ReportsCORONAVIRUS

Persistence and decay of human antibody responses to the receptor binding domain of SARS-CoV-2 spike protein in COVID-19 patients

See allHide authors and affiliations

Science Immunology  08 Oct 2020:
Vol. 5, Issue 52, eabe0367
DOI: 10.1126/sciimmunol.abe0367
  • Fig. 1 Measurement of IgG, IgM, IgA against SARS-CoV-2 spike protein receptor binding domain among pre-pandemic controls and PCR positive cases.

    Each dot represents a unique measurement of an isotype (Row A: IgG, Row B: IgM, Row C: IgA) in pre-pandemic controls (left panels) and PCR positive cases (right panels). The blue line is a loess smooth nonparametric function. Black dashed lines indicate the maximum concentration (μg/mL) found among pre-pandemic controls (IgG: 0.57, IgM: 2.63, IgA: 2.02). Horizontal jitter was introduced into the pre-pandemic controls. The limit of detection (μg/mL) was 0.04 for IgG, 0.28 for IgM, and 0.30 for IgA.

  • Fig. 2 Parametric and nonparametric model estimates of time to seroconversion and seroreversion for each isotype.

    A) The isotype cut-offs chosen for seroconversion were the maximum concentration (μg/mL) found among pre-pandemic controls (IgG: 0.57, IgM: 2.63, IgA: 2.02). The solid line represents the estimated cumulative distribution function of the time to seroconversion or reversion with 100 bootstrapped fits shown as transparent lines. The parametric accelerated failure time models assume a log-normal time-to-event distribution. Nonparametric estimates shown in grey were calculated using the Turnbull method. Only 3 individuals seroreverted for IgG, so no model is included. B) The table indicates the estimated average number of days since onset of symptoms it takes for a percentage of cases to seroconvert or serorevert. Bootstrap 95% confidence intervals are shown in parentheses.

  • Fig. 3 SARS-CoV-2 pseudovirus neutralization antibody titers in symptomatic PCR positive cases and correlation with anti-RBD IgG responses.

    A) Each point represents a measurement of 50% neutralizing titer (NT50). Lines connect measurements from the same individual and a loess smooth function is shown in blue. B) The overall repeated measures correlation coefficient (r) is shown. Lines represent simple linear models for each time period.

  • Table 1 Individual characteristics of PCR-positive SARS-CoV-2 cases and pre-pandemic controls.

    CharacteristicPre-pandemic
    Controls*
    (N=1,548)
    PCR-positive
    Cases
    (N=343)
    Age
    Median [IQR]37 [30–54]59 [45–71]
    <65 years (%)1,386 (90)213 (62)
    65+ years (%)162 (10)130 (38)
    Female (%)1,024 (66)132 (38)
    Race or ethnic group¥
    White (%)NA125 (36)
    Black or African American (%)NA34 (10)
    Hispanic or Latino (%)NA121 (35)
    Asian, American Indian, Alaska
    Native or Other (%)
    NA30 (9)
    Immunosuppressed (%)NA26 (8)
    Severity
    Not Hospitalized (%)NA24 (7)
    Hospitalized, no ICU (%)NA138 (40)
    Hospitalized, required ICU (%)NA137 (40)
    Died due to COVID-19 (%)NA43 (13)

    *Pre-pandemic controls included healthy adults (n=274), patients undergoing routine serology testing (n=1241), and patients presenting with other known febrile illnesses (n = 33), including 13 with bacteremia (e.g., S. aureus, S. pneumoniae, E. coli, or K. pneumoniae confirmed by standard microbiologic techniques), 4 with babesiosis (confirmed by microscopy and/or PCR), 1 with presumed scrub typhus, and 15 with viral respiratory infections (e.g., influenza [7], parainfluenza [4], respiratory syncytial virus [3], and metapneumovirus [1] confirmed by PCR or direct fluorescent antibody test).

    ¥Data available for 310 cases.

    Data available for 342 cases.

    • Table 2 Predictive accuracy of individual isotypes for classifying controls and cases across time.

