%)10(45.5%)0.0005
Thrombo-embolic1(3.3%)1(3.7%)9(40.9%)0.0002
Osteoporosis4(13.3%)9(33.3%)13(59.1%)0.002
  • EIM, extra-intestinal manifestations; IBD, inflammatory bowel disease.

In the PSC + OLT group, the indication for OLT was end-stage liver disease secondary to PSC. The type of OLT was determined to be 52 (65.8%) patients cadaveric transplant, 2 (2.5%) patients with LDLT and 21 patients (26.6%) with living donor liver transplantation.

In the PSC + OLT + UC group, 23 (85.2%) patients had UC before OLT and 4 (14.8%) had OLT prior to UC diagnosis. The median time from OLT to UC diagnosis was 90 months, while the median time from UC diagnosis to OLT was 216 months. The indication for OLT was PSC with end-stage liver disease in all 27 patients.

In the PSC + OLT + UC + IPAA group, 9 patients (40.9%) had IPAA-then-OLT and 13 (59.1%) had OLT-then-IPAA, while 21 (95.5%) had UC before OLT and 1 (4.5%) had OLT before UC. The mean duration from IPAA to OLT in the 9 patients was 5.2 ± 1.98 years and the mean duration from OLT to IPAA for the 13 patients was 6.0 ± 3.7 years. The indication for OLT in this group was end-stage liver disease secondary to PSC in 21 (95.5%) patients and 1 (4.6%) patient with cholangiocarcinoma. Indications for colectomy were CRC (n = 2, 9.1%), dysplasia (n = 16, 71.7%), or refractory UC (n = 4, 18.2%). The mean age for colectomy in patients with PSC + OLT + UC + IPAA was 46.6 ± 11.5 years old. Pouch configurations were J pouches (n = 19, 86.4%), S pouch (n = 1, 4.6%) and other pouches (n = 2, 9.1%). Fifteen (88.2%) patients had 2-stage pouch surgery and 2 (11.8%) had 3-stage pouch surgery. The stages for 2 remaining patients with J pouches were unknown. The average interval from pouch reconstruction to ileostomy closure was 6.19 ± 4.41 months.

In the 9 patients that had IPAA before OLT, the mean age for colectomy was 40.9 ± 10.8 years and interval to pelvic pouch construction to ileostomy takedown was 4.25 ± 2.0 months. The mean age at OLT was 46.6 ± 10.7 years and the mean duration of IPAA to OLT was 5.2 ± 20.0 years.

3.2 Patient and graft survivals and post-operative complications for patients with PSC + OLT + UC + IPAA compared with PSC + OLT and PSC + OLT + UC

There were 10 (33.3%) patient deaths in the PSC + OLT group, 4 (14.8%) patient deaths in PSC + OLT + UC group, and 2 (9.1%) patient deaths in the PSC + OLT + UC + IPAA (P = 0.07). The Kaplan–Meier patient survival curve showing the comparison between the 3 groups can be seen in Fig. 1 . One of the 2 deaths in the PSC + OLT + UC + IPAA group was a result of septic complication (the patient had OLT before IPAA) and the other unknown (the patient had IPAA before OLT). In the UC + PSC + OLT group, 3 of the 4 deaths were sepsis-related (2 of 3 with sepsis had UC before OLT), while the cause of the 4th death unknown. In the PSC + OLT group, 2 deaths were sepsis related, 2 deaths related to coagulopathy, while the cause was unknown in 6 patients.

Figure 1

Comparison of patient survival between the 3 groups.

There were 10 (33.3%), 10 (37.0%) and 10 (45.5%) patients with acute rejection, and 2 (6.7%), 3 (11.1%) and 1 (4.6%) with chronic rejection in the PSC + OLT, PSC + OLT + UC and PSC + OLT + UC + IPAA groups respectively (Table 2). For graft survival, 3(10.0%), 6(22.2%) and 0 (0%) had re-transplantation in the PSC + OLT, PSC + OLT + UC and PSC + OLT + UC + IPAA groups, respectively (Table 2). The Kaplan–Meier graft survival curve between the 3 groups is illustrated in Fig. 2 .

Figure 2

Comparison of graft survival between the 3 groups.

View this table:
Table 2

Univariate analysis of post-OLT complications between the 3 groups.

