FDA should employ more realistic timelines for drug/diagnostic co-development when it crafts draft guidance on the topic, stakeholders argue in written comments to the agency.

The remarks were submitted in response to a CDRH/ CDER concept paper published in April as a precursor to draft guidelines (¹ Also see "Drug/Diagnostic R&D Guidance Will Accommodate Biomarker Data Diversity" - Medtech Insight, 18 April, 2005., p. 11).

The document is one of several recent FDA releases that aim to clarify regulatory requirements in the fast growing areas of pharmacogenomics and personalized medicine. Others include CDER's March final guidance on pharmacogenomic data submissions, and an April 2003 CDRH draft guidance on multiplex DNA tests (² Also see "FDA Genetic Array Submissions Guidance Dovetails With NCCLS Effort" - Medtech Insight, 4 October, 2004., p. 22).

The concept paper envisions a parallel process of drug and assay development in which a test would be established and analytically validated during Phase I and Phase II drug studies, and clinically validated in Phase III trials.

However, AdvaMed and the Pharmaceutical Research and Manufacturers of America (PhRMA) both maintain, in separate comments, that although this is an ideal situation, it is "rarely the case."

The trade groups explain that the data necessary to qualify a biomarker or determine the need for a diagnostic that will help predict drug response is typically not available until late in Phase II or even Phase III trials.

"To have a more significant practical impact, the eventual guidance should address co-development involving a new drug and diagnostic test in which diagnostic test development may not be realized until late during the drug development," AdvaMed says in ^FDL-1 July 15 comments.

Roche agrees. In ^FDL-2 July 8 comments based on input from its diagnostic and drug units, the firm urges FDA to explicitly acknowledge the significantly shorter development times for in vitro diagnostics compared to drugs.

This will make the draft guidance "more realistic and flexible," says Roche.

Roche manufactures the Amplichip CYP450 , which predicts therapy response by measuring drug metabolism gene mutations (³ Also see "Regulatory News In Brief" - Medtech Insight, 14 March, 2005., In Brief).

Stakeholders are also concerned about potential evidence standards for "co-development" diagnostics. The FDA concept paper states that "normally data from two or more adequate and well-controlled trials would be collected to confirm clinical effectiveness, thereby establishing clinical utility of the test."

The paper does acknowledge that "the amount and extent of clinical trial data to verify the clinical utility of a test will differ" from case to case, depending on various factors.

The two-trial standard "presents a major burden for the diagnostic partner," Roche says. Although typical for drug development, "this is not a requirement for a PMA, and this is an excessive request," the firm contends.

Roche notes that additional studies for diagnostics "are commonly considered as post-approval requirements."

At the annual meeting of the American Association for Clinical Chemistry in Orlando July 26, a Roche Diagnostics rep asked FDA Office of In Vitro Diagnostics Evaluation & Safety Director Steve Gutman whether the evidence requirements might be satisfied with post-marketing conditions.

Gutman, however, expressed concern that this approach would not distinguish clearly enough between approved co-developed tests and investigational devices.

"We owe the patients...and the physicians in this country transparency and honesty, and we should not pretend that investigational products are actual products," Gutman said. "It is disingenuous."

CDER Office of Clinical Pharmacology & BioPharmaceutics Director Lawrence Lesko said at the AACC meeting that he hopes the co-development draft guidance will be out by the end of this year.

AACC, Genzyme and GlaxoSmithKline also submitted comments to FDA on the concept paper.

Industry Seeks Changes In FDA Drug/Diagnostic Co-Development Model

More from Archive

More from Medtech Insight