New England Journal of Medicine May 14, 2020

Atezolizumab plus Bevacizumab in Unresectable Hepatocellular Carcinoma

Richard S. Finn, Shukui Qin, Masafumi Ikeda, Peter R. Galle, Michel Ducreux, Tae-You Kim, Masatoshi Kudo, Valeriy Breder, Philippe Merle, Ahmed O. Kaseb, Daneng Li, Wendy Verret, Derek-Zhen Xu, Sairy Hernandez, Juan Liu, Chen Huang, Sohail M. Mulla, Yulei N. Wang, Ho Yeong Lim, Andrew X. Zhu, Ann-Lii Cheng

Bottom Line

The combination of atezolizumab and bevacizumab significantly improved overall and progression-free survival compared to sorafenib as a first-line treatment for unresectable hepatocellular carcinoma.

Key Findings

1. The hazard ratio for death was 0.58 (95% CI, 0.42 to 0.79; P<0.001) in favor of the atezolizumab-bevacizumab group.
2. Overall survival at 12 months was 67.2% (95% CI, 61.3 to 73.1) with atezolizumab-bevacizumab compared to 54.6% (95% CI, 45.2 to 64.0) with sorafenib.
3. Median progression-free survival was 6.8 months (95% CI, 5.7 to 8.3) with atezolizumab-bevacizumab versus 4.3 months (95% CI, 4.0 to 5.6) with sorafenib (hazard ratio for disease progression or death, 0.59; 95% CI, 0.47 to 0.76; P<0.001).
4. Grade 3 or 4 adverse events occurred in 56.5% of patients receiving atezolizumab-bevacizumab (out of 329 evaluable) and 55.1% receiving sorafenib (out of 156 evaluable).
5. Grade 3 or 4 hypertension occurred in 15.2% of patients in the atezolizumab-bevacizumab group, while other high-grade toxic effects were infrequent.

Study Design

Design
Phase 3 RCT
Open-Label
Sample
501
Patients
Duration
Median 8.6 mo
Median
Setting
Global, multicenter
Population Patients with unresectable hepatocellular carcinoma who had not previously received systemic treatment.
Intervention Atezolizumab plus bevacizumab administered until unacceptable toxic effects occurred or loss of clinical benefit.
Comparator Sorafenib administered until unacceptable toxic effects occurred or loss of clinical benefit.
Outcome Overall survival and progression-free survival in the intention-to-treat population, assessed according to RECIST 1.1 by an independent review facility.

Study Limitations

• The open-label design may have introduced bias, particularly concerning subjective adverse event reporting and investigator-assessed metrics.
• Strict enrollment criteria requiring preserved liver function (Child-Pugh class A) limits generalizability to patients with more advanced cirrhosis.
• Because bevacizumab carries a risk of hemorrhage, patients were required to undergo screening endoscopy and were excluded if they had untreated or incompletely treated high-risk esophageal varices.
• Follow-up duration was relatively short at the time of the primary data cutoff (August 2019).

Clinical Significance

The IMbrave150 trial established atezolizumab plus bevacizumab as a new standard-of-care first-line therapy for unresectable hepatocellular carcinoma. It was the first treatment regimen to demonstrate a statistically significant overall survival benefit over sorafenib, fundamentally shifting the treatment paradigm after more than a decade of stagnation in frontline options.

Historical Context

Following the landmark SHARP trial in 2008, the multikinase inhibitor sorafenib remained the exclusive frontline standard of care for advanced hepatocellular carcinoma. Over the ensuing 11 years, numerous targeted therapies and single-agent immunotherapies failed to demonstrate superiority to sorafenib in Phase 3 trials. IMbrave150 successfully broke this efficacy ceiling by combining a PD-L1 inhibitor (atezolizumab) with a VEGF inhibitor (bevacizumab), leveraging synergistic immune activation and angiogenesis inhibition.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

What is the mechanistic rationale for combining a VEGF inhibitor (bevacizumab) with a PD-L1 inhibitor (atezolizumab) in hepatocellular carcinoma, rather than using immunotherapy alone?

Key Response

VEGF promotes angiogenesis but also creates an immunosuppressive tumor microenvironment by inhibiting dendritic cell maturation, increasing regulatory T cells, and decreasing cytotoxic T-cell infiltration. Bevacizumab reverses this immunosuppression and normalizes tumor vasculature, which synergistically enhances the ability of the PD-L1 inhibitor atezolizumab to activate an anti-tumor immune response.

Resident
Resident

Before initiating the atezolizumab-bevacizumab regimen in a patient with newly diagnosed unresectable HCC, what specific gastroenterological screening procedure is mandatory, and why?