FactorPSC + OLT (N = 30)PSC + OLT + UC (N = 27)PSC + OLT + UC + IPAA (N = 22)P value
Survival to date20(66.7%)23(85.2%)20(90.9%)0.07
Bacterial infection7(23.3%)13(48.2%)7(31.8%)0.14
Bacteremia7(23.3%)13(48.2%)12(54.6%)0.04a
Intra-abdominal abscess1(3.3%)5(18.5%)6(27.3%)0.03a
Bile leak1(3.3%)4(14.8%)5(22.7%)0.08
Acute renal failure5(16.7%)4(14.8%)9(40.9%)0.07
Chronic renal failure3(10.0%)1(3.7%)2(9.1%)0.6
Hepatic thrombosis2(6.7%)5(18.5%)6(27.3%)0.1
Ascitic leak002(9.1%)0.07
Stricture5(16.7%)3(11.1%)4(18.2%)0.75
Rejection12(40.0%)12(44.4%)11(50.0%)0.77
Acute rejection10(33.3%)10(37.1%)10(45.5%)0.67
Chronic rejection2(6.6%)3(11.1%)1(4.5%)0.67
  • a The individual comparison of UC + PSC + OLT patients with PSC + OLT patients did not show any statistically significant difference for both bacteremia and infection.

The results from univariable comparisons of post-OLT complications between the 3 groups are further listed in Table 2 . There was a numerical trend towards greater bile leak, ascitic leak, and acute renal failure in the PSC + OLT + UC + IPAA group than in the PSC + OLT group.

Bacteremia was identified in 7 patients (23.3%) in the PSC + OLT group, 13 (48.2%) in the PSC + OLT + UC group, and 12 (54.6%) in the PSC + OLT + UC + IPAA group (P = 0.04). Intra-abdominal abscess was identified in 1 patient (3.3%) in the PSC + OLT group, in 5 (18.5%) in the PSC + OLT + UC group, and in 6 (27.3%) in the PSC + OLT + UC + IPAA group (P = 0.03). Bacterial infection, although not statistically different between the study groups, was found in 7 (23.3%) in the PSC + OLT13 (48.2%) with PSC + OLT + UC and 7 patients (54.5%) with PSC + OLT + UC + IPAA (P = 0.14).

Univariable comparison of post-OLT complications comparing only the patients with IPAA before OLT in the PSC + OLT + UC + IPAA group with the two other groups are shown in Table 3 . There were no statistical differences in post-OLT infection, bacteremia, intra-abdominal abscess, acute and chronic renal failure, hepatic thrombosis and acute and chronic rejections between the 3 groups.

View this table:
Table 3

Univariate analysis of post-OLT complications for IPAA-then-OLT patients with other 2 groups.

FactorPSC + OLT (N = 30)PSC + OLT + UC (N = 27)PSC + OLT + UC + IPAA (N = 9)P value
Survival to date20(66.7%)23(85.2%)8(88.9%)0.16
Bacterial infection7(23.3%)13(48.2%)4(44.4%)0.12
Bacteremia7(23.3%)13(48.2%)5(55.6%)0.07
Intra-abdominal abscess1(3.3%)13(48.2%)2(22.2%)0.12
Bile leak1(3.3%)4(14.8%)3(33.3%)0.06
Acute renal failure5(16.7%)4(14.8%)3(33.3%)0.49
Chronic renal failure3(10.0%)1(3.7%)1(11.1%)0.58
Hepatic thrombosis2(6.7%)5(18.5%)2(22.2%)0.29
Ascitic leak001(11.1%)0.13
Biliary stricture5(16.7%)3(11.1%)1(11.1%)0.81
Rejection12(40.0%)12(44.4%)6(66.7%)0.37
Acute rejection10(33.3%)10(37.0%)6(66.7%)0.20
Chronic rejection2 (6.7%)10(37.0%)0(0.0%)0.39

3.3 MELD score association with infectious complications

MELD scores were calculated from the pre-transplant serum total bilirubin, INR and serum creatinine. For all patients in the cohort, the mean MELD score was 18. A comparison of the three groups (14.8 ± 4.5 for PSC + OLT, 19.5 ± 7.7 for PSC + OLT + UC and 19.2 ± 6.8 for PSC + OLT + UC + IPAA) showed that the mean MELD differed significantly across the 3 groups (P = 0.019). However, a one-way analysis of variance to determine association of MELD with infectious complications — intra-abdominal abscess, infection, or bacteremia showed no statistically significant association.

