Comparison of one single-antigen assay and three multi-antigen SARS-CoV-2 IgG assays in Nigeria

Abstract

Objectives

Determining an accurate estimate of SARS-CoV-2 seroprevalence has been challenging in African countries where malaria and other pathogens are endemic. We compared the performance of one single-antigen assay and three multi-antigen SARS-CoV-2 IgG assays in a Nigerian population endemic for malaria.

Methods

De-identified plasma specimens from SARS-CoV-2 RT-PCR positive, dried blood spot (DBS) SARS-CoV-2 RT-PCR positive, and pre-pandemic negatives were used to evaluate the performance of the four SARS-CoV-2 assays (Tetracore, SARS2MBA, RightSign, xMAP).

Results

Results showed higher sensitivity with the multi-antigen (81% (Tetracore), 96% (SARS2MBA), 85% (xMAP)) versus the single-antigen (RightSign (64%)) SARS-CoV-2 assay. The overall specificities were 98% (Tetracore), 100% (SARS2MBA and RightSign), and 99% (xMAP). When stratified based on <15 days to ≥15 days post-RT-PCR confirmation, the sensitivities increased from 75% to 88.2% for Tetracore; from 93% to 100% for the SARS2MBA; from 58% to 73% for RightSign; and from 83% to 88% for xMAP. With DBS, there was no positive increase after 15-28 days for the three assays (Tetracore, SARS2MBA, and xMAP).

Conclusion

Multi-antigen assays performed well in Nigeria, even with samples with known malaria reactivity, and might provide more accurate measures of COVID-19 seroprevalence and vaccine efficacy.

1. Introduction

Good performance of the SARS-CoV-2 assay is paramount to generating accurate data on COVID-19 seroprevalence. Most validations are performed in Europe and North America, where the serological panel characterization is uniquely different from Africa. These populations may not represent the other biological variables seen in Africa or across the world [1]. Therefore, validating SARS-CoV-2 assays using specimens from populations of interest is crucial to avoid biased COVID-19 seroprevalence estimates [2,3] and ensure that the tests meet a minimum requirement for diagnostic performance in the population. For example, studies from African countries demonstrated limited specificity with SARS-CoV-2 assays [3,4], and the differences in specificities may impact their positive predictive values [5]. Additionally, several studies have shown cross-reactivity of SARS-CoV-2 serological assays with pre-pandemic African [3,4,6,7] and Cambodia specimens [2], which could be a result of previous exposure to malaria [[2], [3], [4],7], other human coronaviruses [6,8], dengue exposure [9,10], HIV [11,12], and Zika virus infection [13].

Evaluation of commercially available SARS-CoV-2 serologic tests is imperative for adequately detecting antibody responses to SARS-CoV-2 infections in individuals [14]. Previous studies suggested that combining two or more antigens in a serological assay is better than a single antigen target for COVID-19 seroprevalence studies [15], [16], [17]. Additionally, multi-antigen SARS-CoV-2 IgG assays provide a reliable serological classification of individuals that had been infected with SARS-CoV-2 [18]. The use of dried blood spots (DBS) for the detection of antibodies against SARS-CoV-2 has been documented [19] and modified for Luminex xMAP assay [20]. DBS is easier to collect in the large-scale survey, and it provides an alternative to plasma and serum for seroprevalence of large populations, but the use has not been assessed for SARS-CoV-2 seroprevalence in Nigeria.

