NEW YORK – A Medicare contractor's proposed coverage criteria for next-generation sequencing tests for personalizing cancer treatment may significantly improve reimbursement prospects for some academic cancer centers and health systems with large in-house testing panels.
National Government Services issued a draft local coverage determination, or LCD, in September outlining the circumstances in which it will cover NGS tests that assess more than 50 genes for patients with advanced non-small cell lung cancer, colorectal cancer, and other solid tumors. The proposal stands to positively impact NGS testing reimbursement for labs in National Government Services' jurisdiction, which spans New York, several New England states, Illinois, Wisconsin, and Minnesota, though stakeholders have questions and concerns about certain aspects of the policy as written and have requested clarifications.
"For the last several years, reimbursement for molecular testing in clinical labs was really bad," said Eric Loo, a pathologist at Dartmouth-Hitchcock Medical Center, a large academic health system in New Hampshire that conducts an in-house 170-gene solid tumor panel on the Illumina TruSight platform. "[Reimbursement] has improved in the last few years, but reimbursement for the larger sequencing panels is still something of a gray area."
Medicare coverage for NGS cancer tests began improving after the Centers for Medicare & Medicaid Services' national coverage determination, or NCD, in 2018. That decision established Medicare coverage for NGS tests that the US Food and Drug Administration has approved as companion diagnostics to guide treatment decisions for advanced cancer patients. However, CMS left it up to its Medicare contractors to come up with policies covering laboratory-developed NGS cancer tests without FDA approval.
The NCD made reimbursement for large, FDA-approved NGS panels, like those offered by Foundation Medicine and Guardant Health, more reliable, spurring their use, though local Medicare coverage of lab-developed NGS panels without regulatory approval continues to vary around the country based on contractors' policies. For example, the MolDx program, which provides genetic testing coverage policies followed by Medicare contractors Palmetto, Noridian, CGS and Wisconsin Physicians Services, has issued fairly broad criteria for large NGS cancer panels that gauge more than 50 genes in solid tumors and blood cancers. These four Medicare contractors manage coverage for molecular testing that occurs in more than half the US states.
National Government Services, by comparison, presently covers smaller genetic testing panels for advanced NSCLC and colorectal cancer patients who meet certain criteria. That's why the National Government Services' draft LCD is a "big deal," said Rob Dumanois, director of reimbursement strategy at Thermo Fisher Scientific, which sells distributed NGS kits, ranging for smaller targeted panels to a 500-gene panel, to health systems and labs to implement in house.
"The only NGS testing that's covered [by National Government Services] for somatic mutations for late-stage solid tumor indications are either small panels [for] five to 50 genes for non-small cell lung cancer or colorectal cancer, or [Memorial Sloan Kettering's] MSK-IMPACT test," Dumanois said. "That's it. So, the starting point in Medicare … is certainly well behind where lots of other jurisdictions are."
In its draft LCD, however, the contractor seems to be "guaranteeing" coverage of NGS panels that analyze between five and 50 genes for advanced NSCLC and colorectal cancer patients, said diagnostic reimbursement expert and consultant Bruce Quinn. But the contractor is now also saying any patient with an advanced solid tumor, including advanced NSCLC and colorectal cancer patients, "may" be covered for testing on a larger panel that gauges more than 50 genes, if smaller panels are "insufficient."
The document has garnered the attention of lab professionals, particularly in New York and New England, where there is a high concentration of academic institutions and health systems with in-house NGS capabilities. Although most oncologists still order NGS panel testing from outside commercial labs, more healthcare institutions around the country are installing in-house testing out of a desire to lower costs, shorten result turnaround time, and improve precision oncology treatment and trial access for the local community.
Massachusetts General Hospital, for example, runs a number of targeted NGS cancer panels that may be impacted by National Government Services' proposed coverage policies. At a public meeting hosted by the contractor last week, Joe Lennerz, medical director at MGH's Center for Integrated Diagnostics, extended support for the policy. "Instead of running separate individual tests in individual cancers one at a time, comprehensive genotyping is the most efficient and scalable solution that will keep pace with new advancements in oncology and cancer therapeutics to improve patients' outcomes," Lennerz said.