Key Response

An upper endoscopy (EGD) is mandatory within 6 months prior to treatment. HCC patients frequently have underlying cirrhosis and portal hypertension, placing them at high risk for esophageal varices. Bevacizumab, a VEGF inhibitor, significantly increases the risk of severe mucosal hemorrhage, necessitating the prophylactic evaluation and management of varices before therapy begins.

Fellow
Fellow

The IMbrave150 trial exclusively enrolled patients with Child-Pugh class A liver function. How should a medical oncologist approach a patient with unresectable HCC and Child-Pugh B cirrhosis who desires this combination therapy?

Key Response

The safety and efficacy of atezolizumab/bevacizumab in Child-Pugh B patients were not established by this trial. Because CP-B patients have a much higher baseline risk of hepatic decompensation, bleeding, and drug toxicity, fellows must weigh the lack of high-level evidence carefully. The standard approach requires multidisciplinary tumor board discussion, often favoring alternative TKIs or best supportive care, as off-label use of bev/atezo in CP-B carries substantial, unpredictable risks.

Attending
Attending

How does the introduction of atezolizumab and bevacizumab fundamentally alter the sequencing of systemic therapies for HCC, and what challenges does it pose for interpreting second-line therapy trials designed in the post-sorafenib era?

Key Response

This trial displaced sorafenib as the standard first-line therapy. Consequently, the vast majority of established second-line data (e.g., regorafenib, cabozantinib, ramucirumab) is based on progression post-sorafenib. Attendings must navigate a major evidence gap when choosing second-line treatments, as there is limited prospective data on sequencing multi-kinase inhibitors after disease progression on upfront immune checkpoint and VEGF inhibition.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

The trial utilized both standard RECIST v1.1 and HCC-specific modified RECIST (mRECIST) criteria to assess objective response rate and progression-free survival. What is the methodological advantage of this dual assessment in HCC trials involving anti-angiogenics?

Key Response

Standard RECIST measures total tumor diameter, which may not decrease even if therapy induces massive central necrosis (a common effect of anti-angiogenics like bevacizumab). HCC mRECIST specifically measures the viable, arterially enhancing portion of the tumor. Using both ensures a rigorous, standardized baseline for regulatory comparability (RECIST v1.1) while accurately capturing the true biological and radiographic efficacy in hypervascular tumors (mRECIST).

Journal Editor
Journal Editor

The IMbrave150 trial utilized an open-label design, which inherently risks introducing bias. As a reviewer, how would you critically evaluate the impact of this open-label design on the study's primary endpoints (OS and PFS) and patient-reported outcomes?

Key Response

Overall survival is a hard, objective endpoint highly resistant to open-label bias. However, progression-free survival and patient-reported outcomes (such as time to deterioration of quality of life) are highly susceptible to expectation bias when patients and investigators know the treatment assignment. A rigorous reviewer would confirm the presence of a blinded independent central review (BICR) for imaging to mitigate PFS bias, while flagging that the PROs must be interpreted with caution due to the open-label nature.

Guideline Committee
Guideline Committee

Based on the survival benefit demonstrated in the IMbrave150 trial, how should international guidelines update the algorithm for first-line systemic therapy in unresectable HCC, and what specific patient population caveats must be explicitly integrated into the recommendation?

Key Response

The trial provides Level 1 evidence to elevate atezolizumab plus bevacizumab to the preferred, Category 1 first-line treatment for unresectable HCC, superseding sorafenib or lenvatinib. However, guidelines must explicitly restrict this strong recommendation to patients with Child-Pugh A liver function, ECOG performance status 0-1, and adequately treated esophageal varices. This reflects the exact study population and safeguards against catastrophic bleeding events in unselected, real-world decompensated cirrhotic patients.

Clinical Landscape

Noteworthy Related Trials

2008

SHARP Trial

n = 602 · NEJM

Tested

Sorafenib 400 mg twice daily

Population

Patients with advanced hepatocellular carcinoma

Comparator

Placebo

Endpoint

Overall survival

Key result: Sorafenib improved median overall survival to 10.7 months compared to 7.9 months with placebo.
2018

REFLECT Trial

n = 954 · Lancet

Tested

Lenvatinib

Population

Patients with untreated advanced hepatocellular carcinoma

Comparator

Sorafenib 400 mg twice daily

Endpoint

Overall survival

Key result: Lenvatinib was non-inferior to sorafenib in overall survival and showed statistically significant improvements in progression-free survival.
2022

HIMALAYA Trial

n = 1171 · NEJM Evid

Tested

Tremelimumab plus Durvalumab

Population

Patients with unresectable hepatocellular carcinoma

Comparator

Sorafenib 400 mg twice daily

Endpoint

Overall survival

Key result: The STRIDE regimen significantly improved overall survival compared to sorafenib with a median OS of 16.43 months versus 13.77 months.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis
↑