The FDA’s weekly town hall on testing for the COVID-19 pandemic included the usual range of concerns about test performance, but concerns regarding swabs and sample sites continue to mount. The predicament has led to the announcement that the FDA along with the National Institutes of Health (NIH) will hold a May 15 town hall regarding swabs, with a particular interest in swabs produced via additive manufacturing.
The weekly testing town hall hosted by the FDA’s Center for Devices and Radiological Health has played a key role in keeping test developers up to date on the agency’s expectations, including those associated with serological tests. Tim Stenzel, director of the FDA’s Office of In Vitro Diagnostics and Radiological Health, said the FDA added a template for antigen testing and updated the template for molecular testing for viral RNA following the May 4 guidance update. Stenzel said the agency is still working on a home specimen collection template as well, adding, “we see a lot of continued and growing interest in home collection.”
Sara Brenner, associate director for medical affairs at CDRH, said 3D-printed nasopharyngeal swabs “and swabs in general, have been a very hot topic” of late among stakeholders, including manufacturers. Brenner said the agency has been hinting at the possibility of a town hall dealing with swabs “for many weeks,” an event that will take place May 15 from 1 to 2 pm U.S. Eastern time. Representatives with the Veterans Healthcare Administration (VHA) will join the NIH and FDA staff on the call, Brenner said.
The agency’s policy for 3D printed swabs is encompassed by an overarching policy for additive manufacturing in connection with the pandemic. The FDA has formed a collaboration with the NIH and the VHA, which provides design templates for various items. Most of these fall under the category of personal protective equipment, however. The template for nasopharyngeal swabs comes with a disclaimer regarding safety risks, as does the design for a throat swab.
Stenzel said the agency sees the nasopharyngeal (NP) swab as superior to the nasal (NS) swab for most testing considerations, although the agency has authorized the use of swabs for other anatomical locations. As for the best site for samples for asymptomatic testing, he said, “the science hasn’t been established yet,” adding that the answer probably depends in part on the viral load that have been established in that asymptomatic patient. A number of developers are interested in this and other questions, and Stenzel said the FDA and these sponsors are working on an appropriate study design.
Patient samples in, contrived samples out for RNA testing
Stenzel said the most recent FDA update regarding molecular testing revolves largely around the question of positive patient samples. He said the agency intends to shift away from the use of contrived samples to actual patient samples for demonstrations of clinical performance and, corollary to that, will also seek a comparison to a previously authorized molecular test to assess the accuracy of the candidate test.
Stenzel reiterated the performance expectations regarding direct antigen tests, which includes a minimum sensitivity of 80% relative to a high-sensitivity molecular assay. The FDA also recommends that specificity “be very high” so that a positive result is reliable, he said. Lower test sensitivity raises concerns regarding false negative, and thus there is language in the authorization for the first antigen test that recommends that a negative result for a patient that is symptomatic be referred for a molecular test.
Stenzel said the NIH will publish the results of the National Cancer Institute’s (NCI) reviews of serological tests once the FDA has made a regulatory decision on those results. He said the NIH will post the results “essentially contemporaneously with those [regulatory] decisions.”
The FDA’s policy for serology tests for antibodies generally calls for a prospective evaluation of the test, and Stenzel said the agency prefers that developers compare the results for the candidate test to another sample (e.g., whole blood or serum) rather than to a previous molecular test result. A sponsor that would like to steer a different tack for validating a serology test should discuss this with the agency, and Stenzel said, “there is a little risk using a previous molecular result.”
This is driven largely by the prospect that test performance may not correlate strongly between the two test types. This hinges in part on the amount of time that has elapsed since that patient turned positive, as well as the elapsed time since the patient developed antibodies and/or developed symptoms. Stenzel noted that a sponsor will have to factor in the timing of the respective dates of emergence of the antibody isotypes as well.
Stenzel said the FDA is taking part in an interagency panel of samples for serology tests, which is spearheaded by the Biomedical Advanced Research and Development Authority (BARDA). A public announcement about such a panel will be made in the near term, and he noted that a growing number of entities outside of this federal agency consortium are also looking into development of such a panel. The unreliability of some commercially available samples is one of the principle drivers of the interagency effort.
There is a lot of interest among developers in the NCI serology test validation program, dubbed the umbrella EUA program, although Stenzel said a backlog may develop as that interest converts to actual applications. An expedited pathway to authorization is available for serological tests that do not differentiate between the antibody isotypes, and Stenzel explained that the review will check for consistency between sponsor data and data from the NCI test validation regime. If the test distinguishes between IgM and IgG, the NCI will ask for data supporting the accuracy of the distinction between the two. Stenzel added that 95% or greater isotype specificity will win the sponsor a waiver for cross-reactivity evaluation.