As reported in The New England Journal of Medicine by Richard S. Finn, MD, and colleagues,  the phase III IMbrave150 trial has shown that anti–programmed cell death ligand 1 (PD-L1) plus anti-VEGF therapy with atezolizumab plus bevacizumab improved progression-free and overall survival vs sorafenib in patients with unresectable hepatocellular carcinoma who had received no prior systemic therapy.

Richard S. Finn, MD

Study Details

In the open-label trial, 501 patients from 111 sites in 17 countries with unresectable hepatocellular carcinoma were randomly assigned 2:1 between March 2018 and January 2019 to receive atezolizumab plus bevacizumab (n = 336) or sorafenib (n = 165). Treatment consisted of atezolizumab at 1,200 mg and bevacizumab 15 at mg/kg intravenously every 3 weeks or oral sorafenib at 400 mg twice daily, with treatment continuing until loss of clinical benefit or unacceptable toxicity.

Randomization was stratified by geographic region (Asia [excluding Japan] vs the rest of the world), macrovascular invasion or extrahepatic spread of disease, baseline alpha-fetoprotein (AFP) level (< 400 vs ≥ 400 ng/mL), and Eastern Cooperative Oncology Group performance status (0 vs 1). The co-primary endpoints were overall survival and progression-free survival in the intention-to-treat population as assessed at an independent review facility using Response Evaluation Criteria in Solid Tumors, version 1.1.

Concerning patients in the atezolizumab/bevacizumab vs sorafenib groups: 40% vs 41% were from Asia (excluding Japan) and the remainder were from the United States, Australia, New Zealand, and Japan; all patients had ECOG performance status of 0 (62% in both) or 1 (38% in both); 72% vs 73% had a Child-Pugh score of A5; 82% vs 81% had Barcelona Clinic Liver Cancer stage C disease; 38% vs 37% had AFP ≥ 400 ng/mL; and 77% vs 73% had macrovascular invasion (38% vs 43%), extrahepatic spread (63% vs 56%), or both. Among 182 patients with known status, 64% vs 67% had PD-L1–positive tumors. 

Progression-Free and Overall Survival

At data cutoff (August 2019), median follow-up was 8.6 months. At the time of the primary analysis, 28.6% of patients in the atezolizumab/bevacizumab group and 39.4% of patients in the sorafenib group had died, with overall survival being significantly improved in the atezolizumab/bevacizumab group vs the sorafenib group (stratified hazard ratio [HR] = 0.58, 95% confidence interval [CI] = 0.42–0.79, P < .001).

Overall survival at 12 months was 67.2% vs 54.6%. Median progression-free survival was 6.8 months vs 4.3 months (stratified HR = 0.59, 95% CI = 0.47–0.76, P < .001), with 6-month rates being 54.5% vs 37.2%. Objective response rates were 27.3% vs and 11.9% (P < .001), with complete response in 18 (5.5%) vs 0 patients. Duration of response was > 6 months in 87.6% vs 59.1%.

In overall survival subgroup analyses, hazard ratios for atezolizumab/bevacizumab vs sorafenib were: 0.53 (95% CI = 0.32–0.87) in Asia (excluding Japan) and 0.65 (95% CI = 0.44–0.98) for the rest of the world; 0.67 (95% CI = 0.43–1.06) for performance status of 0 and 0.51 (95% CI = 0.33–0.80) for performance status of 1; 0.69 (95% CI = 0.29–1.65) for those without and 0.55 (95% CI = 0.39–0.77) for those with macrovascular invasion, extrahepatic spread, or both; and 0.52 (95% CI = 0.34­–0.81) for those with AFP < 400 ng/mL and 0.68 (95% CI = 0.43–1.08) for those with AFP ≥ 400 ng/mL.  

In progression-free survival subgroup analyses, hazard ratios were: 0.46 in Asia (excluding Japan) and 0.70 for the rest of the world; 0.57 for performance status of 0 and 0.63 for performance status of 1; 0.72 for those without and 0.53 for those with macrovascular invasion, extrahepatic spread, or both; and 0.49 for those with AFP < 400 ng/mL and 0.79 for those with AFP ≥ 400 ng/mL.

For patient-reported outcomes assessed with the EORTC quality-of-life questionnaire QLQ–C30, patients receiving atezolizumab/bevacizumab reported statistically significantly longer median times to deterioration of overall quality of life (11.2 vs 3.6 months), physical functioning (13.1 vs 4.9 months), and role functioning (9.1 months vs 3.6 months).

Adverse Events

Median durations of treatment were 7.4 months with atezolizumab, 6.9 months with bevacizumab, and 2.8 months with sorafenib. Grade 3 or 4 adverse events occurred in 56.5% of patients in the atezolizumab/bevacizumab group vs 55.1% of patients in the sorafenib group. The most common adverse events in the atezolizumab/bevacizumab group were hypertension (15.2% vs 12.2%), increased aspartate aminotransferase (7.0% vs 5.1%), and increased alanine aminotransferase (3.6% vs 1.3%).  

Serious adverse events occurred in 38.0% vs 30.8% of patients; no specific serious adverse event had an incidence that differed in the two groups by ≥ 2%. Adverse events led to discontinuation of treatment in 15.5% (both atezolizumab and bevacizumab in 7.0%) vs 10.3% of patients. Adverse events led to death in 15 patients (4.6%) vs 9 patients (5.8%). Grade 3 or 4 adverse events of special interest in the atezolizumab/bevacizumab group included potential atezolizumab-related (immune-mediated) events in 25.8% (vs similar events in 30.1% of the sorafenib group), with the most common being hepatitis (21.3% as diagnosis or laboratory abnormality) and infusion-related reactions (2.4%). Potential bevacizumab-related (VEGF inhibitor–related) grade 3 or 4 events occurred in 23.1% (vs similar events in 18.6% of the sorafenib group), with the most common being hypertension (15.2%) and bleeding/hemorrhage (6.4%).

The investigators concluded, “In patients with unresectable hepatocellular carcinoma, atezolizumab combined with bevacizumab resulted in better overall and progression-free survival outcomes than sorafenib.”

Disclosure: The study was funded by F. Hoffmann–La Roche/Genentech. For full disclosures of the study authors, visit nejm.org.