Skip to main content

Labs Mull New Regulatory Path for NGS Tumor Panels in Wake of FDA Decision on MSK-IMPACT

Premium

NEW YORK (GenomeWeb) – The US Food and Drug Administration's authorization last week of a 468-gene tumor profiling test puts into play several regulatory strategies that the agency has been mulling for some time, but it remains to be seen if other cancer centers and diagnostic companies with similar cancer panels will be willing to pursue the new pathway.

The agency authorized Memorial Sloan Kettering Cancer Center's MSK-IMPACT (Integrated Mutation Profiling of Actionable Cancer Targets) through the de novo premarket review pathway as a Class II, moderate-risk device, and simultaneously made the New York State Department of Health a third-party reviewer of IVDs, including similar tumor profiling assays. Subsequently, other labs can apply to the FDA for 510(k) clearance — a less onerous path than premarket approval — for their tumor profiling NGS panels. Or, if the test already has approval through the NYSDOH, sponsors can submit that application to the FDA and ask the state regulator to forward its review documents and recommendations.

As a cancer center serving New Yorkers, MSK had extensive experience with the state regulator, having garnered its approval for some 350 lab tests. NYSDOH approved an earlier version of MSK-IMPACT that gauged 341 genes three years ago, and MSK continued to submit more validation data as it added more genes to the panel. "From the perspective of additional work involved, it was less for us than it would be for a laboratory that does not have pre-existing New York state approval," said David Klimstra, chair of the pathology department at MSK.

FDA's authorization for MSK-IMPACT carves a path forward for how the agency will evaluate other NGS tumor profiling tests, which are increasingly used by oncologists to guide treatment strategies for cancer patients. Whether labs will take that route is not clear at this point, particularly since FDA is continuing to exercise enforcement discretion over lab-developed tests, leaving such tests to be overseen by the Centers for Medicare & Medicaid Services in line with federal standards in the Clinical Laboratory Improvement Amendments (CLIA).

"As long as there is a parallel LDT pathway, it is unclear to me what the advantage would be to pursuing FDA clearance for a tumor profiling panel," said Gail Javitt, a lawyer in Epstein Becker Green's healthcare and life sciences practice. Beyond the considerations of the added time and expense of taking a test through FDA, diagnostics developers are making a "lifetime commitment" in becoming an FDA-regulated manufacturer, she noted, which obligates them to activities far beyond a one-time submission, including registration, inspection, adverse event reporting, and maintaining a quality system.

At the Personalized Medicine Conference in Boston last week, Sean Khozin, associate director of FDA's Oncology Center of Excellence, said the agency may try to resolve the longstanding kerfuffle over LDT regulation through legislation based on the outcome of pre-certification pilot programs.

"There's a shift that’s occurring right now and this is being piloted [in the digital health space] … where we turn the focus away from the product itself and look at the manufacturer and the processes employed to develop the product," Khozin said, explaining the idea that when manufacturers become certified to provide certain products, then the products don't necessarily have to go through the analytical validation process. "The pre-certification would serve as that stamp of approval." 

If the pilot program being conducted in the digital health space is successful, then "this may be a framework that can be used in a legislative solution for regulation of … laboratory-developed tests," he added. The FDA will provide additional details on this approach in the coming months, he said.

A path forward

MSK's decision to pursue FDA authorization for its test came out of broader discussions with the agency around regulation of LDTs, a hotly debated topic within the lab industry for 20 years. The debate became particularly contentious three years ago when the agency decided to release draft guidelines laying out a framework for bringing LDTs under its purview. Right after the 2016 presidential elections, the FDA announced it wouldn't issue final guidelines and instead released a paper that contained several ideas for regulating LDTs that it had come up with after extensive discussions with industry and other stakeholders. 

The document, while not enforceable, was received favorably by the American Clinical Laboratory Association, perhaps the most vocal critic of the agency's initial draft plan to regulate LDTs. In authorizing MSK-IMPACT, the FDA tested out many of the ideas in the discussion paper, such as pre-specifying a plan for making changes to the test that would limit supplemental regulatory submissions, and making NYSDOH a third-party accrediting body. 

