A 10-year-old girl wakes up in the night, wheezing and struggling for breath. She uses her asthma inhaler, but as usual it doesn鈥檛 help. In the morning, her mother takes her to the doctor鈥檚 surgery, seeking an explanation. The doctor asks a few questions but seems flummoxed. Then she has an idea. She looks at them closely, and after a moment鈥檚 hesitation asks: 鈥淵ou wouldn鈥檛 be Puerto Rican, by any chance?鈥�
The question seems ridiculous, even racist. What have the girl鈥檚 ethnic origins got to do with anything?
But people from Puerto Rico do suffer unusually badly from asthma, and new studies are suggesting that their genetic make-up may be at least partly responsible. At the moment only a few clued-up doctors are likely to be aware of the asthma research, but in the future, doctors may base a growing number of their treatment decisions on people鈥檚 ethnic origins. Next Thursday, a panel of the US Food and Drug Administration (FDA) is scheduled to decide whether or not to approve BiDil, the first ever race-specific medicine. BiDil is a treatment for heart failure that the manufacturer wants to market for black people only. Its argument is that BiDil works much better in African Americans than in white people because of a key biological difference between the groups.
Advertisement
On the face of it, the manufacturer鈥檚 clinical results support its claims. In some quarters BiDil is portrayed as part of a new wave of personalised medicine in which patients get carefully targeted drugs to ensure maximum response rates and minimum side effects. But others say that the drug is a distraction from the real causes of heart disease among African Americans 鈥� poverty, ill health and poor access to healthcare.
BiDil looks set to spark off one of the most explosive rows the FDA has ever faced. For a start, the science of race is hugely contentious: some people dispute whether there is any such thing as race in genetic terms. In addition, the reasons why BiDil seems to work better in black people than in white are hotly contested. 鈥淪ome people talk about heart failure as a different disease in blacks in what I can only describe as an irresponsible manner,鈥� says Jonathan Kahn of Hamline University in St Paul, Minnesota, who studies the legal and ethical issues surrounding ethnicity and medicine. 鈥淚t fuels talk about how races are genetically different. This is very unfortunate and dangerous.鈥�
Over the past few decades a growing body of research has probed the genetics of ethnicity. A seminal paper came in 1972, when Richard Lewontin, a geneticist at Harvard University, reported that there was far less variation between different races than there was between individuals of a single race. His research seemed to show that any differences between ethnic groups were literally skin-deep. Since then, however, various other studies have both supported and refuted Lewontin鈥檚 conclusions, leaving the genetic significance of ethnicity an open question (see 鈥淩ace in the genes鈥�).
In this field politicians, special-interest groups and media commentators seize on almost every new piece of research to dissect its implications for social policies, particularly in the US, where racial inequalities have deep historical roots. Anything to do with race and medicine can generate controversy. People from some ethnic minorities, particularly African Americans, tend to have worse health and die younger than white people. It is widely accepted that this is at least partly due to these groups having less money and poorer access to healthcare. The extent to which biology could also play a role is fiercely debated. 鈥淲hen you start talking about races being genetically different it can distract us from seeing these very real health inequalities,鈥� says Kahn.
This is the arena into which BiDil steps. So what are the implications of a drug that is supposed to work much better in black people than it does in white people? Should it affect our notions of race 鈥� or is the truth more complex?
While the very idea of dividing people into races remains controversial, doctors have long known that certain diseases are more common in some populations than others. Some of these disparities undoubtedly stem from differences in people鈥檚 environments, but for others, genes are at least partly responsible (see 鈥淗ealth by race鈥�).
鈥淭he manufacturer wants to market the treatment for black people only鈥�
Researchers are even finding clinically relevant differences between ethnic groups within a single race. Puerto Ricans, for example, are known to suffer more from asthma than Mexicans, although these two populations are usually seen as belonging to the same race, Hispanics. Puerto Ricans are also less likely to respond well to the most widely prescribed asthma medicine, salbutamol (known as albuterol in the US), with one of the genes responsible being identified in March (American Journal of Respiratory and Critical Care Medicine, vol 171, p 563). So for the fictitious asthma patient described earlier, the doctor鈥檚 question about her ethnic origin was in fact highly relevant.
It is not an asthma medicine, however, but the heart failure drug BiDil, that could make history by becoming the first pharmaceutical licensed for use in a single ethnic group. Heart disease is one of the biggest causes of death in the west. In heart failure, the organ cannot pump blood round the body forcefully enough, leading to fatigue and breathing difficulties. The condition is usually treated with several medicines that help the heart in different ways, but patients generally deteriorate over time, often developing further heart problems, and have an annual death rate of 10 to 20 per cent.
