Background
Phase III open-label RCT, N=668 patients with high-risk HCC after curative-intent resection or ablation. High-risk defined as resected T2 or higher, MVI-positive disease, multifocal tumors, or ablated tumors ≥3cm or ≥2 nodules. Recruited from 134 sites in 26 countries. Randomized 1:1 within 12 weeks of surgery/ablation.
Interventions and follow up
Arm A: Atezolizumab 1200mg IV + bevacizumab 15mg/kg IV q3wk for 17 cycles (12 months)
Arm B: Active surveillance
Primary endpoint: Recurrence-free survival (RFS) by independent review
Median follow up: 17.4mo (prespecified interim analysis)
Arm B: Active surveillance
Primary endpoint: Recurrence-free survival (RFS) by independent review
Median follow up: 17.4mo (prespecified interim analysis)
Results
RFS: NE vs NE (median not reached in either arm), HR 0.72 (adjusted 95% CI 0.53–0.98), P=.012
12-mo RFS rate: 78% vs 65% (atezo+bev vs surveillance)
12-mo RFS rate: 78% vs 65% (atezo+bev vs surveillance)
Adverse events
Overall: Grade 3–4 AEs in 41% vs 13% (atezo+bev vs surveillance); atezo and bev both discontinued for AEs in 9% of treated patients.
Serious/fatal: Grade 5 AEs in 6 patients (2%; 2 treatment-related) in the atezo+bev group and 1 (<1%) in the surveillance group.
Serious/fatal: Grade 5 AEs in 6 patients (2%; 2 treatment-related) in the atezo+bev group and 1 (<1%) in the surveillance group.
Conclusions
Adjuvant atezolizumab plus bevacizumab significantly improved recurrence-free survival versus active surveillance in high-risk resected or ablated HCC, the first phase III adjuvant study to report a positive result in hepatocellular carcinoma. The trial provided proof-of-concept that immune checkpoint inhibition can reduce post-curative HCC recurrence.
Key Limitations
Interim analysis based on median follow-up of only 17.4 months, with median RFS not reached in either arm — the absolute magnitude of benefit is not fully characterised. Updated analyses showed the RFS benefit attenuated over time (HR moving toward 1.0), raising questions about durability. OS data remain immature. The open-label design may introduce surveillance bias in the active surveillance arm.
Clinical Context
IMbrave050 established atezo+bev as the first FDA-approved adjuvant therapy for HCC (2023). The attenuating RFS benefit at longer follow-up has tempered enthusiasm, and OS benefit has not been demonstrated, fueling debate over whether RFS is a validated surrogate for OS in adjuvant HCC. STORM (sorafenib adjuvant) had previously been negative, making this the first positive adjuvant signal in HCC.