In what could prove to be a landmark event in pharmacogenomic medicine, US firm Clinical Data has launched a phase III clinical trial for the depression drug vilazodone and will concurrently develop a diagnostic test.

"We believe the concept of the blockbuster drug is a failed one," says Clinical Data president and CEO Israel Stein. In what he believes is an industry first, Stein says his firm, the parent company of pharmacogenomics biotech Genaissance, intends to explore "the simultaneous development of the biomarker and the drug."

Depression is in many ways an attractive disease for biomarker development. There is widespread variation in patient response to anti-depression drugs, with half of patients failing to achieve satisfactory results with first-line treatments. The total market is valued at US$10 to $13 billion.

Clinical Data hopes to complete the phase III trial of vilazodone by mid-2007. It must also conduct a long-term safety analysis, the terms of which have yet to be reviewed by the FDA. If all goes well, the company hopes to file its NDA by the end of 2008.

Clinical Data obtained an exclusive worldwide license to vilazodone from Merck KGaA in 2004. The German pharma had previously licensed the drug to GlaxoSmithKline, but the rights were returned following equivocal phase II trial results.

"Vilazodone is a molecule that should be a very effective anti-depressive agent," says chief medical officer Carol Reed. "If you look at its behaviour in receptor binding, in vivo and in vitro, animal models, and even in the phase II trials, it does appear to be an effective agent. Genetics will help clearly demonstrate people for whom the drug will work."

But Reed says the drug encountered a not uncommon problem in its phase II studies -- "the issue of increasing placebo response. Active competitors were included in some trials. The active [comparison drugs] also failed to beat placebo. It's a statistical failure. Phase II studies aren't designed to clearly demonstrate efficacy." Reed notes that the comparison drugs, all of which are approved, did beat placebos in earlier trials, or they would not have been included.

400 patients

Clinical Data plans to recruit 400 patients, all in the US The search for genetic biomarkers that can predict drug efficacy and response will be conducted as the trial progresses.

"Genaissance has developed its reputation in haplotype mapping and candidate gene study. We also have the ability to apply whole genome scans and expression analysis. All of these options are available," says Reed. "We will lean heavily on proprietary information on SNPs [single-nucleotide polymorphisms] and haplotypes across diverse ethnic populations."

A lot is already known about biomarkers for antidepressants. Vilazodone is known to bind to two proteins -- the serotonin transporter, for which there is published data on promoter polymorphisms, and the 5HT1A receptor. "So any downstream pathway genes are obvious candidate genes," says Reed.

Among the genes of interest will be variations in cytochrome enzymes, which she refers to as the "state-of-the-art understanding of how drug metabolising enzymes affect individual response of therapeutics." (Genotyping at Genaissance relies heavily on the Sequenom platform, while expression analysis involves Affymetrix microarrays.)

Stein says: "It's only in the last six months that we identified this opportunity in molecular diagnostics. This is a central focus for us going forward." Another area of interest is the development of safety markers for the anti-psychotic drug clozapine -- an effective drug against schizophrenia, but one with limited use because of serious side effects in some patients, including a severe reduction in white blood cell count).