Abstract
Background Estimates of seroprevalence to SARS-CoV-2 vary widely. We ascertained IgG levels across a single US metropolitan site, Chicago, over the 2020 summer, a period when restrictions on activities had been lifted.
Methods We utilized a self-sampled dried blood spot assay to quantitatively monitor antibodies to the receptor binding domain (RBD) of the spike glycoprotein of SARS-CoV-2 in 1545 participants, with return of blood spot cards either by mail or in-person drop-off.
Results Seroprevalence was 19.8%, with no significant difference between method of contact, or between essential and non-essential workers. Only a small number (1.2%) of participants reported having had a diagnosis of COVID-19 based on virus detection, consistent with a 16-fold greater exposure to SARS-CoV-2 measured by serology than detected by viral testing. Only a modest correlation was observed between having antibodies to SARS-CoV-2 nucleocapsid compared to RBD, with many only having detectable anti-RBD antibodies. From a subset of those who participated in repeat testing, three-quarters of seropositive individuals retained detectable antibodies for at least 120 days. One seropositive individual experienced a strong boost in IgG levels following a symptomatic illness, suggestive of potential re-exposure.
Conclusions These data underscore the importance of a self-collected, quantitative assay with adequate sensitivity to detect antibodies at the lower levels among non-hospitalized persons with community-acquired exposure to COVID-19.
Introduction
The presence of serum antibodies to SARS-CoV-2, the virus that causes COVID-19, reflects prior exposure, and some antibodies may confer protection against subsequent infection. Seroprevalence, or seropositivity, estimates range from 3-50% depending on the population surveyed, method of testing, as well as including the viral antigen target (1-13). Most tests measure antibody content in blood or serum, which often necessitates contact with a health care facility, which can introduce sampling bias. Point-of-care lateral flow devices have also been employed but lack both the sensitivity and specificity of a laboratory-performed measurement (3, 14).
Dried blood spots (DBS) can be easily collected in the home setting using a simple finger prick method. We previously described the development, sensitivity and specificity of a DBS test to qualitatively measure IgG to the receptor binding domain of the SARS-CoV-2 protein using a laboratory-based enzyme-linked immunosorbent assay (ELISA)(15). This assay indexes IgG level to an antibody with known affinity to yield a concentration in micrograms per ml (µg/ml). Here, a combination of DBS and laboratory-based ELISA was utilized in a sample of 1545 participants across Chicago, Illinois in the United States. Approximately half of the participants were recruited from throughout Chicago through web-based advertising, received and returned their DBS test kits through the United States Postal Service. The remaining participants were primarily non-healthcare-providing students, staff, and faculty affiliated with a medical school who received and returned DBS kits to have minimal contact with the study staff.
The total cohort included a mix of individuals who were deemed “essential” defined by their report of working outside the home during the shelter-in-placer order in spring of 2020, and “nonessential”, indicating they reported worked from home during this time. Chicago and much of Illinois were under stay-at-home orders that extended from March 21 through May 30, 2020. The sampling began in late June 2020, in a period when local shelter-in-place orders were partially relaxed and continued through September 6, 2020. Overall, the seroprevalence of IgG was 19.8%, with similar seroprevalence among samples ascertained through mail and those obtained through onsite DBS kit distribution. Additionally, we observed no significant difference between seroprevalence in participants self-reporting being “essential” and “non-essential” participants. Three-quarters of seropositive cases retained antibodies at 4 months, and there was serologic evidence of re-exposure in at least one individual. These data highlight the importance of quantitative, self-collected seroprevalence in monitoring the COVID-19 pandemic response.
Methods
Study approvals
All research activities were implemented under protocols approved by the institutional review board at Northwestern University (#STU00206652, #STU00212371, #STU00212457, #STU00212472, and STU00212515).
