key: cord-290796-x9xqqcj6
authors: Stefanelli, P.; Bella, A.; Fedele, G.; Fiore, S.; Pancheri, S.; Benedetti, E.; Fabiani, C.; Leone, P.; Vacca, P.; Schiavoni, I.; Neri, A.; Carannante, A.; Simmaco, M.; Santino, I.; Zuccali, M. G.; Bizzarri, G.; Magnoni, R.; Benetollo, P. P.; Brusaferro, S.; Rezza, G.; Ferro, A.
title: Longevity of seropositivity and neutralizing titers among SARS-CoV-2 infected individuals after 4 months from baseline: a population-based study in the province of Trento
date: 2020-11-13
journal: nan
DOI: 10.1101/2020.11.11.20229062
sha: 
doc_id: 290796
cord_uid: x9xqqcj6

Background. There are conflicting results about the duration of antibodies induced by SARS-CoV-2, but several studies show a rapid decay in a few months after infection. To evaluate antibody decline, we re-evaluated the presence of anti-SARS-CoV-2 antibodies among individuals found seropositive in a first population survey conducted 4 months before. Methods. All individuals above ten years of age resident in 5 municipalities of the Autonomous Province of Trento, northern Italy, who resulted IgG positive for anti-SARS-CoV-2 nucleocapsid (NC) antibodies in a serosurvey conducted on May 2020 were retested after 4 months. Anti-SARS-CoV-2 antibodies were detected using the Abbott SARS-CoV-2 IgG assay (Abbott Diagnostics, USA) detecting anti-NC antibodies. Samples that gave a negative result were re-tested using the same test plus Liaison SARS-CoV-2 IgG assay (DiaSorin, Italy) to assess anti-spike (S) S1/S2 IgG antibodies. Seroprevalence was calculated as the proportion of positive people on the total number of tested. A neutralizing assay was performed on a subgroup of formerly positives sera using fifty-percent tissue culture infective dose (TCID50) as endpoint dilution to produce a cytopathic effect in 50% of inoculated Vero E6 cells culture. In all the analyses a p value < 0.05 were considered statistically significant. Statistical analysis was performed by STATA version 16.1 (STATA Corp., College Station, Texas, USA). Findings. Overall, 1159 out of 1402 initially anti-NC seropositive participants were enrolled in the study. Of them, 480 (41.1%) became seronegative for anti-NC IgG antibodies. When 479 negative sera were tested for anti-S IgG, 373 samples (77.9%) resulted positives. A functional neutralization assay was performed on 106 sera showing high concordance with anti-S antibodies positivity. Interpretation. A decline of anti-NC IgG values was recorded 4 months after the first evaluation. Worth of note, a high proportion of anti-NC seronegative individuals were positive for anti-spike IgG antibodies, which appear to persist longer and to better correlate with neutralization activity.

The presence of neutralizing antibodies induced by natural infection or by an effective vaccine is likely to be predictive of protection. Thus, whether antibody response to the administration of vaccines may confer protection is still undefined, and cases of reinfection by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) have been sporadically reported1.

The duration of the immune response after infection is also under investigation. The duration of protection against infection with common human coronaviruses appears to be rather short 2, 3 , and there are studies showing declines in IgG antibodies against SARS-CoV-2 among both symptomatic and asymptomatic individuals 4, 5 .

Whether memory-B-cell and T-cell responses may still confer protection in individuals experiencing antibody decline to undetectable levels is unknown 6 . In general, controversies exist on the possibility that SARS-CoV-2 infection induces sustained humoral immune responses in convalescent patients following symptomatic COVID-19 [7] [8] [9] .

The type of antibody response may also play a role. Experimental vaccination against SARS-CoV with NC can induce strong antibody responses that were found to be non-neutralizing 10 . While nonneutralizing antibodies might still exert antiviral activity, for example via the Fc-Fc receptor-based effector function, non-neutralizing NC antibodies may lead to enhanced disease for some vaccine candidates in animal models when neutralizing antibodies are absent 10 . Instead, studies conducted on SARS-CoV-2 and other coronaviruses have shown that the spike protein is the main target for neutralizing antibodies [11] [12] [13] .

