Ischemic-type biliary lesions are a major complication following orthotopic liver transplantation. They occur in up to 26% of liver transplant recipients. Among other factors, unknown immunologic factors have always been assumed to be partly responsible for these lesions. CC-chemokines and their receptors play a key role in postoperative immunomodulation after liver transplantation. The non-function CC-chemokine receptor 5Δ32 polymorphism (CCR5Δ32) has been shown to lead to a lower rate of acute rejection after kidney transplantation; in liver transplantation the role of CCR5Δ32 is unclear. We investigated the influence of the CCR5Δ32 after liver transplantation with special regard to ischemic-type biliary lesions. The CC-chemokine receptor-5 (CCR5) of 146 recipients was analyzed by polymerase chain reaction to detect CCR5Δ32 in blood samples of patients after liver transplantation. One hundred twenty patients with wild-type CCR5 and 26 patients with CCR5Δ32 (1 homozygote, 25 heterozygote) were identified. Ischemic-type biliary lesions occurred in 14 of 120 patients with wild-type CCR5 and in 8 of 26 patients with CCR5Δ32 polymorphism (P = = 0.01). 5 year patient survival with CCR5Δ32 and CCR5 was 70% and 85%, respectively (P = .0067). Our results show that the CCR5Δ32 is a significant risk factor for the development of ischemic-type biliary lesions after liver transplantation and leads to a reduction in 5-year survival. In conclusion, the CCR5 status should be screened prospectively before liver transplantation. (Liver Transpl 2004;10:434–439.)
Ischemic-type biliary lesions (ITBL) following orthotopic liver transplantation (OLT) are a life-threatening complication.1, 2 Like graft loss due to surgical problems, chronic rejection, and recurrent disease, they are one of the most important reasons for liver retransplantation.3, 4 ITBL are reported to occur in up to 26% of all patients undergoing OLT and lead to reduced graft and patient survival.5, 6 Concerning the high risk of developing ITBL for all candidates undergoing OLT, the impact on graft loss due to this complication seems to be even higher than the impact on graft loss in recurrent hepatitis C.7 The sequelae, namely cholestasis and cholangitis, are often treated by repeated interventional or endoscopic procedures8, 9 and reoperations.10 However, due to the biliary destruction, all of these approaches are of limited effectiveness in preventing retransplantation in many cases. Features of ITBL are bile duct stenoses, dilatations, and cast formation.2 With regard to therapeutic consequences, two major types can be distinguished:11 In type A lesions, the complete biliary system is affected, and in type B lesions, only the major extrahepatic bile ducts are involved. Pathomorphologically, these alterations are caused by epithelial and muscular necrosis of the biliary system. Periductal connective tissue is usually remarkably well preserved. This leads to the assumption that merely the bile duct itself is damaged by the responsible mechanism. The features of ITBL have been repeatedly categorized and biliary cast or biliary sludge formation have been described in other studies. The lesions were properly characterized as a particular entity by the Mayo group and briefly thereafter by Li et al. and Kadmon et al.1, 12, 13 According to the literature, the rate of patients with ITBL ranges between 15 and 26%. By definition, ITBL comprise all diffuse (intrahepatic as well as extrahepatic) bile duct destructions after OLT without any known cause. The underlying cause of ITBL remains unclear despite numerous studies.14 Our group reported a significant reduction of ITBL by additional arterial back-table pressure perfusion (APP) for organ procurement.15 Since ITBL could not be totally prevented by that procedure, we searched for other immunologic risk factors to explain ITBL.16
Chemokines play a key role in the postoperative immunomodulation,17 especially during rejection18 but also in post-ischemic and inflammatory damage. CC-chemokine ligand (CCL) 3 (macrophage inflammatory protein 1 alpha) and CCL4 (macrophage inflammatory protein 1 beta) are overexpressed by infiltrating leukocytes.19 Both ligands bind to CC-chemokine receptor 5 (CCR5). CCR5Δ32 polymorphism is a nonfunctional mutant allele of CCR5 with an internal deletion of 32 base pair (bp) (CCR5Δ32). It leads to a resistance to human immunodeficiency virus-1 as well as to an improved survival of kidney allografts,20 Schröppel et al. showed no differences in patient survival, rejection rates, retransplantation rates, and deaths in patients with CCR5Δ32 following OLT, compared with patients with wild-type CCR5.21 In kidney transplantation, reduced rejection rates were demonstrated under CCR5Δ32.21 The influence of CCR5Δ32 on diffuse bile duct lesions following OLT has not been investigated yet.
