Effectiveness and immunogenicity of SARS-CoV-2 booster vaccine in immunosuppressed systemic autoimmune disease patients: A prospective study

Abstract

Introduction and objectives

Patients with systemic autoimmune rheumatic disease (SARD) are a vulnerable population for severe COVID-19 and worse response to vaccination, prompting the need of a booster vaccine. Data regarding its response is limited and inconsistent. The aim of this study was to assess the effectiveness and immunogenicity of the third dose of the SARS-CoV-2 vaccine in immunosuppressed SARD patients.

Materials and methods

We conducted a prospective study in immunosuppressed SARD Portuguese patients, who received a SARS-CoV-2 booster vaccine, from October 2021 to August 2022. We evaluated COVID-19 incidence in the following 6 months, as well as vaccine immunogenicity through anti-Spike IgG titers and T-cell reactivity to the Spike protein.

Results

We included 131 patients with a mean age of 54.9 ± 12.2 years. Almost 40% (n = 52) developed COVID-19 within 6 months after the booster, but 51 (98.1%) were mild infections. Median post-booster antibody levels and antibody variation were 9540.7 (14,724) and 8937.9 (11,561.3) AU/mL, respectively, and 73.3% (n = 96) of the patients showed post-booster T-cell reactivity. Antibody variation was significantly lower in the COVID group (p = 0.015). Although post-booster antibody levels and T-cell reactivity were statistically significantly lower in the patients under biologic DMARD, there was not a significant increase in COVID-19 incidence.

Conclusions

This study shows that a booster vaccine elicits strong immunogenicity and reduces COVID-19 severity, highlighting its importance in immunosuppressed SARD patients. Larger and more homogeneous cohorts are needed to guide periodic booster administration in this susceptible population.

Resumen

Introducción y objetivos

Los pacientes con enfermedad reumática autoinmune sistémica (ERAS) son una población vulnerable a la COVID-19 grave, y a una peor respuesta a la vacunación, necesitando de un refuerzo. Los datos sobre su respuesta son limitados e inconsistentes. El objetivo de este estudio fue evaluar la efectividad y la inmunogenicidad de la tercera dosis de la vacuna SARS-CoV-2 en los pacientes inmunodeprimidos con ERAS.

Materiales y métodos

Realizamos un estudio prospectivo en pacientes portugueses inmunodeprimidos con ERAS, que recibieron un refuerzo de la vacuna SARS-CoV-2, de octubre de 2021 a agosto de 2022. Evaluamos la incidencia de la COVID-19 en los 6 meses siguientes, y la inmunogenicidad mediante el nivel de IgG anti-Spike y la reactividad de las células T con la proteína Spike.

Resultados

Se incluyeron 131 pacientes, edad media de 54,9 ± 12,2 años. Casi el 40% (n=52) desarrollaron COVID-19 en los 6 meses siguientes al refuerzo, pero 51 (98,1%) pacientes fueron infecciones leves. Los niveles medios y la variación posrefuerzo de anticuerpos fueron 9.540,7 (14.724) y 8.937,9 (1.1561,3) AU/ml, respectivamente, y el 73,3% (n = 96) mostraron reactividad de células T. La variación de anticuerpos fue significativamente menor en el grupo COVID (p = 0,015). Aunque los anticuerpos posrefuerzo y la reactividad de las células T fueron estadísticamente significativamente inferiores en los pacientes sometidos a bDMARD, no hubo un aumento significativo en la COVID-19.

Conclusiones

Este estudio muestra la fuerte inmunogenicidad y reducción de la gravedad de la COVID-19 por el refuerzo, destacando su importancia en los pacientes inmunodeprimidos con ERAS. Se necesitan cohortes más grandes y homogéneas para guiar la administración periódica de refuerzos en esta población susceptible.

