When a giant of the drug industry claims that most medicines don't work for most people who take them, is it time to ditch our prescriptions and take up prayer? Dr Allen Roses, the worldwide vice-president of genetics at GlaxoSmithKline, was yesterday quoted saying that the vast majority of drugs - more than 90% - only work in 30% or 50% of the people they are prescribed for. At first glance, such an admission may seem alarming. But is it so surprising?
Not according to Richard Ley, spokesman for the Association of the British Pharmaceutical Industry, the trade association for companies in the UK producing prescription medicines - more than 90% of those prescribed through the National Health Service. Ley told the Guardian: "It's not really news to anyone that not all medicines work in all the people all of the time. What is certainly true is that we often don't know why." And he agrees with Roses that the answer is likely to lie in people's genetic make-up.
But for those patients who depend heavily on their prescriptions, the admission will come as more of a shock. Not least because of the soaring cost of the NHS drugs bill to the taxpayer. Currently the annual bill stands at £7.2bn and is rising. So why is the efficacy rate of some drugs so poor, and why does one drug work for one patient and not for another?
As scientists unravel the secrets of the genetic code, they are finding out more about the importance of variations between the proteins for which genes code in determining how individuals respond to drugs. The two disciplines of pharmacogenetics and pharmacogenomics are focusing on, respectively, the roles of single genes and the whole collection, or genome.
Work in these areas aims to enable better targeting of drugs to people who would most benefit, and indeed suffer fewer side-effects, and also allow drug companies to refine products based on knowledge of the genetically determined ways in which people respond to them.
David Goldstein, professor of genetics at University College London and an international expert in this field, points out: "We actually know of about 42 genetic differences which affect how we respond to drug treatments, and this body of knowledge is growing. What is less clear is how to get such knowledge translated into real improvements in drugs and hence improve treatments."
Ley stresses that alongside genetic bases of drug responses, the lack of knowledge about the fundamental processes underlying many conditions explains in part why it can be so difficult to find effective treatments. "Take asthma for example. We're not yet sure what causes it, and it's probably a complex mix of genetic and environmental factors. Knowing more about these basics will help researchers to find new and better drug treatments."
He also points out: "While it's an over-simplification, it's important not to get too worried when we hear that a drug helps, for example, 30% of those who take it. As long as you have three drugs for that condition, you may well find something for almost everyone who needs it. This is why it is important to go on researching and developing more medication, even though this is sometimes criticised as making useless variations on a theme. It's the only way to extend coverage so that we have effective treatments for all who need them."
It certainly seems as though pharmacogenetics has promise, but does not hold all the answers. Dr Mike Clarke, Director of the UK Cochrane Centre, part of a renowned international collaboration which helps people to make better decisions about their healthcare (www.cochrane.org) is adamant that better and bigger drug trials are an immediate way forward. As he says: "People are more complex than their genes, though they're complex enough. We need to do bigger trials which enable us to look at patient subgroups, and to look at the quality of the evidence emerging from these trials in a systematic way." He adds: "With all this pharmacogenetic research around, we'll also have to ensure that the research itself is systematically reviewed to basically, be sure that it's good."
How well do your prescription pills perform?
Heart disease
This is certainly an area where knowledge would increase power against the UK's biggest killer. Professor Steve Humphries, chief executive officer of the London Ideas Genetic Knowledge Park (www.londonideas.org) and British Heart Foundation Professor of Cardiovascular Genetics (bhf.org.uk) agrees that more drugs for the same condition are often needed. "For high blood pressure, we currently have three main drug classes, and we try one first then move on to the next if it doesn't work. Knowing more about interactions between drugs and genes might certainly help us to tailor better treatments."
Humphries says that it is an obvious area for drug companies to be channelling their efforts, and may well lead to better medications for a range of conditions. However, he also sounds a note of caution. "As the field progresses, drug companies may not only be in the business of selling you a drug, but also ultimately the tests that will tell you whether that drug suits you or your patient best. One of the problems could be the emergence of a "genetic underclass". What will happen if the test tells you that drug X will decrease cholesterol by 20% in one patient, but 50% in another - will the first patient be denied it because they are not considered to respond well enough?"
Humphries also warns that many drugs are currently tested on restricted ethnic or racial groups. "For example, drugs are often assessed primarily in whites or Japanese people because they ultimately make up the big markets for them, which means that we end up knowing little about how they work in people from the Indian subcontinent or Africa and people in these regions go on being disadvantaged. It's an important political issue."
Cancer
Many cancer drugs have a particularly low success rate. According to Dr John Toy, medical director of Cancer Research UK, this phenomenon is due to the genetic complexity of the condition, and the difficulty of attacking cancer without using treatment so aggressive that the patient cannot survive. "It's such a delicate balance. A chemotherapy drug may hit 99% of the cancer cells, but leave 1% that are resistant, but you don't realise it at the time of treatment. Any new diseased cells that subsequently develop arise from that tough 1%, and may be especially hard to treat. We also know too well that cancer treatment is often very toxic. Finding out more about the genetic makeup of who fares best with what treatments will help us to not only give the most beneficial treatments, but not give it to those who would react badly."
Neurology and psychiatry
The complexity of the brain and human behaviour makes the development of drugs across neurology and psychiatry fraught with difficulty.
However, there are now successful treatments for many conditions, and while some welcome genetic advances, others fear what pharmacogenetics may bring. Julia Cream, of the Alzheimer's Society (www.alzheimers.org) points out: "Currently available drugs for Alzheimer's disease seem to work miracles in about 10% of people who take them, while they only work at all in about 50%. Despite a lot of work in genetic profiling, we haven't yet been able to identify who benefits. While we obviously don't want people to try things which don't work for them, we are equally sure that anything which denies access to some who could benefit mustn't be allowed."
Cliff Prior, chief executive of Rethink, which works with people affected by severe mental illness (www.rethink.org), is optimistic about the power of pharmacogenetics, but says that there are more immediate ways to ensure that those with mental illness are treated well. "People have been telling us for years that different medicines work for different people. The idea of pharmacogenetics, that you might have a clue as to which would work best before prescribing it, is excellent. But it's still years away from reality. Meanwhile, doctors must listen to people taking medicines and be ready to try a different one if the side effects are bad."
