Abstract
Background There is considerable debate about the rate of antibody waning after SARS-CoV-2 infection, raising questions around long-term immunity following both natural infection and vaccination. We undertook prospective serosurveillance in a large cohort of healthy adults from the start of the epidemic in England.
Methods The serosurveillance cohort included office and laboratory-based staff and healthcare workers in 4 sites in England, who were tested monthly for SARS-CoV-2 spike protein and nucleoprotein IgG between 23rd March and 20th August 2020. Antibody levels from 21 days after a positive test were modelled using mixed effects regression models.
Findings In total, 2247 individuals were recruited and 2014 (90%) had 3-5 monthly antibody tests. Overall, 272 (12.1%) of individuals had at least one positive/equivocal spike protein IgG result, with the highest proportion in a hospital site (22%), 14% in London and 2.1% in a rural area. Results were similar for nucleoprotein IgG. Following a positive result, 39/587 (6.6%) tested negative for nucleoprotein IgG and 52/515 (10.1%) for spike protein IgG. Nucleoprotein IgG declined by 6.4% per week (95% CI, 5.5-7.4%; half-life, 75 [95% CI, 66-89] days) and spike protein IgG by 5.8% (95% CI, 5.1-6.6%; half-life, 83 [95% CI, 73-96] days).
Conclusions Over the study period SARS-CoV-2 seropositivity was 8-10% overall and up to 21% in clinical healthcare workers. In seropositive individuals, nucleoprotein and spike protein IgG antibodies declined with time after infection and 50% are predicted to fall below the positive test threshold after 6 months.
Funding PHE
Introduction
SARS-CoV-2, the virus responsible for coronavirus disease 2019 (COVID-19), first emerged in Wuhan, China, in December 2019 and spread rapidly across the continents, leading the World Health Organization to declare a global pandemic on 11 March 2020.(1) SARS-CoV-2 infection may be asymptomatic,(2) or manifest along a wide clinical spectrum, from mild upper respiratory tract illness to severe pneumonia, multiorgan failure and death.(3,4) Risk factors for COVID-19 include age, male gender, ethnicity, underlying comorbidity and occupation, especially being a healthcare worker. (5)
SARS-CoV-2 infection is usually confirmed by identifying viral RNA on RT-PCR of a nasal, nasopharyngeal or throat swab, but the sensitivity of this test may be as low as 70%,1 which will significantly underestimate the true extent of infection in a population.(6) Antibody tests potentially provide a more accurate assessment of SARS-CoV-2 exposure and are a useful tool for understanding transmission dynamics and pandemic progression.(7,8) Antibodies against the spike protein of SARS-CoV-2 demonstrate neutralising activity in vitro and may protect against re-infection, possibly allowing assessment of both individual and population-level immunity.(9,10) In the absence of an effective vaccine, COVID-19 control is likely to rely, at least in part, on herd immunity arising from a proportion of the population being exposed to the virus and developing a functional antibody response.(11) There are, however, concerns that SARS-CoV-2 antibodies may be short-lived.(12–14) If confirmed, antibody waning raises concerns about future protection from re-infection.(15)
In England, the first imported cases of COVID-19 were identified in late January 2020 and started increasing rapidly from early March, resulting in implementation of national lockdown on 23rd March, which included closures of schools and non-essential businesses.(16) Cases continued to increase until mid-April before plateauing and then declining to low levels by end May 2020, after which lockdown measures since eased gradually. Studies of adult blood donors in particular have estimated a seroprevalence of 5-10% across England, and 15% in London during June 2020. (6,17,18)
In order to monitor seroprevalence and the course of antibodies over the course of the pandemic, Public Health England (PHE) initiated a monthly seroprevalence study in March 2020 across three English regions. Participants included healthcare workers with direct patient contact, those with public-facing but non-clinical roles, and non-clinical office and lab workers, who all continued working throughout the lockdown period. Repeated monthly testing allowed measurement of incidence as well as antibody seroconversion and seroreversion rates as the pandemic unfolded, thus providing an ideal opportunity to study SARS-CoV-2 antibody waning in a large cohort of adults with asymptomatic infection and mild-to-moderate disease.