Study Recruitment
Participants were recruited through two mechanisms. Community-based participants were recruited from ten zip codes in Chicago through social media advertising and news articles. Alternatively, staff, students and faculty from the Northwestern University Feinberg School of Medicine in Chicago IL were sent an email describing the study with a link to the website. Participants were screened for eligibility (zip code and demographics or affiliation to Northwestern). Eligible participants were then invited to complete a questionnaire regarding health status, including COVID-19 symptoms. Community participants received materials for DBS collection through the USPS and returned their test kits using prepaid USPS envelopes provided to them by the study team. Those affiliated with Northwestern were given a specific time to collect DBS kits in person and were instructed to return their completed kits to a secure, unmanned collection box. Sample collection occurred between June 24 and September 6, 2020. Samples from 2018 (n=23) were used as negative controls. Comparator samples derived from 40 COVID-19 non-hospitalized cases, 22 hospitalized cases, and 110 samples from a healthcare worker study selected from a larger cohort of 1790 samples (16). These 110 samples were selected based on a high expected probability of concordance and discordance between the nucleocapsid IgG and RBD assays. For example, the sample was enriched for participants specifically who were nucleocapsid IgG+ and PCR+ (concordance) and for participants with COVID-19 symptoms and exposures who were nucleocapsid IgG negative (discordance). Twenty-four seropositive participants were used from a study starting April 24, 2020 (15) that was resampled through October 5, 2020 and 30 confirmed SARS-CoV2 positive samples.
Serological assay
The ELISA protocol was validated for DBS and previously described (15, 17). As in the prior assay, samples were run in duplicate and reported as the average. Results were normalized to the CR3022 antibody with known affinity (18). Participant sample anti-RBD IgG concentration (µg/ml) was calculated from the 4PL regression of the CR3022 calibration curve. A value >0.39µg/ml CR3022 was considered positive.
Statistical Analysis
Statistical analyses were performed with Prism (GraphPad, La Jolla, CA) or R 4.2 (The R Foundation for Statistical Computing, http://www.R-project.org). The difference between groups used Wilcoxon-Mann-Whitney Test or the Kruskal-Wallis test if more than one group was tested. Two-sample Kolmogorov-Smirnov test was used to compare essential vs non-essential samples. Changes over time were depicted with first order locally weighted regression and smoothing (Lowess model). An unpaired two-tailed t-test was used to compare two groups, where appropriate. Pearson’s chi-squared test statistic was used to compare proportions.
Results
Twenty percent of SCAN participants had IgG antibodies to SARS-CoV-2
The Screening for Coronavirus Antibodies in Neighborhood (SCAN) study uses an at-home testing strategy to measure IgG antibodies to the receptor binding domain (RBD) of the SARS-CoV-2 spike protein (15). Participants received a kit to provide a dried blood spot sample (DBS) from a finger prick, and DBS cards were used in a quantitative laboratory-performed ELISA. Between June 24 and September 6, 2020, 1545 SCAN samples were collected. Of SCAN participants, 19 (1.23%) reported having COVID-19 from a prior, positive diagnostic test for SARS-CoV-2 virus. Eight hundred and fifteen participants were recruited through web-based recruiting, who subsequently received and returned DBS test materials through the mail (no-contact method). An additional 730 were solicited through email and an in-person contact to provide and retrieve the DBS materials (contact method). Seropositivity in SCAN participants utilizing the no-contact method was 19.3% (n=157), while the seropositivity among the group who used the contact method was 20.4% (n=149) (odds ratio 0.93; p= 0.6; CI 0.72-1.20).
Of the 1545 SCAN participants, 19 (1.23%) reported having COVID-19 with a prior positive virus test, with 18 of 19 (94.7%) being RBD IgG seropositive. Of these 1545 participants, 306 (19.8%) were seropositive and 1239 (80.2%) were seronegative. This calculates to 16x more seropositive samples than confirmed SARS-CoV-2 nucleic acid positivity in the SCAN cohort. Seropositivity was similar between males 19.7% (124 of 629) and females 19.9% (182 of 916) (Table 1). Seropositivity by age group varied slightly from the average positivity of 18.9% with 23.5% (18-29 yrs), 17.7 % (30-39 yrs) 23.5% (40-49 yrs), 18.2% (50-59 yrs), and 12.7% (60+ yrs) (Table 2).
Overlapping IgG serum levels in SCAN with known COVID-19 cases
The CR3022 antibody has known affinity for the target antigen, making it possible to quantify IgG directed at RBD. Seronegative samples had a similar median concentration of IgG to pre-COVID-19 samples (0.11 µg/ml vs 0.09 µg/ml). Of the seropositive samples from SCAN, the IgG median concentration was 0.62 µg/ml. As a comparison, the median IgG from those with a COVID-19 diagnosis based on self-reported virus detection (n=96), but who did not require hospitalization, was 5.2 µg/ml, while the median IgG from those requiring ICU hospitalization was 98.5 µg/ml (Figure 1). The considerable overlap between the IgG range seen in the community-acquired SCAN samples and non-hospitalized COVID-19 cases suggests a similar immune response between these two seropositive groups (Figure 1).
