Industry Insider
Regulatory, scientific, and commercial hurdles may block biomarker implementation.
Pharmaceutical companies have long trumpeted the rise of personalized medicine, where patients would be selected for tailored therapies based on biomarkers expressed either by their genetic profile or that of their disease. In truth, however, the big pharmaceutical companies always lacked an incentive to pursue this ideal. Even the most effective blockbuster drugs do not work equally in all patients. But, companies have an economic incentive to promote prescribing a drug to everybody rather than limiting a drug’s potential market by predicting in advance who will benefit.
This trend is changing as blockbusters are increasingly harder to find and new capabilities in genetic and proteomics make biomarker-based development more feasible. Perhaps the area where biomarkers have had the greatest impact to date is in oncology. Treating a cancer patient with the wrong drug can literally be a death sentence, and researchers have found several biomarkers that can effectively predict patient response to therapies. The most successful example may be found in the treatment of breast cancer where about 25% of tumors are positive for the HER2 biomarker, which normally connotes a more aggressive and deadly tumor. However, Genentech’s drug Herceptin targets HER2 and offers patients who test positive for the biomarker a much better chance at cure or extended survival with Herceptin treatment.
In addition to developing drugs that target tumor cell receptors like Herceptin, oncology researchers are using biomarkers in three ways to improve odds of trial success:
- Retrospectively searching for markers of differential response: Using retrospective biomarker analysis to identify responder subsets in trials that perhaps otherwise miss primary endpoints.
- Conducing trials in biomarker patient sub-sets: Predefining biomarker sub-sets and powering trials to draw conclusions about potential sub-segments.
- Predicting patient prognosis: Separating those patients whose genetics or tumor characteristics predict that their cancer will progress rapidly who can obscure a drug’s efficacy in better prognosis patients. For example, in breast cancer, 100+ gene panel arrays can do a better job of predicting if a patient is likely to benefit from chemotherapy maintenance.
Developers are also using biomarkers as surrogate trial endpoints, which allow companies to show efficacy for an oncology drug without having to track a trial population for several years to show a survival benefit. While biomarker endpoints are not yet routinely accepted in registration trials by the FDA, they do have utility in earlier-stage trials. For example, in ovarian cancer, levels of the CA-125 protein are used in trials to show tumor response and relapse after first remission.
|
Clinical trial challenges
Despite the increased interest on the part of researchers in the utilization of biomarkers in clinical trials, the issue remains that these types of trials require new competencies in design, data analysis, and ultimately commercialization.
For instance, the search for biomarkers starts with retrospective analyses of early stage trial results to identify biomarkers that play a role in response. This has several risky implications.
- Time and money: Biomarkers need to be routinely included in all trials, even early ones, which can dramatically increase expense and analysis time.
- Sampling and consistency concerns: Tumor and patient DNA samples must be routinely collected and stored, which limits patient and site participation. There are also issues with consistency. These tests require highly-skilled manipulation that must be performed across multiple sites.
- Trial design: For example, investigators must decide how broad of a biomarker analysis to perform. Do you analyze the tissue from the tumor, surrounding host tissue, or all of the host’s genetic markers?\
Conducting a biomarker trial adds complexity as trial sites either must have instrumentation and competencies in biomarker analysis—or developers need to contract with vendors to centrally provide these services. Attention needs to be paid to sampling methods, preservation, and transport to testing facilities. Greater statistical and computing power is also necessary to mine data for complicated relationships between genetic, biological, and environmental factors.
On top of the scientific and infrastructure issues outlined above, there are also regulatory and commercial hurdles. Receiving regulatory approval based on biomarker data is still at a nascent phase in regulatory bodies and thus can require significant extra validation. Approval paths are not well established and can be highly variable, adding another layer of risk to drug development.
Chicken or egg?
Finally, biomarker commercialization suffers from a classical chicken and egg problem. For a pharmaceutical company to succeed in marketing a drug that requires that a physician test biomarkers before prescribing, the test needs to be widely and easily available to practicing clinicians. This requires a major campaign on the part of diagnostic manufacturers to develop and launch the test.
However, there is likely to be little demand for the test, unless there is a treatment available for physicians to prescribe with a positive result. Until a drug is approved, diagnostics manufacturers have little incentive to market a test. The validation, optimization and marketing of a new diagnostic marker takes a non-trivial investment on the part of the diagnostics manufacturer, and yet these companies bear the same risk as the developers of the drug as their success is dependent on the drug’s approval. Since diagnostics typically have much lower returns, many large diagnostic manufacturers have done the math and decided that personalized medicine was not a top priority.
About the Author
Jeffrey S. Aroy has over 18 years of experience as a consultant and executive in Life Sciences including work with The Wilkerson Group/IBM Healthcare Consulting, Deloitte Consulting, WellCheck, and Berkeley HeartLab.
Filed Under: Drug Discovery