Methods
Study design
ESCAPE (Enhanced SeroIncidence for COVID-19 Antibodies among PHE and NHS Staff) is a prospective surveillance cohort that collects blood samples for SARS-CoV-2 antibodies and questionnaire data at monthly intervals. Surveillance was initiated in March 2020 and included staff from Public Health England (PHE), National Health Service (NHS) and volunteers across four sites. Sites were (1) London (PHE London and Colindale, primarily office and laboratory staff), (2) PHE Porton Down (Southwest England, primarily office and laboratory staff in rural Wiltshire plus family and friends), (3) PHE Manchester and Manchester Royal Infirmary (MRI) (office and laboratory staff as well as healthcare workers) and (4) NHS Wythenshawe Hospital, Manchester (primarily frontline healthcare workers). After providing written consent, participants completed a brief questionnaire and provided ∼10 mls of blood by venepuncture. Samples were processed in PHE laboratories and frozen at -70°C or below until they were batch-tested for SARS-CoV-2 antibodies. Participants who missed an appointment continued to be invited for subsequent appointments, as did participants who became unwell with a respiratory or any other illness.
Laboratory methods
Samples were tested using two assays using the sample methodology for each assay and in the same laboratory: the Abbott chemiluminescence enzyme immunoassays for qualitative detection of SARS-CoV-2 nucleoprotein (Np) IgG antibodies (Abbott Diagnostics, IL, USA) and EUROIMMUN enzyme-linked immunosorbent assay (ELISA) for semi-quantitative detection of IgG antibodies using recombinant S1 domain of SARS-CoV-2 spike protein (EUROIMMUN Medizinische Labordiagnostika AG, Lubeck, Germany). Assay sensitivity and specificity were assessed by PHE using panels of convalescent and negative sera.(19) The ranges of detection for the assays were negative (<0.8), equivocal (0.8-1.4) or positive (>1.4) for nucleocapsid IgG and negative (<0.8), equivocal (0.8-1.1) and positive (>1.1) for spike protein IgG
Statistical analysis
Data for nucleoprotein IgG and spike protein IgG were analysed separately. For nucleoprotein IgG the threshold for positivity was 0.8, while for spike protein IgG, we defined a positive test as being above the threshold of 1.1, with results in the 0.8-1.1 classed as equivocal. The proportion positive at each visit for the four sites was calculated with 95% binomial confidence intervals (CI). Seroconversion was defined as having a positive test subsequent to a negative test at the previous visit. Sero-reversion was defined as having a negative test subsequent to a positive test at the previous visit. The proportion of seroconversions and reversions was calculated with 95% CIs across study visits. Differences in positivity according to site, calendar month, age (18-29, 30-39, 40-49, 50-59, ≥60 years) and sex were analysed using logistic regression. Interactions between covariates were assessed via likelihood ratio (LR) tests, and confidence intervals based on cluster robust standard errors to account for the repeated measurements in individuals.
To assess trends in antibody responses following probable infection, we included individuals with at least one positive or equivocal test. We classified individuals according to whether they were negative at their first study visit and had a positive/equivocal test subsequently, or were positive/equivocal from their first study visit. For the former group, any seroconversion is likely to have taken place between visits, while for those positive/equivocal, the timing of seroconversion is unknown but based on the epidemiology of COVID-19 in England should still be recent.(6) We further classified positive/equivocal individuals according to whether their first non-negative test was positive or equivocal. For spike protein IgG the standard range for equivocal tests described above was used; for nucleoprotein IgG we classified those in the range 0.8-1.4 as “equivocal”. Spike protein IgG results above 15 (2.71 on the log scale) were censored, this being the maximum value for the test. The mean log antibody and 95% CI was then calculated for each group at each visit.
To account for potential short-term changes in antibody levels following infection, we examined test results starting from the subjects’ next test, including only tests that were performed at least 21 days after the initial non-negative test. A linear mixed effects model was fitted to the log of the quantitative results, with time as a continuous variable. Random effects representing individual variability in response were included for the intercept (baseline log result) and slopes (change over time). Correlation between the random effect for the intercept and the slope was allowed for in the variance-covariance matrix.
Based on the fitted model, we calculated antibody half-life as log(0.5)/r, where r is the estimated change in antibody levels per day. To quantify individual variability in trends, we compared the estimated weekly change at -1 and +1 standard deviations from the mean, based on the estimated variance of the random effect for weekly change. Individual-level trends were extrapolated to predict future antibody levels, and the predicted time until negativity for those with declining antibodies. The predictions were then used to determine the proportion that would revert to a negative result 3, 6 and 12 months after their first positive test, and the time at which 50% of individuals would revert to a negative test. Data were analysed using Stata v14.2 (StataCorp, College Station, Texas).
