Combining antibody markers to improve serosurveillance for SARS-CoV-2

By Nidhi Saha, BDSReviewed by Benedette Cuffari, M.Sc.Sep 19 2021

Seroepidemiological studies can provide accurate insights into viral infection fatality rates and the degree of transmission in different populations. Such studies also provide an insight into disparities in infection rates without health-seeking behavior biases. Serosurveillance data is useful in estimating the burden of disease and amplitude of population immunity.

Study: Combining antibody markers for serosurveillance of SARS-CoV-2 to estimate seroprevalence and time-since-infection. Image Credit: Yurchanka Siarhei / Shutterstock.com

This news article was a review of a preliminary scientific report that had not undergone peer-review at the time of publication. Since its initial publication, the scientific report has now been peer reviewed and accepted for publication in a Scientific Journal. Links to the preliminary and peer-reviewed reports are available in the Sources section at the bottom of this article. View Sources

About the study

A recent study supported by the National Institutes of Health found that a single spike (S) or receptor-binding domain (RBD) antibody was adequate for confirming prior severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection. Additionally, the estimate of time since coronavirus disease 2019 (COVID-19) infection could be better deduced through a combination of two antibody biomarkers than any single marker.

A new study posted to the medRxiv* server aimed to highlight the best antibody biomarkers for the estimation of seroprevalence and the time elapsed since SARS-CoV-2 infection. Additionally, the researchers of the current study were interested in determining whether a combination of antibody biomarkers could improve such estimations.

The tudy entailed building models to examine which biomarker(s) are best for identifying prior SARS-CoV-2 infection and for prediction of time-since-infection. For this, serological data from five studies, including confirmed SARS-CoV-2 infected patients, was used.

Studies measuring multiple antibody responses at varying time-points that were greater than a median of 50 days after polymerase-chain-reaction (PCR)-confirmed SARS-CoV-2 infection, were selected. The antibody responses examined included immunoglobulin G (IgG), IgM, and IgA responses against S, RBD, and nucleocapsid (N) antigens.

For each serologic measurement, the time-since-infection was recorded. Random forest models were used to determine the order of importance of the biomarkers and to make final predictions.

As the biomarkers were highly correlated, their importance was measured using the conditional permutation importance algorithm. Two independent models were developed. The first model was used to identify PCR-confirmed infections using biomarkers, whereas the second estimated the time since infection among those previously infected.

For the first model, PCR-confirmed cases and pre-pandemic controls were selected to train a random forest model. The performance of these models was assessed using the cross-validated area under the curve.

For the second model, only confirmed cases were included to train a random forest model for estimating time-since-infection. Performance was assessed using cross-validated mean absolute error (MAE), which depicted the mean of absolute differences of predictions from the true time since infection.

The findings of this study revealed that S/RBD IgG antibodies alone are adequate for the discrimination and classification of individuals infected with SARS-CoV-2. Meanwhile, combinations of antibody markers may be best for estimation of time since infection.

In fact, the use of three or more biomarkers could confer a slightly better estimate of the time elapsed since infection than only two biomarkers. It was also noted that clinical demographic factors such as age and severity of disease were potentially important predictors that should be considered in model-building.

Implications

Identifying prior infections with SARS-CoV-2 in vaccinated populations will aid in deducing the rate of spread. Such population-level surveys can guide predictions on the probability and timing of future waves of SARS-CoV-2 infection. Additionally, these methods can measure the impact of preventive measures or vaccination, and may even confirm the absence of transmission as and when that is achieved.

Taken together, such results can inform the design of serosurveys for SARS-CoV-2, including decisions regarding the number of antibody biomarkers measured.

The results of this study contribute towards guiding the choice of antibody responses for future seroepidemiological investigations of SARS-CoV-2. Distinguishing vaccine-induced immune responses from those triggered by natural infection is important in the design of seroepidemiologic studies.

The most widely used antibody markers for SARS-CoV-2 serosurveillance include the IgG to S or RBD, which are also the target of currently approved vaccines in the United States and the European Union. Thus, future serosurveillance efforts may depend on the nucleocapsid antibody.

Despite the inadequacy of nucleocapsid-specific IgG alone in classifying infection, combinations of N antibodies may provide better results in distinguishing the background of SARS-CoV-2 infection.

The two best N antibodies fared better in estimating the time-since-infection than any single S or RBD antibody. Hence, assessing nucleocapsid-specific antibodies may be useful in differentiating infection-related antibodies from vaccine-induced antibody responses.

Strengths and limitations

The results of this study highlighted SARS-CoV-2 antibody responses across diverse populations. For the estimation of time-since-infection, the researchers found that a combination of biomarkers should be preferred. Moreover, the use of three or more biomarkers could confer a slightly better estimate than only two biomarkers.

The present study analyzed adults in high-income countries. Therefore, the results discussed here cannot be generalized to low- and middle-income countries or for pediatric populations.

Conclusions

The current study found that combining N antibodies with nucleocapsid-specific IgG facilitates a better classification of SARS-CoV-2 infection. While monitoring nucleocapsid-specific antibodies can effectively distinguish infection-related antibodies from vaccine-induced antibody response, it can also aid in estimating the time-since-infection.

This news article was a review of a preliminary scientific report that had not undergone peer-review at the time of publication. Since its initial publication, the scientific report has now been peer reviewed and accepted for publication in a Scientific Journal. Links to the preliminary and peer-reviewed reports are available in the Sources section at the bottom of this article. View Sources