How Does How Accurate is Sars-cov-2 Antibody Igg Work?

Background

Accu-Tell Product Page

COVID-19 is an infectious disease caused by the SARS-CoV-2 virus that spreads easily between people in a similar way to the common cold or &#;flu&#;. Most people with COVID-19 have a mild-to-moderate respiratory illness, and some may have no symptoms (asymptomatic infection). Others experience severe symptoms and need specialist treatment and intensive care.

In response to COVID-19 infection, the immune system develops proteins called antibodies that can attack the virus as it circulates in their blood. People who have been vaccinated against COVID-19 also produce these antibodies against the virus. Tests are available to detect antibodies in peoples' blood, which may indicate that they currently have COVID-19 or have had it previously, or it may indicate that they have been vaccinated (although this group was not the focus of this review).

Why are accurate tests important?

Accurate testing allows identification of people who need to isolate themselves to prevent the spread of infection, or who might need treatment for their infection. Failure of diagnostic tests to detect infection with COVID-19 when it is present (a false negative result) may delay treatment and risk further spread of infection to others. Incorrect diagnosis of COVID-19 when it is not present (a false positive result) may lead to unnecessary further testing, treatment, and isolation of the person and close contacts. Accurate identification of people who have previously had COVID-19 is important in measuring disease spread and assessing the success of public health interventions.

To determine the accuracy of an antibody test in identifying COVID-19, test results are compared in people known to have (or have had) COVID-19 and in people known not to have (or have had) COVID-19. The criteria used to determine whether people are known or not known to have COVID-19 is called the &#;reference standard&#;. Many studies use a test called reverse transcriptase polymerase chain reaction (RT-PCR) as the reference standard, with samples taken from the nose and throat. Additional tests that can be used include measuring symptoms, like coughing or high temperature, or &#;imaging&#; tests like chest X-rays. People known not to have COVID-19 are sometimes identified from stored blood samples taken before COVID-19 existed, or from patients with symptoms confirmed to be caused by other diseases.

What did the review study?

We wanted to find out whether antibody tests:

- are able to diagnose infection in people with or without symptoms of COVID-19, and

- can be used to find out if someone has already had COVID-19.

The studies we included in our review looked at three types of antibodies. Most commonly, antibody tests measure two types known as IgG and IgM, but some tests only measure a single type of antibody or different combinations of the three types of antibodies (IgA, IgG, IgM).

What did we do?

We looked for studies that measured the diagnostic accuracy of antibody tests to detect current or past COVID-19 infection and compared them with reference standard criteria. Since there are many antibody tests available, we included studies assessing any antibody test compared with any reference standard. People could be tested in hospital or in the community. The people tested may have been confirmed to have, or not to have, COVID-19 infection, or they may be suspected of having COVID-19.

Study characteristics

We found 178 relevant studies. Studies took place in Europe (94), Asia (45), North America (35), Australia (2), and South America (2).

Seventy-eight studies included people who were in hospital with suspected or confirmed COVID-19 infection and 14 studies included people in community settings. Several studies included people from multiple settings (35) or did not report where the participants were recruited from (39).

One hundred and forty-one studies included recent infection cases (mainly week 1 to week 3 after onset of symptoms), and many also included people tested later (from day 21 onwards after infection) (117).

Main results

In participants that had COVID-19 and were tested one week after symptoms developed, antibody tests detected only 27% to 41% of infections. In week 2 after first symptoms, 64% to 79% of infections were detected, rising to 78% to 88% in week 3. Tests that specifically detected IgG or IgM antibodies were the most accurate and, when testing people from 21 days after first symptoms, they detected 93% of people with COVID-19. Tests gave false positive results for 1% of those without COVID-19.

Below we illustrate results for two different scenarios.

If people were tested for IgG or IgM antibodies during the third week after onset of symptoms and only 20 (2%) of them actually had COVID-19:

- 26 people would test positive. Of these, 8 people (31%) would not have COVID-19 (false positive result).

- 974 people would test negative. Of these, 2 people (0.2%) would actually have COVID-19 (false negative result).

If people with no symptoms for COVID-19 were tested for IgG antibodies and 500 (50%) of them had previously had COVID-19 infection more than 21 days previously:

- 455 people would test positive. Of these, 6 people (1%) would not have been infected (false positive result).

- 545 people would test negative. Of these, 51 (9%) would actually have had a prior COVID-19 infection (false negative result).

How reliable were the results of the studies of this review?

