Mixed Opinions On How To Regulate Laboratory Developed Tests
A trio of sessions at the 2019 Next Generation Dx Summit in August focused on laboratory developed tests (LDTs)—in vitro diagnostic tests designed, manufactured and used within a single laboratory—and incited lively debate about proposed regulation under the VALID (Verifying Accurate, Leading‑edge IVCT Development) Act, whether current oversight approaches have been enhancing or restricting innovation, and how to ferret out bad actors.
A regulatory and reimbursement consultant even offered up a top 10 list of what labs and software developers need to know about the U.S. Food and Drug Administration (FDA) when making LDTs or software for consumers and patients.
In a plenary keynote discussion on proposals and solutions for diagnostic reform, representatives from six organizations presented their differing view on the VALID Act, which proposes a new framework for clinical lab diagnostics. The legislation includes a marketing pathway enabling eligible technology developers to obtain precertification for a single test method.
Julie Khani, MPA, president of the American Clinical Laboratory Association (ACLA), says the ACLA’s top priorities are to protect Medicare clinical lab reimbursement provisions, represent members on Capitol Hill and in the courts, and foster a legislative solution to comprehensive reform for clinical lab diagnostics. The VALID Act is currently a discussion draft and the ACLA has provided detailed comments on every version of it.
The current VALID draft has provisions that are “critical to success” from the perspective of labs, says Khani, including grandfather provisions and precertification for lower-risk tests. But a reasonable transition period is needed.
So is avoiding duplication of the Clinical Laboratory Improvement Amendments (CLIA) under which the Centers for Medicare & Medicaid Services (CMS) oversees laboratory operations for LDTs, says Khani, noting some overlap with VALID in terms of purchasing control and personnel standards. It’s good policy to have public confidence in all tests, both LDTs and FDA-approved test kits, and the conversation needs to include the FDA, CMS, CLIA and legislators on the Hill.
AdvaMedDx, a trade association for manufacturers of in vitro diagnostics (IVD), has likewise commented extensively on VALID, says Executive Director Susan Van Meter. The group favors legislation that is comprehensive, risk-based, diagnostic-specific and predictable. The precertification program looks promising, she says, but may prove to be less advantageous than expected. “We see significant areas where improvements can be made,” she adds, noting the impact on patient access to tests is uncertain.
Precertification is “intriguing, but the risk categories are a little unclear,” says Tara Burke, senior director of public policy & advocacy at the Association for Molecular Pathology (AMP). She says AMP appreciates that the sponsoring legislators are open to stakeholder input but wants to understand the “nuances” of what’s being proposed.
One concern is how the VALID Act aligns with or duplicates provisions of the CLIA, Burke says. Another is that the list of “mitigating measures” fails to account for the role of medical professionals in designing and validating LDTs. CMS has also been noticeably absent in the dialogue.
“Even if VALID passes, CLIA is still critical to how labs get regulated, so it’s important to understand where there is duplication,” Burke says. Adverse events reporting is one such area—CLIA requires it and VALID, as written, makes it difficult to parse out what the FDA expects should be reported. No amount of regulation would likely stop the gross level of fraud perpetuated by Theranos, she says.
The diagnostic reform interests of Friends of Cancer Research (Friends) stem from work in targeted therapy development that requires companion diagnostics, says Laura Lasiter, a science policy analyst. “Patients don’t need access to a drug but to the right drug and that depends on an accurate test result.”
Evidence suggests there may be tangible clinical and economic value to having LDTs subject to FDA approval, bolstering the case for uniform diagnostics regulatory oversight, Lasiter points out. “It’s really difficult to quantify if harm is being done to patients under the current regulatory framework. We don’t even know how many LDTs are out there.
“I’m not saying laboratory developed tests are bad tests,” Lasiter continues. “At a reputable cancer center an LDT may be better than an FDA-approved test, but currently there’s no way to know that. We need transparency… and more information.”
Friends would like to see a regulatory pathway that will align with the development timeline for both the FDA’s Breakthrough Therapy designation and Fast Track process, says Lasiter. The VALID Act proposes that diagnostic developers “pre-specify” planned device modifications with the FDA so they don’t have to later re-apply for additional premarket approvals, and that will be one of the topics of conversation with the agency in November.
Group labeling of companion diagnostics appropriate for use with a specific group or class of therapeutic products may help reduce some of the confusion, says Lasiter. The FDA has draft guidance to this effect that was issued last December.
