Materials and Methods
This was an institutional review board (IRB)-approved, single-center and retrospective analysis of all deceased donor LT procedures performed between January 2001 and December 2010 at our hospital. During the study period, a total of 929 LTs were performed on 883 patients. After exclusion of 46 re-transplantations, 10 ABO-incompatible procedures and 31 LTs for acute liver failure (ALF), 842 primary procedures were included in the outcome analysis. Eighty-five patients (10.1%) were transplanted with grafts ≥80 years; 233 (27.7%) with grafts 70–79 years; 176 (20.9%) with grafts 60–69 years and 348 (41.3%) with grafts <60 years.
Variables and Measurement
Deceased Donors. Donor data were obtained from the clinical charts at the LT unit coordination office. Eligibility to liver donation was evaluated as per our institutional policy and according to the Italian National Transplant Agency (Centro Nazionale Trapianti [CNT]) guidelines. The variables included were: age, gender, body weight, height, BMI, cause of death (trauma vs. cerebrovascular), liver and renal function tests at procurement, comorbidities (with focus on diabetes mellitus, hypertension, cardiovascular disease, renal disease, hemodialysis and dyslipidemia), use of inotropic agents, serology for hepatitis B virus (HBV) infection, HCV, human immunodeficiency virus (HIV), cytomegalovirus (CMV), Epstein–Barr virus (EBV), toxoplasma and cultures when appropriate.
The evaluation policy for octogenarian donors in place at our center is summarized in Table 1. Based on national guidelines, reasons for discarding a reported deceased liver graft were donor HIV-positivity, history of melanoma or lymphoproliferative disease, or any intractable systemic infection. History of malignancy within 5 years (10 years for breast cancer), HCV-positivity and hepatitis B surface antigen (HBsAg)-positivity required donor–recipient matching and evaluation of urgency and benefit of transplantation. Liver graft biopsy was performed on demand based on surgical evaluation at procurement and histologies were reviewed at the Pathology Institute of our hospital. At our center, a liver graft was discarded in the presence of macrovescicular steatosis ≥30%, necrosis ≥5%, fibrosis ≥2 as per Ishak's score, severe micro-angiopathy (with arteriolar thickening >60%) and macro-angiopathy precluding arterial anastomosis, unless otherwise indicated as per the recipient's clinical status. All grafts were routinely evaluated on the back table before LT for vessel patency and anatomical variants.
For the purpose of the current analysis, donor hemodynamic instability was defined as any donor experiencing cardiac arrest or requiring noradrenaline or more than one vasopressor to maintain an acceptable blood pressure before procurement.
Recipients. All LT recipients were evaluated in the pretransplant clinic and followed up after transplantation according to our institutional policy. Based on recommendations from the Italian Association for the Study of Liver Disease (AISF), in the absence of hepatocellular carcinoma (HCC) a Child-Pugh (CP) score ≥7 (until 2008) or a Model for End-Stage Liver Disease (MELD) score ≥10 (since 2011) was used as the minimal listing criterion for nonurgent candidates. Patients with a CP score <7 or a MELD score <10 were referred to LT in the presence of refractory ascites, severe or recurrent encephalopathy, severe or recurrent gastro-esophageal bleeding, recurrent bacterial cholangitis, hepatorenal, hepatopulmonary and portopulmonary syndromes (www.webaisf.org). Data included in the current analysis were demographics (age at transplantation, gender, body weight, height, BMI, donor-to-recipient gender mismatch); indication to LT; clinical status at transplantation as per MELD score and CP status; posttransplant surgical complications (type and severity according to the modified Clavien's classification system); immunosuppression; clinically relevant adverse events; incidence of treated or biopsy-proven acute rejection (BPAR); graft and patient survival. MELD scores were retrospectively re-calculated for patients transplanted before 2002. Time-dependent data were censored at time of event, latest available follow-up, or as December 31, 2012.
Graft loss (early ≤30 days vs. late >30 days) was defined as need for re-transplantation and censored at the time of re-listing or re-transplantation, at our center or elsewhere. A severe vascular complication was any posttransplant abnormality in hepatic artery, portal vein or vena cava associated with symptoms or signs and requiring a radiological or surgical procedure. Bleeding was defined as any blood loss in the first 30 days after LT and requiring re-operation. A severe posttransplant biliary complication was any abnormality in the biliary tree associated with symptoms or signs and requiring an endoscopic or surgical procedure.
