Immunoglobulin G production in COVID-19 - associations with age, outcome, viral persistence, inflammation and pro-thrombotic markers

Abstract

Age represents the major risk factor for fatal disease outcome in coronavirus disease (COVID-19) due to age-related changes in immune responses. On the one hand lymphocyte counts continuously decline with advancing age, on the other hand somatic hyper-mutations of B-lymphocytes and levels of class-switched antibodies diminish, resulting in lower neutralizing antibody titers. To date the impact of age on immunoglobulin G (IgG) production in response to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection is unknown. Therefore, we investigated the impact of age on the onset of IgG production and its association with outcome, viral persistence, inflammatory and thrombotic markers in consecutive, hospitalized COVID-19 patients admitted to the Clinic Favoriten (Vienna, Austria) between April and October 2020 that fulfilled predefined inclusion criteria. Three different IgGs against SARS-CoV-2 (spike protein S1, nucleocapsid (NC), and the spike protein receptor binding domain (RBD)) were monitored in plasma of 97 patients upon admission and three times within the first week followed by weekly assessment during their entire hospital stay. We analyzed the association of clinical parameters including C-reactive protein (CRP), D-dimer levels and platelet count as well as viral persistence with the onset and concentration of different anti-SARS-CoV-2 specific IgGs. Our data demonstrate that in older individuals anti-SARS-CoV-2 IgG production increases earlier after symptom onset and that deceased patients have the highest amount of antibodies against SARS-CoV-2 whereas intensive care unit (ICU) survivors have the lowest titers. In addition, anti-SARS-CoV-2 IgG concentrations are not associated with curtailed viral infectivity, inflammatory or thrombotic markers, suggesting that not only serological memory but also other adaptive immune responses are involved in successful viral killing and protection against a severe COVID-19 infection.

Introduction

The coronavirus disease (COVID-19) pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has led to severe morbidity and mortality worldwide with more than 630 million registered infections and around 6.57 million deaths by October 2022 (https://www.worldometers.info/coronavirus/), particularly in adults above the age of 60 and individuals with age-related metabolic and cardiovascular diseases [1], [2]. The dramatic shift in the COVID-19 risk profile across age groups depends on multiple factors involving innate and adaptive immune responses [3], [4]. The dynamics and the nature of immune responses to SARS-CoV-2 infection are still incompletely understood but have substantial impact on identifying vulnerable subpopulations with an elevated risk for adverse outcome (either mortality or intensive care unit (ICU) admission), testing and therapeutic strategies as well as public health issues worldwide.

Several studies revealed that poor adaptive immunity to SARS-CoV-2 represents a critical risk factor for the elderly [5], [6], [7], [8]. Adaptive immunity in response to SARS-CoV-2 has been a major scientific focus, as serological memory provides the basis for current vaccination strategies. Adaptive immunity critically controls viral infections via three broad arms: while B-lymphocytes produce neutralizing antibodies, T-lymphocytes produce antiviral cytokines, with CD8 + T cells directly killing infected cells and CD4 + T cells aiding B-lymphocytes in antibody production. With age, humoral immune responses decline as somatic hyper-mutations of B-lymphocytes and plasma levels of class-switched antibodies diminish, resulting in lower neutralizing antibody titers [6]. Further, defective germinal center formation caused by elevated levels of tumor necrosis factor (TNF), which are associated with age as well as severe SARS-CoV-2 infections, diminished CD40 ligand (CD40L) expression and lymphocytopenia in aged individuals contribute to impaired B cell responses in this vulnerable population [6], [7].

Limited pre-existing immunity against SARS-CoV-2 and emerging mutations cause persistent and explosive increases in COVID-19 cases worldwide. Virus-specific antibody responses showed that IgM is consistently detected before IgG, which peaks between three to seven weeks post-symptom onset and persists for at least eight weeks. Neutralizing antibodies are detectable within seven to 15 days of disease onset and levels further increase until days 14–22, before plateauing and subsequently decreasing [9]. Similar to B cell immunity, T cell immunity develops over a period of at least 10–20 days post-symptom onset.

As age represents the major risk factor for adverse outcome in COVID-19 and immune responses are subject to age-related changes like immunosenescence and inflammaging [10], we aimed at investigating the impact of age on the onset of IgG production and its association with outcome, viral persistence, inflammatory and thrombotic markers in hospitalized COVID-19 patients.

