Objectives: Biliary complications such as an ischemic-type biliary lesion can increase morbidity and mortality after liver transplant. Former studies have investigated several risk factors, but the underlying pathomechanism remains unclear. The focus of this study was to investigate factors causing early-onset (< 12 mo after liver transplant) versus late-onset ischemic-type biliary lesions (> 12 mo after liver transplant).
Materials and Methods: This retrospective study included 641 patients. Patients were grouped to those who developed ischemic-type biliary lesion and those who did not. Patients developing ischemic-type biliary lesions were further subgrouped into those diagnosed early (< 12 mo) and late (> 12 mo) after liver transplant. We analyzed demographic data, characteristics, and comorbidities of the recipients and donors, operative variables, and postoperative course, as well as laboratory values.
Results: The incidence of ischemic-type biliary lesions was 4.9%. Retransplant was performed more frequently in patients developing ischemic-type biliary lesions. The number of transfusions of blood products was higher in ischemic-type biliary lesion patients, especially in the early-onset ischemic-type biliary lesion group. Bilirubin levels were higher in patients with ischemic-type biliary lesions starting from day 7 after the operation, particularly in the early-onset group. Survival tended to be best in the late-onset ischemic-type biliary lesion group; however, this difference was not significant.
Conclusions: This study serves as a supplement to current data and the understanding of ischemic-type biliary lesions with emphasis on the relevance of disease onset and causes. We could in fact determine transfusion of blood products as a determinant of an early onset of ischemic-type biliary lesion. Bilirubin could be a surrogate marker for ischemic-type biliary lesions, especially in its early-onset form.
Key words : Biliary complications, Cholangitis, Graft failure, Nonanastomotic stricture, Postoperative complications
Liver transplant (LTX) is the first-line therapy for end-stage liver failure. Continuous technical and perioperative progress have improved outcomes of patients, but LTX is still associated with several vascular and nonvascular challenges. Biliary complications have an incidence of up to 35% and range from biliary leakage and strictures to dysfunction, leading to repeated interventions, readmissions, and sometimes acute or chronic transplant failure.1-3 Biliary strictures are classified into anastomotic strictures, situated close to the site of the anastomosis, and nonanastomotic strictures, which occur all over the biliary tree. Up to 50% of nonanastomotic strictures are related to thrombosis of the hepatic artery and mainly appear within the first year after LTX.2,4 The other 50% of nonanastomotic strictures manifest without arterial thrombosis, but with a similar pathomorphological and radiological presentation, and are called ischemic-type biliary lesions (ITBL).5
Ischemic-type biliary lesions have an incidence of 1.4% to 26% in patients after LTX and become apparent with bile duct destruction, subsequent constricting formations and sequestrations, followed by biliary abscesses or hepatic necrosis.6-8 According to the anatomic site, lesions can be classified into type I (extrahepatic), type II (intrahepatic), and type III (extra- and intrahepatic).9 Ischemic-type biliary lesions provoke uncharacteristic inflammatory reactions and intracellular cholestasis, which cause proliferation of the bile ducts, fibrosis, and destruction of liver tissue.6,10-12 Clinical symptoms are nonspecific and range from malaise, subfebrile temperatures, elevated alkaline phosphatase, gamma-glutamyltransferase, and bilirubin to fever, jaundice, or septic shock.6 The pathophysiology of ITBL is complex. The donor’s age, sex, or ABO incompatibility, as well as technical circumstances like ischemia time, preservation solutions, or arterial back-table perfusion, donation after cardiac death, portal vein thrombosis, bile salt toxicity, and some underlying diseases have been detected as sug-gestive risk factors.6,7,13 The development and prediction of ITBL remain unknown.
Besides immunological factors, blood circulation and oxygen supply seem to play an important role during the development of ITBL. Furthermore, the time span of manifestation of ITBL ranges from early after LTX to very late-onset disease. Data on the onset of disease are scarce, and information on the precise onset of ITBL is lacking in the current literature. However, we assumed that ITBL (being more the description of a clinical presentation than a pathomechanism) occurring right after LTX might be a condition different from ITBL manifestation presenting years after transplant.
Materials and Methods
This study included 641 patients who received LTX from deceased donors at our clinic from January 2005 until December 2009. No donations after cardiac death were included. All patients gave written informed consent, and the study was performed in accordance with the ethical standards of the responsible committee on human experimentation, following approval from the institutional review board and in accordance with the Declaration of Helsinki of 1975. Exclusion criterion was age under 18 years for recipients.
