Binding Antibodies Responses to SARS-COV-2 Infection in Hospitalized Patients and Vaccinated Subjects: A Longitudinal Prospective Observational Study

Abstract

Background:

Prevalence of seropositivity following SARS-CoV-2 infection is vital in evaluating herd immunity. However, depending on illness severity, it remains unclear whether the breadth and magnitude of immune response to SARS-CoV-2 infection is for short or long term.

Objective:

To test the persistence of humoral antibody responses after SARS-CoV-2 exposure in patients with different illness severity and among volunteers who had been vaccinated.

Methods:

This study was conducted in two Saudi Arabian tertiary hospitals. Participants were categorized as critically ill COVID-19 patients, non-critically ill COVID-19 patients, or vaccinated volunteers. We collected demographic data, COVID-19 exposure history, symptoms, vaccination details, and serum samples to analyze antibody persistence. We evaluated SARS-CoV-2 antibody concentrations in COVID-19 patients with varying disease severity and age groups, as well as in BNT162b2-vaccinated individuals, focusing on IgG levels against the S.FL and S1 domains of the spike protein.

Results:

The study included 172 adults: 92 unvaccinated hospitalized COVID-19 patients and 80 vaccinated volunteers. All vaccinated subjects demonstrated seropositivity to the SARS-CoV-2 spike protein, with nearly 80% having a median antibody titer of 13,500 AU/mL. Notably, vaccinated subjects exhibited significantly higher IgG levels than naturally infected patients (P < 0.001), including higher S.FL and S1 titers, regardless of severity. Age, comorbidities, and previous infections influenced S-specific antibody levels. Among hospitalized patients, 58% required intensive care, with 28- and 90-day mortality rates of 23% and 43%, respectively.

Conclusion:

These findings shed light on the immune response dynamics following SARS-CoV-2 infection compared to vaccinated individuals, where the latter showed significantly higher level of antibodies response, providing crucial insights for evaluating short-term herd immunity and the effectiveness of natural infection-induced immunity.

INTRODUCTION

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) has a wide spectrum of severity, with the majority of cases ranging between mild to no symptoms, 15%–20% requiring hospitalization, and about 3%–5% of severe cases requiring admission to critical care, with a 30%–80% mortality for mechanically ventilated patients.[1,2,3,4,5,6,7]

Immunity against coronavirus infection primarily develops through humoral antibody responses targeting the spike (S) protein, with a lesser extent for the nucleocapsid (N)-protein. Previous studies on SARS-CoV-1 and Middle East respiratory syndrome (MERS-CoV) found S-specific antibodies emerging within weeks of infection and persisting for months.[8,9,10,11] Similarly, SARS-CoV-2 studies have shown variable seroconversion in recovered patients likely correlated with symptom severity.[12,13,14] Hains et al. found that in a pediatric dialysis unit, asymptomatic patients had lower and less sustained S-specific antibody responses compared with symptomatic patients.[13] Similarly, a US cohort study observed higher and more durable antibody responses in symptomatic individuals.[14]

In the study by Robbiani et al. that included 149 patients who had recovered from SARS-CoV-2, a significant proportion (79%) of the patients had low S-specific antibody titers (≤1:1,000) at 39 days post-symptom onset.[15] Long et al. reported a median viral shedding duration of 19 days in asymptomatic patients, compared with 6 days in symptomatic patients. Despite this, asymptomatic patients frequently developed S-specific IgG antibodies, with seroconversion rates comparable with symptomatic patients. However, asymptomatic patients often exhibited lower antibody levels during the acute and convalescent phases.[16] Critically ill patients who survived the SARS-CoV-2 infection may have a more robust and potentially longer-lasting immune response compared with those with milder infection. Further research is needed to elucidate the factors influencing the durability of immune responses and their implications for long-term protection against reinfection.[17,18]

The occurrence of seropositivity following SARS-CoV-2 infection might have potential benefits in predicting the validity of herd immunity. It is well established that SARS-CoV-2 infection would have a short- or long-term immune response, depending on the severity of the illness. If the evidence on the persistence and magnitude of antibodies accurately reflects the effectiveness of the immune response post-infection, serodiagnosis could be a valuable tool for identifying individuals at varying risks of infection.

