By Malorye A. Branca
November 15, 2002 | To help pharmacogenomics gain smooth passage into the clinic, industry representatives are banding together to develop data standards.
James Freeman, a bioinformatics consultant at Biogen Inc., and Paul DiBello, a software development group leader at Genaissance Pharmaceuticals Inc., lead the Interoperability Informatics Infrastructure Consortium that is spearheading the effort. This working group, known as I3C, is an industry-driven, nonprofit organization trying to develop standards for data exchange and interoperability in the life sciences. The group’s aim is to find a model that can facilitate pharmacogenomic data exchange between companies, clinical laboratories, and the FDA.
Such standards could not only speed approvals but also many basic research processes. The standards would cover genotypes, haplotypes, phenotypes, correlations between these, and their relation to clinical outcomes. “I3C has a powerful process in place for developing standards,” Freeman says.
The working group first gathers use cases, then designs requirements, and finally decides upon a reference implementation that satisfies the requirements.
Freeman has already found one good reference model in the form of Visible Genetics Inc.’s application for its TRUGENE HIV-1 Genotyping Kit and OpenGene DNA Sequencing System. This was one of the first pharmacogenomic tests approved, clearing the FDA in September 2001.
The TRUGENE test was reviewed by the Center for Biologics Evaluation and Research because it is an HIV test. It was approved as a completely integrated device, including chemistry, hardware, software, an interpretive patient report, and training certification and service for the technicians. The company can only sell to facilities in which technicians have been trained and certified by Visible Genetics.
Pleasing the FDA will be the trickiest part of developing standards. “The FDA will not want to pick winners or losers,” says Freeman. The agency may provide input, but it is unlikely to officially “bless” any particular format.
“We are in a very preliminary stage, but we have a great group of people involved,” says Tim Clark, chairman of I3C and vice president of informatics at Millennium Pharmaceuticals Inc. The I3C working group includes representatives from Genaissance, Biogen, the Clinical Data Interchange Standards Consortium (CDISC), Health Level Seven, Eli Lilly, and several clinical research organizations. Clark says there is also an effort to get some FDA representatives on board.
“Today, clinical laboratories support 1,300 to 1,500 different interfaces,” says Rebecca Kush, president of the Austin, Texas-based CDISC, a nonprofit collaboration working on standards for electronic submissions. “Often they are asked to use different formats even for data from the same trial.”
The group has just published Version 1.1 of their Clinical Lab Data Standards (see Clinical Data Standards Face Uphill Battle, page 22), but those do not include standards for pharmacogenomic data. “The team wants to add that,” Kush says, citing the main driver behind CDISC’s interest in seeing I3C pursue this project.
“Today, companies have to be able to function as an extended enterprise,” Clark says, “and to do that you have to have an electronic interface. Moving forward, pharmacogenomic and genomic data will be some of the key data being exchanged.”