Randy Scott, a biochemist who previously co-founded another genomics company, Incyte (NASDAQ: [[ticker:INCY]]), was motivated to start something new— in part because he saw that all the genomic information he was helping produce couldn’t do much for a friend with cancer.
With early support from Kleiner Perkins Caufield & Byers, Versant Ventures, and TPG, among others, he and Genomic Health’s chief medical officer Steve Shak were off and running. They hit on the idea of obtaining breast cancer tumor samples that in some cases were 20 years old, and for which the ultimate outcomes of patients were already known. They looked at the tissue samples to see which genes were active, and how that was connected to the better or worse outcomes. The study was doable at a company because it didn’t require taking 20 years to run a clinical trial and wait and around to see results. Breast cancer was an obvious choice because it’s a big market—about 207,000 women in the U.S. are diagnosed with the condition each year, according to the American Cancer Society.
The company whittled down its original list of 250 candidate genes to a standard 21-gene test. The test was introduced on the U.S. market in January 2004, offering women and doctors a chance to see if their tumor was the kind that had a low, intermediate, or high risk of relapsing in the future.
Early on, the company realized that the information was useful, but not really enough, Popovits says. More work needed to be done to find out which patients were likely to benefit from chemotherapy. For example, a patient might have a low risk of recurrence, say 7 percent, but if preventive chemotherapy could drive that risk rate down to 1 percent, some women and doctors might opt for the chemo anyway. Data on that question, which made the findings more “actionable,” was presented at a medical meeting in December 2004, Popovits says. “Now all of a sudden the story came together,” she says.
That helped with the doctors, but none of it meant anything to the business until insurers got on board. Genomic Health had already spent $100 million on R&D for the test by the time it was introduced on the market, and had decided it was going to charge for the test based on the health economic value it delivered to the system, rather than just a slim mark-up beyond the cost of raw materials, and marketing. This was a radical notion.
“When you look at how diagnostics are reimbursed today, it’s activity-based. It’s the same as if you went to Genentech and said ‘we’re going to pay you for Avastin based on the number of steps you have in your manufacturing process.’ That’s how diagnostics work. We said that’s not going to work,” Popovits says. “We put $100 million plus into developing these tests. They will not be sold at $300 per test.”
This took a lot of convincing among private insurers. It took a lot