Material and methods

Participants and inclusion criteria

This clinical study (ClinicalTrials.gov NCT04351724) represents a sub-study of previously published cohorts [11], [12], [13] and was conducted under approval of the local ethics committee (EK1315/2020) according to the Declaration of Helsinki. All symptomatic COVID-19 patients admitted to the Clinic Favoriten (Vienna, Austria) between April 17th and October 28th, 2020, participated and gave informed consent. A medical history was taken at the day of hospital admission. COVID-19 severity was classified according to the World Health Organization (WHO) scores into mild, moderate, severe and critical at admission [14]. Clinical outcomes including uncomplicated disease, ICU requirement and death were documented up to 21 days after symptom onset. The inclusion criteria were confirmed SARS-CoV-2 infection by real-time PCR of naso- or oropharyngeal swab, onset of symptoms ≤ 5 days and ≥ 18 years of age. Exclusion criteria were severe comorbidities (late stage cancer patients or terminally-ill patients), pregnancy or breast-feeding and anemia (hemoglobin < 11 g/dl). While Austria was hit hard by the COVID-19 pandemic, triage was never necessary due to the high hospital bed per inhabitant ratio, allowing for earlier and unrestricted admission to hospitals, resulting in a broad distribution in patient age and disease severity. Early hospital admittance allowed for analysis of neutralizing antibodies at a very early stage to determine the impact of three different SARS-CoV-2 specific IgGs on various aspects of disease progression. We assessed patients as early as possible and monitored them during their entire hospital stay.

Sample collection

Clinical and laboratory parameters including platelet count and C-reactive protein (CRP) were assessed at hospital admission or within 72 h after admission and monitored over the entire hospital stay (full list of parameters listed in Table 1). D-Dimer levels were recorded daily during the entire hospital stay or up to 30 days in cases of longer hospitalization. Blood samples for IgG quantification were collected at the day of hospital admission (day 0) and every 2–3 days during the first week followed by weekly measurements over the entire hospital stay or in cases of longer hospitalization up to 21 days.

Table 1.

Statistical analyzes of clinical data recorded at hospital admission according to outcome.

