Markkula Center of Applied Ethics

Pharmacogenomics, Ethics, and Public Policy

By Karen Peterson - Iyer

ABSTRACT

The advent of pharmacogenomics - the study of how the human genome influences drug response within a person or population - has begun to drive the development of pharmaceuticals in Western medicine today. Although pharmacogenomics promises dramatic improvement in drug safety and efficacy, the field also raises a host of ethical questions. The need to protect informed consent and confidentiality and to promote justice and equity - both nationally and globally - requires that one approach pharmacogenomics with an enthusiastic, yet critical, eye. Drawing on the normative values of respect for persons (as both autonomous and relational), human well - being, socioeconomic justice, and human solidarity and the common good, this article offers several concrete suggestions for public policy to help ensure that pharmacogenomics develops in a way that promotes the good of both individuals and the broader society.

When my first child was a newborn, he had a problem: He threw up. A lot. Of course, spitting up is a common, indeed normal, activity of most newborns but this was not an ordinary case. After regular feeding sessions, literally cups of milk would come back out of his mouth, sometimes projecting across a table or even a room. One of the doctors we consulted about the problem prescribed a drug designed to aid in propelling food through the digestive tract. We tried one dose on our son; and, almost instantly, our otherwise normal, gurgling, alert baby became so zombie - like that we barely recognized him - sleeping nearly 24 hours a day and almost entirely unable to focus or interact with us. Needless to say, we stopped the drug immediately and, thankfully, our baby bounced back within a few days to his normal, bubbly, mostly - awake self.


Of course, what my son was experiencing was a very mild adverse drug reaction (ADR). Something about his biological make - up did not allow him to tolerate well a drug that was designed to help him. According to some studies cited by the Federal Drug Administration, more than two million serious ADRs take place every year, and these account for more than 100,000 deaths annually among hospitalized patients in the United States alone. Eliminating, or at least reducing, ADRs is one of the primary goals of those pursuing what is commonly referred to as "personalized medicine." The idea behind personalized medicine is simple. Rather than accepting the "one-size-fits-all" approach of ordinary drug therapy-where dosage is the only primary variant from patient to patient-researchers envision that medications would be tailored to fit the profile, especially the genetic profile, of each individual patient.


Scientists have recognized for years that there is indeed some genetic basis for individualized drug response, an insight that has evolved into the burgeoning field of pharmacogenomics, the study of the correlation between an individual patient's genotype and his or her response to drug treatment. Although pharmacogenomics promises fantastic benefits, it also raises a host of ethical issues. Promises first: A better understanding of the genetic basis for drug response could enable scientists to hone in on the most critical aspects of drug action on the body, improving drug safety and making it easier to prescribe the right dose for each person. Moreover, drug development itself could be expedited through a more streamlined, safer, and less expensive clinical trial process, wherein only those individuals with particular genetic profiles-those apparently most suited to the drug in question-would be included in phase III clinical trials. Smaller and more targeted trials-so the argument goes-would result in quicker, safer, and more successful drugs on the market. These new drugs could be marketed only for those most likely to experience their full benefit. Drugs previously determined to be too unsafe for use could be reissued to a more limited population for whom they would be safer.


Most importantly, dangerous adverse drug reactions, and all the costs associated with them-to individuals, to health care providers, and to society more generally-could be better avoided. As Emilio Mordini (2004, p. 378) writes, "at a time when harmful drug reactions are thought to rank just after strokes as a leading cause of death in the U.S., the potential benefits of tailoring drugs to a patient's genetic makeup should not be underestimated even from an ethical point of view."


Yet, as with many new technologies, pharmacogenomics is a field plagued by ethical concerns. These pertain first to the need to protect informed consent and confidentiality. Additionally, questions of justice and equity of access bedevil the field of pharmacogenomics, especially in terms of the distribution of its burdens and benefits nationally and globally.

INFORMED CONSENT IN PHARMACOGENOMIC RESEARCH

Among the more troubling of ethical concerns vis-à-vis pharmacogenomics are those having to do with the practice of informed consent1. The move toward drugs that are tailored to individual genomes will necessitate that clinical drug trials be redesigned to include-perhaps extensive-genetic information about the participants. Moreover, the data collected as a part of these trials could be stored and utilized for future research into genetically-based disease. In other words, DNA likely will be collected and stored in large quantities, and that DNA will include not just the known genetic variables that are directly related to a particular research study, but also information about the individual subject's entire genome. Hence, these DNA samples could provide a host of other information about the subject, information that could be entirely unrelated to the original study but yet might prove useful for other genetic research.


When DNA research samples are collected this way, who owns them? It is very unlikely that subjects who participate in this kind of research will have considered the uses to which their DNA might be put; even the researchers may not have considered such uses. Thus, the process of obtaining informed consent may be clouded by the unknown quality of the (future) uses for this DNA-uses beyond the immediate drug research yet still possibly related to the "disease" gene in question. One possible route is to "reconsent"-i.e., seek to obtain the renewed informed consent of-the subjects for every subsequent use of their DNA. This approach, however, is cumbersome and in many cases impossible. Some IRBs have required "tiered consent," wherein subjects proactively specify allowable uses of donated specimens-for example, consent to future research involving specific disease categories. Of course, the more detailed and specific variations of consent that exist, the more difficult it is for researchers to track and manage informed consent directives.


These issues are particularly sensitive in drug trials, where the research subject and the patient may in fact be one and the same person, and thus where the motivation may be high for a sick patient to participate. Yet if pharmacogenomics is to proceed in a way that respects the autonomy of human research subjects and patients, some meaningful guarantee of informed consent, for all participants, is essential.