Thus, assessing the duration of detectable antibody response and changes in the titer of neutralizing antibodies is an important step in order to better understand their dynamics and to predict the duration of protection against SARS-CoV-2 infection.

In order to evaluate the persistence of SARS-CoV-2 antibodies, we repeated a serosurvey in five municipalities of the Autonomous Province (AP) of Trento, Italy, recruiting those individuals who had resulted positive in a large population-based seroprevalence study conducted 4 months before 14 .

Moreover, in a subsample of seropositive participants, the antibody neutralizing titre was also evaluated.

As already reported 14 , the study was conducted in 5 municipalities of the AP of Trento, in the northern of Italy, with the highest incidence of COVID-19 confirmed cases.

The Department of Prevention of the Azienda Provinciale per i Servizi Sanitari (APSS) sent a letter of invitation to participate at a second study to all the citizens who resulted to be positive for anti-SARS-CoV-2 antibodies in the serosurvey conducted 4 months before, between May 5 and 15, 2020.

Blood samples (5 ml) were collected in Serum Separator Tubes (BD Diagnostic Systems, Franklin Lakes, NJ, USA) and centrifuged at room temperature at 1600 rpm for 10 min. Aliquots were transferred to 2ml polypropylene, screw cap cryotubes (Sorfa, Zhejiang, China) and immediately frozen at -20 °C. Frozen sera were then shipped to the Istituto Superiore di Sanità (ISS) as national reference laboratory for COVID-19, in dry ice following biosafety shipment condition. Upon arrival serum samples were immediately stored at -80 °C. 14 

Two commercial CLIA assays, employing either NC or S antigens and designed for high throughput in healthcare settings, were used. In particular, all the serum samples were evaluated by using the Abbott SARS-CoV-2 IgG assay (Abbott Diagnostics, Chicago, IL, USA); sera resulting negative were retested using the DiaSorin Liaison SARS-CoV-2 IgG assay (DiaSorin, Italy). The Abbott Diagnostics anti-NC IgG assay was performed on the Architect i2000SR automated analyser. The analyser automatically calculates SARS-CoV-2 NC IgG antibody concentration expressed as an is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229062 doi: medRxiv preprint index value. According to the manufacturer's instructions, the results were interpreted considering as positive an index of ≥1.4 and as negative an index of <1.4.

The DiaSorin SARS-CoV-2 IgG assay is also a two-step CLIA assay for the detection of IgG antibodies against S1/S2 antigens of SARS-CoV-2. The assay was performed on the LIAISON® XL fully automated chemiluminescence analyzer. The analyser automatically calculates SARS-CoV-2 S1/S2 IgG antibody concentrations expressed as arbitrary units (AU/mL). The assay range is up to 400 AU/mL. According to manufacturer's instructions, values ≥ 15 AU/mL were interpreted as positive, and values ≤ 12 AU/mL as negative; in case of results falling within an equivocal zone in between 12 AU/mL and 15 AU/mL, the test was repeated.

In vitro neutralising activity provides quantitative results as a measure of a functional humoral immune response against SARS-CoV-2. A known amount of SARS-CoV-2 (code 77III, isolated and cultivated at ISS, titre 1x10 5,4 ; GISAID accession ID: EPI_ISL_412973) was incubated with different dilutions of the serum sample to determine the dilution at which cytopathic effect on Vero E6 cells (ATCC® CRL-1586) is observed in 50% of infected wells (MN 50%). The detailed protocol is described below: two-fold serial dilutions of serum samples starting at 1:8 dilution up to 1:512 in cell culture medium EMEM (Sigma) supplemented with 1X pen/strep and 2% fetal bovine serum (FBS;

Corning) were added to 96-well plates. The mixture of virus (100 TCID50) and serum was incubated at 37ºC for 1 hour for a total 100 µl. After this incubation period, a solution of 22,000 cells per well in a total volume of 100 µl was added and incubated at 37ºC for 5 days.

Finally, the neutralization titer was calculated and expressed as the serum dilution capable of reducing the cytopathic effect to 50% (MN 50%). Positive and negative sera samples and cell culture control together with the virus were added in each test.

