While the test is commercially available, few doctors are using it yet. David Mrazek, chair of the psychiatry department at the Mayo Clinic in Rochester, MN, started using a version of the test in 2003, after studying it in his clinic. "As more clinicians understand the potential benefit, I think it will be used more," says Mrazek.

The Roche test predicts who will experience side effects with a particular drug; but scientists are also searching for a trickier diagnostic: a genetic marker that can predict which drug will relieve depression most effectively in a particular patient.

Peter McGuffin, director of the Social, Genetic & Developmental Psychiatry Centre at the Institute of Psychiatry in London, is running a large, multicenter trial of 1,000 patients taking either citalopram, an antidepressant that acts on the neurotransmitter serotonin, or nortriptyline, which acts on the neurotransmitter norepinephrine.

Scientists are searching for genetic variations in the serotonin or norepinephrine pathways that predict how well a person responds to each drug. McGuffin and collaborators also plan to run a whole genome scan on each patient to search for genes outside these pathways that may also play a role in drug response to these two antidepressants. Experts say patients can expect to see tests derived from this type of trial in five to ten years.

Although the study is only partly completed, McGuffin has already had some success. His team confirmed that a variation in the gene for the serotonin transporter makes people less likely to respond to serotonin-based antidepressants.

Some drug companies are already starting to incorporate genetic testing into their drug development process. Clinical Data, a diagnostics and therapeutics company based in Newton, MA, announced plans this week for a clinical trial of a novel antidepressant that acts on the serotonin system. The drug failed earlier clinical trials; but Clinical Data researchers plan to develop a diagnostic test to identify a subset of people who do respond to the drug. If successful, the test would be marketed along with the drug.

Indeed, searching for genes that predict who will respond to a drug may actually be a more effective way to study depression than searching for genes that are implicated in the diseases itself, according to Steven Hamilton, a psychiatrist and geneticist at the University of California, San Francisco, who is involved in a large-scale study of depression in the United States. It is simpler, he says, to look for genes involved in drug response than at the complex genetic and environmental factors implicated in depression.

Caption for home page image: The AmpliChip CYP450 Test analyzes variations in two genes that play a major role in the metabolism of many widely prescribed drugs. Image courtesy of Roche Diagnostics.