CONFIDENTIALITY

In a related vein, the large-scale collection and storage of vast amounts of genetic information raises obvious issues regarding subject or patient confidentiality. Again, who owns this information? Who should have access to it? Obviously, in the case where a patient's genetic information has been collected as a part of a clinical diagnosis, that patient him- or herself may have a strong interest in keeping the information confidential. Yet others also may believe that they have a claim over such information. For instance, family members who are genetically related to the individual may have a need to know of certain genetic data-propensities toward various inherited diseases, for example-that may exist in the person's genetic profile. Third-party health insurers also may argue that they have some right to the information, if they are to be able accurately to assess the risk they take on by insuring the individual. Finally, employers may desire access to information if they are to bear the cost of hiring and providing health insurance for that employee. Indeed, the fear of losing jobs and/or health insurance is a force behind many people's unwillingness to undergo certain forms of genetic testing today. According to the Council for Responsible Genetics (2004), incomplete and inconsistent laws that seek to guard against these dangers "hardly scratch the surface of the problem."

In the case of research subjects who are not clinical patients, the confidentiality issues look slightly different. Here there arise questions related to the anonymity of collected samples and how such samples are stored. Although space prohibits doing so here, it is worthwhile considering more deeply the growing body of literature on biobanks, repositories of large quantities of human biological specimens that may be stored, analyzed, and distributed.2 Biobanks enable large-scale analysis of various diseases and health phenomena, but they also represent a link between abstract genomic data and concrete patient medical records (Rothstein and Knoppers 2005). The degree of anonymity of such stored samples therefore raises a host of confidentiality issues that space prohibits my addressing here.

JUSTICE AND EQUITY IN PHARMACOGENOMICS

Perhaps the most disturbing ethical questions vis-à-vis pharmacogenomics fall into the category of justice-particularly distributive justice, which demands that the burdens and benefits of these new technologies be shared in an equitable fashion. The idea that someone or some group would enjoy significantly less access to medical treatment simply by virtue of race or economic status, or even by virtue of their "draw" in the genetic lottery, violates a deep-seated sense of fairness.

Of course, many would argue that this sort of injustice is already deeply entrenched in the American health care system. In the United States in 2006, 47 million people-almost 16 percent of the total population-were uninsured for basic health care. This includes 11.7 percent of all children in the United States (DeNavas-Walt, Proctor, and Smith 2007). Lack of basic health insurance is dramatically more common if one is poor, or if one is African-American or Hispanic (Center on Budget and Policy Priorities 2006). These inequalities in access to health care often coexist with deep inequalities in health status; by most indicators, certain racial/ethnic minorities-notably African-Americans and Native Americans/Native Alaskans-and the poor are in worse health than their Caucasian or non-poor counterparts (Munson 2004, pp. 216-24). These facts serve as reminders that market forces do not guarantee justice in the distribution of health care. When it comes to pharmacogenomics, one must ask not only whether researchers can achieve adequate levels of informed consent or patient confidentiality but also whether pharmacogenomics will rectify or exacerbate the profoundly disturbing inequalities that already exist today in the U.S. health care system. To this end, it is especially germane to examine issues related to economic status and race. Intertwined with these issues, however-and more specific to pharmacogenomics-is consideration of differences that are based directly in genotypes themselves. Beyond that, one also must address the international context, asking whether and how justice can be well-served between developed and underdeveloped countries.

Genotype and Access to the Fruits of Pharmacogenomics

One point of general agreement within the pharmaceutical world is that the cost of genetically-tailored drugs likely will be high. Increased research must be done, both in order to identify particular genetic profiles that might call for "tailor-made" drugs, and to develop those drugs and the genetic tests that would accompany them. Indeed, many wonder whether pharmaceutical companies themselves will be motivated to pursue these developments at all; as Dorothy C. Wertz (2003, p. 194) writes, "In view of the costs of drug development and the limited market, it is questionable whether major pharmaceutical companies are interested in tailor-making drugs." The more limited the market, of course, the less appealing it will be for drug companies to participate in personalizing medicine. Perfectly personalized medicine would, by definition, not be worthwhile to drug manufacturers; there is little profit to be made in developing a drug for a market of one. Hence, some sort of balance must be struck between the good of individualized medicine and the cost concerns of pharmaceutical companies as well as the economics of the health care system more generally.

A particularly sticky problem arises with respect to which diseases (or genotypes) merit attention from the pharmaceutical industry. As the Human Genome Project continues to bear fruit, more "disease" genes may be identified as possibilities for pharmacogenomic targeting. Inevitably, those most attractive to the pharmaceutical industry will be those with the largest market potential-relatively common genetic variants within the overall population. Incentives thus may be needed in order to promote research and drug development targeted at less common diseases or genotypes-that is, for those possessing so-called "orphan" genotypes. An "orphan" population may have a genotype leading to a condition for which there is currently no effective therapy; or, such a population may have been defined-for market purposes-by the pharmaceutical industry as too small to be attractive as a drug market (Smart, Martin, and Parker 2004, p. 328). The design of clinical drug trials, as well as pharmaceutical company decisions about whether to pursue drug development for "orphan" populations, potentially could create whole new categories of people underserved by personalized medicine because they possess relatively less common genotypes. Indeed, existing orphan drug legislation is targeted at just this sort of difficulty, and the development of pharmacogenomics may increase the problem and require drawing upon or possibly expanding such legislation.