The economic impact of the VALID Act on labs was top of mind for Donald E. Horton, Jr., senior vice president, global government relations & public policy, at Laboratory Corporation of America Holdings (LabCorp). “We process 500,00 specimens every day,” he says.
Much work remains to be done on the VALID draft, says Horton, echoing the other panelists. The financial fallout from the legislation will look different for labs depending on whether they offer a handful of tests or thousands of them and if they operate in a state (notably New York) that already subjects them to an added layer of regulatory scrutiny.
The VALID Act means labs will need to develop a lot of new standard operating procedures, spend a long time on employee training and make “difficult choices about the tests they offer because they may be economically unfeasible to do,” Horton says. “A lot of policymakers have not considered this, which is why we need balance [between regulation and innovation] and to ensure patient access to valid tests.”
At a separate panel session at the Next Generation Dx Summit, the former director of the Office of In Vitro Diagnostics and Radiological Health in the FDA’s Center for Devices and Radiological Health offered a view from inside the agency. Alberto Gutierrez, now a consultant with NDA Partners, says that the FDA’s bar for measuring risk with IVD screening devices is based on what happens to patients if they get a wrong result and how often that happens. Clinical validity, for better or worse, is gauged partly by the judgment of physicians.
Which tests have utility for individual patients is another matter, adds Paul Gerrard, chief medical officer and MolDX (Molecular Diagnostic Services) director of clinical science at claims clearinghouse Palmetto GBA. Medicare beneficiaries have no copay for lab tests, so they may be easily swayed to get relatively noninvasive tests without any real utility. The challenge is to “not get in the way” of individual physicians who have the best information on patients, which comes with some downside financial risk.
“As an academic physician, I worry about tests without utility,” says Eric Konnick, assistant professor, associate director, Genetics and Solid Tumors Laboratory, University of Washington. “My medical license is on the line. The problem is we have a closed box system that no one is really monitoring. We get results we don’t expect and can’t talk about the utility of the test sometimes.” Over-regulation, however, would simply result in 5% of patients eating up 90% of physicians’ time.
The FDA sometimes “goes too far,” Gutierrez says, adding that the agency has no desire to limit the use of diagnostic tests and drugs. The issue is having enough controls to stem abuses. “We think there’s a happy medium here. You can’t separate the mechanics of tests from what it does if you don’t know how well the test performs.”
When it comes to unmet needs, the perspective of test developers is that diagnostics can provide patient benefit even if they hit a lower level of specificity, says Brad Spring, vice president of regulatory affairs at BD Diagnostics. Companies rely on key opinion leaders to decide what’s “good enough.” Market access is also critical to meeting unmet needs, because for-profit companies aren’t going to make tests they can’t get paid for.
Hospital systems handle all comers, Konnick interjects, and so they don’t always have the luxury of saying no to things that don’t make economic sense.
Industry, for its part, spends “tens of millions” on clinical studies to develop often low-volume devices addressing unmet needs, notes Spring. As the FDA is aware, and expected to factor into its regulatory approach moving forward, the costs of research and development, lack of profitability for industry, and challenges of conducting trials in small, heterogeneous populations stand in the way of progress in this area.
FDA mandates and policy can sometimes hamper efforts to protect the public, according to Gutierrez. For example, the agency has instructed companies to remove information about susceptibility test interpretive criteria (“breakpoints”) from the labels of antimicrobial drugs and testing devices. Critical updates are instead being posted online to inform the prescribing decision of healthcare professionals. “[Companies] can’t change labels fast enough… especially when technology is changing rapidly like it is in diagnostics.”
Gerrard, who believes some labs may be driving business to themselves, says “identifying bad actors is itself a challenge.” MolDX attempts to minimize the burden by categorizing molecular and genetic tests into a registry so they can be quickly identified, along with the evidence backing their use—”what the good actors do anyway.”
Precertification, as allowed in the VALID Act, is “a way to reward the good actors” and perhaps be an incentive for “competitors to tell on other competitors” behaving fraudulently, says Spring. “I don’t like it when I see one bad actor [upend] an entire regulatory process,” citing the breast implant scandal that prompted the new European Medical Devices Regulation (MDR). “[MDR] will eliminate entire companies and not help the problem it is trying to solve.”
