As reported in The Lancet Oncology by Robin Kate Kelley, MD, and colleagues, the phase III COSMIC-312 trial has shown improved progression-free survival with cabozantinib plus atezolizumab vs sorafenib in patients with advanced hepatocellular carcinoma; interim analysis of overall survival showed no significant difference.

Robin Kate Kelley, MD

Study Details

In the open-label trial, 837 patients from sites in 32 countries were randomly assigned 2:1:1 between December 2018 and August 2020 to receive cabozantinib at 40 mg once daily plus atezolizumab at 1,200 mg every 3 weeks (n = 432), sorafenib at 400 mg twice daily (n = 217), or cabozantinib alone at 60 mg once daily (n = 188).

The primary endpoints were progression-free survival on blinded independent radiology committee assessment in the first 372 patients (progression-free survival intention-to-treat [ITT] population) randomly assigned to cabozantinib/atezolizumab (n = 250) or sorafenib (n = 122), and overall survival among all patients in the combination group vs sorafenib group.

Progression-Free and Overall Survival   

Median follow-up was 15.8 months in the progression-free survival ITT population and 13.3 months in the total ITT population. In the progression-free survival ITT population, median progression-free survival was 6.8 months (99% confidence interval [CI] = 5.6–8.3 months) in the combination group vs 4.2 months (99% CI = 2.8–7.0 months) in the sorafenib group (hazard ratio [HR] = 0.63, 99% CI = 0.44–0.91, P = .0012). Rates at 6 and 12 months were 54.5% vs 40.0% and 28.5% vs 18.0%, respectively.

On interim analysis, median overall survival was 15.4 months (96% CI = 13.7–17.7 months) in the combination group vs 15.5 months (96% CI = 12.1 months–not estimable) in the sorafenib group (HR = 0.90, 96% CI = 0.69–1.18, P = .44). Rates at 6 and 12 months were 81.4% vs 76.1% and 61.8% vs 58.2%, respectively.

In prespecified exploratory subgroup analyses, progression-free survival appeared to be longer with combination treatment in patients with hepatitis B virus (HBV)-related disease, those with extrahepatic disease or macrovascular invasion, and patients enrolled in Asia, but not in other prespecified subgroups. In the interim overall survival analysis, survival appeared to be longer with combination treatment in patients with HBV-related disease, but not in other prespecified subgroups.

On interim analysis of progression-free survival in the total cabozantinib monotherapy group and sorafenib group, median progression-free survival was 5.8 months (99% CI = 5.4–8.2 months) in the cabozantinib group vs 4.3 months (99% CI = 2.9–6.1 months) in the sorafenib group (HR = 0.71, 99% CI = 0.51–1.01, P = .011).

Adverse Events

Among all patients, grade 3 or 4 adverse events occurred in 64% of those in the combination group, 46% of the sorafenib group, and 60% of the cabozantinib group. The most common were increased alanine aminotransferase (9% in combination group, 3% in sorafenib group, 6% in cabozantinib group), hypertension (9%, 8%, 12%), increased aspartate aminotransferase (9%, 4%, 10%), and palmar-plantar erythrodysesthesia (8%, 8%, 9%). Serious treatment-related adverse events occurred in 18%, 8%, and 13% of patients, respectively. Treatment-related adverse events led to death in six patients (1%) in the combination group (due to encephalopathy, hepatic failure, drug-induced liver injury, esophageal varices hemorrhage, multiple organ dysfunction syndrome, and tumor lysis syndrome), one patient (< 1%) in the sorafenib group (due to general physical health deterioration), and one patient (< 1%) in the cabozantinib group (due to gastrointestinal hemorrhage).

The investigators concluded, “Cabozantinib plus atezolizumab might be a treatment option for select patients with advanced hepatocellular carcinoma, but additional studies are needed.”

Dr. Kelley, of the Department of Medicine (Hematology/Oncology), UCSF Helen Diller Family Comprehensive Cancer Center, is the corresponding author for The Lancet Oncology article.

Disclosure: The study was funded by Exelixis and Ipsen. For full disclosures of the study authors, visit thelancet.com.