Main

The continued spread of coronavirus disease 2019 (COVID-19) has prompted widespread concern around the world, and the World Health Organization (WHO), on 11 March 2020, declared COVID-19 a pandemic. Studies on severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS) showed that virus-specific antibodies were detectable in 80–100% of patients at 2 weeks after symptom onset1,2,3,4,5,6. Currently, the antibody responses against SARS-CoV-2 remain poorly understood and the clinical utility of serological testing is unclear7.

A total of 285 patients with COVID-19 were enrolled in this study from three designated hospitals; of these patients, 70 had sequential samples available. The characteristics of these patients are summarized in Supplementary Tables 1 and 2. We validated and used a magnetic chemiluminescence enzyme immunoassay (MCLIA) for virus-specific antibody detection (Extended Data Fig. 1a–d and Supplementary Table 3). Serum samples from patients with COVID-19 showed no cross-binding to the S1 subunit of the SARS-CoV spike antigen. However, we did observe some cross-reactivity of serum samples from patients with COVID-19 to nucleocapsid antigens of SARS-CoV (Extended Data Fig. 1e). The proportion of patients with positive virus-specific IgG reached 100% approximately 17–19 days after symptom onset, while the proportion of patients with positive virus-specific IgM reached a peak of 94.1% approximately 20–22 days after symptom onset (Fig. 1a and Methods). During the first 3 weeks after symptom onset, there were increases in virus-specific IgG and IgM antibody titers (Fig. 1b). However, IgM showed a slight decrease in the >3-week group compared to the ≤3-week group (Fig. 1b). IgG and IgM titers in the severe group were higher than those in the non-severe group, although a significant difference was only observed in IgG titer in the 2-week post-symptom onset group (Fig. 1c, P = 0.001).

Fig. 1: Antibody responses against SARS-CoV-2.
figure 1

a, Graph of positive rates of virus-specific IgG and IgM versus days after symptom onset in 363 serum samples from 262 patients. b, Levels of antibodies against SARS-CoV-2 in patients at different times after symptom onset. c, Comparison of the level of antibodies against SARS-CoV-2 between severe and non-severe patients. The boxplots in b and c show medians (middle line) and third and first quartiles (boxes), while the whiskers show 1.5× the interquartile range (IQR) above and below the box. Numbers of patients (N) are shown underneath. P values were determined with unpaired, two-sided Mann–Whitney U-test.

Source Data

Sixty-three patients with confirmed COVID-19 were followed up until discharge. Serum samples were collected at 3-day intervals. Among these, the overall seroconversion rate was 96.8% (61/63) over the follow-up period. Two patients, a mother and daughter, maintained IgG- and IgM-negative status during hospitalization. Serological courses could be followed for 26 patients who were initially seronegative and then underwent seroconversion during the observation period. All these patients achieved seroconversion of IgG or IgM within 20 days after symptom onset. The median day of seroconversion for both IgG and IgM was 13 days post symptom onset. Three types of seroconversion were observed: synchronous seroconversion of IgG and IgM (nine patients), IgM seroconversion earlier than that of IgG (seven patients) and IgM seroconversion later than that of IgG (ten patients) (Fig. 2a). Longitudinal antibody changes in six representative patients of the three types of seroconversion are shown in Fig. 2b–d and Extended Data Fig. 2a–c.

Fig. 2: Seroconversion time of the antibodies against SARS-CoV-2.
figure 2

a, Seroconversion type of 26 patients who were initially seronegative during the observation period. The days of seroconversion for each patient are plotted. b–d, Six representative examples of the three seroconversion type: synchronous seroconversion of IgG and IgM (b), IgM seroconversion earlier than that of IgG (c) and IgM seroconversion later than that of IgG (c).

IgG levels in the 19 patients who underwent IgG seroconversion during hospitalization plateaued 6 days after the first positive IgG measurement (Extended Data Fig. 3). Plateau IgG levels varied widely (more than 20-fold) across patients. IgM also showed a similar profile of dynamic changes (Extended Data Fig. 4). We found no association between plateau IgG levels and the clinical characteristics of the patients (Extended Data Fig. 5a–d). We next analyzed whether the criteria for confirmation of MERS-CoV infection recommended by WHO, including (1) seroconversion or (2) a fourfold increase in IgG-specific antibody titers, are suitable for the diagnosis of COVID-19 (using paired samples from 41 patients). The initial sample was collected in the first week of illness and the second was collected 2–3 weeks later. Of the patients whose IgG was initially seronegative in the first week of illness, 51.2% (21/41) underwent seroconversion. A total of 18 patients were initially seropositive in the first week of illness; of these, eight patients had a fourfold increase in virus-specific IgG titers (Extended Data Fig. 6). Overall, 70.7% (29/41) of the patients with COVID-19 met the criteria of IgG seroconversion and/or fourfold increase or greater in the IgG titers.