In cases where there’s an intent to defraud, as was the case with Theranos, “the system worked but slower than it should have,” says Konnick. “On the ‘omics side, the vast majority of labs do a good job,” with higher risk correlating with greater information opacity. “Why not just update the thing [CLIA] that has been working for decades versus a new regime [VALID Act]?”
“I don’t think CLIA works very well,” says Gutierrez, because the FDA lacks the expertise to determine if a test is worth doing—regardless of the articles published or certification by the College of American Pathologists. “Part of the problem is there is no consensus in the lab community or between industry and the FDA as to what are sufficient controls even in terms of the analytical validity of a test.” Many issues, including how to clinically validate a test, are simply left up to the lab director to work out.
Mitigation measures will be important when patients are taking tests at home between doctor appointments, says Spring, responding to a question from the floor. Patients’ education level, the risk of a wrong result and what gets done with the results are all considerations. Possible measures might be counseling services or instructions-for-use that include more visual content.
The risk of an incorrect result can vary widely, notes Konnick. Being a “little off” with glucose meters is less concerning than a test that misses metastatic cancer.
“It’s not always easy to regulate,” says Gutierrez. When the first direct-to-consumer (DTC) tests for pregnancy and HIV were making their way to market, reviewers were concerned about what people would do with the results, leading to charges that the agency was being “paternalistic.” The FDA’s approach moving forward will be informed by the impact of special controls in place for the first DTC test for BRCA breast cancer genes.
Medicare isn’t necessarily going to pay for a diagnostic test, says Gerrard, noting that physicians are often as unfamiliar with the performance characteristics of tests as their patients. “They’re looking for someone else to certify tests for them, so they don’t have to.”
One ongoing problem is the disincentive for manufacturers to update their tests, says Konnick. “We had an HIV test that wasn’t updated for 10 years and missed certain strains of HIV.”
The 23andMe BRCA test, which misses over 90% of BRCA mutations, was market-approved after Gutierrez’s tenure at the agency. But he agrees there are problems, noting the FDA’s position that labeling patients as the intended users of the BRCA test was “good enough.”
Gerrard says he wonders if physicians ought to be paid more via evaluation and management coding if they stay current with research about diagnostic tests. He also worries where that might lead, such as lots of incremental changes to tests and exorbitant price hikes rather than innovations that are worthy of higher reimbursement.
The diagnostics community has an appetite for patient metrics but probably doesn’t realize false-positive rates can vary from 9% to 90% among tests, says Gutierrez.
To avert a huge false-positive problem, adds Gerrard, manufacturers need to specify the target population in which their device has been tested and proven.
The University of Washington provides education every year to its new crop of residents, using the lab as a resource, says Konnick. Test use is monitored, and interventions happen if they’re not used as intended.
Top 10 List
What labs and software developers most need to know about the FDA’s role with LDTs was summed up with a “top 10” list, courtesy of Sheila Walcoff, JD, founder and CEO of Goldbug Strategies:
#10 – The FDA defines LDTs as a type of IVD medical device. To qualify, it must be designed and developed in single clinical lab.
#9 – Enforcement discretion means the scope of policy is discretionary. It does not mean exemption.
#8 – The FDA grants authority for LDT/IVD manufacturers to market testing systems either as “for professional/prescription use only” or “over-the-counter (DTC) access.”
#7 – The FDA and CLIA regulations still co-exist. Under CLIA, all testing performed in a CLIA-certified lab is “clinical” grade and all tests in those labs must be FDA-cleared.
#6 – The agency does not regulate the practice of medicine, but things used in the practice of medicine. Providers are free to speak about off-label use of an LDT; it is forbidden only in commercial speech about the device.
#5 – The practice of lab medicine is not the same as clinical consulting about patient-specific test results.
#4 – The FDA can limit misbranding and off-label statements by a lab.
#3 – The FDA regulates software products if they’re used in a medical device context, including SaMD.
#2 – The agency has defined an over-the-counter, patient-initiated genetic test system in its sub-regulatory policies, and there’s a pathway for labs that want to do these kinds of tests. For example, the label on a DTC pharmacogenetic assessment system would read, “This device is for in vitro diagnostic use only and is intended for access testing by lay user.”
#1 – Enforcements against LDTs are most often done by non-public communications with clinical labs and software developers. “Since 2008, the FDA has increasingly narrowed the scope of its enforcement discretion policies through incremental policymaking and regulations applicable to both LDTs and software IVD products and services,” says Walcoff. The objective is to bring labs into FDA regulatory compliance.