Lab Industry Says CMS Draft National Coverage Decision for NGS Cancer Tests Bad for Patients

NEW YORK (GenomeWeb) – The Centers for Medicare & Medicaid Services' recent draft national coverage determination for FoundationOne CDx and similar next-generation sequencing tests has some lab industry players concerned that if the policy is finalized, it would significantly limit late-stage cancer patients' access to critical testing. 

In comments to the draft NCD so far, academic medical centers and providers of lab-developed tests (LDTs) have criticized the policy for the very limited circumstances in which CMS is proposing to cover NGS tests that are performed in a CLIA-certified lab but lack US Food and Drug Administration approval or clearance. By tying a broad national coverage decision for NGS cancer tests to FDA approval/clearance, these groups have accused CMS of advancing FDA oversight of LDTs, which the lab industry and pathologists' groups have aggressively fought against. 

"I expect CMS will hear from practitioners, professional societies, and others who are apprehensive about this possibility during the comment period," said Roger Klein, chair of the professional relations committee at the Association for Molecular Pathology. "It seems likely that the issue will require clarification."

One commenter expressed suspicion that CMS' coverage criteria "appear to be contrived to support a single vendor's operations," namely Foundation Medicine's. If these restrictive criteria are finalized, it would foster a monopoly, others wrote, hindering access to patients in the community who rely on LDTs developed and provided by local labs. The vast majority of these tests don't have FDA approval or clearance, though this year the agency provided marketing authorization for several NGS cancer panels. 

The most recent action was for FoundationOne CDx, for which Foundation used the Parallel Review pathway to simultaneously garner a national coverage determination from CMS and FDA approval. In issuing the draft NCD, however, CMS said it wanted to advance "equitable" and "predictable" coverage for the entire class of NGS tests, and not just limit its decision to Foundation's test. 

Recognizing that NGS cancer tests simultaneously assess many genetic variants with varying levels of clinical validity, the draft NCD parses coverage into three pathways for recurrent, metastatic, or advanced stage IV cancer patients. CMS is willing to grant top-level coverage for the detection of genetic alterations that the FDA has approved as companion diagnostics, required for the safe and effective use of certain therapies. 

Then, CMS proposes a coverage with evidence development (CED) option for FDA approved/cleared tests that don't have a CDx indication, but provides biomarker results to help manage patients' cancer when used with other tools and information. In order to qualify for this second level of coverage, tested patients must be enrolled in a prospective registry that tracks their overall survival, progression-free survival, response rate and other data.

The last CED option proposed by CMS is the only pathway open to LDTs without FDA clearance or approval. For this third level of coverage, tests must be included in an National Cancer Institute study within its National Clinical Trial Network .

In addition to the narrow circumstance in which LDTs are covered, industry players are also confused about how the requirements for CMS's three coverage pathways align with a framework the FDA has developed in reviewing recent NGS tumor panel tests. In that framework the agency outlines the evidence needed to support three types of biomarker claims (see pdf for larger version of graphic below).

"We certainly support pathways for the coverage of innovative molecular diagnostic tests," said American Clinical Laboratory Association President Julie Khani. "However, the issuance of an NCD with such a broad impact on patient access to emerging and current NGS testing deserves careful consideration by CMS. We're concerned that the time frame under the NCD to really engage with stakeholders is inadequate."

It appears that CMS has listened to these concerns. This week, the agency extended the public comment period, originally slated to expire on Dec. 29, to Jan. 17, 2018.

Shutting out LDTs?

CMS' emphasis on FDA approval has raised red flags among lab industry players. When a patient or a test doesn't meet the criteria proposed in the NCD, the draft document explicitly states that NGS testing is not covered.

Prior local coverage determinations from Medicare contractors, which a final NCD would supersede, didn't restrict coverage for NGS testing on the basis of FDA approval. But in this proposal, CMS "seems almost to go out of its way to say [non-FDA approved LDTs] are not covered," said Joydeep Goswami, president of clinical NGS and oncology at Thermo Fisher Scientific, adding that limiting coverage for LDTs is a bad decision for patients.

Many patients rely on NGS tests provided by CLIA/CAP certified labs in their vicinity and at their local hospitals, and the lack of Medicare coverage of these tests could hinder access to potentially life-saving information for a late-stage cancer patient. "Our patients depend on these tests being readily available in order to get diagnosed and treated," wrote Michelle Afkhami, medical director of clinical molecular diagnostics at City of Hope Cancer Center. "Our cancer center is the last hope for many patients in the west region of US."