Nigeria is a malaria-endemic country, with a malaria prevalence of 23% (by microscopy) in children under five years based on the 2018 Demographic and Health Survey [21]. Such high disease prevalence may affect the performance of serological SARS-CoV-2 assays in Nigeria. Previous validation of commercial SARS-CoV-2 single antigen immunoassays in Nigeria indicated lower sensitivity than the manufacturers’ results [22]. In addition, another study showed substantial cross-reactive with two single-antigen targets in Nigeria [7]. Therefore, it is essential to validate commercially available SARS-CoV-2 IgG assays in this setting for serosurveillance purposes among the population [23]. Low sensitivity with SARS-CoV-2 rapid immunoassays is a concern, and rigorous assay validation using standardized sample sets is imperative [24]. Specificity is an essential characteristic of SARS-CoV-2 serological assay, and low incidences of SARS-CoV-2 infection in sub-Saharan Africa may be due to cross-recognition of other human coronaviruses [6]. Therefore, an accurate estimate of SARS-CoV-2 seroprevalence in Nigeria will help inform decisions about which tests to use in Nigeria for SARS-CoV-2 serosurveillance. The objective of this study was to evaluate the performance of four SARS-CoV-2 assays using SARS-CoV-2 reverse transcriptase-polymerase chain reaction (RT-PCR) positive and pre-COVID-19 pandemic plasma and DBS samples from Nigeria, where malaria is endemic.

2. Materials and methods

2.1. Study design and sample collection

The laboratory analyses were performed at the Center for Human Virology and Genomics, Nigeria Institute of Medical Research (NIMR), Yaba, Lagos, and the Multiplex Bead Assay (MBA) Laboratory, National Reference Laboratory (NRL), Nigeria Center for Disease Control (NCDC), Gaduwa, FCT, Nigeria. The study design and descriptions have been published elsewhere [7,22]. NIMR prepared well-characterized plasma panels collected from SARS-CoV-2 positive (n = 100) patients taken at various time points post-RT-PCR confirmation and a negative panel comprising of pre-pandemic archived frozen plasma samples of Hepatitis B positive (n = 50) and HIV positive patients (n = 50), which were used for the previous validation. The diagnosis was done using the Cobas SARS-CoV-2 test on the Cobas 6800 system (Roche Diagnostics, Basel, Switzerland) RT-PCR methodology on the swab specimens collected on day 0. Subsequently, the SARS-CoV-2 RT-PCR positive patients that consented were invited back on various time points on days; 0-3, 4-7, 8-14, 15-28, and ≥29 for whole blood collection. A sub-set of the RT-PCR samples (n=55) was spotted on DBS for comparison with plasma. The RT-PCR procedure was performed according to the manufacturer's instructions. The pre-pandemic samples were archived plasma specimens collected before October 2019 with consent from the participants for future testing.

2.2. Further assessment of cross-reactivity

In 2018, Nigeria conducted a population-based HIV survey called Nigeria HIV/AIDS Indicator and Impact Assessment (NAIIS). With written consent from the participants for future studies, these samples were stored at -70^°C at the Biorepository of the NRL, NCDC, Gaduwa. The NAIIS samples, under a different protocol, “Nigeria Multi-disease Serologic Surveillance using Stored Specimens (NMS4)”, were characterized for other pathogens using the MBA. The selected NAIIS samples had both plasmas (200) and matched DBS (100) specimens and were utilized for the validation. In addition, the NAIIS samples with known malaria reactivity characterized using the MBA were included for cross-reactivity evaluation. All specimens were tested using the four SARS-CoV-2 IgG assays: Tetracore® FlexImmArray™ 7-Plex SARS-CoV-2 Human IgG Antibody Test (RUO), a SARS-CoV-2 multi-antigen IgG Bead Assay developed by CDC (SARS2MBA), RightSign^TM COVID-19 IgG/IgM Rapid Test Cassette, and Luminex xMAP® SARS-CoV-2 multi-antigen IgG Assay.

2.3. Assays

2.3.1. Tetracore® FlexImmArray™ 7-Plex SARS-CoV-2 Human IgG Antibody Test (RUO)

The Tetracore® FlexImmArray™ 7-Plex SARS-CoV-2 Human IgG Antibody detects Human IgG Antibodies to SARS-CoV-2 in serum or plasma samples. Three immobilized SARS-CoV-2 recombinant antigens; Receptor Binding Domain (RBD-CoV-2), Nucleocapsid Protein (NP-CoV-2), and RBD-NP Hybrid (RBD-NP), are included in the 7-plex microsphere cluster for three unique microsphere regions for detecting IgG antibodies to the virus [25].