This prospective clinical study aimed to determine the prevalence of seropositivity following infection by assessing S-specific IgG antibodies in hospitalized COVID-19 patients (critically and non-critically ill) and compare them to healthy volunteers who received two doses of the BNT162b2 vaccine.

METHODS

Study design, setting, and participants

This multi-center, prospective observational, longitudinal study included unvaccinated critically and non-critically ill patients with SARS-CoV-2 and healthy volunteers who had received two doses of the BNT162b2 vaccine. This study was approved by the Institutional Review Board of Imam Abdulrahman Bin Faisal University and registered on ClinicalTrials.gov (NCT04520880). All participants/their relatives provided informed consent.

Between August 2020 and August 2021, inpatient adult (aged ≥18 years) patients were recruited from the isolation wards, high dependency units, and intensive care units of two of the largest tertiary academic hospitals in the Eastern Province of Saudi Arabia. For all patients, SARS-CoV-2 infection was confirmed by RT-PCR testing of nasopharyngeal or oropharyngeal swabs. Healthy volunteers were vaccinated healthcare workers from the two hospitals. All volunteers had received two doses of the BNT162b2 vaccine: with the second dose at least 21 days before inclusion in the study. Serum samples were collected from all patients 2 and 8 weeks after the onset of symptom to allow comparison of antibody responses among different groups.

We excluded pregnant, asymptomatic patients who are PCR positive during routine screening upon admission; patients who had received immunoglobulins within 3 preceding months, including COVID-19 convalescent plasma; patients with do not resuscitate orders; and patients with terminal illnesses, regardless of the severity of SARS-CoV-2 infection.

Sample size calculation

Based on the previous study of Whitman et al., the average proportions of seropositive patients using nine different kits for measurement of antibody level >20 days after SARS-COV-2 infection was 72.5%. An a priori power analysis suggested that 158 patients for each group were sufficiently large enough to detect 30% changes in the proportions of seropositive symptomatic and critically ill patients (50.8%), with a type-I error of 0.05 and a power of 80%. An online software (http://powerandsamplesize.com/Calculators/Compare-2-Proportions/2-Sample-Equality) was used for the sample size calculation.

Anticipating a decline in infected cases due to increasing vaccination rates and considering potential loss to follow-up from ICU mortality, long-term complications, or emigration during the pandemic, we estimated a sample size of approximately 80 patients per study arm.

Data collection and variables

Data collected included demographics, clinical characteristics, ICU admission status, complication occurrence and timing, vaccination history, and 28- and 90-day outcomes. Variables included age, weight, and sex. In addition, comorbidities including malignancy, patients who were receiving chemotherapy or immunosuppressive drugs, chronic kidney disease, and end-stage renal failure were included.

Primary study parameter/endpoint

The primary endpoint of this study was to assess the levels of S-specific antibodies elicited within 8 weeks post-symptom onset in SARS-CoV-2 patients. We compared critically ill patients with those with milder illness and healthy vaccinated individuals. Specifically, we analyzed antibody responses to the full-length S (S.FL) protein and the globular domain (S1) in patients with laboratory-confirmed SARS-CoV-2.

Similarly, we correlated different parameters of humoral antibody responses with length of ICU stay and mortality rates. S-specific antibody responses between vaccinated subjects and the patients (ICU vs. non-ICU patients) was compared.