One hundred forty-six patients transplanted between September 1997 and April 2003 at the Johannes Gutenberg University of Mainz were included in this retrospective study. The following data were extracted form the hospital records: Donor-related factors (age, blood group, gamma glutamyltransferase [GT]), transplantation-related factors (preservation technique, warm- and cold-ischemia time), and recipient-related factors (age; gender; blood group; original disease; perioperative aspartate aminotransferase [AST]; alanine aminotransferase [ALT]; alkaline phosphatase; gamma glutamyltransferase [gamma GT]; bilirubin; prothrombin time; creatinine; rejection and cytomegalovirus [CMV] rates; ITBL; and intensive care unit [ICU] and hospital stay). All patients gave their written informed consent for this investigation. All patients and all organ donors included in this investigation were Caucasian.
Ischemic-Type Biliary Lesions, Rejection and CMV Infection
The development of ITBL was the endpoint of the study. Diagnosis of ITBL was based upon endoscopic retrograde cholangiography in every case. Clinical symptoms of biliary obstruction and/or increased alkaline phosphatase and gamma GT led to the diagnostic procedure. In every case, anastomotic bile duct complications, arterial thrombosis, and chronic rejection were excluded by angiography and liver biopsy. Acute rejection, indicated by a rise in the aminotransferases, was biopsy proven in all cases. Acute rejection was treated with bolus steroids; OKT-3 was necessary only in one case. CMV infection was treated by intravenous ganciclovir, if pp65 tested positive.
For analysis of the CCR5 genotype, genomic DNA was prepared from 200 μL peripheral blood using the QIAamp DNA blood kit (Qiagen, Cologne, Germany). 2.5 μL of DNA were amplified by PCR using the following CCR5 specific primers: CCR5 sense, 5′-CAAAAAGAAGGTCTTCATTACACC-3′; and CCR5 antisense, 5′-CCTGTGCCTCTTCTT CTCATTTCG-3′). The PCR mixture was composed of 2.5 μL 10× PCR buffer (Roche Molecular Systems, Mannheim, Germany), 0.5 μL 12.5 mmol/L dNTP (Peqlab, Erlangen, Germany), 2.5 μL of each sense and antisense primer (10 μmol/L) and 1.25 U AmpliTaq DNA polymerase (Roche Molecular Systems) in a total volume of 25 μL. Forty PCR cycles were run on a Genius thermocycler (Techne, Cambridge, UK), using the following temperature profile: initial denaturation, 94°C 3 minutes; amplification, 94°C 1 minute, 64°C 1 minute, 72°C 1 minute (40 cycles); terminal elongation, 72°C 9 minutes. The size of the wild-type PCR product was 189 bp, and the CCR5Δ32 allele yielded a product of 157 bp. PCR products were analyzed by 2% agarose gel electrophoresis.
Statistical analyses were performed using SPSS software version 11.0.0 (SPSS Inc., Chicago, IL). Values were expressed as means ± SD. P values lower than .05 were considered significant. The chi-square test, Fisher exact test, and Student's t test were used to compare the different mean values; the Kaplan-Meier method and the log-rank test were used to compare survival rates. The multivariate analysis was performed as a logistic regression. Only those risk factors that reached P values <.05 in the univariate analysis were included in the multivariate analysis.
The indications for liver transplantation were alcoholic cirrhosis in 21%, viral hepatitis in 28 % (hepatitis C virus, 19%; hepatitis B virus, 9%), hepatocellular carcinoma in 14%, autoimmune liver disease in 15%, graft dysfunction in 6%, acute liver failure in 6%, and other indications in 10%. The average donor age was 45 ± 14 years. Cold ischemia time was 539 ± 141 minutes, and warm ischemia time (piggyback technique, sequential reperfusion) was 37 ± 10 minutes. The bile duct anastomosis was performed end-to-end in interrupted suture technique with 6-0 PDS. The average recipient age was 50 ± 11 years. Acute rejection occurred in 33%. The immunosuppressive protocol included tacrolimus in combination with steroids in 72%; in 28%, a triple therapy containing cyclosporine A, azathioprine, and steroids was used. CMV infections occurred in 22%. The average ICU stay was 7 ± 12 days, which were included in the total hospital stay of 27 ± 29 days.
