NEW YORK – Researchers have begun enrolling patients in an adaptive clinical trial aimed at uncovering new treatments for glioblastoma.
There has been little change in recent decades in treatment options for glioblastoma or in patient outcomes. About half of the patients diagnosed with glioblastoma die within the first 15 months of being diagnosed and 95 percent die within five years of being diagnosed.
The Glioblastoma Adaptive Global Innovative Learning Environment (GBM AGILE) clinical trial, which recently opened its first clinical site at the Henry Ford Cancer Institute in Detroit, is taking what it says is a more efficient approach to its clinical trial to evaluate therapies for glioblastoma.
"We got together and tried to figure out what was going to be the best thing to do … for our field because we weren't making progress," said Tim Cloughesy, a neuro-oncologist at the University of California, Los Angeles, and the global PI of the GBM AGILE trial.
The trial, which is sponsored by the nonprofit the Global Coalition for Adaptive Research, will also examine biomarkers that might act like companion diagnostics to gauge whether a drug might work in a particular patient population. Regorafenib (Bayer's Stivarga), which has shown promise as a glioblastoma treatment in a small study, will be the first drug examined in this Phase II/Phase III trial.
The first stage of the trial will be a learning stage that's akin to a randomized Phase II study, Cloughesy said. Patients will undergo randomization to either the control group, which will receive the standard of care, or the experimental drug arms, as the researchers outlined in Clinical Cancer Research in 2017. The aim of this stage is to identify treatments that affect overall survival.
How that randomization occurs will change as the researchers generate data. They are using a Bayesian algorithm to assign patients to the various arms, and that randomization process will be tweaked based on how well each of the three glioblastoma subtypes — patients with recurrent glioblastoma, newly diagnosed MGMT promoter methylation-positive glioblastoma, or newly diagnosed methylation negative glioblastoma — respond to treatment.
"Instead of having a flat randomization that you might start off with initially, you get into this adaptive randomization that occurs based upon how that particular subtype of patient is functioning within each of those different therapies," Cloughesy added.
Drugs that appear effective based on pre-set thresholds and pre-set patient numbers will then move forward into the second stage of the trial, which includes confirmation testing, while drugs that don't appear effective will be dropped.
For a third group that falls into a gray zone, where the treatment seems possibly effective, but the arm has already enrolled the maximum number of patients, the researchers will provide the data back to the companies, which can then determine if they want to pursue additional testing.
The researchers argue that this type of approach is more efficient, as it is expected to be faster, cheaper, and require fewer patients. It's faster, Cloughesy said, because the drugs are being evaluated at the same time and because the transition from Phase II to Phase III will occur more quickly. Then, because it is relying on Bayesian statistics rather than more typical power calculations that often overpower studies, and as there's a common control group, it will enroll fewer patients which will also decrease costs.
At the same time, the experimental arms will also evaluate biomarkers to identify any that distinguish patients who might respond to a particular treatment. For instance, if the arm is assessing an EGFR inhibitor, it would also test for EGFR mutations within patients in that arm. That arm then would have six populations: biomarker-positive or -negative recurrent glioblastoma, biomarker-positive or –negative methylation-positive glioblastoma, and biomarker-positive or negative methylation-negative glioblastoma.
This too will play into the adaptive nature of the trial. If, for instance, a new patient is being enrolled who has recurrent glioblastoma and is positive for a certain biomarker the algorithm will evaluate how recurrent, biomarker-positive glioblastoma was doing in all of the arms to determine how to weigh that patient's randomization.
These different signatures, Cloughesy added, can also be used to determine whether a treatment has reached its threshold in that subset of patients to move on to the second stage of the study.
He noted that they would likely also conduct a number of broader biomarker evaluations that won't be predefined, but would be more exploratory in nature.
In the end, he said he hopes they will identify biomarker-based disease subtypes linked to effective therapies. "This idea of having a biomarker-positive group all of a sudden begins to maybe break us down, like in breast cancer with its triple-negative, hormone-receptor positive, HER2-positive [groups]," he said.
The first drug in the trial, regorafenib, has been approved by the US Food and Drug Administration and the European Medical Agency to treat metastatic colorectal cancer, gastrointestinal stromal tumors, and hepatocellular carcinoma.
A 2018 report appearing in Lancet Oncology from the Phase II trial REGOMA found that regorafenib increased overall survival, as compared to the chemotherapy drug lomustine, in patients whose disease had progressed after radiotherapy and temozolomide chemoradiotherapy.
But, Cloughesy said, Bayer was unable to start a new drug application based on this data and instead sought to join GBM AGILE. By joining the trial, he said the company will be able to not only examine how regorafenib fares among patients with recurrent disease, but also among patients with newly diagnosed unmethylated or methylated disease. Bayer will both provide the drug and support the clinical trial at the sites conducting this arm.
Other drugs, evaluated by a trial committee, will be added into the trial as it takes off.
By the end of July, 20 trial sites are to be open, with it expanding to include more than 40 academic medical and community-based centers in the US by the end of September. The researchers are also seeking approvals to expand into Europe, China, Canada, and Australia.