NEW YORK (GenomeWeb) – The US Food and Drug Administration and the Centers for Medicare & Medicaid Services presented a unified front before Congress today, making the case that the FDA needs to play a leading role in evaluating the safety and efficacy of laboratory-developed tests (LDTs) and that CMS doesn't have the expertise or resources to take up the task as some in the lab and pathology community have suggested.
To drive this point home, the FDA release a report highlighting 20 examples when LDTs that weren't regulated by the agency caused patients harm or could have. For example, the agency pointed out how a false-positive result from a non-invasive cell-free DNA prenatal test had resulted in an unnecessary abortion and suggested that insufficiently validated ovarian cancer screening tests could have resulted in patients receiving needless surgery to remove their ovaries.
At a hearing of the House Energy & Commerce Committee, legislators questioned officials from the two federal agencies about whether regulation of LDTs should be shared between the FDA and CMS, whether one agency should take the brunt of the responsibility, or if the government should just stay out of it.
"Legislation is both appropriate and necessary to modernize clinical laboratory diagnostics oversight," Representative Raymond Green (D-Texas) said. "Legislation is a sure way of establishing a framework that will be embraced by stakeholders, avoid litigation and extended uncertainty, and foster innovation of new diagnostic tests."
Representative Michael Burgess (R-Texas), previously a practicing doctor, had reservations about a greater FDA role, however. "It's unclear to me how we can separate the practice of medicine from lab processes," said Burgess, suggesting requiring premarket approval of LDTs would potentially impose unnecessary requirements and costs.
Similarly Representative Joe Barton (R-Texas) was unconvinced that the current oversight system for LDTs under CMS' Clinical Laboratory Improvement Amendments (CLIA) was insufficient. "I'm not being a horse's rear on this," Barton said. "If it's not broke, don't fix it. It just looks to me like we're looking for ways to give FDA and CMS more authority."
The risk-based framework that FDA proposed last year in its draft guidance for regulating LDTs would phase in requirements for tests over nine years. The higher the risk of the LDT to public health, the greater the regulatory burden for the lab. According to the FDA, of the more than 11,000 LDTs on the market, 50 percent would fall into the low-risk category and wouldn't require premarket approval, while 2 percent of high-risk and 48 percent of moderate-risk tests would have varying levels of commitments in this regard.
During the hearing, Jeffrey Shuren, director of FDA's Center for Devices and Radiological Health, highlighted several innovative approaches that the agency has taken to ease the path for LDTs. For example, the agency granted 510(k) clearance in 2013 to Illumina's MiSeqDx next-generation sequencing platform with special controls as a Class II (moderate risk) device. As a result, Thermo Fisher Scientific could simply list its Ion PGM Dx next-generation sequencing system with the FDA without having to seek clearance.
Similarly, when the FDA cleared 23andMe's Bloom syndrome test earlier this year, the agency simultaneously deemed certain carrier screening tests 510(k) exempt, meaning that such tests wouldn't require premarket review if labs met special requirements. And last week, at a workshop to discuss regulation of NGS tests, the agency launched a preliminary version of a cloud-based platform, called PrecisionFDA, where researchers and test developers can compare genetic variant calls and test out data analytics platforms. In the future, FDA hopes to use this platform as a regulatory tool through which it can review information on NGS diagnostics.
"Some labs have already been working with us, and we congratulate them for crossing that picket line," Shuren told the House E&C. "But our message and invitation to the rest of the lab community is to put down the swords. For the sake of patients, it's time to end the saber rattling."
FDA pushes for the lead
Despite these attempts to show its regulatory flexibility toward LDTs, the lab industry and pathologists' groups have not warmed to the idea of a greater FDA presence in their realm. Pathologists have lobbied legislators that the agency's heavy-handed regulations would stifle patient access to critical tests and would seek to control the practice of medicine. The American Clinical Laboratory Association (ACLA) has hired lawyers to make a legal argument against FDA's statutory authority to regulate LDTs.
The FDA contends that the law gives the agency authority to regulate in vitro diagnostics, of which LDTs are a subset. However, for 40 years, the agency has regulated diagnostic kits (performed at multiple labs), leaving oversight of LDTs (performed at a single lab) to CMS under CLIA, because these tests were relatively simple and performed locally for small numbers of people. This is no longer the case, particularly in fields like cancer where across the nation oncologists are relying on LDTs to home in on the right treatments for patients.