Standard for biliary anastomoses was a single-stitch side-to-side anastomosis. The primary endpoint was the incidence of ITBL within 5 years after transplant. Secondary endpoints were death or graft loss. Ischemic-type biliary lesion was diagnosed when nonanastomotic strictures (bile duct stenosis and/or destruction, constricting formations, and/or sequestrations) occurred after transplant. Diagnosis was established via endoscopic retrograde cholangiography or percutaneous transhepatic cholangiography by an experienced endoscopist or magnetic resonance cholangiopancreatography. Ischemic-type biliary lesion was only diagnosed in the absence of arterial malperfusion. Patients were grouped according to those who developed ITBL and those who did not. Patients developing ITBL were further subgrouped into those diagnosed early (< 12 mo; early onset) and late (> 12 mo; late onset) after LTX. Retransplant was considered as a new case. Data were retrospectively analyzed and extracted from the patient documentation system of our clinic. All patients underwent a standard preoperative evaluation process.
We evaluated parameters of donors, recipients, surgical procedures, and postoperative management that might influence the development of ITBL. The recipients’ characteristics were assessed by sex,
age (years), weight (kg), height (cm), body mass
index (BMI, calculated as body weight in kilograms divided by height in meters squared), blood group, retransplant, underlying disease, and Model for End-Stage Liver Disease (MELD) score. The MELD score was calculated according to the date of transplant by a common formula, as follows: (0.957 × ln(creatinine) + 0.378 × ln(bilirubin) + 1.120 × ln(INR) + 0.643) × 10), where INR is the international normalized ratio. Donors were characterized by sex, age, weight, height, BMI, and blood group, as well as by mismatch
in the donor’s and recipient’s sex and blood group. Surgical procedure was evaluated by ischemia time (cold, warm, and total), preservation solution (University of Wisconsin or histidine-tryptophan-ketoglutarate), type of biliary anastomosis (side-to-side or hepaticojejunostomy), the duration of operation (minutes), and graft size (full size or split size).
The postoperative course was analyzed by the amount/number of transfusions (erythrocyte concentrates and fresh frozen plasma), length of hospital stay (days), and duration of stay in the intensive care unit (days). Blood was taken on days 1, 2, 3, 7, 10,
and 14 after LTX, and we measured and recorded the following details: alanine aminotransferase (U/L), aspartate aminotransferase (U/L), glutamate dehydrogenase (U/L), bilirubin (mg/dL), gamma-glutamyltransferase (U/L), alkaline phosphatase (U/L), hemoglobin (mg/dL), leukocytes (counts/μL), C-reactive protein (mg/dL), creatinine (mg/dL), urea, glomerular filtration rate (mL/min), international normalized ratio, partial thromboplastin time (seconds), and thrombocytes (counts/μL).
Postoperative complications were grouped according to biliary complications (anastomotic stenosis, nonanastomotic stenosis, anastomotic insufficiency, biliary leakage/biloma, intrahepatic abscess, cholangitis) and acute rejection (according to the Banff classification14). The follow-up course was examined during routine consultations at 6, 12, 36, and 60 months after LTX (clinical status, laboratory, sonography, biopsy).
Treatment of ITBL consisted of endoscopic treatment of stenosis and cholestasis (dilatation, stenting, external percutaneous bile drainage). In case of associated infections, antibiotic treatment and/or adjustment of immunosuppressant therapy was performed. In case of complete loss of the transplant, retransplant was performed.
We performed statistical analyses with Excel (version 15.32; Microsoft) and XLSTAT-Biomed (version 2017.1; Addinsoft) software. The level of significance was defined as P < .05. Nominal data were tested by cross-tabulation (univariate, Fisher exact test). Nonnormally distributed data were tested by the Mann-Whitney U test; for more than 1 sample, the Kruskal-Wallis test was used. Survival rates were estimated by Kaplan-Meier analyses.
From 2005 until 2009, there were 641 patients who underwent LTX at our clinic. Of these, 34 patients (4.9%) developed an ITBL within 5-year follow-up. The prevalence was 3.74% within the first 12 months (n = 24) and 1.56% from 12 to 60 months (n = 10) after LTX (Figure 1). The median period to diagnosis of ITBL was 262.5 days after LTX. The demographic data and recipient characteristics are given in Table 1. There were 408 men (63.65%) and 233 women (36.35%) included in this study, with no significant differences in sex, age (54 vs 56.5 y), weight (77 vs 78 kg), height (174 vs 176 cm), and BMI (28.7 vs 26.0) among recipients. The most frequent blood group within the recipients was blood group A (44.6% in patients without ITBL, and 50% in patients with ITBL), but the distribution of the blood groups was without significance. Retransplant (9.7% in patients without ITBL, and 26.5% in patients with ITBL) was more frequently performed in patients who developed ITBL afterward. The underlying disease did not differ significantly between the groups. The MELD score tended to be higher in patients with ITBL (20 vs 21) but without significance.
