A variety of clinical scores have been developed with the intent to predict early allograft failure after liver transplantation. With the present study the investigators aim to validate the recently published L-GrAFT Score on a multicenter cohort from 14 liver transplant centers in Italy.

Secondly, after identifying coefficients which are peculiar for the Italian transplant population, the investigators aim to develop a novel, simplified model for the estimation of early allograft failure (EASE Score).

Thirdly, the investigators plan to validate the EASE Score on a population from two liver transplant centers in the United Kingdom.

Liver transplantation (LT) is the gold standard treatment for end-stage liver disease. The broadening of indications has caused a growing gap between patients on the waiting list and those who receive a transplant with the consequence of patients still dying while awaiting to be transplanted.

This phenomenon has led the transplant community to expand the donors pool, thus including organs with a higher risk profile. The so called marginal organs carry a higher risk of failure especially in the early post-transplant phase. Early allograft failure (EAF) is known as a poor prognostic factor for patient survival.Treatment of graft failure is based on re-transplantation (Re-LT).

However, there are no clear-cut clinical/biochemical parameters to base the decision of Re-LT on. In addition, to which extent EAF is irreversible is not entirely predictable. Such prediction has been the objective of extensive research and debate as it can guide the physicians through the decision whether or not re-transplanting a recipient of a failing graft.

The availability of an easy algorithm to quickly identify the cases who are irreversibly heading towards graft failure and need re-LT is highly desirable.

Various definitions of EAF have been introduced but they all share the same limitation of being based on a dichotomous evaluation of biochemical parameters (e.g. AST, INR, bilirubin, etc. below or above a certain cut-off level).

Recently, a new score has been developed with the aim of overcoming this limitation: the Liver Graft Assessment Following Transplantation (L-GrAFT). This score not only provides a tool to diagnose EAF but also assesses the severity and the evolution of EAF using the kinetics of a set of biochemical parameters. However, L-GrAFT is predictive of EAF at 90 days, is based on 31 biochemical determinations and has not been validated in a multicenter setting.

With the present study the investigators aim:

1. to evaluate the performance of the L-GrAFT score and its ability to predict graft loss in the early post-operative phase (i.e. 90 days after LT) in a cohort of patients who received a primary LT from 2016 to 2017 in one of 14 LT Centers based in Italy, analyzing their prospectively maintained databases with a minimum follow up of 6 months;
2. to develop a simplified algorithm, derived from the L-GrAFT algorithm, based on the Italian LT population, reducing the number of determinations allowing an easier data entry, which is predictive of EAF at 90 (90dEASE Score) and also at 30 days post LT (30dEASE Score);
3. to validate the EASE scores on a population internal to the Italian database, using bootstrap methodology;
4. to validate the EASE scores on an external UK liver transplant Cohort. The investigators are planning to present the results of the study in international congresses and meetings and after that to produce a manuscript.

Inclusion Criteria:

• Primary liver transplants from deceased donors

Exclusion Criteria:

• combined transplants (liver transplanted simultaneously with another organ transplant)
• living donor liver transplant
• domino liver transplant
• pediatric transplants
• HIV+ recipients

A total of 1800 recipients of a primary liver transplant (LT) from 14 LT Centers from Italy (Training set) and 550 recipients of LT from 2 LT Centers from the United Kingdom (Validation set).