The IgG values were summarized by the median and interquartile range (IQR). The differences among IgG values between the first and the second survey were evaluated by the Wilcoxon test. The is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229062 doi: medRxiv preprint < 0.20 = "poor", 0.20-0.40 = "fair", 0.40-0.60 = "moderate", 0.60-0.80 = "good", and 0.80-1.00 = "very good").

A multivariable logistic regression model was used to determine the relationship between persistent anti-NC IgG in the second serosurvey (positive versus negative) and a set of explanatory variables.

The following variables that were significantly associated (p< 0.10) at the univariate analysis were included in the multivariable model: gender, age group, geographical area, presence of symptoms, working in contact with the public and household size, IgG positivity group (weak, medium, high) olfactory and gustatory dysfunctions, fever, weakness, cough, dyspnea, arthralgia, diarrhoea, and abdominal pain and vomit. The likelihood ratio test was used to compare different models.

A subset of anti-NC IgG positive samples was tested with the neutralization test. Assuming a positive proportion of 95% and precision of 4%, 106 samples are required with an alpha error of 5%.

In all the analyses a p-value < 0.05 was considered statistically significant. Statistical analysis was performed by the STATA version 16.1 (STATA Corp., College Station, Texas, USA).

Informed consensus for blood collection was obtained from all the participants. The study was approved by the Ethical Committee of the ISS (Prot. PRE BIO CE n.15997, 04.05.2020),

Overall, 1159 individuals of the 1402 individuals who resulted seropositive in the first survey (82.7%) were enrolled in the study. All age groups were well represented. The proportion of those who were retested ranged between 72.6% in the age group 20-29 years and 93.1% in the age group 60-69 years.

Of the 1159 individuals who resulted initially seropositive 480 (41.4%) seroreverted at the second evaluation. As shown in Figure 1 , a statistically significant reduction in the median value was observed in the second survey, from a median of 5.7 (IQR = 3.5) to 1.9 (IQR = 2.8) (p-value < 0.0001 using the non-parametric Wilcoxon signed-rank test). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted November 13, 2020. ;

As shown in Figure 2 , when the participants were stratified in three groups in accordance with their anti-NC IgG level at the baseline [i.e., weak positive (with a value between 1.4 and 3), medium positive (between 3 and 5), and high positive (greater than 5)], the median value of the weakly and moderately positive groups decreased below the assay cut-off after 4 months, while the median of the highly positives remained above the cut-off.

The samples resulting negative for antibodies against NC in the second study were tested to evaluate the presence of antibodies against the S protein. Since for one sample the available amount of serum was not sufficient for the analysis, 479 available serum samples were tested, and 373 of them (77.9%) resulted positive (Figure 3 ).

A subgroup of 106 sera that were positive for anti-NC-IgG at the first testing were tested for anti-NC IgG, anti-S IgG, and their neutralizing activity 4 months after the baseline. Of the 106 sera, 97 (91.5%) showed neutralizing activity (TCID50 ≥ 1/8), and 9 sera (8.5%) had a TCID50 titer <1/8; 57 (53.8%) were anti-NC positive and 93 (87.7%) were anti-S positive. (Figure 4 ). Table 1 High and significative agreement (94,3%) was found between anti-S and TCID50 (k=0.70; p<0.0001) ( Table 1 ). To further confirm the concordance, when IgG levels were considered, a good correlation between anti-S and TICD50 was observed (rho-Spearman: 0.84, p < 0.0001) compared with anti-NC/anti-S (rho-Spearman: 0.61, p < 0.0001) and anti-NC/TICD50 (rho-Spearman: 0.56, p < 0.0001).

The multivariable logistic regression model showed that age group, gender, anti-NC IgG level in the first serosurvey, and cough were factors associated with the persistence of anti-NC seropositivity ( Table 2 ). In particular, the individuals with high anti-NC IgG levels in the first serosurvey had the highest probability to be seropositive after four months (OR=69.2). Age above 70 years and cough, as reported during the first survey, were also strongly associated with persistent anti-NC IgG levels. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229062 doi: medRxiv preprint

