By Deborah Borfitz
December 17, 2019 | The development of biomarkers for running therapeutic trials is a relatively risky endeavor for an organization to undertake, but collaboration can improve the odds of success for the benefit of all—including, most notably, federal drug regulators. “If five companies come in with the same biomarker rather than five completely different biomarkers or approaches to biomarker analysis, it helps the FDA [Food and Drug Administration] make better decisions,” says Joseph Menetski, Ph.D., associate vice president of research partnerships who leads the Biomarkers Consortium of the Foundation for the National Institutes of Health (FNIH).
The FNIH Biomarkers Consortium is a public-private partnership and almost one-third of its governing executive committee are influential leaders with the FDA, says Menetski. Consortium members from nearly 60 private organizations thereby gain access to pivotal advice and oversight “hard to find anywhere else.”
FDA representatives include Janet Woodcock, director of the Center for Drug Evaluation and Research (CDER), and Chris Leptak, co-director of CDER’s Biomarker Qualification Program. Other members of the consortium hail from the National Institutes of Health and its National Center for Advancing Translational Sciences, National Institute of Child Health and Human Development and National Cancer Institute.
Most projects undertaken by the group have regulatory intent, Menetski says, including one on kidney safety that late last year resulted in the first clinical safety biomarker for acute kidney injury (AKI) to be qualified by the FDA.
When biomarker development is taken off the plate of individual companies they can “just focus on the drug,” says Menetski. “That’s how we get our support and why we do these pre-competitive projects. Companies just want a good tool and if it’s somebody else’s tool so be it. If it’s everyone’s tool, even better.”
The FNIH is laser-focused on potentially transformative drug development tools that require group devotion to “not-so-sexy” science, Menetski says. Four steering committees—one each for inflammation and immunity, metabolic disorders, cancer, and neuroscience—drive the science that gets done and they’re seeded with experts in the biomarker field who know what’s missing but necessary to bring good therapeutics to market. “We try not to do what everyone else is doing,” he adds.
Organizations that run therapeutic intervention trials provide the researchers and funding for selected projects. Currently, the FNIH has 22 active projects and they run the gamut from multimodal and fluid biomarkers to clinical measures and what are loosely termed “digital” biomarkers, says Menetski.
“Certain words carry very specific meaning to the FDA,” he notes. A digital device measuring data “associated with how a patient feels, functions, or survives,” for example, it is a clinical outcome measure. But if the measurement is something like heart rate, it is considered a biomarker.
That’s an important definitional point when the goal is to translate interesting ideas into defined, usable tools, Menetski says. “From a regulatory point of view, it’s never good to confuse the FDA. It just slows things down.”
Some biomarkers can be as difficult and expensive to develop as a novel drug and “absolutely” require a consortium—surrogate endpoints, for example, which predict an outcome irrespective of the drug and patient, says Menetski. “That is going to require a lot of different types of data from a lot of different drugs and a lot of analysis.”
On the other end of the spectrum are patient selection or stratification markers to, for instance, predict which asthmatics will respond best to a CRTH inhibitor based on a simple sputum test. The amount of data needed would be based on the comfort level of an individual company and FDA input wouldn’t even be necessary initially, Menetski says.
Tools being developed by the Biomarkers Consortium tend to fall between those two extremes.
Measuring Tumor Burden
Projects of the Biomarkers Consortium are all well-defined and address a knowledge gap in running therapeutic trials, says Menetski. The end goal is a specific tool, resource, or definition. One of its biggest and most ambitious projects to date is Non-Invasive Biomarkers of Metabolic Liver DiseasE (NIMBLE), aimed at developing a composite biomarker for diagnosing nonalcoholic steatohepatitis in lieu of a biopsy.
“Most of our projects are not like that,” continues Menetski, and not likely to be funded by NIH because they’re focused on translational rather than clinical research. He’s particularly proud of Vol-PACT (Advanced Metrics and Modeling With Volumetric Computed Tomography for Precision Analysis of Clinical Trial Results) for its standout level of collaboration across multiple therapeutic areas to come up with a better way to measure the tumor burden of a patient.
It would be a simple exercise if it were only a matter of counting tumors, Menetski explains. But tumor size, shape, and location—and whether they’re growing—are other critical considerations.
Menetski’s tongue-in-cheek take on the current gold-standard RECIST (Response Evaluation Criteria in Solid Tumors) image analysis standards is that they’re “tin at best.” One big problem is the absence of a standard way to identify the longest axis of a detected mass.
Common practice is to make a crude estimate of length using a ruler, he says. But those eyeball estimates can differ slightly person to person and create ambiguity as to whether a tumor is or isn’t growing.
A better standard for measuring tumor burden requires lots of data, and the Vol-PACT project team convinced six private-sector companies to share raw data from 12 of their phase III clinical trials, says Menetski. The donations included imaging data on pembrolizumab, one of the foundations of Merck’s cancer therapy portfolio.
