Background: Although many previous studies on local ablation outcomes for hepatocellular carcinoma (HCC) have dichotomized tumor size with a 3-cm cutoff to determine prognostic significance, a growing number of reports describe excellent outcomes for larger tumors. To address the logic of this 3-cm cutoff beyond small single-center experiences, we stratified patients by 1-cm tumor size intervals and hypothesized that disease-specific survival (DSS) would not vary significantly between adjacent groups.
Methods: Patients treated with local ablation for T1 HCC (≤8 cm) were identified from the surveillance, epidemiology, and end results database (2004-2008). Log-rank tests and multivariable Cox proportional hazards models were used to compare DSS curves of adjacent study groups.
Results: There were 1,083 patients included in the study (26 % female, median age: 62 years). The 3-year DSS was significantly lower in patients with 3- to 4-cm tumors compared to 2- to 3-cm tumors (58 vs. 72 %, p = 0.002). In adjusted models, DSS did not vary significantly between any size intervals up to 3 cm. Patients with 3- to 4-cm tumors, however, had a poorer prognosis compared with patients with 2- to 3-cm tumors (hazard ratio: 1.6, 95 % confidence interval: 1.18-2.18, p = 0.002). DSS also fell when tumor size increased from 5-6 to 6-7 cm (53 vs. 21 %, 0.006).
Conclusions: This study emphasizes the 3-cm size, and possibly the 6-cm size, as informative predictive thresholds when ablating HCC, because variability of DSS occurred specifically at these tumor sizes. Future research in this field should either adopt a 3-cm breakpoint or provide evidence for alternative thresholds.