Enzyme-linked immunosorbent assay

ELISA was used to quantify the antibody’s specific binding full-length spike protein (S.FL) and the globular domain (S1) of SARS-CoV-2. Briefly, flat-bottom 96-well plates (Catalog # 442404; Thermo Scientific, Denmark) were coated with 1 μl/ml S.FL (Catalog # 40589-V08B1; Sino Biological, Beijing, China) or S1 (Catalog # 40591-V08B1; Sino Biological, Beijing, China) antigens and were incubated for overnight at 4°C. Using a microplate washer, plates were washed five times with 300 μl of 1× PBS (Catalog# CH-M-077, MoleQule-ON, Auckland, New Zealand). Plates were blocked with 200 μl of 5% skimmed dry milk (Catalog # CH-M-457, Molequle ON, Auckland, New Zealand) in 1× PBS and incubated for 1 hour at room temperature. Following five washes with 300 μl of 1× PBS, 100 μl of 1:500 serially diluted serum was added to each well and incubated for 1 h at room temperature. After another set of washes, horseradish peroxidase (HRP)-conjugated rabbit anti-human IgG secondary antibody (Catalog # PA1-28587; Invitrogen, IL, USA) was diluted at 1:5000 in PBS and 100 μl of this was added to each well, followed by incubation for 1 h at room temperature. After the five final washes with 300 μl of 1× PBS, tetramethylbenzidine substrate (Catalog # ab171524; Abcam, USA) 100 μl was added to all wells. Then, 100 μl of 2M sulfuric acid (Catalog # CAS#7664-93-9; Fisher Scientific, New Jersey, USA) was added to stop the reactions. The optical density values were read at 450 nm as a last step using an ELISA reader. All experiments were performed at least in duplicate.

Data analysis

Continuous data distribution was assessed by visual inspection of histograms. For both arms, the baseline characteristics were expressed as counts and percentages, means and standard deviations (SD), or medians and interquartile ranges (IQR), whenever appropriate. Hypothesis tests were two-sided with a significance level of 0.05. Analyses were performed using the IBM SPSS statistics version 26.

A generalized linear model was used to analyze the primary outcome. Independent student t-tests and Mann–Whitney tests were used for continuous and non-continuous data, respectively. Pre-hoc multivariate analysis examined Pearson’s correlations between seroconversion rate and patient characteristics, hospital and ICU stays, admission status, and 28-and 90-day mortality. Post hoc analysis of the differences in S1 and S.FL antibody responses among the hospitalized patients of the three age groups (youth group: aged 15–47 years; middle-aged: 48–63 years; elderly: ≥64 years)[19] were examined using the Mann–Whitney U test.

Variables were analyzed using a multivariate logistic regression model in a progressively increasing manner to identify the independent predictors for seroconversion rate during the two time-points for measurement of S-specific antibodies, with entry and retention set at a significance level of P <0.05.

RESULTS

A total of 172 participants were enrolled in the study: 92 COVID-19 patients and 80 vaccinated volunteers. Patients were significantly older than volunteers (mean difference: 18.0 years; 95% CI: 14.4–21.7, P < 0.001). Most patients (60%) were symptomatic, primarily with mild or moderate symptoms [Table 1]. Table 2 shows differences in the baseline characteristics between critically and non-critically ill patients.

Table 1.

Study participant’s characteristics

Variables	COVID-19 patients (n=92), n (%)	Vaccinated volunteers (n=80), n (%)	P
Age (years)	56.1±14.71	37.9±9.53	<0.001
BMI (kg/m2)	31 (25.1–39.9)	27.68 (24.7–38.1)
Gender (male/female)	57/35	47/33
Previously tested positive	16 (17.4)	0	<0.001
Asymptomatic	37 (40)
Symptomatic	55 (60)
Mild	7 (13)
Moderate	47 (85)
Severe	1 (2)
Comorbidities
Asthma	10 (11)	0	0.006
Dyspnea	55 (60)	0	2.020
Heart failure	13 (14)	1 (1.2)	0.005
Persistent ventricular tachycardia	10 (11)	0	0.006
Smoking	10 (11)	0	0.006
Diabetes mellitus	48 (52)	1 (1.2)	5.555
Type I	16 (33)	0	0.711
Hypertension	48 (52)	2 (2.4)	2.810
Recent mechanical ventilation within 30 days	10 (11)	0	0.006
Stroke	2 (2.2)	0	0.541
Active hematological malignancy	1 (1)	0	0.944
Receiving chemotherapy	1 (1)	0	0.944
Receiving immunosuppressive therapy	1 (1)	0	0.944
Solid organ transplantation recipient	2 (2.2)	0	0.541
Chronic kidney disease	5 (5.4)	0	0.097
Type of vaccine received
BNT162b2	0	80 (100)
Disease status
Need for hospital admission	92 (100)	0	2.020
Ward	39 (42.4)
ICU	53 (57.6)
ICU stay (days)	5.5 (0–17.8)
Hospital stay (days)	14 (8.0–28.3)