2.4. SARS-CoV-2 multi-antigen IgG bead assay developed by CDC (SARS2MBA)

Samples were screened in single-well with microspheres coupled to spike (S) protein, receptor-binding domain (RBD)541, RBD591, and N (nucleocapsid) protein. Briefly, 50 µl of plasma and DBS pre-diluted to 1:400 in Buffer B (1X PBS, 0.5% casein, 0.5% polyvinyl alcohol [PVA], 0.8% polyvinylpyrrolidone [PVP], 0.3% Tween-20, 0.02% sodium azide and 3 µg/mL Escherichia coli extract) was incubated with beads (1250 µl/well/antigen) for 90 min in a 96-well assay flat bottom plate. The beads were then washed with 150 µl of phosphate-buffered saline (PBS) with Tween 20 (PBST) three times using a handheld manual magnetic plate washer to hold the beads during the washing process. Then after, 50 ng IgG and 20 ng IgG4 were added to the beads and incubated for 45 min to detect bound antibodies, after which they were washed three times with PBST. The beads were further incubated for 30 min with phycoerythrin-conjugated streptavidin (SA-PE) and washed three times with PBST. To remove any loosely bound antibodies, 50 µl assay buffer (1X PBS, 0.5% BSA, 0.05% Tween-20, and 0.02% NaN3) was added to the beads and incubated for 30 min after which they were washed once with PBST. The beads were then suspended in PBS and kept at 4^°C overnight. The MAGPIX® instrument was used to analyze the beads the next day, and the median fluorescence intensity (MFI) for each antigen was determined for each sample. A sample was considered positive for SARS-CoV-2 antibodies if tested positive for antibodies to both S protein and RBD591 [26].

2.5. RightSign^TM COVID-19 IgG/IgM rapid test cassette

The RightSign^TM COVID-19 IgG/IgM Rapid Test Cassette detects SARS-CoV-2 antibodies in venous blood, serum, and plasma. The RightSign test contains the recombinant SARS-CoV-2 antigen (spike RBD domain antigens of SARS-CoV-2) and was interpreted based on the manufacturer's instructions [27].

2.6. Luminex xMAP® SARS-CoV-2 multi-antigen IgG assay

The xMAP® SARS-CoV-2 multi-antigen IgG Assay is intended for the detection of IgG antibodies directed against the nucleocapsid protein (NCP), RBD of the spike protein, and the S1 subunit of the spike protein (S) of SARS-CoV-2 in human serum or plasma [28]. DBS (diluted 1:400) or plasma (dipotassium EDTA) samples were incubated with the multiplexed microspheres in which the SARS-CoV-2 specific IgG antibodies bind to the S1, RBD, and NCP antigen-specific microspheres. The xMAP Multi IgG CoV-2 multi-antigen IgG assay software was used to determine the presence or absence of SARS-CoV-2 specific IgG antibodies in each sample. The microsphere counts and MFI of the controls were checked against predefined threshold values for each tested specimen. If the MFI values of the NCP target antigen control were above the set threshold with at least one of the MFI values of other target antigen control S1 subunit or the RBD of the spike protein, the sample was considered SARS-CoV-2 IgG positive [28].

2.7. Statistical analysis

The sensitivities for all the four SARS-CoV-2 IgG assays were calculated, with the corresponding 95% confidence intervals (CI), using the proportion of positive SARS-CoV-2 RT-PCR test results over the total number of post-pandemic samples tested. Sensitivities were also estimated based on days of post-RT-PCR confirmation of SARS-CoV-2 infection (days: 0-3, 4-7, 8-14, 15-28, and ≥29). Specificities, with the corresponding 95% CI, for all the assays were calculated using the ratio of negative tests over the total pre-pandemic specimens tested. Specificity was further analyzed by sex. All data cleaning, management, and analysis were conducted in the R statistical environment (R Core Team, 2022).

