American Association for Clinical Chemistry
Better health through laboratory medicine
July 2007 Clinical Laboratory News: The Push for Integrating Biomarkers into Drug Development

July 2007: Volume 33, Number 7

The Push for Integrating Biomarkers into Drug Development
Will a New Collaborative Accelerate Clinical Use?
By Julie McDowell

Biomarkers have long been celebrated as the future of personalized diagnostic medicine—not only for detecting early forms of disease, but also for monitoring a patient’s disease and response to therapy. This potential has yet to be realized, however, and biomarkers remain stalled in research, rather than applied in clinical settings. But a recently announced joint effort by the American Association for Cancer Research (AACR), the Food and Drug Administration (FDA), and the National Cancer Institute (NCI) is seeking to carve out a regulatory pathway for biomarkers—taking them from the research phase through the FDA approval process and then on to use in clinical laboratories. Called the Cancer Biomarkers Collaborative (CBC), this initiative is focused on developing recommendations on assay validation and bioinformatics that would integrate biomarkers into drug development and clinical trials, a move many leaders in the laboratory medicine community believe will provide much-needed standardization for biomarker development, as well as realize the potential of personalized medicine.

“The driving force behind forming the CBC was the shared recognition that to realize the acceleration of drug approvals based on modern cancer biology, we really need to have guidelines that the FDA can use as they develop the guidance for industry to get biomarkers approved,” said AACR President William Hait, MD, PhD, who is also Senior Vice President and Worldwide Head for Hematology Oncology Research Development at Johnson & Johnson (Raritan, N.J.). “It’s one thing for individual laboratories out in the academic world or industry to develop biomarkers, but it’s quite another thing to get them to the standards that would be acceptable to the FDA. The impact will ultimately be some clear guidelines from the FDA for laboratories trying to develop diagnostic tests in this area, so they can get them approved.”

Accelerating Information Sharing

Announced in March at the AACR Annual Meeting in Los Angeles, the CBC will make recommendations to the FDA’s Critical Path Initiative, which was launched in March 2004 to ensure that scientific discoveries translate rapidly into better medical treatments, as well as create new tools to evaluate the safety and effectiveness of medical devices at a faster rate, lower cost, and with better information.

The CBC will be comprised of four committees that will focus on developing recommendations for guidelines in four areas: biospecimens, bioinformatics, assay validation, and information sharing, according to another CBC leader, Samir Khleif, MD, a Senior Investigator with NCI and Special Assistant to the FDA Commissioner. “In the area of biospecimens, the plan is to come up with recommendations for standards and specifications on how to collect biospecimens and integrate them into the clinical trials to reach the desired endpoint of biomarker measurement and be able to compare these endpoints among clinical trials,” he explained. “In terms of validation, we want to identify and define how to validate a biomarker assay and how to consider a biomarker assay eligible to be included and integrated into clinical trials.” The third focus area, data or information sharing, involves developing a model for expanding the “precompetitive” space. This means that companies, sponsors, and developers would be able to share data without jeopardizing their intellectual property, enabling research and development to advance in a more cooperative and collaborative manner. “For the last focus area, bioinformatics, we want to determine how to utilize IT in the integration of biomarkers within clinical trials and advance the field,” said Khleif.

Need for Standards, Validation

Many involved in cancer biomarker research and development are excited about forthcoming standards and guidance in this area. “There are still a lot of questions as to the best way to develop biomarkers, and I think getting some guidelines put together with the FDA’s involvement as well as the NCI and clinical groups is a fantastic initiative,” said Kenneth R. Evans, PhD, President and CEO, Ontario Cancer Biomarker Network (OCBN, Toronto, Canada), a non-profit organization created by the provincial government to further biomarker research. “One of my biggest concerns in this field is the lack of existing standards in a variety of different areas. As a result, people are often doing things that are very different, using different methods, different approaches, and as a result, they can’t replicate each other’s results. The world of drug development is much better defined and has clear endpoints and targets, as well as clear regulatory rules, but that’s not the case in the diagnostics world.”

