Every year in the UK about 10,000 people die from adverse drug reactions and tens of thousands more become seriously ill. Yet an estimated 20 per cent of those cases are the direct result of people getting the wrong drugs for their particular genes.

For years, scientists have been predicting that once we have mapped all human genes a new era of personalised medicine will follow. Treatments will be more effective, scientists say, because they can be suited to our individual make-up. The recent claim, by the head of a leading drug company, that drugs benefit only 50 per cent of those who use them, would no longer be true as pharmaceutical companies would develop several varieties of each drug, each suited to a particular genetic “ type”.

Personalised medicine took a step closer to reality this month as researchers at the Dana-Farber Cancer Institute in Boston, Massachusetts, reported that breast cancer patients with a certain gene variation (called FVL) had an increased risk of blood clots when treated with tamoxifen.

This is the field of pharmacogenetics. Some experts believe that regular gene tests before prescribing certain drugs could well become common in hospitals, and even doctors’ surgeries, within a few years. This summer a report into pharmacogenetics is due to be published by the Royal Society and, according to the chairman Sir David Weatherall, Emeritus Regius Professor of Medicine at Oxford University, it is likely to conclude: “Genetic testing before prescribing certain drugs should be done, provided that pilot trials show it is effective.”

This report could put pharmacogenetics firmly on the health agenda and give impetus for proper large-scale trials to be conducted. The information needed to tailor drugs to particular patients is already available in the databases of drug companies and academic researchers. Before a drug can be licensed, the pharmaceutical company has to provide data on what happens to it in the body; crucially, which human enzymes are involved in breaking down the drug. In parallel, gene research centres have information about which variations in human genes produce more or less efficient versions of these enzymes. Put the two together and, in theory, you could be tested to see if your particular genes are producing the version of the enzyme that is more likely to cause problems.

This information could help in a lot of ways. One person in ten of European descent, for instance, does not make the right enzyme to turn the painkiller codeine into morphine in the body, the process that makes it effective as an analgesic. Do you know if you are one of the 10 per cent for whom taking codeine is a waste of time? Almost certainly you don’t — but you would do if your doctor tested for it.

Gene testing could also cut out months of frustrating and possibly damaging trial-and-error prescribing to find the right dose for epilepsy drugs. This year, researchers at both University College London and Duke University Medical Centre, North Carolina, reported that particular variations in two genes were found in patients who needed high doses of two of the most widely used epilepsy drugs (see box).

“The range of doses taken by patients at epilepsy clinics is great,” says Professor David Goldstein, of Duke University. “The study uncovers facts that might determine in advance which patients will need the higher dose.”

Even more intriguing is another study which found that one Caucasian male in four with a particular gene combination has ten times the risk of developing prostate cancer if he doesn’t have adequate levels of antioxidants (see box). If he does have high levels, then his risk goes right down — 65 per cent lower than the others. Wouldn’t it be useful to know about that? Many of the most widely used “blockbuster” drugs can affect different people quite differently. Those who metabolise drugs fast need up to four times the amount of Prozac that slow metabolisers require; only women with a particular gene variant benefit from taking aspirin to lower their risk of colon cancer; and so on.

Gene testing could make a real difference. So why aren’t we getting it already? Part of the problem is that no one has actually put financial and political muscle behind it. And the studies that have been done are quite small and not broad enough to influence policy — yet.

“There has not yet been enough research to determine the clinical implications,” said Professor Goldstein in New Scientist in March. “What are needed are large-scale studies to see how testing would work in practice.”

In the end governments may have to take the lead. “Neither market forces nor academic incentives are likely to produce the quality of data needed,” concluded an editorial in the Journal of the American Medical Association on using pharmacogenetics to improve drug safety and efficacy.

But that does not mean that people who are worried about a possible mismatch between their genes and prescription drugs can’t have a test right now. The UK biotech firm Genovar (01795 436688) says it can test individuals for problematic variations in the ten or so genes most commonly involved in drug metabolism. “It would cost about £500 for them all or £30 for specific ones,” says Dale Smith, the managing director. He believes that a one-off test for everyone, with proper safeguards, could save about 2,500 lives a year. “At the moment, doctors base drug dosage on factors such as weight and sex,” he says. “Knowing the gene variants could help to make prescribing more precise.”

GENE-DRUG LINKS

You need to watch your antioxidant intake if you have one of three possible variants of the gene MnSOD, which produces an enzyme that works as an antioxidant in human cells. In a trial of more than 1,200 men, followed for 13 years, those with a low level, who also had the gene variant known as AA, had an 89 per cent rise in the risk of developing aggressive prostate cancer.