Background: We sought to measure the impact of model for end stage liver disease (MELD) score, tumor staging, and microvascular invasion (MVI) on the relative survival benefit of liver transplantation (LT) versus liver resection (LR) for hepatocellular carcinoma (HCC).
Methods: The study population comprised 1,106 HCC patients with cirrhosis undergoing LR from one Eastern (n = 424) and two Western (n = 682) surgical units. Exclusion criteria were very large (>10 cm) tumors, macrovascular invasion, and metastases. We identified three tumor stages: stage I (within Milan, n = 806), stage II (beyond Milan within Up-to-7, n = 123), and stage III (beyond Milan and Up-to-7, n = 177). Patient survival after LR was compared to that predicted after LT by the Metroticket calculator in relationship with staging, MVI, and MELD score using Monte Carlo simulation.
Results: Two hundred eighty-three patients (26 %) with a MELD score of ≥10 had an acceptable 5-year survival after LR of 47 %, while that of patients with a low MELD score was 67 % (p < 0.0001). Mean 5-year LT benefit was -4.50 months (95 % confidence interval [CI] -4.73 to -4.27) for patients with a MELD score of <10, and 0.81 months (95 % CI 0.58 to 1.04) for those with a MELD score of ≥10. MELD score and MVI were the strongest predictors of transplant survival benefit. LT reached a survival benefit, versus LR only in HCC patients with a MELD score of ≥10 and without MVI (3.08 months, 95 % CI 2.78 to 3.39), whatever the tumor stage.
Conclusions: LT proved to be harmful in patients with resectable HCC with a low MELD score (<10) or with aggressive tumors (with MVI). As a result of a shortage of donors, only selected resectable tumors with a MELD score of ≥10 should be considered for transplantation.