“Wouldn’t it be great if we could come up with a noninvasive diagnostic assay to detect pancreatic cancer at an earlier, more treatable stage?” asked Lori Aro of Myriad Genetics. Her company has been trying to do so for years. So why hasn’t it happened?
Aro, senior director for new product planning at Myriad, outlined the business obstacles at a recent panel hosted by Boston Children’s Hospital’s Technology and Innovation Office (TIDO).
First, who are the target patients for a pancreatic cancer test? Skinny diabetics, patients with chronic pancreatitis, patients with hereditary cancer risk — or all three? “Those three patient types all sit in different doctor’s offices,” said Aro. Simultaneously reaching endocrinologists, gastroenterologists and high-risk patients would be an insurmountable challenge, Myriad concluded.
Second, the assay would likely need to be validated in all three patient populations, with confirmatory imaging. Could the test populations be large enough to make the results statistically significant?
Third, a new test wouldn’t change care, as there is no treatment for pancreatic cancer. In fact, no current data show that earlier diagnosis improves survival. So who would pay for it?
Aro’s story exemplifies just some of the challenges in developing a new diagnostic test. Supreme Court rulings, such as Myriad vs. Association for Molecular Pathology and Mayo vs. Prometheus, have made it hard to patent diagnostic correlations. The Centers for Medicare & Medicaid Services (CMS) and, to some extent, the U.S. Food and Drug Administration (FDA) are requiring long and costly trials to demonstrate clinical utility and better patient outcomes before they will pay for a new test. And Medicare has decreased reimbursement rates for tests in general.
How do scientists interested in developing diagnostics navigate such hurdles? The panel provided six takeaways.
Intellectual property is important, but doesn’t top companies’ technology assessment lists.
“The first thing we evaluate is whether or we believe the test is reimbursable and how quickly it can get to market,” said Aro. “Next, can clinical utility be demonstrated? Can it be commercialized in a high-throughput laboratory? Can it be reproduced three times?”
Jerry Casey, MBA, diagnostics executive and former president and COO of Sekisui Diagnostics, noted that strong science and the potential clinical utility of a biomarker test are most important. Having a robust data package to support a clinical trial is next. Casey also noted that a test should be compatible with existing diagnostic instrumentation or platform technologies
You can still get diagnostic claims in a patent — if you think outside the box.
“A typical diagnostic patent claim is a correlation between a biomarker and a disease, which is very difficult to patent because of the Supreme Court decisions on such claims as a natural phenomenon,” said panelist Roque El-Hayek, MD, JD, shareholder, Wolf, Greenfield & Sacks. “It needs to involve something more significant. The challenge is to find the additional elements that get you from the ineligible to eligible for patenting.”
For example, a panel of five to six biomarkers or a particular biomarker signature could be patentable. Casey was able to get patentable claims around a synthetic calibrator used to measure the biomarker. The process by which the diagnostic antibodies or the immunogens are produced may also be patentable, if necessary to fully develop and produce the final product.
Think of CMS as a regulatory agency.
Although the FDA has the purview to regulate laboratory diagnostic tests (LDTs), it is only regulating a small segment of them. The buck really stops with CMS: you have to prove that doctors will use your test and that it will change the standard of care and improve patient outcomes.
To even consider paying for an assay, CMS requires at least two clinical validation studies and two clinical utility studies — prospective studies in which patients who get your test actually have better outcomes. These studies are long and expensive, and reaching CMS’s milestones doesn’t guarantee that it will reimburse. It just opens up the conversation.
CMS coverage doesn’t guarantee coverage by other payers — and coverage doesn’t guarantee a viable product.
“[CMS] is the starting point and does set the standards, but each payer is different,” said Aro. The Blues are the most conservative in terms of how they evaluate your data package, she added.
Under the current reimbursement system, you can’t petition the American Medical Association for a Current Procedural Terminology (CPT) code until the FDA has reviewed the clinical trials and given regulatory clearance.
Once your diagnostic is reimbursable, payment level becomes the issue. Reimbursement may not necessarily reflect the cost benefit or value of the test, which could hurt its financial viability. This problem is causing a shakeout in the industry, but could resolve in the next few years, believes Casey. Under the recently passed Protecting Access to Medicare Act (PAMA) Regulations, laboratory test payment will change to a more market-based payment system in 2018, meaning that commercial pricing will eventually determine CMS payment. The diagnostic industry hopes this will be better than the current system.
Be open to new relationships with industry.
There is a spectrum of relationships companies can have with early-stage discoveries, depending on the disease and how hard it will be to establish clinical utility of the product. “Academic scientists coming up with amazing ideas will always outpace our ability to commercialize,” says Aro.
How much risk a company will bear depends on clinical utility and the test’s commercial potential. To assume more of the risk, Myriad has to see a very well-defined clinical utility endpoint and be able to visualize a path to getting the data package payers require for reimbursement.
Licensing may be the approach when companies are fully confident in the technology. When they are not, companies may prefer to collaborate, asking the academics to take on some of the risk and continue to develop the concept.
Think beyond the science.
Good science at the bench isn’t enough. El Hayak recommends that academics think beyond data correlations and consider how the product or assay would be used — and then think about the additional elements that will get over the patent hurdle.
“I would say spend time in a clinical laboratory, see how that lab operates and get a good understanding of what labs are looking for in terms of technological advancements,” said Casey. “Then talk to doctors about the real clinical unmet need. We can always come up with good science, but it may not pan out in the physician’s office.”