Genomic diagnostics are moving into a wealth of new application areas, but speakers at a recent conference underscored a number of challenges facing companies looking to commercialize novel tests.
At the Future of Genomics Medicine conference hosted earlier this month by the Scripps Translational Research Institute, several speakers discussed the rapid progress they have made in developing genomic tests to detect disease and guide treatment. But while the science behind such tests is advancing rapidly, bringing them to market poses a number of hurdles that can slow the commercialization timeline.
Eric Topol, director of the STRI and a co-chair of the conference, remarked that recent advances in genomics — in combination with other new technologies like smart phones and social networking — are likely to result in the "biggest shakeup in medical history" as new discoveries find their way into patient care.
But these new discoveries can sometimes take their time getting into the hands of doctors. For example, according to presenter Matthew Price, even though the link between CYP2C19 genotype and clopidogrel response is well-established, and the FDA has added information about genetic testing to the drug's label, "outside of Vanderbilt, Scripps, and other forward-thinking centers, CYP2C19 genotyping has not taken huge hold in the community at large."
Price, an associate professor at STSI and director of the cardiac catheterization laboratory at the Scripps Clinic, presented several reasons for why this is the case, with the key reason being the turnaround time for current tests, which can take up to 48 hours to get results. This is an unreasonable timeline given that most treatment decisions for heart attack patients and those undergoing percutaneous coronary intervention are made in the acute setting — generally no more than 90 minutes after a myocardial infarction, he noted.
Even though there are a range of available CYP2C19 genotyping tests from companies like Roche, Quest, and Autogenomics, these tests "are not practical," Price said. "They're complex, they're specialized, they require infrastructure, and, importantly, they're economical only when batched."
He noted, however, that adoption might pick up as new technologies emerge for "bedside genotyping." He cited Nanosphere's Verigene point-of-care CYP219 test, which returns results within four hours, as a step in the right direction. "It can run single samples, and that's important because if you're the one with a heart attack or getting a stent, you don't want to wait until next Thursday when [the lab] can get enough samples to run your genotype."
The Verigene test is currently being used in a clinical trial, called “Thrombocyte Activity Reassessment and Genotyping for Percutaneous Coronary Interventions," or TARGET-PCI, which aims to demonstrate a 30 percent relative risk reduction in post-PCI ischemic events with PGx-guided clopidogrel treatment compared to standard treatment (PGx Reporter 8/11/2010).
In addition, Price said that he and his colleagues at Scripps are working with Focus Diagnostics, a subsidiary of Quest Diagnostics, on a new system that will enable CYP2C19 genotyping within 30 to 45 minutes.
"That, clearly, will be a real advantage," he said.
For other speakers, the obstacles to commercializing their tests were not so evident.
Geoffrey Ginsburg, director of the Center for Genomic Medicine at the Duke Institute for Genome Science and Policy, discussed his team's work with CardioDx to develop Corus CAD, a 23-gene expression test to determine whether stable, symptomatic patients are likely to need further assessment for obstructive coronary artery disease. There is a clear need for such a test, he noted, since it's estimated that approximately two-thirds of patients who undergo invasive angiographic procedures have no evidence of CAD.
The company published a validation study for the test last year in the Annals of Internal Medicine, which described the outcome of the 1,300-patient "Personalized Risk Evaluation and Diagnosis in the Coronary Tree," or PREDICT, trial. The investigators found that Corus CAD demonstrated a sensitivity of 85 percent and a negative predictive value of 83 percent.
"This test is available, it's being used, but it's still trying to find its market niche," Ginsburg said. "There is clear evidence of the overuse of elective coronary angiography, and the question is, 'How does a test of this complexity get integrated into the normal clinical workflow, which has been in place and dictated by clinical guidelines for many years?'"
[ pagebreak ]
Ginsburg, who is on the board of CardioDx, said that the company is still determining how best to position the test.
"Could it be that Corus CAD will be a filter to rule out obstructive coronary disease, since it has a high negative predictive value? Could it be used in combination with imaging tests to augment the ability to diagnose obstructive coronary disease? These are some of the critical questions that are trying to be answered right now by CardioDx and colleagues in order to position it in the marketplace and get it clinically adopted," he said.
One issue facing CardioDx and other firms developing genomic diagnostics is that there "seems to be a moving target — or an undefined target — about the evidentiary threshold that is necessary for clinical adoption, regulatory approval, and reimbursement," Ginsburg said. "This is something that I think has really stymied the field."
Ginsburg called for the "harmonization of regulatory and reimbursement standards" in order to help these complex diagnostics move forward.
Speaking from the perspective of a firm that has successfully commercialized a genomic test — and gained reimbursement for it — Steve Shak, chief medical officer of Genomic Health, shared some advice for "what one needs to do if one is interested in developing a test and then having that test go all the way forward, be used in clinical practice, be reimbursed in clinical practice, and ultimately transform medicine."
The "number one point," he said, is for test developers to focus, "even from the very beginning, about what patients, physicians, regulators, and payors are going to need."
Shak stressed that "tests must be fit for purpose," noting that the requirements for a direct-to-consumer test would be "very different, for example, from testing that might be used in the transplant setting."
Proper validation is also crucial. "The reason most biomarker studies never get incorporated [into clinical practice] is that most biomarker efforts stop at the very first stage, … which are very small convenience studies with readily available samples, and they're not hypothesis-generating."
Developers must also prove that they can get consistent results across multiple studies, Shak said.
In addition, "the test must be shown to have value beyond traditional measures — what the healthcare community is already practicing," he said. "Doctors don't go around thinking that what they do every day isn't working, so what you've got to do is build on that. You've got to evaluate what's normally done and then directly, in a head-to-head way, compare your new test. Comparative effectiveness is just as important for diagnostics as it is for drugs."
Another key aspect to a successful commercialization strategy is a standardized implementation for a test, he said. "One of the things that's different about diagnostic tests and drugs is that most people believe that they probably shouldn't brew up their own drug on a bench in their laboratory and then use it to treat patients," he said. "And yet it's just too easy to brew up an assay in your lab and then say, 'Let's take a look at it now in clinical research.'"
Finally, he noted that even though reimbursement is one of the biggest challenges for the field, it is not an insurmountable hurdle.
"Getting reimbursement is not trivial," he said. "We have many, many payors and each of them has to be addressed individually."
Genomic Health currently employs more than 50 people to help secure reimbursement for its tests. So far, "more than 90 percent of lives in the US are covered" for the company's Oncotype Dx test in estrogen-receptor positive node-negative breast cancer. However, the company is now pursuing a node-positive indication for the test. "So we now have to go back again — even though our test has been shown to be valid in node-positive disease — one by one, and get, again, reimbursement there as well," Shak said.
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at btoner [at] genomeweb [.] com.