By Nancy Weil, IDG News Service
January 12, 2004 | The challenges confronting the healthcare industry in translating human genome knowledge into clinical practice will require a cultural shift that could take decades to achieve. That was the bleak message from a two-day MIT symposium on “The Future of the Pharmaceutical Industry.”*
"You really have to believe passionately in this concept of personalized medicine to bang your way through the challenges," said Donald Anderson, global head of pharmacogenomics and clinical affairs at Aventis Pharmaceuticals.
For personalized medicine to become reality, much has to happen to bring about the necessary shift. Doctors have to be trained in genetics, even though such training is a low priority at most medical schools, said Glenn Miller, scientific director at Genzyme Genetics. "It may, in fact, take a full generation," before genetics becomes part of clinical practice, Miller said.
The media hype that accompanies genetic research creates its own problems. "Every day, there are patients showing up at doctors' offices with stacks of information from the Internet … saying, 'I want this test' or 'Here's how I want my baby delivered,’" Miller said. Patients are driving the transformation because many want genetic testing and medical care based on genetics.
Aside from education issues, several speakers pointed to a lack of IT tools, including the need for computational models and applications that perform specific functions such as genome analysis. These applications often have to be developed by individual researchers who cannot find what they want on the commercial market.
Even more vexing from an informatics perspective is that reams of data on everything from patient histories to clinical trial results are spread globally on disparate databases that cannot communicate with each other. Vast amounts of data are still in paper form. Even if electronic data could be easily accessed, the task of organizing and mining such data is perturbing, complicated further by HIPAA privacy regulations.
The promise of systems biology, which uses quantitative and experimental methods to produce data that can be mined, has met with limited enthusiasm from the pharmaceutical industry, still recovering from genome hype and over-investment. The systems approach may be more rapidly embraced among academics and at research institutions, but it has to take hold in the pharmaceuticals industry for pharmacogenomics to become widespread.
One example of information-sharing using a Web-accessible database is the Shared Pathology Informatics Network (SPIN), a five-year project begun in 2001 by the National Cancer Institute with the aim of using informatics to develop a virtual database for sharing and locating human tissue specimens. Hospitals can join SPIN and then search the database for information, said Martha Gray, director of the Harvard-MIT Division of Health Sciences and Technology.
Each participating hospital, including institutions in Boston, Los Angeles, and Pittsburgh, maintains its own (anonymized) information. While this network is a step in the right direction, "there are really significant challenges to overcome so that we can say with a straight face that we have realized the promises [of the genome project],” Gray said.
On the positive side, the National Institutes of Health, which Gray said serves as "the steward of medical and behavioral research," understands that the transformation in healthcare is cultural and requires various disciplines to integrate.
Of greater concern for pharmaceutical companies is whether the FDA understands this. Janet Woodcock, director of the FDA’s Center for Drug Evaluation and Research (CDER), acknowledged the suspicions evident among audience members. The agency has issued draft guidelines for pharmacogenomics regulation. "It's very important that scientists at the FDA develop an understanding of this field," she said of pharmacogenomics. In return, the agency pledges that such information will be cloaked in anonymity.
Woodcock also expressed concern that doctors "are not trained well in drug metabolism," which results in their being insufficiently informed about toxicity issues.
Drugs developed through genetics help some patients and not others (see The Road to Personalized Medicine, Dec. 2003 Bio-IT World). The better clinical trials are, especially early on, the more researchers are able to identify populations that can most benefit from drugs. Several speakers stressed the need for better clinical trials, notably Phase I. This would benefit people earlier, keep people who would be adversely affected out of trials, and dramatically curb costs.
The most abiding concept to emerge from the symposium was the need for collaboration in order to streamline drug discovery and pave the way for pharmacogenomic advances. Major pharmaceutical companies have the money to develop expensive new drugs; IT tools must be developed and offered commercially; insurance companies must accept innovations; and the FDA has to help make the process easier.
"We all have to work together on this," said Aventis’ Anderson of drug development and the future of pharmacogenomics. "That's the challenge."
*The Future of the Pharmaceutical Industry (MIT): Cambridge, Mass., Dec. 9-10.