The data is being reanalyzed in hopes of identifying computable measures of tumor response and progression, Menetski says. Everything learned will be published for the benefit of everyone in the field. If intellectual property is involved, which is rare, licensing is also available to all.
Vol-PACT typifies the Biomarkers Consortium at its best, says Menetski. In addition to the collaborating companies sharing their raw data, a similar number of companies are providing financial support to the project.
Image analysis is being done by experts from multiple academic institutions, including the University of Chicago, Columbia University, Memorial Sloan Kettering Cancer Center, Harvard University and Inova, he continues. The American Society of Clinical Oncology and the Society for Immunotherapy in Cancer are in discussions at the steering committee level.
One certain benefit of robust imaging biomarker criteria is the ability to reduce the size of clinical trials—perhaps by upwards of 50%—"but that is only the beginning,” says Menetski. “Initially we were talking about measuring the longest axis of a blob, and now we can ask if it has fractured or smooth edges or if it’s wavy… things impossible to consider in the past.” Metric development is also being augmented with radiomic outcomes clinical correlation using machine learning to identify features within tumors difficult to recognize by the naked eye.
Breaking Barriers
A highly collaborative Kidney Safety Biomarkers project is just wrapping up that resulted in a set of protein-based urinary biomarkers of AKI that can be used in support of early clinical drug development, Menetski says. “It made a lot of sense to do as a consortium because no company makes a drug that they want to induce acute kidney injury, but everyone wants to know if it does and they want to know as soon as possible.” Existing biomarkers detect AKI only after the kidneys have suffered quite a bit of damage, he notes.
In October 2018, the new AKI biomarker became the first-ever clinical safety biomarker qualified by the FDA. “Qualification is the gold ring of biomarker development in the regulatory space,” says Menetski. “It’s a big deal to physicians wanting to know if someone’s kidney is injured and clinical trialists wanting to know if an experimental drug is putting participants at risk.”
Another completed project is advancing the quest for a biomarker that indicates sarcopenia. The first order of business was to agree on a clinical definition of the condition, which broadly speaking is age-related frailty that restricts mobility, says Menetski. “People who have this diagnosis tend to fall more and have increased hip fractures, which has mortality associated with it.”
During initial discussions, it became clear that “everyone was using different words to mean different things, and no one had a clear idea of what the disease was,” he says. So, during the first phase of the project, the team pulled together all the public data as the basis for generating a consensus definition of sarcopenia.
Hundreds of papers have published where findings from earlier studies have been revisited considering the new criteria, Menetski says. The second phase of the project helped quantify the definitional criteria, notably for grip strength (DOI: 10.1093/gerona/glz081) and muscle mass (DOI: 10.1093/gerona/gly129), generating response markers so trialists know when interventions are working.
The Biomarkers Consortium is likewise taking a two-stage strategy to addressing fundamental obstacles to the development of new treatments for osteoarthritis, one of the leading causes of disability in the world. “We have no drugs for it other than NSAIDS [nonsteroidal anti-inflammatory drugs],” says Menetski, a biochemist who previously worked in the private sector putting together osteoarthritis study programs.
Invariably, trialists hit a “brick wall” when it came to translating study results into the clinic because the clinical measures for assessing response to disease-modifying drugs were abysmal, Menetski says. “The things we were measuring would take three years to run and we would trial hundreds if not thousands of patients and still not really get a good result.”
Leaders in the field are now endeavoring to convert technologies like magnetic resonance imaging, together with biochemical markers such as collagen turnover and synovial inflammation, into a practical measurement tool for clinical trials. Goals include shorter studies—ideally, six months—involving fewer patients, he says. Data collection is underway, and a qualification letter of intent has been accepted by the FDA.
Tackling Digital Health
The steering committees all expressed interest in pursuing digital biomarkers last year but none of them knew where to start first in terms of developing a project, says Menetski. That’s when he was tapped by the executive committee to plan the upcoming workshop, Remote Digital Monitoring and Drug Development Trials, to figure out “how to turn interesting technology into a useable drug development tool that can be used to make billion-dollar decisions.”
About 150 leaders in the digital healthcare field will be able to personally attend the Feb. 18-19, 2020, workshop in Bethesda, Maryland, and invitations will go out to representatives from the FDA, European Medicines Agency and consumer industry, he says. Others can attend via a webcast. Registration will open by the end of the year on the website of the FNIH Biomarkers Consortium.
Jennifer Goldsack, executive director of the Digital Medicine Society, is assisting with event planning, Menetski says (see, “Digital Medicine Society Launches To Advance Digital Medicine Tools”). The workshop will focus on promoting the use of remote technologies in interventional drug development trials as well as presenting ongoing case studies in the context of regulatory decision-making and the meaning and implications of a product being “FDA cleared.”
The case studies will cover remote digital monitoring technologies for measuring heart rate, cognition, behaviors and possibly mobility, he says. One goal is to identify differences between remote monitoring technologies and standard medical devices used in physician offices, and to discuss what is gained by taking measurements 24/7 outside the clinic environment. Another is to achieve consensus between industry and regulators on basic vocabulary.