Regulatory Beat: Pharmacogenomics Discoveries Shape Biotech R&D

Published on: 
BioPharm International, BioPharm International-10-01-2004, Volume 17, Issue 10

Even if FDA does not require pharma companies to co-market a diagnostic test, insurers and health plans may pressure industry to do so, according to Robert Temple of CDER.

New knowledge of genetic biomarkers for disease is spurring development of pharmacogenomic (PG)-based drugs and biological products and raising new challenges for FDA and manufacturers. Genomics information promises to spur discovery of more individualized treatment regimens by identifying patients most likely to respond to certain medicines — as well as individuals more susceptible to adverse events.

Manufacturers can use this information to select high-response individuals for clinical trials, an approach that promises to reduce clinical research costs, accelerate drug development, and promote appropriate and safe use of complex new treatments.

FDA is eager to encourage more industry use of pharmacogenomic data in drug development in order to achieve these benefits. Last November (2003), the agency issued a draft guidance designed to prompt manufacturers to voluntarily share with FDA internally developed PG information that can help the agency and the research community keep abreast of genomics developments and establish new genetic biomarkers. Most companies, though, are reluctant to submit this data to FDA for fear of raising questions that could delay product approval or restrict labels.

The guidance, which FDA plans to finalize by next year, recommends that sponsors send in voluntary genomic data submissions (VGDSs) for analysis by a special FDA review group, separate from the established new drug review process. Some companies already are sharing PG data with FDA, and the agency is clarifying procedures to encourage other firms to do the same.



Manufacturers may be reluctant to share PG data with FDA, however, if they fear that information on new biomarkers might prompt the agency to require a diagnostic test in order to bring a new treatment to market. The discovery of more genomic markers that can identify high- and low-responding patients is prompting regulators to seek diagnostic tests to determine which patients should be treated with a therapy, the appropriate dose, and the risks associated with treatment.

FDA acting Deputy Director Janet Woodcock is spearheading an effort to issue guidance clarifying how FDA will regulate diagnostics and pharmacologic interventions that traditionally are considered separate entities for regulatory, reimbursement, and clinical practice purposes. Different FDA centers regulate medical devices, drugs, and biologics, but the growing number of medical products that involve multiple components prompted the establishment of a central Office of Combination Products (OCP) two years ago. While some FDA staffers would like to require the use of a diagnostic for any therapy with an identifiable biomarker, manufacturers seek a more flexible approach that recognizes significant differences in the processes and procedures for developing diagnostics as compared to drugs and biologics.


Manufacturers and FDA officials discussed these issues and options for policy development at a July workshop on "Co-Development of Drug, Biological and Device Products" co-sponsored by FDA and the Drug Information Association (DIA). While Woodcock believes that codevelopment provides a real opportunity to improve medical outcomes, the overriding concern for manufacturers is that FDA might require a diagnostic test in order to gain approval of a new therapy.

Industry representatives maintained that PG therapies and diagnostics should not be regarded as combination products, noting that such products usually are developed and brought to market separately. Lois Hinman, director of regulatory affairs at Hoffman-LaRoche referred to linked therapies as "partner products" instead of combinations. "Flexibility is key," Hinman said; she is looking for FDA to develop guidance that describes a range of codevelopment scenarios.

Manufacturers also insist that any codevelopment arrangements involving drug and diagnostic firms must be voluntary. Industry wants FDA to present substantial evidence of medical need and of a diagnostic's ability to improve safety or efficacy in a target population before requesting diagnostic codevelopment. And both FDA and manufacturers desire a clear process for undertaking such codevelopment, including a schedule of critical meetings and clarification of which FDA offices will be involved and how they will coordinate their work.

Woodcock recognizes that it would be difficult for FDA to require codevelopment of drugs and diagnostics. But if a therapy such as Genetech's Herceptin (trastuzumab) works only for certain patients, she believes the manufacturer needs to identify that subgroup (see "Herceptin Provides Codevelopment Model"). Although Woodcock sympathizes with industry's concern that codevelopment requirements could increase the cost and time involved in bringing a new therapy to market, she maintains that diagnostics will be key to launching more targeted treatments and that industry can benefit greatly from such approaches. If a manufacturer can identify the genetic factors related to a product's efficacy, for example, the firm may be able to get very high response rates from very small clinical trials, she explained. And the ability to test for a genetic pre-disposition to serious adverse events can eliminate those individuals from studies and treatment and enhance a product's safety profile. Such approaches may allow FDA to approve some therapies that otherwise might never reach the market.

Even if FDA does not require pharma companies to co-market a diagnostic test, insurers and health plans may pressure industry to do so, added Robert Temple, director of CDER's Office of Medical Policy. For very expensive therapies, he observed, payers will want a diagnostic to better target such treatments to the most appropriate patients.

Woodcock emphasized how important it is for FDA policy makers to "get this right." These issues are slated for further discussion at an industry workshop on PG issues scheduled for April 2005. FDA and the Pharmaceutical Research & Manufacturers of America held meetings on this topic in May 2002 and November 2003 and recognize that another session is warranted, particularly when FDA unveils a final guidance on VGDSs and a draft guidance on drug-diagnostic co-development.

Herceptin Provides Codevelopment Model

A prime example of the issues involved in drug-diagnostic codevelopment is the process followed by Genentech in bringing Herceptin (trastuzumab) — a monoclonal antibody that treats about 25% of breast cancer patients with a genetic abnormality leading to HER2 protein overexpression — to market. FDA and Genetech recognized during product development that appropriate treatment required a diagnostic test to identify HER2 positive individuals. Genetech devised an assay for selecting patients for clinical trials, but to gain market approval, FDA wanted the firm to provide a HER2 diagnostic test suitable for commercial use. Genetech initially partnered with DAKO to provide such a test; the two manufacturers simultaneously filed applications for coordinated use of the drug and the diagnostic and gained FDA approval in late 1998.

After Herceptin was on the market, though, Genentech found that the medical community remained uncertain about when it was appropriate to test patients for HER2, explained Cheryl Madsen, Genentech senior manager for regulatory affairs. Also, further research found that a fluorescence in situ hybridization (FISH) test could better select those patients who could clearly benefit from Herceptin. Genetech partnered with the diagnostic firm Vysis to gain FDA approval of a new FISH diagnostic and applied to FDA to include FISH testing as an alternative.

Diagnostic manufacturers are developing more tests to identify suitable participants for trials of new cancer therapies and other drugs. Such collaborations are most successful when they begin early in the drug development process, pointed out DAKO Clinical Affairs Director Gretchen Murray. This allows drug manufacturers to incorporate diagnostic statistical requirements into development programs and to ensure that clinical study specimens are banked and stored appropriately.

Jill Wechsler is BioPharm International's Washington editor, 7715 Rocton Avenue, Chevy Chase, MD 20815, 301.656.4634.