Data were obtained from our retrospectively maintained database of orthotopic LT recipients at Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea. A total of 2,112 consecutive patients who underwent liver allograft transplantation between January 1, 2000 and December 31, 2010, were evaluated. The indications for LT were hepatitis B virus (HBV)-associated liver disease (n=1,759), toxic or alcoholic liver disease (n=215), hepatitis C virus (HCV)-associated liver disease (n=84), autoimmune hepatitis (n=7), primary biliary cirrhosis (PBC; n=18), primary sclerosing cholangitis (PSC; n=2), and metabolic disease (n=26). Our standard immunosuppression regimen was either tacrolimus or cyclosporine with steroids.
If allograft liver biopsies were performed more than 3 months post-transplantation because of abnormalities in follow-up liver function test results (elevated alanine aminotransferase and aspartate aminotransferase levels, with or without elevated gamma-glutamyl transferase levels), the case was defined as late allograft dysfunction. Using an electronic database, we selected 174 recipients from the 2,112 LT recipients as the study population. The indications for LT in these 174 patients were HBV-associated liver disease (n=134), toxic or alcoholic liver disease (n=18), HCV-associated liver disease (n=15), PBC (n=2), PSC (n=2), and metabolic disease (n=3). The demographics of the study population are summarized in .
Demography of the study population
The total number of liver biopsies was 361 and on average each patient underwent 2 biopsies (range, 1 to 10). If multiple biopsies performed in the same patient in a short period resulted in the same diagnosis, the biopsy results were considered to represent the same episode of late allograft dysfunction and counted as one event. If biopsies were performed over an extended duration of time, each biopsy was counted separately. A total of 245 episodes of late allograft dysfunction were identified including: biopsies of HBV-associated liver disease (n=182), HCV-associated liver disease (n=25), toxic or alcoholic liver disease (n=29), PBC (n=3), PSC (n=2), and metabolic liver disease (n=4).
The causes of late allograft dysfunction were determined by reviewing histological, serological, and radiological findings at the time of late allograft dysfunction and each patient's clinical course after allograft liver biopsy. Transplant recipient data were retrieved from electronic medical records. Data included each patient sex, age at diagnosis of liver disease, treatment, clinical course, and serological test results, including HBsAg, HBcAg, HBV DNA, and HCV RNA. All patients underwent Doppler ultrasound evaluation of the liver, biliary tree, and hepatic vasculature and some underwent computed tomography and endoscopic retrograde cholangiopancreatography. We recorded any evidence of biliary tract dysfunction or blood flow abnormality revealed by these ancillary tests. Biliary complication was defined as a biliary obstruction or stricture proved radiologically or endoscopically. Supporting histologic evidences for biliary complication included portal fibrosis, bile ductular proliferation, neutrophilic or eosinophilic infiltration, as well as centrilobular cholestasis to some degree. Biliary problems caused by other events, such as acute or chronic rejection and recurrent viral hepatitis were not included.
For the purpose of classification and discussion, we have arbitrarily classified liver allograft dysfunction into three time periods after LT: 1) early-late period (ELP) allograft dysfunction (>3 months but ≤6 months after LT); 2) mid-late period (MLP) allograft dysfunction (>6 months but ≤12 months after LT); and 3) late-late period (LLP) allograft dysfunction (>12 months after LT).
Biopsy specimens of allograft livers were fixed in 10% phosphate-buffered formalin and stained with hematoxylin and eosin using the Masson trichrome technique and for cytokeratin 19 (monoclonal mouse anti-human cytokeratin 19 antibody; Dako North America Inc., Carpiteria, CA, USA). If necessary, in situ hybridization study for detection of cytomegalovirus or Epstein-Barr virus was performed. We decided the histologic diagnosis based on variable histological findings including: 1) the distribution, severity, and composition of portal inflammation; 2) the presence and type of interface hepatitis; 3) bile duct inflammation, damage, and bile ductular proliferation; 4) biliary epithelial senescence changes and small bile duct loss; 5) perivenular mononuclear cell inflammation or hepatocyte dropout; 6) lobular findings, including necroinflammatory activities; 7) the pattern of fibrosis; and 8) cholestasis.
Additional comparisons of histological features were performed between the three primary causes of late allograft dysfunction: acute cellular rejection, recurrent HBV and biliary complication. The compared histological features included duct damage, lobular activity, endothelialitis, piecemeal necrosis, fibrosis, cholestasis, bile ductular proliferation, and composition of inflammatory cells. Two pathologists blinded to patients' clinical histories reviewed all slides. After interpreting the histological findings, conflicting findings were discussed with both pathologists. If the pathologic finding was not conclusive or discrepant with the clinical finding, final diagnosis was determined based on clinical findings and the diagnosis remained indeterminate, when decision of final diagnosis was difficult in spite of all effort. Groups were compared using Pearson's χ2 and student t-tests. The results were considered statistically significant at p<0.05. All statistical analyses were performed using SPSS ver. 12.0 (SPSS Inc., Chicago, IL, USA).