Introduction

The emergence of the SARS-CoV-2 virus and the COVID-19 pandemic prompted a global effort to develop effective and safe vaccines to halt the spread and decrease the severity of the disease.^1–3 Several vaccines with different mechanisms of action, including messenger RNA (mRNA) and adenovirus vector-based vaccines,^2–4 have since been developed and approved by the World Health Organization (WHO), having had a major impact on the global burden of this disease.^1,5–7 Real-world studies have shown vaccine effectiveness in reducing the incidence and the severity of COVID-19,^3,7,8 by eliciting a humoral and cellular response, both necessary for a strong long-lasting immunity.^4,6,9 However, immune response and antibody levels seem to wane over time, namely in immunosuppressed patients.^7,10–12 Patients with systemic autoimmune rheumatic disease (SARD) may represent a vulnerable population for COVID-19, with some evidence pointing to a higher risk of severe infection.^1,13–16 In addition, although there is still conflicting evidence, several studies have revealed impaired response to SARS-CoV-2 vaccination, especially if under immunosuppression.^{1,5,6,11,12,17,18} Therefore, administration of a booster dose in this population could enhance immune response^10,17,19,20 and is recommended by the European Alliance of Associations for Rheumatology (EULAR) and the American College of Rheumatology (ACR).^21,22 In Portugal, certain immunosuppressed patients were among the first to receive a third dose of the SARS-CoV-2 vaccine, as recommended by the Directorate-General for Health (Direção-Geral da Saúde – DGS) in September 2021.²³ However, data regarding vaccine effectiveness and development of COVID-19 after a booster dose, as well as risk factors for breakthrough infection, in immunosuppressed SARD patients, is still limited and inconsistent.^2,8,9,20 The role of specific immunosuppressive therapies as well as the potential need to withhold them and in which time frame is also unknown.^1,5 Furthermore, most studies focus solely on the humoral response to the vaccine, while T-cell response has been less extensively studied.^9,12,18,24

The main aim of this study was to assess the effectiveness and immunogenicity (humoral and cellular) of the third dose of the SARS-CoV-2 vaccine in patients with systemic autoimmune rheumatic disease under immunosuppression. The secondary objective was to identify risk factors for COVID-19 after booster vaccine in this population.

Material and methods

We conducted a prospective cohort study in patients with systemic autoimmune rheumatic disease under immunosuppressive therapy regularly followed in the Systemic Autoimmune Diseases Outpatient Clinic of the Internal Medicine department of Unidade Local de Saúde de Coimbra, Portugal. The study was carried out from October 2021 to August 2022. All patients received a third dose (or booster) of SARS-CoV-2 vaccine, either BNT162b2 from Pfizer-BionTech or mRNA-1273SPIKEVAX from Moderna, at least 3 months after full primary vaccination. Primary vaccination was completed with one of the four approved SARS-CoV-2 vaccines (BNT162b2 from Pfizer-BionTech, mRNA-1273SPIKEVAX from Moderna, ChAdOx1-S from AstraZeneca/Oxford or Ad.26.COV2.S from Janssen). Consequently, some patients received a homologous booster while others a heterologous booster. Immunosuppression therapy at the time of the administration was not withheld. The study was performed in accordance with the Declaration of Helsinki of the World Medical Association and was approved by the Ethics Committee of our institution. All patients signed written informed consent.

Participants were recruited from all patients aged ≥ 18 years with confirmed SARD, according with the respective international classification criteria, treated with immunosuppressive therapy. Exclusion criteria included refusal to receive a third dose of SARS-CoV-2 vaccine, absence of full primary vaccination, previous history of SARS-CoV-2 infection, other causes for immunosuppression, or refusal to participate in the study.

Variables were obtained from hospital electronic clinical records and included age, sex, comorbidities, SARD and year of diagnosis, usual medication, and immunosuppressive therapy, including glucocorticoids, hydroxychloroquine, conventional synthetic, targeted synthetic and biologic disease-modifying antirheumatic drugs (DMARDs), and duration of treatment of the latter. Type of primary SARS-CoV-2 vaccination and third dose as well as dates of administration were also recorded. C-reactive protein, erythrocyte sedimentation rate, leukocytes and hemoglobin levels were collected within 3 months before the booster vaccine. Primary outcome was development of SARS-CoV-2 infection after the third dose, diagnosed with real-time reverse transcription-polymerase chain reaction (rRT-PCR) test. Patient COVID status was followed-up until 6 months after the booster vaccine, and time until the infection was recorded. Disease severity was evaluated and classified according to the World Health Organization COVID-19 classification criteria as mild (symptomatic patients without evidence of viral pneumonia or hypoxia), moderate (clinical signs of pneumonia, including fever, cough and dyspnea, but without severe pneumonia signs, including SpO₂ over 89% on room air), or severe (clinical signs of pneumonia plus either respiratory rate > 30 breaths/min, severe respiratory distress or SpO₂ < 90% on room air).²⁵