ITBL occurred in 22 (15%) of 146 grafts with available DNA samples. Table 1 contains the results of the univariate analysis comparing graft and recipient characteristics in patients with and without ITBL. In patients with ITBL, donor age was significantly higher (P= .04) and gamma GT was significantly lower (P = .001). There were no major differences regarding the other factors. One hundred twenty patients had a wild-type CCR5, whereas 26 patients had CCR5Δ32 (25 heterozygote, 1 homozygote). ITBL occurred in 8 of 26 patients with CCR5Δ32, compared with 14 patients with ITBL out of 120 patients with wild-type CCR5 (P = .01; Table 2). Mean patient survival with CCR5Δ32 was 44.7 months (1,341 days), compared with 64. 1 month (1,923 days) in patients with wild-type CCR5. Five-year patient survival with CCR5Δ32 and CCR5 was 70% and 85%, respectively (P = .0067; log-rank test and Kaplan-Meier method, Fig. 1). Of the 26 patients with CCR5Δ32, 7 died after OLT. Reasons for death were ITBL with severe cholangitis in 4 cases, cardiac arrest in 2 cases, and cerebral bleeding in 1 case. Three out of 26 patients with CCR5Δ32 underwent retransplantation due to ITBL and are alive. In the group with wild-type CCR5, 11 out of 120 patients died as a result of hepatitis B reinfection in 4 cases, septic complications in 3 cases, cardiac arrest in 2 cases, recurrent hepatocellular carcinoma in 1 case and a car accident in 1 case. Six patients in this group underwent retransplantation, 2 because of primary nonfunction, 2 because of ITBL, 1 because of chronic rejection, and 1 because of arterial problems.
Table 1. Univariate Analysis of Donor and Recipient Factors Influencing ITBL
Liver grafts procured by our own group received APP in order to prevent ITBL. Table 3 gives an overview of the incidence of ITBL in relationship with the CCR5 status and the perfusion mode. In patients with wild-type CCR5 in combination with APP, none developed ITBL. The risk increased both by CCR5Δ32 status and by use of standard perfusion techniques. Patients suffering from CCR5Δ32 in combination with standard perfusion had the highest risk of developing ITBL later. A multivariate analysis (Table 4) was performed to determine factors influencing the incidence of ITBL. APP (P = .004) and CCR5Δ32 (P = .003) had a significant influence on the development of ITBL. Donor gamma GT (P = .076) and donor age (P = .122) did not reach statistical significance in the development of ITBL.
Table 3. CCR5 Status and Perfusion Mode in Patients With ITBL
Abbreviations: SP, standard perfusion; APP, arterial back-table pressure perfusion.
1 (out of 6) (16%)
0 (out of 37) (0%)
7 (out of 20) (35%)
14 (out of 83) (17%)
Table 4. Multivariate Analysis of Factors Influencing ITBL
Diffuse bile duct strictures have been observed since the early days of OLT, introduced by Thomas Starzl.22–24 They are known in patients with chronic rejection,25 in ABO incompatible grafts,26, in patients with hepatic artery thrombosis,27, 28 and in recurrent primary sclerosing cholangitis.29 Therefore, the classification of bile duct damage after OLT was divided in ischemic biliary lesions and bile duct lesions caused by immunologic problems. The remaining group of bile duct strictures of unknown origin were named ischemic-type biliary lesions due to the same morphological features.
There is no doubt that the bile ducts are the Achilles' heel of the liver graft, reflected in the many different types of organ damage leading to the same morphological symptom. Over the past years, many different groups have tried to determine risk factors for the development of ITBL.1, 6, 14 The type of biliary reconstruction, the primary liver disease, CMV infection, repeated acute graft rejection, positive lymphocyte crossmatch, poor match, initially high aminotransferases reflecting the preservation damage, high donor age, and the experience of the procuring surgeon were all claimed to be causal. A significant increased incidence of ITBL could be demonstrated in grafts with prolonged cold ischemia time over 12 hours.30 The incidence of ITBL increased with the widespread use of the University of Wisconsin (UW)solution for liver preservation.5 The high viscosity of the UW solution presumably leads to an inadequate perfusion of the small arteries of the biliary tree and therefore can cause an insufficient preservation state. Our own group demonstrated an improved preservation technique (APP) that resulted in a significant reduction of the incidence of ITBL after OLT.15 High donor age had a significant influence on the development of ITBL as well. Sclerosis of the arterial system of the aged organ donor may impair the regular perfusion of the bile duct capillaries and can therefore lead to additional damage. Liver grafts form non-heart-beating donors were recently shown to have a high risk for ITBL.31 Pathophysiologically, this setting is linked with a high risk for microvascular clogging before perfusion can be started. This reflects the same mechanism as in aged donors or insufficient perfusion under UW solution.
