Coronavirus disease 2019 (COVID-19) is an emerging infectious disease, and the ongoing quick spread of COVID-19 cases has become a global pandemic (1). Detecting serum-specific antibodies has become one of the key approaches for the identification of COVID-19 virus infection. However, the interpretation of antibody detection largely depends on understanding host antibody responses during infection, but the profiles of the antibody responses and the relationship between IgM, IgA, IgG, and the neutralizing antibodies (NAb) among COVID-19 patients with different disease severity is not yet clearly understood. Therefore, the purpose of this study was to supplement the knowledge regarding the human immune response to COVID-19 virus infection.
In this study, the confirmed COVID-19 patients by real-time reverse transcription polymerase chain reaction (real-time RT-PCR) were included. A total of 210 sera were collected from 160 COVID-19 patients based on varying course of disease in Gansu and Heilongjiang provinces of China during January to March in 2020. The age of the cases ranged between 1–98 years (median age: 46 years). The sampling time of all the sera ranged from 0 to 64 days after onset of illness (median days: 28 days). According to disease classifications outlined in the “Guidelines on the Novel Coronavirus-Infected Pneumonia Diagnosis and Treatment (Seventh Edition)” issued by the National Health Commission of China (NHC), the majority of the COVID-19 cases (76.7%) belonged to the mild (39 cases/45 samples) and moderate (83 cases/109 samples) categories, and the remainder were classified as either severe (34 cases/52 samples) or critical (4 cases/4 samples); 81% of the severe and critical cases were older than 40 years old.
After the sera were inactivated at 56 °C for 30 minutes, the IgM, IgA, and IgG antibodies against COVID-19 virus were detected by using a commercial magnetic chemiluminescence enzyme immunoassay (MCLIA) kit (Bioscience, China) (2). The luminescence value of each sample was positively correlated with the antibody concentration to evaluate the level of IgM/IgA/IgG antibodies against COVID-19 virus in the serum samples. NAb was evaluated using the microneutralization assay (3). Antibody titers greater than or equal to 1∶8 indicated a positive result in this study. To calculate the geometric mean titer (GMT), antibody titers of <1∶8 and >1∶256 were assigned as 1∶4 and 1∶256, respectively, and the 95% confidence interval (95% CI) was calculated. The median and interquartile range (IQR) were used for statistical analysis of IgM/IgA/IgG/NAb levels. The Kruskal-Wallis test and Pearson chi-square test were used to test the differences among groups, including age, days after onset of disease, and clinical classification by using R software (version 3.5.2, Lucent Technologies, FL, USA). A p-value less than 0.05 was considered statistically significant.
The dynamics analysis for four types of antibodies showed that the positive rate of IgM/IgA was slightly higher than that of IgG/NAb within the first two weeks after onset. The positive rate of the 4 antibodies successively reached 100% after two weeks. The dynamics of positive rates of IgM and IgA was generally consistent reaching 44.4% and 38.9%, respectively, within 4–7 days followed by a peak after 2 weeks. The positive rate between IgG and NAb was consistent throughout the course of the disease, except for the 6th week, when the positive rate was 81.8% and 95.5% for NAb and IgG, respectively. The low positive rate of NAb in the 6th week might be associated with the low level of IgG in three cases with negative NAb (Table 1). The levels of IgM/IgA/IgG/NAb were also analyzed based on days after onset. All antibodies could be detected at low levels within 0–3 days. The levels of the 4 antibodies were similar within 2 weeks, followed by a rapid increase and the maintenance of a high level from the 3rd and 4th weeks. Similar to the positive rate, the levels of IgM and IgA were consistent throughout the course of the disease. In contrast, the NAb antibody rapidly decreased from a GMT of 1∶62 during the 5th week to 1∶31 by the 9th week, while IgG remained at a relatively stable level (Figure 1).