"On behalf of industry, we view this as a really good thing, and you never want to be too critical of what's fundamentally a good thing," Dumanois agreed in an interview. "It's just the way that this thing was written, depending how you want to look at it, there are [several] areas that have most of us scratching our heads."
Define 'comprehensive'
In the draft LCD, the contractor states that "genomic sequential analysis panels" will be considered reasonable and necessary for newly diagnosed and previously treated advanced NSCLC patients, as well as certain metastatic colorectal cancer patients who are considering anti-EGFR treatment. National Government Services then goes on to propose coverage for bigger NGS panels, commonly called comprehensive genomic profiling, or CGP tests, for advanced or refractory cancer patients with solid tumors when single-analyte tests or five to 50 gene panels are "insufficient."
The contractor defines "comprehensive" genomic profiling tests as those that detect single nucleotide variants, insertions/deletions, copy number alterations, and structural and splice-site variants in "hundreds of genes." Such tests can assess mutational patterns, such as mismatch repair deficiency (dMMR), microsatellite instability (MSI), and tumor mutational burden (TMB), and may even conduct RNA sequencing, according to the draft LCD.
"AMP commends National Government Services for recognizing the clinical value of comprehensive genomic sequencing in certain patients," Samuel Caughron, chair of the Association for Molecular Pathology's Economic Affairs Committee and CEO of MAWD Pathology Group, said in an interview. However, the association is worried that the lack of details in the draft LCD on when smaller panels are "insufficient" may lead to overuse of large panels. "It is not clear from the policy how a provider would justify the insufficiency of smaller panels," Caughron said.
Although the draft LCD language is "vague" on this key point, one interpretation could be that "any patient who needs TMB is a patient for whom a five- to 50-gene panel test is insufficient," Quinn said. "If that was the intention, it could be said clearly in just a few words."
More oncologists are ordering large NGS panels that can calculate MSI, dMMR, TMB, and NTRK fusions, since refractory cancer patients with these tumor biomarkers can now receive immunotherapy and targeted drugs regardless of the type of cancer they have. National Government Services' draft LCD recognizes the need for broader NGS testing in light of the FDA's approval of several cancer drugs with tissue-agnostic indications, including Merck's PD-1 inhibitor pembrolizumab (Keytruda), and TRK inhibitors larotrectinib (Bayer's Vitrakvi) and entrectinib (Genentech's Rozlytrek).
"NGS CGP testing may especially benefit the 25 percent of patients with rare cancer types, those with no effective alternative treatment options, and underserved minority populations with less access to tumor molecular profiling or off-label therapies," the contractor wrote in the draft policy, adding that although there is conflicting literature on the value of such testing, due to the "widespread societal support for NGS CGP testing in advanced cancer, National Government Services reservedly deems such testing appropriate for advanced somatic cancers."
However, at the meeting to discuss the draft LCD, lab professionals and pathologists suggested that the contractor better define what it considers "comprehensive" genomic profiling for the purposes of coverage and billing. Currently, the draft LCD appears to deem tests "comprehensive," if they gauge hundreds of genes and enable calculation of biomarkers like TMB. At the same time, the contractor specifies in the draft LCD that circulating tumor DNA and germline testing are out of the scope of its policy.
At the meeting, Calvin Chao, senior VP of medical affairs at Tempus, which provides a range of NGS services for cancer patients, pointed out that the Chicago-based lab conducts comprehensive genomic profiling of patients' tumor and normal samples to better distinguish inherited alterations from those driving the tumor. The LCD, as written, likely would not reimburse the lab for germline analysis, even when it is conducted to better distinguish targetable somatic alterations.
Additionally, "ctDNA can be used to assess [actionable biomarkers in] solid tumors, especially in instances with insufficient tissue obtained and [when] the patient is unable to undergo a procedure to obtain the tumor sample," Caughron said, adding that AMP will ask the contractor to remove the coverage limitation for ctDNA testing.
Coding clarity
Loo from Dartmouth-Hitchcock, who also spoke at the public meeting on behalf of the New Hampshire Society of Pathologists, further pointed out that the CPT code National Government Services is asking labs to bill for comprehensive NGS panels (81455) describes tests that gauge more than 50 genes but may not align with the contractor's conception of a comprehensive panel as one that can calculate TMB.
