Limits and Opportunities of SARS-CoV-2 Antigen Rapid Tests: An Experienced-Based Perspective

Abstract

Background: Due to the steadily rising case numbers of SARS-CoV-2 infections worldwide, there is an increasing need for reliable rapid diagnostic devices in addition to existing gold standard PCR methods. Actually, public attention is focused on antigen assays including lateral flow tests (LFTs) as a diagnostic alternative. Therefore, different LFTs were analyzed regarding their performance in a clinical setting.

Material and methods: A pilot sample panel of 13 bronchoalveolar fluids (BALFs) and 60 throat washing (TW) samples with confirmed PCR results, as well as eight throat washes invalid by PCR, were tested with the BIOCREDIT test (RapiGEN), the Panbio^TM assay (Abbott), and the SARS-CoV-2 rapid antigen test (Roche).

Conclusion: The analyzed antigen test showed an interassay correlation of 27.4%, with overall specificities ranging from 19.4% to 87.1%, while sensitivities of the respective tests ranged between 33.3% and 88.1%. Because these assays did not entirely meet all high expectations, their benefit has to be carefully evaluated for the respective test strategy and setting.

Keywords: COVID-19; SARS-CoV-2; rapid antigen tests.

Figure 1

Controls of the Panbio^TM COVID-19 Ag test (Abott). (A) Supplied positive controls (PCs) (2, 4, 6, 8) and negative controls (NCs) (1, 3, 5, 7) of four kits were performed, according to the manufacturer’s protocol; (B) Controls of approach (A) were pooled and applied to the antigen tests by Roche (9, 10) and RapiGEN (11, 12). While the PC by Abbott also became positive with Roche (10), the BIOCREDIT test did not recognize the positive control (11). PC, positive control; NC, negative control.

Figure 2

SARS-CoV-2 antigen detection. To evaluate the suitability of specimen others than nasopharyngeal swabs, two PCR positive throat washes, i.e., (Sample I) Ct_E-Gen = 15.6 and Ct_S-Gen = 14.8 and (Sample II) CT_E-Gen = 14.7 and Ct_S-Gen = 14.9, and one BALF, i.e., (Sample III) Ct_E-gen = 13.1 and Ct_S-Gen = 12.6 were used for initial evaluation of lateral flow test performance. SARS-CoV-2 antigen was detected with the BIOCREDIT test in all samples, but dilution in assay buffer (d 1, 5, and 7) decreased sensitivity as compared with original fluid (o 2, 6, and 8). Undiluted samples I and III were also used for an initial analysis with the Panbio^TM COVID-19 Ag test (Abott) (3 and 9) and the SARS-CoV-2 rapid antigen test (Roche) (4 and 10) also delivering positive results in both tests. As shown by the control line, the specimens allowed a proper test performance and did not contain any inhibitory substances. d, 1:2 diluted with supplied buffer and o, original.

Figure 3

Correlation of SARS-CoV-2 antigen detection with RNA load. This figure shows longitudinal SARS-CoV-2 detection by PCR in two asymptomatic individuals (patient A, dark grey and patient B, light grey). In both cases, SARS-CoV-2 RNA could be detected for about five weeks before recovering in week six. The RapiGEN assay (red) detected SARS-CoV-2 antigen only for patient A in three samples with Ct > 30. Although the assays by Abbott (green) and Roche (yellow) identified 4 and 7 samples as positive, respectively, one sample remained antigen negative (Sample 1 of patient A). Except for the negative sample, there was no coincidence between the LFTs at all. * Only tested negative by RapiGEN, due to insufficient amount of specimen left for the assay.