To investigate whether serology testing could help identify patients with COVID-19, we screened 52 suspected cases in patients who displayed symptoms of COVID-19 or abnormal radiological findings and for whom testing for viral RNA was negative in at least two sequential samples. Of the 52 suspected cases, four had virus-specific IgG or IgM in the initial samples (Extended Data Fig. 7). Patient 3 had a greater than fourfold increase in IgG titer 3 days after the initial serology testing. Interestingly, patient 3 also tested positive for viral infection by polymerase chain reaction with reverse transcription (RT–PCR) between the two antibody measurements. IgM titer increased over three sequential samples from patient 1 (<4-fold). Patient 4 had 100-fold higher IgG and tenfold higher IgM titers than the cutoff values in two sequential samples. Patient 2 tested positive for both virus-specific IgG and IgM. An increase of IgG and/or IgM in sequential samples or a positive result in a single sample collected 2 weeks after symptoms suggest that these three patients were infected with SARS-CoV-2.

We further demonstrated the application of serology testing in surveillance in a cluster of 164 close contacts of patients with known COVID-19. Sixteen individuals were confirmed to be infected with SARS-CoV-2 by RT–PCR, with three cases reporting no symptoms. The other 148 individuals had negative RT–PCR results and no symptoms (Extended Data Fig. 8). Serum samples were collected from these 164 individuals for antibody tests ~30 days after exposure. The 16 RT–PCR-confirmed cases were all positive for virus-specific IgG and/or IgM. Moreover, 7 of the 148 individuals with negative RT–PCR results had positive virus-specific IgG and/or IgM, indicating that 4.3% (7/164) of the close contacts were missed by the nucleic acid test. Ten of the 164 close contacts who had positive virus-specific IgG and/or IgM were asymptomatic.

Our study showed that the criteria for the confirmation of MERS-CoV infection are suitable for most patients with COVID-19. However, a collection of the first serum sample as early as possible is required for some patients to meet these criteria, because 12.2% (5/41) of the patients had already plateaued in IgG titer within 7 days of symptom onset (Extended Data Fig. 6). For those patients who were not sampled during the ideal window, repeated serological tests would be needed to confirm an antibody response to SARS-CoV-2 infection.

Our study has some limitations. First, we did not test samples for virus neutralization and therefore the neutralizing activities of the detected IgG antibodies are unknown. Second, due to the small sample size of patients in severe and critical condition, it is difficult to determine the association between antibody response and clinical course.

RT–PCR-based viral RNA detection is sensitive and can effectively confirm early SARS-CoV2 infection8. Our data indicate that virus-specific antibody detection for COVID-19 could be important (1) as a complement to nucleic acid testing for the diagnosis of suspected cases with negative RT–PCR results and (2) in surveying for asymptomatic infection in close contacts. Confirming suspected COVID-19 cases as early as possible with the help of serological testing could reduce exposure risk during repeated sampling and save valuable RT–PCR tests. In our small-scale survey, seven cases with negative nucleic acid results and no symptoms showed positive IgG and/or IgM. This highlights the importance of serological testing to achieve more accurate estimates of the extent of the COVID-19 pandemic.

Methods

Study design

A total of 285 patients with COVID-19 were enrolled in this cross-sectional study from three designated hospitals in Chongqing, a province-level municipality adjacent to Hubei Province, which was the starting point and epicenter of the COVID-19 epidemic. These three hospitals—Chongqing Three Gorges Central Hospital (TGH), Yongchuan Hospital Affiliated to Chongqing Medical University (CQMU) (YCH) and Chongqing Public Health Medical Center (CQPHMC)—were assigned by the Chongqing municipal people’s government to admit patients from the three designated areas. All enrolled patients were confirmed to be infected with SARS-CoV-2 by RT–PCR assays on nasal and pharyngeal swab specimens. The median age of these enrolled patients was 47 years (IQR, 34–56 years) and 55.4% were males. Among them, 250 patients had an epidemiological history, while 262 patients had a clear record of symptom onset and 70 patients had multiple serum samples. A total of 363 serum samples from patients with a clear symptom onset history were included in the analysis. Of the 285 patients, 39 were classified as in a severe or critical condition according to the Diagnosis and Treatment Protocol for Novel Coronavirus Pneumonia (Trial Version 7), released by the National Health Commission & State Administration of Traditional Chinese Medicine. For the follow-up cohort, serum samples from 63 patients at YCH were taken at 3-day intervals from 8 February 2020 until hospital discharge. To analyze whether the serological criteria for MERS-CoV confirmation recommended by WHO were suitable for the diagnosis of COVID-19, two inclusion criteria were set to screen patients: (1) first serum sample collected within the first week of illness onset or (2) first serum sample collected within at least 7 days of illness onset but with negative IgG. Thirty-four patients met criterion 1 and seven patients met criterion 2.

