to treat the malignancy, and help influence whether they should get chemotherapy or not.
Diagnostics are traditionally considered cheap commodities, but when Genomic Health and a competitor called Agendia came along, they needed to build a mountain of evidence to convince doctors that the test provided valid, useful information. Insurers—tired of hearing lots of grandiose claims that often lack substance—balked at the price. They required companies to perform long complex analyses that essentially show that spending several thousand dollars on a diagnostic test like this can save more down the road on costly and unnecessary rounds of chemotherapy and other aggressive treatments.
Cowens’ job at Genomic Health was, as Gray said, to build up the pipeline of tests after the original hit in breast cancer, to diversify with predictive tests for relapse of colon, prostate, and kidney cancers. But in his last year at the company, Cowens really focused on the critical question of health economics, particularly trying to amass evidence to convince payers in Europe that the Genomic Health test was worth the money. It’s often a line of thinking that companies worry about after they pour energy into generating clinical trial results to win over doctors. At NanoString, Cowens says he hopes to incorporate this broader worldview into the company’s strategy right at the moment it’s designing clinical trials to prove its device works as a diagnostic tool.
“You have to ask the question at the very beginning, ‘will you have the chance to prove that the test is cost effective before you even invest in it?” Cowen says. “One of the things we learned early [at Genomic Health] was that having forethought on health economic analysis is as important as having forethought in your clinical analysis. They are intimately intertwined.”
NanoString, as I reported here in early December, is seeking to make its case for a new breast cancer test that looks for patterns in a set of 50 different genes, known as the PAM50 signature. This is too many genes for other machines to analyze efficiently, but it’s thought to be a meaningful measurement of a patient’s prognosis, and it is something NanoString believes its machine can do simply and reliably. The company, now with Cowens’ input, is working on crafting the clinical trials that it thinks it will need to win over doctors and insurers. What’s interesting about NanoString, Cowens says, is that since it’s early in the game, it’s in a position to lay the groundwork now to win over both vital constituencies.
The plan is to gather data for the PAM50 diagnostic and seek FDA approval of its device as a diagnostic tool. Then Cowens will think hard about how to use this gene expression tool to come up with new diagnostic tests for other types of cancer. And from the get-go, the idea is to prove that it’s not only useful but cost effective in this era of cost-containment.
“Health economics is part of the game at the conception of the pipeline, it’s not an afterthought that you try to graft on later,” Cowen says. “Clinical data needs to be generated to influence physicians to change their behavior. And the health economics should be attractive to the payers. That’s a high bar for a new test to pass, but one could argue that it’s an appropriate bar to clear in this business.”
