Prevalence of IgG Antibodies to SARS-CoV-2 in Wuhan – Implications for the Longevity of Antibodies Against SARS-CoV-2

Background It is to be determined whether people infected with SARS-CoV-2 will develop long-term immunity against SARS-CoV-2 and retain long-lasting antibodies after the infection is resolved. This study was to explore the outcomes of IgG antibodies to SARS-CoV-2 in four groups of individuals in Wuhan, China. Methods We conducted a cross-sectional study on the following four groups who received both COVID-19 IgM/IgG tests and RT-PCR tests for SARS-CoV-2 from February 29, 2020 to April 29, 2020: 1470 hospitalized patients with COVID-19 from Leishenshan Hospital, Zhongnan Hospital of Wuhan University, and Wuhan No. 7 Hospital, 3832 healthcare providers without COVID-19 diagnosis, 19555 general workers, and 1616 other patients to be admitted to the hospital (N=26473). COVID-19 patients who received IgM/IgG tests <21 days after symptom onset were excluded.


Introduction
Currently, coronavirus disease in 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 1-4 has become a global pandemic. The virus was freely transmitted among residents in the communities [5][6][7][8] in Wuhan, China from late November 2019 till several days after the lockdown of the city on January 23, 2020. Most of SARS-CoV-2 infections do not require medical attention [9][10][11] and only about 5% COVID-19 cases in China need intensive care 12  The presence of SARS-CoV-2 in COVID-19 patients is usually con rmed using real-time reversetranscriptase polymerase chain reaction (RT-PCR) method. [15][16][17] Due to sample collection method, discrepancies in personnel skills training and low virus load in throat swabs after symptom onset 18 , RT-PCR methods have high false-negative test results 17,19,20 . Several days after symptom onset when infected patients start to seek medical attention, virus load in clinical specimens of upper respiratory tract will become relatively low 18,21 . It often takes two or three repeatedly collected specimens to get a positive test result in patients. SARS-CoV-2 enters respiratory epithelial cells via interactions with angiotensin converting enzyme 2 (ACE2) 22 . The spike (S) protein of SARS-CoV-2 mediates binding of the virus with Spike protein's receptor ACE2 and promotes fusion of viral and host cell membranes and subsequent virus entry into the COVID-19 host cell. In patients infected with SARS-CoV-2, IgM antibodies are detectable around 7 days post infection and IgG antibodies usually take two weeks to develop. 18,[22][23][24] Recently, several COVID-19 IgM/IgG rapid tests have been developed around the world 25 . It is reported that in some recovered COVID-19 patients who had two negative RT-PCR tests on nasal or throat swabs taken at least 24 hours apart, segments of virus RNA were still detected in other types of clinical samples, especially in fecal swabs 26 . Additionally, some patients had recurrent positive RT-PCR tests on nasal or throat swabs after recovered from COVID-19. 26 It still unknown whether COVID-19 patients will develop long-term immunity against SARS-CoV-2 and retain long-lasting antibodies after the infection is resolved.
Large-scale sero-epidemiological studies are also needed to assess infection attack rates and disease incidence in the population and herd immunity. In this study, we reported the experiences in COVID- 19 IgM/IgG testing in Wuhan. We assessed prevalence of IgG antibodies against SARS-CoV-2 in hospitalized patients with COVID-19 from Zhongnan Hospital of Wuhan University and Leishenshan Hospital (set up on an emergency basis to admit COVID-19 patients and managed by Zhongnan Hospital of Wuhan University), and Wuhan No. 7 Hospital, healthcare providers without a con rmed COVID-19 diagnosis working in Zhongnan Hospital of Wuhan University, and people from the general population in Wuhan. Considering that Wuhan was the early epicenter of COVID-19 outbreak, Zhongnan Hospital of Wuhan University is a major hospital caring for COVID-19 patients and screening for people with COVID-19 symptoms. Prevalence of IgM antibodies to SARS-CoV-2 were also assessed as a secondary outcome.
Most of these healthcare providers would be inevitably exposed to SARS-CoV-2 during the early days of the outbreak (from late November 2019 to January 20, 2020) when person-to-person transmission was not suspected and little personal protection against this virus was employed among medical personnel, a large proportion of whom would get infected with the virus.