To evaluate the potential of the serological test in COVID-19 diagnosis, we enrolled 52 patients with suspected COVID-19 admitted to Wanzhou People’s Hospital (Chongqing, China) who had respiratory symptoms or abnormal pulmonary imaging, but negative RT–PCR results in at least two sequential tests. Serum samples were collected at the time indicated in Extended Data Fig. 7 and antibodies against SARS-CoV-2 were tested.

A serological survey was performed in a cluster of close contacts composed of 164 individuals, identified by the local center for disease control and prevention (Wanzhou, Chongqing). A couple who had traveled back from Wuhan city, and who were confirmed to be SARS-CoV-2 infected on 4 February 2020, were deemed the first-generation patients in this contact network. All other cases in this cohort had close contact (either directly or indirectly) with this couple in the period from 20 January to 6 February 2020. On 1 March, serum samples were collected from these 164 cases for antibody tests.

Definitions

Patients with epidemiologic history were defined as follows: Wuhan residents; recently been to Wuhan (30 days preceding symptom onset); local resident who had contact with confirmed cases. Seroconversion was defined as a transition of the test results for IgG or IgM against SARS-CoV-2 from negative to positive results in sequential samples. Antibody levels were presented as the measured chemiluminescence values divided by the cutoff (absorbance/cutoff, S/CO): S/CO > 1 was defined as positive and S/CO ≤ 1 as negative.

Detection of IgG and IgM against SARS-CoV-2

To measure the level of IgG and IgM against SARS-CoV-2, serum samples were collected from the patients. All serum samples were inactivated at 56 °C for 30 min and stored at −20 °C before testing. IgG and IgM against SARS-CoV-2 in plasma samples were tested using MCLIA kits supplied by Bioscience Co. (approved by the China National Medical Products Administration; approval numbers 20203400183(IgG) and 20203400182(IgM)), according to the manufacturer’s instructions. MCLIA for IgG or IgM detection was developed based on a double-antibody sandwich immunoassay. The recombinant antigens containing the nucleoprotein and a peptide from the spike protein of SARS-CoV-2 were conjugated with FITC and immobilized on anti-FITC antibody-conjugated magnetic particles. Alkaline phosphatase conjugated anti-human IgG/IgM antibody was used as the detection antibody. The tests were conducted on an automated magnetic chemiluminescence analyzer (Axceed 260, Bioscience) according to the manufacturer’s instructions. All tests were performed under strict biosafety conditions. The antibody titer was tested once per serum sample. Antibody levels are presented as the measured chemiluminescence values divided by the cutoff (S/CO). The cutoff value of this test was defined by receiver operating characteristic curves. Antibody levels in the figures were calculated as log2(S/CO + 1).

Performance evaluation of the SARS-CoV-2-specific IgG/IgM detection assay

The precision and reproducibility of the MCLIA kits were first evaluated by the National Institutes for Food and Drug Control. Moreover, 30 serum samples from patients with COVID-19 showing different titers of IgG (range 0.43–187.82) and IgM (range 0.26–24.02) were tested. Each individual sample was tested in three independent experiments, and the coefficient of variation (CV) was used to evaluate the precision of the assay. Finally, 46 serum samples from patients with COVID-19 were assessed using different batches of the diagnostic kit for SARS-CoV-2-specific IgG or IgM antibody; reproducibility was calculated based on the results from two batch experiments.

Cross-reactivity of antigens from SARS-CoV and SARS-CoV-2

Two recombinant SARS-CoV nucleocapsid (N) proteins from two different sources (Sino Biological, cat. no. 40143-V08B; Biorbyt, cat. no. orb82478), the recombinant S1 subunit of the SARS-CoV spike (Sino Biological, cat. no. 40150-V08B1) and the homemade recombinant N protein of SARS-CoV-2 were used in a chemiluminescence enzyme immunoassay (CLEIA), respectively. The concentration of antigens used for plate coating was 0.5 μg ml−1. The dilution of alkaline phosphatase conjugated goat anti-human IgG antibody was 1:2,500. Five serum samples from patients with COVID-19 and five serum samples from healthy controls were diluted (1:50) and tested using CLEIA assays. The binding ability of antibody to antigen in a sample was measured in relative luminescence units.

Statistical analyses

Continuous variables are expressed as the median (IQR) and were compared with the Mann–Whitney U-test. Categorical variables are expressed as numbers (%) and were compared by Fisher’s exact test. A P value of <0.05 was considered statistically significant. Statistical analyses were performed using R software, version 3.6.0.

Ethical approval

The study was approved by the Ethics Commission of Chongqing Medical University (ref. no. 2020003). Written informed consent was waived by the Ethics Commission of the designated hospital for emerging infectious diseases.

Reporting Summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.