In order to qualify for the two CED pathways, CMS is also asking that NGS tests be listed in the NIH's Genetic Testing Registry, so that their analytical validity, clinical validity, and clinical utility data are publicly available. According to Adriana Malheiro, director of the GTR, the repository currently includes information on 282 tests where somatic targets are evaluated via NGS. In order to be registered, labs must fill out minimal fields, including information on test methodology, analytical validity, and the markers assessed by the test, and 75 percent of labs complete some of the optional fields.

However, the vast majority of the more than 54,000 registered tests are LDTs and have blank fields that ask for FDA and NY State Department of Health approval information, Malheiro said, because they don't have it.

The draft NCD is "essentially shutting out LDTs that have been the backbone of patient access to NGS testing," Goswami said. Though CMS' proposal would benefit NGS tests with FDA-approved CDx indications, like FoundationOne CDx and Thermo's Oncomine Dx Target Test, few tests hold this distinction because of the resources and time required to meet the agency's regulatory requirements.

The FDA has also approved companion diagnostic indications for FoundationFocus CDx BRCA and Illumina's Praxis RAS panel. And through a more streamlined pathway, the agency recently authorized Memorial Sloan Kettering Cancer Center's 468-gene MSK-IMPACT test as an IVD that can be used to manage patient care alongside other information, which makes it the type of test that could qualify for CED if patient outcome data were collected in a prospective registry.

However, MSK also offers 13 other NGS tests that don't yet have FDA's blessing, for which the cancer center may seek regulatory clearance, but in the meantime these tests potentially would not be covered for Medicare patients.

"If you take this coverage decision to its extreme, it's sort of a backdoor way that the FDA oversight of LDTs, which was proposed and then tabled, is being resurrected," said David Klimstra, chair of MSK's pathology department.

Several years ago, the FDA said it would lift its longstanding enforcement discretion over LDTs and issued draft guidance on how it planned to regulate such tests currently overseen by CMS under the Clinical Laboratory Improvement Amendments. The agency felt the need to step in because standards under CLIA were inadequately protecting the public health. The lab industry and pathologists pushed back against the agency's efforts, arguing as they now are in response to CMS' draft NCD, that FDA oversight would stifle innovation and hinder patient access to tests.

Industry players were largely relieved when, after the 2016 presidential election, the agency said it would hold off finalizing the guidance, and instead issued a discussion paper outlining some of its regulatory thinking. Since then, FDA Commissioner Scott Gottlieb has discussed the possibility of a legislative solution to this issue.

"All of the issues that came up as part of [FDA's LDT regulatory] proposal come to bear here," said Klimstra. "You essentially disenfranchise academic labs," where much of the research used to improve patient care happens, and "drive all the work into a small number of commercial labs or to a single lab."

He reflected that CMS may have advanced these requirements not fully appreciating how much NGS testing is currently going on in the oncology space, and the extent to which it is considered standard of care by specialty oncologists. "No one is prepared for the avalanche of FDA clearances that would have to happen to move tests even to level two coverage with evidence development," he said.

It remains to be seen how private payors will react to CMS's NCD once finalized. Commercial payor coverage for NGS testing is already inconsistent. After two large private payors, Anthem and UnitedHealthcare, instituted broad prior authorization policies for genetic testing, healthcare providers said coverage denials increased, particularly for NGS panel tests. 

CMS' position on NGS cancer testing may not lead to more robust coverage from private payors, Klimstra fears. "We were actually hoping this would go the other way," he said. "That CMS would agree these are necessary and critical parts of cancer care in the 21st Century, and based on their acceptance of this, private insurers would more reliably cover these tests."

Real-world challenges

Test developers have often complained that payors' demand for data from randomized studies is too high a bar for tests used to detect tumor markers that show up extremely rarely in the population. Registries are one alternative for collecting data on how patients fare after receiving drugs based on mutations identified through NGS testing.

CMS is proposing CED for FDA-approved or -cleared tests that track patient characteristics and outcomes in a prospective registry, though experts in the field aren't clear from the draft NCD if existing resources would meet the registry criteria. For example, MSK has created the open-access repository cBioPortal, which contains large-scale cancer genomic datasets. A subset of the cases in this database have clinical information associated with the genomic data, though it's not yet clear to Klimstra if this could serve as the type of registry CMS is asking for, or if MSK would have to create something new.

