Overview of Prognostic Systems for Hepatocellular Carcinoma and ITA.LI.CA External Validation of MESH and CNLC Classifications

SIMPLE SUMMARY: This review proposes a comprehensive overview of the main prognostic systems for HCC classified as prognostic scores, staging systems, or combined systems. Prognostic systems for HCC are usually compared in terms of homogeneity, monotonicity of gradients, and discrimination ability. However, despite the great number of published studies comparing HCC prognostic systems, it is rather difficult to identify a system that could be universally accepted as the best prognostic scheme for all HCC patients encountered in clinical practice. In order to give a contribute in this topic, we conducted a study aimed at externally validate the MESH score and the CNLC classification using the ITA.LI.CA database. ABSTRACT: Prognostic assessment in patients with HCC remains an extremely difficult clinical task due to the complexity of this cancer where tumour characteristics interact with degree of liver dysfunction, patient general health status, and a large span of available treatment options. Several prognostic systems have been proposed in the last three decades, both from the Asian and European/North American countries. Prognostic scores, such as the CLIP score and the recent MESH score, have been generated on a solid statistical basis from real life population data, while staging systems, such as the BCLC scheme and the recent CNLC classification, have been created by experts according to recent HCC prognostic evidences from the literature. A third category includes combined prognostic systems that can be used both as prognostic scores and staging systems. A recent example is the ITA.LI.CA prognostic system including either a prognostic score and a simplified staging system. This review focuses first on an overview of the main prognostic systems for HCC classified according to the above three categories, and, second, on a comprehensive description of the methodology required for a correct comparison between different systems in terms of prognostic performance. In this second section the main studies in the literature comparing different prognostic systems are described in detail. Lastly, a formal comparison between the last prognostic systems proposed for each of the above three categories is performed using a large Italian database including 6882 HCC patients in order to concretely apply the comparison rules previously described.