Validation of serological models for staging and prognostication of HCC in patients from a Japanese nationwide survey.

BACKGROUND: Two serology-based scoring models for prognostication of patients with hepatocellular carcinoma (HCC), the BALAD and BALAD-2 models, were applied to a Japanese cohort of a nationwide follow-up survey of HCC. The ability of these models to predict the progression of HCC and the deterioration of liver function and to assess prognosis was evaluated. METHODS: BALAD and BALAD-2 scores were calculated in 24,029 patients from a cohort of Japanese nationwide survey based on the serum levels of five markers (bilirubin, albumin, lens culinaris agglutinin-reactive alpha-fetoprotein, alpha-fetoprotein, and des-gamma-carboxy prothrombin) measured at the time of HCC diagnosis. The associations of these scores with the progression of HCC and liver function and with survival rates were analyzed. RESULTS: There were good correlations between BALAD and BALAD-2 scores and the progression of HCC and Child-Pugh class. Both scores accurately categorized patients into risk groups with different survival rates. BALAD-2 showed superior discrimination of patient survival compared with the original BALAD. CONCLUSIONS: Serology-based scoring models for prognostication, especially the BALAD-2 model, were useful for staging and prognostication of survival in a cohort of Japanese patients with HCC from a nationwide survey.