GBM AGILE – A Story of Convergence, Commitment, Collaboration and Compassion

By 18th June 2016 June 20th, 2016 Brain Tumour Magazine 2016/17

This article was written by Dr Anna D. Baker, Executive Committee Chair, GBM AGILE Trial, and was first published in the 2016/17 edition of Brain Tumour, the IBTA's annual magazine that is distributed globally and free of charge in 113 countries and at international neuro-oncology and cancer conferences around the world. Read the full issue online and find out more here.

Glioblastoma is Unacceptable

GBM AGILE logoTens of thousands of people are diagnosed with glioblastoma multiforme (GBM) on a global basis each year. With a median survival of 14 months and only 5% of patients living five years or longer, GBM is a grim diagnosis. Despite hundreds of GBM clinical trials, testing hundreds of agents, there have only been four approved GBM treatments, and the dismal prognosis for GBM patients has not changed in decades.

In addition, despite significant progress in molecularly characterizing GBM through projects such as the Cancer Genome Atlas (TCGA), the stark reality for GBM patients is that there are no treatment-associated biomarkers and no viable therapies with a long-term, positive treatment effect. While a number of cancer types are celebrating the dawn of precision medicine, the lack of GBM biomarkers to guide therapy (and effective treatments) means that GBM patients could be left behind – a future that is unacceptable.

Coalition of the Willing

United States Vice President Joe Biden, who lost his son Beau Biden to GBM in May 2015, joins GBM AGILE leaders at a reception following the National Press Club announcement.

United States Vice President Joe Biden, who lost his son Beau Biden to GBM in May 2015, joins GBM AGILE leaders at a reception following the National Press Club announcement.

Disillusioned with the status quo for GBM patients, Dr. Web Cavenee (Director of Strategic Alliances at the Ludwig Institute, USA), Dr. Al Yung (Director of the Brain Tumor Center at MD Anderson Cancer Center, USA), and I joined forces to organize a series of think tanks to determine what we know and don’t know about GBM – and, more importantly, to create new research models and projects that could really make a difference for patients. These first think tanks convened about 40 leaders with varied backgrounds, who had reputations for being innovators and were unified in their commitment to provide a new source of hope to GBM patients and their families. The meetings identified a number of “big ideas”, including how we might identify effective therapies for GBM and overcome barriers to progress to date. Although traditional randomized controlled trials are considered the gold standard for drug testing, they have significant limitations – especially for rare diseases.

Traditional trials require hundreds of patients to achieve statistical significance, (at a cost of over a billion dollars per agent) and take years to complete before determining if drug A works better than drug B. If the drug succeeds, we often learn little as to why, and if it fails, we often learn nothing. In all of these traditional trials, there is little if any learning from single patients and these trials cannot change to accommodate new knowledge.

Taken together, this bleak picture drove a decision from the think tank participants to form a global coalition that would plan, design, and implement the first adaptive clinical trial for GBM. And while everyone embraced this “big idea,” we were all acutely aware that such an unprecedented trial would challenge dogma and require the integration of what we know about GBM both scientifically and clinically from around the globe. GBM AGILE was born – and became in every sense a “coalition of the willing.”

“Crowdsourcing” Knowledge

Dr Anna D. Barker, Executive Committee Chair, GBM AGILE Trial

Dr Anna D. Barker, Executive Committee Chair, GBM AGILE Trial

Given the historical lack of success for GBM clinical trials, we knew that identifying effective therapies for GBM would require a “learning system” – a system that is more efficient, faster, and patient centric than traditional trials. Thus began GBM AGILE’s quest to “crowdsource” knowledge from leaders in GBM research, innovative clinical trials designers, biomarker experts, clinicians, basic researchers, and patient advocates. Fortunately, several members of the group working on GBM AGILE were part of the team that designed and implemented the I-SPY 2 trial in breast cancer, one of the most transformative clinical trials performed to date. Although there are significant differences between a rare cancer like GBM with essentially no successful therapies or biomarkers and breast cancer, lessons learned from this innovative trial have been helpful to the AGILE Global team. It is our hope that other rare cancers can similarly benefit from GBM AGILE.