Study Design and Participants
Page 4/23 The study was approved by the institutional ethics board at Zhongnan Hospital of Wuhan University. Requirement for written informed consent was waived by the institutional ethics board for emerging infectious diseases. We conducted a cross-sectional study by including the following four groups of individuals who received both COVID-19 IgM/IgG tests and RT-PCR tests for SARS-CoV-2 from February 29, 2020 to April 29, 2020: hospitalized patients with COVID-19 from Leishenshan Hospital, Zhongnan Hospital of Wuhan University, and Wuhan No. 7 Hospital who received these tests before discharge from hospital, healthcare providers (doctors, nurses, and nursing workers) without a con rmed COVID-19 diagnosis working in Zhongnan Hospital of Wuhan University who received these tests before resuming normal clinical services for patients without COVID-19, general workers without a con rmed COVID-19 diagnosis in Wuhan before returning to work, and other patients without a con rmed COVID-19 diagnosis who received these screening tests before being admitted to Zhongnan Hospital of Wuhan University ( Figure 1). There were 1603 hospitalized patients with COVID-19 who received COVID-19 IgM/IgG tests from February 29 to April 5, 2020 (the last test result was used for analyses). Eighty-four patients were transferred from Zhongnan Hospital of Wuhan University to Leishenshan Hospital and were only counted once each. We excluded 133 patients with COVID-19 whose IgM/IgG tests were less than 21 days after symptom onset to allow enough time for IgG antibodies against SARS-CoV-2 to develop. There were 4099 healthcare providers working in Zhongnan Hospital of Wuhan University, of whom 118 were diagnosed of COVID-19 before March 16, 2020 and 3835 healthcare providers without diagnosed COVID-19 received both tests before resuming normal clinical services. Three healthcare providers who were tested positive for SARS-CoV-2 by RT-PCR tests in their throat swabs were also excluded from the analyses. Before returning to work, 19570 general workers without a diagnosis of COVID-19 in Wuhan who received both tests at Zhongnan Hospital of Wuhan University, of whom 15 were tested positive for SARS-CoV-2 by RT-PCR tests in their throat swabs and were excluded from the analyses. Before admitted to Zhongnan Hospital of Wuhan University for other conditions, 1628 patients without COVID-19 diagnosis received both tests for screening for SARS-CoV-2, of whom 12 were tested positive for SARS-CoV-2 by RT-PCR tests in their throat swabs and were removed from the analyses. In total, we included 1470 patients with COVID-19, 3832 healthcare providers, 19555 workers, and 1616 other patients in the nal analyses (N=26473).
Death among hospitalized COVID-19 patients during hospitalization was ascertained from electronic medical records. Two physicians extracted the following data using data collection form from electronic medical records: demographic information such as age and sex, RT-PCR test date and results, COVID-19 IgM/IgG test date and results, date of symptom onset for COVID-19 patients, treatments received, and clinical outcomes. Another physician in the research team reviewed the collected data. Main outcome of interest was presence of IgG antibodies to SARS-CoV-2. Presence of IgM antibodies to SARS-CoV-2 was a secondary outcome of interest in this study.

Page 5/23
Diagnosis of COVID-19 was based on epidemiological history, clinical manifestations and presence of SARS-CoV-2 in clinical samples con rmed by using real-time RT-PCR method. 11 There were changes in diagnosis of COVID-19 in China, and the case de nition was gradually broadened to allow for detection of milder cases. 57 The con rmed cases were estimated to be 4 times less than that if the later broader case de nition had been adopted earlier. Severity of status of patients with COVID-19 at admission was de ned as moderate, severe, or critical. Patients with mild diseases were not admitted to the above three hospitals and were generally admitted to Fangcang Hospitals (makeshift hospitals).

RT-PCR test for SARS-CoV-2 virus RNA
Clinical specimens were collected for RT-PCR test for SARS-CoV-2. Clinical specimens in COVID-19 patients included nasal swabs, throat swabs, sputum, anal swabs, and bronchoalveolar lavage (BAL), and clinical specimens in the other three groups of people without con rmed COVID-19 diagnosis were only throat swabs. In brief, clinical specimens were collected from these people by trained nurses or physicians wearing proper personal protection equipment. RT-PCR tests for SARS-CoV-2 were performed using a nucleic acid detection kit following the manufacturer's protocol. The lower limit of detection (LOD) for the RT-PCR test is 500 copies/ml. The test simultaneously ampli es and detects two target genes, including open reading frame 1ab (ORF1ab) and nucleocapsid protein (N). Primers used for those two target genes are as follows: ORF1ab: forward primer CCCTGTGGGTTTTACACTTAA, reverse primer