Ethics Approval
This study was approved by PHE Research Support and Governance Office (R&D REGG Ref NR 0190).
Results
Seroprevalence
In total, 2247 individuals were recruited (London, 537; Manchester, 594; Wythenshawe, 598 and Porton, 518) and 2014 individuals (90%) had at least 3 visits with up to 5 visits between 23rd March 2020 and 20th August 2020. The median interval between study visits was 28 days, with 90% of visits occurring between 23 and 41 days. Of the scheduled intervening visits between the first and last visit, 95.8% of visits were attended. In total, 7093 nucleoprotein IgG tests were conducted, of which 817 (11.5%) were positive, and 8952 spike protein IgG tests were conducted, of which 712 (8.0%) were positive. Nucleoprotein IgG tests were not conducted for Porton participants, who had the lowest seroprevalence. Excluding the Porton site increased the seropositivity rate for spike protein IgG to 10.3% (683/6637), which was more comparable to nucleoprotein IgG results.
Positivity by either assay was highest among Wythenshawe hospital employees (18-20%) followed by PHE London (11-12%), Manchester (6-8%) and PHE Porton (1-2%) staff (Tables 1 & 2, Figure 1). Proportions were slightly higher for nucleoprotein than for the spike protein IgG. Seropositivity rates were lowest in March and increased in April at all sites (Figure 1). This is also reflected in the seroconversion rates (negative to positive) where paired results were available, with seroconversion rates between first and second visits ranging from 0.8% at PHE Porton to 8% at Wythenshawe hospital and PHE London (Tables 1 & 2). Seroconversion rates between later visits were lower, with all areas being <1% from July onwards for both assays.
Logistic regression showed the odds ratios for SARS-CoV-2 antibody positivity were nearly double for Wythenshawe hospital compared to London, around a half of London in Manchester and substantially lower for Porton. The odds of seropositivity were a quarter or less in March compared to May (base month), increased in April, and were not significantly different for June, July and August: odds ratios for July and August compared to May were <1 but not statistically significant. Females had lower odds of seropositivity than males for SARS-CoV-2 spike protein, but this was not significant for nucleoprotein IgG. For both assays, 18-29 year-olds had around 50% higher odds of antibody positivity than 30-39 year-olds. There were no differences for other ages, except for a lower odds in ≥60 year-olds; the number of tests in the older age-group were smaller and with wide confidence intervals, but the results were strongly significant.
Assessment of potential interactions for spike protein IgG revealed a significant interaction between age and site (p<0.001) but not with other variables (minimum p-value 0.097, for age and sex). For nucleoprotein IgG, there was a borderline significant interaction between age and site (p=0.032). The minimum p-value for other interactions was p=0.326, between site and sex. For spike protein IgG, the increased odds of positivity for 18-29 year-olds was greater in London and Porton compared to Manchester and Wythenshawe hospital, but there was no obvious pattern for nucleoprotein IgG.
Antibody waning
Overall, 266/2179 (12.2%) and 272/2247 (12.1%) individuals had at least one positive/equivocal nucleoprotein IgG and spike protein IgG result, respectively. Following a positive result at one study visit, 39/587 (6.6%) tests were subsequently negative for nucleoprotein IgG and 52/515 (10.1%) for spike protein IgG. Sero-reversion was more common in later months (chi-squared p-values: p<0.001 for nucleoprotein, p=0.012 for spike protein). Table 4 shows bivariate results following the first non-negative test in 308 individuals tested with both assays. At the first positive test, 200 (65%) were positive with both assays, 61 (19.8%) tested positive for nucleoprotein IgG but not spike protein IgG, and 47 (15.3%) tested positive for spike protein IgG but not nucleoprotein IgG, potentially indicating earlier detection of seroconversion with nucleoprotein IgG. Proportions testing positive declined at subsequent visits for both assays.
Figure 2 shows trends in log antibody levels for the two assays. IgG antibodies against both nucleoprotein and spike protein declined following the first non-negative test, although the pattern was less clear for individuals whose first test result was equivocal. After 8-16 weeks, mean IgG antibody levels in individuals whose first test was in the equivocal range were still lower than those with a clear positive at their first non-negative test, possibly because some were infected earlier in the pandemic and their antibodies were already declining.
