Abstract
INTRODUCTION: Liver transplantation (LT) is the only curative option for patients with unrespectable hepatocellular carcinoma (HCC). In the United States. current organ allocation policies grant the same priority to patients with tumors within the Milan criteria. This uniform approach leads to higher waitlist dropout among candidated with more advanced tumors of with more aggressive tumor biology. A model to stratify HCC candidates into different risk groups could optimize organ allocation by providing priority to patients within transplantable criteria but at increased risk of dropout. METHODS: Data from 30,565 adult HCC LT candidates within the Scientific Registry of Transplant Recipients (SRTR) (2002-2022) were used. Inclusion criteria were age ≥18 years and tumors within Milan criteria. Recipients of previous transplants, multi-visceral grafts, and those with missing exception applications for HCC were excluded. The population was randomly divided into development (n = 15,282) and validation (n = 15,283) cohorts. The primary outcome was 5-year LT survival benefit, defined as the difference in survival with and without LT. RESULTS: C-MELD 3.0, serum AFP, and tumor burden score (TBS) were the strongest predictors of LT survival benefit. The HCC-Liver Transplant Survival Benefit model was defined as HCC-LTSB = 0.65 × (C-MELD 145 3.0 - 6) + 1.99 × (TBS - 2.25) + 0.68 × log2(AFP). Validation demonstrated strong performance (Pearson's r = 0.93; 95% CI: 0.93-0.94; R(2) = 0.87; C-index = 0.91). CONCLUSION: The HCC-LTSB model accurately predicted the survival benefit provided by LT in candidates listed with unresectable HCC within UNOS criteria.