Further, on a more personal level, and in a private health care system such as that in the United States, real troubles will exist for those who end up genetically categorized as "difficult to treat" or as "nonresponders" to a given drug therapy. These individuals may then carry a stigma with them that could affect their ability to obtain adequate access to health care-similar to the "preexisting condition" problem. Not only could this elicit personal economic crises, but it also could entail psychological troubles for people-or communities of people-as they begin to think of themselves as untreatable or as a "drain" on the health care system. At a minimum, this danger makes more urgent the need for orphan drug legislation or other measures to try to minimize the category of "nonresponders" who might face this level of insecurity vis-à-vis their health care.

Pharmacogenomics and Economic Status

Distinct from, though related to, the issue of orphan genotypes-and associated genetic stigma-is that of economic access to the fruits of pharmacogenomics. From the standpoint of justice, one of the most disturbing possibilities raised by pharmacogenomics is that it will further entrench the already deep socioeconomic divisions that characterize modern U.S. society. It is widely anticipated that pharmacogenomic drugs will be more expensive; and the resultant premium pricing of drugs may mean that those who are economically worse-off also will have less access to these new therapies (Smart, Martin, and Parker 2004, pp. 333-34). Indeed, will Medicaid-or even private insurance companies-be reluctant to pay for the drug therapies, or the tests needed to prescribe them? If so, the benefits of pharmacogenomics will be available only to those able to pay for them out-of-pocket. Even in societies that have comprehensive national health insurance, the question remains as to how expensive drug therapies will fit into overall health care budgets. One way or another, limits will need to be set; and justice demands that those limits do not shortchange those who already struggle at the economic and health margins of society.

Pharmacogenomics and Race

One of the most controversial of the issues raised by pharmacogenomics is whether and how to integrate the category of race into drug development and marketing. The background for such a question is an insidious history of race-based differences in health status in the United States. This is most pronounced between the white majority and self-defined African-Americans; as Laurie Nsiah-Jefferson (2003, p. 271) writes, "At no time in the history of the United States has the health status of communities of color equaled or even approximated that of white Americans." Moreover, according to at least one study, even when insurance status, income, age, and severity of conditions are comparable, racial/ethnic minorities tend to receive a lower quality of health care than whites (Nsiah-Jefferson 2003, p. 276).

Such differences may be due in part to an historical research bias in favor of white males, leaving many drugs under-tested on women and minorities. Furthermore, when studies have focused on minorities, they sometimes have ended in debacle, as was the case with the infamous Tuskegee Syphilis Study, conducted by the U.S. Department of Public Health from 1932 to 1970 (and condemned on moral grounds in 1973). The unfortunate legacy, of course, is an understandable mistrust and reluctance on the part of many members of racial/ethnic minorities to participate in clinical drug trials at all. The danger thus exists that these populations will be left under-studied and underserved by the drugs that result from such research.

Yet one must not be too quick to commit the mistake of equating race with genotype, for they are by no means the same thing. Race is an imprecise proxy for genotype; in fact, there is enormous genetic diversity among any given racial/ethnic group. Moreover, to wrongly label certain "races"-a rough social, and not a biological, category-as more prone to certain diseases runs the risk of wrongly stigmatizing members of those racial/ethnic populations. As Pilar Ossorio and Troy Duster (2005, p. 116) argue, "while attempting to provide medical benefit, or market products, scientists and the pharmaceutical industry may reinvigorate the very notions of biological difference that have resulted in racially disparate treatment and racially disparate health." Hence, although it may be easy and financially lucrative for pharmaceutical companies to market drugs to specific racial/ethnic communities, it is very often neither sound scientific policy nor wise ethical practice to do so.

Yet recent history reveals that the pharmaceutical industry is indeed prone to taking the path of race-based medicine. The June 2005 FDA approval of BiDil, a heart failure drug, represented the first such approval for use by a self-identified racial group-African-Americans. Although there is no solid evidence that BiDil would be ineffective for the general population, BiDil's developers had much to be gained by marketing the drug to a specific racial population. Excellent analysis has been done already on BiDil (Kahn 2006; Sankar and Kahn 2005; Brody and Hunt 2006); here I wish only to raise it up as an example of how race can easily eclipse genotype as the most convenient-even when less effective-way to "personalize" medicine.

INTERNATIONAL CONTEXT

This essay focuses its discussion on the field of pharmacogenomics as it has developed in the wealthier parts of the world, and primarily the United States. However it is worth briefly considering the impact pharmacogenomics could have on the international setting, and in particular how it may affect the existent inequalities between the developed and the developing worlds. Because of the relatively high costs that are anticipated with the growth of pharmacogenomics, it is likely that its fruits will flow primarily into countries where a significant portion of the population can afford them.

An obvious potential problem is that "orphan" patient populations could wind up disproportionately concentrated in underdeveloped countries. Even if a genotypic variation is relatively common in an underdeveloped country, the market potential for a drug addressing that variation is unlikely to be very strong. Hence, pharmaceutical companies will have relatively little incentive to develop drugs for these populations, unless they have sizable-and wealthy enough-counterparts in the rest of the world. Even then, the drugs developed may be so expensive as to be effectively unavailable to the world's poor. In the face of this kind of disparity, it is not an unlikely scenario that genetically undifferentiated medications, even those with known safety issues, which could theoretically be solved via pharmacogenomics, could be cast off from wealthier countries for eventual use by the populations of poorer countries (Smart, Martin, and Parker 2004, p. 334).