The nonfunctional CCR5Δ32 polymorphism was shown to be beneficial in renal allografts recipients due to a significant lower incidence of acute and chronic rejections.20 The investigation in our liver transplant patients did not show any differences in rejection rate but did show reduced survival in patients after OLT with CCR5Δ32 (Fig. 1). Since ITBL is one major reason for graft loss besides rejection, we correlated ITBL and CCR5Δ32 and saw a significant influence on the development of ITBL. The mechanism of the destruction of the bile ducts in patients with CCR5Δ32 is unclear at the moment. There is no doubt that the regulatory processes of the immunmodulation following OLT are in a certain imbalance under CCR5Δ32.32–34 This imbalance might lead to an immunoreaction with a normal integrity of the graft function and without clinical apparent rejection but destruction of the bile ducts. Indeed, biliary epithelial cells have been shown to be able to produce CC-chemokines that bind specifically to CCR5.35 Nevertheless, these results are of high clinical importance, especially because of the high number of patients undergoing liver retransplantation for ITBL. As the multivariate analysis showed a significant influence on ITBL by APP and CCR5Δ32, we analyzed the risk for ITBL at different combinations with APP and CCR5 status (Table 3). The lowest risk of ITBL is in patients with APP and wild-type CCR5. In contrast, standard perfusion in combination with CCR5Δ32 has a very high risk (7 out of 20) for the development of ITBL. Grafts with improved preservation state under APP with a significant decreased ischemia-reperfusion injury might also have less perioperative immunoactivation. The second risk factor, CCR5Δ32, might be overcome in those grafts; in grafts with intensive ischemia reperfusion damage, the regulatory system under CCR5Δ32 might collapse with the result of bile duct destruction.36 With regard to other well-defined immunolgic risk factors for diffuse biliary injury following OLT, such as ABO-incompatibility,37, 38 recent publications have shown beneficial effects of intensified immunosuppressive protocols to compensate for the negative implication of the risk factor.39
Concerning CCR5Δ32, further investigations must determine whether modification of the immunosuppression can prevent ITBL even in patients with this polymorphism. The role of CCR5Δ32 polymorphism in ITBL should be screened in prospective studies by other groups. Currently in our center, the CCR5 status is screened before liver transplantation. This presents the opportunity to stratify high-risk and low-risk patients for ITBL prior to OLT. In our center all organs are preserved with APP to reduce the incidence of ITBL. In the case of organs allocated from remote centers without APP, these organs are accepted for patients with wild-type CCR5 in our center but still are at higher risk for ITBL because of worsened preservation state. In patients with CCR5Δ32, we accept only liver grafts with excellent preservation state, such as with APP. There is still discussion about whether preservation with the low viscous Histidine-Tryptophan-Ketoglutarat solution can show the same results in preventing ITBL as APP with UW University of Wisconsin solution. So far, studies show the same rate of ITBL.
Regarding these results, a new classification of diffuse bile duct injury after OLT is justified (Fig. 2). Diffuse bile duct injury can be divided into ischemic biliary lesions and immunogenetic biliary lesions . Ischemic biliary lesions are bile duct injuries due to arterial stenosis or thrombosis (macroangiopathy), aged donors (micorangiopathy by arteriosclerosis), and insufficient preservation (microangiopathy by vascular clogging under preservation without adequate pressure), and in non-heart-beating donors (insufficient preservation by vascular clogging under delayed perfusion). In contrast, immunogenetic biliary lesions can be observed in ABO incompatibility, chronic rejection, recurrent primary sclerosing cholangitis, and CCR5Δ32. However, cross-links between the two groups that influence each other are likely.
In conclusion, CCR5Δ32 in liver transplantation is associated with a high risk for ischemic-type biliary lesions. It seems important to use low-risk grafts (young donor age; use of APP) in liver transplantation candidates with CCR5Δ32 polymorphism. Furthermore, intensified induction therapy after transplantation should be studied in these patients. We therefore suggest screening for CCR5 status of all patients on the waiting list.