A negative coverage recommendation proposed by the Oregon Health Authority could restrict access to next-generation sequencing tumor profiling for some of the most vulnerable residents in the state, cancer patient advocates fear.
In June, the health authority’s Health Evidence Review Commission (HERC) issued a draft guidance strongly recommending against covering NGS testing of solid tumors, reasoning that the available evidence doesn’t show that the intervention is clinically useful in improving patient outcomes, or contributes to cost savings by avoiding unnecessary treatments.
Recommendations from HERC guide allocation of state budgets.
“Although personalized (precision) cancer therapy is of significant interest currently, our recommendation for non-coverage of next-generation sequencing tests is strong because there is no direct evidence of benefit, and the best available evidence does not yet establish survival advantage with targeted cancer therapies,” wrote members of HERC and its subcommittees involved in preparing the coverage guidance.
Advocacy organizations are concerned that if HERC finalizes this recommendation, it could be used to deny coverage for NGS tumor profiling for Medicaid beneficiaries, even in cancer types such as non-small cell lung cancer, where testing has shown to be beneficial and is backed by professional society guidelines. A broad non-coverage policy from HERC in Oregon could also end up influencing private payors and other state health authorities.
Members of the public had until Aug. 13 to comment on the draft recommendation, and LUNGevity and several other cancer patient advocacy organizations told HERC that they disagree with its assessment. They’ve pointed out that HERC’s recommendation is in direct contrast to the Centers for Medicare & Medicaid Services’ decision to grant national coverage when advanced cancer patients are tested on NGS cancer panels with FDA-cleared or -approved companion diagnostic indications. The national coverage determination was triggered by Foundation Medicine’s application for CMS and FDA to simultaneously review its 324-gene FoundationOne CDx panel.
While most of the cancer community has been pushing for broad coverage of NGS cancer panels, Vinay Prasad, hematologist/oncologist and an assistant professor of medicine at Oregon Health and Science University, has fashioned himself as the leading naysayer against overzealous adoption of precision oncology. He penned a commentary in the Annals of Oncology earlier this year arguing that the NCD puts the public health at risk because tests like Foundation’s have not demonstrated clinical utility for most of the genes on the panel. He maintains CMS should have asked Foundation to demonstrate clinical utility for its NGS panel in a randomized-controlled study, and in his commentary even suggested two study designs that can be conducted in a reasonable time frame without incurring exorbitant costs.
Prasad did not respond to an email requesting comment.
It hasn’t gone unnoticed by patient advocacy groups that Prasad is a member of HERC’s Health Technology Assessment Subcommittee (HTAS). Walter Shaffer, a medical consultant at Oregon Health Authority and an author of the coverage guidance, said that Prasad is influential on HTAS but also noted these are draft recommendations that could still change in the final version.
The subcommittee is slated to discuss comments submitted by more than a dozen stakeholders during a Sept. 27 meeting. The recommendation from HTAS will go to the Value-based Benefits Subcommittee, which develops guidelines for state Medicaid plans, and then will proceed to the full Health Evidence Review Commission, which will issue the final coverage guidance.
If stakeholders bring forward convincing evidence that HERC hasn’t considered and which shows that NGS testing is clinically useful broadly or in specific indications, then that might be reflected in the final recommendations, Shaffer suggested. “Vinay Prasad is quite influential in HTAS, but everybody has a vote, and we take all the public comments seriously, and issue written responses to those,” he said. “If there is additional evidence to be looked at, then one person can’t stand in the way of that.”
According to Shaffer, the health authority does cover testing for specific biomarkers, such as EGFR in non-small cell lung cancer to predict response to EGFR inhibitors. “We took up this topic of NGS testing to see if this technology had its own evidence of benefit in terms of selecting targeted therapies that improved outcomes,” he said.
The draft recommendation is based on an evidence review, prepared by OHSU’s Center for Evidence-based Policy, that discusses three studies — one randomized-controlled trial and two meta-analyses.
The SHIVA trial, the only published randomized-controlled trial looking broadly at the clinical utility impact of NGS, found that patients matched to a molecularly targeted therapy did not experience improved median progression-free survival compared to the control group, but had more serious adverse events. The study authors noted that patients’ ability to benefit from genomically informed therapies may have been impacted by multiple co-occurring mutations, treatment with single targeted drugs (instead of a combination therapies blocking multiple targets), and limited access to targeted agents.
The committee considered a meta-analysis involving some 638 studies, which showed a 31 percent response rate among patients receiving biomarker-guided personalized treatments and a 10.5 percent response rate in those in non-personalized treatment arms. The study authors have said this research was limited by the fact that it didn’t adjust for factors that could have predisposed patients with targetable mutations toward better outcomes compared to those without targetable mutations.
The second meta-analysis considered by HTAC involved 57 randomized and 55 non-randomized studies, of which a proportion employed biomarker-based treatment strategies. Personalized treatment studies were associated with greater relative response rate ratios compared to non-personalized studies, as well as longer median progression-free survival, but not longer overall survival. There was also no difference between the two types of studies in terms of treatment related deaths.