We have limited confidence in the evidence for several reasons. The number of samples contributed by studies for each week post-symptom onset was often small, and there were sometimes problems with how studies were conducted. Participants included in the studies were often hospital patients who were more likely to have experienced severe symptoms of COVID-19. The accuracy of antibody tests for detecting COVID-19 in these patients may be different from the accuracy of the tests in people with mild or moderate symptoms. It is not possible to identify by how much the test results would differ in other populations.

Who do the results of this review apply to?

A high percentage of participants were in hospital with COVID-19, so were likely to have more severe disease than people with COVID-19 who were not hospitalised. Only a small number of studies assessed these tests in people with no symptoms. The results of the review may therefore be more applicable to those with severe disease than people with mild symptoms.

Are you interested in learning more about How Accurate is Sars-cov-2 Antibody Igg? Contact us today to secure an expert consultation!

Studies frequently did not report whether participants had symptoms at the time samples were taken for testing making it difficult to fully separate test results for early-phase infection as opposed to later-phase infections.

The studies in our review assessed several test methods across a global population, therefore it is likely that test results would be similar in most areas of the world.

What are the implications of this review?

The review shows that antibody tests could have a useful role in detecting if someone has had COVID-19, but the timing of test use is important. Some antibody tests may help to confirm COVID-19 infection in people who have had symptoms for more than two weeks but who have been unable to confirm their infection using other methods. This is particularly useful if they are experiencing potentially serious symptoms that may be due to COVID-19 as they may require specific treatment. Antibody tests may also be useful to determine how many people have had a previous COVID-19 infection. We could not be certain about how well the tests work for people who have milder disease or no symptoms, or for detecting antibodies resulting from vaccination.

How up-to-date is this review?

This review updates our previous review. The evidence is up-to-date to September .

Donate

If you found this evidence helpful, please consider donating to Cochrane. We are a charity that produces accessible evidence to help people make health and care decisions.

ABSTRACT.

Determination of previous SARS-COV-2 infection is hampered by the absence of a standardized test. The marker used to assess previous exposure is IgG antibody to the nucleocapsid (IgG anti-N), although it is known to wane quickly from peripheral blood. The accuracies of seven antibody tests (virus neutralization test, IgG anti-N, IgG anti-spike [anti-S], IgG anti&#;receptor binding domain [anti-RBD], IgG anti-N + anti-RBD, IgG anti-N + anti-S, and IgG anti-S + anti-RBD), either singly or in combination, were evaluated on 502 cryopreserved serum samples collected before the COVID-19 vaccination rollout in Kumasi, Ghana. The accuracy of each index test was measured using a composite reference standard based on a combination of neutralization test and IgG anti-N antibody tests. According to the composite reference, 262 participants were previously exposed; the most sensitive test was the virus neutralization test, with 95.4% sensitivity (95% CI: 93.6&#;97.3), followed by 79.0% for IgG anti-N + anti-S (95% CI: 76.3&#;83.3). The most specific tests were virus neutralization and IgG anti-N, both with 100% specificity. Viral neutralization and IgG anti-N + anti-S were the overall most accurate tests, with specificity/sensitivity of 100/95.2% and 79.0/92.1%, respectively. Our findings indicate that IgG anti-N alone is an inadequate marker of prior exposure to SARS COV-2 in this population. Virus neutralization assay appears to be the most accurate assay in discerning prior infection. A combination of IgG anti-N and IgG anti-S is also accurate and suited for assessment of SARS COV-2 exposure in low-resource settings.

INTRODUCTION

The coronavirus disease (COVID-19) continues to affect the global community at an unprecedented rate. Although vaccines designed against SARS COV-2 have been the cornerstone of the global strategy against COVID-19, limited vaccine supply, poor vaccine uptake, and vaccine hesitancy has led to delays in achieving population-level immunity in selected populations,1,2 especially across the sub-Saharan African region. One study previously showed that almost half of healthcare workers and one-third of unvaccinated participants were previously infected with SARS COV-2, using anti-nucleocapsid (anti-N) seropositivity prior to national vaccine rollout in Nigeria and Ghana, respectively.3

Exposure to SARS COV-2 antigens driven by previous exposure shapes immunity to SARS COV-2 in individuals and at the population level.4 Anti-N and anti-spike (anti-S) markers appear after infection and are used to differentiate between immune responses elicited by infection or vaccination.5 The natural course of SARS COV-2 infection involves the appearance and persistence of anti-N and anti-S markers for several months post-infection.6&#;9 It has been observed that the anti-N antibody has a shorter half-life and persistence than anti-S antibodies,7 with the anti-N antibody peaking at around 30 days post-infection.6 Further, evidence from East Africa has shown that the anti-N antibody did not appear among one in four study participants within 1 month of a confirmed polymerase chain reaction (PCR) test.10

