May 1, 2006 (Vol. 26, No. 9)

Firm’s Normal Reference Gene Expression Values Track Patients’ Molecular Responses

The new molecular diagnostic tools created at Source MDx ( in Boulder, CO, are bringing personalized medicine closer to reality and advancing the new field of theranostics, or companion diagnostics. Source reportedly holds the first-known intellectual property claim for a normal range of gene-expression values. Awarded in November 2005, the patent lays the foundation for the use of RNA-based molecular diagnostics to monitor an individuals disease status and response to therapy based on the Healthy Normals Reference Dataset.

The founders of Source worked previously at Amgens Boulder facility, where they used molecular tools to create new drugs. When the drugs passed into clinical development, however, &#8220we had to rely on subjective clinical end points that are not sensitive enough to track molecular responses, critical for determining a drugs dose and schedule,&#8220 explains Dave Trollinger, vp and director of molecular technologies.

Trollinger, Danute Bankaitis-Davis, Ph.D., and others co-founded Source in 1998. The companys goal is to apply its molecular methods to clinical development and patient care. The Source approach analyzes gene expression in whole blood, which reflects systemic processes associated with disease and response to drug treatment.

When Source was founded, the mantra was to look at changes in patient status over time. However, &#8220we wanted to establish a normal reference range of gene expression in humans,&#8220 says Trollinger. Then those normal values could be used to determine whether patients with active diseases fall inside or outside the range, and whether a drug returns them toward normal values.

A major hurdle was the improvement of quantitative PCR methods to reliably and reproducibly measure RNA from whole blood for valid molecular monitoring. RNA must be amplified prior to analysis, but traditional amplification techniques distort the original proportions existing among genes. &#8220Very rarely will one gene tell you what you need to know. It takes three to twelve genes for reliable results,&#8220 says Karl Wassmann, CEO.

Hence, scientists at Source designed primers and probes that do not affect gene proportionality. Sources platform relies on precise, calibrated measurements of the relationships among genes that are analyzed by multigene algorithms. For instance, not only can gene A be compared before and after a drug treatment but it also can be compared to gene B that is known to be three times more abundant than gene A in healthy individuals. &#8220The proportional relationship between gene A and gene B remains accurate in our measurements,&#8220 says Wassmann.

After the highly sensitive and reproducible techniques for measuring genes in RNA were established, they prompted basic questions: How homogenous is the normal healthy population? and How do genes vary in certain disease states? &#8220In order to characterize that we had to understand whether there is a normal reference range for gene expression,&#8220 says Trollinger.

Source thus selected more than 100 representative genes involved in inflammation and immune responses that are linked to arthritis, cancer, heart disease, and other common illnesses. In whole blood collected from hundreds of healthy blood donors, Source discovered that these genes show a narrow range of expression, and that their activity does not vary significantly with age, race, or gender. The Healthy Normals Reference Dataset comprises the normal expression levels of these representative genes.

Finding the Right Medicine

Sources normal reference values are the basis for molecular tests to diagnose disease and evaluate response to drug therapy. &#8220We want to help find the right medicine for the right person at the right time,&#8220 says Wassmann.

Some biological therapeutics, such as TNF-alpha inhibitors for RA, cost up to $15,000 annually, yet one-third of all patients do not respond to therapy. Wassman and Trollinger envision $100 blood tests that will screen patients for potential effectiveness and adverse events. &#8220The idea that blockbuster drugs will help everyone with a disease is no longer accepted by the medical community.&#8220

Source scientists identify subsets of genes to finetune disease diagnosis. The company has patents pending on more than 1,000 genes, covering 30 diseases. Although microarrays may reveal that 200 genes are up- or down-regulated in a disease, &#8220the medical community is not attracted to this type of signature profile of a few hundred genes,&#8220 says Trollinger. In contrast, the process at Source centers on a few well-characterized genes.

For instance, a panel of just five genes is highly informative in discriminating multiple sclerosis (MS), a disease that often goes undiagnosed for years because of the lack of tests. &#8220We saw biological activity in 100 genes in MS, but our statistical methods narrowed it down to five key genes,&#8220 says Wassmann. In fact, &#8220many diseases can be narrowed down to fewer than 10 genes,&#8220 he adds.

Sources R&D services business provides high-precision Biomarker Panels, which are focused sets containing six to 96 genes and 20 patented disease and pathway-specific Biomarker Panels, to pharmaceutical companies for use in drug development. The disease-specific panels include breast cancer, diabetes, lupus, and MS, whereas the pathway-specific panels include apoptosis, chemokines, and inflammation. In addition, custom-designed gene panels are available.

Theranostics Pioneer

The technology at Source supports companion diagnostics, which are encouraged by the FDA. Largely due to the problems surrounding Vioxx, the FDA issued guidelines regarding new drugs and appropriate diagnostic tests to screen patients for drug safety and efficacy. Theranostics involves not only the diagnosis of a disease but also choosing the correct treatment regime and monitoring patient response to therapy. The development of companion diagnostics should start during Phase I testing, and Sources patented molecular tools can help companies create them.

The normal reference values also can track individual therapeutic responses at the gene level. This will improve patient care by quickly determining whether a drug is effective, then guide personal adjustments in dosage and schedules.

Assays based on precise, quantitative measurements of gene expression will expand treat-to-normal protocols. This approach, in which doses of drugs are adjusted based on clinical outcomes, effectively controls hypertension and diabetes. Treat-to-normal molecular diagnostic practices can be applied to cancer, infectious diseases, arthritis, and other autoimmune disorders.

Previous articleDr. Leroy Hood M.D. Co-founder of the Institute for Systems Biology in Seattle, Washington talks with GEN Editor-in-Chief John Sterling about Systems Biology.
Next articleDr. Raju Kucherlapati Scientific Director, Harvard-Partners Center for Genetics and Genomics & Paul Cabot Professor of Genetics, Harvard Medical School