This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

P E R S P E C T I V E :
B i D i l
 
11 October 2005  
  

The Case Of BiDil: 
A Policy Commentary 
On Race And Genetics

The debate between biologists and sociologists
is just beginning, when it comes to genomics and race.


by Rick J. Carlson


ABSTRACT:

The Food and Drug Administration (FDA) approval of BiDil unleashed a vigorous commentary, mostly critical of the decision. The FDA was soon caught between biologists, who see research and clinical utility in using racial classifications, and social scientists (and many politicians), who fear the adversities of greater discrimination. Analyses rarely addressed the practical factors the FDA had to consider in reaching a decision. Much of the literature simply assigned the question to the domain of racial politics, failing to consider the ethics of professional care, the Hippocratic oath, and the marketplace efficiency in moving drugs quickly to those who might most benefit.

The role that race plays in genetics is an alluring, even intellectually intoxicating subject—lethal if you don’t approach it with self-awareness. The furor that arose after BiDil was approved by the Food and Drug Administration (FDA) for use in African Americans wasn’t about its clinical status but, rather, because the agency permitted racially based marketing of the drug.¹ This is a tempting target for both serious thinking and soapbox politicking. Anticipation of the BiDil decision had academic word processors at the ready.

The take-off point for my comments is an analysis of the FDA decision by Pamela Sankar and Jonathan Kahn.² Although I found their analysis to be reasonable, as far as it went, I also found it incomplete and misleading; the authors failed to locate their subject in the landscapes of population genetics and evolutionary biology. In other words, Sankar and Kahn got stuck on the “race” issues and never quite made it across to the “genetics” issues, which is like explaining purple by only talking about red and not blue.

A Foolhardy Generalist View

I have no particular expertise on this subject, save a growing strategic understanding of the many implications of genomics for health and health care delivery.³ My perspective then is a policy perspective—a generalist view. From that perspective, what can be learned from the literature and research about this very complex subject, and what about BiDil? Are the analyses of this subject of much help in resolving the controversies?

Scholars in these fields seem more exasperated than reflective; to quote Ruth Hubbard, an eminent biologist, “It is beyond comprehension…[that] educated persons…would come forward…in complete ignorance.”⁴ This might be expected in fields like race and genetics, short on facts and long on received wisdom. As to BiDil, many just invoked the catechism of distributive justice and assigned the case to the courts of racial politics without considering the decision-making challenges that confronted the agency—in effect saying, “It’s a racially sensitive topic if we say it is.”⁵ This result was probably inevitable, but throwing a blanket like racial politics on the subject blotted out the sun on many other issues, such as medical ethics, that should have gotten some attention.

Of course, the impact of racial factors would have a high priority in any comprehensive review of all of the scholarly literature about race and genetics.⁶ But those factors shouldn’t be played as trump cards. Since the FDA has already made a decision, what were factors influenced its decision, and would a fair reading of the literature have helped frame the issues so as to assist the FDA in finding a circumscribed political and moral basis for its decision? And, if not, why not?

A Volatile Subject

What should policymakers know about race and genetics to help them make decisions? This is a complex subject. Dogmatism often surfaces, but it shouldn’t suffice. A review of much of the analysis provides only patchy consensus on most of these themes. Opinion varies widely, ranging from Craig Ventner’s oft-quoted observation that “there is no basis in the genetic code for race,” to Neil Risch, a highly regarded geneticist at Stanford, arguing that “identifying genetic differences between races…is scientifically appropriate.”⁷ The subject is volatile but still typical of the policy environment: imperfect information and lots of opinions.

The central concerns raised by the social science community, including many health care services researchers, fall into two buckets. The first is the potential use of racial classifications to further reinforce stereotypes, especially those associated with disease. Second are fears that disparities in access will be perpetuated rather than alleviated by using these classifications. There also is a minority report: Some argue that it is essential to use racial categories to illuminate the adverse consequences of those very disparities.

