The U.S. FDA’s recent decision to pass on any emergency use authorization (EUA) filings for tests for the COVID-19 pandemic is well known, but the agency had a chance to lend some additional information on that question in the Oct. 21 town hall. Despite the opportunity to clarify some of the underlying questions, Tim Stenzel, director of the Office of In Vitro Diagnostics and Radiological Health (OIR), said little more than that the change was made “largely because the FDA cannot require an EUA, according to the HHS statement.”
Stenzel seemed to suggest during the Oct. 14 town hall that an FDA review of an EUA for a lab-developed test (LDT) was not entirely foreclosed, stating that the agency is working on some clarifications for the FAQ. Toby Lowe, associate director of OIR, said during the Oct. 14 event that the agency is “not generally prioritizing review” of LDTs, although an FDA spokesperson had advised BioWorld that additional clarification may be in the offing.
During the Oct. 21 town hall, however, Stenzel said the agency is currently declining to review LDTs for the reasons stated on the FAQ page. He was answering a query as to whether the agency would be willing to review an EUA filing by a lab that could offer high-throughput testing, which the agency has previously said it would prioritize.
After making reference to the HHS rescission order – which indicated only that the FDA lacked the statutory authority to mandate a review of LDT, and did not stipulate that the agency could not conduct such reviews – Stenzel reiterated that the agency wants to focus its resources on tests that require an EUA filing. He said many high-throughput tests available for use in labs rely on kits distributed by test manufacturers, which are still subject to the EUA process.
Interpretation of false positives still an issue
Stenzel said on the Oct. 21 session that the agency has hired additional staff and has shifted some resources to handle the volume of EUA filings, and he urged test developers to assist the FDA in its efforts to educate customers on the meaning of a false positive test.
“We continue to hear about challenges relating to getting false positive results, even in situations where it is totally expected,” Stenzel said, adding that if one assumes that a molecular or antigen test’s specificity is 99%, only one time out of 100 will someone who is negative test positive.
Stenzel characterized that as a solid performance metric, but noted that when used for large-scale screening for populations with a low positivity rate, a perfectly functioning test that was appropriately administered will still fairly routinely yield a 1% rate of false positives. This would hold true regardless of whether the patient is symptomatic, and regardless of whether the test was labeled for use as a screening mechanism.
The scenario in which population screening generates more false positives than true positives “is a totally expected situation, and we planned for it,” Stenzel said, adding, “I would recommend you get some sort of orthogonal confirmatory test done.” The agency’s preference would be that the secondary test be a molecular test of high specificity and sensitivity, although the physician could prescribe the use of another direct antigen test, assuming the index test was also a direct antigen test.
However, there is no guarantee that the second such test could be relied upon to yield a more accurate result than the first. The agency’s concern in this context is that a false positive might place a patient in a high-risk situation via admission to a hospital COVID-19 ward, or force the patient to isolate, possibly cutting off care for other conditions, although the loss of livelihood is another consideration to which the FDA is attuned, Stenzel said.
The FDA is interested in any data industry might have on the repeat test scenario described by Stenzel, but he noted that he does not have a specific recommendation for the choice of follow-up test.
The FDA has processes to ensure that the National Cancer Institute testing validation program follows the directions in package inserts, and that the correct sample types are used in the NCI validations. However, Stenzel said, “we don’t have the luxury of getting a wide variety of panels for each kind of sample type,” and therefore the FDA will take a closer look at those situations where a less-than-ideal sample type is used to handle the validation effort.
One approach is to rely on the subset of NCI results that correlate with the instructions for use with the test, but Stenzel said the FDA will attempt to resolve any negative findings with the test sponsor. The NCI has access to both serum and plasma to validate a test, and uses widely available anticoagulants as well in an effort to avoid any aberrant outcomes.
Certificate of deactivation may not be available
Sponsors who wish to ship the FDA’s SARS-CoV-2 panel overseas for test development may run into a few hurdles, such as a request from the government of the destination country for a certificate of viral deactivation. Stenzel said the agency receives live virus at FDA campus, but said he does not know whether the FDA can provide a certificate of safety for the foreign government despite the agency’s confidence that the end product is safe for distribution as a biosafety level (BSL) 2 panel.
One option might be for the sponsor to request that the FDA draft a letter generated to the effect that the product meets BSL 2 requirements, but Stenzel noted that the limit of 10 kilograms of dry ice for the shipment might not suffice to avoid spoilage. One of the callers to the Oct. 14 town hall noted that this was a limit imposed by another national government, and Stenzel said a work-around for these dilemmas may be to contact a domestic lab that can provide the test validation services.