We also analyzed the booster vaccine immunogenicity through the measurement of anti-SARS-CoV-2 IgG titer, more specifically, anti-Spike (S) IgG titers, and through the evaluation of T-cell reactivity to the Spike protein, both between one and four months after the third dose. Anti-S IgG was also evaluated before the third dose and was detected using the chemiluminescent microparticle immunoassay (CMIA) SARS-CoV-2 IgG II Quant, on Alinity i (Abbott Laboratories), with an antibody titer cut off above 50 AU/mL considered reactive. For each measurement, a volume of 5 mL of peripheral blood was collected in a chemistry tube with silica particles. T-cell reactivity to the Spike protein was evaluated with the T-SPOT.COVID test (Oxford Immunotec). Peripheral blood mononuclear cells (PBMCs) were separated from a whole blood sample (7 mL of peripheral blood collected in a sodium heparin tube), using density gradient centrifugation with Lymphoprep (Stemcell Technologies, Vancouver, Canada), and were then washed in a serum free cell medium (Gibco™ AIM V™ Medium). PBMCs were counted using an automated hematology analyzer, and 250,000 were placed in each well of a 96-well microtiter plate. After adding antigens from the SARS-CoV-2 virus, as well as incubation and washing of the wells, a subtract solution (BCIP/NBT^plus) was added to produce the spots. Spot-forming cells (SFCs) were manually counted and T-SPOT.COVID test results were expressed as reactive, if the number of SFCs was ≥8 (per 250,000 PBMCs), non-reactive, if SFC ≤ 4, or equivocal, if SFC between 5 and 7. A nil and a positive control well was used for each patient. Both tests were performed in accordance with the respective manufacturers’ recommendations. Due to lack of patient compliance, post-third dose measurements were not performed in all the patients included in the study. Equivocal results in the T-SPOT.COVID test were excluded from the comparative analysis. No imputations of missing data were performed.

Statistical analysis was performed using IBM SPSS version 24, considering a two-tailed p value of less than 0.05 statistically significant and a 95% confidence interval (CI). Categorical data are presented as absolute and relative frequencies [Number (%)]. Normally distributed quantitative variables are presented as mean ± standard deviation (SD), and non-parametric quantitative variables as median (interquartile range [IQR]). Comparison between groups was achieved using the Chi-square or Fisher's exact tests for categorical variables, or the Student's t-test or Mann–Whitney U-test, for quantitative variables. The Wilcoxon signed-rank test was used for paired samples. Survival curves were obtained using the Kaplan–Meier method and the log-rank test was used to compare time to COVID-19 after the booster according to type of primary vaccination and bDMARD use. Potential risk factors for COVID were evaluated using univariate and multivariable logistic regression (forward:Wald method) to calculate, respectively, the unadjusted and adjusted Odds Ratio (OR).

Results

Of the 337 eligible patients, identified from the group regularly followed in the Systemic Autoimmune Diseases Outpatient Clinic, 206 were excluded, leaving a final cohort of 131 patients with SARD under immunosuppressive therapy and no exclusion criteria. All patients received a third dose of the SARS-CoV-2 vaccine. Sixty patients did not perform post-third dose measurements of anti-S IgG and T-cell reactivity due to lack of compliance, and a further 11 patients had equivocal responses in T-SPOT.COVID test. Hence, they were not considered in the respective immunogenicity analysis (Fig. 1).

Fig. 1.

Flowchart of the patient inclusion process.