Recognizing unvaccinated individuals with previous infection is an important component of managing SARS COV-2 infection as it relates to contact tracing, isolation, and refining the appropriate intervention strategies for vaccination and case management, especially where PCR testing is not readily available and accessible at the population level. It should be noted that vaccination coverage is still patchy; for example, only 59.3% in Nigeria and 44% in Ghana of the eligible population are fully vaccinated.11 In this study, using cross-sectional samples collected prior to vaccination rollout in Kumasi, Ghana, in , we characterized four markers of SARS COV-2, including serum virus neutralization titer, total IgG binding antibodies for anti-S, IgG anti-N, and IgG anti&#;receptor binding domain (anti-RBD), with the aim of evaluating the ideal marker or combination of markers of characterizing previous SARS COV-2 infection.

MATERIALS AND METHODS

Study design and participants.

Our study was a retrospective cross-sectional comparative analysis of markers of SARS COV-2 on bio-archived anonymized samples. The study cohort comprised 502 participants from the general population of the Kumasi area as previously described.3 These participants were earmarked to receive the SARS COV-2 ChadOx-1 (AstraZeneca) vaccine in early . Study participants provided written informed consent, and blood samples were collected according to the protocols approved by the Committee of Human Research, Publication and Ethics of KNUST (CHRPE/AP/091/21).

Binding antibody and viral neutralization antibody testing.

We measured binding IgG antibodies against SARS-COV-2 receptor-binding domain (RBD), total trimeric spike protein (S), and nucleocapsid protein (N) using the Luminex-based SARS-CoV-2-IgG assay as we previously detailed.12,13 We defined the cutoff of each antibody using an analysis of &#;true&#; positive (convalescent) and negative pre-pandemic samples as we previously described.3 In brief, positive binding antibodies were defined using a threshold of 1,896, 456, and 6,104 mean fluorescence intensity for IgG anti-S, anti-RBD, and anti-N IgG, respectively.3 For plasma-neutralizing antibody measurement, SARS-CoV-2 virus (pseudotyped virus [PV]) was prepared by transfecting HEK293T cells with Wu-1-614G wild type using p8.91 HIV-1 gag-pol expression vector.14 Virus neutralization was performed on Hela-ACE2 cells using SARS-CoV-2 spike PV-expressing luciferase. Briefly, plasma samples were heat inactivated at 54°C for 1 hour, serially diluted in duplicate, and incubated with PVs at 37°C for 1 hour prior to addition of Hela-ACE2 cells.15 The plasma dilution/virus mix was incubated for 48 hours in a 5% CO2 environment at 37°C, and luminescence was measured using the Bright-Glo Luciferase assay system (Promega). All neutralization assays were repeated in two independent experiments containing two technical replicates for each condition. Neutralization was calculated relative to virus-only controls as a mean neutralization with standard error of the mean. The half maximum inhibitory dose (ID50) was calculated in GraphPad Prism version 9.3.1; ID50 > 20 was considered positive. 293T cells (ATCC: CRL-) and HELA-ACE2 cells were a kind gift from Dr. James Voss, SCRIPPS.

Definition of &#;composite gold standard&#; and analyses.

In the absence of a &#;gold standard&#; for evaluating previous SARS COV-2 infection, we defined composite gold standard as positivity to either IgG anti-N or viral neutralization (ID50 > 50). The sensitivity of each &#;index test&#; was evaluated as the proportion of positive result over the positive specimens using the composite reference standard, and the specificity of each index test was evaluated as the proportion of negative results over the negative specimens using the composite reference standard using the defined thresholds. Uncertainty was quantified using 95% CIs, and corresponding receiver operating characteristic (ROC) graphs were plotted for each of the index tests.

RESULTS

The study population of 502 participants had a median age of 33 years (interquartile range [IQR]: 25&#;47), of which the majority participants were male (280/502; 56.0). No significant difference in age was observed between the two population groups in terms of positivity using the composite gold standard as strata for previous exposure (P = 0.32). Of 502 pre-vaccination participant samples, 262 (52.2%) were positive as per composite gold standard and used as the denominator for sensitivity analysis, and 240 (47.8%) were negative and used as the denominator for specificity analysis.

Using the composite standard as reference and evaluating the accuracy of IgG anti-N as the most accepted marker of previous exposure, we found high percentages of false negatives (132/262, 50.4%). The proportion of participants who were true positives using the different markers either singly or in combination is shown in Table 1. It is important to note that viral neutralization assay (250/262; 95.4%) and IgG anti-N + anti-S (226/262; 86.3%) had the highest proportions of true positives (Table 2). The overall accuracy of index tests is summarized in Table 2. The most sensitive tests were the neutralization test and the combination of IgG anti-N + anti-S, with sensitivity of 95.4% and 79.0%, respectively; the least sensitive test was IgG anti-N alone at 52.7% (Table 2).