Others wondered if CMS has thought through its registry requirements in terms of how cancer patients actually receive care, which is often not at a single institution. "Even if all the patients you test are supposed to be in the registry, if you live in the real world you find that patients don't often return for follow up at the site where testing was performed," said Carl Morrison, president of OmniSeq, a company that markets a 144-gene tumor profiling panel and Immune Report Card, which analyzes biomarkers that can help doctors make decisions about immunotherapy. "They move on to somewhere else."

OmniSeq, a subsidiary of Roswell Park Cancer Institute, is exploring the more streamlined 510(k) clearance pathway the FDA introduced with the clearance of MSK-IMPACT for its tumor profiling panel, and is considering the Parallel Review pathway for Immune Report Card. Although the company can leverage a patient registry through the cancer institute, if it chooses to seek CED for its tests, Morrison wondered what percentage of patients would have to fit all the criteria and have all the outcomes data CMS is asking for.

CureOne is a non-profit that has been trying to address some of these concerns by launching a prospective, open-access registry for the collection of outcomes data from cancer patients tested on NGS panels. CureOne is currently collaborating with Foundation (on its FoundationOne panel), Guardant Health (which sells the liquid biopsy test Guardant360), and Washington University in St. Louis to enroll patients into its N1 Registry.

"We have largely built what it seems that CMS is requesting," CureOne CEO Dane Dickson said. "If coverage with evidence development survives in the final draft, we hope to be one of the groups that CMS identifies as an approved registry," said.

During his time working for Medicare contractors, including Palmetto GBA's MolDX program, Dickson said he saw it would be challenging for payors to advance coverage for multiplex panels. "The only rational way for a payor to introduce this technology and advance care was to provide testing as part of coverage with evidence development," he said. "Legally and scientifically, it's the only way that fits."

Three years ago, he started building the N1 Registry, where labs and clinics could submit the types of data that would inform the utility of NGS. The ideal registry would have to be precompetitive, protect patients, and be equitably available to everyone, Dickson said, but there was no obvious organization that could navigate the concerns of many stakeholders. He started CureOne as a non-profit that could act as an independent broker that could address different stakeholders' needs, and, at a low operating cost, encourage collaboration, data sharing and research.

Over the summer, CureOne announced the association of the N1 Registry and the American Society of Clinical Oncology TAPUR trial — a non-randomized basket study investigating the safety and efficacy of commercially available cancer drugs when they are given off label and based on the genomic characteristics of patients' tumors. CureOne has also reached out to the American Association of Cancer Research and the National Cancer Institute to discuss how its N1 registry can be integrated into other efforts currently underway. 

Ultimately, Dickson hopes that these different groups will come together to advance what would benefit everyone, "a national database of genomics data and outcomes, where we will not be sequestering and squabbling over data." FDA and CMS' parallel review decision for FoundationOne CDx has provided the cancer community with the opportunity to build something that will dramatically advance patient care, he reflected. "If we as a cancer community cannot do it, shame on us."

Not everyone matches

Industry players also have concerns about the proposed CED requirement that LDTs without FDA approval or clearance must be part of an NCI trial. Given that only around 4 percent of cancer patients eligible to partake in clinical trials actually participate, experts in the field felt the requirement might be a significant roadblock to patient access.

While there are several NCI-sponsored trials involving NGS platforms, the majority of cancer patients receiving treatment in the community setting may not have access to such trials, and even if they do, may not ultimately make it onto the study. For example, within NCI-MATCH, patients can have their tumors genetically profiled through Foundation, Caris Life Sciences, MD Anderson Cancer Center, and MSK.

As of June this year, although more than 5,482 patients had received NGS results within NCI-MATCH, only 983 had a genetic abnormality matching a study treatment and 660 patients, or around 12 percent, had enrolled. Investigators estimate that ultimately 72 percent of patients assigned to a treatment arm will enroll in the study. Still, these numbers are an important reminder that while patients are eager to partake in these new types of precision oncology trials, not everyone gets the chance. 

Dickson acknowledged that in the final NCD, CMS will have to address stakeholders' concerns about how to address patients who are screened for mutations but don't make it on a NCI-sponsored clinical trial. "Shouldn't they be covered under CED just like everyone else is?" he posited.