Baseline characteristics before booster vaccine

Baseline characteristics of the patients, immediately before receiving the SARS-CoV-2 booster vaccine, are presented on Table 1. Patients’ mean age was 54.6 ± 12.2 years, 75.6% (n = 99) were female, had a mean number of comorbidities of 5 ± 3 and a mean number of medications of 7 ± 4. The two most frequent SARD diagnosis were rheumatoid arthritis (n = 45, 34.4%) and systemic lupus erythematosus (SLE) (n = 27, 20.6%). Median disease duration was of 8.6 (5.6–14.4) years (since the diagnosis). Regarding immunosuppressive therapy, 47.3% (n = 62) of the patients were under prednisolone (7.5–20 mg per day), 31.3% (n = 41) under hydroxychloroquine, 57.3% (n = 75) under a conventional synthetic DMARD (csDMARD) or other immunosuppressive drug, 1.5% (n = 2) under a Janus kinase inhibitor (JAKi), and 37.4% (n = 49) were under a biologic DMARD (bDMARD). Regarding primary SARS-CoV-2 vaccination, over 70% of the patients (n = 94) had received a mRNA vaccine, either from Pfizer or Moderna. Median time between primary vaccination and booster administration was 181.0 (156.0–197.0) days, and median anti-S IgG levels before booster vaccine were 836.1 (212.3–2021.5) AU/mL.

Table 1.

Baseline characteristics of the patients before receiving the SARS-CoV-2 booster vaccine.

Variables	Total (n = 131)
Demographic characteristics
Age, years (mean ± SD)	54.9 ± 12.2
Female sex [n (%)]	99 (75.6)
N of comorbidities (mean ± SD)	5 ± 3
N of medications (mean ± SD)	7 ± 4
SARD [n(%)]
Axial spondyloarthritis	13 (9.9)
Behçet's disease	6 (4.6)
Giant cell arthritis	1 (0.8)
Granulomatosis with polyangiitis	1 (0.8)
Mixed connective tissue disease	6 (4.6)
Psoriatic arthritis	16 (12.2)
Rheumatoid arthritis	45 (34.4)
Sarcoidosis	8 (6.1)
Sjögren's syndrome	3 (2.3)
Systemic lupus erythematous	27 (20.6)
Systemic sclerosis	5 (3.8)
SARD duration, years [median (IQR)]	8.6 (5.6–14.4)
N of immunosuppressive agents [median (IQR)]	2 (1–2)
Immunosuppressive therapy [n(%)]
Prednisolone	62 (47.3)
Hydroxychloroquine	41 (31.3)
csDMARDs or others	75 (57.3)
Azathioprine	6 (4.6)
Cyclosporine	1 (0.8)
Leflunomide	8 (6.1)
Methotrexate	48 (36.6)
Mycophenolate mofetil	7 (5.3)
Sulfasalazine	5 (3.8)
JAK inhibitors	2 (1.5)
Biologic DMARDs	49 (37.4)
Anti-TNF	45 (34.4)
Adalimumab	15 (11.5)
Etanercept	19 (14.5)
Infliximab	11 (8.4)
Others^a	4 (3.1)
Time under bDMARD, months [median (IQR)]	69.2 (10.8–111.5)
Primary SARS-CoV-2 vaccination [n(%)]
mRNA (Pfizer or Moderna)	94 (71.8)
Others (Oxford/AstraZeneca or Janssen)	37 (28.2)
Time to booster after primary vaccination, days [median (IQR)]	181.0 (156.0–197.0)
Pre-booster anti-S IgG, AU/mL [median (IQR)]	836.1 (212.3–2021.5)

a

Rituximab, tocilizumab, secukinumab.

csDMARD – conventional synthetic disease-modifying antirheumatic drug; IQR – interquartile range; N – number; SARD – systemic autoimmune rheumatic disease; SD – standard deviation; TNF – tumor necrosis factor.

Effectiveness of the SARS-CoV-2 booster vaccine and COVID development

From the 131 patients, 39.7% (n = 52) developed SARS-CoV-2 infection confirmed by rRT-PCR within 6 months after the booster vaccine. Only one patient (1.9%) developed moderate disease according to the WHO classification criteria. The other 98.1% (n = 51/52) developed mild disease. Median time to COVID-19 after the booster was 114.5 (52.5–152) days, and 40.4% (n = 21/52) of the patients contracted the infection in the first 90 days after the third dose. No patients were hospitalized or died in the follow-up period.

Looking at effectiveness by type of primary vaccination (mRNA versus non-mRNA), there was no statistical difference in COVID-19 incidence, although its relative frequency was slightly lower in the patients who received a mRNA primary vaccination scheme [mRNA 36.2% (n = 34/94) versus non-mRNA 48.6% (n = 18/37), p = 0.235]. Furthermore, survival curves (Supplementary material, Fig. 1) and log-rank test did not show a statistical difference in time to COVID-19 after booster (p value = 0.264), although patients with a homologous scheme (i.e., mRNA primary vaccination) showed a tendency for lower median time to COVID, namely in the first three months. This can, however, be explained by a significantly higher number of patients in this group.

