Lack of detrimental effect of corticosteroids on antibody responses to SARS-CoV-2 and viral clearance in patients hospitalized with COVID-19.

Dear Editor,

Recent articles in this Journal have described the beneficial effects of corticosteroids on outcome of COVID-19, and have also suggested that corticosteroids may not impair the natural immune response.1, 2 – Corticosteroids are the only immunomodulatory agents that have shown so far a reduction in mortality in a randomized multicenter trial, and have accordingly been recommended for moderate-to-severe COVID-19. Besides the benefits derived from their potent anti-inflammatory properties, the potential negative consequences of corticosteroids’ immunosuppressive effects on SARS-CoV-2 dynamics remain to be characterized. Particularly, information about their impact on the humoral immune response against the virus is limited. Short courses of corticosteroids have been associated with a decrease in serum IgG and IgA concentrations. Short-term and long-term reduction in antibody production might have negative effects on viral clearance and protection against reinfection. In addition, data regarding the effect of corticosteroids on SARS-CoV-2 clearance remain controversial. ,

We analyzed the longitudinal impact of therapy with corticosteroids on the antibody response to SARS-CoV-2 and viral clearance in patients admitted with COVID-19.

A prospective study was carried out in hospitalized patients with COVID-19 confirmed through real-time polymerase chain reaction. Serial nasopharyngeal and plasma samples were obtained at different time points for SARS-CoV-2 RNA and antibody measurement during hospital stay and after discharge. IgG antibody plasma levels against the SARS-CoV-2 internal nucleo-capsid protein (N-IgG) and surface S1 domain of the spike protein (S-IgG) (Anti-SARS-CoV-2 IgG ELISA, Euroimmun, Lubeck, Germany) were determined. Of 210 adults admitted with COVID-19, 77 participants with positive SARS-CoV-2 RNA in more than one nasopharyngeal sample, more than two plasma samples at least 14 days apart, and who did not receive other sole immunomodulatory agents were selected; of them, 27 received corticosteroids given in daily pulses of 250–500 mg during 3 days. Participants receiving corticosteroids had higher severity of disease and tended to be older (see clinical characteristics in Table 1 ). Median follow-up for antibody detection was 71 (62–83) days. Detectable titers of both S-IgG and N-IgG were observed in 23 (92%) patients receiving corticosteroids and 25 (62.5%) not on corticosteroids (p = 0.009) after a median (Q1-Q3) of 13 (10.5–14.5) days vs 16 (13–24) days from symptom onset, respectively, for S-IgG (p = 0.008); and of 10 (9–13) days vs 14 (8–21) days, respectively, for N-IgG (p = 0.043). Kaplan-Meier curves showed a higher cumulative proportion of patients with detectable S-IgG (p<0.001) and N-IgG (p = 0.012) levels among those receiving corticosteroids. After Cox regression adjustment for the significant variables associated with S-IgG and N-IgG response in the univariate analysis (specifically, SARS-CoV-2 viral load, Charlson comorbidity index and C-reactive protein levels), no significant differences in antibody response were observed between the two groups (Fig. 1 A and B). Median (Q1-Q3) peak S-IgG titers were 6.5 (5.4–7.4) vs 4.5 (0.1–6.5) absorbance/cut-off (S/CO) in patients with and without corticosteroids, respectively, (p = 0.005), and 4.9 (4.0–5.4) vs 3.8 (0.1–5.5) S/CO for N-IgG, respectively (p = 0.037). Temporal changes in S-IgG and N-IgG titers analyzed with local polynomial regression did not differ according to corticosteroid therapy group (Fig. 1C). SARS-CoV-2 viral clearance occurred in 21 (77.8%) patients receiving corticosteroids and 44 (88%) not on corticosteroids after a median (Q1-Q3) of 30 (22–46) days from the first positive sample (p = 0.325). Kaplan–Meier curves exhibiting the probability of SARS-CoV-2 clearance by treatment group are shown in Fig. 1D.

Table 1

Clinical data of patients admitted with COVID-19 confirmed with real-time polymerase chain reaction.

