RESUMO
BACKGROUND & AIMS: Liver transplantation is a curative treatment option for patients with hepatocellular carcinoma (HCC) and liver cirrhosis. To date, patient selection for transplantation is based on size and number of nodules as assessed by imaging before listing. We hypothesized that changes in tumour features resulting from pre-transplant transarterial chemoembolisation (TACE) is a superior criterion to predict tumour recurrence. METHODS: 136 patients with HCC in cirrhosis with two or more cycles of pre-transplant TACE were included in this study. According to the surgical specimens, 46 patients exceeded the Milan criteria. RESULTS: Tumour recurrence occurred in 21 patients (15%). Classification of Milan criteria according to the imaging at referral was not predictive for recurrence (p=0.58), whereas the Milan criteria in the imaging immediately before transplantation reflected changes after pre-transplant TACE and were highly predictive (p<0.0001). Of the 99 patients constantly within Milan or downstaged to within Milan, 88% were recurrence-free after 5 years, compared to 55% of the patients exceeding the Milan criteria despite pre-transplant TACE. Five-year absence of recurrence was better predicted by the criterion "Progressive Disease" according to RECIST (p<0.0001). If progression was defined as any progression (including less than 20% of the sum of target lesions or new measurable lesions), predictability of recurrence in the receiver operating characteristic was 0.86. CONCLUSIONS: Imprecise assessment of size and number of tumour lesions limits prognostic importance of initial imaging. Characteristics of tumour response to TACE are reliably recognized and allow identification of suitable patients for transplantation. Future selection criteria for LT in HCC should consider this aspect.