ReACT: Vaccine Improves Long-Term Survival in Recurrent EGFRvIII-Positive Glioblastoma

Adding rindopepimut to bevacizumab therapy was associated with improved long-term survival in patients with recurrent EGFRvIII-positive glioblastoma, according to data from the randomized, double-blind phase II ReACT (CDX-110) study. The findings were reported at the 20th Annual Scientific Meeting of the Society for Neuro-Oncology, held November 19-22 in San Antonio, Texas.

“Mature, randomized survival data continue to show a marked benefit with long-term survival in this small trial,” reported lead ReACT study investigator and Society for Neuro-Oncology President David A. Reardon, MD, clinical director of the Center for Neuro-Oncology at the Dana-Farber Cancer Institute, Boston. “Advantage to rindopepimut therapy was seen across multiple endpoints, including long-term progression-free survival, objective response rate, and steroid requirement.”

Patients with EGFRvIII-mutation-positive recurrent glioblastoma have poor prognoses compared to other glioblastoma patients. Rindopepimut is an investigational EGFRvIII-specific vaccine therapy consisting of EGFRvIII peptide conjugated to keyhole limpet hemocyanin (KLH), Dr. Reardon explained. It is delivered as an intradermally injected formulation of 500 µg rindopepimut and 150 µg GM-CSF, as an adjuvant. “VEGF may mediate immunosuppression,” noted Dr. Reardon. “Bevacizumab enhances immune-mediated antitumor effects in tumor models” and could also optimize EGFRvIII-specific immune response.

The study enrolled a total of 70 bevacizumab-naive patients diagnosed between May 2012 and May 2014 with first- or second-relapse EGFRvIII-positive glioblastoma. Patients were randomized 1:1 to receive either bevacizumab plus KLH (the control group) or bevacizumab plus rindopepimut.

The study’s primary endpoint was progression-free survival at 6 months (PFS6). Investigator-assessed PFS6 was 28% vs 14% for the rindopepimut and control groups, respectively (per-protocol hazard ratio [HR], 0.53 [95% confidence interval (CI), 0.31–0.91]; P = .0186), though a separate expert review found this difference did not quite reach statistical significance.

In a secondary analysis, median overall survival (OS) was 11.3 months with rindopepimut vs 9.3 months for the control group (per-protocol HR, 0.53 [95% CI, 0.32–0.88]; P = .0137), he reported. The 24-month OS rates were 25% vs 0%, respectively. Over time, the survival curves have continued to separate, “becoming even more pronounced,” he said.

Researchers identified a “remarkable frequency and level of anti-EGFRvIII immune response despite prior chemotherapy and growing tumor,” Dr. Reardon reported. EGFRvIII antibody response correlated with improved clinical outcomes, he noted.

“We see a very robust humoral response to this vaccine, and this seemed to correlate with survival,” he said.

Adding rindopepimut to bevacizumab was associated with a reduced rate of steroid use, Dr. Reardon reported. “Thirty-three percent of patients in the rindopepimut arm who were receiving steroids at baseline, were able to stop steroids for 6 months or longer, compared to none in the control arm,” he said.

Rindopepimut plus bevacizumab did not significantly worsen toxicity and was well-tolerated, “with no unexpected toxicity associated with bevacizumab and no severe adverse events attributed to rindopepimut,” Dr. Reardon reported.

There were no treatment discontinuations due to rindopepimut treatment-related adverse events. Grade 1/2 injection-site erythema and pruritus were frequent. Patients receiving rindopepimut were more likely to experience grade 3 or greater back pain (6% vs 0%) and convulsion (11% vs 0%). One grade 2 hypersensitivity reaction was reported.

“There was no evidence of increased cerebral edema,” Dr. Reardon said.

Rindopepimut was granted Breakthrough Therapy Designation by the US Food and Drug Administration (FDA) in February 2014. ACT IV, a phase III trial that completed accrual in 2014, is under way and interim results are expected in early 2016, he reported.