BOSTON - Getting investors and health insurance providers to buy into the personalized medicine business will require more reliable diagnostics and more impressive collections of data, according to a panel session at the Biotechnology Industry Organization's annual meeting here.

With the completion of the Human Genome Project, the concept of personalized medicine - usually defined as a customized therapy based upon an individual patient's disease and genetic makeup - has kept the life sciences world buzzing. Yet the true potential of tailored drugs and treatments has been a little harder to pin down. Gathering in a meeting room in the Boston Convention & Exhibition Center, a handful of industry experts reviewed some of the challenges of developing personalized medicine.

Despite speculation as to its potential, all four panelists agreed that being able to create "one pill for one patient" remains unlikely, though they believe the ability to create more targeted drugs aimed at smaller patient populations are within reach, pending the development of diagnostic tests to identify those patients. To date, there have been a few successes in that area - most notably South San Francisco-based Genentech Inc.'s Herceptin (trastuzumab), which targets HER2-positive breast cancer.

The use of a genetic test to identify those patients helps both with treatment and in reducing the costs of Herceptin therapy per patient, but Paul Keckley, executive director at Deloitte Center for Health Solutions, argued that if the Herceptin story is the best financial case "we can make for personalized medicine, then we've got a long way to go."

Isaac Kohane, as associate professor of health sciences and technology at Harvard, agreed. "Herceptin is the example everyone uses, but really it's the only one." While there have been genetic markers identified for a number of other diseases, such as ApoE for Alzheimer's disease, they are most often used in disease prognostication and classification, rather than for diagnostic purposes, he added.

The best way to promote the use of genetic testing probably is to develop a diagnostic product in tandem with a drug, said Stephen Eck, vice president of global R&D molecular profiling at Pfizer Inc., which is in registrational trials with the CCR5 antagonist maraviroc in HIV patients. Patient enrollment in those trials involves the use of the HIV Co-Receptor Tropism Assay from South San Francisco-based Monogram Biosciences Inc. New York-based Pfizer licensed global rights to the assay last May, the same time it agreed to make a $25 million investment in Monogram. If Pfizer succeeds in getting maraviroc approved, it is likely that Monogram's assay would continue to be needed for patient treatment. (See BioWorld Today, May 9, 2006.)

Having companion diagnostic products will "not just tell you who should get the medications, but also who shouldn't get the medications," Eck said, adding that that precision in treatment will help lower costs for individual patients.

But whether it's possible for a diagnostic to be developed in conjunction with every drug in the pipeline is doubtful, he said, owing to the fact that the mechanisms of some disease pathways might not be understood well enough to create a companion diagnostic test. For example, Eck said, the melanoma drug MDX-010 (ipilimumab) from Princeton, N.J.-based Medarex Inc. and partner Bristol-Myers Squibb Co., of New York, which is in Phase III testing, has shown promising efficacy so far, though it's unlikely researchers will be able to pinpoint precisely how the drug acts before analyzing data from pivotal studies.

That brings up another barrier to personalized medicine: the collection of data that would both support development and help gain confidence from investors, physicians and consumers. Right now, Harvard's Kohane said "health care systems and EMRs [electronic medical records] are woefully inadequate," and it's difficult gathering data from small populations without having to go across different regions. "We need to find a way to share data."

While all panelists agreed on the need for more compelling data, there was some difference of opinion on how far the industry has advanced in reaching the personalized medicine goal. Mara Aspinall, president of Genzyme Genetics, said she believed "we have just passed the tipping point," with today's clinical trials collecting genetic material for further testing and the generation of substantial biomarker data. But Kohane argued that "we're not at that point yet" in terms of "big changes in treatment."

Eck said the underlying science will have to progress a little before personalized medicine can become a reality. Right now, "we're knee deep in the human genome trying to determine what that means," he said.