DNA microarrays are standing on the doorstep of clinical research: Every other week, it seems a scientific study in a high-profile journal like Nature or The Lancet reports the discovery of a new set of biomarkers, based on gene expression data from microarray experiments that may help, for example, predict the clinical outcome of cancer.
But what does it take to transform these results into routine clinical tests, and what will the role of microarrays be in diagnostics in general?
“I think there are going to be three or four major platforms for diagnostics,” said Jorge Leon, vice president of applied genomics at Quest Diagnostics, a major US reference lab. “One of them is going to be [DNA] arrays,” he said. The others will include immunoassays, traditional ELISAs and antibody arrays, as well as fluorescent in situ hybridization (FISH) and protein mass spectrometry.
Roche Diagnostics is at the crest of this wave. Ten years after introducing PCR, the company is working on its first prognostic DNA chip in collaboration with Affymetrix. The p450 array detects polymorphisms in the gene for the drug-metabolizing enzyme p450, and may be able to predict how individuals metabolize certain drugs. Its launch is scheduled for early next year, said Heino von Prondzynksi, global head of Roche Diagnostics, at a meeting in New York City two weeks ago.
After marketing the p450 chip to pharmaceutical companies, Roche is planning to market it more widely within the clinical field. Von Prondzynski also said that Roche is currently developing a sepsis array, which will take another two to three years to reach the market.
But others are not so sure that the microarray platform, which has yielded such a rich vein of results in the research arena, is the appropriate tool for diagnostics. Charles Perou from the University of North Carolina at Chapel Hill, who has used microarray gene expression data to prognostically classify breast cancer tumors, is now developing with his colleagues a PCR-based clinical test from his results, because he does not think microarrays with their specialized equipment will reach the clinic anytime soon.
Heiner Dreismann, president and CEO of Roche Molecular Systems, admitted that diagnostic DNA microarrays may take a decade to become mainstream, but they “will be used by cutting-edge labs earlier,” he said. It will be a challenge though, he said, to equip the majority of testing labs with the new platform.
In any case, the road to developing a diagnostic test from a set of biomarkers is long. New biomarkers need to be validated in initial clinical studies, Leon said. Next, a reference laboratory — often in collaboration with the researchers who discovered the biomarkers — can develop a so-called home-brew test. This does not require FDA approval and can be marketed directly to physicians as a service. Typically, Leon said, this development takes about three years. Genomics companies value reference labs not only for their large numbers of patient samples: “We could license the technology, we could put equity into the company, we could fund the projects, but most important, we can ... commercialize the test and pay them royalties,” Leon said. Alternatively, an in vitro diagnostics company might license the biomarkers and develop them into a test kit, which needs FDA approval and takes about six years to develop.
In the case of multiple biomarkers derived from genomics data, Leon thinks the timescale for both routes might increase because validation could require more extensive clinical trials. Payors need to be convinced of its value, and be willing to reimburse providers. This requires not only scientific evidence that it is at least as good as the current “gold standard,” and that it is “cost-effective,” meaning that using the test — even if it is more expensive — saves money elsewhere or in the long run.
These factors are key, said Leon, because “you can’t go out with a test if you cannot get paid.”