Background
Retrospective comparative analysis (University of Michigan). 224 patients with HCC meeting Milan criteria (single ≤5 cm or ≤3 lesions each ≤3 cm) who received either SBRT or RFA as definitive or bridge-to-transplant therapy. Propensity score matching used to adjust for baseline differences. 63 SBRT vs 161 RFA patients (matched 63:63).
Interventions and follow up
Arm A: SBRT (median 40 Gy, 3–5 fractions; 75% prescribed to 67.5 Gy in 3 fx equivalents)
Arm B: RFA (image-guided, standard practice)
Primary endpoint: Local progression-free survival, overall survival
mFollow up: 4.5 years (SBRT) vs 3.8 years (RFA)
Arm B: RFA (image-guided, standard practice)
Primary endpoint: Local progression-free survival, overall survival
mFollow up: 4.5 years (SBRT) vs 3.8 years (RFA)
Results
2-year OS (matched): 53.9% (SBRT) vs 60.3% (RFA), P=.32 — not significant
Local progression-free survival: Similar, HR ~0.7 (SBRT vs RFA), P=.06
Treatment failure (unmatched SBRT vs RFA): Fewer local-first failures with SBRT (1.5% vs 5.6%)
Tumors >2 cm: SBRT associated with better LC (P<.05)
Local progression-free survival: Similar, HR ~0.7 (SBRT vs RFA), P=.06
Treatment failure (unmatched SBRT vs RFA): Fewer local-first failures with SBRT (1.5% vs 5.6%)
Tumors >2 cm: SBRT associated with better LC (P<.05)
Adverse events
Main adverse events: SBRT: median treatment time 2 weeks, no procedural complications. RFA: standard interventional toxicity (1 bleeding, 1 bile duct injury). Both treatments well tolerated. SBRT advantage: non-invasive, feasible for any location (central, subcapsular).
Conclusions
SBRT and RFA achieved similar overall survival and local control for HCC meeting Milan criteria, but SBRT was associated with fewer local-first treatment failures — particularly for tumors >2 cm — and avoids procedural risks, supporting SBRT as an equivalent alternative to RFA for HCC.
Key Limitations
Key Limitations: Retrospective design with propensity matching — residual confounding possible. SBRT patients had more adverse baseline features (larger tumors, higher AFP) despite matching. Follow-up durations differed between arms. Not a randomized trial — selection bias cannot be fully excluded. Local control definitions differ between modalities (RECIST-based for SBRT vs ablation-zone criteria for RFA). No bridge-to-transplant subgroup analysis reported.
Clinical Context
This analysis supports SBRT as a valid alternative to RFA for HCC ≤5 cm, especially for tumors >2 cm or in locations difficult to ablate (perihilar, subdiaphragmatic). SBRT may be preferred for centrally located HCC (risk of bile duct heat injury with RFA), tumors adjacent to major vessels, or patients with bleeding risk. RFA remains preferred for small (<2 cm) peripheral tumors given cost and logistical considerations. Prospective randomized trials (RTOG 1112 successor) comparing SBRT vs RFA are ongoing.
References