Introduction

As SARS-CoV-2 infection has spread worldwide, cross-sectional serological surveys have provided initial data about exposure to the SARS-CoV-2 virus at a particular epidemic time, place, and the population studied.^1,2 Currently, follow-up studies that focused on profiling both the cellular and humoral response are emerging with new evidence. Since SARS-CoV-1 and MERS-CoV share a genetic and clinical background with SARS-CoV-2, it may be possible to expect a similar immune response in terms of seroconversion, timing, and duration.³ SARS-CoV-2 interacts with target cells by binding the ACE2 receptor. Viral entry in the lung cells results in the typical Covid-19 associated symptoms of fever, nasal discharge, cough, and sore throat. Since the ACE2 receptor is expressed in several tissues, patients may experience additional gastrointestinal and cardiovascular symptoms.⁴ Cytokine dysregulation, coagulopathy, ischemic complications, and multiorgan dysfunction comprise a broad Covid-19 spectrum of clinical manifestations. Exposure to SARS-CoV-2 leads to an immunologic response, including antibody production against viral proteins, while antibodies’ duration and protection capacity remain unknown. Meanwhile, vaccines development (viral vector-based and mRNA-based) represented a critical effort during the last year, and vaccination is ongoing worldwide, including in our country. However, viral variants of clinical concern arising from natural selection might compromise current vaccine efficacy, and additional redesign and adjustment may be required.^5,6 To date, there is a lack of evidence about the efficacy and long-term duration of the humoral response in the natural SARS-CoV-2 infection. This information will be of prime interest in defining epidemic dynamics, maintain current preventive strategies, and guide vaccination programs.⁷

Several studies have highlighted a persistent and robust humoral response in most people infected with SARS-CoV-2,^{8,9,10,11,12,13} while others have reported a decline in humoral immunity that may not be long-lasting.^{14,15,16,17,18,19} These controversial and inconclusive results are based on studies with small sample sizes, variable follow-up times, and different immunoassays. Furthermore, they included specific populations, such as health-care workers with a wide range of disease severity, symptoms, and underlying conditions.

Accordingly, in this study, we aimed to determine the persistence of IgM/IgG antibodies against SARS-CoV-2 among 193 previously detected seropositive subjects in a community-based serological survey, at least 112 days from infection.

Materials and methods

The Institutional Ethics Committee approved the study protocol and written informed consent was obtained from all participants who attended. In this follow-up study, the 193 government employees previously identified with positive anti-SARS-CoV-2 antibodies were invited to a second test around 112 days after the first sample.

Briefly, positive anti-SARS-CoV-2 cases (n= 193) were identified in a community level serological survey performed with 3 268 participants in Guadalupe, Nuevo Leon, Mexico, during July 2020.²⁰ This serological follow-up occurred during a Phase 3 epidemic (World Health Organization, WHO). According to local health authorities, at the time of the follow up tests (October 31, 2020), Nuevo Leon and Guadalupe reported 82 021 and 10 987 PCR confirmed cases, respectively. Of note, participants did not report current symptoms during the baseline and follow-up antibody tests. All answered a standardized electronic questionnaire, and follow-up samples were collected between October 29-31, 2020.

Results

Follow up assessment of IgM/IgG antibodies

A total of 174 serum samples were tested in the follow-up. Among these, 171 (98.3%) were still positive, and only 3 (1.7%) were negative to IgM/IgG antibodies (table II). Also, the COI value (electrochemiluminescence signal) of the immunoassay significantly increased from the first to the second test (P<0.001) (figure 1). The COI value was not different when analyzed by gender (P= 0.52), age >45 years (P= 0.17), obesity (P= 0.07), or self-reported persisted symptoms (P= 0.56) (table III). Non-reactive subjects to the follow-up test reported no known contact with a Covid-19 confirmed case, and only one had mild and unspecific symptoms. They were young adults (<45 years), not obese, and had a medium risk of infection by occupational exposure (supplementary table B).²³

Table II Follow up assessment of anti-SARS-CoV-2 antibodies. Nuevo León, México, 2020 

	n (%)	COI (RLU)*
Positive baseline test	174 (100)	23.3 (9.1-55.5)
Positive follow up test	171 (98.3)	100 (63.9-133)‡
Negative follow up test	3 (1.7)	0.651 ± 0.46

^*Median (IQR) of electrochemiluminescence signal; COI: Cut off index; RLU: Relative luminescence units

^‡Wilcoxon signed rank test P<0.001, baseline vs. follow up positive COI values

COI: Cut off index; RLU: Relative luminescence units

Figure 1 A) Quantitative luminescence values at first and second IgM/IgG tests. COI is expressed in RLU. Wilcoxon signed rank test p < 0.001. B) Individual changes from baseline to follow-up in antibody levels. Nuevo León, México, 2020 

Table III COI value by demographic and clinical characteristics. Nuevo León, México, 2020 

Characteristic	COI (RLU)	p-value*
Sex
Males	100.5 (59.6-133)	0.52
Females	86.2 (55.7-134)
Age (years)
< 45	95.7 (46.2-128.5)	0.17
> 45	108 (76.7-137.5)
Covid-19 symptoms
Asymptomatic	99.8 (48.4-135)	0.56
Symptomatic	99.4 (76-132.5)
Comorbidities
Non-obese	95.7 (42.7-128)	0.07
Obese	102 (77.2-141)

^*Mann-Whitney test

COI: Cut off index; RLU: Relative luminescence units. RLU are expressed as median (IQR)

Discussion

We found that about 98.3% of the subjects with SARS-CoV-2 infection still presented IgM/IgG antibodies 112 days after a positive baseline test. A SARS-CoV-2 infection could happen weeks before initial sampling. Moreover, we observed a statistically significant increase in the quantitative luminescence signal between the baseline and the follow-up test, reflecting increasing antibody titers. Therefore, there was a persistent, robust, and sustained humoral response up to 112 days after the first antibody detection in subjects with mild and asymptomatic SARS-CoV-2 infection. Our results are the first serology follow-up study reported in Mexico.