Immunogenicity of the SARS-CoV-2 booster vaccine

Global markers of the booster vaccine immunogenicity, as well as by non-COVID and COVID status in the following 6 months, are presented in Table 2. Comparison between anti-S IgG levels before and after the booster revealed a statistically significant increase [836.1 (212.3–2021.5) versus 9540.7 (4262.6–18986.6), p < 0.001], and only one patient (with SLE, under hydroxychloroquine) did not present anti-S IgG seroconversion after the booster. The post-vaccine T-SPOT.COVID test was reactive in 44 (73.3%) patients. The patient with no antibody seroconversion was also non-reactive in the T-SPOT.COVID test.

Table 2.

Immunogenicity of the SARS-CoV-2 booster vaccine in the total sample and distributed by non-COVID and COVID status in the following 6 months.

Variables	Totaln = 71	Non-COVIDn = 44	COVIDn = 27	p value
Anti-S IgG seroconversion [n (%)]	70 (98.6)	44 (100.0)	26 (96.3)	–
Post-booster anti-S IgG, AU/mL	9540.7 (4262.6–18986.6)	11,004.7 (4262.6–24944.8)	8320.8 (3307.8–11843.4)	0.059
Anti-S IgG variation^a, AU/mL	8937.9 (4459.4–16020.7)	13,341.7 (6551.5–32906.0)	5278.3 (1913.2–9860.5)	0.015

	n = 60	n = 37	n = 23
Post-booster T-cell reactivity [n (%)]	44 (73.3)	30 (81.1)	14 (60.9)	0.133

All quantitative values are presented as median (IQR).

a

Anti-S IgG variation represents the difference between post- and pre-booster vaccine anti-S IgG levels for each patient.

p values obtained using Chi-square and Mann–Whitney U tests, accordingly.

Looking at humoral immunogenicity by COVID status (Table 2 and Fig. 2), although the difference in the post-booster anti-S IgG between the non-COVID and COVID groups was not statistically significant (p = 0.059), median anti-S IgG variation was lower in the COVID group with a statistically significant p value of 0.015. There was no statistically significant difference in the T-cell reactivity between the two groups.

Fig. 2.

Boxplots representing post-booster anti-S IgG and anti-S IgG variation in non-COVID and COVID groups, respectively. o – Outliers; * – Extreme outlier.

Analyzing booster humoral response by type of primary vaccination, patients who had received a mRNA vaccination scheme presented higher median levels of pre-booster [1106.4 (379.1–2307.7) versus 249.8 (68.2–614.8), p < 0.001] and post-booster anti-S IgG [10,900.5 (4918.4–22,279.8) versus 7166.3 (1447.8–13,362.5), p = 0.018], as well as a greater median antibody variation after the booster [12,382.6 (6069.3–27,857.4) versus 6551.5 (916.8–9480.2), p = 0.017]. There were no statistically significant differences in T-cell reactivity [mRNA 75.0% (n = 27/36) versus non-mRNA 70.8% (n = 17/24), p = 0.771].

Effect of biologic DMARDS on the effectiveness and immunogenicity of the SARS-CoV-2 booster vaccine

Patients using bDMARD had a higher percentage of COVID-19 after the booster [49.0% (n = 24/49) versus 34.1% (n = 28/82)] and were infected more frequently within the first 90 days [45.8% (n = 11/24) versus 35.7% (n = 10/28)], in comparison with patients under non-biologic immunosuppression. These differences, however, were not statistically significant (p = 0.101 and p = 0.573, respectively). Survival curves (Supplementary material, Fig. 2) confirm the latter tendency, but the log-rank test did not show a statistically significant difference (p = 0.625).