BiDil is a combination of two drugs, isosorbide dinitrate and hydralazine, that together raise the concentration of a signalling molecule, nitric oxide, in blood vessel walls. This lowers blood pressure by dilating the vessels and so reduces strain on the heart. Trials involving white and black people suggested that the combination therapy was less effective than ACE inhibitors, a front-line medicine for heart failure. But last year a trial involving 1050 people who identified themselves as African American showed that when BiDil was combined with standard therapies, the annual death rate fell by an impressive 43 per cent compared with a placebo (The New England Journal of Medicine, vol 351, p 2049).
鈥淏lack people may suffer from a different kind of heart failure鈥�
These are the results that the FDA panel will consider next week, when it decides whether or not to recommend approval of BiDil. The manufacturer, a Boston biotech firm called NitroMed, will try to convince the committee that BiDil is fulfilling an unmet medical need. The firm points out that African Americans are more likely to develop heart failure than whites. Some estimates suggest their increased risk is at least 25 per cent; the firm claims it is nearer to 100 per cent. NitroMed acknowledges that social inequalities may be involved, but it says they cannot account for all the increased risk. Blacks are also less likely to respond well to ACE inhibitors, meaning that new treatment options for this patient group are badly needed. 鈥淗eart failure is a catastrophe,鈥� says NitroMed chief medical officer and cardiologist Manuel Worcel. 鈥淪o when you have a drug that saves around half [of patients], this is huge progress.鈥�
So why does BiDil work better in black people than whites? NitroMed suggests that black people generally have lower levels of nitric oxide in their blood vessels than white people. Blacks may even suffer from a different kind of heart failure altogether: their heart failure tends to be a consequence of long-term high blood pressure, or hypertension, while in white people it usually stems from narrowed arteries or a heart attack damaging cardiac muscle. Low nitric oxide levels can contribute to hypertension, which is also more prevalent in African Americans. 鈥淭here has been a lot of controversy, but we know we have to take ethnic origin into account,鈥� says Worcel.
But look closer and the case for BiDil鈥檚 approval specifically for black people is by no means cut and dried, and may have as much to do with economics as medicine.
BiDil is no new therapy. Its two constituent drugs have been prescribed to people of all races for decades, both separately and in combination, to lower blood pressure and to treat heart failure. They are usually prescribed to patients who do poorly on ACE inhibitors. In the 1980s, the two drugs鈥� efficacy in combination was tested in two trials involving both white and black people with heart failure. The trials suggested the drug combination reduced the death rate, although not to the same extent as ACE inhibitors. In 1987 Jay Cohn, one of the trials鈥� principal investigators, applied for a patent on a single-dose combination of the two drugs for use in heart failure, which was granted. The combination therapy, which acquired the name BiDil, would allow patients to take both drugs in a single pill, a significant advantage when many patients were already taking three or four other medicines daily. But in 1997 the FDA rejected an application for approval of BiDil on the grounds that the trials鈥� design made it impossible to draw any firm conclusions.
The prospects for BiDil were looking bleak, until Cohn and his colleagues reanalysed the results and found that the black participants in the studies 鈥� around 30 per cent of the total 鈥� seemed to respond better to the medicine than the white ones. Crucially, in 2000 he was granted a patent for use of BiDil in African Americans. Shortly afterwards NitroMed acquired the rights to BiDil, and carried out the recent trial in African Americans.
Many commentators, including Gregg Bloche, a professor of law specialising in healthcare financing at Georgetown University in Washington DC, believe the explanation for NitroMed鈥檚 race-specific marketing strategy is obvious. The general patent for BiDil expires in 2007. The more recent one that covers the medicine鈥檚 use in African Americans lasts until 2020. 鈥淭he patent for the old use was about to run out, but now they鈥檝e got a patent that will keep them doing well for 15 years,鈥� says Bloche.
So was NitroMed鈥檚 decision to test BiDil on blacks more about profit than hard science? 鈥淭hat鈥檚 absolutely false,鈥� says Worcel. In the reanalysis of the earlier studies, the combination was found to improve survival by 47 per cent in black patients but only 15 per cent in whites, he points out. 鈥淲hat came first was this medical knowledge, and it鈥檚 not unusual to translate medical knowledge into patents. It鈥檚 part of how we progress in medicine.鈥�
But some doctors believe that the way the company has chosen to interpret the science is misleading. For one thing, environmental factors, not genetic ones, are increasingly being fingered as the chief cause of African Americans鈥� high blood pressure. A recent study found enormous variations in the prevalence of hypertension within races (BMC Medicine, DOI: 10.1186/1741-7015-3-2). Among black people, the lowest prevalence was found in rural Nigerians (13.5 per cent) and the highest in African Americans (44 per cent). Rural Nigerians were considerably less likely to have hypertension than even European Americans (26.8 per cent). Perhaps African Americans have more hypertension because their lower income levels tend to go along with unhealthier lifestyles and worse healthcare, not because of an innate predisposition to low nitric oxide levels.