According to experts, panels must assess alterations in at least 300 genes in order to calculate TMB. The 170-gene test that Dartmouth-Hitchcock performs in house is technically a comprehensive genomic panel that can be billed using CPT code 81455 because it gauges more than 50 genes, but the panel cannot determine patients' MSI or TMB.
"Tests that are coded under 81455 can be extremely heterogeneous in size and composition," said Caughron over email, echoing similar concerns on behalf of AMP. "While some larger panels are capable of assessing things like microsatellite instability, copy number, and tumor mutational burden, it is incorrect to assume all assays under the 81455 designation have the same capabilities."
Once a lab bills Medicare for a "comprehensive" test using CPT code 81455, according to the draft LCD, it cannot bill for other molecular tests for the same patient, even if the first test wasn't comprehensive enough to gauge MSI or other biomarkers. In the billing and coding instructions that accompany the draft LCD, National Government Services states it will "automatically deny" any other molecular testing CPT codes that labs bill alongside CPT code 81455 on the same date of service for a patient.
For example, if Dartmouth-Hitchcock's lab performs its 170-gene panel on a patient's sample and other PCR or immunohistochemistry tests to determine MSI or dMMR, National Government Services would deny coverage for those additional tests under the proposed policy. "That doesn't seem quite fair, given the heterogeneity of testing that would fall under this particular CPT code," Loo said at the meeting.
Moreover, large NGS panel tests can take around a month to return results, and amid such delays, doctors may want to do single-gene tests in case those can help get their patients quickly on a treatment. Michael Kluk, a pathologist at Weill Cornell, told National Government Services at the meeting that its policy prohibiting reimbursement for such ancillary testing could be detrimental to patient care.
Duplicative requirements
The lab and pathology community also asked National Government Services to provide more clarity on the process for determining if a comprehensive NGS panel analyzes the minimum, therapeutically actionable genes for coverage. According to the draft LCD, the test must at least assess biomarkers listed in Category 1 or 2A of the latest version of the National Comprehensive Cancer Network's Biomarkers Compendium. While NCCN can change guidelines overnight and recommend oncologists test for new actionable genes, it takes labs much longer to add them to existing panels and re-validate tests.
"The addition of a new biomarker [to guidelines] could potentially invalidate [coverage for] a clinical assay," Caughron said. "AMP would like to see this requirement refined to ensure that the contractor would allow clinical laboratories sufficient time to prepare or revalidate an assay if a new biomarker was added to this compendium."
Lastly, many in the lab community have raised concerns about National Government Services' proposal that labs must publish analytical validation data on their NGS panels in peer-reviewed journals if the tests aren't FDA approved. The publication of validation data is not a requirement under CLIA — the federal standards that labs must meet to perform high-complexity testing on humans — and AMP plans to ask National Government Services to remove this requirement.
"Regulatory requirements stipulated in CLIA already provide strict validation requirements that must be followed before an assay can be offered to patients," Caughron said. "Mandating publication of peer-reviewed studies as a condition to coverage is burdensome and duplicative to existing requirements under CLIA and third-party accreditors, such as the College of American Pathologists."
New York has some of the most stringent premarket evaluation processes for lab tests among US states, and Weill Cornell's Kluk and others working in labs in the state told National Government Services at the meeting that its requirement to publish validation data is onerous. "Our validations are reviewed by the clinical laboratory evaluation program of the New York State Department of Health," said Mahesh Mansukhani, a pathologist at Columbia University Irving Medical Center. "And that should be considered as equivalent to a peer-reviewed publication."
Despite the requests for more details and clarity on the draft LCD, industry observers seem generally optimistic about National Government Services' willingness to provide broader coverage of large NGS panels, since the policy may encourage further adoption of in-house NGS cancer testing within health systems and academic institutions, allowing them to offer precision cancer treatment to more patients in the communities they serve.
Loo noted that even in a difficult reimbursement environment for large NGS panels, institutions lose less money with in-house testing than when doctors order it from an outside commercial lab. Moreover, in-house testing allows the next generation of providers to become comfortable with precision medicine. "If we don't provide the testing in house, the physician or pathologist trainees will not get practical exposure to the testing methodology. They really need to be exposed to all the nuts and bolts and testing pitfalls you may run into when providing high-complexity tests," Loo said. "Otherwise, they definitely won't be qualified to provide these services in the future."