Since SARS-CoV-2 was first described a year ago, there is limited evidence about the elicited immunological response, and ongoing studies focus on defining the long-term duration of humoral immunity after symptomatic and asymptomatic natural infection. Ibarrondo and colleagues described a rapid decay of anti-SARS-CoV-2 antibodies in mild Covid-19 at 37 and 86 days in 34 subjects.¹⁴ These findings may support a concern about a limited duration of the humoral immunity in people with mild illness. Similarly, Liang and colleagues evaluated 76 healthcare workers in a prospective observational follow-up study of Covid-19 survivors from Wuhan (28% males, 41.3 ± 13.8 years). They found that IgM and IgG were negative in 87 and 13% three-months after discharge.¹⁵ Additional evidence in asymptomatic subjects supporting the antibody decline is provided by Milani and colleagues. In this study, about 80% of asymptomatic subjects did not present immunoglobulins against SARS-CoV-2 after eight weeks after a positive PCR test.¹⁶ However, only 31 subjects were evaluated after the initial antibody test during a relatively short follow-up period, and only antibodies against the spike protein were assayed. Also, the decline of anti-SARS-CoV-2 antibodies at 60 days among healthcare workers, who may be at a high exposure risk, was reported in three individual studies.^17,18,24 IgG levels in 93.3% (28/30) of the asymptomatic group declined during the convalescent phase, eight weeks after a positive PCR test and discharge from hospital isolation, according to the study by Long and colleagues. The median percentage decrease was 71.1% (32.8-88.8) for IgG levels in asymptomatic subjects.²⁵

Our results contrast with the reports mentioned above. Conversely, our findings showed a positive IgM/IgG test in 171 of 174 subjects evaluated at the 112-day follow-up visit. The discrepancy in the persistence of antibodies could be explained by the different methods of serological assays (enzyme-linked immunosorbent assays, immunochromatography, or Luminex) and by a difference in the antigenic targets. Furthermore, in our study, the quantitative immunoassay result (COI) was significantly higher compared to the baseline signal. The study population mainly consisted of asymptomatic patients who were not eligible for a PCR test under the Covid-19 study protocol in Mexico. Also, we performed a qualitative and highly sensitive luminescent assay available as a routine test in our daily practice. Since we identified seropositive subjects using an IgM/IgG baseline test, it may be likely that patients had been exposed to SARS-CoV-2 at least several weeks before the first screening. These data suggest that cases had sustained antibody-mediated immunity. Given the conflicting results observed, additional prospective studies with longer follow-up periods in particular populations are needed.

The relationship between Covid-19 severity and the resulting antibody titers is currently unclear. Previous observations indicate that severe Covid-19 may result in a strong humoral response. Wu and colleagues suggested that the humoral immune response to SARS-CoV-2 in symptomatic Covid-19 patients is rather prototypical for viruses having an early expansion phase followed by an intermediate contraction phase and a sustained memory phase, similar to other common cold viruses.²⁶ We speculate that the persisting and increasing antibody levels determined in our study population may be due to the early expansion phase previously described. However, further assessment and characterization of the persisting antibodies will be needed to define a long-lasting humoral response.

In agreement with our findings, Malfertheiner and colleagues detected an IgG response in 77.8% of healthcare workers up to 12 weeks using the Elecsys Anti-SARS-CoV-2 test kit.¹⁰ Interestingly, 22% of Covid-19 patients did not show humoral evidence of infection within 60 days. Marklund and colleagues showed that patients with severe Covid-19 seroconvert earlier and develop higher IgG titers than patients with mild illness.¹¹ Also, another study at Mount Sinai Health System recalled 121 plasma donors at various baseline titer levels and found IgG evidence against the viral spike protein at 82 days post-symptom onset.¹³ Between days 40 and 199, 90% seropositivity in previous SARS-CoV-2-PCR positive subjects was reported by Figueiredo-Campos and colleagues in the longest follow-up study currently available.⁹ Iyer and colleagues reported little to no decrease in IgG antibodies over 75 days since symptom onset in 343 North American patients (93% hospitalized) infected with SARS-CoV-2.⁸ All authors agreed that not all patients develop traceable levels as we did in exceptional cases.

Some limitations should be taken into account. Antibodies were determined around 100 days after the baseline IgM/IgG test, and longer follow-up times may be evaluated. We used an immunoassay exclusively designed to detect antibodies against the nucleocapsid virus protein located in the viral core that leads to the production of non-neutralizing antibodies.^13,27 However, recent observations by Batra and colleagues suggest that antibodies against N protein could be a useful prognostic factor for the clinical course of the disease.²⁸ As for strengths, we followed a large sample size of cases with SARS-CoV-2 infection. All of them did not refer current symptoms during the baseline test. Also, we had a high follow-up rate (90%) at 112 days, and a longer follow-up of these patients would be likely as they are an institutionalized population.

In conclusion, we determined the antibody persistence up to 112 days in previously seropositive cases of SARS-CoV-2 infection. Additional research is needed to profile the complete immunological response in the long term because of the future implications in the serological survey data and immunity-based strategies.