Performance Assessment of First-Generation Anti-SARS-CoV-2 Serological Assays

Introduction

The global pandemic of COVID-19 caused by Severe Acute Respiratory Syndrome Corona virus 2 (SARS-CoV-2) affected Pakistan in mid-March¹ and till mid-July 2020, a total of 254,000 cases were diagnosed and 5,500 (2.1%) deaths reported². According to the Center of Disease Control (CDC), the recommended method for the diagnosis of COVID-19 is to detect the RNA of SARS-CoV-2 by real time Reverse Transcriptase Polymerase Chain Reaction (RT-PCR) with samples collected from nasopharynx and / or oropharynx^{3, 4}.Commercial manufacturers developed serological testing kits using different methodologies⁵.FDA gave approval for the use of these kits as Emergency Use Authorization (EUA)⁶. Verification and validation of these assays is required to achieve the clinical accuracy of test results^{7, 8}.

There are several advantages of serological testing. Pre analytical and analytical parts of serological assays have comparatively easy sample collection, simple to analyzeand requires less technical expertise as compared to nucleic acid detection. Serological testing can be performed in basic and small clinical laboratory settings^8,9. Post analytical important uses of antibody testing include; to determine the exposure of this disease among tested population, to assess the immune response against this virus and to identify the actual number of people who had acquired this virus but were asymptomatic therefore they didn’t get tested on RT-PCR. The results of this testing will be significant for epidemiological studies, disease surveillance and later in monitoring of response to the vaccine^{9, 10}. It plays vital role for identification of convalescent plasma donors for therapeutic use in treating moderate to severely ill COVID-19 patients¹⁰. There are several methodologies available for the antibody testing which include immune chromatographytechnique(ICT),electrochemiluminescenceimmunoassay(ECLIA),chemiluminee (CLIA) and enzyme linked immunosorbant assay (ELISA)^8,11 etc. To-date very little data available about clinical and analytical performance of SARS-CoV-2 antibody assays; we planned to analyze the clinical and analytical performance differentCommercial Serological assays of SARS-CoV-2, compared three commercial anti-SARS-CoV-2-IgG ELISA assays and to see the Seroconversion of anti-SARS-CoV-2 in four different population groups.

MATERIAL &METHOD

RESULTS

Total 404 subjects were recruited in the study. Out of these, 342 (84.6%) were males.Mean age of the subjects was 36.79±11.95 years. There were 239 CPDs; of them, 202 (84.5%) had a positive anti-SARS-CoV-2 by ECLIA.The qualitative ELISA performed in 174/239 CP donors, found to be positive in 142/174(82%) with a mean serum dilution of >1:160 (Range >1:80 to >1:320 serum dilution).Quantitative IgG by ELISA performed on total 252 subjects, of these 239(94.8%) were CPDs; 180 of these 239 (75.3%) were positive for IgG and their mean anti-SAR-CoV-2IgG level was 56.7±39.7U/ml. Seroprevalence was found to be 84.5% in CPDs at 21 days post onset of symptoms.

A total of 44 HCPs were recruited, their RT-PCR performed 3 to 5 days after the onset of symptoms. Their anti-SARS-CoV-2 antibody levels were tested by ECLIAon 7-8days and 12-14 days;seven of 13(53.8%) developed anti-SAR-CoV-2 within 10 days while 14 of the 18 (77.77%) developed antibodies by the end of second week of symptoms. Seroprevalence was found to be 40.9% in HCPs at 21 days post onset of symptoms.

Of the 70 HBDs screened, 15(21.4%) developed this antibody. These 15were also analyzed by quantitative IgGELISA and 14(93.3%) of them showed an antibody level of 27.2 ±19.95 U/ml. Seroprevalence was found to be 21.4% in asymptomatic HBDs.

In community group 12 patients were PCR positive and out of these 12, three patients had antibodies positive on ECLIA. Seroprevalence could not be calculated in this group since these patients were lost to follow up.

Analytical Performance of Serological Assaysshown in table no. 3.

The diagnostic performance(sensitivity, specificity, PPV & NPV) of all the serological assays shown in table no. 4.

DISCUSSON

We evaluated the performance of three different serological assays infourdifferentgroups. This study included convalescent plasma donors’whorecovered from COVID-19 infection.The performance characteristics of different kits e.g., sensitivity claimed by their manufacturers, fell short since our calculated sensitivities were below thatofthe manufacturers’ claim. This finding is in accordance with an Australian report published on 29^th April 2020 that assessed post-market validation of three serological assays for COVID-19 where they tested the serological assays on CPDs and found out their own diagnostic performance for each kit¹⁴.