Regarding immunologic response to the booster (Fig. 3), use of bDMARD resulted in statistically significantly lower median levels of post-anti-S IgG [10,900.5 (7279.4–23,991.5) AU/mL versus 5100.5 (1638.2–12,600.4) AU/mL, p = 0.006] and anti-S IgG variation [13,105.3 (8269.7–30,875.2) AU/mL versus 4446.8 (150.0–6551.1) AU/mL, p < 0.001], as well as lower percentage of T-cell reactivity [83.8% (n = 31/38) versus 56.5% (n = 13/22), p = 0.034]. All patients under bDMARD presented anti-S IgG seroconversion. In addition, median time under bDMARD treatment was 69.2 (10.8–111.5) months and was lower in the patients who contracted COVID-19 [59.0 (10.5–118.9) versus 46.7 (13.3–111.1) months], although this difference was not significant (p = 0.912).

Fig. 3.

Boxplots representing post-booster anti-S IgG and anti-S IgG variation in non-bDMARD and bDMARD groups, respectively. o – Outliers; * – Extreme outlier.

Risk factors for SARS-CoV-2 infection in SARD patients under immunosuppression

Univariate logistic regression analysis of potential risk factors for COVID-19 in SARD patients under immunosuppression is presented in Table 3. Younger age, chronic liver disease and infliximab showed statistically significant association with COVID-19 development. On the other hand, post-booster anti-S IgG, antibody variation and T-cell reactivity did not show association with COVID-19 development in the univariate analysis. Multivariable analysis including the identified risk factors, and adjusting for sex, kept statistical value: younger age OR 0.97, 95% CI 0.93–1.00 (p = 0.046); chronic liver disease OR 8.53, 95% CI 1.53–47.63 (p = 0.015); infliximab OR 7.91, 95% CI 1.69–37.05 (p = 0.009).

Table 3.

Potential risk factors for COVID-19 in SARD patients under immunosuppression.