Table 1.

Test accuracy of seven index tests using different markers of SARS COV-2 infection either singly or in combination

Test Positives, n (% of total sample) True positive, sensitivity (%) Negatives, n (% of total sample) True negatives, specificity (%) Virus neutralization 250 (49.8) 250 (95.4) 252 (50.2) 240 (100) IgG anti-N 138 (27.5) 138 (52.7) 364 (72.5) 240 (100) IgG anti-S (total) 218 (43.4) 199 (76.0) 284 (56.6) 221 (92.1) IgG anti-RBD 195 (38.8) 181 (69.1) 307 (61.2) 226 (94.2) IgG anti-N + anti-RBD 213 (42.4) 199 (76.0) 289 (57.6) 226 (94.2) IgG anti-N + anti-S 226 (45.0) 207 (79.0) 276 (55.0) 221 (92.1) IgG anti-RBD + anti-S 223 (44.4) 201 (76.7) 279 (55.6) 218 (91.0) Open in a new tab

Table 2.

Test accuracy of seven index tests using different markers of SARS COV-2 infection either singly or in combination showing AUC values, specificity, sensitivity, PPV, and NPV with 95% CI

Test AUC (95% CI) Specificity (95% CI) Sensitivity (95% CI) PPV (95% CI) NPV (95% CI) Virus neutralization 0.98 (0.96&#;0.99) 100 (98.5&#;100) 95.4 (93.6&#;97.3) 100 (98.5&#;100) 95.2 (93.4&#;97.1) IgG anti-N 0.76 (0.73&#;0.79) 100 (98.5&#;100) 52.7 (48.3&#;57.0) 100 (98.5&#;100) 66.0 (62.0&#;70.0) IgG anti-S (total) 0.84 (0.81&#;0.87) 92.1 (89.7&#;94.5) 76.0 (72.2&#;80.0) 91.3 (88.8&#;93.6) 77.8 (74.1&#;81.5) IgG anti-RBD 0.82 (0.78&#;0.85) 94.2 (92.1&#;96.2) 69.1 (65.0&#;73.1) 92.8 (90.6&#;95.1) 73.6 (69.8&#;77.5) IgG anti-N + anti-RBD 0.85 (0.82&#;0.88) 94.2 (92.1&#;96.2) 76.0 (72.2&#;79.7) 93.4 (91.3&#;95.6) 78.2 (74.6&#;81.8) IgG anti-N + anti-S 0.86 (0.83&#;0.89) 92.1 (89.7&#;94.5) 79.0 (75.5&#;82.6) 91.6 (89.2&#;94.0) 80.0 (76.6&#;83.6) IgG anti-RBD + anti-S 0.84 (0.80&#;0.87) 91.0 (88.3&#;93.4) 76.7 (73.0&#;80.4) 90.1 (87.5&#;92.7) 78.1 (74.5&#;56.6) Open in a new tab

Regarding specificity, IgG anti-N and virus neutralization were the most specific tests (100%) according to our composite gold standard, followed by the combination of IgG anti-N + IgG anti-RBD and IgG anti-N at 94%. The other index tests were also highly specific at > 90%. The virus neutralization test and IgG anti-N had the highest positive predictive values (PPVs) at 100%, and the serum neutralization test (95.2%) and IgG anti-N + anti-S (80.0%) had the highest negative predictive values (NPVs) (the PPV and NPV are dependent on the population prevalence of prior COVID-19 infection). Comparing the ROC curves for each index test&#;s ability to discriminate previous exposure (Figure 1), virus neutralization test still had the highest area under the curve (AUC: 0.98), reflective of excellent discriminatory ability to exposure to SARS COV-2. Other index tests had acceptable AUC values above 0.82, with IgG anti-N antibodies giving an AUC value of 0.76 (Figure 1).

Figure 1.

Open in a new tab

Receiver operating characteristic analyses showing the area under the curve across seven index tests used to evaluate the best markers of SARS COV-2 exposure. Positivity to either IgG anti-N or detectable virus neutralization was used as composite gold standard.

DISCUSSION

Accurate characterization of previous exposure to SARS COV-2 and other related respiratory pathogen exposure is a critical component of pandemic management, including accurate evaluation of population-level seroprevalence to inform public health interventions/policy16 and vaccination strategies, including possible fractional dosing vaccination approaches.17,18 We show definitive evidence from a West African population of the indiscriminatory capacity of the IgG anti-N to accurately characterize previous exposure to SARS COV-2, which is likely to lead to an underestimation of seroprevalence and previous exposure.