3. Results

3.1. Characteristics and assay seropositivity using pre-pandemic plasma and DBS from NIMR

The first set of samples was collated by NIMR and included pre-pandemic samples (n=100) collected before the advent of the COVID-19 pandemic and SARS-CoV-2 positive plasma specimens (n=100) collected during the COVID-19 pandemic (Table 1 ). The median age for the pre-pandemic plasma, pandemic plasma and pandemic DBS specimens were 39.0, 36.5, and 39.1 years, respectively). The percentage of males (pre-pandemic plasma, pandemic plasma and pandemic DBS specimens) were 43%, 45%, and 49%, respectively. The SARS-CoV-2 positive plasma samples were confirmed using RT-PCR. Additional DBS samples (n=55) were prepared from the RT-PCR positive samples for the serological assay evaluation. Similar to the NAIIS pre-pandemic samples, the seropositivity to NIMR pre-pandemic samples for N protein was highest (8.1%, 14.2%, and 7.0% for Tetracore, SARS2MBA, and xMAP assays, respectively) and lower for the RBD, S, and RBD/N hybrid protein targets (all under 4.0%) (Table 1).

Table 1.

Characteristics of NIMR sample set.

	Pre-Pandemic (Plasma)				Pandemic (Plasma)				Pandemic (DBS)
	TET (n=99)	SARS2MBA (n=98)	xMAP (n=100)	RS (n=100)	TET (n=87)	SARS2MBA (n=98)	xMAP (n=100)	RS (n=100)	TET (n=53)	SARS2MBA (n=53)	xMAP (n=55)	RS
Age, years Median (IQR)	39.0 (33.0-75.0)				36.5 (29.0-44.0)				39.1 (34.2-49.3)
Gender (% Male)	43%				45%				49%
RBD (% Positive)	2.0%	-	-	-	80.5%	-	-	-	71.7%	-	-	-
N Protein (% Positive)	8.1%	-	-	-	93.1%	-	-	-	80.4%	-	-	-
RBD/N Protein Hybrid (% Positive)	1.0%	-	-	-	82.8%	-	-	-	71.4%	-	-	-
S protein (% Positive)	-	1.0%	-	-	-	95.9%	-	-	-	77.4%	-	-
RBD 541 (% Positive)	-	3.1%	-	-	-	93.9%	-	-	-	71.7%	-	-
RBD 591 (% Positive)	-	3.1%	-	-	-	95.9%	-	-	-	83.0%	-	-
N protein (% Positive)	-	14.2%	-	-	-	94.9%	-	-	-	79.3%	-	-
Nucleocapsid (% Positive)	-	-	7.0%	-	-	-	90.9%	-	-	-	63.6%	-
RBD (% Positive)	-	-	2.0%	-	-	-	87.9%	-	-	-	58.2%	-
S1 (% Positive)	-	-	0%	-	-	-	22.2%	-	-	-	7.3%	-
Spike RBD (% Positive)	-	-	-	0%	-	-	-	64.0%	-	-	-	-

TET = Tetracore, SARS2MBA = SARS-CoV-2 assay, xMAP = Luminex xMAP® SARS-CoV-2, RS = RightSign, IQR = Interquartile range, RBD = receptor binding domain, N protein = Nucleocapsid protein, S = Spike, n = Number.

3.2. Characteristics and assay seropositivity using pre-pandemic plasma and DBS from NAIIS

The second sample set was the pre-pandemic NAIIS survey plasma (n=200) and matched DBS (n=100) specimens collected in 2018 (Table 2 ). The median age for the DBS and plasma samples was similar (30.0 versus 32.5 years, respectively). The percentage of males in each was also similar (59% versus 56%, respectively). Seropositivity to the different SARS-CoV-2 antigen targets varied by assays. For the RBD targets tested using the Tetracore assay, 5.9% of DBS and 5.4% of plasma samples were seropositive, respectively. For the RBD targets from the SARS2MBA, 3.9% of DBS were seropositive to RBD 541 and 0.98% seropositive to RBD 591, while that of plasma was 5.4% for RBD 541 and 4.0% for RBD 591. In contrast, the xMAP assay showed that 0.0% of DBS samples were seropositive to their RBD target and 4.0% seropositive with plasma (Table 2). In addition, the result showed 0.0% seropositivity with RightSign assay for its plasma samples against the spike RBD target.

Table 2.