“It’s important to have these debates about the quality of the evidence [on NGS cancer profiling] but our hope would be to do so in a way that results in better consistency in access,” said Jeff Allen, CEO of Friends of Cancer Research. He pointed out that CMS conducted a far more extensive evidence review of NGS cancer panels, and concluded that they are clinically useful when the panels are FDA approved or cleared as companion diagnostics.
CMS considered more than 200 published studies in its clinical utility review, including the SHIVA trial and several studies that failed to find that NGS testing improved patient outcomes. Ultimately, though, the government payor concluded that when the FDA has approved or cleared an NGS-based companion test for an advanced cancer patient, “it is sufficient … to expect meaningful improvement in their health outcomes, such as progression-free survival.”
HERC, however, didn’t accept FDA clearance or approval as a proxy for clinical utility. The committee issued a blanket non-coverage determination despite the NCD, the availability of FDA approved/cleared tests, and clinical practice guidelines recommending NGS testing for colorectal and non-small cell lung cancer.
HERC indicated the need for randomized-controlled trials demonstrating the utility of NGS testing. “Randomized clinical trials utilizing next-generation sequencing are certainly feasible and appropriate and would not need to be of long duration for the most prevalent types of cancer,” the group wrote.
Friends of Cancer Research wrote in comments to the committee that while such studies are the gold standard for evaluating drugs, they “often do not evaluate appropriate measures for diagnostic tests” and may be unethical in certain circumstances. The group suggested HERC consider real-world evidence, which would depend on patients having access to and coverage for NGS testing.
Upal Basu Roy, director of LUNGevity’s translational science research program, criticized HERC for failing to consider how NGS has impacted individual cancer indications since the level of clinical utility evidence varies depending on whether it’s used in a breast, lung, colon, or prostate cancer patient. In lung cancer, for example, NGS testing is recommended in practice guidelines since there are multiple biomarker-informed, FDA-approved treatment options, and access to tissue is difficult, especially in advanced patients.
In comparison, “the utility in breast cancer is really behind, and prostate cancer doesn’t even have a targeted therapy,” Basu Roy said. “Comparing NGS in prostate cancer to NGS in lung cancer is like comparing apples to oranges.”
A blanket non-coverage policy for NGS cancer panels can be further detrimental to innovation, patient advocates observed, at a time when most big pharmaceutical firms are exploring biomarker-driven strategies for the majority of molecules in their oncology portfolios and taking a tissue-agnostic approach toward target discovery using basket and umbrella trials that rely on NGS.
“If you don’t provide NGS coverage right now, you’re setting a very bad precedent for companies to not use precision medicine-based diagnostics when they’re conducting clinical trials,” Basu said. “You’re giving them the message that for every new drug approved in the future, you’ll have to go through the reimbursement process again and again for every new [biomarker-guided] indication that you add.”
Risk of harm?
In the absence of evidence supporting broad use of NGS testing, HERC stated there is potential for patient harm because such testing may “promote use of costly targeted therapies when equally effective (or more effective) conventional chemotherapy might be available.”
CMS’ national coverage decision for NGS cancer panels has raised similar concerns about unnecessary spending and patient harms. The terms of the NCD grant automatic national coverage for a broad panel of genes with varying or no clinical utility evidence as long as there is one gene on the panel with FDA approval or clearance as a companion diagnostic. Some experts in the oncology community, among them Prasad, predict that such a broad coverage policy for NGS testing will increase off-label use of pricey targeted therapies that won’t benefit patients but cause adverse events and further drive up costs.
However, Basu Roy doesn’t see much risk of ballooning cost or adverse risks precisely because the clinical utility of NGS varies across tumor types. The community oncologist is more likely to send a lung cancer patient for NGS testing but for a breast cancer patient, the probability is a lot lower.
“Just saying there is coverage for NGS testing doesn’t mean there will be more uptake from oncologists,” he said. “That’s an assumption some payors are making but the reality is that a doctor doesn’t know how to use NGS data in a breast or a prostate cancer setting, but does know how to use NGS data in the lung cancer setting.”
It remains to be seen if such arguments will persuade HERC to soften its stance. According to Anna Pugh, director of public policy initiatives at LUNGevity, the Oregon committee is currently an outlier among state health authorities in considering a broad non-coverage recommendation for NGS cancer panels. If the guidance is finalized as is, however, Pugh fears private payors and other states might follow Oregon’s lead, impeding access for Medicaid beneficiaries around the country.
Currently, Oregon Medicaid plans tend to lean toward non-coverage because there is no guidance on NGS cancer panels, but it can decide to grant coverage a case-by-case basis. Once HERC issues a final recommendation, it is generally adopted by the Oregon Medicaid program though the guidance defines the minimum level of coverage. “If HERC decides that next-generation sequencing is not covered at this point, state Medicaid plans could still cover it for individual cases, but as a general rule, they’ll follow these guidelines,” Shaffer said.
Meanwhile, commercial plans aren’t bound to HERC’s recommendations. “Some commercial payors are interested in our evidence reviews, while others aren’t,” he said.