¹Section of Hematology/Oncology, Department of Medicine, Committees on Clinical Pharmacology and Pharmacogenomics and Molecular Medicine, and Cancer Research Center, University of Chicago, Chicago, Illinois, USA

Correspondence: MJ Ratain, (mratain@medicine.bsd.uchicago.edu)

While lost on the back roads of Long Island en route to a recent meeting of the National Institutes of Health (NIH) Pharmacogenetics Research Network, my passenger and collaborator suggested that I needed a global positioning system receiver to solve our current challenge related to pharmacogenetics (Mary Relling, personal communication). We frequently know where we need to go but often have no idea how to get there, and we thus rely on technology to lead us to our destination. Personalization of medical care, particularly in the prescription of drugs, is our goal. In a utopian implementation, whole-genome data would be used in every prescribing intervention. Furthermore, the data would be used in a way that neither burdened the prescriber nor put patients at risk for unintended consequences from release of their genomic information. To achieve this vision, we need to build a genomic prescribing system (GPS) that will both improve outcomes and lower societal medication costs.

What are the elements of the GPS? We need inexpensive whole-genome typing, rapid access to the results, knowledge of the significance of the results in the context of a drug or indication (i.e., valid pharmacogenetic tests), a method for determining the appropriate prescription, and protection of the genetic data from inappropriate use. In other words, we need lower genotyping costs, more research, and appropriate software—all of which are within our means. Genotyping costs are plummeting, and whole-genome typing can be obtained for less than $500, a trivial cost if the data are useful for life. Access to the results could be provided through the Internet, with appropriate security measures similar to those provided for Internet banking.

What are the challenges in developing a commercial market for pharmacogenetic tests? Unlike many newer diagnostic tests (e.g., HIV viral load), pharmacogenetic tests would need to be determined only once in an individual's life. This greatly limits the opportunity for revenues and thus the incentive to innovate through the usual commercial approach. In addition, one cannot patent a particular allele or haplotype, only the use of genomic information relating to the use or dosage of a particular medication. Such patents can be difficult to obtain and even more difficult to defend from infringement. In addition, the future regulatory environment is unknown and is likely to become more rigorous if genotyping results are indicated to select a specific medication. The overall result is that the commercial markets have little interest in developing pharmacogenetic tests, a necessary component of the overall GPS.

But who could profit from a GPS? Companies involved in genotyping and/or sequencing will have plenty of business, but the opportunities for large profits will be limited by the commoditization of the technology. In addition, such companies may also develop new services, such as management of the data for individual patients. The actual use of the data will probably require the development of complex systems for computer-assisted prescribing. Notably, it is unlikely that companies currently involved in diagnostics or therapeutics development will have a significant role in the GPS of the future (Figure 1).

Figure 1.

GPS (Genomic Prescribing System).

Full figure and legend (83K)

Can a complete GPS be created through a market-based approach? We generally assume that medical innovations (e.g., new drugs and devices) will emanate from opportunities to profit in our capitalistic economy. In the usual path, companies are formed around one or more inventions, and they attract capital because of the billions of dollars per year in revenues often associated with success. However, the business model required is markedly different from the usual models for developing drugs or devices, with which the primary risk is regulatory risk. Developing a GPS requires a business model that combines service and technology, but it is complicated by indeterminate regulatory and patent risk. In addition, the market is unknown, because the economic benefit of widespread pharmacogenetic testing is unknown. Unfortunately, developing the market requires either funding the development of pharmacogenetic knowledge or reliance on development of such knowledge by others. Whether a new industry will emerge to provide pharmacogenetic services is unknown.

Are there alternative models for development of a GPS? Given that pharmacogenetic tests will be performed only once in an individual's lifetime and that their evaluation may require randomized trials, it is unlikely that the potential revenues (at least as based on current pricing of pharmacogenetic tests) associated with successful development will justify the costs and risks. Because the public and payers will gain from the development of a GPS through both improved outcomes and lower medication costs, it is reasonable to suggest that they should fund the development of the pharmacogenetic knowledge required to fuel the system.

What are the options for public funding of a GPS? The most obvious option is to build the GPS under the auspices of the federal government. Given the increasing federal stake in medication costs, this opportunity should be strongly considered. Might the NIH create a new National Institute of Personalized Medicine? Or should the Department of Health and Human Services (HHS) create a new agency for this purpose, perhaps called the Genomic Prescribing System Administration?

One such solution has been proposed by Senator Barack Obama in the "Genomics and Personalized Medicine Act of 2006," which was introduced in the US Senate on 3 August 2006 and was referred to the Committee on Finance. The proposed bill finds that more than 100,000 deaths in the United States are attributed annually to adverse drug reactions, and as many as 15% of hospitalized patients (2.2 million) have serious adverse drug reactions every year. The bill goes on to expand on positive effects of pharmacogenomics in therapy solutions, listing as examples imatinib, trastuzumab, warfarin, 6-mercaptopurine, and others.¹ Obama proposes the creation of a Genomics and Personalized Medicine Interagency Working Group (IWG) within the HHS. The IWG would be charged with coordinating HHS activities related to pharmacogenomics and personalized medicine. Obama also proposes a national Biobanking Research Initiative, citing similar efforts in other countries (e.g., Sweden). In addition, he proposes funding for training in applied genomics and economic incentives for development of pharmacogenomic diagnostics, making provisions for $150 million to sponsor research on genomics, and a 100% tax credit for the development of companion diagnostic tests that can improve drug safety.

Although an act of Congress bolstering personalized medicine would be a dream come true, it is not the only approach to building the requisite GPS. An alternative would be to create a GPS foundation, not with federal involvement, but with the support of payers suffering from the rapidly increasing costs of medications. Such a foundation could be started with philanthropic funds and then funded through annual contributions from payers, who would be members of the foundation. Members would receive all the services of the foundation at a low cost, and it could also provide services to nonmembers at market rates. These services would include genotyping, data management (including protection of patients' genetic information), education of health-care providers, and assistance with pharmacogenetic-guided dosing. Continuing funding of the foundation would be supported by the annual payers' contributions as well as by revenue from services provided to both members and nonmembers. The founding members would include insurance companies and large self-insured companies (e.g., General Motors), and might eventually expand to include the governments of countries that provide medications to their citizens.

The goal is clear, but the best route for attaining it is unknown. Thus, we should begin debate on the various options for a GPS, which will require input from scientists, physicians, economists, payers, philanthropists, and the federal government. Ideally, some of this debate will occur on the floor of the US Senate. Continuing to focus only on the research without discussing the implementation of the results will eventually lead to frustration of investigators desperate to see their research translated into clinical practice. In addition, the public will be frustrated by an apparent failure by the scientific community to translate research dollars into better (and more affordable) pharmacotherapy. The legacy of this generation of pharmacogenetic researchers needs to be more than publications in the medical literature. Otherwise we are doomed to continue to prescribe expensive medications for large populations, as opposed to truly personalizing medicine.