Variables	Total (n = 131)	Non-COVID (n = 79)	COVID (n = 52)	OR (95% CI)	p value
Age, years (mean ± SD)	54.9 ± 12.2	56.7 ± 11.5	52.0 ± 12.6	0.97 (0.94–1.00)	0.031
Female sex [n(%)]	99 (75.6)	59 (74.7)	40 (76.9)	1.13 (0.50–2.57)	0.770
N of comorbidities (mean ± SD)	5 ± 3	5 ± 3	5 ± 2	0.99 (0.86–1.14)	0.887
Osteoporosis [n(%)]	22 (16.8)	14 (17.7)	8 (15.4)	0.84 (0.33–2.18)	0.726
Arterial hypertension [n(%)]	46 (35.1)	29 (36.7)	17 (32.7)	0.84 (0.40–1.75)	0.638
Dyslipidemia [n(%)]	43 (32.8)	25 (31.6)	18 (34.6)	1.14 (0.54–2.40)	0.723
Type 2 diabetes [n(%)]	19 (14.5)	13 (16.5)	6 (11.5)	0.66 (0.24–1.87)	0.436
Obesity [n(%)]	31 (23.7)	19 (24.1)	12 (23.1)	0.95 (0.42–2.16)	0.898
Hyperuricemia [n(%)]	6 (4.6)	1 (1.3)	5 (9.6)	8.30 (0.94–73.21)	0.057
Heart failure [n(%)]	9 (6.9)	5 (6.3)	4 (7.7)	1.23 (0.32–4.83)	0.763
Chronic kidney disease [n(%)]	8 (6.1)	4 (5.1)	4 (7.7)	1.56 (0.37–6.55)	0.541
Thyroid disease [n(%)]	29 (22.1)	17 (21.5)	12 (23.1)	1.09 (0.47–2.53)	0.834
Anemia [n(%)]	22 (16.8)	16 (20.3)	6 (11.5)	0.51 (0.19–1.41)	0.197
Chronic liver disease [n(%)]	10 (7.6)	2 (2.5)	8 (15.4)	7.00 (1.42–34.43)	0.017
Solid neoplasm [n(%)]	7 (5.3)	4 (5.1)	3 (5.8)	1.15 (0.25–5.35)	0.861
Depression [n(%)]	53 (40.5)	33 (41.8)	20 (38.5)	0.87 (0.43–1.78)	0.706
Smoking [n(%)]	25 (19.1)	19 (24.1)	6 (11.5)	0.41 (0.15–1.11)	0.081
Rheumatoid arthritis [n(%)]	45 (34.4)	31 (39.2)	14 (26.9)	0.57 (0.27–1.22)	0.148
Spondyloarthropathies [n(%)]	29 (22.1)	17 (21.5)	12 (23.1)	1.09 (0.47–2.53)	0.834
Systemic lupus erythematosus [n(%)]	27 (20.6)	15 (19.0)	12 (23.1)	1.28 (0.54–3.01)	0.572
SARD duration, years	8.6 (5.6–14.4)	9.4 (5.5–16.4)	8.5 (5.7–13.3)	0.97 (0.92–1.01)	0.154
N of medications (mean ± SD)	7 ± 4	7 ± 4	8 ± 3	1.04 (0.94–1.15)	0.428
More than 5 drugs [n(%)]	84 (64.1)	50 (63.3)	34 (65.4)	1.10 (0.53–2.28)	0.807
N of immunosuppressive agents	2 (1–2)	2 (1–2)	2 (1–2)	0.89 (0.56–1.41)	0.627
Prednisolone [n(%)]	62 (47.3)	39 (49.4)	23 (44.2)	0.81 (0.40–1.64)	0.565
Hydroxychloroquine [n(%)]	41 (31.3)	25 (31.6)	16 (30.8)	0.96 (0.45–2.05)	0.916
csDMARDs or others [n(%)]	75 (57.3)	49 (62.0)	26 (50.0)	0.53 (0.25–1.11)	0.092
Methotrexate [n (%)]	48 (36.6)	32 (40.5)	16 (30.8)	0.65 (0.31–1.37)	0.259
bDMARDs [n(%)]	49 (37.4)	25 (31.6)	24 (46.2)	1.85 (0.90–3.81)	0.095
Anti-TNF [n (%)]	45 (34.4)	23 (29.1)	22 (42.3)	1.79 (0.86–3.72)	0.121
Adalimumab [n (%)]	15 (11.5)	8 (10.1)	7 (13.5)	1.38 (0.47–4.07)	0.559
Etanercept [n (%)]	19 (14.5)	12 (15.2)	7 (13.5)	0.87 (0.32–2.38)	0.784
Infliximab [n (%)]	11 (8.4)	3 (3.8)	8 (15.4)	4.61 (1.16–18.27)	0.030
Time under bDMARD, months	69.2 (10.8–111.5)	59.0 (10.5–118.9)	46.7 (13.3–111.1)	1.00 (0.99–1.01)	0.873
CRP, mg/dL	0.2 (0.1–0.5)	0.2 (0.1–0.4)	0.3 (0.1–0.6)	1.07 (0.62–1.85)	0.808
ESR, mm/h	14.5 (9–22)	13.0 (9.0–20.5)	17.0 (10.0–23.0)	1.03 (0.98–1.08)	0.252
Leukocytes, ×10^9/L	6.5 (5.2–7.7)	6.1 (5.1–7.6)	6.7 (5.6–8.0)	1.16 (0.97–1.39)	0.114
Neutrophils, ×10^9/L	3.7 (2.9–4.8)	3.7 (2.8–4.9)	3.6 (3.0–4.8)	1.05 (0.84–1.31)	0.682
Lymphocytes, ×10^9/L	1.7 (1.4–2.3)	1.6 (1.2–2.3)	1.9 (1.4–2.3)	1.09 (0.69–1.73)	0.705
Hemoglobin, g/dL	13.5 (12.5–14.2)	13.4 (12.4–14.2)	13.8 (12.6–14.3)	1.11 (0.84–1.46)	0.460
mRNA primary vaccination [n(%)]	94 (71.8)	60 (75.9)	34 (65.40)	0.60 (0.28–1.29)	0.191
Time to booster after primary vaccination, days	181.0 (156.0–197.0)	183.0 (156.0–198.0)	175.0 (147.5–195.5)	0.99 (0.99–1.00)	0.169
Time to booster > 6 months [n(%)]	70 (53.4)	45 (57.0)	25 (48.1)	0.70 (0.35–1.43)	0.319
Post-booster anti-S IgG, AU/mL	9540.7 (4262.6–18986.6)	11,004.7 (4262.6–24944.8)	8320.8 (3307.8–11843.4)	1.00 (1.00–1.00)	0.077
Anti-S IgG variation^a, AU/mL	8937.9 (4459.4–16020.7)	13,341.7 (6551.5–32906.0)	5278.3 (1913.2–9860.5)	1.00 (1.00–1.00)	0.057
Post-booster T-cell reactivity^b[n(%)]	44 (73.3)	30 (81.1)	14 (60.9)	0.35 (0.11–1.13)	0.080