By a crude head count, most analysts favor sharp restrictions on the use of racial classifications, or even a prohibition in some cases. Some researchers argue that the costs of using racial classifications outweigh whatever arguable clinical benefits are claimed. Others take the opposite view. These differences in view are significant, but they also reflect a more fundamental and stubborn difference between social and biological models for understanding our simultaneous biological and social lives. Those who study race as social scientists appreciate the power of genomics but deplore its perverted applications that make harsh social and economic discrimination sting even more. Those who study genetics grasp the social and economic issues social scientists raise but can’t understand why they’re attacked as racists when they create racial classifications for research findings to facilitate translation of those findings into clinical services that they believe will benefit patients.

When it considered the new drug application (NDA) for BiDil, the FDA sat right at the intersection of those two models—both of which have limitations. Social scientists have the problem that although most people believe that discrimination persists and that its outcomes are poisonous, the evidence supporting that position is mostly associational, inferential, and generally squishy. Racial discrimination, however, is a fact of American life and is commonly believed to occur. In fact, race and the reactions it elicits were all brought home again in the harsh aftermath of Hurricane Katrina in September 2005.

On the other hand, biologists have a problem justifying the use of racial classifications in terms of demonstrable benefits to patients. So far, they can’t make a very strong case for race as a useful classification in explaining much about our health or how to improve it. From a distinctly genomics point of view, what we call racial characteristics is shaped partly by genes, of course, but not in ways that appear to indisputably further medical science. The few racial distinctions in health care that are recognized (if not always observed) arose directly out of patient care experience, such as drug intolerances or metabolic reactions, or resulted from diagnoses based on Mendelian genetics, as in the case of sickle cell disease (a simple genetic disease rather than a complex one, such as most cancers). These occasional classifications were incorporated into practice because they had practical clinical utility—and they continue to do so.

The questions about BiDil arose in a regulatory rather than practice environment. The FDA’S decision that drew fire wasn’t about safety or efficacy, or even clinical utility (those decisions had already been made); it was about the business of selling drugs—specifically, the scope and content of permissible racially based marketing for promoting a drug. The question asked here is, given prevailing biological and social science thinking about race and genetics, was that a sound decision, or at least a supportable one?

Deep, core support comes from population geneticists and evolutionary biologists, looking at the subject from a mountaintop, examining human development from the historical record DNA leaves behind. This might sound detached from the policy world, but these researchers at least offer an explanation for why BiDil proved to be more effective in a racially defined group (and, for the same reasons, why there will be more demographically specific product profiles ahead). The explanation is complex, but it boils down to whether information gleaned about our health from research about continent of origin, ancestry, ethnicity, and sometimes race of our species is reliable—and, if so, whether it also has clinical utility—given the evolving state of genomics research. Most scientists believe that as these fields of research further mature, more probative classifications for health care applications other than race will evolve. In the meantime, many in these fields believe that ancestry is the more useful classification but that there are instances—perhaps this case—where race remains a valid proxy in the absence of better information. That position is likely to change over time. As Mark Rothstein suggests, “The idea…is not to eradicate or ignore [racial] differences, but to redefine or move beyond race to more precise categories of difference with justification for establishing such differences.”⁸

On balance, however, when taken together, the social and biological literature urges broad caution in using racial classifications. Most agree that racial information adds value but that the costs of its use are perhaps unacceptably high. In the future, however, as genomics matures as a science, in as little as five to ten years, risk-related data will become much more detailed and gain greater clinical utility. At that time, a different balancing of risks and opportunities will undoubtedly make sense.

The BiDil case reminds us that we remain far away from the day when the subject of race can be discussed more dispassionately. In an unadulterated scientific environment, racial variables would be weighted by the same measures applied to other variables, such as temperature. But human beings run labs, and diminishing discrimination is a paramount social policy for us all. In short, we are going to need a new vocabulary of “difference,” as genetic data continue to confound existing categories. Perhaps we could fashion the social-policy equivalent of the precautionary principle characteristically applied to the environmental issues: Use of race/ethnicity data should not be undertaken if the long-term consequences are costly and the gain is marginal relative to those costs.