Ironically, pharmacogenomics may in fact have great potential for success in developing countries. The World Health Organization (2002, p. 89) points out that individual variation among persons in drug response could have substantial implications for the control of common communicable diseases in the developing world. If these countries could harness the information made possible by pharmacogenomics and transmit it into meaningful services and policy change, a powerful tool might be gained to further the cause of public health in these settings. Market-driven drug development alone, however, will be unlikely to do the job; public policy must be altered to encourage focused attention on countries that, on the surface, do not provide lucrative markets for personalized medicine.

NORMATIVE FRAMEWORK

In approaching the various ethical issues raised by the advent of pharmacogenomics, it is perhaps most helpful to start with the very basic tenet of respect for persons. Throughout the history of Western bioethical thought, this tenet has taken on a distinctly Kantian tenor, with a strong emphasis on respect for autonomy as the key component of human dignity. Respect for autonomy underlies the requirement for informed consent, the protection of patient/subject confidentiality, and the need to promote the safety of human subjects. It has been the hallmark of Western bioethics, particular during the past 30 or 40 years, since the rise of patients' rights along with a broader social emphasis on individual civil liberties.

In recent years, however, voices from various corners have challenged the near-dominance of autonomy as a concept in bioethics. Among these voices are those who question whether respect for autonomy, at least as it historically has been conceived, is overly-individualistic in its depiction of the human person. In order to take full account of the human person in all her concreteness, the concept of respect for persons should be expanded to include not only autonomy but also human historicity and relationality. Margaret Farley (1993, p. 182), for instance, has written from a feminist perspective that moral theory "needs both autonomy and relationality. … It can show that autonomy is ultimately for the sake of relationship; … [and] that relationships without respect for individuality and autonomy are destructive of persons."

Concretely, one can see the limitations of an individualistic and overly-abstract emphasis on respecting autonomy when one examines concepts such as informed consent and confidentiality in the light of genomic developments. Informed consent-a concept that for decades has stood for the protection of individual choice regarding medical treatments or research-begins to break apart with the advent of certain genetic technologies. For instance, the gathering of genetic data via large amounts of genetic screening and testing will mean that such data potentially could be used for any number of research purposes, many unrelated to the original purposes for which the data were gathered. What can it possibly mean to "inform" a patient or research subject of the risks and benefits of participation when scientists and clinicians themselves may not be aware of all the potential uses of the genetic information gathered? The patient here no longer should be conceived as an individual existing in an isolated moment in time, but rather as a concrete historical person, with interests that will stretch forward in time and be intertwined with the interests of others-who may themselves benefit from the information embedded in that patient's DNA.

Similarly, with confidentiality, genomic developments seem to push the envelope. Without protections of confidentiality in place, many patients likely will feel reluctant to submit their genetic material for testing in the first place, for fear that the dissemination of the information could lead to the loss of a job, the loss of health insurance, or negative social stigma. Yet the information gathered from such tests could have enormous value to family members or others who share a genetic heritage with the patient. To view the patient only as an isolated individual who "owns" her genetic information would seem to deny the reality that she is an historical individual with concrete relationships, relationships that partially constitute who she is.

In fact, although persons' ability to determine their own course of action continues to deserve respect and protection, so should their embeddedness in concrete historical circumstances and relationships. Ultimately, of course, this embeddedness may place limits upon autonomous choice; the trick is to accept such limitations in a way that preserves as much as possible the essential dignity and well-being of the human person. To be clear: the argument here is not that autonomy itself is an outdated concept. Rather, autonomy must be understood as one-very important-quality of a person whose dignity also is marked by his relationships and concrete, historical situatedness.

In the face of this scaled-back understanding of autonomy, beneficence, or the promotion of persons' well-being, rises in prominence. In recent years, a surge of interest both in virtue ethics and in neo-Aristotelian versions of human flourishing has fueled a renewed emphasis on beneficence and what exactly it means to foster human well-being. Finite social resources place limits around what society can and cannot achieve with respect to promoting human health; but taking human well-being seriously does require consideration of the potential of pharmacogenomics for its advancement. The promise of personalized medicine, although perhaps still further off in time than one would like, is indeed substantial. Moreover, the gains in safety and drug efficacy-for subject populations in drug trials as well as for the general population-could be enormous.

Promoting human well-being vis-à-vis pharmacogenomics, however, requires looking further than the technical promise of pharmacogenomics. One also must consider the emotional and psychological aspects of human well-being. What will it mean, for instance, for individuals to be tagged as "nonresponders" or "poor responders" to ordinary drug therapies, in a society where only some people are guaranteed adequate health care? What will it mean for entire communities to be so labeled? Further, how will children who are tagged with this label come to understand themselves, as they grow? Policymakers must take seriously the ways in which persons and groups might subtly be affected by the widespread use of genetic profiling, particularly in a society where health care is a limited and unguaranteed resource.

Taking beneficence seriously means taking utility seriously. Thus policymakers must squarely face the tradeoffs that are inevitable in the course of promoting well-being in a finite world. In the case of pharmacogenomics, policymakers will need to make concrete choices about where society's resources will be spent: Which diseases will physicians screen for? Which drugs should researchers develop? How much of the overall health care budget should be devoted to pharmacogenomic research?

These questions, of course, feed directly into concerns about justice. Justice demands that society develop and distribute out the benefits of personalized medicine so that they are available to people of all racial and socioeconomic backgrounds. In particular, policymakers must keep in view the dramatic injustices (disparities) that exist in health care today. If personalized medicine develops in such a way that its beneficiaries are primarily or exclusively wealthy Caucasian males, justice has not been well-served.

