Phase I Study Considers Potential for Personalized Immunotherapy in Glioblastoma

A peptide-based personalized vaccine demonstrated encouraging signals of efficacy and was well tolerated in patients with newly diagnosed glioblastoma, according to the results of the phase I GAPVAC-101 trial.

Hideho Okada, MD, PhD

A peptide-based personalized vaccine demonstrated encouraging signals of efficacy and was well tolerated in patients with newly diagnosed glioblastoma, according to the results of the phase I GAPVAC-101 trial.

“In tumors with low mutation load, such as glioblastoma, inclusion of carefully selected off-the-shelf nonmutated antigens is reasonable,” said Hideho Okada, MD, PhD, who presented the trial’s findings Friday afternoon at the 33rd Annual Meeting of the Society for Immunotherapy of Cancer (SITC 2018). Okada is a professor of neurological surgery and director of the University of California, San Francisco Brain Tumor Im- munotherapy Program.

Members of the Glioma Actively Personalize Vaccine Consortium (GAPVAC) conducted the multicenter, open-label, single-arm, first-in-human phase I trial to investigate the use of a personalized vaccine for treating patients with gliblastoma. The investigators sought to integrate the personalized immunotherapy into standard chemoradiation treatment for glioblastomas.

“We think that personalized vaccines should exploit the full repertoire of antigens, both mutated and non-mutated antigens, which are overexpressed in glioma cells,” Dr. Okada said.

Patients with newly diagnosed glioblastoma who had undergone a gross total 2 resection, defined as less than one cm of residual tumor found on follow-up MRI scan within 48 hours of surgery, were eligible for participation in the trial. A Karnofsky performance status of at least 70 percent and HLA-A*02:01 or HLA-A*24:04 positivity was also necessary for enrollment.

The primary endpoints for the study were safety, tolerability, immunogenicity, and feasibility. Secondary endpoints included overall and progres-sion-free survival (PFS).

Fifteen patients were treated with either a non-mutated “off-the-shelf ” peptide (APVAC1) or a synthesized patient-specific mutated peptide vaccine (APVAC2). The peptides were carefully selected based on expression profile, immunogenicity, etc.

Standard chemoradiation therapy was given for six weeks followed by temozolomide (Temodar) for four weeks. For patients receiving APVAC1, selection was done based on patient’s tumor expression profile, peptidomes from their tumor, and immunogenicity pretesting, and was followed by formulation of the non-mutated glioblastoma-associated antigens from Immatics’ peptide warehouse. Among those receiving APVAC2, patients went through mutation discovery, peptide selection, synthesis, and formulation of the personalized vaccine. APVAC2 vaccines were ready for adminis- tration approximately six months after enrollment.

Okada gave an example of peptide PTP-013 that was used in one patient that showed presentation on both normal and tumor tissue, according to both peptidome and messenger RNA analysis. The immunogenicity of this peptide was also tested in the patient’s peripheral blood mononuclear cells and found that the patient’s T cells could respond to PTP-013, so it was used for this patient’s APVAC1 vaccine. The immunomodulators poly-ICLC and GM-CSF were given in combination with either vaccine.

The APVAC1 vaccine was composed for pa- tients of up to seven nonmutated HLA class I peptides that were selected and formulated from a pre-manufactured warehouse, along with two pan-DR binding HLA class II-restricted tumor-associated peptides and one HLA class I viral marker peptide. The vaccination was given 11 times within 21 weeks, starting day 15 of the first cycle of temozolomide.

“However, we have to mention that MS-peptidome analysis did not detect any of the predicted mutated epitopes for APVAC2 from over 600 mutations,” Okada noted.

APVAC2 de novo vaccines were composed of two peptides through mutation-containing HLA class I or III peptide with MHC presentation confirmed, a mutation-containing predicted HLA presentation and immunogenicity, or nonmutated HLA class I peptides found to be expressed in the patient. The vaccination was given eight times within nine weeks, starting day 15 of the fourth cycle of temozolomide.

After a median follow-up of 31.5 months, the median overall survival (OS) was 29.0 months, and the median PFS was 14.2 months.