Days after onset No. of samples IgA IgM IgG Neutralization antibody No. of positive samples Detection rate (%) No. of positive samples Detection rate (%) No. of positive samples Detection rate (%) No. of positive samples Detection rate (%) 0–3 26 7 26.9 10 38.5 5 19.2 4 15.4 4–7 18 7 38.9 8 44.4 6 33.3 7 38.9 8–14 13 8 61.5 8 61.5 6 46.2 6 46.2 15–21 21 20 95.2 21 100.0 20 95.2 20 95.2 22–28 27 25 92.6 27 100.0 27 100.0 26 96.3 29–35 33 33 100.0 30 90.9 31 93.9 31 93.9 36–42 22 20 90.9 22 100.0 21 95.5 18 81.8 43–64 50 48 96.0 45 90.9 50 100.0 47 94.0 Total 210 168 80.0 171 81.4 166 79.0 159 75.7
Table 1. The positive detection rate of antibodies against COVID-19 Virus in different course of disease.
Figure 1. The changes in levels of antibodies, including IgM, IgA, IgG, and NAb, against COVID-19 virus after symptom onset. COVID-19 patients from the cross-sectional cohort were grouped based on days after disease onset (0–3 days, 4–7 days, 8–14 days, 15–21 days, 22–28 days, 29–35 days, 36–42 days, 43–64 days). The left Y-axis indicated the S/CO values of the IgM, IgA, and IgG antibodies converted by Log2(S/CO+1), and the antibody levels of all the groups were shown with the median values. The right Y-axis indicated the geometric mean titer (GMT) level of the neutralizing antibody.
All 160 cases in this study were divided into 3 age groups: 1–39 years, 40–59 years, and ≥60 years. Due to the limited number of cases aged 1–19 years, all cases younger than 40-year-old were classified into 1 group for statistical analysis in this study. The levels of IgA, IgG, and NAb antibodies showed significant differences among the 3 age groups between the 3rd and 4th weeks (p<0.05), and the level of IgM and NAb showed significant differences among 3 age groups between the 5th and 9th weeks (p<0.05). No obvious differences were observed for all four types of antibodies among different age groups within two weeks after onset (Figure 2).
Figure 2. The changes in levels of antibodies, including IgM (B), IgA (C), IgG (D), and NAb (E), against COVID-19 virus after disease onset based on the age distribution of the COVID-19 cases. (A) showed the number of samples in age groups including those aged 1–39 years, those aged 40–59 years, and those over 60 years of age. The symbol “**” indicated statistical significance (p<0.05) across the age groups. Scatter and histogram plots were created with GraphPad Prism (version 8.0.2, GraphPad Software, San Diego, CA, USA). GMT=geometric mean titer.
Variable responses among the four different disease types were observed for the four types of antibodies. The antibody levels of IgM and IgA showed significant differences among the different disease types within two weeks of onset (p<0.05), and higher antibody levels were observed in the mild and moderate groups than those in the severe and critical groups. While the antibody levels of IgM, IgA, IgG, and NAb showed significant differences among the different disease severities after 4 weeks of onset (p<0.05), higher levels were observed in the severe and critical groups. Persistently higher levels of IgG and NAb were observed from the beginning of disease onset among the four critical-type cases in this study compared to other types of disease classification (Figure 3).
Figure 3. The antibody response of IgM (B), IgA (C), IgG (D), and NAb (E) against COVID-19 virus among COVID-19 patients with different disease severity. (A) showed the number of samples in the different disease severities of COVID-19 cases, including mild (Mi), moderate (Mo), and severe and critical types (S/C). (F) showed the antibody levels of critical-type cases based on number of days after disease onset. The level of NAb was shown by the titer of the NAb, instead of geometric mean titer (GMT), due the limited number of critical cases. The symbol “**” indicated statistical significance (p<0.05) across age groups. Scatter and histogram plots were created with GraphPad Prism (version 8.0.2, GraphPad Software, San Diego, CA, USA).