Related to justice, of course, is the concept of the common good, an idea with a rich history in the Catholic tradition. In many ways, promoting the common good stands as a corrective to the overemphasis on individual choice previously described. The common good "rests on a vision of society in which all people join in the pursuit of shared values and aims" (McLean 2006). Here, the common good is not simply the sum of individual goods, but rather is a good worth pursuing in its own right. Although it is tempting to think of the common good as standing in tension with individual well-being, in fact, the common good is best seen as inextricably bound up with individual good, such that the individual can truly flourish only in the context of a healthy larger community. In the context of health care, a focus on the common good demands that policymakers lift their eyes beyond the horizon of individual choice to consider the broader well-being of the larger community and its many members and groups. A basic level of health care for all, including certain guarantees that no one will live in fear of being denied access to such basic health care, is in the common interest and should be considered part of the common good.

Bound up with the common good is the notion of solidarity. Pope John Paul II (1987) described solidarity as "a firm and persevering determination to commit oneself to the common good … to the good of all and of each individual, because we are all really responsible for all." According to this view, individual human well-being actually depends in part upon moving beyond oneself in order to assist others in realizing their full potential within the human community. Therefore, attention to the common good requires more than a vague sense of responsibility for the good of society; rather, it requires sustained and specific commitment to stand with others in the pursuit of a larger, communal vision that will benefit all.

The Catholic tradition has often held that justice and solidarity sometimes will require that a disproportionate amount of effort be put into "favoring" those who are socially or economically marginalized by the current system-that is, those who are victims of injustice. This is often referred to as a "preferential option for the poor;" that is, special care is given to the poor "because their needs are greater" (Locatelli 2005, p. 10). Hence, in a sense, one must start with the injustice that exists in order to achieve a more complete version of justice (Lebacqz 1987). In the context of pharmacogenomics, this means that one must pay particular and sustained attention to correcting the health disparities and inequalities in access to health care that exist in the United States, and indeed around the globe, today. Insofar as the strength of pharmacogenomics can be harnessed to that end, it can be a powerful tool.

RECOMMENDATIONS FOR PUBLIC POLICY

With these considerations in mind, I turn now to the concrete realm of public policy. Pharmacogenomics represents an exciting new phase in the application of human genetics, one that should be approached with measured enthusiasm and cautious encouragement. Policy choices cannot eliminate entirely the ethical hurdles I have described; but they certainly can help guide the field's development and implementation in more ethically suitable directions. My aim here is to spark individual and corporate ethical imaginations as to how, concretely, the moral considerations described above might be realized in the public sphere.3

1. Informed Consent for the Use of Genetic Samples

Pharmacogenomics raises new issues related to informed consent because tissues or blood collected today for one test or trial also might be useful in some currently unthought-of way as scientists learn more about the human genome. For example, the soldier whose autopsy sample recently provided the clue to the DNA profile of the 1918 influenza virus never could have given his consent for such testing, as it was unknown to doctors at the time of the pandemic.

How might genuine informed consent look, if one were to take account of persons as both autonomous and relational? One possible approach to this problem is to allow for more general forms of advanced consent in cases where the research is relatively noncontroversial and entails no or only minor risk for the donating individual, and where the potential gain to society is great. Unfortunately, this approach may run into problems under the HIPAA Privacy Rule, which took effect in 2003. This rule among other things prohibits authorizations of an identifiable individual's private health information for unspecified or nonspecific future research, thereby raising cautions or even roadblocks for research involving nonanonymous stored tissue samples, even as it (rightly) seeks to protect the heart of informed consent.

Nevertheless, to encourage consenting subjects to participate-via the "donation" of genetic material-in future research takes seriously their participation in a society where persons depend on each other for the development of social goods such as more sophisticated medicines or disease research. For any more controversial uses of genetic samples, however-such as screening for the propensity toward a particular serious disease-specific consent should indeed be sought, thus hopefully safeguarding and respecting the dignity of the individual person and preserving the heart of the requirement of informed consent. Institutional review boards could be used to determine when more general consent is acceptable and when more specific forms should be required.

2. Improvements to Patient/Subject Confidentiality

Related to questions about informed consent are those regarding patient confidentiality. Given the high degree of sensitivity that marks much genetic information that might be collected in an age of pharmacogenomics, how should one think about confidentiality? If pharmacogenomic testing showed that a person had an increased risk of a serious illness such as cancer, should that information be available to insurers? What if scientists someday uncover genetic propensities toward antisocial behavior? Who should have access to such information? A breach of confidentiality in the case of genetic information could translate into a potentially devastating loss of privacy for the individual.4

On one hand, society must do all it can to ensure that patient confidentiality is protected, a concern that touches the heart of respect for personal dignity, at least as it is understood in Western bioethics today. Careful design of genetic tests-where possible, choosing biomarkers, for instance, that carry minimal secondary information-and the use of highly sophisticated coding techniques could help to preserve, to some degree, patient confidentiality (Buchanan et al. 2002, p. 10). Further, federal and state medical privacy legislation-including HIPAA-should continue to be updated and policed to ensure that records are indeed kept anonymous or accessible only to the patient and his or her medical caregivers. Finally, professional education of health care providers should emphasize the ways that pharmacogenomics raises the stakes on patient confidentiality-that is, the ways in which increased levels of genetic testing could uncover genetic information that is likely to be considered by patients as highly personal and sensitive.
Yet these are only band-aid solutions. As legal bioethicist Mark A. Rothstein points out, one fundamental issue has to do with whether there ever exist social interests that are strong enough to warrant universal access to genetic information. If so, he argues, then a system for genetic risk-sharing should be devised so that particularly at-risk individuals do not face unfair discrimination vis-à-vis their health care. This could be accomplished, for instance, by mandatory group-based health insurance with built-in risk-spreading mechanisms (Rothstein 2005, p. 31). It also could be accomplished by a national health system that guaranteed a basic level of health care for all, regardless of individual risk. At a minimum, legislation must be designed to eliminate employer- or insurance-based discrimination that might be rooted in the disclosure of individual genetic information. Such health care reform legislation will not erase the issue entirely, of course, for there exist other reasons that some individuals may want to keep their genetic profiles private-fear of social stigma, for instance. Yet legislative reform that reduces the fear an individual may feel regarding his or her access to health care certainly will eliminate a large part of the problem.