Samples were acquired through the Screening for Coronavirus Antibodies in Neighborhood (SCAN) between June 24, 2020 and September 6, 2020 (n= 1545). A) Overlap between the IgG range seen in the community-acquired SCAN seropositive samples (light purple) and non-hospitalized COVID-19+ seropositive samples (dark purple, far right). B) 19.8% (306 of 1545) of SCAN samples were seropositive with a median IgG concentration of 0.62 µg/ml SCAN for the seropositive group. The median concentration of the seronegative SCAN group was 0.11 µg/ml. As a comparator, shown is the range of IgG detected from 96 non-hospitalized and 22 ICU hospitalized individuals with COVID-19 documented by a positive nucleic acid test for SARS-CoV-2 virus. The median IgG concentration was 5.2 µg/ml for the non-hospitalized COVID-19+ group and 98.5 µg/ml for the ICU hospitalized COVID-19+ group. The SARS-CoV-2 RBD IgG ELISA seropositive threshold is marked by the red line at 0.39µg/ml. Two hundred and forty-four seronegative samples with an IgG concentration below 0.001 were plotted at 0.001. Comparing seropositive groups * p<0.0001 by Wilcoxon-Mann-Whitney Test. Both seronegative groups are significantly different than all seropositive groups.
Similar levels of SARS-CoV-2 RBD IgG antibodies in essential and non-essential workers
Between March 21 and May 30, 2020 the state of Illinois and the city of Chicago were under a shelter-in-place order, except for essential workers and essential trips. Between June and August 2020, these orders were gradually relaxed but limits on gatherings of more than 50 people and restrictions on indoor dining/bars remained in effect. We compared those who reported leaving their place of residence and interacting with others at the workplace (essential workers) or not (non-essential workers) who provided DBS testing between June and September 2020. SCAN participants reported whether they were essential workers, which was defined as leaving their place of residence and interacting with others at the workplace during the shelter-in-place interval. Essential (n=681) and non-essential (n=864) groups had similar percent seropositivity at 20.4% and 19.3%, respectively (Figure 2A and 2B). The two groups were well matched by design for age, gender, and race (Table 3). The median RBD IgG level and distribution was not different (p=0.81) between the seropositive essential (n = 139, median 0.65 µg/ml) and seropositive non-essential groups (n = 167, median 0.59 µg/ml) (Figure 2C). Because this survey reflects collection during the summer months, it is possible seropositivity increased after the shelter-in-place order was relaxed.
Fifteen hundred and forty-five unique SCAN community-acquired samples was acquired between June 24 - September 6, 2020 in the Chicagoland area. Participants self-reported essential worker status during the months of March-September; defined as leaving the residence for work and interacting with co-workers / public. A and B) Essential and non-essential reported groups have similar percent seropositivity at 20.4 % and 19.3%, respectively. C) Essential (n=139) and non-essential (n=167) groups have similar distributions of SARS-CoV-2 RBD IgG seropositivity with a median of 0.65 µg/ml and 0.59 µg/ml, respectively. The SARS-CoV-2 RBD IgG ELISA positivity threshold is denoted with the red dotted line at 0.39 µg/ml. Dashed purple lines represent quartiles. Statistics: Two-sample Kolmogorov-Smirnov test p=0.81.
