EGFR VACCINE NOT EFFECTIVE IN GLIOBLASTOMA

The phase 3 study ACT IV showed that rindopepimut, an investigational vaccine against EGFRvIII, did not improve outcomes when added onto temozolomide (Temodar, Merck), the current standard of care. The combination of rindopepimut plus temozolomide provided a median survival of 20.0 months compared with 20.1 months for temozolomide alone in patients with newly diagnosed EGFRvIII glioblastoma and minimal residual disease (MRD) after surgical resection and adjuvant chemoradiation.

The announcement that the trial results were negative was made some time ago, but full details of those results were published online August 22 in Lancet Oncology.

The ACT IV investigators noted that "despite the strong anti-EGFRvIII immune response generated in patients, the primary study analysis did not show a survival benefit for patients with MRD." However, they observed that the study challenged the view that MRD is a requirement for immunotherapy to work, noting a potential long-term survival benefit in patients with significant residual disease (SRD).

Also noteworthy was the observation that the control group did better than historical control groups, suggesting that outcomes have improved for this patient population.

"ACT IV was a well-conducted study that both answered questions about glioblastoma treatment but also posed intriguing new ones to be explored, reinforcing that a negative trial can still provide useful information," comments Elizabeth R. Gerstner, MD, of the Massachusetts General Hospital, Boston, in an accompanying commentary. She notes that overexpression of EGFRvIII, the truncated form of the EGFR gene, is seen in 20% to 50% of glioblastomas showing EGFR amplification, which occurs ~40% of all primary glioblastomas.

She also points out that more than 20 clinical trials have used different EGFR-targeted therapies, yet no trial in this tumor type has shown clinical benefit sufficient to warrant a US Food and Drug Administration approval.

Details of ACT IV

ACT IV was a randomized, double-blind, phase 3 study conducted in 754 patients with EGFRvIII-expressing glioblastoma (found after screening 4519 glioblastoma patients). All patients had undergone surgical resection and had received adjuvant chemoradiation therapy. They were then randomnly assigned to receive either the combination of rindopopimut and temozolomide (the treatment arm, n = 371) or temozolomide alone (control arm, n = 374).

Just more than half of the patients (405 of the 745 patients, 195 in the treatment arm, and 210 in the control arm) were centrally determined to have MRD and were included for the primary analysis of overall survival (OS) in patients with MRD.

Because rindopepimut was designed as a peptide vaccine in which a novel amino acid sequence from the EGFRvIII deletion was conjugated with keyhole limpet hemocyanin, patients in the control arm received keyhole limpet hemocyanin in addition to temozolomide.

The trial was terminated early when a preplanned analysis conducted in 2016 showed that patients treated with the vaccine were unlikely to have significant survival benefits with rindopepimut.

In this final analysis, OS for the intent-to-treat population, which was the entire cohort of patients, was also not significant ― 17.4 months for both the treatment and control groups. In patients with SRD, median OS was 14.8 months for the rindopepimut group and 14.1 months for the control group ― also not significant. But an exploratory prespecified long-term survival analysis showed a significant 2-year OS of 30% for the treatment group and 19% for the control group (P = 029).

All other clinical endpoints were also not significant, including progression-free survival, median duration of response, and objective tumor response.

However, a robust humoral response was seen with rindopepimut in patients with MRD as well as those with SRD. But high antibody titers did not consistently correspond with a clinical response.

Adverse events were not significantly different between the two groups.

Why ACT IV Failed 

ACT IV was predicated on data accrued from 100 patients who had undergone complete resection of EGFRvIII-expressing glioblastoma and had demonstrated MRD in three phase 2 studies (ACTIVATE, ACT II, and ACT III), which used rindopepimut alone or in combination with temozolomide in newly diagnosed patients.

Given this background, "why rindopepimut was uneffective is not clear," Dr Gerstner comments in his editorial. ACT IV investigators suggest that their data raise questions as to whether targeting a single tumor antigen with immunotherapy in a tumor with heterogeneous tumor expression is appropriate.

Dr Gerstner suggests that studies need to be undertaken to tease out the interplay between the different contributors to immune system activation and to identify effector cells that mediate the immune response.

In that regard, ACT IV does not bring the curtain down on trials of rindopepimut in EGFRvIII-expressing glioblastoma. Rational combinations of rindopepimut need to be evaluated, suggests Dr Gerstner. One of particular interest combines rindopepimut with bevacizumab (Avastin, Roche/Genentech). A phase 2 study in patients with recurrent disease showed promise in the ReACT study.

"There might still be a future for rindopepimut if the correct concomitant agent can be identified," Dr Gerstner commented. However, she also pointed out that multicenter and multinational trials are needed to successfully enroll enough patients when specific antigens are targets of therapy. "[ACT IV] was a monumental effort that screened more than 4000 patients to enroll 745 patients," she said.

She also noted that because the control arm in ACT IV fared better than historic controls, statistical design of trials is more challenging, "because the benchmark to beat might have to evolve."

Several authors on the ACT IV study have reported ties with industry, details of which are included in the published article. Dr Gerstner has disclosed no relevant financial relationships. 

Lancet Oncol. Published August 22, 2017