The number of tests of various types for the SARS-CoV-2 virus are still growing, but the emergency use authorization (EUA) program is not the only option for developers. Timothy Stenzel, director of the Office of In Vitro Diagnostics and Radiology, said on the agency’s May 27 town hall that the agency would like to see test developers file for a 510(k) for their tests when the data are sufficient to support an application, adding that any such clearances will not affect the U.S. FDA’s interest in keeping plenty of similar tests on the market via the EUA mechanism.
The town hall was held a day after the CDC publicly acknowledged that in some circumstances, a serological test is of limited value in an area with a low-prevalence population. The CDC’s May 26 statement noted that a serological test offering 90% sensitivity and 95% specificity might offer a positive predictive value of only 49% if the prevalence is only 5%. That part of the CDC statement drove extensive media coverage of the agency’s update, but that metric was widely understood well before the CDC made that dynamic explicit.
Stenzel said the agency had received an inquiry into when the EAU program will end and when the associated tests will have to move into the 510(k) premarket pathway, although he said the state of emergency declaration is “unlikely to be terminated anytime soon.”
Nonetheless, Stenzel said, “we always do encourage developers to work toward a routine application,” which might entail a petition for de novo classification of the test. “We encourage these applications as soon as the developers are ready, and we’re willing to work with them right now” on converting EUA test dossiers into conventional premarket filings, he said. The FDA has some discretion to keep EUA tests on the market despite the filing of a 510(k) for one test of a particular type, and thus the agency does not anticipate that the need for EUA tests will disappear anytime soon.
Stenzel said the FDA has added a new extraction option under the EUA for molecular testing, an extraction reagent by Beckman Coulter Life Sciences of Brea, Calif. This may be a reference to a May 18 announcement by the company, which states that the RNA Advance Viral XP reagent is available as high-throughput labs “worldwide ramp up RNA extraction capabilities.”
The Rapid Acceleration of Diagnostics (RADx) challenge has “significant funding” for developers of rapid antigen and molecular diagnostic tests, Stenzel observed, adding that the program is for established test developers as well as companies new to the market.
Rapid antigen assays must still be matched with molecular assay
Stenzel said the FDA is sticking with its policy that developers of rapid antigen tests for the SARS-CoV-2 virus should use a high-sensitivity molecular assay as a comparator rather than another rapid antigen test. As for the role of a point-of-care test for asymptomatic presurgical patients, he said “we are open to receiving data that would provide an indication and appropriateness” for such a test, adding that labs are allowed to perform such tests under the existing EUA. Stenzel reminded those in attendance that at-home collection kits must already appear on the EUA list prior to distribution, a different approach than is used for many tests for COVID-19.
There is substantial interest among developers of molecular testing in the pooling of samples, and Stenzel said the FDA recognizes that the limited supply of reagents makes sample pooling an attractive approach to back-to-work and back-to-school testing. To date, no EUA authorization has been granted for a molecular test that was validated in this manner, and Stenzel said, “we are very open to pooling schemes,” but would like to discuss the required validation with the sponsor. One of the considerations is the effect of pooling on low-positive samples, while the impact of the limits of detection of a test is another. He said the FDA will update its website with some information regarding pooling “as soon as possible.”
The FDA is not averse to the use of banked blood samples for use as negative controls and/or to evaluate the cross-reactivity of a test, Stenzel said, but he noted that such a proposal would have to be accompanied by a number of details, such as whether the test would be a qualitative versus a quantitative test. Information on the stability of the banked sample would also be crucial, and Stenzel said he would be willing to be involved directly in questions of this nature, thus highlighting the importance of banked samples for test validation.
Cycle threshold values emerge as an issue
Some clinical lab directors are receiving requests from the physician for the cycle threshold (Ct) value for a test, presumably because such information can be useful in clinical decision-making. Stenzel said patients may remain molecularly positive even after symptoms dissipate, and physicians may be of the view that a high Ct count can lend some clarity as to what course of action to recommend to the mildly symptomatic or asymptomatic patient.
Still, Stenzel said, “we still don’t know very much about that,” given that there is considerable uncertainty regarding whether respiratory samples yield consistent viral load results for a given patient, even when sampled on consecutive days. Tests for other infectious diseases may rely on blood or blood products, and so Ct thresholds are more reliable in those scenarios, but Stenzel observed, “this is probably a conversation that will not go away because of the need to figure out what to do with some patients.” He noted that this is another matter in which he intends to take a personal interest.