Nivolumab vs Bevacizumab in Recurrent Glioblastoma

Get Permission

As reported in JAMA Oncology by David A. Reardon, MD, and colleagues, the phase III CheckMate 143 trial showed no difference in overall survival among adult patients with a first recurrence of glioblastoma treated with nivolumab vs bevacizumab following standard radiation and temozolomide therapy.

David A. Reardon, MD

David A. Reardon, MD

Study Details

In the open-label trial, 369 patients from sites in 12 countries were randomly assigned between September 2014 and May 2015 to receive nivolumab at 3 mg/kg (n = 184) or bevacizumab at 10 mg/kg (n = 185) every 2 weeks until disease progression or unacceptable toxicity. The MGMT promoter was methylated in 43 patients (23.4%) in the nivolumab group and 42 patients (22.7%) in the bevacizumab group and unmethylated in 59 (32.1%) vs 67 (36.2%), with status unknown in the remaining patients. A total of 73 patients (39.7%) vs 79 patients (42.7%) were using corticosteroids at baseline. The primary endpoint was overall survival in the intent-to-treat population.  

Overall Survival

Median follow-up was 9.5 months at data cutoff (January 2017). Median overall survival was 9.8 months (95% confidence interval [CI] = 8.2–11.8 months) in the nivolumab group vs 10.0 months (95% CI = 9.0–11.8 months) in the bevacizumab group (hazard ratio [HR] = 1.04, 95% CI = 0.83–1.30, P = .76), with 12-month overall survival rates of 42% vs 42%.

In exploratory hypothesis-generating subgroup analyses, multivariate analysis indicated that no baseline corticosteroid use vs baseline use (HR = 0.59, 95% CI = 0.36–0.95) and methylated vs unmethylated MGMT promoter status (HR = 0.47, 95% CI = 0.29–0.78) were associated with longer overall survival in the nivolumab group. Methylated MGMT promoter status (HR = 0.54, 95% CI = 0.32–0.89) was also associated with longer survival in the bevacizumab group. Among 31 patients treated with nivolumab vs 22 patients treated with bevacizumab with methylated MGMT promoter status and no baseline corticosteroid use, median overall survival was 17.0 vs 10.1 months (HR = 0.58, 95% CI = 0.30–1.11). Among patients with corticosteroid use at baseline, overall survival was poorer with nivolumab treatment (HR = 1.41, 95% CI = 1.01–1.97). 


  • No difference in overall survival was observed for nivolumab vs bevacizumab.
  • Median overall survival was 9.8 vs 10.0 months.

Median progression-free survival was 1.5 months vs 3.5 months (HR = 1.97, P < .001). Objective response rates in patients evaluable for response were 7.8% vs 23.1%, with median response durations of 11.1 vs 5.3 months.

Adverse Events

Treatment-related grade 3 or 4 adverse events occurred in 18.1% of patients in the nivolumab group, with the most common being fatigue (3.3%), vs 15.2% of the bevacizumab group, with the most common being hypertension (7.9%). Neurologic treatment-related adverse events of any grade occurred in 13.7% (grade 3 or 4 in 4.4%) vs 9.7% (grade 3 or 4 in 1.2%). The most common immune-mediated adverse events were diarrhea (14.8% vs 7.9%), increased alanine aminotransferase (8.2% vs 5.5%), and rash (9.3% vs 4.2%). No treatment-related deaths were observed.

The investigators concluded, “To our knowledge, the CheckMate 143 randomized clinical trial is the first phase III study investigating the use of a [programmed cell death protein 1] inhibitor in patients with recurrent glioblastoma. The study did not meet the primary endpoint of overall survival. The safety profile of nivolumab in patients with glioblastoma was consistent with that in other tumor types. Patients with methylated MGMT promoter glioblastoma and no baseline corticosteroid use may potentially derive benefit from treatment with immune checkpoint inhibition.”

Dr. Reardon, of Dana-Farber/Harvard Cancer Center, is the corresponding author for the JAMA Oncology article.

Disclosure: The study was supported by Bristol-Myers Squibb. For full disclosures of the study authors, visit