Characteristics of NAIIS sample set.

	DBS				Plasma
	TET (n=100)	SARS2MBA(n=99)	xMAP (n=100)	RS (n=100)	TET (n=200)	SARS2MBA(n=198)	xMAP (n=200)	RS (n=200)
Age, years Median (IQR)	30.0 (25.0-43.50)				32.5 (23.25-44.75)
Gender (% Male)	59%				56%
RBD (% Positive)	5.9%	-	-	-	5.4%	-	-	-
N Protein (% Positive)	42.2%	-	-	-	21.8%	-	-	-
RBD/N Protein Hybrid (% Positive)	7.8%	-	-	-	5.9%	-	-	-
S protein (% Positive)	-	0%	-	-	-	0.5%	-	-
RBD 541 (% Positive)	-	3.9%	-	-	-	5.4%	-	-
RBD 591 (% Positive)	-	0.98%	-	-	-	4.0%	-	-
N protein (% Positive)	-	0.98%	-	-	-	4.0%	-	-
Nucleocapsid (% Positive)	-	-	6.9%	-	-	-	14.9%	-
RBD (% Positive)	-	-	0%	-	-	-	4.0%	-
S1 (% Positive)	-	-	0%	-	-	-	0%	-
Spike RBD (% Positive)	-	-	-	-	-	-	-	0%

TET = Tetracore, SARS2MBA = SARS-CoV-2 assay, xMAP = Luminex xMAP® SARS-CoV-2, RS = RightSign, IQR = Interquartile range, RBD = receptor binding domain, N protein = Nucleocapsid protein, S = Spike, DBS = Dried blood spot, n = Number.

For the SARS2MBA, the seropositivity of the S protein target was 0.0% with DBS and 0.5% seropositivity with plasma. The S1 target for the xMAP assay also showed 0.0% seropositivity with DBS and plasma. However, the seropositivity for the N protein target for the Tetracore assay was much higher than the SARS2MBA and xMAP assays, at 42.2% for DBS and 21.8% for plasma samples. For the N protein target from the SARS2MBA, the seropositivity was 0.98% with DBS and 4.0% with plasma. For the xMAP assay, the seropositivity to its N protein target was 6.9% with DBS and 14.9% with plasma. In addition, the seropositivity to the hybrid RBD/N protein target for the Tetracore assay was 7.8% for DBS and 5.9% for plasma samples.

3.3. Assay sensitivity using NIMR plasma and DBS specimens

In general, plasma sensitivity was high (average of 81.5%), while DBS sensitivity was much lower (63% average). Since algorithms for an overall SARS-CoV-2 seropositive call are dependent on IgG positivity to multiple targets (outlined in Methods), specificity was also high for both plasma and DBS, with a very narrow range (94%-100%). Based on the results, SARS2MBA multi-antigen IgG assay had the highest overall sensitivity (96%) and 100% specificity for both plasma and DBS samples for NIMR and NAIIS specimens. The overall sensitivities were 81% (95% CI: 70% – 88%) for Tetracore, 96% (95% CI: 90% – 99%) for the SARS2MBA, 64% (95% CI: 54% – 73%) for RightSign, and 85% (95% CI: 76% – 91%) for xMAP (Table 3 ). In addition, sensitivity increased across the four assays with additional days post-RT-PCR confirmation of infection. When comparing specimens collected <15 days to ≥15 days post-RT-PCR confirmation, the sensitivities increased from 75% (95% CI: 60% – 86%) to 88.2% (95% CI: 71.7% – 96.2%) for Tetracore; from 93% (83% – 98%) to 100% (95% CI: 89% – 100%) for the SARS2MBA; from 58% (95% CI: 45% – 71%) to 73% (95% CI: 56% – 85%) for RightSign; and from 83% (95% CI: 71% – 91%) to 88% (95% CI: 72 – 95) for xMAP (Table 3). With DBS, there was no positive increase after 15-28 days for the three assays (Tetracore, SARS2MBA, and xMAP).

Table 3.

Overall diagnostic sensitivity, by test and days since RT-PCR confirmation.

