Background: Recently, the American Joint Committee on Cancer (AJCC) released its 8th edition changes to the staging system for hepatocellular cancer (HCC). We sought to validate the 8th edition staging system and compare the performance to the 7th edition using a population-based data set.
Methods: Using the Surveillance, Epidemiology and End Results (SEER) database (1998-2013), patients undergoing resection or transplant for non-metastatic HCC were identified. Overall survival was estimated using the Kaplan-Meier method and compared using log-rank tests. Concordance indices (c-indices) were calculated from Cox proportional hazards models to evaluate discriminatory power.
Results: The study included 8918 patients resected (63%) or transplanted (37%) for HCC. Nodal staging was performed in 19%, of whom 5% had positive nodes. The c-index for the AJCC 8th edition staging system was 0.60, similar to that for the 7th edition (0.59). Survival was better for solitary tumors >2 cm with vascular invasion than for multifocal tumors <5 cm (median not reached vs 57 months, P < 0.0001), although the staging system groups these tumors together as T2. For multifocal tumors ≤5 cm, those with vascular invasion had worse survival than those without (median 42 vs 50 months, P < 0.001), although the staging system draws no such distinction.
Conclusion: The AJCC 8th edition staging system for HCC performs similarly to the 7th edition. Future revisions should consider substratification of early HCC, specifically by distinguishing solitary tumors >2 cm from multifocal tumors ≤5 cm, and by considering the prognostic impact of vascular invasion in multifocal tumors ≤5 cm. Future studies should aim to validate these findings.
Keywords: SEER; hepatocellular cancer; staging; surgery; survival.
© 2017 Wiley Periodicals, Inc.