The long-term outcomes of deceased-donor liver transplantation for primary biliary cirrhosis: a two-center study in China

Patients

We retrospectively studied data from all 69 patients who underwent DDLT for PBC from March 2006 to July 2018 at the organ transplantation center of the First Hospital of Jilin University and the First Central Hospital of Tianjin. The diagnosis of PBC is based on established diagnostic criteria of EASL (2017). The 69 patients received UDCA therapy (13–15 mg/kg/day) pre- and post-transplant, and had no OCA and immunosuppressisve treatment preoperatively. The AMAs of 69 patients were all postive before liver transplantation. We declare that this study has been approved by the Ethics Committee of the First Hospital of Jilin University and the First Central Hospital of Tianjin, written informed consent was obtained from all participants.

Calculation formulas

The model for end-stage liver disease (MELD) score was calculated using the following formula: R = 3.8*ln (bilirubin in mg/dl)+11.2*ln (international normalized ratio)+9.6 ln (creatinine in mg/dl)+6.4* etiology (etiology:0 for cholestatic or alcohol-related liver disease; 1 for others). And, the formula for aspartate aminotransferase-to-platelet ratio index (APRI): R= (aspartate aminotransferase/ the upper limits of normal)*100/(platelet count in 10 ∧9/L). The updated Mayo risk score (Murtaugh et al., 1994) was calculated by the formula: R = 6.843+0.051*[age] +1.029*ln(bilirubin in mg/dl)+1.020**ln (international normalized ratio)-3.304*ln(albumin in g/dl)+0.675* (edema score)(edema score:0 for no edema and diuretic therapy, 0.5 for edema present without diuretic or edema eliminated by diuretic therapy, 1.0 for edema despite diuretic therapy).

Histologic examination

Liver biopsy was performed when patients had unexplainable abnormal liver function post-transplant. The diagnosis of acute transplant rejection and chronic transplant rejection were based on the international Banff schema (Demetris et al., 2000). Recurrent PBC was diagnosed when either of the following findings were observed: (i) granulomatous cholangitis or epithelioid granulomas without any sign of infection or (ii) dense lymphoid aggregates in the portal tracts not associated with perivenular inflammation or endotheliitis (Haga et al., 2007).

Serologic liver function

For each patient, liver function tests [aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), gamma-glutamyl transpeptidase (GGT), total bilirubin (TBIL), direct bilirubin (DBIL), albumin (ALB)] were performed immediately before the transplant, daily for 4 days after the transplant, and weekly for 4 weeks posttransplant. If AST or/and ALT was greater than 1500 U/L at 1 week after transplant, it was classified as initial poor graft function (IPGF) (Kocbiyik et al., 2009). Of the initial 69 patients, 5 patients were excluded from the analysis of liver function results (5 due to uncompleted data). In total, the liver function results of 64 patients were analyzed.

Diagnosis of complications

Complications were diagnosed based on symptoms, laboratory results, and radiological findings. Infection-related complications were diagnosed based on established diagnostic criteria (Horan, Andrus & Dudeck, 2008). Infections can be classified by anatomy (such as respiratory system, renal system, biliary system) or by microbiology (such as bacteria, fungus, virus). In this study, most infection complications were based on diagnostic microbiology; some infection complications with negative diagnostic microbiology were diagnosed based on clinical presentation and radiology.

Biliary complications included strictures (anastomotic and nonanastomotic), leakage, and bile tumors. Anastomotic stricture was diagnosed as a segmental stricture located within one cm of the biliary anastomosis as shown by magnetic resonance cholangiopancreatography (MRCP), endoscopic retrograde cholangiopancreatography (ERCP), or percutaneous cholangiography. Ischemic-type biliary lesions were diagnosed when either MRCP or ERCP showed typical signs of segmental or diffuse intrahepatic or extrahepatic strictures or biliary tree destruction. Biliary leakage was determined by abdominal fluid drainage tube or bile drainage under ultrasound-guided puncture. Color Doppler ultrasound was performed the second day after liver transplant or with abnormal liver function. Abdominal computed tomography with contrast or angiography was performed when complications of the portal vein or hepatic artery were detected by color Doppler ultrasound.

Diabetes mellitus was defined as a fasting glucose higher than seven mmol/L or a random glucose higher than 11.1 mmol/L and a hemoglobin A1C higher than 6.5%.

Statistical analysis

Descriptive statistics are presented as median and range for continuous variables and counts and percentages for categorical variables. Overall survival and PBC recurrence-free curves were estimated according to the Kaplan–Meier method. The log-rank test was used to evaluate whether each variable significantly affected postoperative patient survival and PBC recurrence. The hazard ratio and 95% confidence interval were calculated by multivariate Cox regression analysis. A P-value of 0.05 was considered to be statistically significant.

