771 - Liver specific graded prognostic assessment can predict the outcome for patients with brain metastases from hepatocellular carcinoma

Date 28 September 2012
Event ESMO Congress 2012
Session Publication Only
Topics Hepatobiliary Cancers
Presenter Seungtaek Lim
Authors S. Lim1, S. Lee1, K. Han2, H. Choi1
  • 1Medical Oncology, Severance Hospital, Seoul/KR
  • 2Gastroenterology, Severance Hospital, Seoul/KR



After the introduction of sorafenib which showed prolongation of survival in patients with advanced HCC, the incidence of brain metastasis seemed to increase. Assessment of prognostic factors might useful to decide treatment in patients with brain metastases. Although diagnosis-specific Graded Prognostic Assessment (GPA) has been identified for several cancer types, optimal treatment strategy for brain metastasis from HCC has not been well established.


A total of 128 HCC patients were newly diagnosed with brain metastasis at Yonsei University Health Center between 1995 and 2011. Using SPSS program, univariate and multivariate analyses of the prognostic factors were performed. With the P value less than 0.10 as cutoff value, the significant prognostic factors were used to develop the HCC specific-GPA (HCC-GPA).


The median overall survival after brain metastasis in all patients was only 6.1 weeks (95% confidence interval: 4.8-7.4 week). Significant prognostic factors were the presence of extracranial lesion, and Child-Pugh-Class score. Using those variables, we newly developed HCC-GPA: extracranial metastasis (Yes: 0, No: 1 point), and Child-Pugh-Score (C:0, B:1, A:2 point). The median survival time from brain metastasis were significantly separated by HCC-GPA grouping (3.1 weeks in group 1 (HCC-GPA score 0 or 1.0, N = 50), 8.0 weeks in group 2 (HCC-GPA score 2.0, N = 55), and 22.0 weeks in group 3 (HCC-GPA score 3.0, N = 13); p< 0.001 by log rank test).


Although the overall prognosis of patients with brain metastasis from HCC is dismal, the present data showed that newly developed HCC-GPA score can discriminate the prognosis of HCC patients with brain metastasis. It can be used as a valuable tool to select patients who can be good candidates for active local treatment.


All authors have declared no conflicts of interest.