Prognostic impact of model for end-stage liver disease score in patients undergoing liver transplantation with suboptimal livers.

Abstract: Background/aims. The aim of this retrospective study is to analyze the prognostic impact of Model for End-Stage Liver Disease (MELD) score in patients undergoing liver transplantation (OLT) with suboptimal livers. Methods. Between January 2002 and January 2006, 160 adult patients with liver cirrhosis received a whole liver for primary OLT at our institution including 81 with a suboptimal liver (SOL group) versus 79 with an optimal liver (group OL). The definition of suboptimal liver was: one major criterion (age > 60 years, steatosis > 20%) or at least two minor criteria: sodium > 155 mEq/L, Intensive Care Unit stay > 7 days, dopamine > 10 mu g/kg/min, abnormal liver tests, and relevant hemodynamic instability. Results. Baseline recipients characteristics were comparable in the two study groups. The SOL group had a significantly greater number of early graft deaths (< 30 days) than the OL group, while the 3-year Kaplan-Meier patient survivals were similar. Using logistic regression, MELD score was significantly related to patient death only in the SOL group (P = .01), and the receiver operator characteristics curve method identified 17 as the best MELD cutoff with the 3-year survival of 93% versus 85% for MELD <= 7 versus > 17, respectively (P > 05). In comparison, it was 94% and 72% in the SOL group (P < .05). Similarly, MELD > 17 was significantly associated with early graft death rates only in the SOL group. Conclusion. This study advised surgeons to not use suboptimal livers for patients with advanced MELD scores, thus supporting a donor-recipient matching policy.