Going back to the future in personalized medicine

Shaping the future of personalized medicine is not all about developing expensive new drugs -- it will also mean revisiting older, cheaper medicines armed with new genetic knowledge.

Recent discoveries of genetic clues as to why medicines work better in some patients than others suggests combining new tests with old drugs will be a cost-effective approach -- attractive to governments and insurance companies, experts say.

There are two sides to personalized medicine -- there is work in looking for new gene clues for the design of new drugs, and we are also doing a lot of work on currently used medications, said Colin Palmer of Dundee University, whose role as head of pharmacogenomics puts him at the heart of work to use genetic information to personalize medicine.

We're trying to get rid of the one-size fits all approach ... and create more effective drugs tailored to the individual.

Few believe it is possible to make all drugs work for all patients all the time, but experts say the current situation -- where many patients do not get any benefit -- demands action.

Its easy to see why. According to a report by PricewaterhouseCoopers earlier this year, patient response rates to medicines can be very low -- varying from 20 percent to 75 percent, depending on the drug.

It is no surprise that industry is under pressure to improve efficacy and safety, thereby making drugs more cost-effective.

It is already the case that more and more new drugs, particularly for cancer, are coming to market with a so-called companion diagnostic -- a test allowing doctors to determine if a patient has the right genetic makeup to respond to treatment.

In Europe, there are around a dozen drugs -- including GlaxoSmithKline's Ziagen for HIV and AstraZeneca's lung drug Iressa -- that require the use of companion diagnostics.

And in the United States, the Food and Drug Administration requires patients be tested for genetic variants before taking Pfizer's HIV drug Selzentry, Eli Lilly and Bristol-Myers Squibb's Erbitux for colorectal cancer and Roche's Herceptin for breast cancer, among others.

GENETICS AND GENERICS

But these are new and highly pricey drugs -- and experts say payers may be more encouraged by recent studies which show genetic clues being found for response rates in generic drugs.

In the last year or so people have been beginning to find gene markers in much more common areas, said Donald Singer, a professor of clinical pharmacology and therapeutics at the University of Warwick's medical school. We are really on the cusp at the moment in terms of the cost effectiveness.

While pharmaceutical companies would rather promote new drugs, he believes a better approach for payers may be to revisit old drugs armed with greater genetic knowledge.

A study published last week showed that common asthma drugs -- salbutamol, a popular inhaler medicine also known as Ventolin, and salmeterol, an ingredient in Glaxo's Advair -- do not work in patients with a particular genetic make-up and may make things worse.

Another study showed that about half of patients given the generic drug tamoxifen as a hormone therapy in breast cancer have a genetic variation which helps them metabolize the drug -- meaning they are likely to respond well -- but 8 percent have a gene type which means it will not work.

Palmer's team is also investigating the genes involved in defining whether a patient can respond well to statins, a class of drugs used by millions of people to try to lower cholesterol.

In some of these areas, scientists say a relatively cheap and easy test, such as a cheek swab or blood test, could be carried out to see ahead of time whether a patient is likely to respond well to the medicine usually prescribed.

From the point of view of governments, testing and then going for the older off-patent drugs could be more cost-effective, rather than plowing money into new ones, Singer said.

Experts in this field point to rapid acceleration in genetic technology since 2003 when scientists completed the Human Genome Project -- a decade-long race to sequence all the DNA in people.

Some companies already offer a genotyping service, where you can send in a DNA sample and, for a fee, they give you a typing for as many as a million genetic variants.

For now such information is not widely useful without the ability to act on it, but as studies on common medicines reveal more about how and when they work, clinical knowledge about how to exploit those genetic variations to best effect is growing.

In real life what you really want is to be able to go to your doctor, get a blood test which could lay out a genetic map, and then they prescribe based on the test results, said Singer.