FIND evaluations of SARS-CoV-2 molecular tests

In February 2020, FIND launched an expression of interest (EOI) for test developers of in vitro diagnostics (IVDs) that detect SARS-CoV-2 nucleic acid (molecular tests) to participate in independent evaluation studies. Over 200 submissions were received and applications were reviewed according to the following scoring criteria (Table 1).

Table 1

[ADD TABLE]

As of August 2020, FIND is no longer accepting applications to evaluate molecular tests.

FIND conducted independent evaluations at the University Hospitals of Geneva (HUG) to verify the limit of detection (LOD) – as reported by the manufacturers – and the clinical performance of 22 manual molecular test kits in comparison to an in-house PCR protocol that was optimized based on the Tib Molbiol assay. The LOD analysis was performed using cultured viral stocks from a clinical isolate from Switzerland that was quantified using an E gene standard. The clinical performance analysis was conducted on extracted samples from individuals suspected to have COVID-19, 50 of which were reference PCR positive and 100 of which were reference PCR negative. Data are summarized in Table 2.

Additionally, a limited clinical performance evaluation of the Cepheid Xpert Xpress SARS-CoV-2 assay was also performed at HUG. A second collaborating site, the Translational Health Science and Technology Institute (THSTI) conducted a similar limited clinical performance evaluation of the Molbio TrueNat SARS-CoV-2 assay. Both studies were performed using frozen, stored respiratory samples from COVID-19 suspects. Results on the performance of these automated near-POC assays are shown in Table 3.

COVID-19 molecular assay evaluation protocol

Table 2

[ADD TABLE]

Table 2 footnotes

*  Clinical specificity: Further investigation is needed to determine if apparent false positives are truly false positives or whether they are due to a false negative reference standard result
** PCR platform: All products were evaluated on a PCR platform recommended by the supplier, listed in this table. Each test can be performed on other PCR systems, detailed in the product’s instructions for use.
[a] The two false negative samples tested positive with the second PCR (PCR 2) that targets E gene of SARS, SARS-COV-2 and/or SARS-like coronaviruses.
[b] Samples for both analytical and clinical analyses were from already-extracted specimen, therefore the methods varied from those recommended by the supplier as the internal control was not included.
[c] Samples for both analytical and clinical analyses were from already-extracted specimen, therefore the methods varied from those recommended by the supplier as the internal control was added to the master mix.
[d] Evaluation procedure varied from recommended protocol. In order to achieve the recommended sample input volume, a 2.5 fold dilution of the samples was used.
[e] Sansure claims a lower LOD of 6.4 cp/rxn, which has been independently verified.
[f] Evaluation procedure varied from recommended protocol as source material was already-extracted RNA; extracted MS2 control was added directly to the master mix.

Table 3

[ADD TABLE]

Table 3 footnotes

* Clinical specificity: Further investigation is needed to determine if apparent false positives are truly false positives or whether they are due to a false negative reference standard result
[1] Note: evaluation performed at THSTI
[2] RdRP is only used as a reflex test; the results are for combined E+RdRP positives

More information

For questions relating to the evaluation of molecular tests, please contact our Emerging Threats team.