Up until now, it has been unclear how the US Food and Drug Administration will regulate multi-analyte diagnostic tests, including those comprising protein biomarkers.
This may soon change: In April, the FDA issued a draft guidance for industry on how to prepare submissions for so-called multiplex tests that assay more than one analyte simultaneously. But the jury is still out on how relevant the final guidance, which is geared towards nucleic acid-based tests like DNA microarrays, will be for protein-based assays.
The FDA decided to issue the guidance, entitled “Multiplex Tests for Heritable DNA Markers, Mutations and Expression Patterns; Draft Guidance for Industry and FDA Reviewers,” in response to mounting requests from industry on how the agency would evaluate microarray-based and other multiplex tests. “At the same time, we weren’t getting any submissions, and we wanted to make sure we had something in place,” said Elizabeth Mansfield, the official from the FDA’s Office of In Vitro Diagnostic Device Evaluation and Safety who wrote the draft. While the document was conceived with nucleic acid-based tests in mind, “we felt that any kind of test ... that generates multiple datapoints would fit generally into this guidance,” said Mansfield — including those using protein arrays and protein mass spectrometry.
The draft, which is available for download at http://www.fda.gov/cdrh/ oivd/guidance/1210.pdf, is open for comments from industry. After July 21, the FDA will start revising the document, based on these comments, and issue a second draft. After another comment period, a final guidance is expected to publish sometime in 2004. This cautious approach reflects both the FDA’s stated inexperience with techno-logy like microarrays and the technology’s rapid development: “We are not sure that we can get it right in a single draft, partly because things are changing so fast right now,” Mansfield said.
The guidance is intended for tests the FDA classifies as “high risk,” such as a cancer diagnostic test, or “moderate risk,” such as a cystic fibrosis test, Mansfield said. These tests require so-called premarket approval (PMA) or 510(k) submissions, respectively. However, the guidance is not a regulation, but merely a set of recommendations on how to prepare a submission, with advice on matters such as what information to include, or what clinical population to study. “If you have another way of showing that your device is safe and effective, that may be acceptable as well,” said Mansfield.
As of this week, only one company — a diagnostics firm — had submitted a comment, but “the expectation is that we will get more comments as the deadline nears,” Mansfield said. However, “we have had limited interactions with companies that are seeking protein or proteomics applications,” she added.
One of the reasons might be that technology used to measure protein biomarkers, such as protein arrays, is still at an early stage of its development compared to DNA microarrays and similar technology. “It’s not in its infancy anymore, but it’s not yet totally mature,” said Michael Natan, vice president of business development at Mountain View, Calif.-based SurroMed, which uses both mass spectrometry and multiplexed immunoassays to study protein biomarkers. “Maybe in a couple of years, it might make sense to start issuing guidances [for proteomics],” he said. There is some overlap, he acknowledged — for example regarding analysis tools used to interpret both gene chip and protein chip data — so the guidance may become more relevant as protein arrays mature.
But the current draft guidance does not deal with specific issues of protein analysis technologies, such as mass spectrometry, or quantitative measurements of proteins. As a result, SurroMed is not planning to comment: “It’s just too early,” Natan said.
This view is echoed by other proteomics players. “In proteomics, people are just seeing the first prototypes emerge. Clinical experience is extremely limited at best,” said Robert Maurer, vice president of business development at Fremont, Calif.-based Ciphergen, which uses chip-based mass spectrometry to analyze protein biomarkers. Like SurroMed, Ciphergen does not intend to comment on the draft guidance. Rather than relying on a guidance that does not fit its approach, “our strategy is to wait and see what develops with respect to our tests and then go in and have a dialogue directly [with the FDA],” Maurer said.
This strategy is a viable alternative, according to the agency. “We encourage companies to come and talk to us first, so they don’t waste time and money,” said Mansfield, acknowledging that it might be difficult to cover all of the protein analysis technologies in a guidance. However, she encourages protein biomarker companies to raise their voices now: “If somebody thinks that we need a separate proteomics guidance that specifically goes over things that are only relevant to proteomics, we would like to know that,” she said.