Random effects models were used to estimate trends from 21 days after the first positive test (Figure 2). The estimated mean change for nucleoprotein IgG was -0.067 log units per week (95% CI: -0.077 to -0.056) representing a 6.4% decline per week (95% CI: 5.5-7.4%), or a half-life of 75 days (95% CI: 66-89 days). There was substantial between-individual variability in trends, with weekly changing ranging from a 0.3% increase to a 12.7% decline per week within 1 standard deviation from the mean. For spike protein IgG, the mean estimated change was -0.060 log units per week (95% CI: -0.068 to -0.052) representing a 5.8% decline per week (95% CI: 5.1-6.6%), or a half-life of 83 days (95% CI: 73-96 days). Here, too, there was substantial between-individual variability in trends, with weekly decay ranging from 2.2% to 9.3% per week within 1 standard deviation from the mean.
The random effects for baseline and trends showed significant negative correlation, indicating a faster decline in those with a higher baseline result (Supplementary Table S1). For nucleoprotein IgG, there was less of a decline in those whose first non-negative test was in the equivocal range (0.8-1.4) with a 4% decline per week in those who were non-negative at the start of the surveillance and 1.3% decline per week in those initially negative and then seroconverted during the surveillance period, although confidence intervals were wide. There was little difference in trends by baseline status for spike protein IgG. For both assays there was no clear pattern to differences in baseline status and trend by age and sex (Supplementary table S1). For spike protein IgG those age 40-49 had somewhat lower baseline results (p=0.052), but a slower decline (p=0.002), compared to those age 30-39, but little difference for those age 18-29 or 50+ (minimum p-value=0.143). There were no obvious differences according to age for nucleoprotein IgG (minimum p-value=0.156), but possibly a slightly faster decline in females (p=0.05).
The full multivariable model was used to predict time to negativity assuming continuation of the estimated individual-specific linear trends (i.e. accounting for both the fixed covariates and random effects). The proportions predicted to revert to a negative test at 3, 6 and 12 months were 9%, 51% and 95% for spike protein IgG, and 20%, 43% and 77% for nucleoprotein IgG. The predicted time at which half of the cohort would be negative was 26 weeks for spike protein IgG and 29 weeks for nucleoprotein IgG (Supplementary Figure S2). Given the variability in individual trends, the proportion of individuals remaining non-negative levels off over time because some individuals are estimated to have a very slow decline, or none at all. However, this proportion was low, particularly for spike protein IgG. Although both assays showed a similar average decline, the greater variation in response for nucleoprotein IgG meant that antibody levels in some individuals are predicted to decline very slowly, whereas the trend is more consistent for spike protein IgG.
Discussion
We analysed data from a monthly prospective longitudinal cohort study of 2247 PHE and NHS employees in three English regions, with up to five sampling time-points from the start of the COVID-19 epidemic in March 2020 until mid-August 2020. Seroprevalence by region and time was consistent with other sources, with a higher prevalence in London, a rise in prevalence from March to May 2020 and a slight decrease in seropositivity from June to August.(6) Wythenshawe hospital had nearly double the seroprevalence of PHE London staff due to most participants being healthcare workers with direct patient contact. (5) Conversely, seroprevalence among PHE Porton office and laboratory staff in rural southwest England was <2%. Antibodies against nucleoprotein were detected earlier than spike protein in adults exposed to SARS-CoV-2. Seroconversion rates for nucleoprotein and spike protein were highest between the first and second blood sampling visit and then stabilised. We observed seroreversion to antibody negativity in both assays, with measured antibody levels declining at an estimated weekly rate of 6.4% for nucleoprotein IgG and 5.8% for spike protein IgG. The predicted time at which half the cohort would be negative 29 weeks for nucleoprotein IgG and 26 weeks for spike protein IgG.
The strength of this surveillance was the early recruitment of large numbers of seronegative participants at the start of the epidemic in England. We recruited both clinical and non-clinical staff in three English regions with varying community prevalence, most of whom continued to work during national lockdown. The monthly follow-up with blood sampling allowed assessment of seroprevalence throughout the course of the first peak of the epidemic and the course of antibodies against multiple viral antigens in a large population of healthy adults who were exposed to SARS-CoV-2, most of whom were asymptomatic or only developed mild-to-moderate disease.