In this present study, we showed the most accurate test based on sensitivity and specificity was PV neutralization, with 95% sensitivity and 100% specificity. This was consistent with the PPV and NPV of 100% and 95%, respectively. Accurate detection of different markers of SARS COV-2 for characterization of previous exposure has its advantages and drawbacks. Although we used a highly sensitive Luminex approach to measure binding antibodies, we believe that serological markers measured directly at the bench using ELISA should produce similar results. These tests have a quick turnaround time and accuracy (sensitivity and specificity), although the inability to discern between coronaviruses responses resulting from cross-reactivity due to viral sequence homology19,20 is a particular concern.

As the SARS COV-2 pandemic continues, novel variants of concern with mutations enhancing transmissibility, replication, and immunity evasion capacity continue to emerge as observed in the Delta21&#;25 and Omicron variants,15 deriving from chronic infections.26 These novel variants may be associated with a difference in the antibody kinetics of markers, such as the anti-N observed with neutralization responses. Likewise, as vaccination coverage expands globally, the emergence of novel variants that are likely to lead to breakthrough infection in already vaccinated or previously infected individuals will become increasingly common,27,28 which further affects antibody kinetics. It is noteworthy that during the early phase of vaccination scale-up, there were limited reported cases of breakthrough infections in high-income settings;29,30 however, this changed with the arrival of the immunity-evasive Delta variant.21,25,31,32 Similarly, in our previous analysis from Nigeria, we observed a breakthrough infection rate of 16% following two doses of the Chad-Ox1 vaccine in a healthcare worker cohort3 during the Delta variant wave of and relatively similar rates were observed in Uganda33.

Data on the longitudinal trajectories of SARS COV-2 IgG antibodies in western cohorts are conflicting. Two reports showed sustained responses to IgG spike and nucleocapsid up to 125 days after exposure,34,35 whereas other reports reported declines in antibody levels over the same time period,36&#;38 with limited antibody data. In our previous analysis in a West African population comprising 140 and 527 participants from Nigeria and Ghana and using IgG anti-N, we reported seroprevalence rates of 44% and 28%, respectively, which increased to 59% and 39% when IgG anti-RBD was used as an additional marker of previous infection due to its specificity for the SARS COV-2 epitope.3 Our data from Ghana are consistent with evidence from Lagos in participants from the general population prior to vaccination rollout in early , which showed underestimation of previous infection when SARS-CoV-2 IgG anti S + anti-N or IgG anti&#;S-only responses were measured by T-cell interferon-γ assay for use as a diagnostic biomarker for previous exposure.39 T cell responses have also been observed in antibody negative participants in Kenya.40 This indicates that anti-N is not an ideal independent marker and that additional markers are required for accurate discernment of previous exposure, although time since exposure event is likely to play a role because the antibody response wanes with time.

Our study size is robust with over 500 participants, although demographic data were limited to sex and age. Previous evidence has shown age-related predictors (likely driven by immune senescence) of waning humoral response to vaccination and, in principle, humoral responses triggered by previous SARS COV-2.40&#;43 We found no difference between the ages of participants classified positive or negative by composite gold standard (i.e., median age of 34 [IQR: 26&#;47] and 33 [IQR: 25&#;47], respectively; P = 0.32); this is reassuring and suggests that study participants likely have robust immune systems, with similar waning rate between groups.

This study was subject to limitations. It was advantageous to have access to a large cohort of samples collected pre-vaccination within a West African setting, which is understudied. We note the collection of limited study demographic data and the lack of administered questionnaires, which could have included self-reported measures of potentially previous exposure. Finally, our use of a Luminex high-performance assay for IgG antibody measurement should be considered more than a simplified approach especially in the absence of a comparative analysis with standard on the bench ELISA kits of comparable accuracy.

We conclude that IgG anti-N alone is an inadequate marker of prior exposure to SARS COV-2 in an unvaccinated population with an underestimation of previous exposure by almost 50%. In our hands, virus neutralization assay was the most accurate assay in discerning prior infection in unvaccinated populations. In resource-limited settings where virus neutralization assay is not easily available, a combination of IgG anti-N and IgG anti-S is potentially a suitable alternative.

ACKNOWLEDGMENTS

We are thankful to the volunteers who participated in the study. The American Society of Tropical Medicine and Hygiene has waived the Open Access fee for this COVID-19 article.

If you want to learn more, please visit our website viral transport swab.