Validation of the AFP model as a predictor of HCC recurrence in patients with viral hepatitis-related cirrhosis who had received a liver transplant for HCC

Background & Aims The AFP model was shown to be superior to the Milan criteria for predicting hepatocellular carcinoma (HCC) recurrence after liver transplantation in a French population. Our aim was to test the AFP model in a non-French, post-hepatitic cirrhosis-based population of HCC candidates. Methods 574 patients transplanted for HCC in four Italian centers were studied. AFP score was assessed at the last evaluation before liver transplantation (LT). Probabilities of recurrence and survival were estimated by the log-rank test or competing risk analysis and compared according to the AFP model. Results 24.7% patients were beyond Milan criteria. HCC complicated hepatitis C virus (HCV) and hepatitis B virus (HBV) cirrhosis in 58.7% and 24% of the cases, respectively. Five-year probabilities of recurrence differed according to AFP score ⩽2 vs. >2 in the whole population (13.2 ± 1.8% vs. 49.8 ± 8.7%, p <0.001, HR = 4.98), in patients within Milan criteria (12.8 ± 2.0% vs. 32.4 ± 12.1%, p = 0.009, HR = 3.51), beyond Milan criteria (14.9 ± 4.2% vs. 58.9 ± 11.5%, p <0.001, HR = 4.26), HCV patients (14.9 ± 2.5% vs. 67.6 ± 14.7%, p <0.001, HR = 6.56) and HBV patients (11.6 ± 3.4% vs. 34.3 ± 12.5%, p = 0.012, HR = 3.49). By net reclassification improvement analysis AFP score significantly improved prediction of non-recurrence compared to Milan criteria. Overall five-year survival rates according to AFP score ⩽2 or >2 were 71.7 ± 2.2% vs. 42.2 ± 8.3% (p <0.001, HR = 2.14). Conclusions The AFP model identifies HCC candidates at low risk of recurrence, otherwise excluded by Milan criteria in a population with a predominance of post-hepatitic-related HCC. The AFP score can be proposed for selection of HCC candidates in programs with a high proportion of viral/HCV-related cirrhosis. Lay summary Selection criteria for liver transplantation of patients affected with hepatocellular carcinoma (HCC) are based on the Milan criteria, which have been shown to be too restrictive, precluding access to liver transplantation for some patients who might be cured by this operation. Recently, a French group of researchers developed a new selection model called the AFP model, or AFP score, allowing some patients with HCC not meeting Milan criteria to be transplanted with excellent results. In the present work, the AFP score was tested in a population of non-French patients transplanted for HCC occurring mainly on post-hepatitic (HCV or HBV) cirrhosis. The results confirm that in this specific population, as in the original French population of patients, the AFP model better selects patients with HCC eligible for transplantation, compared to Milan criteria. We conclude that the AFP score, which has been officially adopted by the French organization for Organ Sharing for HCC patients, can also be implemented in countries with an important burden of HCC occurring on post-hepatitic cirrhosis. The AFP model was shown to be superior to the Milan criteria for predicting hepatocellular carcinoma (HCC) recurrence after liver transplantation in a French population. Our aim was to test the AFP model in a non-French, post-hepatitic cirrhosis-based population of HCC candidates. 574 patients transplanted for HCC in four Italian centers were studied. AFP score was assessed at the last evaluation before liver transplantation (LT). Probabilities of recurrence and survival were estimated by the log-rank test or competing risk analysis and compared according to the AFP model. 24.7% patients were beyond Milan criteria. HCC complicated hepatitis C virus (HCV) and hepatitis B virus (HBV) cirrhosis in 58.7% and 24% of the cases, respectively. Five-year probabilities of recurrence differed according to AFP score ⩽2 vs. >2 in the whole population (13.2 ± 1.8% vs. 49.8 ± 8.7%, p <0.001, HR = 4.98), in patients within Milan criteria (12.8 ± 2.0% vs. 32.4 ± 12.1%, p = 0.009, HR = 3.51), beyond Milan criteria (14.9 ± 4.2% vs. 58.9 ± 11.5%, p <0.001, HR = 4.26), HCV patients (14.9 ± 2.5% vs. 67.6 ± 14.7%, p <0.001, HR = 6.56) and HBV patients (11.6 ± 3.4% vs. 34.3 ± 12.5%, p = 0.012, HR = 3.49). By net reclassification improvement analysis AFP score significantly improved prediction of non-recurrence compared to Milan criteria. Overall five-year survival rates according to AFP score ⩽2 or >2 were 71.7 ± 2.2% vs. 42.2 ± 8.3% (p <0.001, HR = 2.14). The AFP model identifies HCC candidates at low risk of recurrence, otherwise excluded by Milan criteria in a population with a predominance of post-hepatitic-related HCC. The AFP score can be proposed for selection of HCC candidates in programs with a high proportion of viral/HCV-related cirrhosis.