Patient characteristic

In our study, the follow-up period ranged from 0.5 to 107 (median, 32) months. The baseline characteristics of the 69 patients are summarized in Table 1. The median age was 55 years (range, 39–67), and 60 patients (87%) were female. The MELD and APRI were 19 (range, 6–31) and 2.26 (range, 0.11–42.22), respectively. The median updated Mayo risk score was 9.53 (range, 4.34–12.44). The median levels of AST and ALT were 75 (range, 15–513) and 45 (range, 6-229), respectively. The median ALP and GGT levels were 186.5 (range, 54–732)and 93.5 IU/L (range, 14–779), respectively. The median value of TBIL was 9.83 (range, 0.41–41.6 mg/dl). The median value of ALB was 3 g/dl (range, 2.3–4).

Of these cases, 26 patients were classified as Child-Pugh grade B; 43 patients, grade C. One patient had a second liver transplant due to thrombosis of the hepatic artery 2 weeks after the first liver transplant.

APRI greater than 2 associated with poor survival of PBC patient survival after liver transplantation

During the follow-up, three patients died. One died from abdominal hemorrhage after transplant, and the other two died from infection and liver failure at 14 months posttransplantation. Overall patient survival in the study population was 98.6% at 1 year, 95.1% at 3 years, and 95.1% at 5 years after transplantation (Fig. 1). Multivariate Cox regression analysis were failed to perform due to the small samples of death.A log-rank test of the Kaplan–Meier life table analysis showed that a pre-transplant APRI greater than 2 was related to poor survival (P = 0.0018) (Fig. 2). The Kaplan–Meier life table analysis with a log-rank test failed to find that age, gender, the MELD score, Child-Pugh score, Mayo risk score , pre- and post- operative liver function were statistically significant in predicting the survival of liver transplantation.

Younger age at liver transplantation associated with a greater risk for recurrent PBC

Of 69 patients, 7 patients who had unexplained abnormal liver function underwent percutaneous liver biopsy postoperatively. Only 1 patient was diagnosed rejection reaction, and another 6 patients were diagnosed recurrent PBC. The incidence of recurrence increased with time after transplantation (Fig. 3), leading to relapse rates of 3.5% at 1 year, 8.1% at 3 years, and 21.8% at 5 years after transplantation. The median duration from liver transplantation to the diagnosis of recurrent was 21.5 months. The Kaplan–Meier life table analysis with a log-rank test revealed a preoperative factor of recipients aged 48 years or younger as the only significant risk factor of disease recurrence (Fig. 4). In a multivariate Cox regression analyse model adjusted for gender, pre-operative Mayo risk score, APRI score, MELD score, a preoperative age of 48 years or younger remained the only independent risk factor of patient recurrence (hazard ratio 0.028, 95% confidence interval 0.01–0.71, P = 0.03)

Normal liver function at 4 weeks posttransplantation

The curves of AST and ALT changes posttransplantion were similar—reaching a peak on the first day, dramatically decreasing on the second day (AST declined more than ALT), returning to normal by 1 week, and remaining stable with normal levels at 4 weeks (Fig. 5A). The curves of ALP and GGT changes were also somewhat similar after transplant (Fig. 5B)—both returned to normal on the first day but then slowly increased. However, ALP peaked at the seventh day and then returned to the normal range; GGT peaked at 2 weeks (2 times higher than the upper cutoff) and then slowly decreased to normal by 4 weeks.

The levels of TBIL and direct bilirubin (DBIL)decreased after transplantation to normal by 4 weeks (Fig. 5C). The level of ALB eventually increased to normal by 4 weeks (Fig. 5D). The peak levels of ALT and/or AST were greater than 1500 U/L in 15 patients. The frequency of IPGF was 21.7%.

Infection and biliary complications

Postoperative complications of the 69 patients are summarized in Table 2. A total of 43 patients (62%) developed 63 episodes of infections in our follow-up period. Of these patients, 16 developed multiple infectious complications. The vast majority of infections were microbiological infection. Among the cases of infection without evidence of microbiology, there were four cases of pneumonia, one upper respiratory infection, one bile duct infection, and one urinary infection. The 56 microbiologically defined episodes were categorized by the following causes: 42 (75%), bacteria; 6 (11%), fungus; 6 (11%), mixture of bacteria and fungus; and 2 (4%), virus.

During the follow-up, 18 (26%) patients had biliary complications, including 7 anastomotic strictures (10%), 10 nonanastomotic strictures (14%), 3 leakage (4.3%), and 1 biloma (1.5%). One patient had biliary leakage and nonanastomotic strictures; two patients had anastomotic strictures secondary to leakage. Among the five patients with vascular complications, two(2.9%) had portal vein thrombosis, and three(4.3%) had hepatic artery thrombosis. One patient had a second liver transplant due to liver failure caused by nonanastomotic strictures secondary to hepatic artery thrombosis. Only one (69) patient had a rejection reaction, which occurred at 33 months. Other complications included osteoporosis (5), hepatitis B infection (3), and diabetes (2).