Prediction of survival after liver transplantation by pre-transplant parameters

Objective. Score-based medical urgency criteria are used for necessity-oriented liver transplantation (OLT) but lead to an increasing number of complications in patients with reduced post-OLT survival. A prediction of outcome would improve preoperative patient selection and management. Material and methods. One-hundred-and-thirty-three consecutive adult patients (63.9% men, mean age 47.4 +/- 11.2 years) given transplants between May 2004 and November 2005 at the Hannover Medical School were analysed retrospectively using univariate and multivariate methods. Results. Indications were: 27.1% viral hepatitis, 19.6% primary sclerosing cholangitis, 15.0% alcoholic liver disease, 7.5% metabolic liver disease, 6.8% primary biliary cirrhosis. Overall, 12-month patient survival was 81.2%. The mean MELD score at OLT was 14.55.3 and 12-month survival with MELD > 16 (71.7%) and < 16 (86.2%) differed significantly (p=0.041). Predictors of 12-month mortality included age (53.2 +/- 9.4 versus 46.1 +/- 11.2 years; p=0.004), lower cholinesterase (2.9 +/- 1.88 versus 3.7 +/- 2.02 kU/l; p=0.026) and serum creatinine (160.4 +/- 186.8 versus 77.7 +/- 31.6 mol/l; p=0.007), with creatinine and cholinesterase as independent parameters. Based on these parameters, a model for predicting patient survival after liver transplantation was calculated and validated in a second independent cohort of 87 OLT patients. This score identified a high-risk group and a low-risk group (overall survival 47.4 versus 91.2%; p < 0.001) with a specificity of 87.3% and a sensitivity of 68.75%. Conclusion. Age, pre-OLT creatinine and cholinesterase are predictors of short-term post-OLT survival and may be helpful as a bedside score in pre-OLT clinical management, outcome prediction and decision-making.