Another concern is that the picture presented by NitroMed of different types of heart failure in different ethnic groups may be an oversimplification. While heart failure undoubtedly tends to arise from different causes in different ethnic groups, hypertension is the trigger in some white patients and heart attacks in some blacks. And nitric oxide levels are low in the blood vessels of all patients with hypertension, regardless of race.
BiDil鈥檚 advocates deny that testing it in a black-only trial set an unfortunate precedent for drug development. 鈥淭his trial simply says that there is a mechanism of progression of heart failure that is treated by this combination, which is more prevalent in an African American population,鈥� says Anne Taylor, a cardiologist at the University of Minnesota and the lead researcher on the BiDil trial.
But if BiDil gets its race-specific approval next week, many doctors may conclude that skin colour is a good guide to whether a particular patient will do well on this drug. They could still legally prescribe the therapy for whites 鈥渙ff-label鈥� but in practice, they may be much less likely to. 鈥淭he problem is, when you racialise, there are going to be many white people who could have benefited from the drug not being prescribed it, and many black people who have a different kind of response to the drug being given it,鈥� says Troy Duster, a sociologist at New York University.
鈥淎 patient鈥檚 clinical history could be a better guide to response than race鈥�
What many people wonder is whether NitroMed could have used patients鈥� clinical history as a better guide to BiDil response than race. To its credit, the firm has been trying to pinpoint clinical and genetic markers that could eventually indicate how well a particular patient will respond to BiDil. If the firm succeeds, the therapy could become a genuine 鈥減harmacogenomic鈥� medicine, one that is prescribed according to a patient鈥檚 genetic make-up (see 鈥淛ust for you鈥�). 鈥淥nce we have this kind of information we could use genomic markers to identify patients who could benefit from the drug regardless of the colour of their skin,鈥� says Worcel. NitroMed is likely to announce preliminary findings from this research at a cardiology meeting in September. But a commercially available genetic test is likely to be many years away.
In the meantime, the FDA must use the existing data to make its decision. At the meeting next week an advisory panel will hear arguments from NitroMed, and its supporters and critics. The panel鈥檚 recommendation is usually, although not always, followed by FDA approval. The agency has promised NitroMed a final decision by 23 June. The FDA has an unenviable task; whatever it decides it could be accused of failing African American people.
Most commentators predict that the agency will approve the therapy in some form. Despite all the caveats, BiDil is an effective medicine that works well in a currently under-treated patient group. That is not an argument that convinces Kahn, however. 鈥淵ou can鈥檛 have a kind of 鈥榚nds justifies the means鈥� mentality,鈥� he says. 鈥淵ou can鈥檛 just say, 鈥榃e鈥檙e saving people by using race,鈥� and imply races are genetically different when they鈥檙e really not.鈥�
Then again, if NitroMed had not gone down the race-specific track, the final trial would never have been done and nobody with heart failure, white or black, would have had a chance to benefit. 鈥淭hat,鈥� says Kahn, 鈥渋s the great irony.鈥�
Race in the genes
HOW much are societal concepts of race reflected in the human genome? In 1972, Richard Lewontin of Harvard University turned conventional thinking on its head by suggesting the answer 鈥渧ery little鈥�.
Lewontin analysed variations in blood proteins across different populations and discovered that at least 90 per cent of the diversity among humans is found within races and only 10 per cent between them. These results were supported in 1997 by a more sophisticated analysis of DNA variations by Guido Barbujani from the University of Ferrara in Italy.
Two years ago, however, the statistician and geneticist Anthony Edwards pointed out that Lewontin鈥檚 approach ignored a crucial source of information in DNA: patterns of genetic variation spread throughout the genome (BioEssays, vol 25, p 798). If you compare genes one by one there may seem to be more diversity within populations than between them. But if you look at several genes together, correlations emerge that correspond to particular ethnic groups.
Edwards鈥� argument has been borne out by several studies, including one published in February that looked for such patterns among 326 stretches of highly variable DNA known as microsatellite markers, in 3636 people. This analysis grouped people into clusters that corresponded to four commonly recognised racial groups: white, African American, East Asian and Hispanic (American Journal of Human Genetics, vol 76, p 268).
Significantly, the genetic categorisations showed 鈥渘ear perfect鈥� correspondence to the racial groups into which the people classed themselves. Only five people showed genetic cluster membership different from their self-identified race.