Data are presented as mean±SD, median (IQR), and n (%). ICU – Intensive care unit; COVID-19 – Coronavirus disease of 2019; IQR – Interquartile range; BMI – Body mass index; SD – Standard deviation

Table 2.

Characteristics of critically and noncritically ill COVID-19 patients

Variables	Critical (n=53), n (%)	Noncritical (n=39), n (%)	P
Age (years)	58.6±13.80	52.4±15.40	0.413
Body mass index (kg/m2)	31 (25.8–40.1)	30 (23.9–38.8)	0.377
Gender (male/female)	34/19	23/16
Comorbidities
Asthma	8 (15)	2 (5)	0.238
Heart failure	8 (15)	5 (12.8)	0.995
Persistent ventricular tachycardia	6 (11)	4 (10.3)	0.861
Smoking	5 (9.4)	5 (12.9)	0.911
Diabetes mellitus	26 (49)	22 (56.4)	0.631
Hypertension	32 (60)	16 (41)	0.104
Mechanical ventilation	10 (19)	0	0.011
Active hematological malignancy	1 (2)	0	0.881
Receiving chemotherapy	1 (2)	0	0.881
Receiving immunosuppressive therapy	1 (2)	0	0.881
Solid organ transplantation recipient	1 (2)	1 (2.6)	0.614
End-stage renal disease	5 (9.4)	2 (5.1)	0.711
Dialysis dependent	5 (9.4)	2 (5.1)	0.711

SARS-COV-2 S-specific antibodies

Persistent humoral antibody responses measured by S.FL and S1 antibodies levels, the primary outcomes, were significantly higher in the volunteers than in the hospitalized non-critically ill and critically ill COVID-19 patients [S.FL: 13500 AU/mL [4500–13500 AU/mL] vs. 4500 AU/mL [750–13500 AU/mL] and 4500 AU/mL [500–13500 AU/mL], respectively, P <0.001; and S1; 13500 AU/mL [13500–13500 AU/mL] vs. 4500 AU/mL [2250–13500 AU/mL] and 4500 AU/mL [1500–13500 AU/mL], respectively, P <0.001] [Figures 1-3 and Tables 3–5].

Figure 1.

Comparison of full-length S specific IgG levels between critically and noncritically ill COVID-19 patients

Figure 3.

Comparison of (a) full-length S specific and (b) S1 specific IgG levels between the COVID-19 patients and vaccinated subjects

Table 3.

Comparison of COVID-19 antibody levels and mortality between critically and noncritically ill COVID-19 patients

Primary outcome	Critical (n=53)	Noncritical (n=39)	P
COVID-19 antibody levels (AU/mL)
S.FL	4500 (500–13,500)	4500 (750–13,500)	0.215
S1	4500 (1500–13,500)	4500 (2250–13,500)	0.663
Mortality (days), n (%)
28	12 (22.6)	0	0.004
90	23 (43.4)	0	6.582

S.FL – Full-length S

Table 5.

Comparisons of changes in S1 and full-length S responses in critically and noncritically ill COVID-19 patients

Age groups (years)	S1 antibodies		S.FL antibodies
	SE	P	SE	P
Youth (15–47)	13.31	0.111	13.80	0.808
Middle age (48–63)	50.31	0.796	51.80	0.196
Elderly (≥64)	16.81	0.934	17.13	0.846

SE – Standard error; S.FL – Full-length S

Notably, the non-critically and critically ill COVID-19 patients had comparable persistent immune responses with respect to S.FL (P = 0.215) and S1 (P = 0.663) antibody levels [Tables 3, 4 and Figures 1, 2]. Post-hoc analysis showed non-statistical significance differences in S1 and S.FL antibody responses among critically and non-critically ill youth, middle-aged, and elderly patients [Table 5].