Variable	Corticosteroids N = 27	Non-corticosteroids N = 50	P
Sex, male	18 (66.7)	25 (50.0)	0.229
Age, years	71 (58–81.5)	63.5 (46.8–74.0)	0.059
Active smoking	17 (70.8)	27 (57.4)	0.311
Charlson comorbidity index	3.0 (1.0–5.5)	3 (1–5)	0.490
Days from symptom onset to admission	7 (3–10)	6.5 (3–11)	0.797
SOFA score on admission	3 (2–3)	2 (2–3)	0.035
SpO2/FIO2 on admission	344.6 (321.4–350)	353.6 (343.8–380.8)	0.035
SARS-CoV-2 RNA, copies/sample	3.9 (3.4–4.4)	2.2 (2.0–3.7)	<0.001
Peak S-IgG, S/CO	6.5 (5.4–7.4)	4.5 (0.1–6.5)	0.005
Peak N-IgG, S/CO	4.9 (4.0–5.4)	3.8 (0.1–5.5)	0.037
Interleukin-6, pg/mL	35.3 (17.4–97)	13.4 (8–29.8)	0.021
Ferritin, ng/mL	299.5 (190–640)	180.5 (115.5–333)	0.056
C-reactive protein, mg/L	80.1 (35.1–141.7)	34.5 (4.9–53.5)	0.001
Fibrinogen, mg/dL	614 (429.7–851.3)	443 (323.1–552.5)	0.028
Lymphocytes, x103/μL	1.0 (0.8–1.2)	1.4 (1.2–2.1)	<0.001
Hospital stay, days	19 (13.5–24.5)	9 (6–12)	<0.001
Death	1 (3.7)	1 (2.0)	1
ICU admission	2 (7.4)	4 (8.0)	1
HCQ-based combinations	27 (100.0)	49 (98.0)	1
Azithromycin	27 (100.0)	44 (88.0)	0.085
Lopinavir/ritonavir	26 (96.3)	39 (78.0)	0.047
Remdesivir	0	1 (2)	1
Interferon-β−1b	2 (7.4)	7 (14.0)	0.481
Concomitant tocilizumab use	24 (88.9)	0	<0.001

Categorical variables are expressed as no. and (%), and continuous variables as median (Q1-Q3). Mann-Whitney-Wilcoxon test was used to compare continuous variables, and Fisher's exact test to compare categorical variables. SOFA, Sequential Organ Failure Assessment; TCZ, tociluzumab; SpO2/FIO2, peripheral blood oxygen saturation/fraction of inspired oxygen rate; S/CO, absorbance/cut-off; ICU, Intensive Care Unit; HCQ, hydroxychloroquine.

Fig. 1

Effects of corticosteroids on antibody responses and viral clearance. A, Adjusted Kaplan Meier curve to estimate the cumulative proportion of patients with negative titers of S-IgG according to therapy with corticosteroids. B, Adjusted Kaplan Meier curve to estimate the cumulative proportion of patients with negative titers of N-IgG according to therapy with corticosteroids. C, Temporal changes in N-IgG titers (left) and S-IgG titers (right) analyzed with local polynomial regression. D, Kaplan Meier curve to estimate the cumulative proportion of patients with detectable viral RNA according to therapy with corticosteroids.

In contrast to other studies, we analyzed the effects of corticosteroids on both viral kinetics and the humoral immune response to SARS-CoV-2. We did not find a detrimental effect of corticosteroid pulses on the intensity and duration of antibody responses, and the same was observed with time to viral clearance in this cohort of patients who were thoroughly investigated with multiple sequential samples. Although differences in antibody response even favored patients receiving corticosteroids, probably due to their higher severity of disease and initial SARS-CoV-2 viral load, such differences vanished after adjustment. Remarkably, most patients on corticosteroids also received anti-interleukin-6 (IL-6) therapy with tocilizumab. Our results show that, even in combination with IL-6 blocking agents, corticosteroids do not negatively impact viral clearance or the humoral immune response against SARS-CoV-2. Further carefully designed studies are warranted to confirm these findings.

Funding sources

No funding for this work.

Declaration of Competing Interest

All authors: no conflict of interest.