BiDil is a recently approved therapy indicated for congestive heart failure patients already taking standard treatments. It is not a new drug. It is a combination of two existing drugs—isosorbide dinitrate and hydralazine—already marketed seperately as generics. What is new about BiDil is its scientifically dubious distinction as the first drug ever approved in the United States for exclusive use in African-Americans.

But BiDil was not always destined for African-American hearts. Its transformation into an 'ethnically targeted' medicine is relatively recent and follows a long history of clinical testing for heart disease in the general population. Far from “illuminating the future of medicine,” as Sally Satel of the American Enterprise Institute has characterized it, BiDil's development says more about commercial opportunism, clever marketing and the primacy of political correctness over science than it does about the value of selectively targeted drugs.

BiDil is an example of the old adage 'third time lucky.' It failed to make the regulatory grade in two separate clinical trials in general, nonapartheid populations in the 1980s and 1990s. In the first, it showed only marginal, nonsignificant benefits and, in the second, poor trial design and statistical uncertainty dashed its hopes. That could have been the end of the story.

But in 1999, NitroMed in-licensed the drug on the basis of a reanalysis of the clinical data and other health statistics on mortality rates in blacks with heart disease, having filed a patent for the use of BiDil specifically in African-Americans. It argued that the rate of deaths from heart failure in blacks is twice as high as in whites in the 45–64 age group: in other words 'blackness' is positively associated with mortality rate in cardiac disease. The company sought a race-specific indication from the FDA and, in March 2001, they received an approvable letter from the agency pending the results of a late-stage trial in African-Americans. The trial was duly undertaken: 1,050 black patients of median age 57 with heart failure received either BiDil or placebo, plus standard therapy. The results were so impressive (a 43% improvement in survival in the treatment arm) that the trial was prematurely halted last summer, results were released in November and FDA approval followed on June 23 this year.

But closer scrutiny of NitroMed's rationale reveals that the 2:1 ratio is questionable on at least two counts, as Jonathan Kahn, a law professor at Hamline University in Minnesota, has pointed out. The first is that the ratio is based on data from the National Center for Health Statistics dating from 1981. More recent data from the Centers for Disease Control in 1995 showing that individuals age 65 and older have mortality rates from heart failure nearly equal for blacks and whites (a 1.1:1 ratio) were not included. Moreover, a whopping 94% of all heart failure mortality occurs over the age of 65. And yet the NitroMed reanalysis focused on data from patients aged from 45 to 64 years, thus boosting the black to white differential. The interpretation of these data suggests that BiDil is more effective in blacks merely because the black population studied is more prone to die of heart disease at a younger age.

On the positive side, black patients now have an effective drug that might not have been available at all had the efficacy-diluting effects of white people in the trial not been excluded. What may also be true, however, is that white patients may have been deprived of an effective drug by the same fallacious analysis. NitroMed has both neatly avoided the need for a large, costly, multiracial clinical trial and gained 13 extra years of market exclusivity through its racially targeted patent (the original BiDil patent, which does not refer to race, expires in 2007).

The broader concern is that an ostensibly science-oriented body like the FDA did not reject NitroMed's analysis. Or worse, the agency felt it should take the path of affirmative ethnic drug approval in the face of political pressure. Whatever the case, the decision to approve this drug with a race-specific indication represents a regressive step in medicine. Race is simply a poor proxy for the environmental and genetic causes of disease or drug response (Nat. Biotechnol. 20 637, 2002). And it ignores a multitude of other characteristics that might be much more relevant to the disease. An article published last year in Nature Genetics (Suppl., S36–S37, 2004) reports that of 29 medicines where differential racial responses have been 'identified,' only 9 show a reasonable underlying physiological basis and even fewer (4) show any evidence of a genetic cause.

Proponents will argue race is a placeholder—a crude marker of variation that might as well be used until better markers (genetic or environmental) for differential responses are identified and cost-effective genetic screening technologies become available. But scientifically, race is a meaningless marker of anything. Pooling people in race silos is akin to zoologists grouping raccoons, tigers and okapis on the basis that they are all stripey.

The fact is race has been, and is still, used in patient segmentation simply because it can be. NitroMed's reanalysis was possible because the early trials collected data on self-assessed racial grouping. But the black people in the trials didn't respond to BiDil because their skins contained more melanin: they responded because they were more ill (than white people of the same age).

BiDil is thus a poor model for targeted medicine. Drugs like it will not redress the inequities in US healthcare, which the Institute of Medicine estimates account for the deaths of 80,000 or so African-Americans every year. These disparities are due to a host of factors, including genetics, diet and lifestyle, social discrimination, economic inequality and even geographic location. They do not result from an absence of drugs targeted to blacks. And they will never be solved by a brave new era of racially targeted therapies.