Table 4.

Multivariate analysis of the independent factors for the COVID-19 full-length S antibodies

Variables	SE	95% CI	P
Age	8.6	−115.5–202.8	0.043
Gender	0.44	−1.21–0.52	0.515
Tested positive with RT-PCR before enrolment	1.12	2.51–6.90	<0.001
Asthma	1.14	−3.31–1.21	0.368
Heart failure	1.74	−6.90–−0.10	0.047
Persistent ventricular tachycardia	3.80	−1.71–13.20	0.130
Current smoking	2.10	−7.43–1.51	0.194
Diabetes mellitus	0.72	−2.97–−1.43	0.031
Hypertension	0.73	−2.01–0.84	0.421
Stroke	2.71	−14.50–−3.96	<0.001
Chronic kidney disease	1.92	−30.21–−22.64	<0.001

SE – Standard error; CI – Confidence interval; RT-PCR – Reverse transcription polymerase chain reaction

Figure 2.

Comparison of S1 specific IgG levels between critically and noncritically ill COVID-19 patients

Subgroup analyses

Subgroup analyses demonstrated that compared with volunteers, patients were older (P <0.001) [Table 6]. Compared with volunteers, patients had a higher frequency of comorbidities (including asthma), persistent ventricular tachycardia, smoking, and recent mechanical ventilation (11% vs. 0%, P = 0.006, for each) and heart failure (14% vs. 1.3%, P = 0.005) [Table 1]. In terms of mortality, 22.6% and 43.4% of the critically ill patients died within 28 and 90 days after admission [Table 3].

Table 6.

Comparisons between noncritically and critically ill COVID-19 patients and vaccinated volunteers

Variables	Noncritical (n=39)	Critical (n=53)	Vaccinated volunteers (n=80)	P
Age (years)	52.4±15.40*	58.6±13.80*	37.9±9.53	<0.001
Gender (male/female)	23/16	34/19	47/33	0.107
S.FL antibodies levels (AU/mL)	4500 (750–13,500)*	4500 (500–13,500)	13,500 (4500–13,500)	0.001*
S1 antibodies levels (AU/mL)	4500 (2250–13,500)*	4500 (1500–13,500)*	13,500 (13,500–13,500)	<0.001
Mortality (days), n (%)
28	0	12 (22.6)	0
90	0	23 (43.4)	0

*Statistically significant compared with the vaccinated healthy volunteers. S.FL – Full-length S

Factors affecting seropositivity

Multivariate analysis identified heart failure, stroke, and chronic kidney disease as independent factors for S.FL and hypertension for S1 [Table 4]. In addition, age, having a positive test with RT-PCR before enrolment, and diabetes mellitus were predictors for both S.FL and S1 [Tables 4 and 7].

Table 7.

Multivariate analysis of the independent factors for COVID-19 S1 antibodies

Variables	SE	95% CI	P
Age	56.7	−226.76–−2.50	0.023
Gender	0.71	−1.43–1.20	0.860
Tested positive with RT-PCR before enrolment	2.03	6.31–14.31	<0.001
Asthma	1.50	−321–2.63	0.844
Heart failure	2.44	−2.14–7.41	0.279
Persistent ventricular tachycardia	5.70	−20.10–2.50	0.127
Current smoking	2.74	−10.50–0.31	0.063
Diabetes mellitus	0.90	−3.51–−0.21	0.031
Hypertension	1.11	−4.30–−0.10	0.043
Stroke	4.55	−3.10–14.70	0.200
Chronic kidney disease	5.62	−1.91–7.71	0.231

SE – Standard error; CI – Confidence interval; RT-PCR – Reverse transcription polymerase chain reaction

DISCUSSION

This study found that age was a predictor for both S.FL and S1. This correlates with heart failure, stroke, and chronic kidney disease being identified as independent factors for S.FL and hypertension for S1, all of which are more common with older age and increase the risk of severe COVID-19.[5,6] Notably, 17.4% of COVID-19 patients had prior infections, suggesting potential impact on disease progression. As previous infections can influence long-term immunity and disease outcomes, they should be considered in future studies.