	Tetracore			SARS2MBA			RightSign			xMAP
	N	%	95% CI	N	%	95% CI	N	%	95% CI	N	%	95% CI
Plasma Overall 0-3 days 4-7 days 8-14 days 15-28 days 29+ days	82 7 13 28 17 17	80.5 28.6 61.5 92.9 94.1 82.4	(70.3, 88.4) (3.7, 71.0) (31.6, 86.1) (76.5, 99.1) (71.3, 99.9) (56.6, 96.2)	98 10 20 28 20 20	95.9 60.0 100 100 100 100	(89.9, 98.9) (26.2, 87.8) (83.2, 100) (83.2, 100) (83.2, 100) (83.2, 100)	100 10 20 30 20 20	64.0 30.0 40.0 80.0 70.0 75.0	(53.8, 73.4) (6.7, 65.2) (19.1, 63.9) (61.4, 92.3) (45.7, 88.1) (50.9, 91.3)	99 9 20 30 20 20	84.8 33.3 80.0 100 90.0 85.0	(76.2, 91.3) (7.5, 70.1) (56.3, 94.3) (88.4, 100) (68.3, 98.8) (62.1, 96.8)
<15 days	48	75.0	(60.1, 85.9)	58	93.1	(82.5, 97.8)	60	58.3	(44.9, 70.7)	59	83.1	(70.6, 91.1)
≥15 days	34	88.2	(71.7, 96.2)	40	100	(89.1, 100)	40	72.5	(55.9, 84.9)	40	87.5	(72.4, 95.3)
DBS Overall 0-3 days 4-7 days 8-14 days 15-28 days 29+ days	50 20 5 7 14 7	66.0 30.0 40.0 57.1 100 100	(51.2, 78.8) (11.9, 54.3) (5.3, 85.3) (18.4, 90.1) (76.8, 100) (59.0, 100)	53 15 6 9 15 8	75.5 46.7 66.7 66.7 100 100	(61.7, 86.2) (21.3, 73.4) (22.3, 95.7) (29.9, 92.5) (78.2, 100) (63.1, 100)	- - - - - -	- - - - - -	- - - - - -	55 17 6 9 15 8	47.3 11.8 33.3 44.4 80.0 75.0	(33.7, 61.2) (1.5, 36.4) (4.3, 77.7) (13.7, 78.8) (51.9, 95.7) (34.9, 96.8)

CI = Confidence Interval, N = Number

3.4. Assay specificity using NIMR plasma, NAIIS plasma, and DBS specimens

The overall specificities (combined NIMR and NAIIS) were 98% (95% CI: 96% – 99%) for Tetracore, 100% (95% CI: 98% – 100%) for SARS2MBA and RightSign assays, and 99% (95% CI: 96% – 100%) for xMAP assay (Table 4 ). Specificity for DBS NAIIS pre-pandemic specimens was 100% (95% CI: 96% – 100%) for both the SARS2MBA and xMAP assays, and 99% (95% CI: 95% – 100%) for Tetracore (Table 4). The specificity for plasma samples from NAIIS was 100.0% (95% CI: 98% – 100%) for the SARS2MBA and RightSign assays, 99.0% (95% CI: 97% – 100%) for the xMAP assay, and 98% (95% CI: 94% – 99%) for Tetracore (Table 4).

Table 4.

Overall diagnostic specificity, by test and sex.

	Tetracore			SARS2MBA			RightSign			xMAP
	N	%	95% CI	N	%	95% CI	N	%	95% CI	N	%	95% CI
Overall (NIMR + NAIIS)
Plasma	297	98.3	(95.9, 99.4)	296	100	(98.4, 100)	300	100	(98.4, 100)	300	98.7	(96.4, 99.6)
NIMR Specificity
Plasma	97	100	(96.3, 100)	98	100	(96.3, 100)	100	100	(96.4, 100)	100	98.0	(93.0, 99.8)
NAIIS Specificity
Plasma Overall Male Female	200 11288	97.5 96.4 98.9	(94.3, 99.2) (91.1, 99.0) (93.8, 100)	198 112 86	100 100 100	(98.2, 100) (96.8, 100) (95.8, 100)	200 112 88	100 100 100	(98.2, 100) (96.8, 100) (95.9, 100)	200 112 88	99.0 99.1 98.9	(96.5, 100) (95.2, 100) (93.8, 100)
DBS Overall Male Female	100 60 40	99.0 100 97.4	(94.6, 100) (94.1, 100) (86.5, 100)	100 60 40	100 100 100	(96.4, 100) (94.0, 100) (91.4, 100)	- - -	- - -	- - -	100 60 40	100 100 100	(96.4, 100) (94.1, 100) (91.4, 100)