"This is still a draft policy, and this was a very good first shot over the bow," he said. "We're going to see a lot of refining in the final version."

Coding clarity

The draft NCD currently doesn't provide any guidance as to how the tests that fall into three coverage pathways would be coded and priced. "We don't know what exists with regard to coding or pricing," said Rob Dumanois, manager of Thermo Fisher's reimbursement strategy. "That's one of the great unknowns right now … The economics are essential to our success."

In comments to the draft NCD, private payor Humana suggested CMS include billing guidance, and recommended CPT codes 81450, 81445, and 81455 be utilized when seeking payment for covered NGS services. CPT code 81450 describes a sequencing panel gauging 5-50 genes for blood cancers and is priced at $648.50; code 81445 describes a sequencing panel gauging 5-50 genes for solid tumors and is priced at $602.10; and code 81455 describes sequencing panels assessing 51 or more genes, priced at $2,920. (These prices are the 2018 national limitation amounts.)

It's unclear if CMS will use these codes, since Medicare contractors, such as Palmetto, have specifically asked NGS panel tests gauging somatic variants to not use them, except in very specific circumstances. 

Another unknown, said reimbursement and regulatory expert Girish Putcha, is whether CMS will use different codes to identify coverage for markers that fall into the three coverage pathways. For FoundationOne CDx, "if coding is based on genes with companion diagnostic claims only, one could argue that 81445, priced at around $600 is appropriate," said Putcha, who is director of laboratory science at Palmetto, but didn't provide comments for this article in that capacity.

"But if including all 324 genes, then code 81455 paid at $2,920 may be used, and, obviously, these are meaningfully different," he said. "Alternatively, Foundation could choose to use their own unique identifier," such as Proprietary Laboratory Analyses or other advanced diagnostic laboratory test (ADLTs) test codes.

Thermo Fisher is of the view that CMS shouldn't peg payment based on the number of genes included in a panel, but reimburse FDA-approved IVDs at a higher rate to reward the additional effort put forth by labs that pursue that path. In the past, labs have been able to negotiate higher Medicare reimbursement rates for FDA-approved, single-gene companion diagnostics, such as KRAS and EGFR testing.

Applied to the present NCD, "this would retain the incentive for innovation in the area for introducing new tests for oncology through LDTs and provide a rapid way of getting them to the market," Thermo Fisher's Goswami said. "But it would also retain the incentive to submit PMAs and go through the extensive and highly expensive validation required before these tests are put on the market." Labs would be more willing to pursue this higher bar, he noted, "if the reimbursement was commensurate with the level of work they go through."

Others recommended CMS increase payment under the CED pathway requiring a prospective registry. Tracking patient outcomes in a prospective registry requires substantial investment, noted Margot Schoenborn, OmniSeq's chief administrative officer. "We're certainly concerned with how pricing will be determined under the NCD, and hope CMS will take this into consideration when structuring pricing," she said.

With more granular CPT coding available to payors, differential payment of IVDs and LDTs is possible, said Putcha, though to date payors have largely been unwilling to recognize and reward differences in the evidence base underlying tests. While implementing a new Medicare pricing law for clinical lab tests, CMS proposed using a new code set — Proprietary Laboratory Analyses codes — to track test utilization and payment for ADLTs, which are multi-analyte algorithm tests or FDA approved/cleared tests performed at a single lab. The American Medical Association began accepting applications for PLA codes last year, and a number of labs wishing to more specifically identify their tests in billing have garnered these unique codes. 

The availability of these codes, which provide a level of transparency previously unavailable with CPT codes, means private and government payors are potentially out of excuses, Putcha said. "After all, shouldn’t there be a single transparent and consistent evidentiary bar for all tests with the same intended use? At this point, to do otherwise is a choice," he said.

Ultimately, with this draft NCD, CMS is trying to bring much needed transparency and consistency in the diagnostics space with regard to the analytical validity, clinical validity, and clinical utility data underlying LDTs and IVDs, reflected Putcha. "Whatever one may think about this draft NCD, which will undoubtedly be roundly criticized by multiple stakeholders and certainly has its share of challenges and required clarifications, one has to acknowledge what CMS’ coverage analysis group has done here," said Putcha. "I for one commend them for at least trying to address these structural challenges within the US diagnostics ecosystem."