A PROMISING VACCINE FOR GBM

New Orleans, Louisiana — Patients with newly diagnosed glioblastoma multiforme (GBM) treated with the experimental heat shock protein-peptide complex-96 (HSPPC-96) vaccine plus chemoradiation showed a 146% increase in progression-free survival (PFS) and a 60% increase in overall survival (OS) compared with those receiving the standard of care alone, in a preliminary analysis of a phase 2 trial of this vaccine.

"Cancer vaccines offer the hope of benefit without toxicity for patients with glioblastoma," said Andrew T. Parsa, MD, PhD, senior investigator from the University of California San Francisco (UCSF) and the newly appointed chair of Neurological Surgery at Northwestern University, Chicago, Illinois. "Data from the HSPPC-96 trials have been consistently positive in both recurrent and newly diagnosed GBM settings, supporting the premise that this vaccine may one day become part of the standard of care," he told the media.

Preliminary results of the phase 2 trial were presented here at a Plenary Session of the American Association of Neurological Surgeons (AASN) 81st Annual Scientific meeting by Orin Bloch, MD, from the Brain Tumor Research Center at UCSF.

Dr. Bloch noted that, in addition to minimal toxicity, immunotherapy for GBM offers high tumor specificity and sustained efficacy. However, there are challenges: appropriate antigen selection, appropriate antigen delivery, and glioma-induced immunosuppression.

The HSPCC-96 vaccine (Prophage G-100, Agenus Inc) is an autologous treatment derived from dendritic cells extracted from the patient's tumor. The vaccine, therefore, contains a precise antigenic "fingerprint" of a patient's particular cancer and is designed to reprogram the body's immune system to target only cells bearing this fingerprint, reducing the risk for off-target effects. The vaccine is being studied in both recurrent and newly diagnosed GMB, Dr. Bloch explained to Medscape Medical News.

Study Design

The phase 2 open-label, single-group trial included 46 patients with newly diagnosed GBM from 8 centers who underwent gross-total resection, had good performance status, and had sufficient tumor resected to generate a vaccine. After resection, patients underwent chemotherapy with temozolomide plus radiation therapy, then began the immunotherapy. This involved injections of HSPP-96 weekly for the first 4 weeks, then monthly thereafter, until the supply was exhausted. The time to first vaccine dose was 14 weeks, and the median number of vaccine doses was 9.

The study's primary endpoint is OS, but the data are premature for that. Safety data were also not presented, but Dr. Bloch indicated that no serious adverse events occurred; the main adverse effect was injection site reaction.

At a median follow-up of 13.5 months, 27 of 46 patients had progressed (59%) and 19 (41%) had died. In the UCSF cohort of 32 patients, 21 had progressed (66%). Median PFS was 16.9 months for the whole study population and 18.0 months in the UCSF cohort, Dr. Bloch reported.

"Progression is much farther out than we see with conventional treatment," he said in an interview withMedscape Medical News.

Although cautioning against cross-trial comparisons, Dr. Bloch noted that the median PFS was shorter in other key trials of GBM treatments: 6.9 months with temozolomide (N Engl J Med. 2005;352:987-996), 13.6 months with upfront bevacizumab plus temozolomide (J Clin Oncol. 2011;29:142-148), and 14.2 months with the epidermal growth factor receptor variant III (EGFRvIII) peptide vaccine (J Clin Oncol. 2010;28:4722-4729).

"However, our population was a slightly more selected group of patients. They had good functional status and good surgical resection," Dr. Bloch noted. "Nevertheless, in comparative studies of GBM treatments, we still see median PFS on the order of about 10 to 12 months with immunotherapy, and our PFS — around 17 months — is longer than anything published."

The data are preliminary, but at this point median OS for the whole population is 23.2 months. Most enrolled patients in the trial are still being followed, and it is expected that PFS and OS will continue to mature as more data are analyzed. For the standard of care — chemotherapy and radiation — median OS is about 14 months, he said.

B7-H1 Antigen Expression

The 32 patients treated at UCSF also underwent testing for expression of the B7-H1 antigen in blood samples taken before surgery. Glioblastoma induces systemic immunosuppression through stimulation of B7-H1 expression, which could affect the efficacy of immunotherapy.

These exploratory analyses showed that patients with low expression of B7-H1 (53%) had better median PFS than those with high B7-H1 expression (47%): 21.6 months vs 11.4 months.  This finding may have the potential to help identify a more responsive patient population for future trials, Dr. Bloch said.

Researchers will also attempt to determine whether peripheral monocytes in GBM patients express significant amounts of B7-H1, whether the tumor drives expression of By-H1, and whether B7-H1–expressing monocytes promote systemic immunosuppression, he said.

In addition to the presentation of data on newly diagnosed GBM at the AANS meeting, the Cancer Therapy Evaluation Program of the National Cancer Institute (NCI) is supporting a study of the HSPPC-96 vaccine in a large randomized phase 2 trial sponsored by the Alliance for Clinical Trials in Oncology (ALLIANCE), a cooperative group of the NCI. This study represents the largest brain tumor trial ever funded by the NCI and the largest vaccine study ever conducted with bevacizumab.

The three-group ALLIANCE trial will investigate the benefits of combining HSPPC-96 and bevacizumab in 222 patients with surgically resectable recurrent GBM using a primary endpoint of OS. The study will compare efficacy of the HSPPC-96 vaccine administered with bevacizumab either concomitantly or at progression vs treatment with bevacizumab alone. This study design is supported in part by previous research indicating a potential synergistic effect between HSPPC-96 and bevacizumab.

Promising, But With Caveats

"The study of HSPPC-96 is promising and we are learning a number of things from the study, but the definitive results are not in and there are other caveats," said James Markert Jr, MD, chief of neurosurgery at the University of Alabama at Birmingham, who served as the paper's formal discussant.

"The biggest caveat is that the patients were highly selected. They had high performance status and had tumors that could be largely resected. Both of these herald a better prognosis," he pointed out. Patients who had tumor growth of 10% or more after chemoradiation were not eligible.

Other trials involving different therapies preselected or took "all comers," he explained to Medscape Medical News. But in fairness, he added, "I understand why they designed it this way and it was appropriate."

Vaccine trials tend to preselect well-resected patients because immunotherapy works best in those with minimal glioma cells, which are immunosuppressive and dampen the effect of the vaccine, he explained. "Still, at the same time we have to be careful about comparing the results with a different group of patients," he cautioned.

Dr. Bloch said another important discovery from the trial is the differential effect of the vaccine in patients overexpressing the B7-H1 antigen. "The B7-H1 antigen presents on a large fraction of circulating monocytes, and these patients do worse," he noted. It will be important to determine the mechanism underlying the difference and whether it is universal. "This finding may be important for gliomas in general," he said.

The larger study of the vaccine plus bevacizumab will be the way to determine whether the therapy is effective. "At the end of this trial, we will have a feeling as to whether the vaccine makes a difference or not," he predicted.

The study was not industry-sponsored. The vaccine was supplied by Agenus Inc. Dr. Bloch and Dr. Markert have disclosed no relevant financial relationships.

American Association of Neurological Surgeons (AANS) 81st Annual Meeting. Abstract 801. Presented May 1, 2013.