Personalized medicine (PMx), medical treatment tailored to specific patient populations based on their genetic or molecular biology profiles, has long been heralded as the next big thing in healthcare. It’s been about 16 years since Genentech launched Herceptin, a drug for breast cancer patients with a specific genetic mutation. At the time, Herceptin seemed to usher in a revolution for how drugs would be developed and patients would be cured.
In that new version of care, drugs could be tailored to a patient’s specific biochemical profile, dramatically improving efficacy rates and reducing the system-wide costs and complications associated with one-size-fits-all medications. For pharmaceutical manufacturers, this approach had the potential to improve sales and profits through a radically new business model: differentiated products for segmented populations (see “A Strategist’s Guide to Personalized Medicine,” by Avi Kulkarni and Nelia Padilla McGreevy, s+b, Winter 2012).
But despite the occasional success story, PMx is largely seen today as the dog that did not bark. With a few exceptions, such as Herceptin, there are few PMx success stories. This is true for several reasons.
First, health insurers remain unconvinced of PMx’s merits. One would expect these companies to push hard for personalized medicine, considering that they are the main beneficiaries of more efficient healthcare. Yet most payors seem to believe that the economic benefits of PMx are relatively small. The few PMx-based therapeutics now on the market are much more expensive than conventional therapies—and the prices don’t always translate to proportionately better outcomes, such as higher survival rates. For example, Bristol-Myers Squibb released a new metastatic melanoma therapy called Yervoy in the U.S. in 2011. Yervoy costs US$120,000, but in Phase III trials, it added only about 3.7 months of survival time.
In addition, many pharma companies have been hesitant to make the necessary investments in personalized medicine. The steep costs required, including best-in-class PMx development and commercialization capabilities, seem out of proportion to the small markets for each drug. Cancer drugs are the exception, but pharmaceutical companies have focused less on the genetic causes of other diseases. That makes PMx a costlier and riskier proposition.
More broadly, the technologies required to support PMx (to identify and quantify all the molecular markers and mutations in the body that are linked to specific diseases) are still in their infancy. The cost of sequencing the human genome has decreased, but the analysis needed to interpret the data is still expensive. And even the truest of believers are forced to admit that the next step—the molecular analysis of proteins and our understanding of the human proteome (the protein makeup of individual cells and genomes)—is many years from completion.
Even when molecular markers are identified, their absolute clinical relevance is hard to establish. Diagnostic technologies can alert scientists to certain biomarkers in the body, but not how they interact with one another and their environment to cause disease. Currently, clinical relevance has been established for an infinitesimally small number of the millions of biomarkers that the human body is capable of generating.
Finally, the reason success stories are so rare is a notable reluctance among physicians to adopt PMx. Medicine is a cautious discipline, understandably, and in some cases PMx requires practitioners to dispense diagnoses and treatments based on complex molecular changes. For example, in the 10 years since Genomic Health launched its pivotal Oncotype DX test, which can determine the recurrence risk of breast cancer and assess the likely benefit of certain types of chemotherapies, it has faced steep resistance from the medical community. Even though Oncotype DX has been proven as medically relevant technology, and been widely reimbursed by payors, analysts estimate that it is used on only half of all eligible patients.