"The FDA wanted to provide an example of this new mechanism for LDTs to get 510(k) clearance," said Marc Ladanyi, chief of MSK's molecular diagnostics service. "We were getting a lot of encouragement from the FDA in this process."

MSK-IMPACT is not a companion diagnostic, which FDA considers high-risk tests required for the safe and effective use of drugs. The agency authorized MSK's test as a tool for detecting somatic mutations and microsatellite instability, information that doctors can use according to professional guidelines. However, the agency noted that doctors shouldn't treat test results as "conclusive" information about whether or not to prescribe a drug.

"The results of this test are to be interpreted by knowledgeable treating physicians, in the context of all other medical information about the patient's disease," Klimstra said. "It does require a level of understanding for what these alterations mean for cancer treatment. This is not an 'if this, then that' kind of a scenario."

Under CLIA, only the analytical validation process for LDTs is evaluated, which was one of the reasons FDA wanted to step in, to regulate LDTs for clinical validity as well. NYSDOH states that it reviews analytical and clinical validity of tests, but the evaluation of analytical validity is more intensive, requiring multiple studies. A 57-page FDA decision summary document shows that the agency reviewed study data on MSK-IMPACT's accuracy, analytical sensitivity and specificity, limit of detection, precision, reproducibility, and coverage. 

For clinical validity, the state regulator accepts evidence from the published literature, whereas the FDA has historically required sponsors to conduct clinical trials.

When FDA initially proposed to regulate LDTs, one of industry's concerns was that they'd have to conduct a new study every time a new gene and variant was added to a test. This would be impossible for large panels like MSK-IMPACT. FDA's decision summary document includes a list of common hotspot cancer mutations for all 468 genes on MSK-IMPACT; a list of gene names  on the panel and their transcript IDs; and a list of exons excluded from reporting due to low coverage.

In advancing this new pathway with MSK, the FDA tried to mollify industry's concerns in this regard by creating a three-tier risk framework for biomarkers assessed by tumor profiling tests. The highest level of risk is for companion diagnostics, for which test manufacturers must show analytical validity for each biomarker and perform a clinical study demonstrating the link between the test results and patient outcomes or concordance to another approved CDx. 

Level 2 biomarkers, meanwhile, are "cancer mutations with evidence of clinical significance" that doctors can use in the care of cancer patients in line with guidelines and other information. Sponsors must demonstrate analytical validity and can point to published evidence on clinical validity. Level 3 biomarkers are "cancer mutations with potential clinical significance," which can help direct patients to clinical trials. Sponsors have to submit analytical validation data on a subset of these markers and provide a clinical or mechanistic rationale from peer-reviewed publications or preclinical models.

The FDA will not require additional submissions when sponsors want to add more variants of the same type within genes that are already analytically validated. Further, mutations can be moved from Level 3 to Level 2 without requiring additional submissions.

"The interpretation of sequencing data is always done in the context of a specific patient, and a variant might be a [Level] 3 for one patient, given the other clinical information, but [Level] 2 for another. It is not a situation in which the clinical community wakes up one morning and says, 'Time to move variant X from Level 3 to Level 2,'" Javitt reflected. "Given that reality, I guess it is a good thing that a new submission will not be required to move from [Level] 3 to 2, as that would delay a laboratory’s ability to start reporting a new variant as evidence emerges."

According to Klimstra, the MSK-IMPACT test report will be organized as reporting Level 2 or Level 3 markers. These levels are a condensed version of how MSK discusses the clinical significance of somatic variants in OncoKB, a publicly available evidence curation resource that the cancer center uses to interpret the mutations detected by MSK-IMPACT. 

"This information changes on a weekly basis as additional data come forward. We are in a situation where for certain tumor types, the oncologist has to have a deep understanding of the therapeutic significance of various mutations and stay up to date on the literature," Klimstra said. "We try to provide some direction in that regard."