Although patient confidentiality is indeed important, from one perspective, there are also arguments to be made that close biological family members may have a strong reason to demand a breach of confidentiality in some cases, where their own health and well-being is strongly implicated. Indeed, an understanding of human dignity that recognizes the relational nature of persons cannot avoid this interconnection. Although it may not be wise-or feasible-to allow for breaches of confidentiality as a matter of public policy, it would be prudent in certain cases for health care providers to encourage-strongly-that patients inform their biological relatives of sensitive genetic information that surfaces as a part of their medical care. Professional guidelines could be issued to this effect, with particular emphasis on the growing need for this as pharmacogenomics increases the frequency of genetic testing.

3. Increased Post-Marketing Surveillance

One of the promises of pharmacogenomics is that clinical trials could be made smaller and more targeted, thus ideally increasing the safety of the research subjects and bringing drugs to market more quickly. Yet conducting smaller and more targeted trials also means that drug companies will have less available data about potentially adverse drug reactions-including those provoked by drug-drug interaction-until the drug is marketed more widely. Hence, post-marketing surveillance of drug response-i.e., surveillance beyond phase III clinical trials-will be essential to ensure patient safety as far as possible-a requirement demanded by the ethical principle of beneficence. Britain's Royal Society (2005, p. 42), in its report on personalized medicine, argued that there is a mandate for regular post-market monitoring that links genetic variability with clinical outcomes. In the American setting, Rothstein (2003, p. 325) suggests that new legislation is long overdue that would grant the FDA subpoena power and the authority to impose civil and monetary penalties for failure to conduct such post-marketing studies. Certainly some such measures seem called for if society is to take seriously the need to protect long-term human well-being.

Here it is perhaps worthwhile to add that any such scientific surveillance of clinical outcomes must thoroughly probe the effects of environment on drug response. Rather than resorting to an overly-simplistic genetic determinism, it is incumbent upon scientists and clinicians to dig deeper for the various environmental factors that might contribute to a particular patient's response to any given medication. Moreover, the scientific community in general must be encouraged to conduct research into these factors rather than making assumptions about underlying genetic explanations for clinical outcomes.

4. Increased Incentives for the Development of Orphan Drugs

Justice and a commitment to the common good all demand that society addresses the problem of orphan diseases. Society's resources must be spent not only on relatively common diseases; they must also be directed toward less common, and yet sometimes more debilitating, diseases, very often diseases that can be attributed in large part to nothing more than the genetic lottery. Obviously, finite public resources will force difficult choices in this regard, and there is no formula for how much money should be spent researching diverse diseases or drug responses. But to let pure market principles or pharmaceutical interests decide the matter is insufficient and could result in market indifference toward entire categories of diseases and persons.

The goal of legislation to address this problem would be to induce the development of drugs for less common diseases or genotypes. In this way, the category of "nonresponders" to drugs could be minimized, which is both a good idea for its own sake as well as for reducing the potential psychological stress on those with less common genotypes. Such measures could include research funding/grants through public bodies such as the National Institutes of Health; federal inducements to pharmaceutical or biotech companies in the form of tax incentives, direct grants, or loans; and extension of the period of market exclusivity for drugs designed for "orphan" genotypes. In addition, drug pricing potentially could be controlled to make it more attractive for pharmaceutical companies to invest their resources in such drugs. These measures may draw on or expand existing orphan drug legislation to accommodate the increased need brought about by pharmacogenomics.

5. Revision of Patent Law to Encourage the "Rescue" of Drugs

Along similar lines, pharmacogenomic targeting of drugs could allow the "rescue" of some drugs that had been shelved because of safety risks discovered during clinical trials-or later, for that matter. In this way, drugs that were classified and discarded as unsafe or ineffective after clinical trials could be resurrected and utilized for certain patients-patients for whom they are in fact relatively safe and effective. Indeed, a commitment to beneficence and human well-being surely would support such a move. Yet it is difficult to see the motivation private companies would have for the rescue of such drugs if the original patent on the drug is near expiration. Hence, it may be necessary to revise patent law so that the revitalization of such "rescued" drugs becomes economically viable. Along these lines, the FDA perhaps could award additional time for market exclusivity to pharmaceutical companies who engage in such undertakings (Rothstein 2003, p. 329).

6. Subsidies to Ensure that the Less Wealthy Have Fair Access

In an era of price cutting, it is difficult to see how the increased cost of genetically tailored medicine will be managed by an ailing health care system. Is it acceptable that those with fewer economic resources enjoy significantly worse access to pharmacogenomic medicine, especially if pharmacogenomic drugs become widely utilized as standard treatment? Widely-held societal norms requiring a social safety net-norms that support programs such as welfare and Medicaid-would seem to require that personalized medicine be enjoyed not only by the rich.

