Relationship of Heterologous Virus Responses and Outcomes in Hospitalized COVID-19 Patients

Abstract

The clinical trajectory of COVID-19 may be influenced by previous responses to heterologous viruses. We examined the relationship of Abs against different viruses to clinical trajectory groups from the National Institutes of Health IMPACC (Immunophenotyping Assessment in a COVID-19 Cohort) study of hospitalized COVID-19 patients. Whereas initial Ab titers to SARS-CoV-2 tended to be higher with increasing severity (excluding fatal disease), those to seasonal coronaviruses trended in the opposite direction. Initial Ab titers to influenza and parainfluenza viruses also tended to be lower with increasing severity. However, no significant relationship was observed for Abs to other viruses, including measles, CMV, EBV, and respiratory syncytial virus. We hypothesize that some individuals may produce lower or less durable Ab responses to respiratory viruses generally (reflected in lower baseline titers in our study), and that this may carry over into poorer outcomes for COVID-19 (despite high initial SARS-CoV-2 titers). We further looked at longitudinal changes in Ab responses to heterologous viruses, but found little change during the course of acute COVID-19 infection. We saw significant trends with age for Ab levels to many of these viruses, but no difference in longitudinal SARS-CoV-2 titers for those with high versus low seasonal coronavirus titers. We detected no difference in longitudinal SARS-CoV-2 titers for CMV seropositive versus seronegative patients, although there was an overrepresentation of CMV seropositives among the IMPACC cohort, compared with expected frequencies in the United States population. Our results both reinforce findings from other studies and suggest (to our knowledge) new relationships between the response to SARS-CoV-2 and Abs to heterologous viruses.

Figure 1.

Serum IgG levels specific for SARS-CoV-2 antigens in COVID-19 patients, by clinical trajectory group (1–5, where 1=most mild and 5=fatal). All data are detrended for covariates. MFI = median fluorescence intensity. A. Visit 1 SARS-CoV-2-specific IgG levels tend to increase with severity, from trajectory group 1 to 4, reaching significance for RBD and S-M antigens for trajectory group 4 (vs. group 1). Squares are means. B. Longitudinal trend for NP-specific serum IgG differs significantly for trajectory group 4 (vs. group 1; P=0.003). Longitudinal trends for the RBD- and S-M-specific serum IgG were not significantly different from trajectory group 1 (not shown).

Figure 2.

Serum antibody levels specific for seasonal coronavirus antigens in COVID-19 patients. All data are detrended for covariates. MFI = median fluorescence intensity. A. Visit 1 serum IgG levels for seasonal coronavirus antigens decrease on average from clinical trajectory group 1 to 5 (1=most mild; 5=fatal), with significant differences (vs. group 1) as marked. There were no significant differences by trajectory group for IgM or IgA for any of these antigens (not shown). Squares are means. B. Longitudinal plots of serum IgG (gray) and IgM (red) for these antigens over the course of 28 days show little change with time, although there are spikes for a minority of individuals, especially for IgM. These likely represent cross-reactive responses to SARS-CoV-2. C. Longitudinal trends in serum IgG response to SARS-CoV-2 antigens (NP, RBD, and S-M) in patients with low (black lines) vs. high (red lines) levels of IgG to seasonal coronaviruses at first visit. While there is a trend towards slightly higher starting levels of anti-SARS-CoV-2 IgG in those with low seasonal coronavirus IgG, the overall differences are not significant (P>0.09 for all antigens). All data are detrended for covariates. MFI = median fluorescence intensity.

Figure 3.

Visit 1 serum IgG levels for other respiratory viruses in COVID-19 patients, by clinical trajectory group (1=most mild; 5=fatal). All data are detrended for covariates. MFI = median fluorescence intensity. Squares are means. A. Serum IgG levels for influenza A, parainfluenza 2, and parainfluenza 3 tend to decrease from trajectory group 1 to 5, with statistical significance for the groups marked (vs. group 1). B. Serum IgG levels for influenza strain-specific hemagglutinins also tend to decrease from trajectory group 1 to 5, with statistical significance for the groups marked (vs. group 1). There were no significant differences by trajectory group for IgM or IgA for any of these antigens (not shown).

Figure 4.

Serum antibody levels to other viruses in COVID-19 patients, by clinical trajectory group (1=most mild; 5=fatal). All data are detrended for covariates. MFI = median fluorescence intensity. A. Visit 1 serum IgG levels specific to CMV, EBV, measles, and RSV antigens are not significantly different for any clinical trajectory group (vs. group 1). There were also no significant differences by trajectory group for IgM or IgA for any of these antigens (not shown). Squares are means. B. Longitudinal plots of serum IgG (gray) and IgM (red) for these antigens over the course of 28 days show little change with time, although there are IgM spikes for CMV and EBV for a few individuals.