The current study also found that the frequency of asthma was significantly higher among patients compared with volunteers. Respiratory airway inflammation in asthma worsens the condition of a patient with COVID-19, as the disease also suppresses the activity of the airways. In addition, diabetes was an independent predictor for S.FL and S1 levels. In general, comorbidities can affect the robustness of responses to infection, as found in the current study. It was also found that there were significantly more smokers among patients than volunteers. Smoking weakens the natural respiratory defense system, increasing the risk of having severe symptoms of the disease.

The higher level of S.FL and S1 antibodies found in vaccinated individuals compared with patients suggests the humoral antibody response after vaccination, and thus highlights its effectiveness. However, as the patients were older than the volunteers, it might also indicate that age could be a factor that results in lower levels of antibody production even after vaccination.

Immunoglobulin A is an example of S-specific antibodies in respiratory fluids and saliva.[8] The development of IgA occurs very late to assist in the prevention of viral replication in the mucosa. For this reason, vaccines against SARS-CoV-2 should be modeled to include improving cellular immune responses, which contribute to overall protection.[11]

Future studies can be focused on understanding the antibody response kinetics between vaccinated and symptomatic infected patients with different severity level. This could assist in individualizing therapy, and thus in mitigating the impact of COVID-19.[20,21]

Limitations

The study’s limitations include its small sample size, retrospective nature of some variables, potential for selection bias, confounding factors and the lack of a control group. The retrospective aspect of previous positive RT-PCR test results can introduce bias and limitations in data accuracy, thus influencing the reliability of the findings. The study’s sample size was limited due to declining infection rates and the inclusion of previously infected individuals, which may have influenced the severity of subsequent infections. Excluding these patients would have compromised the generalizability of the results.

Despite multivariate analysis, unmeasured confounding factors such as socioeconomic status and lifestyle factors may influence the observed associations. Clinical data can be subject to documentation bias and interpretation variability, while healthcare provider quality and accuracy can affect data consistency. A control group of unvaccinated individuals would have strengthened the study’s findings.

CONCLUSION

The findings of this study, particularly the proportions of patients with positive S-specific antibodies, at 8 weeks post-symptom onset in both ICU and non-ICU patients contribute to the understanding of the short-term humoral immune response to SARS-CoV-2 infection. By measuring S-specific binding antibody levels, the study sought to provide valuable information on the immune response over different time intervals and implications for predicting the short-term herd immunity of the disease. However, the small sample size and inclusion of previously infected individuals limit the generalizability of the findings. Moreover, age, comorbidities, and prior positive RT-PCR test results influenced the antibody levels in hospitalized patients.

Ethical considerations

The study was approved by the Institutional Review Board of Imam Abdulrahman Bin Faisal University Research (Ref no.: IRB-2020-01-236; date: November 2020), Dammam, Saudi Arabia. All study participants/their relatives provided consent before inclusion in the study. The study adhered to the principles of the Declaration of Helsinki, 2013.

Peer review

This article was peer-reviewed by four independent and anonymous reviewers.

Data availability statement

All data produced and analyzed in this study are included in this article in the Tables and Figures.

Author contributions

MSA, IAM, AAN, and LPA conceived the project. LPA, CMM, RA, NC, ZAC, AAN, MAA, FAM, JAM, SMA, MAM, RAS, MKA, TAB, MSL, MAB, SAA, AAS, MAJ, SMA, AAD, and AAT managed the trial (including recruitment and data collection) with support, input and oversight from MSA and IAM. MET, IAM, CMM, LPA and MSA prepared the data and did the statistical analysis, which was interpreted by all other authors. MSA, IAM, MET and LPA wrote the first draft of the manuscript.

All authors contributed to the design of the study and the revision of the manuscript.

All authors have read and agreed to the published version of the manuscript.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

The study was funded by the Deanship of Scientific Research, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia (Grant no.: Covid19-2020-079-Med).