CI = Confidence Interval, N = Number

4. Discussion

Our results showed that the overall sensitivities of the SARS-CoV-2 multi-antigen IgG assays were higher than the RightSign single-antigen IgG assay. When stratified by days post-RT-PCR confirmation of SARS-CoV-2 infection, the sensitivity increased for ≥15 days compared to <15 days post-RT-PCR confirmation of infection. In addition, the SARS2MBA assay had the highest sensitivity compared to the remaining three evaluated assays, and there was no difference in specificity for either plasma or DBS samples by sex. Compared to the previous validation performed in Nigeria [22], these results indicated that the SARS2MBA multi-antigen IgG assay may have shown better sensitivity and specificity for the integrated serosurveillance study in Nigeria. A previous study demonstrated that multi-antigen assays had lower cross-reactivity with pre-pandemic specimens [29] and can provide a reliable serological classification of individuals with SARS-CoV-2 infections [18]. Furthermore, in areas of high background reactivity with other betacoronaviruses and Plasmodium falciparum malaria antigens, the multi-antigen approach provided a reassuring community seroprevalence estimate [30]. In addition, the multi-antigen assay fulfills the required orthogonal testing algorithm with a better and more critical interpretation of the serological status of a patient [31] and provides an in-depth repertoire understanding of antibody responses to SARS-CoV-2 [32].

Previous studies have shown limited specificity with SARS-CoV-2 assays in the African continent [3,4], and cross-reactivity of SARS-CoV-2 serological assays has been documented, especially with African specimens [3,4,6], including Nigeria [7]. In addition, a previous study revealed heterogeneous responses among African COVID-19 patients and underscored the need for a comprehensive evaluation of commercial immunoassays before use for surveillance [33]. The comparative analysis of 16 commercial SARS-CoV-2 assays in a standard laboratory showed differences in performance outcomes [34]. Additionally, another study compared the clinical performance of three SARS-CoV-2 antibody assays and indicated significant differences in the performance of the assays [35]. These observed differences may impact the choice of an assay for COVID-19 seroprevalence studies in countries due to the selected viral antigens, fluctuation in antibody levels, and the isotypes of human chosen antibodies, which is crucial to improve sensitivity and specificity [36]. The beginning of COVID-19 led to the emergency use authorization and listing of SARS-CoV-2 immunoassays and may have enabled the development of some suboptimal assays; some of these were later revoked [37]. Therefore, appropriate validation of SARS-CoV-2 immunoassays before use for COVID-19 serosurveillance studies is crucial to reflect the population profile. Also, the appropriate use of SARS-CoV-2 immunoassay for serosurveillance will provide reliable data needed for effective public health response within the region [38].

The use of DBS to detect antibodies against SARS-CoV-2 in populations provides an alternative to plasma and serum. However, the performance characterization of DBS requires careful validation on SARS-CoV-2 IgG assays as this specimen type awaits further external quality assessment procedures [19]. Though the sensitivity of the assays were lower with DBS compared to plasma in this validation, the reliability of DBS as an alternative to plasma had been documented [19,20,39]. There is the possibility that three years into the COVID-19 pandemic, many people may have been infected at least once. In our study, the sensitivities of the SARS2MBA and Tetracore assays with DBS were both 100% after 15 days post-RT-PCR confirmation. Though the sample sizes were small for DBS, our results indicated that DBS could serve as an alternate, especially for individuals infected for at least 15 days and above. A similar study suggested that the positivity rate of SARS-CoV-2 antibody testing was relatively low within 14 days of symptom onset than 15 days after symptom inception [40]. Therefore, an antibody-based test for epidemiological surveillance after 15 days is recommended.