COVID-19 IgM/IgG test for antibodies against SARS-CoV-2
Serum samples from these people were collected. Methods for testing serum IgM and IgG antibodies to SARS-CoV-2 were previously described. 58 COVID-19 IgM/IgG test kits contained recombinant SARS-CoV-2 antigens (spike protein and nucleocapsid protein) labelled with magnetic beads (tested on a fully-automated chemiluminescence immunoassay analyzer) or colloidal gold (test card), anti-human IgM monoclonal antibody, and anti-human IgG monoclonal antibody. These test kits were reported to have high sensitivity and speci city 25,58 . According to the manufacturers, the sensitivity and speci city are ~90% and >99% for IgM, and ~98% and ~98% for IgG, respectively. Validation studies of these test kits conducted by the Department of Laboratory Medicine at Zhongnan Hospital of Wuhan University showed that the sensitivity and speci city are ~80% and ~99% for IgM, and ~96% and ~98% for IgG, respectively.

Statistical Analysis
Continuous variables were reported using mean and 95% con dence interval (CI) if normally distributed. Days from symptom onset to IgM/IgG test were reported as median and interquartile because this variable was not normally distributed. Age was handled as both continuous variable and categorical variable (by 10 years of age) in the analyses. Categorical variables were described as frequency rates and percentages. The χ2 test was used for the comparison of categorical variables and Fisher's exact test was used when frequency was too low. Multigroup comparisons of age between the four groups were performed using ANOVA test, following by Tukey test for adjusting for multiple comparisons. Prevalence of positive IgG test results and 95% CI was also reported. Logistic regression models were tted and χ2 tests were used to compare seroprevalence of IgG and IgM antibodies to SARS-CoV-2 between the four groups. For the assessment of IgM/IgG test results in hospitalized patients, the last test result for each patient was used. We only included individuals from the four groups who received both COVID-19 IgM/IgG tests and RT-PCR tests for SARS-CoV-2. Individuals with missing data were excluded from the analyses. Sensitivity analyses were performed in subgroups, such as age groups and different sexes. Statistical analyses were conducted using SAS software version 9.4 (SAS Institute; Carey, NC). A 2-sided p value of <0.05 was considered statistically signi cant.

Characteristics of participants
Mean age was 58.7 years in 1470 COVID-19 patients, 37.1 years in 3832 healthcare providers, 41.6 years in 19555 general workers and 53.3 years in other patients (Table 1). Hospitalized COVID-19 patients were composed of more older people, while healthcare providers and general workers were mostly young adults. Among COVID-19 patients, median time from symptom onset to IgM/IgG test was 41 days (interquartile range 33-50 days). The last RT-PCR tests for SARS-CoV-2 were positive in only eight COVID-19 patients and all these eight patients were tested positive for IgG antibodies to SARS-CoV-2.
IgG antibodies to SARS-CoV-2 and mortality in COVID-19 patients Among COVID-19 patients, mean age was similar between those with IgG antibodies to SARS-CoV-2 and those without ( Table 3). Presence of IgG antibodies to SARS-CoV-2 was not associated with most demographic characteristics, disease severity, presence of comorbidities, treatment received, and clinical characteristics, except for antibiotics treatment, chloroquine/hydroxychloroquine treatment, and needing intubation. IgG prevalence among hospitalized patients with COVID-19 by demographic and clinical characteristics are presented in Table 4. Case mortality rate was 1.3% (95% CI 0.7-1.9%) in those with IgG antibodies to SARS-CoV-2 during hospitalization, similar to 3.3% (95% CI 0.4-6.2%) in those without.
IgG Antibody to SARS-CoV-2 degradation We identi ed 24 hospitalized patients with COVID-19 and multiple COVID-19 antibody tests who lost previously detected IgG antibodies to SARS-CoV-2 (Table 5). These patients all had positive tests for IgG antibodies to SARS-CoV-2. However, their most recent tests for IgG antibodies to SARS-CoV-2 were negative. Mean age of these patients were 56.6 years old and 58.3% were male. All these patients had moderate COVID-19 at admission, 3 of whom were progressed to critical status. Seven (29.2%) of these patients had hypertension, and 3 (12.5%) had diabetes.

