Personalized medicine holds immense promise for the cure and treatment of human disease. By using the genotype to design drugs, we can eliminate much of the delay, suffering, and waste associated with a trial and error process for choosing among available drugs. But important technological advances typically raise social concerns. Here are several to consider. Ideally, we will take steps to address issues of access and privacy sooner rather than later.
A question of access
Will targeted drugs be developed? Economic considerations drive companies to invest most of their resources in the development of drugs for large markets. Pharmacogenomics will facilitate the definition of markets in terms of genotypes, and for economic reasons interest will be greatest in the most prevalent genotypes, or perhaps more precisely, the genotypes most prevalent in populations with the ability to pay for drugs.
It has long been recognized that diseases affecting small numbers of individuals may be neglected. The adjective ''orphan," attached both to diseases with this characteristic and to the drugs that might be developed to treat them, underlines a moral claim to special protection. To satisfy this claim, the United States and other countries have created incentives for companies to engage in research and development activities that, if successful, will yield drugs to treat rare diseases (in the United States, conditions that affect fewer than 200,000 individuals).
The new twist is the recognition that some people with ''common" diseases may be liable to the same neglect experienced by rare disease groups. Regulators have shown a willingness to make room for these cases. For example, Herceptin, a drug that only benefits women with breast cancer whose tumor cells overexpress a particular protein, received an orphan drug designation.
Will insurers pay for these tests and drugs? Some observers have expressed the view that the interests of patients and insurance companies are aligned, since pharmacogenomics promises to lessen both the burdens and the costs associated with trial and error prescribing and adverse drug reactions. Still, there are at least four points at which the interests of patients may diverge from those of insurers:
Will testing be covered? Payers may be reluctant to cover testing if the payoff in terms of cost-savings is unclear, as could be the case with tests developed to spare patients side-effects that are uncomfortable but do not threaten life or limb or require additional funds from the insurer to treat.
Will an expensive but promising therapy be covered if testing indicates it is likely to benefit a patient? The case of Herceptin suggests that insurers will pay for an expensive drug proven to make a significant difference in patient survival. Still, sticker shock could set in as the number of targeted drugs multiplies.
Will an experimental alternative be covered if testing indicates that established therapies are unlikely to benefit a patient? For some patients, gene testing may rule out all established therapies. In the era of personalized medicine, insurers might be more willing to cover some of the costs associated with these patients' participation in research, especially if their numbers are small, but they might also insist that their role is only to pay for proven therapies.
Will an expensive but promising therapy be covered if testing indicates it is unlikely to benefit a patient, but no alternative exists and the patient would choose to assume any risks for even a small chance of benefit? In 2003 the lung cancer drug Iressa made a big splash by shrinking tumors in 10 percent of lung cancer patients in a small clinical trial. A later study suggested that the major predictor of response was a genetic mutation in tumor tissue, but at least one patient who responded did not have the mutation. At the time, questions were raised about whether insurers would or should pay for Iressa for all lung cancer patients, at roughly $2,000 per month.
These are difficult issues. Respect for patient choice and compassion support giving weight to patients' desires, but limiting demand on pooled resources based on evidence-based judgments about benefit can be defended as good stewardship.
A matter of privacy
Many pharmacogenomic tests will generate information that is relatively innocuous, with little potential to serve as a basis for discrimination. At the same time, some pharmacogenomic tests may produce information that could lead to the classification of some individuals as ''more expensive to treat" or ''more susceptible to harm from a workplace exposure." Insurers and employers may be interested in this type of information.
Also, tests performed to determine drug response could reveal information about susceptibility to disease. For example, one genetic variation was identified as significant for patients with heart disease; it was found to be associated with response to statin drugs intended to lower cholesterol. Scientists also discovered that individuals with this genetic variation are at heightened risk of developing Alzheimer's disease, a matter of much greater sensitivity.
The solution here is to strengthen general protections against discrimination in insurance and employment or, even better, to combine antidiscrimination protections with reforms that guarantee all Americans access to affordable healthcare. Pharmacogenomics did not create privacy problems or health disparities, but like any advance it has the potential to compound them.
Mary Anderlik Majumder is an assistant professor at the Center for Medical Ethics and Health Policy at Baylor College of Medicine.