3.4 Multivariable analysis for risk factors for infectious complications

Age, body mass index (BMI), pouch construction, and UC were not found to be associated with bacteremia or intra-abdominal abscess on multivariable analysis. The odds ratios (OR) for age, ileal pouch, UC, and BMI for intra-abdominal abscess were 1.0, 1.1, 0.3 and 1.1 and 1.0, 0.7, 0.1 and 1.1 for bacteremia, respectively (Table 4 ).

View this table:
Table 4

Multivariable analysis of risk factors associated with intra-abdominal abscess or bacteremia.

FactorIntra-abdominal abscessP valueBacteremiaP value
Age1.0 (0.91–1.03)0.341.0 (0.95–1.03)0.61
Presence of ileal pouch1.1 (0.17–7.28)0.920.7 (0.13–3.75)0.68
Presence of ulcerative colitis0.3(0.06–1.25)0.090.1 (0.22–2.21)0.55
Body mass index1.1(0.92–1.33)0.281.1 (0.95–1.16)0.37

3.5 Post OLT medication use

Medications including antibiotics, sulfasalazine, corticosteroids, cyclosporine, ursodiol, mycophenolate, sucralfate, metrodinazole, sirolimus and tacrolimus administered post OLT were compared between the 3 groups. None of the medications used were statistically significant except for antibiotics and sulfasalazine. The number of patients who used antibiotics was 7 (23.3%) in the PSC + OLT group, 16 (59.3%) in the PSC + OLT + UC group and 14 (63.64%) in the PSC + OLT + UC + IPAA group (P = 0.004). The number of patients who used sulfasalazine was 0 (0%) in the PSC + OLT group, (3.7%) in the PSC + OLT + UC group and 14 (22.7%) in the PSC + OLT + UC + IPAA group (P = 0.006) (Table 5). Post OLT medications were compared in the OLT groups before and after the year 2000 time periods. Results showed a statistical trend of significant difference in mortality: 29.3% in the patients who had had OLT before the year 2000 (N = 12) and 10.5% in those who had OLT after year 2000 (N = 4) (P = 0.05). The use of tacrolimus use was significantly less prevalent in patients who had been transplanted before year 2000 (N = 13, 31.7%) than those who had OLT after year 2000 (N = 25, 65.8%) (P = 0.003).

View this table:
Table 5

Post-liver transplantation medications.

MedicationPSC + OLT (N = 30)PSC + OLT + UC (N = 27)PSC + OLT + UC + IPAA (N = 22)P value
Antibiotics7(23.3%)16(59.3%)14(63.6%)0.004
Sulfasalazine0(0%)1(3.7%)5 (22.7%)0.006
Corticosteroids18(60.0%)22(81.6%)12(54.6%)0.09
Cyclosporine2(6.7%)4(14.8%)2(9.1%)0.60
Ursodiol5(16.7%)5(18.5%)4(18.2%)0.98
Mycophenolate7(23.3%)6(22.2%)6(26.3%)0.91
Sucralfate8(26.7%)6(22.2%)6(27.3%)0.91
Metrodinazole2(6.7%)1(3.7%)2(9.1%)0.73
Sirolimus2(6.7%)1(3.7%)4(18.2%)0.20
Tacrolimus11(36.7%)16(59.3%)12(54.6%)0.66

4 Discussion

While IPAA appears to have no negative impact on overall graft and patient survivals in patients with PSC for OLT, there were increased infectious complications with the presence of IPAA on univariable analysis. The tendency for both bacteremia and intra-abdominal abscess was increased in PSC + OLT + UC + IPAA patients compared with PSC + OLT patients. However, these differences were not significant in multivariable analysis. In addition, there was no significant difference in bacteremia and intra-abdominal abscess between the PSC + OLT + UC group and PSC + OLT group. There were no increased infectious complications for OLT for PSC patients that had IPAA before transplantation. When all groups were assessed together in univariable analyses, the results were similar. These findings appear not to support our hypothesis that restorative proctocolectomy with intestinal reconstruction and bacterial overload may increase the risk for post-OLT infectious complications.