Following infection with other seasonal coronaviruses, anti-nucleocapsid antibodies decline within two months and reinfections are common after 12 months. (20) The COVID-19 pandemic is new and longitudinal antibody studies so far have primarily included small numbers of mainly hospitalised patients with severe COVID-19 with a limited follow-up period.(13,21–24) One such study suggested more rapid waning of antibodies in asymptomatic individuals compared to symptomatic patients, with 12 of 30 (40.0%) asymptomatic individuals and 4 of 31 (12.9%) symptomatic individuals becoming SARS-CoV-2 IgG negative within 8 weeks of hospital discharge, which was also associated with a decline in neutralising antibodies over the same period.(25) Another report involving 34 patients followed-up for a mean of 86 days after infection estimated a SARS-CoV-2 antibody half-life of 36 days (95% CI: 26-60 days). (26) A short duration of immunity would suggest that SARS-CoV-2 would enter into regular circulation alongside other coronavirus, with seasonal variation and an annual, biennial or sporadic pattern depending on the length of protection afforded by the initial infection.(15) Our findings, however, which includes more than 2000 healthy individuals with asymptomatic infection or mild-to-moderate disease and, therefore, more representative of the general population, suggest that IgG antibodies against the SARS-CoV-2 spike protein and nucleoprotein do not decline as quickly as earlier estimates in smaller cohorts of patients with shorter follow-up. (13,21–24) Our findings are, however, consistent with the recent Icelandic study which found most hospitalised patients with COVID-19 remained seropositive 120 days after diagnosis with no significant decrease in antibody levels using two different antibody assays.(27)
The humoral response is considered to provide the first line of defence against infection and, therefore, the presence, neutralising ability and persistence of antibodies is likely to correlate with protection against infection,(28) which is consistent with our understanding of the host immune response to respiratory viruses. Similarly, SARS-CoV-2 infections also trigger a cellular immune response, with activation of a range of T cells, including helper (CD4) and cytotoxic (CD8) T cell responses against all major SARS-CoV-2 antigens, which is independent to B-cell mediated development, functionality or duration of SARS-CoV-2 antibodies.(29,30) The demonstration of strong T cell memory responses in patients recovering from COVID-19 could potentially indicate long-term protection against SARS-CoV-2 even in the absence of detectable antibodies.(31)
Restriction of analysis to individuals with two positive tests at least 21 days apart was used to ensure that the tail in any initial short-term antibody response post-infection would not be interpreted as a general decline. If our observed trends were to continue at the same rate, then half of those seropositive would be predicted to fall below the positive test threshold at 6-7 months for both spike protein and nucleoprotein antibodies. Given the estimated variability in the rate of decline, a proportion of individuals would remain antibody positive for at least one year, although this could be very low. However, care must be taken in extrapolating these results to long-term trends, as it is possible that the rate of decline may slow with some individual remaining above thresholds in the long term.
Our study is, so far, the longest prospective serosurveillance of a large cohort of healthy adults with asymptomatic infection or mild-to-moderate disease. Despite little evidence of non-linearity in those followed up so far, trends in log antibody levels may stabilise, or the decline may plateau over time. For both MERS-CoV and SARS-CoV, antibodies were shown to last for at least two years after infection.(32,33) The duration of immunity following SARS-CoV-2 infection is a critical factor in determining the course of the pandemic and whether the virus will be eradicated or enter the general circulation and cause seasonal outbreaks in the coming years.(15)
Conclusions
In adults exposed to SARS-CoV-2, nucleoprotein IgG antibodies were detected earlier than spike protein IgG antibodies. Both antibodies declined with time after infection and half of those with evidence of infection are predicted to revert to a negative test after 6 months. Ongoing serosurveillance is critical for monitoring the course and projection of SARS-CoV-2 antibodies in the longer term. Further studies are needed to assess the role of cellular immunity in long-term protection against SARS-CoV-2.
Data Availability
Sharing of individual-level data is not permitted under the study protocol, although provision of aggregated data for specific purposes will be considered on request.
Author contributions
SL was the chief investigator, had full access to all the data in the study and had final responsibility for the decision to submit for publication. SL, MR, BH, SA, RB and NA conceived and designed the study. EL, TB and KB oversaw data collection and quality assurance. RH and HW analysed the data and carried out data checking. RH, HW and SL drafted the manuscript. All authors contributed to interpretation of the data and critical revision of the final manuscript.
Supplementary material S1: Full results of multivariable mixed effect models
Supplementary material S2: Predicted time to negativity
Acknowledgements
BP is supported by the NIHR Manchester Biomedical Research Centre.
This research was internally funded by Public Health England and carried out at by the NIHR Manchester Clinical Research Facility. The views expressed are those of the authors and not necessarily those of the (funder name), NHS, the NIHR or the Department of Health..
We would like to express our gratitude to all the participants and to the laboratory staff for testing the samples.