Discussion
We analyzed prevalence of IgG antibodies to SARS-CoV-2 in hospitalized COVID-19 patients, healthcare providers without a con rmed COVID-19 diagnosis, general workers, and other patients to be admitted to hospital. The most intriguing nding of this study is that we found in a subpopulation of hospitalized patients with COVID-19 who tested positive for IgG antibodies to SARS-CoV-2 that the IgG antibodies to SARS-CoV-2 became undetectable overtime. It seems that the longevity of the IgG antibodies to SARS-CoV-2 is questionable. Our ndings are consistent with previous reports on the longevity of antibodies to SARS-CoV-2. [27][28][29] Infected patients with no symptoms 27 or mild symptoms 28, 29 experienced rapid decay of antibodies to SARS-CoV-2 in a couple of months.
The observed only 4% prevalence of IgG antibodies to SARS-CoV-2 in healthcare providers without con rmed COVID-19 also raised the possibility of fast decay of these antibodies. Zhongnan Hospital of Wuhan University is a major hospital caring for COVID-19 patients and screening for people with COVID-19 symptoms in the early epicenter of this pandemic. Most of these healthcare providers in our study would be inevitably exposed to SARS-CoV-2 during the early days of the outbreak (from late November 2019 to January 20, 2020) when person-to-person transmission was not suspected and little personal protection against this virus was employed among medical personnel. Two months of daily work with high-volume patient intake but no proper personal protection equipment is more than su cient to expect a rather high proportion of SARS-CoV-2 infection among these healthcare providers. In Zhongnan Hospital of Wuhan University, 2.88% (118/4099) healthcare workers were diagnosed with COVID-19 before March 16, 2020. Seropositivity results consistently showed that only a small portion of infected patients were detected by RT-PCR tests. [30][31][32][33][34] With a moderate estimation, the true infection rate would bẽ ten times that had been con rmed, i.e., >25% of those healthcare providers without diagnosed COVID-19 had been infected. However, only 4% of those healthcare providers without diagnosed COVID-19 tested positive for IgG antibodies to SARS-CoV-2. A plausible explanation would be the lack of persistent antibodies.
Our observed high prevalence of IgG antibodies to SARS-CoV-2 in older groups (60-69 years old and ≥70 years old) among health care workers and general worker in Wuhan also raised the concern about the longevity of IgG antibodies to SARS-CoV-2, as young or middle-aged people usually took more social responsibilities and had higher chances to get infected during lockdown of the city. We also found that >10% of con rmed COVID-19 cases had no detectable serum levels of IgG antibodies to SARS-CoV-2 after Why antibodies to SARS-CoV-2 in infected persons fade so quickly unlike antibodies to SARS-CoV-1? After infection with SARS-CoV-1, patients start to produce SARS--speci c IgG antibody in the second week, which persists for a long time 36,37 . Even after 210 days after symptom onset, neutralizing viral antibodies (anti-viral IgG) are still detectable in recovered SARS patients 38 . It is believed that the Spike protein and nucleocapsid protein play a central role in the antibody production 36,37,39 . Research has focused on developing vaccines and therapeutics targeting the Spike protein 40, 41 and the nucleocapsid protein 42 . However, our ndings indicate that people are unlikely to develop long-lasting antibodies to SARS-CoV-2 after SARS-CoV-2 infection. Infections with some viruses, such as HIV, do not illicit robust protective immunity 43 , while common cold coronavirus only generate partial protective immunity 44 .
Cross-reactive antibodies in convalescent SARS patients' sera can neutralize other human betacoronaviruses 45 , as those viruses share a signi cant B-cell epitope overlapping the heptad repeat-2 region of the Spike protein. SARS-CoV-2 also belongs to betacoronaviruses. However, no populations had shown explicit immunity to SARS-CoV-2 and people were generally susceptible to this virus before it rst jumped from wild animals to humans. 22, 46-50 It might be due to the dramatic differences between SARS-CoV-1 and SARS-CoV-2 in the receptor-binding region of Spike protein and the key amino acid residues involved in the interaction with human ACE2. 22 Neutralizing antibodies to SARS-CoV-1 may not produce reliable immunoprotection against SARS-CoV-2. However, why long-lasting antibodies are not produced after SARS-CoV-2 infection is still to be studied. It also reported that antibody titer seems to be higher in patients with severe COVID-19 diseases and lower in asymptomatic patients and patients with mild symptoms. [27][28][29]51 A recent study characterized the viral spike protein receptor-binding domain-speci c monoclonal antibodies derived from single B cells of patients with COVID-19. 51 In that study, the three most server cases (one dead) had much higher plasma binding activities to SARS-CoV-2 spike protein, spike protein receptor-binding domain, and nucleocapsid protein than the other ve cases with mild symptoms, which casts some doubts on the relationship between antibody response and disease progression and the utilization of neutralizing antibodies as prophylactic and therapeutic SARS-CoV-2 interventions. We also found mortality risks were similar between hospitalized COVID-19 patients with IgG antibodies to SARS-CoV-2 and those without, which indicates that absence of IgG antibodies may not affect clinical end outcome. Whether therapeutic antibodies targeting the spike protein are effective will also be a question. Antibody-dependent enhancement of viral entry via the binding of those neutralizing antibodies and the spike protein is also a big concern to be closely monitored [52][53][54][55] .
COVID-19 IgM/IgG tests in the US and around the world as reported in the news constantly showed that the true infection rate would be 10 to 80 times higher than that had been con rmed by RT-PCR tests for SARS-CoV-2. [30][31][32][33][34] Seroprevalence of antibodies to SARS-CoV-2 in 1021 people before resuming work from April 3 to 15, 2020 in Wuhan was reported to be ~10%, 30 about 20 times higher than the infection attack rate calculated from the con rmed COVID-19 cases. In Los Angeles County, California, the seroprevalence of IgM/IgG antibodies to SARS-CoV-2 was 4.65%, over 40 times higher than that estimated from con rmed cases. 31 In New York City, a record 21.2% positive rate was reported with young and middleaged people having the highest positive rate. Our observed 4.6% prevalence of IgG antibodies to SARS-CoV-2 in general workers is also about 10 times higher than that of the general population in Wuhan (50,000 cases of COVID-19 in 11 million people). The proportion of people infected with SARS-CoV-2 who have no symptom or only mild symptoms that do not need medical attention or hospitalization may account for the majority of SARS-CoV-2 infections. Detecting patients with SARA-CoV-2 infection who are not in urgent need of hospitalization may have important public health applications in tracing their close contacts and preventing these people from spreading the infection to others. Pool testing to enable largescale screening 56 and introducing antibody test into the testing scheme may be considered to detect most of the infected persons.
The main strength of this study is that we analyzed data on COVID-19 IgM/IgG tests on a large cohort of individuals from three hospitals in Wuhan, the epicenter of COVID-19 outbreak in China. Antibody tests on healthcare providers provided valuable information on the longevity of antibodies to SARS-CoV-2 after infection. Limitations of this study include that we used convenient samples from three hospitals. We also do not have long-term follow-up data on recovered COVID-19 patients. Whether or not those patients will lose COVID-19 IgG antibodies in the next few months is still to be studied. Additionally, COVID-19 IgM/IgG test does not have perfect sensitivity and speci city for detecting IgG antibodies to SARS-CoV-2 in the serum. Nevertheless, we observed that long-lasting antibodies to SARS-CoV-2 are unlikely produced after infection. Our ndings may have important implications for herd immunity, antibody-based therapeutics, public health strategies, surveillance of past infections of SARS-COV-2, and vaccine development.
In conclusion, we found in a small sample of hospitalized patients with COVID-19 who tested positive for IgG antibodies to SARS-CoV-2 that these antibodies became undetectable overtime. In addition, very few healthcare providers without con rmed COVID-19 diagnosis in Wuhan have IgG antibodies to SARS-CoV-2, though a substantial portion of them had been infected with the virus. About 90% hospitalized COVID-19 patients still have IgG antibodies to SARS-CoV-2 before discharge from hospital, while more than 10% of COVID-19 patients do not have those antibodies after 21 days post symptom onset. Antibodies against this virus produced in infected people seem to be not long lasting.