For example, NGS tests, largely performed as lab tests, can detect numerous markers at once, while other kinds of tests employ complex algorithms to discern the varying contributions of these markers on patients' health. However, under CLIA, the clinical validity of these tests isn't checked and there is no system in place for tracking adverse events from faulty tests.
But our message and invitation to the rest of the lab community is to put down the swords. For the sake of patients, it's time to end the saber rattling.
Shuren told Congress that the agency has long been aware of the evolving LDT space and tried to regulate tests in the past, but had been thwarted by industry. The lab industry and pathologists now acknowledge that since laws regulating laboratory activities were passed in the 1980s, some labs are developing more complex tests, but maintain that the majority of LDTs use tried and tested methods that should continue to be overseen under CLIA.
Some groups have issued proposals suggesting that regulatory improvements — such as a risk-classification system for LDTs, a method of tracking adverse events, and ways of evaluating clinical validity — be made through CLIA by giving CMS added powers and resources. There would be a very limited role for FDA under these plans.
"What most of these proposals … would do is create a duplicative program under CMS and a bifurcated system leading to more inefficiency, higher costs, and still put patients unnecessarily at risk," Shuren said at the hearing.
Patrick Conway, chief medical officer in CMS's Office of the Administrator, agreed that premarket review of LDTs should be FDA's responsibility, since the agency has experience regulating IVDs, as well as the scientific expertise to evaluate the evidence underlying tests.
Conway noted that CMS is increasingly asked to do more with fewer resources, and plainly admitted that the agency would not be able to review the clinical validity of LDTs in a timely fashion. There are 25 people in CMS's central office overseeing CLIA and around 110 surveyors of labs nationally. But CMS doesn't have the medical officers, PhDs, and biostatisticians needed to review scientific evidence underlying tests.
"The challenge of managing in the CMS environment for the resources we have, for the duty we have for the American people, is by far the hardest job I've ever had," Conway said. He further noted that his agency has not asked for additional statutory powers to regulate LDTs, but given available resources, the CLIA program could continue to do what it does best, which is oversee lab operations and review the performance of tests within these labs in the post-market setting.
"As a practicing physician who works clinically on weekends, I know the importance of tests being assessed for clinical validity, as well as the need for assessment of laboratory standards," he told the House E&C. "FDA and CMS can work together, utilizing their respective authorities and strengths to assess premarket clinical validity and laboratory standards, respectively."
The House has been considering an alternative regulatory proposal from the so-called Diagnostic Test Working Group (DTWG), comprising experts from Becton Dickinson, Roche, Mayo Clinic, LabCorp, Abbott, and ARUP Labs. In contrast to the proposals from pathologists' groups, this plan would spread oversight over FDA, CMS, and the states, and would create a new category of tests, dubbed in vitro clinical tests (IVCTs), which could describe LDTs or kits. FDA would have authority over test development and validation, CMS would remain in charge of traditional lab activities necessary to perform tests, and states would oversee accuracy of test interpretation.
There wasn't much discussion of this particular proposal during the hearing. However, Shuren and Conway repeatedly impressed upon legislators that an oversight plan where CMS takes the lead in regulating LDTs would be unfeasible, since the agency doesn't have the bandwidth to take on the task.
Evidence of harm?
To make the case to legislators that FDA needs to regulate LDTs, the FDA highlighted 20 examples where tests not regulated by the agency have caused harm or could potentially cause harm, based on public reports and the literature. Among the examples are well-known cases such as Duke University researchers' chemotherapy response test, where greater oversight may have caught data analysis errors earlier, or in the case of Yale University and Laboratory Corporation of America's OvaSure ovarian cancer screening test, where more stringent oversight of questionable validation data could have kept the test from coming to market at all.
The agency also took on big names in the molecular testing space, calling out Genomic Health for reporting HER2 status as part of its Oncotype DX breast cancer recurrence testing service and Myriad Genetics' Prolaris prostate cancer aggressiveness diagnostic.
Genomic Health markets the 21-gene expression Oncotype DX as an LDT that determines the risk of breast cancer recurrence and the likelihood that a patient will benefit from chemotherapy. In 2008, the firm began reporting HER2 status as part of test reports, but subsequently, a group of pathologists reported in a study that the Genomic Health test deemed patients equivocal or negative for HER2 status when they were positive by FDA-cleared fluorescent in situ hybridization tests.
So you're going to guarantee that if we let your agency review all these diagnostic tests that something like that will never happen again?