Hereby, we report the results of a repeated serosurvey conducted in five municipalities in the AP of Trento, located in northern Italy 14 These results appear to be consistent with those obtained for other human coronaviruses, such as NL63, 229E, OC43, and HKU1, showing a rapid decay of antibodies directed against the nucleocapsid protein 16 . However, other studies showed different results, with high IgG levels after several months [7] [8] [9] 17 . The inconsistency in the results of previous studies could be explained by differences in the study populations (i.e., patients with mild vs moderate or severe disease) or in the use of different methods (i.e., detection of antibodies directed against the nucleocapsid vs whole spike or the receptor binding domain of the spike) 4 . More recently, Seow et al. 18 In a longitudinal study of RT-PCR confirmed COVID-19 cases, the participants showed a wide range of antibody responses, and a decline in antibodies levels and virus neutralization was observed within three months of the onset of symptoms 18 . For those who developed a low neutralizing antibody response (ID50 100-300) the titers could return to baseline over a relatively short period, whereas those who developed a robust neutralizing antibody response maintained titers >1000 despite the . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229062 doi: medRxiv preprint initial decline 18 . The decline of protective antibodies might be explained, to some extent, by the sporadic COVID-19 reinfection that have been reported [19] [20] [21] . In a more recent study, detectable neutralizing antibody responses were detected for several months after infection 17 .

To this regard, animal models show that SARS-CoV-2 infection protect from re-infection for at least some time 22, 23 . This protection appears to be more pronounced in the lower respiratory tract rather than in the upper respiratory tract 23 .

The types of antibody response may also play a role. Atyeo et al 24 , showed that a predominant humoral response to nucleocapsid protein is associated with poor outcome in patients admitted to hospital, compared to response to spike protein.

Most vaccine candidates elicit responses to S rather than NC protein. Measuring antibodies to S will therefore indicate whether there has been a good response. To this regard, our findings are in good agreement with the study by Wajnberg is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229062 doi: medRxiv preprint Funds: APSS sustained the expenses for the IgG assays on collected sera. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229062 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229062 doi: medRxiv preprint Figure 3 Percentage of anti-spike (S1/S2) IgG antibodies on retested anti-NC IgG negative sera. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted November 13, 2020. ; Table 1 . Concordance between IgG against NC and S proteins and neutralization activity.

Anti-S -p-value* is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229062 doi: medRxiv preprint

Is recurrence possible in coronavirus disease 2019 (COVID-19)? Case series and systematic review of literature

T cell-mediated immune response to respiratory coronaviruses

Duration of Antibody Responses after Severe Acute Respiratory Syndrome. Emerg Infect Dis

Rapid decay of anti-SARS-CoV-2 antibodies in persons with mild COVID-19

Clinical and immunological assessment of asymptomatic SARS-CoV-2 infections

Targets of T-cell responses to SARS-CoV-2 coronavirus in humans with COVID-19 disease and unexposed individuals

SARS-CoV-2 infection induces sustained humoral immune responses in convalescent patients following symptomatic COVID-19

Loss of Anti-SARS-CoV-2 Antibodies in Mild Covid-19

Loss of Anti-SARS-CoV-2 Antibodies in Mild Covid-19. Reply

Vaccine efficacy in senescent mice challenged with recombinant SARS-CoV bearing epidemic and zoonotic spike variants

SARS-CoV-2 Vaccines: Status Report

The receptor binding domain of the viral spike protein is an immunodominant and highly specific target of antibodies in SARS-CoV-2 patients

A human monoclonal antibody blocking SARS-CoV-2 infection

Prevalence of SARS-CoV-2 IgG antibodies in an area of North-eastern Italy with a high incidence of COVID-19 cases: a population-based study

Immunological considerations for COVID-19 vaccine strategies

Seasonal coronavirus protective immunity is short-lasting

Robust neutralizing antibodies to SARS-CoV-2 infection persist for months

Longitudinal evaluation and decline of antibody responses in SARS-CoV-2 infection

Serum antibody profile of a patient with COVID-19 reinfection

Asymptomatic reinfection in two healthcare workers from India with genetically distinct SARS-CoV-2

Genomic evidence for reinfection with SARS-CoV-2: a case study

SARS-CoV-2 infection protects against rechallenge in rhesus macaques

Syrian hamsters as a small animal model for SARS-CoV-2 infection and countermeasure development

Distinct early serological signatures track with SARS-CoV-2 survival

The authors would like to thank the study's participants. 

The authors declare no conflict of interest related to this study.