In this study, there were no statistical differences in overall patient and graft survivals between PSC + OLT patients and PSC + OLT + UC + IPAA or PSC + OLT + UC patients. Patient and graft survivals were not shown to be statistically different on Kaplan–Meier survival analysis. We did not find any difference in the secondary outcomes following OLT including acute and chronic renal failure, hepatic artery thrombosis, acute or chronic rejection, strictures and bile leak. A review of literature shows excellent patient and graft survivals of OLT for PSC. However, graft survival is shown to be affected by a high incidence of acute and chronic rejection and a high incidence of hepatic artery thrombosis.9 In an early study by Marsh et al.10 of 55 patients with PSC undergoing OLT, the patient survival rates at 3, 6, 9, 12, 24 and 36 months were 80%, 76%, 76%, 71%, 57% and 57%, respectively. Deaths in the cohort resulted from direct complications of transplantation, including graft failure, sepsis, biliary sepsis due to cholangiocarcinoma reoccurrence, thrombosis of portal and superior mesenteric veins leading to gastrointestinal bleeding and liver failure. A 12-year single-center study by Goss et al.12 of 127 patients undergoing OLT for PSC showed a 1, 2, and 5 year actuarial patient survival of 90%, 86% and 85% and graft survival of 82%, 77% and 72%, respectively. There were 91.4% of patients without recurrence of PSC, but 8% of patients had biliary complications, 3% had hepatic artery thrombosis and 1.6% had infection. These studies did not address whether and how IPAA affects the outcome of OLT. However, in a cross-sectional study of 32 patients with PSC + OLT + UC + IPAA evaluating surgical outcome of OLT, Mathis et al.13 showed that the 30-day morality was nil, while 1 year and 10 year graft survival was present in 32 and 26 patients, respectively.

There is an important mutual relationship between OLT for PSC and IPAA for UC. Hence, the impact of OLT on the outcome of IPAA has been studied. Our previous published study demonstrated that the presence of OLT and post-OLT immunosuppression may not affect IPAA and pouchitis disease course.7 On the other hand, the impact of IPAA or UC on the outcome of OLT has not been systematically investigated. Graziadei et al.14 studied 150 patients; 78% with IBD, 75% with UC, 13% with Brooke ileostomy and 10% with IPAA that underwent OLT for PSC and showed that the 1, 2, 5 and 10 year patient survival was 93.7%, 92.2%, 86.4% and 69.8%, respectively and 10-year graft survival was 83.4%, 83.4%, 79.0% and 60.5%, respectively. Twenty-three patients (15.3%) died after a mean duration of follow-up of 55 months post OLT, with the major causes being related to infection and sepsis, coronary artery disease and perioperative bleeding. There was at least 1 episode of acute rejection in 68.7% of patients and chronic rejection in 8% of patients in the first allograft of the series. There were 24 retransplantations in the series due to hepatic artery thrombosis, biliary strictures, chronic ductopenic rejection with vanishing bile duct syndrome and primary nonfunction. Their study further showed that OLT for PSC patients with IBD had a significantly higher number for acute cellular rejection (1.2 episodes per patient) compared to OLT for PSC patients without IBD (0.7 episodes per patient). These findings suggest that OLT for PSC patients with IBD are at increased risk for acute rejection. However, Skaked et al.15 studied 36 patients with OLT for PSC and 29 with concurrent UC and showed no effect on allograft function, except elevation of alkaline phosphatase. In our study, we did not observe any difference in rejection rates in OLT for PSC patients with or without UC or with or without IPAA for UC. In addition, there was no statistical difference of acute and renal failure, hepatic artery thrombosis, biliary stricture or bile leak between the groups.

Post-OLT infectious complications are common. In a study by Narumi et al.,16 there was moderate to severe infection and higher retransplantation in IBD patients (70%) compared to a matched control group (37.5%).14 In our study, there was significant difference in bacteremia and intra-abdominal abscess between the PSC + OLT + UC + IPAA patients and PSC + OLT groups in univariable analysis. When patients with IPAA before OLT were compared with the other groups to determine if IPAA may increase the risk of infectious outcomes or other outcomes, the findings were not significant. In addition, there was no significant difference in outcome of bacteremia and intra-abdominal abscess between the PSC + OLT + UC and PSC + OLT groups. It is important to note that although not significant, there were a greater number of patients with the use of tacrolimus, cyclosporine, mycophenolate and corticosteroids (59.3%, 14.8%, 22.2%, and 81%, respectively) in PSC + OLT patients with UC than PSC + OLT patients without UC (36.7%, 6.7%, 23.3% and 60%, respectively). However, greater use of sulfasalazine in the PSC + OLT + UC + IPAA group compared to the PSC + OLT and PSC + OLT + UC was statistically significant. There were a greater number of patients with antibiotic use in PSC + OLT + UC patients (59.3%) and PSC + OLT + UC + IPAA patients (63.6%) than PSC + OLT patients without UC (23.3%). The finding that the greater number of PSC + OLT + UC and PSC + OLT + UC + IPAA patients who were on antibiotics may be explained by the higher number of patients who were treated with immunosuppressives.

