Hepatocellular carcinoma specific graded prognostic assessment can predict outcomes for patients with brain metastases from hepatocellular carcinoma.

Stratifying patients with brain metastasis (BM) from hepatocellular carcinoma (HCC) by prognostic factors can be useful when making treatment decisions. Nevertheless, a diagnosis-specific graded prognostic assessment (GPA) for HCC has not been well established. We retrospectively reviewed the data from 118 HCC patients newly diagnosed with BM at the Yonsei University Health System between 1985 and 2011. After univariate and multivariate analyses of prognostic factors, those shown to significantly affect survival were used to develop a HCC-specific GPA (HCC-GPA) index. The median overall survival after BM in all patients was 6.1 weeks (95% confidence interval 4.8-7.4 weeks). Using the prognostic factors identified via multivariate analysis, we developed a HCC-GPA index, including number of brain metastases (single: 0.5, multiple: 0 points), alpha-feto protein (<400 ng/mL: 0.5, ≥400 ng/mL: 0 points), and Child-Pugh-Score (A: 3, B: 2, C: 0 points). There were no survival differences for age, sex, performance status, and time interval from initial diagnosis to development of BM. Median survival times from BM were discriminable when applying the HCC-GPA scoring system: 1.7, 3.2, 7.9, and 27.0 weeks for HCC-GPA scores of 0-1.0 (N = 16), 1.5-2.5 (N = 32), 3.0-3.5 (N = 49), and 4.0 (N = 21), respectively (P < 0.001). Although the prognoses of patients with BM from HCC are dismal, the newly developed HCC-GPA index can be used to discriminate the expected prognoses thereof. Moreover, the index may hold value as a tool for selecting patients who may be good candidates for active local treatment.