Low agreement between RBD and nucleocapsid serology status
Multiple platforms are used to detect antibodies to SARS-CoV-2. The most common assays measure antibodies to nucleocapsid (N)-protein or receptor binding domain (RBD) of spike (S1) glycoprotein. We assayed 28 samples from individuals who recovered from symptomatic COVID-19 based on reported virus diagnostic testing, and 92 samples not reporting a positive virus diagnostic test result. Agreement between RBD IgG and nucleocapsid classification was low for both the COVID-19-positive (Κ=0.20; 95% CI: 0.00-0.59) and SCAN (Κ=0.21; 95% CI: 0.12-0.32) groups (Figure 3). Specifically, 6 of 28 (21.4%) COVID-19-positive cases had IgG to RBD detected, but no IgG to nucleocapsid was detected. Correspondingly, 21 of 28 (75%) samples had both RBD and nucleocapsid IgG, while one sample was below the limit of detection in both assays. A similar analysis of 92 SCAN samples was conducted; this cohort was selected to skew towards seropositivity based on reported possible exposure or mild symptoms. Of the 92 samples, 65 (70.7%) samples had IgG to RBD. Of these 65 positive samples, 20 samples had IgG to both RBD and nucleocapsid, while the remaining 45 had only IgG to RBD and not to nucleocapsid. In both groups, there were no samples that were RBD negative and nucleocapsid positive. These data illustrate a discordance between nucleocapsid and RBD IgG seropositivity using these two assays, and suggests that RBD IgG assay may have greater sensitivity.
Twenty-eight COVID-19+ viral positive, non-hospitalized samples and 92 SCAN samples with untested/negative COVID-19 status were analyzed for the presence of SARS-CoV-2 RBD IgG or nucleocapsid IgG antibodies. All samples that were nucleocapsid positive were also RBD IgG positive. 6 of 28 (21.5%) known COVID-19+ viral positive samples were nucleocapsid negative but were RBD IgG positive. One of 28 COVID-19+ viral positive samples was seronegative on both platforms. Of the 92 unknown COVID-19 status samples, 20 (21.7%) samples were both nucleocapsid and RBD IgG positive, while 45 (48.9%) samples were nucleocapsid negative and RBD IgG positive. Agreement between RBD IgG and nucleocapsid classification was low for both the known COVID-19 viral positive (Κ=0.20; 95% CI: 0.00-0.59) and unknown COVID-19 status (Κ=0.21; 95% CI: 0.12-0.32) samples. The black diagonal line and gray shaded area represent the simple linear regression of RBD IgG on nucleocapsid ratio and the 95% confidence interval band respectively. The SARS-CoV-2 RBD IgG ELISA positivity threshold is denoted with the red dotted line at 0.39 µg/ml. The SARS-CoV-2 nucleocapsid positivity threshold is denoted with the orange dotted line at ratio 1.4.
Persistence of SARS-CoV2 RBD IgG antibodies over time
We monitored change in RBD IgG concentration in seropositive SCAN participants (n= 286) with serial measurements from June 22 through September 6, 2020. IgG levels in a cross-sectional comparison from different participants at each time point remained similar over 10 weeks of sampling (Figure 4A). We also assessed the longevity of seroprevalence in 24 seropositive SCAN participants who were followed with up repeat samples after 2-5 months from their first detection of having IgG to SARS-CoV-2 (Figure 4B). The median IgG concentration at day 0, the first day of seropositivity, remained fairly constant after 73-166 days (median 0.53 µg/ml before versus 0.58 µg/ml after; p= 0.29), above the 0.39 µg/ml positivity threshold. Eighteen of 24 seropositive samples (75%) remained seropositive over 73-166 days after the first seropositive result. These data show that in the majority of SCAN participants who were seropositive at their first test, IgG antibodies against SARS-CoV-2 RBD remained detectable over 4 months.
A) Two hundred and eighty-six seropositive RBD IgG SCAN sample concentrations plotted as function of calendar week of acquisition. Lowess curve (black dotted line) is steady across 10 weeks of sampling. B and C) Twenty-four seropositive SCAN participants (purple dots) were resampled 73 – 166 days post the first seropositive RBD IgG test. Dotted lines connect the same participant over time. Eighteen of 24 (75%) samples remained seropositive after 3-5 months with 6 samples converting to seronegative (grey dot). Time points were not significantly different p = 0.29.
One individual was observed with a stable seropositive RBD IgG level over a three month period and during the 4th month experienced a significant increase in IgG to RBD. The participant was then queried regarding health status and retrospectively reported having mild symptoms as outlined in Figure 5. The participant sought testing for COVID-19 after experiencing two days of altered taste and smell occurring 25 days after the first non-specific symptom onset. The diagnostic virus test was negative. We hypothesize this seroprevalence pattern reflects re-exposure to SARS-CoV-2. The viral test was 25 days into the course of nonspecific symptoms and may have been conducted beyond the window of viral detection or alternatively, the participant may never have achieved significant viral load.