We found no significant difference with regard to the sex of the donors. Patients with ITBL received grafts from elder donors (52 vs 54 y); however, this was without significant difference. Also, weight (75 vs 78 kg) and height (170 vs 175 cm) were not significantly different. There was no significant difference in sex mismatch of graft and recipient or ABO incompatibility.
No significant differences could be detected in the length of the cold (570 vs 600 min) and warm (45 vs 42 min) ischemia time. There was no significance in type of preservation solutions, surgical technique of the biliary anastomosis, duration of the operation, and the graft size. Notably, patients with ITBL received significantly more transfusions, both red blood cell units (6 vs 11) and fresh frozen plasma units (21 vs 26). The length of hospital stay and duration of intensive care were nearly equally distributed. Biliary complications were significantly associated with ITBL. Anastomotic and nonanastomotic stenosis and intrahepatic abscesses had a significantly higher prevalence in patients with ITBL. Acute rejection did not significantly differ. Regarding the therapy for ITBL, all patients with ITBL received endoscopic retrograde cholangiography (97%) and/or percutaneous transhepatic cholangiography (24.2%) interventions during their postoperative course. Interventions included dilatation (60.6%), stenting (81.8%), and external bile drainage (24.2%).
Transaminases were nearly equal, with an adequate decrease. Bilirubin levels were significantly higher among patients with ITBL after day 7. Glutamate dehydrogenase, gamma-glutamyltransferase, alkaline phosphatase, leukocyte level, and C-reactive protein showed no substantial differences. Hemoglobin was higher in patients without ITBL (significant on day 3). International normalized ratio and partial thromboplastin were equal, but the count of thrombocytes was higher in patients without ITBL (significant on day 7).
The estimated survival rate in patients without ITBL was 94.1 ± 2.1 months versus 88.7 ± 8.6 months in the ITBL group (Figure 1).
When we compared early-onset and late-onset ITBL, donors tended to be older in the early-onset ITBL group (56 vs 49 y); however, this was without significance. Patients developing ITBL early after transplant received more fresh frozen plasma units (27 vs 24; P = .004); they showed more biliary complications such as anastomotic stenosis (16 vs 4; P < .001), nonanastomotic stenosis (14 vs 8; P < .001), and intrahepatic abscesses (3 vs 0; P < .001). Patients with late-onset ITBL showed a tendency for longer survival that those with both early-onset ITBL and no ITBL; however, this was not significant (Figure 2). Patients with early-onset ITBL had significantly higher bilirubin levels on days 7 and 14 compared with patients with late-onset ITBL. The other laboratory values evaluated showed no significant differences (Figure 2).
In accordance with former studies, the characteristics of the recipient (age, blood group, underlying disease, MELD score) did not play a significant role in the development of ITBL, with the one exception that former transplant was significantly associated with ITBL.6,13 The incidence of our cohort (4.9%) corresponded with reports from the literature.2,6
Regarding the characteristics of the graft, we could detect a higher age of the donor as a potential risk factor. Previous studies could show an impact of a higher age of the donor on ITBL, even if one large study demonstrated the opposite.6,13 However, biliary strictures have also been described after pediatric LTX, and this has even been described as the Achilles heel of the procedure.14,15 Biliary complications are even more prevalent in pediatric patients than in adults, with reported rates varying between 15% and 30%.16
Mismatch of sex and blood group might be an individual risk factor for ITBL.13 However, in our study, the donor’s individual blood group did not play a significant role, nor did sex mismatch or ABO incompatibility.
We demonstrated a significantly higher amount of red blood cell transfusions and fresh frozen plasma in patients with ITBL. Of note, patients who developed ITBL within the first 12 months after transplant received more transfusions. This could be explained by the accumulating evidence regarding the immunological impact of blood products; that is, every blood transfusion interferes with the immune system of the recipient. Almost all identified severe immunological reactions toward blood transfusions are mediated by allo-antibodies.17 Those allo-antibodies might, after the transplant, lead to immune-related complications like transplant rejection. Therefore, many transplant centers now try to avoid blood transfusion prior to transplant, to avoid alloimmunization.18 Immunological mechanisms also play a role in LTX and the development of ITBL; that is, preformed antibodies to biliary epithelial cells have been linked to acute rejection of liver allografts.19 Antibodies against the biliary epithelium contribute to posttransplant cholangitis via induction of expression of Toll-like receptors 2 and 3, as well as production of and inflammatory cytokines and chemokines.20 However, to draw a clear link between intraoperative transfusions and ITBL, further studies on preexisting antibodies and de novo antibodies are needed.
