AIM & BACKGROUND: Advanced hepatocellular carcinoma (HCC) (Barcelona clinic liver cancer [BCLC] stage C) needs subclassification to more accurately predict survival. This study aims to establish a substaging system of BCLC stage C HCC patients for accurate prognosis. METHODS: Data from 564 patients with newly diagnosed BCLC stage C HCC from three tertiary-care hospitals affiliated with the Korea University (training set) were assessed retrospectively. Variables affecting overall survival (OS) were analysed, and patients were substaged according to the number of prognostic factors they fulfilled. The substaging system was validated using a nationwide database from the Korean Liver Cancer Association (validation set; n = 742). RESULTS: In the training set, tumour factors such as tumour burden ≥10 cm, major portal vein invasion and distant metastasis, as well as underlying liver function, were independently associated with OS. BCLC stage C was classified into four substages (C1-4) according to the number of prognostic factors. Substages C1, C2, C3 and C4 showed a median OS of 17.50 months (95% confidence interval [CI], 8.57-26.43), 10.13 months (95% CI, 8.17-12.09), 4.20 months (95% CI, 3.42-4.98), and 2.90 months (95% CI, 2.34-3.46) respectively (P < 0.05). This substaging system also had good discriminative ability in predicting survival in the validation set. In addition, it was considered that the BCLC substaging is better than Hong Kong liver cancer substaging in predicting the OS for patients with advanced HCC. CONCLUSION: Our substaging for BCLC stage C might help predict patients' prognosis better.
AIM & BACKGROUND:Advanced hepatocellular carcinoma (HCC) (Barcelona clinic liver cancer [BCLC] stage C) needs subclassification to more accurately predict survival. This study aims to establish a substaging system of BCLC stage C HCC patients for accurate prognosis. METHODS: Data from 564 patients with newly diagnosed BCLC stage C HCC from three tertiary-care hospitals affiliated with the Korea University (training set) were assessed retrospectively. Variables affecting overall survival (OS) were analysed, and patients were substaged according to the number of prognostic factors they fulfilled. The substaging system was validated using a nationwide database from the Korean Liver Cancer Association (validation set; n = 742). RESULTS: In the training set, tumour factors such as tumour burden ≥10 cm, major portal vein invasion and distant metastasis, as well as underlying liver function, were independently associated with OS. BCLC stage C was classified into four substages (C1-4) according to the number of prognostic factors. Substages C1, C2, C3 and C4 showed a median OS of 17.50 months (95% confidence interval [CI], 8.57-26.43), 10.13 months (95% CI, 8.17-12.09), 4.20 months (95% CI, 3.42-4.98), and 2.90 months (95% CI, 2.34-3.46) respectively (P < 0.05). This substaging system also had good discriminative ability in predicting survival in the validation set. In addition, it was considered that the BCLC substaging is better than Hong Kong liver cancer substaging in predicting the OS for patients with advanced HCC. CONCLUSION: Our substaging for BCLC stage C might help predict patients' prognosis better.