	Missing Data	All (n = 97)	Uncomplicated (n = 50)	ICU (n = 34)	Death (n = 13)
Parameter	n	n (%) Median (IQR)	n (%) Median (IQR)	n (%) Median (IQR)	n (%) Median (IQR)	p-Value*
Sex						0.740
Male	-	28 (28.9)	14 (28.0)	9 (26.5)	5 (38.5)
Female	-	69 (71.1)	36 (72.0)	25 (73.5)	8 (61.5)
Age (years)	-	66 (52.00–77.00)	69 (51.25–76.00)	61 (49.75–76.25)	78 (67.00–81.00)	0.062
Comorbidities
Current smoker	25	6 (8.3)	3 (9.1)	3 (10.7)	0 (0.0)	0.733
Obesity (BMI > 25)	7	51 (56.7)	24 (54.5)	20 (58.8)	7 (58.3)	0.924
Diabetes type II	-	22 (22.7)	11 (22.0)	7 (20.6)	4 (30.8)	0.795
Hypertension	1	59 (61.5)	28 (57.1)	20 (58.8)	11 (84.6)	0.180
Cardiovascular disease (any)	-	19 (19.6)	11 (22.0)	4 (11.8)	4 (30.8)	0.284
Coronary heart disease	-	10 (10.3)	5 (10.0)	2 (5.9)	3 (23.1)	0.258
Chronic heart failure	-	6 (6.2)	4 (8.0)	1 (2.9)	1 (7.7)	0.759
Atrial fibrillation	-	6 (6.2)	3 (6.0)	1 (2.9)	2 (15.4)	0.352
Peripheral arterial disease	-	5 (5.2)	3 (6.0)	0 (0.0)	2 (15.4)	0.130
Chronic obstructive pulmonary disease	-	11 (11.3)	7 (14.0)	3 (8.8)	1 (7.7)	0.675
Asthma	-	4 (4.1)	2 (4.0)	0 (0.0)	2 (15.4)	0.061
Chronic renal insufficiency	-	8 (8.2)	7 (14.0)	0 (0.0)	1 (7.7)	0.061
Chronic liver disease	-	5 (5.2)	2 (4.0)	2 (5.9)	1 (7.7)	> 0.999
Malignancy	-	10 (10.3)	4 (8.0)	4 (11.8)	2 (15.4)	0.722
Medication
Anti-platelet therapy	-	20 (20.6)	9 (18.0)	6 (17.6)	5 (38.5)	0.250
Anticoagulation therapy	-	24 (24.7)	13 (26.0)	9 (26.5)	2 (15.4)	0.687
Dexamethasone	-	31 (32.0)	17 (34.0)	12 (35.3)	2 (15.4)	0.411
COVID-19 classification at admission †	1					0.064
Asymptomatic / mild		12 (12.5)	6 (12.0)	4 (12.1)	2 (15.4)
Moderate		48 (50.0)	26 (52.0)	14 (42.4)	8 (61.5)
Severe		24 (25.0)	16 (32.0)	6 (18.2)	2 (15.4)
Critical		12 (12.5)	2 (4.0)	9 (27.3)	1 (7.7)
Clinical Characteristics
Total hospitalization (days)	-	15.0 (11.0–23.0)	13.0 (10.0–21.3)	18.0 (12.8–28.0)	14.0 (10.0–24.0)	0.023#
Invasive ventilation	-	11 (11.3)	0 (0.0)	9 (26.5)	2 (15.4)	0.001#
Laboratory Parameters
Hemoglobin (g/dL)	1	13.6 (12.5–14.7)	13.4 (12.4–14.7)	13.8 (12.8–14.8)	13.0 (11.5–14.1)	0.261
Red blood cell count (x1012/L)	1	4.7 (4.3–5.1)	4.8 (4.3–5.1)	4.8 (4.4–5.1)	4.2 (3.7–4.9)	0.091
Platelet count (x109/L)	1	184.5 (145.3–238.0)	193.0 (151.0–235.5)	193.0 (137.0–224.0)	174.0 (139.5–231.5)	0.707
Leukocyte count (x109/L)	1	6.1 (4.7–8.2)	6.15 (4.4–7.3)	6.1 (4.8–8.9)	6.8 (4.7–10.3)	0.375
Lymphocyte count (x109/L)	8	0.8 (0.6–1.2)	1.0 (0.7–1.3)	0.8 (0.6–1.1)	0.6 (0.5–0.7)	0.008#
Neutrophil count (x109/L)	8	4.6 (3.2–6.9)	4.5 (3.1–6.0)	4.9 (3.5–7.5)	5.4 (3.3–8.2)	0.378
Monocyte count (x109/L)	9	0.3 (0.2–0.5)	0.3 (0.2–0.5)	0.3 (0.2–0.4)	0.3 (0.2–0.4)	0.814
Eosinophil count (x109/L)	9	0.01 (0.00–0.03)	0.01 (0.00–0.06)	0.01 (0.00–0.02)	0.03 (0.01–0.06)	0.212
Basophil count (x109/L)	9	0.02 (0.01–0.04)	0.03 (0.01–0.04)	0.02 (0.01–0.04)	0.03 (0.01–0.10)	0.621
C-reactive protein (mg/L)	1	74.1 (40.8–111.8)	59.5 (29.5–92.3)	83.8 (51.8–120.2)	74.6 (41.7–146.1)	0.101
D-dimer (mg/L)	20	0.7 (0.5–1.0)	0.7 (0.5–1.0)	0.7 (0.5–0.9)	0.7 (0.5–0.9)	0.952
Prothrombin time (%)	5	103.2 (92.1–112.1)	103.9 (94.7–112.9)	100.7 (87.5–108.2)	100.2 (92.2–110.5)	0.299
International normalized ratio	7	1.00 (1.00–1.05)	1.00 (1.00–1.04)	1.0 (1.00–1.10)	1.00 (1.00–1.03)	0.449
Activated partial thromboplastin time (s)	11	32.2 (28.8–35.4)	32.9 (29.3–37.1)	32.0 (28.4–33.8)	29.8 (26.4–31.6)	0.108

† COVID-19 classification according to the guidelines issued by the World Health Organization in mild (fever <38 °C, no dyspnea, no pneumonia), moderate (fever, respiratory symptoms, pneumonia), severe (respiratory distress with respiratory rate ≥30 per minute, oxygen saturation < 93% at rest) and critical (respiratory failure with requirement of mechanical ventilation, requirement of ICU)