One individual in Figure 4B was observed to have a marked increase in IgG to RBD 3.5 months after first having IgG measured. In retrospect, mild symptoms were noted for the 25 days before the increase in IgG was detected. After experiencing mildly altered taste and smell, a COVID-19 nasal swab was performed and was negative.
Discussion
RBD antibodies
We surveyed the presence of antibodies to a very specific component of SARS-CoV-2, the receptor binding domain, among individuals in a single metropolitan US center during the summer of 2020, following relaxation of shelter-in-place orders. This RBD of the spike glycoprotein mediates viral entry into cells, the first step in infection. This domain is highly specific to SARS-CoV-2 with only SARS-CoV-1 having homology among coronaviruses. Some antibodies to the RBD can be neutralizing and protect against cellular invasion. Whether the RBD antibodies detected in the SCAN study are neutralizing or protective in nature is unknown, and ongoing longitudinal studies of SCAN participants will help clarify if infection rates differ between seronegative and seropositive status.
Estimating SARS-CoV-2 spread
In our prior study and in this study, we observed 97% of those with documented COVID-19 developed IgG antibodies to RBD (15). Among individuals with documented and symptomatic COVID-19, there was a lower detection rate for the nucleocapsid antigen, using a completely independent detection platform. This differential sensitivity may represent differences in antibody response, testing platform, or both. Additionally, testing for COVID-19 became much more widely available during May and June 2020, yet only 1.2% of individuals in this survey reported a positive COVID-19 test. In the SCAN cohort, the range of RBD IgG overlapped considerably with the RBD IgG range measured in non-hospitalized people with a positive diagnostic test for COVID-19. However, COVID-19 cases that required hospitalization had greater RBD IgG levels than non-hospitalized COVID-19 cases and that observed in the SCAN cohort. The 19.8% seropositivity rate in this study is consistent with this antibody test identifying 10-20 fold greater exposure than noted with virus diagnostic testing, supporting previous predictions (19). We did observe persistent seropositivity in 75% of these community-acquired samples, which may indicate some degree of lasting protection.
Persistence of IgG directed to RBD
During the summer of 2020, local restrictions imposed during spring 2020 were eased. This included allowing retail stores to open, outdoor dining, permitting larger gatherings, and gym reopening. We observed a similar seroprevalence among participants who self-reported as being “essential” during the spring months and those who remained at home. We expect that some of this seropositivity occurred in the early part of June and appeared sustained in 75% of participants over several months. We observed at least one case of a second “boost” of seropositivity in repeat sampling, and this event was associated with mild symptoms. Despite a negative viral test, we hypothesize this pattern reflects re-exposure to SARS-CoV-2. Whether this pattern portends what might occur with re-exposure generally, or even after vaccines become available, will require population-level surveys, and the ongoing second wave may present an opportunity for this evaluation. The SCAN platform, which relies on simple at home monitoring combined with laboratory precision, is positioned to help address this, and other, knowledge gaps.
Data Availability
Written requests for data should be made by qualified researchers trained in human subject confidentiality protocols to the corresponding author.
Contributions
AD, LV, NR, MV, EB, RH, NK processed samples, performed ELISA. AD, LP, RS, DT managed the data, analyzed data, and generated figures. TM, EM, BM, JW, JH, and MI secured IRB approval and collected samples. AD, BM, TM, RD, EM provided critical input in study design and wrote the manuscript. All authors reviewed and approved the final version of the manuscript.
Conflicts of Interest
All authors declare no conflicts of interest.
Acknowledgments
Supported by NSF 2035114, NIH 3UL1TR001422-06S4, and a generous gift from Dr. Andrew Senyei and Noni Senyei. This work was supported by supplements to the NIH National Center for Advancing Translational Sciences grant UL1 TR001422 and UL1 TR002389 (JFH), the generous support of the Dixon Family Foundation (JFH), a supplement to the Northwestern University Cancer Center P30 CA060553 (JFH), and through a generous contribution from the Walder Foundation’s Chicago Coronavirus Assessment Network (JFH). We acknowledge the efforts of Grace Gallo in sample collection. The funding sources had no role in the study design, data collection, analysis, interpretation, or writing of the report.