Clinical laboratorians involved in biomarker research are also enthusiastic about validations efforts, as well as establishing data sharing protocols. “Validating biomarkers—especially tumor markers or markers for other low-incidence diseases—is problematic because you need large sample sets from well-defined patient groups. With that comes a pretty significant expense, unless you have a large number of banked specimens. Even then, it can be expensive to launch these in trials,” said William Clarke, PhD, Assistant Professor of Pathology at Johns Hopkins Medical Institutions in Baltimore, Md. “The other issue is repeatability. We see these promising studies, but then other labs are not able to reproduce the same results. It would be significant if the CBC could pull together all of the different groups—the IVDs, the regulatory agencies, oncologists, researchers, medical and laboratory medicine researchers—to get a large enough specimen base so they can validate these markers in order to take it to evidence-based practice. This would show that not only can these biomarkers detect cancer, but also show whether they are useful in terms of affecting patient outcomes.”

Developing Companion Diagnostics

By confronting the need for guidelines for cancer biomarker development, the CBC is taking on a recommendation put forth in a report by the Institute of Medicine (IOM), released in late March. The report, Cancer Biomarkers: The Promises and Challenges of Improving Detection and Treatment, states that industry and stakeholders should share precompetitive data on the validation and qualification of biomarkers. The report goes on to suggest that government agencies and stakeholders develop a transparent process to create well-defined consensus standards and guidelines for biomarker development, validation, qualification, and use; and that the FDA and industry should work to facilitate the co-development and approval of diagnostic-therapeutic combinations (See Tables, below).

Use of Biomarkers in Drug Development
Biomarker Use
Drug Development Objective
Target validation Demonstrate that a potential drug target plays a key role in the disease process
Early compound screening Identify compounds with the most promise for efficacy and safety
Pharmacodynamic assays Determine drug activity; select dose and schedule
Patient selection In clinical trials, patient selection (inclusion/exclusion) by disease subset or probability of response/adverse events
Surrogate endpoint Use of a short-term outcome measure in place of the long-term primary endpoint to determine more quickly whether the treatment is efficacious and safe in drug regulatory approval

Source: Cancer Biomarkers: The Promises and Challenges of Improving Detection and Treatment, Institute of Medicine.

Reprinted with permission from the National Academies Press, Copyright 2007, National Academy of Sciences.


Biomarker Validation and Qualification Requires Demonstration Of Fitness for a Specified Purpose
Type of Biomarker Definition Purpose
Exploration Research and development tool Hypothesis generation
Demonstration Probable or emerging biomarker Decision making, supporting evidence with primary clinical evidence
Characterization Known or established biomarker Decision making, dose finding, secondary/tertiary claims
Surrogacy Biomarker can substitute for a clinical endpoint Regulatory approval

NOTE: Shown are four categories of biomarkers used for drug development and their intended purpose.

Source: Cancer Biomarkers: The Promises and Challenges of Improving Detection and Treatment, Institute of Medicine.

Reprinted with permission from the National Academies Press, Copyright 2007, National Academy of Sciences.

Establishing an FDA pathway for companion diagnostics is vital if tests are going to be accepted as meaningful endpoints in drug development, according to David Parkinson, MD, Senior Vice President of Oncology Research and Development for Biogen Idec (San Diego, Calif.), who was also a co-author of the IOM report. “For example, let’s say we have an assay that is critical in determining whether or not a particular patient is likely to benefit from a drug or is likely to have toxicity from the drug, and therefore requires dose modifications,” he explained. “If this assay becomes inextricably linked with the use of the drug, this information will need to be in the registration package and label, and therefore, the test itself needs to be of a quality and reproducibility to satisfy all the criteria used for FDA approval to be incorporated into the use of the drug.”

However, a drug-diagnostic co-development approval process doesn’t exist within the FDA, although officials outlined their preliminary thoughts on approving these devices in a 2005 report titled, “Concept Paper on Pharmacogenomic Drug Diagnostic Co-development”. In addition, officials say they are finalizing a March 2005 final guidance on pharmacogenomics data submissions, although they declined to provide a scheduled release date. Despite the delay in updating both these documents, any future guidance on these issues will certainly dictate the role of pharmacogenomics and the development of new drugs, including how clinical labs will report test results designed for those drugs.

In addition to the potential for companion diagnostics, there is also an important role for biomarkers in drug development, said Parkinson. “Biomarkers are increasingly important in the development of the drug; in determining whether the drug is doing what we designed it to do and determining how the patients react to the drug,” he explained. “There are a number of different areas in which the development of a series of biomarker tools would more accurately characterize a patient’s tumor or the patients themselves in the case of drug metabolism. This could contribute to better and more efficient drug development.”

