Personalized Medicine Not Paying The Bills Yet; More Science Needed

Though the promise of personalized medicine has been widely touted, firms focusing on targeted therapies and companion diagnostics are not getting much bang for their buck - yet.

"The potential is there for more tailored and targeted therapies," said former FDA commissioner and CMS administrator Mark McClellan, now the director of the Brooking Institution's Engelberg Center for Health Care Reform. "But the number of products and the value of products coming out of the pipeline is far below what it's been in the past. The potential remains unfulfilled."

McClellan spoke June 20 about the need to build scientific evidence to support personalized medicine applications before a diverse audience of health care professionals and policymakers at a personalized medicine conference in Washington, D.C., hosted by the American Association for the Advancement of Science and the Food and Drug Law Institute.

So far, the actual impact of personalized medicine, and specifically pharmacogenetics-based therapies that predict a patient's response to drugs based on genetic information, has been limited, he said.

Companion Dx In Drug Label - Now What?

Last August, for example, FDA updated the labeling of the anti-clotting drug warfarin to include information about the usefulness of genetic testing, as specific gene variations are associated with lower tolerance (¹ (Also see "Warfarin Label Change Falls Short of Recommending Genetic Testing" - Pink Sheet, 20 Aug, 2007.), p. 8).

That should have been a turning point for firms such as Nanosphere, whose Verigene warfarin genetic test kit was cleared in September 2007, and AutoGenomics, Nanogen, Third Wave Technologies and Luminex, who sell warfarin genetic marker reagents and are developing test kits.

Unfortunately, uptake of the tests has been tepid.

"At FDA it was a very big deal for us to overcome a lot of barriers to put that into the drug label and also for [the Center for Devices and Radiological Health] to have it in the labeling, but it's not used," commented Shiew-Mei Huang, deputy director of the Office of Clinical Pharmacology in FDA's Center for Drug Evaluation and Research. "It's developed, it's in the label and nobody's using it. I think we need to have [more] communication with health care providers."

The warfarin tests are struggling for several reasons, Third Wave CEO Kevin T. Conroy suggested. He said the labeling is not strong enough, and clinical guidelines do not incorporate warfarin testing.

"And there's not enough data," he conceded.

Third Wave markets a companion diagnostic that tests response to the cancer drug Camptosar , the Invader UGT1A1 molecular assay, cleared in 2005. It also sells analyte specific reagents that detect variations in the CYP450 gene family, including the two variants associated with warfarin response.

Huang said CDER has approached insurance providers to find out why they are not covering the warfarin tests, and they respond that there is not sufficient evidence backing up the technology.

Steve Gutman, director of FDA's Office of In Vitro Diagnostic Device Evaluation and Safety, agrees the scientific evidence supporting targeted therapies is simply not up to par.

"I don't think it's the damned regulation that's holding up the field," he said. "I actually don't even think it's the damned reimbursement, although it is problematic. I think it's the damned science."

For its part, pharmacy benefit manager Medco has been taking a leadership role in gathering evidence and convincing payers to adopt personalized medicine. Medco recently hired former FDA genomics guru Felix Frueh to help develop its Personalized Medicine Research Center (² 'The Pink Sheet,' May 19, 2008, In Brief).

Medco has been quite involved in building support for the warfarin tests specifically. The firm began collaborating on a warfarin research project with the Mayo Clinic in late 2006. According to the PBM, many of its health plan clients are interested in the warfarin testing (³ (Also see "Warfarin Dosing Issues Emerge In Early Results From Medco/Mayo Research" - Pink Sheet, 26 Nov, 2007.), p. 11). Medco also has a partnership with LabCorp on genetic testing for tamoxifen.

It's All About The Evidence

The key to launching a successful companion test is pinpointing a specific link between an assay and a tangible application, McClellan and others at the meeting stressed.

"Diagnostics are not just for the sake of diagnostics anymore," Gutman said. "They can be used in the context of drug candidacy to improve development and to perhaps improve use."

Finley Austin, head of U.S. external research and innovation at Roche, said that it is crucial to have evidence validating not only the biomarker's and the assay's prognostic ability, which refers to its success in identifying likely clinical outcomes independent of therapy, but also its ability to predict response to therapy.

"That's what I'm thinking about when I'm developing a test and taking it to the FDA, because if I don't have the appropriate trial data, then I'm going to get a prognostic claim in the label, only," she said at the AAAS/FDLI meeting.

"Aetna's then going to tell me, 'I don't want to pay for this because it's not changing anything. We already know these patients have a poor prognosis.' Without that predictive element to it ... you don't have a lot of value to offer to patients."

Roche, which has played up its potential for leadership in the personalized medicine space, based on its heritage as both a drug and diagnostic manufacturer, is a prime example of a firm that has learned that biomarker tests without specific and predictive therapeutic claims do not sell.

The firm's AmpliChip CYP450 microarray test, which analyzes variations in two genes that play a role in the metabolism of about 15 to 20 percent of drugs, was approved and launched in 2005 without a specific claim, said Walter Koch, head of research at Roche Molecular Diagnostics.

"FDA said these are valid biomarkers, they're used in drug development today and they impact so many drugs that they assumed physicians would understand how to apply it to a wide range of drugs," he said in an interview. "As it turned out, it's been complicated for physicians to understand exactly how to employ it."

Koch says Roche got the message: "One needs to be much more specific in linking the diagnostic to a therapeutic decision."

The test is primarily used to identify patients who are not likely to respond to tamoxifen treatment for breast cancer. Roche is building the clinical evidence to seek that specific claim, Koch said.

Roche recently unveiled some changes to its R&D organization - such as the addition of "diagnostic liaison managers" - to further facilitate side-by-side development of diagnostics paired with targeted drugs (⁴ (Also see "Roche Makes Early Headway With Pharmaco-Diagnostics" - Pink Sheet, 30 Jun, 2008.), p. 15).

It is this type of approach that personalized medicine supporters like McClellan hope can spread to the broader health industry.

"If there are clearly demonstrated ways of getting safer and more effective treatment to patients - our policies may take a while to figure out, we may limp along ... but we will figure it out," he said. "It is a very challenging environment, but the potential payoff is tremendous."

[Editor's note: This story appears courtesy of the editorial staff of ⁵ "The Gray Sheet" , your source for coverage of devices and diagnostics. For a sample copy, call customer service at 800-332-2181.]

- Jessica Bylander ([email protected])