^# Post hoc analysis with Bonferroni correction for multiple comparison: Total hospitalization (days): uncomplicated vs ICU p = 0.020, ICU vs death p = 0.357, uncomplicated vs death p > 0.999; Invasive ventilation: uncomplicated vs ICU p = p < 0.001, ICU vs death p = 0.702, uncomplicated vs death p = 0.040; Lymphocyte count (x10⁹/L): uncomplicated vs ICU p = 0.572, ICU vs death p = 0.134, uncomplicated vs death p = 0.007.

BMI: body mass index; ICU: intensive care unit; IQR: interquartile range.

^*p < 0.05. Nominal variables were compared using the Chi-square test; if one of the expected frequencies was less than five, Fisher´s exact test of independence was applied. Metric data were compared using the Kruskal-Wallis-Test. The Kolmogorov-Smirnov Test was chosen to assess normality.

Sample preparation and anti-SARS-CoV-2 IgG quantification

Blood was drawn into vacutainer tubes containing citrate and plasma generated by centrifugation at 1000 x g for 10 min at 4 °C, followed by a second centrifugation of the supernatant at 10,000 x g for 10 min at 4 °C [15]. Plasma samples were aliquoted and stored at − 80 °C until further use.

Antibodies against the SARS-CoV-2 spike protein S1, nucleocapsid (NC)-, and the Receptor Binding Domain (RBD)-protein were determined by a LEGENDplex™ bead-based immunoassay kit (BioLegend, United States) according to the manufacturer´s instructions. Briefly, plasma was incubated with capture beads for 2 h while mixing. After washing, samples were incubated with biotinylated detection antibodies for 1 h forming capture bead-analyte-detection antibody complexes. Subsequently, streptavidin-phycoerythrin (SA-PE) was added for 30 min providing a fluorescent signal. Samples were measured on a Cytoflex S cytometer within 3 h and analyzed using LEGENDPlex v8.0 software (https://www.biolegend.com/en-us/legendplex). Absolute concentrations of anti-spike S1, anti-NC and anti-RBD IgGs (ng/ml) were assessed via a standard (BioLegend) concentration curve.

Statistical analysis

Statistical analyzes were performed with IBM SPSS Statistics 28, diagrams were generated with GraphPad Prism 8 and Adobe Illustrator CS6 16.0.0.

Anti-SARS-CoV-2 IgGs were analyzed according to the clinical outcome or patient age. For the IgG analysis according to age the data were divided into two groups, below 75 and 75 years of age and above because the risk of severe disease is elevated in patients above 60–70 and in our cohort COVID-19 patients with fatal outcome were predominantly in the group of patients with 75 years of age and above [16].

Nominal variables were compared using the Chi-square test; if at least one expected frequency in a fourfold table was less than five, Fisher´s exact test of independence was applied. Metric data were compared using the Kruksal-Wallis-Test. The Kolmogorov-Smirnov Test was chosen to assess normality of data (Table 1). Post-hoc analyzes were performed with Bonferroni correction for multiple comparisons. Mean antibody concentrations of the two age groups were compared by (2-sided) Student´s T-test and the duration of viral persistence between age and outcome groups were analyzed by the Mann-Whitney U-Test or the Kruskal-Wallis-Test. P-values ≤ 0.05 were considered as statistically significant. Only two-sided tests were used.

Due to high complexity of clinical data (Supplementary Figure 1) linear mixed models were applied to explore if anti-SARS-CoV-2 IgG titers develop differently over the disease course in COVID-19 patients below and above 75 years of age and whether this development differs between the outcome groups.

To stabilize the residual distribution, all outcomes were log-transformed prior to modelling. The predictors ‘age’ in years and ‘days after symptom onset’ were used as continuous covariates. For each of these predictors an additional quadratic term was included to allow non-linear effects. ‘Outcome’ was included as a factor with three levels, namely uncomplicated, ICU or death. Each patient was included as level of a random factor allowing a random intercept for each patient.