Just how marginal is the gain against the costs, though? For example, if, in a case like BiDil, marketing only to blacks (this is the term the FDA uses in its published approval) is judged to be discriminatory by increasing the likelihood of discrimination against them, then marketing only to whites—should that have been the equally plausible research outcome—would also be judged as prima facie discrimination because it would presumably lead to the same result. The result, in our racially sensitized world, is that those who wish to use genetic information must prudently assume that discrimination—racial or otherwise—might occur, or is sure to be alleged, and hence the data had best not be used unless their transcending clinical value can be justified in differential diagnosis. If that case can’t be made, or if the effort to do so is too costly, then why bother to collect those data at all? Tipping the policy scales to protect against some forms of discrimination, as for example, health insurance, then, is not the slam-dunk policy decision it seems, because it comes with costs: use of ever more applicable genetic risk data to improve health care outcomes.

Finally, through a bioethics lens, another interesting line of thought emerges. What about the Hippocratic oath and especially the obligation of the healer to provide all available care to each patient, independent of economic, social, and racial considerations? There is a great efficiency in applying this still-relevant commandment; it keeps it simple. As Mike Bamshad put it in his recent and an excellent review of this field, “clinicians often [just] want to know whether it is valid and reliable to use race as a proxy to infer an individual’s genetic risk for disease and treatment response.”⁹ From what evidence there is, BiDil seems to fit that profile.

Summing Up

The central social policy questions about race and genetics arise in two ways: First, will genetic data be used to further stigmatize and expand application of social controls, and, second, will the pursuit of genetic causes for disease and poor health further marginalize public health and social programs, as well as skewing the research agenda to develop products for the more numerous and economically advantaged mainstream white population? The evidence suggests that these fears and concerns are often justified, but, of course, that doesn’t lead inexorably to barring the use of racial classifications in health care if there are countervailing considerations.

There is an even deeper, cultural divide at work here with unique implications for health care, but one that also affects other social and commercial sectors: the democratization of evidence. There is a real peril that lowbrow theories wrapped in tendentious and oily slogans will get the public’s ear and gain even footing with scientific proof as worthy of belief. The case in point is the fiery debate (theological dispute) about intelligent design and human evolution. In the case of BiDil, the phenomenon arises in two ways, both more benign but both worth examining closely. First, addressing questions about race and genetics, social science has achieved parity with the “harder” life sciences. Taking the biological perspective alone, although there are genuine differences on the meaning and utility of racial classifications in research, absent the arguments of social scientists, the FDA could have followed a simpler path and sided with the biologists’ consensus view that for now and the foreseeable future, there is sufficient support for use of racial classifications in some instances.

In fact, the FDA did side with the biologic view, but not necessarily because one model—the biological—prevailed over social science perspectives. There is yet another perspective that the FDA had to consider that might have tipped the balance. In addition to accessing available research, the decisions the agency has to make also have to be filtered through a rough-and-tumble marketplace ethic. Some in the research community find this unprincipled. Many scholars seem to think that the only debates at the FDA are about safety and efficacy. That is largely true, but the debate is also about moving product. The FDA is supposed to help move, not just approve, product. Pharmaceutical manufacturers pay the FDA very large sums to do just that. This objective might not be widely admired by scholars, but it shouldn’t be ignored because it seems unsavory. Imagine the challenge facing the BiDil sales force: It might be more right and correct to market based on generic population variables or even ancestry, but such messages aren’t remotely as compelling as using race in a culturally respectful way. This doesn’t have to be crass or offensive, either (certainly no more than ads for the erectile dysfunction drug Cialis). If the product works, there is a stout marketplace ethic based on customs, results, and focus that comes into play. The folks who write ads don’t want to be any more racist than the rest of us. So companies go to work after the FDA says so, and get the product sold to as many people as they can, hoping that most will benefit who want it. That’s just good, effective marketing. Scholars might not like the rough-and-tumble nature of this marketplace, but the FDA has to live in that world, too.

