Payers Evaluate the Clinical Utility of Diagnostic Tests

Diagnostic tests play a critical role in this era of personalized medicine, as they guide physician decision making, particularly in oncology. Molecular tests generate data that are expected to improve patient outcomes. Research has shown that use of molecular diagnostic testing has increased 20% to 30% each year, while the number of therapeutic decisions based on these tests increased 200% during a recent 3-year period.¹However, like with most innovations, molecular diagnostic testing does not come without costs.

What are the factors that influence reimbursement and policy decisions for diagnostic tests? What evidence is needed to ensure payment? How is the value of a diagnostic test determined? These were some of the questions addressed by payer representatives from leading managed care companies during theOncology Stakeholders Summit: Insightsdiscussion organized byThe American Journal of Managed Care. The participants includedJohn L. Fox, MD, MHA,senior medical director and associate vice president of medical affairs at Priority Health;Michael Kolodziej, MD,national medical director for Oncology Strategies, Aetna Inc; andBryan Loy, MD,physician lead—cancer, Humana.Peter Salgo, MD,professor of medicine and anesthesiology at Columbia University and associate director of surgical intensive care at NewYork-Presbyterian Hospital, interviewed the experts individually.

Currently, physicians can choose from 3 tests to assess distant recurrence in early-stage breast cancer: MammaPrint (Agendia, 70 genes), Prosigna (Nanostring, 50 genes), and OncotypeDX (Genomic Health, 21 genes). The 3 gene panels have only 12 common genes, none of which are shared across the board. Both MammaPrint and OncotypeDX are covered by Medicare and by a few commercial carriers; Prosigna is covered by UnitedHealthcare but not yet covered by Medicare. While MammaPrint, a laboratory-developed test (LDT), is a microarray that is offered as a service, OncotypeDX, also an LDT, is an RT-PCR assay that can be conducted in a laboratory but is also offered as a centralized service. Prosigna, which is based on the PAM50 gene signature, is an FDA-approved kit that is decentralized.²

Salgo asked the experts whether they saw any advantage of a decentralized test versus a centralized LDT. All 3 payers agreed that being decentralized is a definite advantage for a diagnostic test, with a faster turnaround time and ease of access. Kolodziej said that OncotypeDX, an LDT that is not FDA approved, requires samples being sent to California where the testing service is offered. He believes that while oncologists have come to trust the proficiency of the test, competition from an FDA-approved diagnostic would be difficult to disregard. Besides a faster turnaround time, Kolodziej also pointed to the ease of access to a local pathologist with a decentralized test like Prosigna. “Getting a pathologist on the phone, talking about the result, is a lot more challenging when the pathologist is in California. Now, I’m not saying that Oncotype doesn’t have good customer service. I’m saying that there’s a big advantage to being able to talk to the pathologist,” he said.

Fox, however, did not see any direct advantage to the patient. The fact that a patient may not immediately be initiated on treatment following the results of the test means that a decentralized test does not necessarily serve them. He went on to add, “I’m indifferent as a payer to what test I pay for if the costs are the same,” and that he did not believe decentralizing would influence the cost of the test.

FDA Guidelines for Diagnostic Tests

The FDA recently announced that it would issue guidance for the development, review, and approval of companion diagnostics, while also publishing a risk-based oversight framework for LDTs.3,4 The decision was based on the increasing reliance of physicians on diagnostic test results to make treatment decisions. Would payers welcome this increased oversight on diagnostics? Would it influence payer policies?

Kolodziej said that payers would definitely welcome the FDA’s “Good Housekeeping Seal of approval.” While they might wish to do so, payers do not have the ability to diligently scrutinize these tests, so an FDA evaluation would be a tremendous help, he said. “I have sat on panels—multistakeholder panels—where patients and patient support groups are represented. And they are horrified by the concept that there is no standard proficiency testing, that they could, in fact, have a test result that’s done by a lab that doesn’t do a very good job; and their whole therapeutic course could be influenced by that,” Kolodziej stated. He added that a community oncologist who’s ordered a particular test cannot dictate where the test will be conducted—whether a tissue sample is evaluated using a test developed in the pathology laboratory or sent to a vendor to conduct. This level of variation, he said, can undermine the oncologist’s confidence in the results as well.