At first, a mixed linear model was calculated with all main effects (outcome, age, days after symptom onset) and all possible interactions including ‘age’, ‘days after symptom onset’ and ‘outcome’ (i.e. 3 main effects, 3 two-way interactions and the 3-way interaction) plus two quadratic terms of ‘age’ and ‘days after symptom onset’. Next, non-significant interactions were omitted from the model beginning with the 3-way interaction, followed by two-way interactions beginning with the one having the highest P-value. This procedure was performed until only significant interactions or only main effects were left. Least-square means (with 95% confidence intervals) were computed to allow visualization of the final models including the uncertainty of estimates. Before plotting, estimated values were back-transformed which allows plotting on the original scale.

Results

Our cohort comprised of 97 COVID-19 patients between 18 and 89 years of age (50 patients with uncomplicated disease, 34 patients requiring intensive care and 13 with fatal outcome, Fig. 1A). As demonstrated in Table 1, no significant differences were observed in terms of sex distribution, smoking status or other comorbidities between the different outcome groups. However, patients with fatal disease showed a strong reduction of lymphocytes comprising T, B and NK-cells compared to patients without complications (uncomplicated vs. death p = 0.007, uncomplicated vs. ICU p = 0.572, ICU vs. death p = 0.134) (Table 1).

Fig. 1.

IgG levels against SARS-CoV-2 spike S1, nucleocapsid (NC) and receptor binding domain (RBD) protein in COVID-19 patients below and above 75 years of age: A) Overview of the study design (green: uncomplicated disease, orange: requiring intensive care and red: fatal outcome), B) modelling of anti-spike S1 IgGs in patients below and above 75 years of age over the first week post symptom onset, C) modelled anti-spike S1 IgG levels in COVID-19 patients below and above the age of 75 years recorded over the first, second and third week post symptom onset according to the clinical outcomes.

Mixed linear model analyzes of our patient cohort indicate that there is a significantly delayed onset of antibody production against the NC-protein (Fig. 1B, middle) (main effect of age p = 0.001) and RBD-protein (Fig. 1B, right) (main effect of age p = 0.003) in patients below 75 years compared to individuals of at least 75 years of age. However, there is no evidence for a delayed onset of anti-spike S1 IgG production in patients below 75 years of age (Fig. 1B, left) (main effect of age p = 0.125).

When we analyzed the impact of age on anti-spike S1 IgG levels over the first, second and third week post symptom onset separately according to the clinical outcome groups, we could see that with increasing age there was a small tendency towards lower antibody titers. We found that patients with uncomplicated disease above the age of 75 had significantly lower concentrations of IgGs in the first week post symptom onset (p = 0.024). In the second (p = 0.298) or third week (p = 0.077) post symptom onset the mean antibody titers against the spike S1 protein did not differ between younger and older individuals. Younger and older patients admitted to the ICU had comparable concentrations of anti-spike S1 IgGs in the first (p = 0.351) or third (p = 0.071) week, but a significant difference was detected in the second week (p = 0.032) post symptom onset. In addition, antibody concentrations of deceased patients differed neither in the first week (p = 0.860), nor in the second week (p = 0.370) post symptom onset (Fig. 1C).

Next, we analyzed outcome specific differences of anti-SARS-CoV-2 IgGs including anti-spike S1, anti-NC and anti-RBD IgGs over the disease progression. Modelling of anti-spike S1, anti-NC and anti-RBD IgGs revealed that patients with fatal disease had the highest concentrations of antibodies against SARS-CoV-2, followed by patients with uncomplicated disease and those requiring intensive care ( Fig. 2A). Levels of anti-spike S1 IgGs were quite similar between deceased patients and patients with uncomplicated disease. However, the concentration of anti-spike S1 IgGs of ICU patients was significantly reduced compared to patients with uncomplicated disease (Fig. 2A, left; main effect of time p < 0.001, main effect of age p = 0.024, main effect of outcome p = 0.040, contrasts: uncomplicated vs ICU p = 0.015, uncomplicated vs death p = 0.932, ICU vs death p = 0.091) (Fig. 2A left, Fig. 2B). No apparent outcome specific differences could be observed in antibody concentrations against the NC-protein (main effect of time p < 0.001, main effect of age p = 0.392, main effect of outcome p = 0.835, contrasts: uncomplicated vs. ICU p = 0.584, uncomplicated vs death p = 0.916, ICU vs death p = 0.669) (Fig. 2A middle, Fig. 2C). In turn, also the concentration of anti-RBD IgGs of patients requiring intensive care was highly reduced compared to patients without complications (main effect of time p < 0.001, main effect of age p = 0.026, main effect of outcome p = 0.157, contrasts: uncomplicated vs. ICU p = 0.060, uncomplicated vs death p = 0.899, ICU vs death p = 0.258) (Fig. 2A right, Fig. 2D).

