 |
| MIT's Mark Trusheim says the industry needs to rethink its approach to companion diagnostics |
Getting around the pitfalls of taking companion diagnostics from the lab to the bedsides of patients will take more than just platitudes, according to experts from pharmas, payers and academics.
At a roundtable at the 2012 BIO International Convention, Mark Trusheim, a professor at MIT's Sloan School of Management, said the current clinical trial model doesn't jibe with companion diagnostics. Most trials seek to understand the average among patients, he said, whereas personalized treatments are meant to understand variance, to identify target populations for smaller-market treatments. "We need to develop trials that account for broader variance" in order to develop effective companion diagnostics, Trusheim said.
Of course, that's just the developmental aspect. After that, drugmakers face the challenge of convincing payers to pony up for the necessary tests.
Bryan Dechairo, senior director at Medco Health Solutions, said most payers are more willing to fund efficacy-based tests--those that identify patients who will respond to a treatment--than safety tests, the ones that tab those most likely to suffer adverse events. The math is simple, Dechairo said: Efficacy-based tests are cost-effective because they identify target patients, while safety-based ones spend up-front money only to limit a drug's treatable population.
But even when payers are all-aboard with an efficacy-based test, it can be impossible to ensure that patients are getting the right ones. James Creeden, chief medical officer at Roche Diagnostics, used the company's melanoma treatment Zelboraf as an example. Roche ($RHHBY) developed the cancer treatment alongside a test for gene mutation that could tag patients who would respond to the drug.
There was just one problem: Roche has no way to ensure that patients get the company's FDA-approved test and not a copycat produced by another firm. Some of the rival tests for Zelboraf have proven hideously unreliable, Creeden said, and payers don't restrict which tests they'll reimburse, allowing doctors to prescribe off-brand diagnostics that can lead to faulty results.
That saps up a lot of the incentive for drugmakers to pursue companion diagnostics in the first place, Trusheim said, as pharmas can't be sure patients are getting effective tests, and if knock-off diagnostics lead to unreliable results, that will cut into prescription rates and profit margins.
There's no quick fix for the companion diagnostics world--all of the panelists agreed on that. But, Trusheim said, if drugmakers, regulators and payers communicate throughout the process--not just when Dx-treatment combos are submitted and approved--it would go a long way toward smoothing over difficulties and creating better outcomes for patients.