Out of the three methodologies, ECLIA assay had sensitivity 97.44%, qualitative ELISA by Omega diagnostics had 96.4% while other two qualitative ELISA had lower sensitivity. This may be due to the fact that electrochemiluminescence assay detects total antibodies (including IgG) against SARS-CoV-2 while ELISA is only IgG specific.

Mei San Tang et al compared Abbott Chemiluminescenceassay and ELISA within 5 days of onset of symptoms, none of the immunoassays was able to detect the antibodies⁹. This leads to an important observation that when to test for the presence of antibodies against SARS-CoV-2 antibodies in our population.The samples which were taken from HCPs and community within 3 to 5 days post the onset of symptoms did not show antibody positivity on either assay. This finding is in contrast with the manufacturer’s claim that for ECLIA by Roche which showed 65.5% (CI 56%-74%) sensitivity when tested within 6 days PCR confirmation. None the less, gold standard for the diagnosis of acute SARS-CoV-2 infection is RT-PCR and we will need larger cohorts of patients to identify correct time for using serological assays in our population.

We found important and interesting finding in our study, out of 239 CPDs the 37 (15.5%) did not develop antibodies against SARS-CoV-2 virusbyany of the testing method i.e. ECLIA, quantitative and qualitative ELISA.Cellular immunity might have been developed in these seronegative CP donors. This remains to be tested in our cohort. Shane et al in June2020 detected CD4+ and CD8+ T cells in 100% and 70% of convalescent COVID patients; they further discovered that T cell responses are not only on spike (S) but also on M, N, and other ORFs of SARS-CoV-2¹⁵.

The two rapid testing devices using ICT methodology showed the sensitivity for IgM/IgG tests ranging from 23.53% to 90.4%. The specificity of both the devices on IgM and IgG tests was found to be 95% to 100%. The results of sensitivity are, however in contrast with the findings of Z.Zainol Rashid et al. They used nine rapid devices based on the principle of colloidal gold-labeled immunochromatography. Their sensitivity for both IgM and IgG tests ranges between 72.7% and 100%, while specificity ranges between 98.7% to 100%¹⁶.

The heathyblood donors had no symptoms of COVID-19 infection, they were healthy and active;a quarter of them seroconverted against COVID-19. This is a significant finding as it highlights the prevalence of this disease in general population. This observation has not been reported until now. We have also correlated the results of ECLIA and ELISA in our groups of CP donors and healthy blood donors which is showing positive correlation between these two assays.

Limitation of study was that we could not perform RT-PCR of healthy blood donors and ELISA of health care professionals and people from community due to cost limitation.

The strength of the study was that weincluded 239 convalescent plasma donors who recovered from COVID-19 infection that is by far the largest number of convalescent plasma donors recruited in any study.Wedid quantitative ELISA in the CPDs’ sample; this is the largest number of CPDs tested on quantitative ELISA till today in Pakistan.

CONCLUSION

In conclusion, the Roche ECLIA assay has higher sensitivity then ELISA and ICT methods in our study. ECLIA and ELISA are not to be used for the acute diagnosis of SARS-CoV-2 infection since these assays have poor sensitivity when tested within 7 days post onset of symptoms. Serological assays are important in determining the prevalence of this disease. In our study, all the commercially available serological assays are detecting antibodies against nucleocapsid protein of the SARS-CoV-2 virus. New serological assays should also be evaluated for the detection of antibodies against receptor binding domain of spike protein that is said to be neutralizing antibody¹⁷. Serial monitoring of IgG titers among patients for the collection of CP as well as pooled human IVIG is also an important query for the future of treatment of COVID-19. Measuring immunity to SARS-CoV-2 is essential for vaccinedevelopment, T cell responses are not only on spike but also on M, N, and other ORFs. These newer methods are being available for measure the T-Cell response, need of validate these assays is highly recommended.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE

This study was starts after the approval of Institutional Review Board (IRB)/Ethical Committee of Hospital.

AVAILABILITY OF DATA AND MATERIALS

Data is confidential.

COMPETING INTEREST

Authors do not have any conflicts of interest to declare.

FUNDING OR FELLOWSHIP GRANT

None