Cardiac troponin (cTn), an indicator of heart tissue damage, has become an essential part of diagnosing myocardial infarction (MI) since testing for it was introduced in the 1990s. New “high sensitivity” (hs-cTn) assays are common in Europe and other parts of the world, but not yet widely available in the U.S. In November, the Food and Drug Administration (FDA) convened a daylong meeting to discuss the next stage in evaluating and introducing hs-cTn assays for clinical use in the U.S.
FDA hosted the workshop to obtain feedback from and open lines of communication with stakeholders, including clinicians, clinical laboratory professionals, and in vitro diagnostic manufacturers, according to Paula Caposino, PhD, branch chief of cardio renal diagnostics in FDA’s Office of In Vitro Diagnostics and Radiological Health. The agency heard from invited panelists and the audience on five broad areas of concern involving its regulation of hs-cTn assays, including: cutoff determination/reference interval studies; clinical trial design; preanalytical and analytical considerations for clinical trials; clinical trials for point-of-care (POC) devices; and use of existing clinical data to support claims. The agency has faced challenges assessing cTn assays for the U.S. market in all these areas.
Defining the 99th Percentile
One issue involves how manufacturers have gone about establishing their assays’ recommended reference intervals. The Third Universal Definition of Myocardial Infarction considers a rise and fall of cTn a hallmark of MI, with at least one value being elevated. The crucial elevation threshold is the 99th percentile of all results in a healthy population. This presents a problem, however, as cTn tests are not standardized or harmonized, Caposino noted. “Test results are not interchangeable from one device to another. This means that each device has its own reference interval information.”
In January 2017, FDA cleared Roche Diagnostics’ Elecsys Troponin T (TnT) Gen 5 Stat, which has a fast turnaround time and higher analytic precision than other cTn tests available in the U.S. Roche has marketed this test as “high sensitivity.”
With more such assays likely to come on the market in the U.S., Fred Apple, PhD, DABCC, director of clinical laboratories at Hennepin County Medical Center in Minneapolis, has argued that all should meet the International Federation of Clinical Chemistry and Laboratory Medicine’s (IFCC) standard for hs-cTn assays. In the case of Roche’s T Gen 5 Stat, FDA appears to have limited the company to reporting concentrations to the assay’s limit of quantitation at 6 ng/L. Yet, the test’s limit of detection is 3 ng/L. “When sponsors choose a different cutoff other than one based on the 99th percentile and their device can measure the 99th percentile,” said Caposino, “we request that the sponsor include clinical performance information using the 99th percentile in the labeling in accordance with the clinical guidelines since it is our understanding that this information is useful for clinical labs and clinicians.”
What’s the Best Population?
Another aspect of establishing a test’s reference range has been having an accurate sample of the patients the assay will be used for. This can be difficult to pin down because a diverse population is needed for the clinical trial to be representative. IFCC lists minimal requirements for clinical studies (300 men and 300 women), but many manufacturers enroll hundreds more in an effort to include enough participants from important subgroups. Further complicating this hurdle is that even if trial subjects declare themselves healthy, subclinical causes for troponin elevation might go undetected and potentially bias the results, noted Jeff Bishop, PhD, senior vice president of research and development at Singulex, which in November filed a 510(k) pre-market notification submission with FDA for its Singulex Clarity cTnI assay.
The panel concluded that until more detailed guidelines are set up, manufacturers should be more transparent in their assays’ labeling, giving more information about the parameters of clinical trials supporting a test’s development. This will help clinicians evaluate how to use the assay. Examples of such data include details about the enrolled population, the subjects’ ages, the distribution of values and whether they were skewed in any way, and how outliers were defined. “Transparency is the name of the game,” emphasized Robert Christenson, PhD, FAACC, professor of pathology, medical and research technology at the University of Maryland School of Medicine in Baltimore.
One potential avenue for clinical trials is to use existing samples or data to evaluate new assays for licensure. But this approach has its own problems: Analytes in frozen samples may not be stable, and data pulled from hospital records is often messy and incomplete. Still, data from existing sources may be useful, but it needs to be carefully vetted if used in applications for new devices or assays. “Using data that exists is great, as it saves some effort, but the devil is in the details,” said Alberto Gutierrez, PhD, the recently retired director of FDA’s Office of In Vitro Diagnostics and Radiological Health.
Some attendees expressed frustration that hs-cTn assays have been slow to be adopted in the U.S. “The reality right now is the U.S. is 10 years behind the rest of the globe on our troponin assays,” said Frank Peacock, MD, associate chief of emergency medicine research at Baylor College of Medicine in Houston. “We are the third world of troponin.” Peacock urged that in designing clinical trials, “the perfect are the enemy of the good” and that good trials can be designed that may not be ideal, but allow clinicians from the U.S. faster access to useful troponin assays.
Same Test, Different Uses
Clinical trials of point-of-care (POC) hs-cTn assays pose particular problems. Trade-offs often have to be made between a test’s accuracy and differing needs of clinicians depending on their practice sites. “Does everybody have to have the same car?” said Christenson. “The answer is no. I’ve come to the realization that we don’t all have to have the same assay.”
Some clinicians rely on cTn results to rule-in MI patients, but others might use them more to rule out MI in order to send patients home safely without further workup, he argued. Conversely, Alan Jaffe, MD, a consultant in cardiovascular medicine and chair of the division of clinical core laboratory services at the Mayo Clinic in Rochester, Minnesota, noted that many institutions, especially those in rural areas, need to use the same assay for both tasks. “It often will be used as a solitary device,” he said, so devices need to be able to handle both kinds of demands.
Some POC cTn devices tested in emergency department settings have limited temperature and humidity ranges. Because nurses in those settings are not likely to have time to read and understand those limitations, it’s important to note when test conditions are not within range. Panelists leaned toward not reporting values in those circumstances out of concern that even if warnings were included with results, they might be ignored. Devices should simply report that test conditions are out of range.
Under the rubric of preanalytical and analytical considerations for clinical trials, panelists also discussed biotin interference in cTn assays. Biotin has become popular in recent years as a beauty supplement and when taken at levels well above the daily recommended intake may cause interference in immunoassays, including for cTn. FDA released a safety communication in November urging more communication among laboratorians, clinicians, and patients about the importance of reporting biotin use to mitigate the risk of clinically significant incorrect lab test results. The agency said it received an increase in the number of reported adverse events related to biotin interference, including one death. This patient had been taking a high dose of biotin, which led to a false negative cTn result. “The obligation is on manufacturers to identify whether there is biotin interference, especially at the 99th percentile because that is the microscope being looked at,” said Apple.
FDA staff members at the meeting didn’t make any declarations about how the agency will proceed on these issues. Instead, they noted that the agency plans to take comments made by panelists and audience members as a starting point for guidelines for troponin assay manufacturers seeking clearance. Caposino emphasized, “We want to work together with all stakeholders so we can increase availability of troponin devices that work well and are innovative.”
Rina Shaikh-Lesko is a freelance science writer in San Jose, California. +Email: firstname.lastname@example.org