Some technical aspects of LTX were previously discussed as risk factors for ITBL. Prolonged ischemia time has been previously reported to influence the development of ITBL.5,6 However, in contrast, other reports (especially some studies from the 1990s) did not show any impact of the ischemia time on ITBL.21-23
The high viscosity of University of Wisconsin solution compared with histidine-tryptophan-ketoglutarate solution might cause an incomplete rinsing of the peribiliary plexus and is associated with ITBL.6,13,24-26 However, the type of preservation solution was not confirmed as a risk factor by our collective. Furthermore, the type of biliary anastomosis and the duration of the operation procedure had no significant impact on ITBL.
Previous studies have examined changes in several laboratory parameters. We analyzed blood tests from the day of transplant until 2 weeks after transplant and could not confirm previous findings of significantly elevated transaminases.13 However, we detected significant changes in bilirubin.13 Bilirubinemia can be an unspecific early sign caused by a general hepatic injury, and this may be related to more frequent biliary complications among patients with ITBL. Patients who developed ITBL early after transplant had significantly higher bilirubin levels than patients developing ITBL later than 12 months or no ITBL. Therefore, bilirubin seems to be a good surveillance marker with which to identify patients at higher risk. Because of the small sample size of patients with ITBL, it was not possible for us to define a valid cutoff value for bilirubin. Patients with ITBL were more often affected by a simultaneous anastomotic stenosis compared with patients without ITBL, and this difference was significant, especially for patients with early onset of the disease. In addition, nonanastomotic strictures, intrahepatic abscesses, and cholangitis were significantly more frequent among patients with ITBL, and these conditions were most frequent in patients with early-onset ITBL.
The retrospective design and limited number of ITBL cases must be considered as limiting factors of this study. However, contrary to other studies, the time frame from 2005 to 2009 does not imply fundamental changes in diagnosis and treatment of ITBL; therefore, ours is a more homogenous cohort compared with inclusion of patients over a longer time frame.
This study could not completely illuminate the origin and pathophysiology of ITBL. We tried to focus on the differences between early-onset ITBL and late-onset ITBL. With regard to our data as well as previously described factors favoring ITBL, we explored the hypothesis that an ITBL event occurring only a few weeks or months after LTX might not be considered to originate from the same pathophysiological mechanism versus ITBL occurring much later (years) after LTX. This can be explained by the fact that ITBL is simply a description of a clinical picture, nothing more. We tried to better define the factors leading to this clinical picture and to draw conclusions that may better differentiate the causes thereof. We did in fact determine that the transfusion of blood products was a determinant of early onset of the disease, as well as biliary complications. Bilirubin could be a surrogate marker for ITBL, especially for early-onset ITBL.
The limitation of this study is that there were only 34 patients with ITBL, which is a small number. Therefore, further investigations are necessary for complete comprehension of ITBL, as it remains a polygenetic disease with the uniform manifestation of a gradual destruction of the bile ducts.
DOI : 10.6002/ect.2020.0032
From the 1Department of Surgery, Charité-Universitätsmedizin Berlin, and the 2Department of Hepatology and Gastroenterology, Charité-Universitätsmedizin Berlin, Berlin, Germany
Acknowledgements: The authors thank all patients for their participation in the study. Moreover, we thank Gabriele Najbar and all nurses and physicians involved in care of the patients. The authors declare that there are no financial relationships with any organizations that might have an interest in the submitted work. All authors do not have any competing interests related to the presented work and research. Rosa Schmuck is participant in the BIH‐Charité Clinician Scientist Program funded by the Charité-Universitaetsmedizin Berlin and the Berlin Institute of Health.
Corresponding author: Eva Maria Dobrindt, Department of Surgery, Charité-Universitätsmedizin, Chariteplatz 1, 10117 Berlin, Germany
Phone: +49 30 652 001
Table 1. Demographic Display of the Characteristics and Comorbidities of the Recipients and Donors, Operative Variables, and Postoperative Course (N = 641)
Figure 1. Prevalence and Survival Rate of Ischemic-Type Biliary Lesion
Figure 2. Characteristics of Early-Onset and Late-Onset Ischemic-Type Biliary Lesion