A Multi-National, Multi-Institutional Study of Comparing an Efficacy of Stereotactic Body Radiation Therapy and Radiofrequency Ablation for Hepatocellular Carcinoma

Abstract

Purpose/Objective(s): Until now, limited evidence has been available for determining the efficacy of SBRT in comparison to RFA. Our early reports of single-institutional retrospective analysis revealed that SBRT and RFA could achieve tolerable local control (LC). However, tumors located in subphrenic region and larger than 2 cm could be appropriate candidates for SBRT. In the present study, we underwent a thorough analysis for comparing the effectiveness of SBRT to RFA in patients treated at 7 tertiary-referral hospitals. Materials/Methods: Patients treated for HCC in 7 tertiary-referral hospitals were retrospectively reviewed. Among these, 1607 patients who underwent RFA of 1758 tumors and 457 patients who underwent SBRT of 479 tumors were included. Median prescribed total dose and fractional dose for SBRT were 48 (interquartile range, IQR, 40-54) Gy and 10 (IQR, 8-12) Gy, respectively. SBRT was delivered using a frameless robotic radiosurgery system (n=158; 33.0%), 3DCRT and IMRT system (n=135; 28.2%), 3-dimensional conformal radiotherapy (n=112; 23.4%), and volumetric modulated arc therapy (n=74; 15.4%). Patients were assessed after the completion of treatment at first month, every 3-6 months thereafter with CT or MRI, liver function tests, and tumor markers. Radiologic responses were evaluated using modified Response Evaluation Criteria in Solid Tumors (mRECIST) to assess LC. Results: At baseline, SBRT-treated tumors were in more advanced BCLC stage (B-C, 61.4% vs. 15.7%, p<0.001), larger (median, 2.8 cm vs. 1.8 cm; p < 0.001), and had a higher incidence of prior liver-directed treatments (median, 2 times vs. 0 time; p < 0.001) than RFA-treated tumors. The median follow-up period for the entire cohort was 28.0 (IQR, 14.1-46.0) months. The 1-year and 3-year LC rates were 77.7% and 66.0% for tumors treated with SBRT and 82.3% and 66.6% for tumors treated with RFA, respectively (p=0.232). The 1- and 3-year overall survival rates were 92.1% and 72.2% for patients in RFA group and 86.6% and 58.8% for patients in SBRT group, respectively (p<0.001). In the univariate analysis for LC, tumor size, advanced stage (B or C), and serum AFP levels were attributed to local progression. However, treatment modality was significantly correlated with LC favoring SBRT (HR 0.65, 95% CI 0.50-0.83, p<0.001). Other independent factors included tumor size, advanced stage and serum AFP level in the multivariate analysis. Though there was statistical difference in overall survival according to treatment modality in the univariate analysis favoring RFA, treatment modality (RFA or SBRT) did not contribute to survival outcomes after adjusting all clinical factors in multivariate analysis. Conclusion: Although SBRT-treatment tumors had more negative prognostic factors at baseline, SBRT provided comparable LC rates to RFA. Overall, SBRT was associated with a better LC rate than RFA in the entire cohort after adjusting clinical factors.