Parkinson’s thoughts on co-development are echoed by Daniel Chan, PhD, Director, Center for Biomarker Discovery, and Director of Clinical Chemistry Division and Department of Pathology at Johns Hopkins Medical Institutions in Baltimore, Md. He is also a Principle Investigator with the NCI’s Early Detection Research Network (EDRN), which is focused on translating new molecular diagnostics into practical clinical tests that identify cancers at the earliest stages. EDRN is also working on evaluating and validating biomarkers. Chan is also on the National Academy of Clinical Biochemistry committee that wrote Tumor Markers 2006: Practice Guidelines and Recommendations For Use Of Tumor Markers In The Clinic, which will be presented in their updated form at this month’s AACC Annual Meeting in San Diego.

Through his experience developing biomarkers with industry, Chan finds that pharmaceutical companies will often develop a drug and proceed into clinical trials, only to discover it is ineffective or toxic to some patients. “Then they ask me if I can find a marker to indicate which group of patients will respond—or not respond—to their treatment,” he said. “Then I try to come up with markers for patients that aren’t performing well, or ones that would cause a lot of toxicity. But that’s the wrong way to develop drugs, in my opinion. These markers need to be developed at the same time as the pharmaceutical agents.”

However, while Chan welcomes any recommendations on standardization and validation from the CBC, he is concerned that there are several ongoing efforts that appear to be working on the same or related issues (See Box, below). In addition to the CBC and the EDRN, there is also the recently announced Severe Adverse Event (SAE) Consortium. This group, made up of representatives from pharmaceutical companies and the clinical research community, will identify and qualify biomarkers, including pharmacogenetic markers, for adverse events, establish appropriate databases and repositories, and develop effective intellectual property and data-pooling strategies. Specifically, the SAE Consortium will look at the role of biomarkers in the adverse events such as hepatotoxicity. “I think that it’s going to be confusing for multiple groups to be working on similar issues,” said Chan. “We should encourage cooperative efforts instead of everyone working on their own. However, the FDA appears to be involved in all of these efforts, which is promising. So this kind of effort like the Cancer Biomarkers Collaborative is great, but it’s important that the right group and the right people get involved in this.”

Moving Biomarkers into the Future

Below are some current initiatives working on biomarker development and validation:

Cancer Biomarker Collaborative
Who: AACR, FDA, and NCI
Mission: To facilitate use of validated biomarkers in clinical trials and evidence-based oncology and cancer medicine.

Early Detection Research Network
Who: National Institutes of Health’s National Cancer Institute
Mission: To translate new molecular knowledge into practical clinical tests that identify cancer at the earliest stages, and to identify individuals at risk of developing cancer.

Serious Adverse Events Consortium
Who: FDA, Pharmaceutical Companies
Mission: To identify and qualify biomarkers for adverse events, including pharmacogenetic markers and identifiers; leverage/establish appropriate databases and repositories; coordinate the development of common research platforms, nomenclature, and standards; and establish effective intellectual property and data-pooling strategies.

Ensuring Lab Involvement

Getting the right groups—such as the clinical laboratorians and oncologists—and the right people involved in creating any biomarker standards is vital, said Chan, who added that in his experience, the right people might not be involved in setting the performance criteria for cancer markers. “When I work with in vitro diagnostic companies who are developing new cancer markers, it’s often the business development or marketing people who set the performance criteria, including specificity, sensitivity, and precision of these assays,” he explained. These things need to be defined by clinical laboratorians, not only in industry but also in whatever biomarker guidelines are defined by regulatory and federal efforts like the CBC. “We will be doing the testing for the biomarkers, and we know what the performance characteristics of the assays should be, as well as the performance criteria,” he said. “Therefore, we should really be the ones who set the performance criteria.”

Evans, from Canada’s OCBN, also believes that laboratorians should be involved in developing cancer biomarker standards, particularly when it comes to issues like how specimens are collected and handled in research and what specific data from clinical trials needs to be reported. “When researchers measure clinical specimens and look for differences among disease states without having adequate information on how that sample is collected and the specifics of the population that they came from, they may be missing some information that is unique to the specimen population,” he explained. “Therefore, when others try to replicate the study, they don’t find the same results, so they ignore the result, even though it can be significant.”

How samples are collected from various populations and differences in sample handling parameters can dramatically impact a trial’s results. Therefore, Evans hopes that the CBC will offer some recommendations in this area, because identifying subtle differences in populations is what drug diagnostic co-development needs to focus on. “When you talk about the future of drug and biomarker co-development, and identifying a population that you are going to treat with that drug, I believe that it’s important for biomarker research to move in that direction,” said Evans. “This is the key to personalized therapeutics, which is what we are all after.”