      IsotypeDays since symptom onsetAUC (95% CI)Sensitivity (95% CI)
      IgG≤7 days0.68 (0.66–0.70)0.07 (0.03–0.12)
      8-14 days0.91 (0.89–0.92)0.51 (0.43–0.58)
      15-28 days0.99 (0.99–1.00)0.95 (0.92–0.98)
      >28 days0.99 (0.99–1.00)0.95 (0.91–0.98)
      IgA≤7 days0.63 (0.61–0.65)0.07 (0.03–0.11)
      8-14 days0.87 (0.85–0.89)0.44 (0.38–0.51)
      15-28 days0.98 (0.97–0.98)0.89 (0.84–0.94)
      >28 days0.98 (0.97–0.98)0.60 (0.51–0.68)
      IgM≤7 days0.60 (0.58–0.62)0.08 (0.03–0.13)
      8-14 days0.87 (0.85–0.89)0.55 (0.48–0.62)
      15-28 days0.98 (0.97–0.99)0.86 (0.81–0.92)
      >28 days0.93 (0.91–0.94)0.51 (0.43–0.59)

      The isotype cut-offs chosen for calculating sensitivity were the maximum value found among pre-pandemic controls (IgG: 0.57 μg/mL, IgM: 2.63 μg/mL, IgA: 2.02 μg/mL). Bootstrap 95% confidence intervals are shown in parentheses.

      Supplementary Materials

      • immunology.sciencemag.org/cgi/content/full/5/52/eabe0367/DC1

        Figure S1: Number of PCR positive cases with a sample taken during each week since symptom onset.

        Figure S2. Smooth average measurements of IgG, IgM, and IgA against SARS-CoV-2 spike protein receptor binding domain among PCR positive cases across time.

        Figure S3. Individual trajectories for 16 randomly selected individuals with 4 or more measurements.

        Figure S4. Measurements of IgG, IgM, and IgA against SARS-CoV-2 spike protein receptor binding domain among pre-pandemic controls and symptomatic PCR positive cases.

        Figure S5. Receiver operating characteristic curve from random forest models and isotype contributions.

        Figure S6. Confusion matrices and out-of-bag error estimates for random forest models.

        Figure S7. Confusion matrices and out-of-bag error estimates for random forest models with downsampled controls.

        Figure S8. Measurements of IgG, IgA, and IgM against the RBD of other coronaviruses among pre-pandemic controls and PCR positive cases.

        Figure S9. Correlation between plasma and dried blood spot measurements (DBS).

        Table S1. Full amino acid sequences of the coronavirus receptor-binding domains (RBDs) used in this study.

        Table S2. Predictive accuracy of multiple isotypes for classifying controls and cases over time since symptom onset.

        Table S3. Parametric estimates of median time to seroconversion for each isotype by different patient characteristics.

        Table S4. Raw data file (Excel spreadsheet).

      • The PDF file includes:

        • Fig. S1. Number of PCR positive cases with a sample taken during each week since symptom onset.
        • Fig. S2. Smooth average measurements of IgG, IgM, and IgA against SARS-CoV-2 spike protein receptor binding domain among PCR positive cases across time.
        • Fig. S3. Individual trajectories for 16 randomly selected individuals with 4 or more measurements.
        • Fig. S4. Measurements of IgG, IgM, and IgA against SARS-CoV-2 spike protein receptor binding domain among pre-pandemic controls and symptomatic PCR positive cases.
        • Fig. S5. Receiver operating characteristic curve from random forest models and isotype contributions.
        • Fig. S6. Confusion matrices and out-of-bag error estimates for random forest models.
        • Fig. S7. Confusion matrices and out-of-bag error estimates for random forest models with downsampled controls.
        • Fig. S8. Measurements of IgG, IgA, and IgM against the RBD of other coronaviruses among pre-pandemic controls and PCR positive cases.
        • Fig. S9. Correlation between plasma and dried blood spot measurements (DBS).
        • Table S1. Full amino acid sequences of the coronavirus receptor-binding domains (RBDs) used in this study.
        • Table S2. Predictive accuracy of multiple isotypes for classifying controls and cases over time since symptom onset.
        • Table S3. Parametric estimates of median time to seroconversion for each isotype by different patient characteristics.

        [Download PDF]

        Other Supplementary Material for this manuscript includes the following:

        • Table S4. Raw data file (Excel spreadsheet).

      Stay Connected to Science Immunology

      Navigate This Article