Fig. 2.

Anti-spike S1, anti-NC and anti-RBD IgG levels in COVID-19 patients with uncomplicated disease, requiring intensive care and fatal outcome over the course of disease: A) modelling of IgG levels against the spike S1-, NC- and RBD-protein over 21 days post symptom onset, visualization of modelled B) anti-spike S1, C) anti-NC and D) anti-RBD IgGs over the first, second and third week post symptom onset according to the three clinical outcome groups.

Next, we investigated if there was an association between anti-SARS-CoV-2 antibody titers and viral persistence. The time until patients first tested negative for SARS-CoV-2 was neither associated with outcome nor with age ( Fig. 3A). Partial regression analyzes of anti-spike S1 IgG titers in the first, second and third week after symptom onset and viral persistence revealed that the time until patients were PCR negative for SARS-CoV-2 was independent of anti-spike S1 IgGs (Fig. 3B).

Fig. 3.

Lack of association of anti-spike S1 IgG levels with the first negative PCR result post symptom onset, age and clinical outcome: A) first negative PCR result post symptom onset of patients below and above 75 years of age according to the clinical outcome groups including uncomplicated disease, requiring intensive care and fatal disease, B) partial regression analysis of anti-spike S1 IgG levels and days until patient’s first negative SARS-CoV-2 PCR result post symptom onset assessed for the first, second and third week post symptom onset. Red dots indicate deceased patients.

Finally, we performed partial regression analyzes between platelet counts, D-Dimer or CRP levels and antibodies against SARS-CoV-2. Very strong positive correlations were found between platelet counts and anti-spike S1 antibody concentrations ( Fig. 4A, left, R = 0.763, p < 0.001) as well as anti-RBD antibodies (Fig. 4A, right, R = 0.722, p = 0.001). Also the correlation coefficient between platelet counts and antibody concentrations against the NC-protein were strong (Fig. 4A middle, R = 0.600, p = 0.008). In contrast, D-Dimer levels (Fig. 4B) and CRP concentrations (Fig. 4C) did not correlate with antibodies against the spike S1-, NC- or RBD-protein.

Fig. 4.

Partial regression analyzes of anti-SARS-CoV-2 IgG levels and platelet counts, D-Dimer or CRP levels over the first week post symptom onset: Regression analyzes of IgG levels against the spike S1-, NC- or RBD-protein and the A) platelet count, B) D-Dimer or C) CRP levels. Red dots indicate deceased patients.

Discussion

Our data demonstrate that in older individuals anti-SARS-CoV-2 IgG production increases earlier after symptom onset and that deceased patients had the highest amount of antibodies against SARS-CoV-2 whereas patients at the ICU had the lowest titers. In addition, anti-SARS-CoV-2 IgG concentrations were not associated with curtailed viral infectivity, determined by PCR for SARS-CoV-2 or the inflammatory marker CRP. When we analyzed the association of SARS-CoV-2 IgGs with thrombotic markers, we found no association with D-dimer, as a marker for an activated coagulation cascade, but a strong positive correlation of all three SARS-CoV-2 IgGs and platelet counts.

While older individuals with COVID-19 have an elevated risk of adverse outcome [17], [18], to date little is known about the production and effectiveness of SARS-CoV-2 specific antibodies in COVID-19 patients above the age of 60. Contrary to our expectations, we observed that independent of clinical outcome the onset of antibody production against SARS-CoV-2 including anti-NC and anti-RBD IgGs was detected significantly earlier in patients above 75 years compared to patients below 75 years of age, suggesting that antibody production against SARS-CoV-2 is not compromised in patients with advanced age. In comparison, IgG production against the spike S1 protein was also initiated earlier in elderly individuals, however differences were not statistically significant. Previous studies suggest that patients with severe disease respond with higher antibody titers compared to patients with mild or moderate course of disease [8]. It is hypothesized that production of IgGs with an afucosylated fragment crystallizable region (Fc) tail during acute SARS-CoV-2 infection is associated with severe COVID-19, as they bear an enhanced pro-inflammatory activity via FcγRIIIa signaling, resulting in an exaggerated response of innate immune cells [19]. However, older individuals in our cohort did not show stronger symptoms or elevated markers of inflammation such as CRP. Our observations might be explained by the hypothesis that cross-reactive antibodies and B-cells derived of prior coronavirus infections causing common, non-severe colds, lead to elevated IgGs in elderly patients [20].