Not everyone is convinced of this study鈥檚 significance. 鈥淲e don鈥檛 know whether these particular markers indicate anything about behaviour or disease manifestation 鈥� you simply have this correlation,鈥� says Troy Duster, a sociologist at New York University. And no matter how noble people鈥檚 intentions, he says, such research may give ammunition to those who want to believe in the innate superiority of one race over another. 鈥淚t gives a kind of scientific halo to their arguments.鈥�
On the other hand, some commentators believe the pitfalls of this approach are being overstated. 鈥淎ll we鈥檙e talking about here is population genetics,鈥� says Sally Satel, a psychiatrist at Yale University School of Medicine and author of PC, M.D.: How political correctness is corrupting medicine. 鈥淧eople who share a common ancestry simply have a higher frequency of certain genes. If you could get the word 鈥榬ace鈥� out of it then even those who get nervous about the social implications of this would calm down.鈥�
Ironically, by focusing on race, researchers can identify the very genes that will eventually allow doctors to ignore it and tailor their treatments to individual patients鈥� genetic make-up. In a strategy called admixture mapping, which has only become technically possible over the past year, geneticists use people of mixed ancestry to track down disease genes. If a disease has a much higher prevalence in one race compared with another, the genome of a patient with mixed ancestry who has the disease can be used like a treasure map to locate the genes responsible.
David Reich of Harvard Medical School, for example, is searching for genes that play a role in multiple sclerosis, which is extremely rare in Africans, and about twice as common in European Americans as in African Americans. In a patient with both African and European ancestry, Reich can draw a map of their genome showing which areas have the most 鈥淓uropean鈥� genes and which the least. A multiple sclerosis gene is most likely to be found in a region with lots of European DNA.
Just for you
IF BiDil is approved for treating African Americans with heart failure, it will become the first drug targeted at a specific race. In future, however, doctors could tailor people鈥檚 treatments to their precise genetic make-up, with the aim of maximising response rates and banishing side effects. This is the new science of pharmacogenomics.
It has long been known that people vary widely in their response to drugs, partly due to differences in their age and general health, but also because of their genes. At the moment, finding out which medicines suit an individual best is largely a process of trial and error. But scientists are starting to unravel the genetic factors involved.
One important set of genes encodes a large group of liver enzymes called the cytochrome P450 (CYP) family, which helps metabolise dietary toxins. The enzymes do the same job for many medicines. Some people carry an enzyme variant that works less well or not at all, meaning certain drugs can build up in high levels in their blood, increasing the chance of harmful side effects. The antidepressant Prozac is one example. There is a genetic test for inability to metabolise Prozac, though it is rarely used.
CYP genes can affect the response to drugs in another way. Some medicines are inactive precursor chemicals that need to be converted into an active form by enzymes like the CYP family. People wih a faulty CYP gene, cannot make the conversion. For example, up to 1 in 10 white people and smaller numbers in other ethnic groups have a mutation that means they are poor converters of the painkiller codeine into its active form, morphine. If you never get any pain relief from codeine, you probably have the mutation.
Other important genes are those involved in the disease process itself. Arguably the first example of a personalised medicine is Herceptin, a breast cancer drug. Herceptin can only work on tumour cells that have extra copies of a gene called HER2, which is the case for about one in four breast cancers. Herceptin is approved for use only in patients with this kind of tumour.
Since Herceptin鈥檚 launch in 1998, only a few other pharmacogenomic medicines have followed it to the market, including Glivec (called Gleevec in the US) for leukaemia and Erbitux for bowel cancer. Some believe the promise of this kind of therapy has been overhyped. 鈥淭here鈥檚 a gap between the expectations being created around this technology and what [it] can definitely deliver,鈥� says Adam Hedgecoe, a sociologist at the University of Sussex in the UK and author of The Politics of Personalised Medicine.
There are many reasons for the slow progress. Some in the industry are sceptical that the benefits of this new approach will outweigh the costs. While pharmacogenomics promises that drug development will no more be wrecked by unpredictable side effects, it could further inflate R&D costs. And it is likely to end the era of blockbuster drugs that dominate their markets, like Prozac and Viagra. After all, a genetic test that shows which people should respond to a medicine will also indicate who shouldn鈥檛 take it.
In an attempt to speed the next wave of personalised medicines to market, the US Food and Drug Administration launched a set of guidelines for the industry in March. The FDA is encouraging firms to submit all available pharmacogenomic data on their products, whether it involves genes that are widely accepted to affect a disease or drug response, or more exploratory research. It has promised not to use exploratory research to restrict how a new drug can be used. 鈥淚t would be a shame if there was data out there that was useful but we weren鈥檛 seeing it,鈥� says Felix Frueh, the agency鈥檚 associate director for genomics. 鈥淭o prevent paranoia, it was important to put these guidelines out.鈥�
Clare Wilson
