Laparoscopic or Open Liver Resection Versus Multibipolar Radiofrequency Ablation of HCC Within Milan Criteria on Cirrhosis

BACKGROUND & AIMS: We compared multibipolar radiofrequency ablation (mbpRFA) with open liver resection (OLR) or laparoscopic liver resection (LLR) in patients with HCCs within Milan. METHODS: This multicenter cohort study included patients treated by mbpRFA, OLR or LLR for HCC within the Milan Criteria, developed on advanced fibrosis in nine French centres between 2008 and 2018. Adjustments on HCC size/number, gender, age, ASA, HCV, HBV, metabolic syndrome, chronic alcohol intake, MELD score, serum AFP level, APRI score, cirrhosis, significant portal hypertension and HCC localization were performed using multivariable regressions with centre-robust variance and propensity score-based matching (3:1 nearest-neighbour with calliper < 0.1; effective ratios 2.18:1 and 2.43:1). RESULTS: 1040 patients were included (median age: 64 years; male 82.9%), with 606, 266 and 168 treated by mbpRFA, LLR and OLR. Most HCCs were solitary (84.8%) and on cirrhosis (89.2%). In months, the follow-up was 50.2 (95% CI: 45.3, 54.3), 43.3 (95% CI: 40.0, 48.6) and 47.4 (95% CI: 40.8, 58.7) for mbpRFA, LLR and OLR patients. After matching, OLR patients had similar overall survival (OS) (HR = 1.05 [CI 95% = 0.65, 1.69], p = 0.853), transplant-free survival (TFS) (HR = 0.91 [CI 95% = 0.56, 1.47]; p = 0.690) or recurrence-free survival (RFS) (HR = 0.99 [CI 95% = 0.69, 1.43]; p = 0.968) compared to mbpRFA patients but with more severe adverse events (RR = 2.58 [CI 95% = 1.16, 5.71]; p = 0.02) and mortality (RR = 4.51 [CI 95% = 1.16, 17.59]; p = 0.03). After matching, LLR patients had better OS (HR = 0.57 [CI 95% = 0.38, 0.88], p = 0.01) but similar TFS (HR = 0.77 [CI 95% = 0.55, 1.1]; p = 0.152) and more severe adverse events (RR = 2.61 [CI 95% = 1.15, 5.96]; p = 0.022) compared to mbpRFA patients. LLR patients had a better RFS (HR = 0.7 [CI 95% = 0.53, 0.91]; p = 0.009) than mbpRFA patients. When distant-only recurrences were considered in the mbpRFA patients, there was no significant difference in RFS (p = 0.087). The LLR survival benefit was consistent across pre-specified subgroups, with no interaction surviving correction for multiple testing. CONCLUSIONS: Minimally invasive modalities (LLR or mbpRFA) should be prioritized when treating early HCC. LLR maximizes oncological outcomes, but mbpRFA maintains optimal TFS with less morbidity.