The concept of combining our knowledge and expertise to create the first global adaptive trial for GBM quickly became something of a “movement,” and in a few months, our original group of 40 visionaries expanded to include over 150 leading neurosurgeons, neuro-oncologists, neurobiologists, imagers, and patient advocates from the US, China, European Union, and Australia. United by a simple goal of bringing better treatments to GBM patients, everyone stepped out of their silos, left their egos, and paid their own way to work with colleagues from around the globe to undertake the planning and development of this remarkable project we named GBM AGILE. AGILE stands for Adaptive Global Innovative Learning Environment – truly a descriptive acronym. Crowdsourcing the knowledge we need for GBM AGILE will continue throughout the trial, ensuring that learning from both inside and outside of the project is incorporated as it evolves over time.

Every Patient Counts

GBM AGILE differs from traditional clinical trials in that it will employ a Bayesian, adaptively randomized trial framework that enables testing many different hypotheses (therapeutic arms) and associated biomarkers simultaneously. Bayesian refers to a statistical model that builds on prior information to determine probability of an outcome. GBM AGILE will be a standing phase II trial, conducted under an FDA master protocol, which provides flexibility to add and drop drugs through mutually-agreed processes. The master protocol reduces time and increases the efficiency of all aspects of the trial. GBM AGILE will include both newly diagnosed and recurrent patients who do not carry the IDH1 mutation.

GBM AGILE announcement November 12, 2015 at the National Press Club in Washington, DC

GBM AGILE announcement November 12, 2015 at the National Press Club in Washington, DC

Several classes of biomarkers will be included (and further investigated) in the trial, but two categories, stratification biomarkersand enrichment biomarkers, are seminal to the design and conduct of the trial. The former will allow initial assignment to subgroups that require different treatment at baseline; and the latter will be used to predict response to specific agent(s). These two classes of biomarkers may ultimately combine to become a biomarker signature for a specific arm of the trial. The primary endpoint for GBM AGILE is overall survival, but a longitudinal model is also being developed to inform and learn what aspects of the disease may predict for survival.

Eligible primary and recurrent GBM patients will enter the trial and primary GBM patients will be stratified based on the presence or absence of a critical genetic marker involved in the repair of DNA (DNA methylation status or MGMT). Patients with methylated MGMT have a longer median survival than un-methylated patients, so MGMT status serves as an effective stratification biomarker for primary GBM patients. Following stratification, patients will be randomized to receive differential standards of care (depending on MGMT status) and further randomized based on enrichment biomarkers that are associated with specific potential therapies. Interestingly, by including both primary and recurrent GBM patients, we will learn if specific agents might work in different GBM settings.

As the study proceeds, the GBM AGILE trial (learning system) will learn from every patient, and similar patients (as determined by subtypes and enrichment biomarkers) will have a higher probability of being assigned to a successful agent. Conversely, patients will no longer be randomized to agents that perform poorly in a specific subtype. As the trial proceeds, arms will come and go – and successful agents will “graduate” with their signatures, fully ready to enter a phase III trial. This means that a pharmaceutical company can pursue a phase 3 trial using a relatively small number of patients, patient and financial resources, and, of course, time.

Adaptive Global Innovative Learning Environment (AGILE)

The National Biomarker Development Alliance (NBDA), a nonprofit organization founded by Arizona State University and focused on the development of biomarkers for precision medicine, hosts the planning and, along with foundations and other donors, provides support for the 12 committees tasked with making AGILE a reality.

The master protocol and data system will be centralized to support all facets of AGILE. AGILE will initially be implemented in the US and Australia, then followed by China and Europe. There are hurdles yet to clear, but we have an ambitious and aggressive timeline to begin enrolling patients in the autumn of 2016.

Although our efforts thus far have been fueled by volunteers and patient advocacy organizations, fundraising efforts are underway to raise much needed capital to implement the AGILE trial. We seek philanthropists, research organizations, and individuals who care about improving outcomes for GBM – and for all cancer patients and other rare diseases.


Author: Anna D. Barker, PhD, Director, National Biomarker Development Alliance; Executive Committee Chair, GBM AGILE Trial; Co-Director, Complex Adaptive Systems, Professor, School of Life Sciences, Arizona State University

For more information about the GBM AGILE Trial or to learn how to get involved, please visit

let us know you agree to cookies

The IBTA website uses cookies to give you the best online experience in accordance with our cookies policy. If you are happy, please confirm by clicking the 'Accept' button below.