In approving Myriad Genetics' BRACAnalysis as a companion diagnostic, the agency reviewed the primary data within the company's proprietary database of BRCA1/2 genetic variants. However, the FDA has also envisioned that public variant databases could be a source of clinical validity information on biomarkers in NGS panels, and issued a draft guidance on the topic. As such, OncoKB is a public variant resource that relies on external data sources, such as drug labeling, guidelines, expert recommendations, and the medical literature. 

MSK described this resource in a Journal of Clinical Oncology – Precision Oncology paper in July. "I believe the FDA reviewed the overall approach for OncoKB and confirmed that its approach to ranking actionability is consistent with [its own], albeit with more granularity and inclusion of more indirect evidence," Ladanyi said. 

The other novel strategy that FDA put into play with the MSK-IMPACT authorization is a streamlined process for updating the test when anticipated modifications are pre-specified. As long as MSK validates these changes and meets "pre-specified success criteria," a new 510(k) won't be required. However, "significant changes such as adding new genes or variant types to the panel would require a new submission with appropriate validation," the agency stated.

MSK-IMPACT utilizes Illumina's HiSeq 2500 sequencer and gauges all variant types. However, the present authorization is only for detection of point mutations and small insertions or deletions less than 30 base pairs in length in the coding regions of 468 genes, as well as microsatellite instability.

The test also gauges copy number variants and certain cancer fusions, Ladanyi said. After NYSDOH approves the test for these variants, MSK will follow up with FDA submissions. In addition, MSK-IMPACT assesses tumor mutation burden, and MSK plans to submit to NYSDOH and eventually the FDA regarding this aspect of the test.

Accrediting NYSDOH

While FDA was reviewing MSK-IMPACT, NYSDOH, which has reviewed more than 11,000 new and modified LDTs over the last decade, became accredited as a third-party reviewer in June. As part of the application process, the agency's Wadsworth Center had to submit information on its demographics, infrastructure, and qualifications, as well as its process for handling conflicts of interest.

The NYSDOH last year issued a three-tier, risk-based framework for evaluating and approving laboratory-developed tests. This was the protocol it used to review and approve MSK's test. 

It remains to be seen how other influential groups in the molecular diagnostics space, such as the College of American Pathologists (which also accredits labs) and the Association for Molecular Pathology will interpret the new pathway paved by MSK and NYSDOH. Both these groups have released LDT regulatory proposals that describe a limited role for FDA regulation of certain high-risk tests and employ a CLIA-centric approach to improving oversight. 

"AMP members share a common goal of putting the patient first and preserving broad access to these types of essential laboratory-developed procedures," AMP Executive Director Mary Steele Williams said in a statement commenting on the FDA's approval of MSK-IMPACT. "We feel our CLIA modernization proposal, which updates the current regulation program and expands the network of third-party medical experts, is still the most effective oversight solution for patients, healthcare providers, clinical laboratories, and the entire field of molecular diagnostics."

Industry advantage?

Labs have given a number of reasons for not seeking premarket review from the FDA, including that the process is burdensome, expensive, and not attuned to lab processes. Others have maintained that because labs are already regulated under CLIA, FDA oversight is duplicative and unnecessary.

The few labs and diagnostic companies that have braved FDA review have described it as an immense undertaking, despite the agency's attempts to work with industry and streamline processes. For example, after garnering approval for a 23-gene companion diagnostic panel, the Oncomine Dx Target Test, Thermo Fisher Scientific's Joydeep Goswami said the entire process culminating in a 220,000-page data package to the agency was "a huge burden." 

Like others in the life sciences space, Ladanyi had expected the experience to be far more challenging than it was. The agency asked for a few things not in MSK's NYSDOH submission, such as additional data on reproducibility at the limit of detection for the assay, using known positives with low variant allele frequencies. But, "on the whole, it was not an onerous process," Ladanyi said.

Klimstra estimated the entire process for FDA market authorization took six months, two of which the agency spent mulling the strategy, and the rest of the time MSK and the agency spent having weekly meetings and intense discussions. "We were learning together," he said. "This was a true collaboration."