In a more recent study, highlighted in FDA's report, the Oncotype DX missed three HER2-positive patients in a study, and as a result, two patients didn't receive the HER2-targeted drug Herceptin (trastuzumab). "The underlying issue is that there is no demonstrated direct correlation between [the] number of RNA copies of the gene, the basis for Oncotype DX HER2 RT-PCR, and the number of protein copies on the cell surface," the FDA wrote in the report. The agency estimated that it costs the healthcare system $775,278 when a patient doesn't receive Herceptin when they should.
Genomic Health spokesperson Emily Faucette told GenomeWeb that the FDA report inaccurately characterized the Oncotype DX test, since the firm doesn't market an LDT called "Oncotype DX HER2 RT-PCR." Moreover, the test isn't intended to guide treatment decisions with Herceptin, and healthcare providers perform independent HER2 analysis before ordering Oncotype DX.
"Because of the well-known discordance in HER2 testing using [immunohistochemistry] and FISH, treating physicians requested that we provide the underlying results for the quantitative RT-PCR measurement of HER2 for the purpose of providing added detail on the interpretation of the [Oncotype DX] Recurrence Score," Faucette said, highlighting that the company has studied the test in a number of clinical investigations involving several thousand patients. She emphasized that less than 1 percent of the company's tests are performed on HER2-positive tumor tissue because the test is explicitly indicated for women who have estrogen receptor-positive, HER2-negative breast cancer.
Similarly, Myriad took issue with FDA's report in which the agency asserted that the company is making claims about its Prolaris test that are not backed by clinical evidence. Specifically, the agency pointed out that although the 46-gene test is being used by urologists to guide patient care, there has been no prospective study to assess whether the decisions made based on test results are improving outcomes.
A prospective investigation to track if the test helps prostate cancer patients live longer or gets them to the right treatment would be challenging and cost-prohibitive because the disease progresses very slowly. Still, Myriad spokesperson Ron Rogers noted that Prolaris has been studied in more than 5,000 patients and is the subject of 10 scientific peer-reviewed publications, of which the FDA cites only one.
"The FDA case study white paper is not a scientific review of Prolaris," Rogers said, highlighting that CMS evaluated test validation data in awarding local reimbursement coverage and that the test is included in treatment guidelines.
Similarly, ACLA took issue with FDA's report, noting that the case studies are not representative of the thousands of LDTs that are performed accurately under CLIA and which benefit patients. ACLA further criticized the report for citing outdated data and for featuring tests no longer on the market. "The current peer-review process among highly trained professionals was effective in many of these cases," the association said in a statement.
Over the years, as the FDA has pushed to regulate LDTs, it has maintained that the lack of oversight has resulted in harm. Previously, the agency provided only a few examples, such as that of a test gauging the KIF6 variant, which some studies had shown to be associated with heart disease risk and statin benefit. A large study ultimately didn't bear out the statin benefit association, but by that time 150,000 tests had been performed, costing the healthcare system $2 billion, according to the agency.
At the hearing, Representative Barton suggested that following peer review, this test tapered off in use "without FDA having to do anything." To that Shuren responded that such a passive system would put patients at risk.
"So you're going to guarantee that if we let your agency review all these diagnostic tests that something like that will never happen again?" Barton questioned. "That y'all are perfect and all-knowing and you're going to do it in a cost effective way and there will be peace and light until judgment day?"
Representative Burgess also questioned whether the scope of the problem was really as dire as FDA presents in its report. "It took you three years to provide us with 20 [examples]," said Burgess, who at prior hearings has asked the agency to furnish evidence that LDTs cause patient harm regulated under CLIA.
Representative Kurt Schrader (D-Ore.) similarly cautioned FDA against giving the public the impression that the agency's stamp of approval means tests will never fail. "A test is a secondary, adjunct [tool] to help establish the diagnosis," Schrader said. "There are false positives all the time in every single test … I just don't want the consumer to be misled that by having FDA approval that test is going to be 100 percent."
"I am concerned that the tone here is that we're going to put certainty in the art of medicine, when there isn't that much certainty," reflected Schrader, who is a veterinarian.
Throughout the hearing, Shuren tried to communicate to legislators that it's not that FDA regulation would ensure that all tests are accurate in every instance, but that a much-needed system would be in place to identify when a test hasn't been validated appropriately, when it fails to perform as it should, and when patients are harmed. "You're right it's not 100 percent," Shuren said. "But what we assure is that information is made available to the practitioner."