In conclusion, our results indicated that SARS-CoV-2 multi-antigen assays had higher sensitivity than RightSign (a single-antigen assay), with SARS2MBA having the highest sensitivity. Additionally, the SARS-CoV-2 multi-antigen assays performed well in Nigeria, even with samples with known malaria reactivity, and might provide more accurate measures of COVID-19 seroprevalence and vaccine efficacy.

PEPFAR/CDC funding acknowledgement statement

The funding was from CDC's COVID-19 International Task Force and CARES Act Funds. This study was also supported by Bill and Melinda Gates Foundation and CDC Foundation funding for the Nigeria Multi-disease Serologic Surveillance using Stored Specimens (NMS4) project. NAIIS was supported by the President's Emergency Plan for AIDS Relief (PEPFAR) through CDC under the cooperative agreement U2GGH002108 to UMB and by the Global Funds to Fight AIDS, Tuberculosis, and Malaria through NACA, under the contract number NGA-H-NACA to UMB. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

CDC disclaimer

The findings and conclusions of this study are those of the authors and do not necessarily represent the official positions of the NAIIS Group or the U.S. Centers for Disease Control and Prevention. The use of trade names and commercial sources is for identification and informational purposes only and does not constitute an endorsement by the U.S. Centers for Disease Control and Prevention or the U.S. Department of Health and Human Services.

Ethical approval

This study received ethical approval from the NIMR Institutional Review Board (IRB) and the National Health Research Ethics Committee of Nigeria (NHREC). In addition, this activity was reviewed by CDC as research not involving human subjects and was conducted consistent with applicable federal law and CDC policy 45 CFR 46.102(e).

CRediT authorship contribution statement

Nnaemeka C. Iriemenam: Conceptualization, Visualization, Funding acquisition, Formal analysis, Supervision, Writing – original draft, Writing – review & editing. Fehintola A. Ige: Conceptualization, Visualization, Validation, Formal analysis, Funding acquisition, Writing – review & editing. Stacie M. Greby: Conceptualization, Visualization, Writing – review & editing. Olumide O. Okunoye: Conceptualization, Visualization, Funding acquisition, Writing – review & editing. Mabel Uwandu: Validation, Formal analysis, Writing – review & editing. Maureen Aniedobe: Validation, Formal analysis, Writing – review & editing. Stephnie O. Nwaiwu: Validation, Formal analysis, Writing – review & editing. Nwando Mba: Writing – review & editing. Mary Okoli: Writing – review & editing. Nwachukwu E. William: Writing – review & editing. Akipu Ehoche: Writing – review & editing. Augustine Mpamugo: Writing – review & editing. Andrew Mitchell: Writing – review & editing. Kristen A. Stafford: Writing – review & editing. Andrew N. Thomas: Validation, Formal analysis, Investigation, Writing – review & editing. Temitope Olaleye: Validation, Formal analysis, Funding acquisition, Investigation, Writing – review & editing. Oluwaseun O. Akinmulero: Validation, Formal analysis, Writing – review & editing. Ndidi P. Agala: Validation, Formal analysis, Writing – review & editing. Ado G. Abubakar: Validation, Formal analysis, Writing – review & editing. Ajile Owens: Formal analysis, Investigation, Writing – review & editing. Sarah E. Gwyn: Formal analysis, Investigation, Writing – review & editing. Eric Rogier: Formal analysis, Investigation, Writing – review & editing. Venkatachalam Udhayakumar: Writing – review & editing. Laura C. Steinhardt: Conceptualization, Visualization, Writing – review & editing. Diana L. Martin: Formal analysis, Investigation, Writing – review & editing. McPaul I. Okoye: Conceptualization, Visualization, Writing – review & editing. Rosemary Audu: Conceptualization, Visualization, Writing – review & editing.

Declaration of Competing Interest

The authors have no conflicts of interest to declare.