Hence, society needs some assurance that Medicaid will cover pharmacogenomic drugs-at least those that are determined to be necessary for supporting some basic level of human health. Insofar as pharmacogenomics pushes society closer to a two-tiered system of health care, society must seek to compensate for that trend by finding ways to spread its benefits to rich and poor alike. Again, in the long term this may mean some version of comprehensive health care reform that streamlines administrative costs and spreads burdens more broadly. In the short term, however, society may need to consider subsidies to allow the poor greater access to personalized medicine.

7. Approval of Gene-Specific Drugs Over Race-Specific Drugs

Given the broad genetic diversity within racial groupings, as well as the insidious history of racially differentiated health care in the United States, it is important to discourage the (mis)equation of race with genotype. This is both for the sake of accuracy-and thus human safety and well-being-and in order to minimize the danger that racial stereotypes, and ultimately racial injustice, will be aggravated by the development and marketing of new pharmacogenomic drugs. The FDA should be extremely wary of applications that target particular races of patients, especially in the absence of thorough and conclusive evidence that there is a significant incremental difference in safety or efficacy for patients of different racial or ethnic backgrounds. Genotyping evidence should be required for new drugs that claim to be more effective for a particular racial/ethnic population. Finally, as Andrew Smart and his colleagues (2004, p. 342) suggest, professional education could be better targeted to encourage prescribing practice to be based more on scientific evidence and less on prejudice.

8. Inclusion of Racial/Ethnic Minority Groups in Drug Research

Yet, simultaneous to resisting of the use of "race" as a category, scientists also must design research studies to include adequate numbers of subjects from racial/ethnic minority populations. This is as true of pharmacogenomic drug design as it is of other forms of research. Moreover, in order to resist oversimplification of the race-genotype connection, subjects from racial/ethnic minorities should be recruited for all appropriate categories of genotypes; they should not simply be associated with a few specific genotypes that may be more common within their populations. Public funding, combined with private sector incentives, can begin to address the research gaps that otherwise would exist with respect to racial/ethnic minority representation in all phases of pharmacogenomic research. Such measures would advance the larger common-good goals of a fair distribution of health care resources and a more accurate and well-informed allocation of society's health care dollars.

In order to help ensure that the interests of racial/ethnic minority groups are protected and promoted, minority community groups could be specifically called upon to review pharmacogenomic research and drug development. Bioethicists Charles Weijer and Paul B. Miller (2004), for instance, suggest that DNA databanks should be subject to consultation by community groups representative of racial/ethnic minorities-or other historically disadvantaged communities-who contribute samples. In most cases, this consultation would not be restrictive on individuals; but rather it would provide the occasion for constructive feedback and, further, could facilitate increased respect for the values and interests of organized communities especially affected by such research. Likewise, racial and ethnic minorities should be well-represented on institutional review boards, so that such boards do not inadvertently overlook or downplay the interests of such minority communities in the ongoing development of pharmacogenomic drugs.

9. Incentives for Pharmaceutical Companies to Invest in and Provide Drugs to Developing Countries

Distributive justice and attention to the common good require that personalized medicine not simply be a benefit enjoyed only by the world's wealthier countries. This is particularly true since a good portion of the research for clinical drug trials-between 10 and 20 percent, according to estimates-is conducted in less developed countries (Barton 2006).

An increasingly globalized world must be especially careful not to allow market forces to be the sole determining force in this regard. Wherever possible, international bodies such as the World Health Organization should develop policies to enable and encourage technology transfer to the developing world, especially where pharmacogenomics could provide a clear benefit to public health-for instance, in the realm of infectious disease. Where markets do not appear to be lucrative, governments as well as nongovernmental organizations might conduct or subsidize research directed at diseases that particularly plague developing countries. Moreover, wealthier countries such as the Unites States could provide incentives for private companies to conduct and implement research directed at these diseases.

Furthermore, international regulatory bodies must be made more inclusive to reflect more diverse and equitable representation, in order to facilitate justice on the international scene. The International Commission on Harmonization (ICH), for example, was initiated in 1991 to facilitate global product approval and includes representatives from the United States, the European Union, and Japan, plus observers from the World Health Organization and Canada. This body should be altered/expanded to include formal participation by less developed countries as well, to help ensure that their own interests are fairly considered. This is particularly important as the practice of conducting clinical trials in less developed countries becomes more common; for justice demands that those participants who bear a substantial part of the burden of pharmacogenomic research also have access to its benefits. International regulation should establish solid procedural safeguards, such as setting up qualified local IRBs in developing countries, in order to help secure the well-being of pharmacogenomic research participants. Finally, such bodies could establish and promote certain baseline standards for international research-such as requiring an up-to-date standard of care for all research participants-that must be observed in all settings. This would go some distance toward protecting the well-being of research participants worldwide, including those who are relatively less wealthy or powerful.

CONCLUSION

There are, to be sure, a host of hurdles, both technical and ethical, standing in the way of widespread clinical implementation of pharmacogenomic progress. Yet in spite of these many hurdles, one should not lose sight of the worthiness of the overall goal: to capitalize on improvements in understanding the human genome in order to develop safer and more effective responses to human disease. If this can be done, and done effectively, respectfully, and fairly, it would represent a major step forward in the progress of medical science.

REFERENCES

Barton, John. 2006. Presentation made at Santa Clara University, 17 March.

Brody, Howard, and Hunt, Linda M. 2006. BiDil: Assessing a Race-Based Pharmaceutical. Annals of Family Medicine 4: 556-60.