Conclusion
IgG antibodies to SARS-CoV-2 in infected people may become undetectable overtime.

Declarations Ethical Approval and Consent to participate
The study was approved by the institutional ethics board at Zhongnan Hospital of Wuhan University. Requirement for written informed consent was waived by the institutional ethics board for emerging infectious diseases.

Consent for publication
The authors have not published any related papers from the same study, and this paper is not under consideration elsewhere. None of this paper's contents have been previously published. All authors have read and approved the manuscript.

Availability of supporting data
For protection of patient privacy, data and analyses codes are available from the corresponding authors on request. All request for raw and analyzed data and materials will be reviewed by the corresponding authors to verify whether the request is subject to any intellectual property or con dentiality obligations. Access will be granted after a signed data access agreement is attained.

Competing interests
The authors have no con icts of interest to report.

Primary Funding Sources
Part of the study was supported by National Key Research and Development Program of China (2020YFC0845500) and Postdoctoral Research Foundation of China (2020M670071ZX). The content is solely the responsibility of the authors and does not necessarily represent the o cial views of the sponsors.

Role of the Funder/Sponsor
The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript, and decision to submit the manuscript for publication.

Financial Disclosures
The authors have no con ict of interest to disclose.

Author Contributions
Dr. X. Wang had full access to all the data in the study and takes responsibility for the integrity of the BriefCommunication. Nature Reviews Immunology. 2020-04-