Loy echoed this, saying that providers may not necessarily understand the rigors of these tests, and they have to rely on the decisions made by the pathologist on where and by whom the tumor sample is tested, resulting in a lack of transparency on the analytical and clinical validity of the test. “I think that the value in having a more rigorous and well-defined FDA process will help in the planning stages,” said Loy. “Many of these folks developing these tests are largely growing out of the laboratory industry, and they don’t necessarily have the deep pockets.” He emphasized that without stifling innovation, streamlining and regulating the process would guarantee reliable tests. Fox believes that FDA scrutiny would ensure the clinical and analytical validity of the test.

Value Proposition of a New Diagnostic Test

Competition is always good, and it helps level the playing field—a monopoly is not ideal for growth and innovation. “Now, all of a sudden, you’ve got a discussion around price point, accessibility, decentralization, how interpretable the information is, ease of doing business, and the ability to generate sufficient turnaround time in order for them to get out in front and not delay therapy. All those become extremely important pieces of the discussion, above and beyond just the evidence base,” stressed Loy.

According to Kolodziej, OncotypeDX has an advantage because it was the first test in the space, and its backers were able to convince physicians of its predictive and prognostic value. In his opinion, OncotypeDX has provided data that can identify the right patient population to receive chemotherapy. On the other hand, Prosigna, with its quicker turnaround time and a more stringent patient prognosis, is yet to provide predictive data on chemotherapy response, said Kolodziej. And that, he thinks, would be the clincher to guide treatment decisions. “If, in fact, [Prosigna] can perform the same functionality, then it will become a very useful test and then we’ll see. Then we’ll have true competition in the marketplace.”

Fox said that a head-to-head comOncology way to determine which one is optimal. A paper published in theJournal of Clinical Oncologyin 2013 assessed the mRNA of 1017 estrogen receptor—positive breast cancer patients treated with anastrozole or tamoxifen in the Armidex, Tamoxifen, Alone or in Combination (ATAC) trial. The trial found that the risk-of-recurrence score provided by PAM50 added prognostic information far beyond the clinical treatment score in all the different subsets evaluated. Additionally, the test categorized fewer patients in the gray “intermediate-risk” category compared with OncotypeDX.⁵

Fox said he is more concerned about patients being placed in the appropriate risk category than about the actual number of patients being classified as “high” versus

“intermediate” versus “low” risk. Proper patient stratification is key, he affirmed.

Kolodziej presented an alternate rationale: that patients stratified in the gray zone pose a challenge for oncologists. “There are actually a couple of really large studies being done right now, such as the TAILORx study, to try to figure out whether or not people in that intermediaterisk category should receive treatment. So we’re going to get at the answer to this question, either by reclassification or by a study that shows the intermediate-risk (population) does or does not benefit from chemotherapy,” he said.

All 3 payers agreed that a test that can determine the response to treatment can be defined as being “predictive.” Kolodziej would like the test to determine the patient’s

EBO

outcome in the long run, while Fox would ideally like to see the test define optimal and personalized therapy for a patient, not just response to chemotherapy. “If I could acquire information that was easily understood by both the patient and the provider and it was clinically meaningful, being able to provide that level of context, I think, would be the most valuable,” said Loy.References

1. Burns J. Health plans deploy new systems to control use of lab tests. Managed Care website. http://www.managedcaremag.com/archives/2014/10/health-plans-deploy-new-systems-control-use-labtests. Published October 2014. Accessed October 15, 2014.

2. Biomarkers and commercial strategy: comparing the three leading breast cancer panels. Biomarker Trends website. http://www.biomarker-trends.com/review-of-three-leading-breast-cancer-panels/. Published May 9, 2014. Accessed October 16, 2014.

3. FDA takes steps to help ensure the reliability of certain diagnostic tests [press release]. Silver Spring. MD: FDA newsroom; July 31, 2014. http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm407321.htm.

4. Caffrey M. FDA moves to regulate thousands of diagnostic tests.Am J Manag Care. 2014;20(SP11):SP338.

5. Dowsett M, Sestak I, Lopez-Knowles E, et al. Comparison of PAM50 risk of recurrence score with Oncotype DX and IHC4 for predicting risk of distant recurrence after endocrine therapy.J Clin Oncol. 2013;31(22):2783-2790.

Payers Evaluate the Clinical Utility of Diagnostic Tests