Finally, there is, for me, a telling point, also marginalized in the literature: the impact of the patient-doctor bond in meeting patients’ needs, independent of social and economic considerations. There is a virtue to the staunch, uncomplicated clarity of that ethic. In times when nearly every professional reward, acknowledgement, and attribute has been devalued in practice, and even though its applicability might be considered quaint, it directly applies in this situation because it focuses our attention on the definable needs of specific and real patients who could benefit from a product that’s available, not on tortured social science posturing. In this case, in my view, this simple ethical commandment trumps an earnest but muddled resistance rooted in legitimate and palpable fears of stigmatization and discrimination.

What should the FDA have done, then, in my generalist opinion? Just what it did. And, although unlikely, my fantasy is that the agency used a variation on Occam’s razor: If something simple explains something, don’t keep looking.

This work was supported in part by Project nos. U35MC02601 and U35MC02602 from the Maternal and Child Health Bureau (Title V, Social Security Act), no. 11223, HRSA DHHS. The views expressed are those of the author alone, although colleagues at the University of Washington assisted with comments.

NOTES

1. U.S. Food and Drug Administration, “FDA Approves BiDil Heart Failure Drug for Black Patients,” Press Release, 23 June 2005, www.fda.gov/bbs/topics/NEWS/2005/NEW01190.html (4 October 2005).
2. P. Sankar and J. Kahn, “BiDil: Race Medicine or Race Marketing?” Health Affairs, 11 October 2005, content.healthaffairs.org/cgi/content/abstract/hlthaff.w5.455.
3. See P.R. Billings et al., “Ready for Genomic Medicine? Perspectives of Health Care Decision Makers,” Archives of Internal Medicine 165, no. 16 (2005): 1917–1919.
4. See R. Hubbard, “Race and Genes,” presented in an online forum, “Is Race ‘Real’?” 20 April 2005, http://raceandgenomics.ssrc.org/Hubbard (15 September 2005).
5. Typical of this literature is a series of commentaries on race and genetics offered as an online forum sponsored by the Social Sciences Research Council, “Is Race ‘Real’?” http://raceandgenomics .ssrc.org; and S.J. Lee, J. Mountain, and B.A. Koenig, “The Meanings of ‘Race’ in the New Genomics: Implications for Health Disparities Research,” Yale Journal of Health Policy, Law, and Ethics (Spring 2001): 53–69.
6. Ibid.; A.E. Shields et al., “The Use of Race Variables in Genetic Studies of Complex Traits and the Goal of Reducing Health Disparities: A Transdisciplinary Perspective,” American Psychologist 60, no. 1 (2005): 77–103; M. Bamshad, “Genetic Influences on Health: Does Race Matter?” Journal of the American Medical Association 294, no. 8 (2005): 937–946; and Institute of Medicine, Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care (Washington: National Academies Press, 2003).
7. N. Risch et al., “Categorization of Humans in Biomedical Research: Genes and Race,” Genome Biology 3, no. 7, 2002, genomebiology.com/2002/3/7/comment/2007 (20 September 2005).
8. See, for example, M. Rothstein and P.G. Epps, “Pharmacogenenomics and the (Ir)Relevance of Race,” Pharmacogenomics Journal 1, no. 2 (2001): 104–108.
9. See Bamshad, “Genetic Influences,” 937.

To read the paper by Sankar and Kahn, please click here.

Rick Carlson (rickjcarl{at}aol.com) is a clinical professor; Policy Programs, and a senior adviser in the Resource Center for Health Policy, Department of Health Services, School of Public Health, at the University of Washington in Seattle.

DOI: 10.1377/hlthaff.w5.464
©2005 Project HOPE–The People-to-People Health Foundation, Inc.