Immunosuppression was with calcineurin inhibitors (CNIs) in all patients. The immunosuppressive protocols for patients included in the current data set varied according to era and availability of drugs. Excluding patient enrolled in clinical study trials, we employed cyclosporine (CSA) for HCV-positive recipients and tacrolimus (TAC) for HCV-negative patients in association with antimetabolites (azathioprine or mycophenolic acid [MPA]), steroids and/or induction with anti-CD25 mAbs for patients requiring staggered or minimized CNI introduction. Whenever clinically indicated, rejection was confirmed histologically and graded according to the Grading for Acute Liver Allograft Rejection and expressed as Rejection Activity Index (RAI). We usually treated rejection episodes ≥7 RAI with steroids in HCV-negative patients and with CNI dose adjustments in HCV-positive recipients.
Posttransplant HCV recurrence was treated after histological confirmation in case of grading ≥4, staging ≥3 according to Ishak et al or in the presence of hepatitis flares with liver function tests (LFTs) ≥3 times the upper limit range and no evidence of BPAR. Treatment varied according to era and availability of drugs, but was with interferon (IFN) and ribavirin (RIBA) in all cases. The current analysis includes a data set of HCV recipients undergoing a pilot study on preemptive antiviral treatment, as indicated elsewhere. A clinically significant HCV recurrence was the histological presence of staging ≥3 according to Ishak et al or of any clinical sign of portal hypertension not related to surgical complications.
Surgery. All donors were procured with aortic flush only with Celsior™ solution (Genzyme Italia-Sanofi, Milan, Italy) and en bloc liver and pancreas procurement, as previously described. All LT procedures were performed using the conventional technique with vena cava replacement and veno-venous bypass. A T-tube was routinely used for duct-to-duct biliary anastomosis. Variables included in the current analysis were CIT and WIT, duration of surgery, blood transfusions at surgery and intra-operative complications.
Graft Allocation Policy. Italy has a region-based donation and transplantation network whereby no national liver graft algorithm is in place with the exception of United Network for Organ Sharing (UNOS) status one patients, pediatric transplantations and patients with MELD scores ≥30 who are granted national priority. Beyond these indications, grafts are allocated within the region(s) of origin based on center-based algorithms. Until 2005, at our center grafts were allocated to adult recipients based on UNOS categories for non-HCC patients and on clinical tumor stage for HCC. In 2005, we implemented a MELD-based allocation algorithm for adult transplantation whereby, in the absence of national priorities, allocation of liver grafts <80 years was as follows: re-transplants >combined liver–kidney transplantation >MELD scores 30–25 >T3-HCC >MELD 24–23 >T2-HCC >MELD scores 22–10. For standard donors (<60 years), competing patients within the same strata were prioritized based on time on the waiting list.
For ECD, priority was granted to non-HCV recipients in order to reduce the burden of HCV graft recurrence. Exceptions to MELD scores were graded according to international guidelines published elsewhere. Donor grafts ≥80 years were deliberately not assigned to recipients with biochemical MELD score >24, except for UNOS status one patients. The transplant center director or his deputy officer made any allocation decision. The staff at the CNT monitors compliance with the proposed algorithm, which is responsible for overseeing and auditing all procedures in place at local transplant centers.
Statistical Analysis. According to type of variables and their level of distribution, descriptive statistics are reported as means ± standard deviations (SD), medians, interquartile ranges (IQR), extremes, frequencies and ranges, as appropriate. The Student's t- and analysis of variance tests (with Bonferroni's correction for multiple measurements) tests were used for continuous variables with normal distribution, while the Mann–Whitney U and Kruskal–Wallis tests were used for continuous variables with nonparametric distribution. A chi-square test was used for categorical variables. Graft and patient survival was with the Kaplan–Meier method and further stratified according to donor's age categories. The survival curves were compared with the log rank test across strata. A Cox proportional hazard model was used for univariate prediction of graft and patient survival, and included interaction terms between donor age and relevant clinical factors. A multivariate Cox regression analysis was eventually performed retaining all factors with a p-value <0.20 at univariate analysis. Hazard ratios (HRs) and 95% confidence intervals were calculated for each variable. A p-value <0.05 was considered statistically significant. All statistical calculations were performed with PASW 20.0 software (Chicago, IL).