Infectious complications are among the most significant frequent causes of morbidity and mortality in OLT recipients.17,18 Although rejection is a reason for concern with liver transplant patients, it is eclipsed by alarming mortality rates associated with infections ranging from 30% to 50%.1921 Complexity of surgical procedure, underlying disease and possible contamination from bowels are factors that increase the risk of infection in OLT patients.22,23 Other risk factors are duration of transplantation, elevated serum bilirubin level and post-transplantation immunosuppressive agents.2427 The most common type of infection inflicting OLT patients are bacterial infections, however, viral, protozoan and fungal infections are also seen in this group.28 Of infectious complications, bacteremia and intra-abdominal infections are the most common infectious findings post OLT.29 Bacteremia is found in 25–35% of all OLT patients and constitutes about 19–30% of all infectious processes in liver transplants.30 Mortality rates for bacteremia range from 24% to 36%; hence it is used as a marker of OLT patient mortality.31 Reported risk factors for bacteremia in liver transplant patients are intravascular catheterization, postoperative hospitalization, longer duration of ICU stay and post-transplant hemodialysis.31,16 Intra-abdominal infections are also associated with OLT.32 The risk factors associated with intra-abdominal infections, including intra-abdominal abscess and peritonitis accounting for 70% of all intra-abdominal infections, are pre-transplant ascites, patients undergoing post-transplant dialysis, non-intra-abdominal infection surgical complications, reoperation and hepaticojejunostomy.28

Bacteremia and intra-abdominal abscess are common causes of morbidity and mortality in OLT recipients. Our univariable analysis shows that the presence of IPAA or UC may increase bacteremia and intra-abdominal abscess. However it does not appear to affect the graft or patient survival of OLT for PSC patients. This study opens doors for intriguing questions with regard to the interaction of the liver and the gut. There is a possibility that bacterial translocation, fecal stasis and immunosuppressant use may predispose OLT for PSC patients to the increased complication rates.

There are several limitations to our study. There might have been referral bias, due to the tertiary care nature of our practice. The study population was small, although it may represent the largest series in literature to date. We were not able to identify any specific factors which increased the susceptibility to infections. Sulfasalazine was the only immunosuppressive medication with higher use in the PSC + OLT + UC + IPAA group that may possibly explain increased susceptibility to infection compared to the other two groups. Data on sepsis, duration of surgery, length of stay in intensive care unit, transfusion requirements and cause of death were not available from chart review in the majority of patients. This missing data would have been useful in determining if these were additional risk factors for bacteremia and intra-abdominal abscess. The effect of immunomodulators and biological agents on the risk of post-OLT patient or graft survivals or infectious complications was not amenable to assessment due to the lack of sufficient data and the overlap in the immunosuppressive use for OLT and for UC. However, we speculate that better immunosuppressive and/or antibiotic regimens and better surgical procedures after 2000 may have an impact on mortality, infectious complications, and organ rejection. A prospective, longitudinal study is warranted to further determine if OLT for PSC patients with IPAA for UC is at increased risk for bacteremia and intra-abdominal abscess.

In conclusion, the presence of concurrent UC with or without IPAA appears not to affect liver allograft or patient survival.

Conflict of interest

The authors declare no conflict of interest.

Abbreviations
CD
Crohn's disease
CDC
Center for Disease Control
CNS
central nervous system
CRC
colorectal cancer
EIM
extraintestinal manifestations
FAP
familial adenomatous polyposis
IBD
inflammatory bowel disease
IC
indeterminate colitis
IPAA
ileal pouch-anal anastomosis
OLT
orthotopic liver transplantation
PSC
primary sclerosing cholangitis
UC
ulcerative colitis

References

View Abstract