Although FDA review is considered an expensive process, requiring several hundred thousand dollars in filing fees depending on the size of the manufacturer, MSK paid nothing in this instance. The agency waived the fee for this initial submission — the waiver contingent on MSK meeting a filing deadline — because it was the pilot for a new pathway, according to Klimstra. "New Yorkers know how to cut a good deal," he quipped, but he noted that submissions for updates to the panel would cost several thousand dollars.

"The process that we went through was much less onerous than what we feared after reading FDA's initial draft guidance," he said. "Perhaps the FDA felt that going through this exercise with a lab like ours, a New York lab with high visibility as an academic leader, would show that this process could be done without a crippling expense or time commitment."

However, labs that have never undertaken NYSDOH approval might feel differently. "Documents that go to New York are like a phone book and take months," Klimstra said.

According to an NYSDOH spokesperson, the state regulator has approved 32 solid tumor panels from 22 labs. These labs may have more reason now to consider garnering FDA's blessing.

Roswell Park Cancer Institute spin-out OmniSeq, for example, has NYSDOH approval for its 144-gene tumor profiling panel and Immune Report Card, which analyzes biomarkers that can help doctors make decisions about immunotherapy. Previously, the vast majority of pathologists have had no interest in pursuing FDA review for LDTs, acknowledged Carl Morrison, OmniSeq's CSO and lab director.

"On the other hand, pharma will see this as the first step in standardization of lab tests," Morrison said. "It appears that we will now have three basic types of lab tests that include FDA reviewed [tests], LDTs [that are] FDA approved by a third-party reviewer, and laboratory-developed procedures that are not reviewed by FDA."

OmniSeq CEO Mark Gardner added that the company is now "seriously considering whether to take our assays to the FDA to enable our pharma partners to garner more assurance of regulatory stability and scalability."

Meanwhile, the Moffitt Cancer Center in Florida was less enthusiastic about pursuing this new path for the 170-gene solid tumor panel and the 54-gene myeloid malignancy panel it performs in its molecular lab. The cancer center currently doesn't offer these tests to New York patients, and therefore has no reason to pursue NYSDOH approval. But if Moffitt wanted to offer testing more broadly, then the cancer center would likely have to garner NY state approval for legal reasons.

"I am not against an institution going through the process of FDA clearance for the test, if they have the extra funds to do so," said Howard McLeod, chair of Moffitt's Department of Individualized Cancer Management. "At the moment, there is not a requirement for us to go that route, there is not an advantage to our patients, and there is not a reimbursement consequence either way. Should any of those aspects change, we are grateful that MSK has provided a regulatory path to use."

For MSK, whether FDA market authorization translates into a reimbursement benefit also remains to be seen. The cancer center has performed MSK-IMPACT on 20,000 patients and the test is "selectively reimbursed at this point," Klimstra noted.

"It is true that insurers do sometimes look for the imprimatur of FDA review in making coverage decisions, but that is not always the case, and I think would be hard to justify in a context where most panels do not have clearance," Javitt said. "Ultimately, insurers want to see evidence that performing a test has a meaningful benefit for a patient, [for example for] selecting an effective drug, or avoiding an ineffective drug."

The possibility of netting national Medicare coverage for its FoundationOne solid tumor profiling panel was likely the draw for Foundation Medicine in pursuing parallel review from FDA and CMS. If the strategy proves successful, FoundationOne will receive simultaneous FDA market authorization and a national coverage determination from CMS for its NGS panel. The company is expecting a decision by year end. 

While some in the molecular diagnostics space do appear more open to seeking FDA approval of advanced tests, and are even willing to credit the agency for thinking forward, the belief that regulation will always play catch-up to innovation endures. "Panels are a pragmatic, necessary evil, but they're a dead end," Marc Williams, director of Geisinger's Genomic Medicine Institute, said at the Personalized Medicine Conference in Boston last week, reacting to the news that FDA authorized MSK-IMPACT.

The integrated health system is sequencing the exomes of participants in its MyCode Community Health Initiative and reporting actionable results in 76 genes. "We're going to be using [panels], but ultimately we need access to genomes, and we need to have them on the tumor and on the germline," Williams said.