Buchanan, Allen; Califano, Andrea; Kahn, Jeffrey; et. al. 2002. Pharmacogenetics: Ethical Issues and Policy Options. Kennedy Institute of Ethics Journal 12: 1-15.

Center on Budget and Policy Priorities. 2006. The Number of Uninsured Americans Is at an All-time High. 29 August. Available at http://www.cbpp.org/8-29-06health.htm, accessed 10 March 2008.

Clayton, Ellen Wright. 2005. Informed Consent and Biobanks. Journal of Law, Medicine & Ethics 33: 15-21.

Council for Responsible Genetics. 2004. Genetic Testing, Discrimination, and Privacy. Available at http://www.gene-watch.org/programs/privacy.html, accessed 10 March 2008.

DeNavas-Walt, Carmen; Proctor, Bernadette D.; and Smith, Jessica. 2007. Income, Poverty, and Health Insurance Coverage in the United States: 2006. U.S. Census Bureau. Available at http://www.census.gov/prod/2007pubs/p60-233.pdf, accessed 12 March 2008.

Farley, Margaret. 1993. Feminism and Universal Morality. In Prospects for a Common Morality, ed. Gene Outka and John P. Reeder, Jr., pp. 170-90. Princeton, NJ: Princeton University Press.

John Paul II. 1987. Sollicitudo Rei Socialis. Available at http://www.vatican.va/holy_father/john_paul_ii/encyclicals/documents/hf_jp-ii_enc_30121987_sollicitudo-rei-socialis_en.html, accessed 10 March 2008.

Kahn, Jonathan. 2006. Race, Pharmacogenomics, and Marketing: Putting BiDil in Context. American Journal of Bioethics 6 (5): W1-W5.

Lebacqz, Karen. 1987. Justice in an Unjust World. Minneapolis, MN: Augsburg.

Locatelli, Paul, S.J. 2005. The Catholic University of the 21st Century: Educating for Solidarity. Unpublished paper.

McLean, Margaret R. 2006. A Framework for Thinking Ethically About Human Biotechnology. Available at http://www.scu.edu/ethics/publications/submitted/mclean/biotechframework.html, accessed 10 March 2008.

Mordini, Emilio. 2004. Ethical Considerations on Pharmacogenomics. Pharmacological Research 49: 375-79.

Munson, Ronald. 2004. Intervention and Reflection: Basic Issues in Medical Ethics. Belmont, CA: Wadsworth/Thomson.

Nsiah-Jefferson, Laurie. 2003. Pharmacogenomics: Considerations for Communities of Color. In Pharmacogenomics: Social, Ethical, and Clinical Dimensions, ed. Mark A. Rothstein, pp. 267-90. Hoboken, NJ: Wiley-Liss.

Ossorio, Pilar, and Duster, Troy. 2005. Race and Genetics: Controversies in Biomedical, Behavioral, and Forensic Sciences. American Psychologist 60: 115-28.

Rothstein, Mark A. 2003. Epilogue: Policy Prescriptions. In Pharmacogenomics: Social, Ethical, and Clinical Dimensions, ed. Mark A. Rothstein, pp. 319-35. Hoboken, NJ: Wiley-Liss.

------. 2005. Genetic Exceptionalism and Legislative Pragmatism. Hastings Center Report 35 (4): 27-32.

------, and Bartha Maria Knoppers. 2005. Introduction. Journal of Law, Medicine & Ethics 33 (1): 6.

Royal Society. 2005. Personalised Medicines: Hopes and Realities. London: The Royal Society.

Sankar, Pamela, and Kahn, Jonathan. 2005. BiDil: Race Medicine or Race Marketing? Health Affairs (web exclusive): 455-63.

Smart, Andrew; Martin, Paul; and Parker, Michael. 2004. Tailored Medicine: Whom Will It Fit? The Ethics of Patient and Disease Stratification. Bioethics 18: 322-43.

Weijer, C., and Miller, P. B. 2004. Protecting Communities in Pharmacogenetic and Pharmacogenomic Research. Pharmacogenomics Journal 4: 9-16.

Wertz, D. C. 2003. Ethical, Social and Legal Issues in Pharmacogenomics. Pharmacogenomics Journal 3: 194-96.

World Health Organization. 2002. Genomics and World Health: Report of the Advisory Committee on Health Research. Geneva: World Health Organization.

NOTES

1Although informed consent and other ethical issues arise both for research subjects and for clinical patients, the issues can in fact be somewhat distinct for the two populations-although, in practice, the populations themselves are not always entirely distinct. At various points in this article I focus on one or the other population-research subjects or patients-simply because the particular ethical issue under discussion seems clearer or more pressing in one or the other case. However, this focus should not be taken to minimize ethical issues for either group.

2See, for instance, the spring 2005 issue of The Journal of Law, Medicine & Ethics, particularly the articles by Clayton (2005) and by Rothstein and Knoppers (2005).

3For another and very helpful discussion of some of these points, see Buchanan and colleagues (2002).

4Here I focus on the confidentiality concerns that arise in clinical settings, rather than those that arise as part of research. The latter raise a rather different set of issues, some of which may be addressed by greater use of biobanks, as already mentioned.

Karen Peterson-Iyer is program specialist in health care ethics at the Markkula Center for Applied Ethics. This article appeared originally in the Kennedy Institute of Ethics Journal, 18:1 (March 2008), published by Johns Hopkins University Press, which is the copyright holder. Research for this article was supported by a generous anonymous donor.

June 2008