Our data further demonstrate that patients with fatal disease had the highest concentration of anti-spike S1, anti-NC and anti-RBD IgGs, which is in line with previous reports [21], [22], [23], [24]. However, in contrast to earlier findings, our data indicate that patients admitted to the ICU had lower anti-spike S1, anti-NC and anti-RBD IgG levels than patients without disease complications. Of note, in our cohort deceased patients were almost exclusively above the age of 60. It has previously been reported that innate and adaptive cellular response are impaired in older individuals facilitating uncontrolled viral entry and reproduction [25], followed by an overproduction of pro-inflammatory mediators [11], [22] potentially resulting in a stronger antibody production in severely ill patients.

Our findings indicate that high anti-SARS-CoV-2 antibody titers at the onset of disease might be predictive for an adverse outcome in patients above 75 years. In line with previous reports, this demonstrates that antibodies alone are not protective against severe disease during acute SARS-CoV-2 infection. Instead adaptive immunity is hypothesized to be associated with a mild disease and important to control a SARS-CoV-2 infection [8]. This could explain why patients admitted to the ICU, who had the lowest amount of antibodies against SARS-CoV-2 still survived the infection. That antibodies do not necessarily protect SARS-CoV-2 patients against severe COVID-19 symptoms could also explain why we could not find any correlations between anti-spike S1 IgGs and the duration of infectivity.

Finally, we found strong positive correlations between platelet counts and anti-spike S1, anti-NC and anti-RBD IgGs. Severe COVID-19 is often accompanied by thrombocytopenia caused by uncontrolled platelet activation, thrombosis or dysregulated megakaryocyte maturation [26]. Platelets can respond early and quickly to pathogens and inflammatory mediators and boost immune responses by binding innate immune cells. Formation of platelet-leukocyte or platelet-lymphocyte aggregates facilitates immune cell trafficking, tissue infiltration and secretion of immune modulators, which are needed for a rapid activation of adaptive immunity and antibody production [27]. Thus, it is hypothesized that a continuously replenished platelet population, which compensates platelet consumption during infection is key for the establishment of adaptive immunity [28]. In contrast, D-Dimer is a surrogate marker for an activated coagulation cascade and the risk for thrombosis. However, we could not find any association between D-Dimer levels and antibody titers.

Notably, our study has certain limitations. Due to the single-center character of the study, which only includes Caucasian individuals, we cannot exclude that our data only reflect the situation in Austria and the virus mutations present. We want to point out that the limited sample size did not allow for investigation of sex-specific differences and while we tried to take various possible confounders, for example comorbidities, into account, no subset analyses were possible due to small sample sizes. Therefore, the character of our study is hypothesis-generating and further studies on larger cohorts are warranted.

Taken together, our data demonstrate that while older individuals show an earlier onset of anti-SARS-CoV-2 IgGs, these IgGs are not associated with curtailed viral persistence, inflammatory or thrombotic markers.

Funding

This work is part of the ACOVACT study of the Medical University of Vienna and is financially supported by the Austrian Federal Ministry of Education, Science and Research, the Medi-cal-Scientific Fund of the Mayor of Vienna (COVID024) and the Austrian Science Fund (P-34783, P-32064; SFB-54).

CRediT authorship contribution statement

A.P., W.C.S. and A.S. collected and analyzed the data, performed flow cytometry analyzes, prepared the figures and wrote the manuscript; S.H., performed statistical analyzes and wrote the manuscript; S.T., D.P., J.S., E.P., M.T., C.S., T.S. and M.K., collected and analyzed the data; B.J., U.R. and A.Z., provided resources and interpreted the data; A.A. conceived the study, analyzed and interpreted the data and wrote the manuscript. All authors have read and agreed to the published version of the manuscript.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Appendix A. Supplementary material

Supplementary material.

.