Except when expressively said otherwise, all quantitative values are presented as median (IQR).

a

Anti-S IgG variation represents the difference between post- and pre-booster vaccine anti-S IgG levels for each patient.

b

Percentage derived from the number of patients who presented valid (reactive or non-reactive) post-booster T-SPOT.COVID tests (n = 60).

bDMARD – biologic disease-modifying antirheumatic drug; CRP – C-reactive protein; ESR – erythrocyte sedimentation rate; JAKi – Janus kinase inhibitor; N – number; TNF – tumor necrosis factor.

p values obtained using univariate logistic regression.

Discussion

To the best of our knowledge, this is the first study to assess the effectiveness and immunogenicity of the booster SARS-CoV-2 vaccine in Portuguese patients with SARD under immunosuppression.

In this study, over one third of the patients developed SARS-CoV-2 infection, an incidence much higher than previously reported in the literature in both non-vaccinated, varying between 0.16% and 18.1%,^15,16,26 and vaccinated SARD patients, varying between 0.1% and 1.1%.^18,19 However, existing published data should be compared with caution since some of the studies were performed during the first wave of the pandemic when strict quarantine and confinement measures were in place, and others were performed in different geographical contexts, in some cases with lower testing rates, facts that could on their own explain this discrepancy. Other possible reasons include lower efficacy against the Omicron variant,^3,7,27,28 which was the main variant in circulation in Portugal during the study timeframe (according to the National Health Institute Doutor Ricardo Jorge), and also the lack of temporary discontinuation of immunosuppressive therapy which has been shown to be potentially advantageous in increasing efficacy and immunogenicity of the vaccine.^{1,17,19,22,29} On the other hand, all but one patient had mild disease and no cases resulted in hospitalization or death, which is in line with other studies on post-booster breakthrough infection,^19,30 and unlike the increased COVID-19 severity and rates of hospitalization (0.11–44%) and death (0–40%) reported in non-vaccinated immunosuppressed SARD patients.^1,13–16,18 Therefore, in comparison with current literature, despite a likely low effectiveness in preventing disease, the booster vaccine may have reduced severity, rate of hospitalization and death in this population. In addition, almost 60% of the COVID patients developed the disease after over 90 days post-booster, which can be a consequence of the waning of immune protection over time, and varying risks of exposure during the time after the booster.

Regarding immunogenicity, it has been previously demonstrated that a SARS-CoV-2 booster vaccine was effective in eliciting seroconversion in previous non-responders and in enhancing humoral response, including in immunosuppressed SARD patients, with seroconversion rates of around 47–100%.^{10,17,19,20,31,32} In our sample, there was an almost universal seroconversion with a statistically significant increase in anti-S IgG levels after the booster. Only one patient, a previous non-responder with SLE solely under hydroxychloroquine, did not seroconvert, once again pointing out to other mechanisms coming into play. On the contrary, cellular response has been much less studied, in spite of its recognized importance for the COVID-19 vaccine immunogenicity,^6,9,31–33 particularly considering humoral waning, and, also, in patients with B-cell depletion.^12,34 In our study, T-cell response was present in a slightly lower percentage than previously reported in immunosuppressed patients with two SARS-CoV-2 vaccine doses, with one study reaching 81.8% of T-cell reactivity, using the T-SPOT method.³⁴ These previous superior results can be the product of a more heterogeneous population and different T-cell reactivity testing timings. Inclusion of patients under bDMARD in our analysis could also have lowered the overall percentage, considering its associated lower T-cell reactivity.

Importantly, correlation of immune response with protection against infection is crucial for the development and improvement of COVID-19 vaccines, especially considering the multiple SARS-CoV-2 variants in circulation.^27,35,36 Current available literature supports a correlation between antibody levels and protection against infection.^{27,32,35–37} However, our results did not show a statistically significant association of these factors with COVID development, in the logistic regression. Hence, other factors may also be relevant in determining the development of COVID-19, namely immunosuppression, specific SARD or other comorbidities. The increased circulation of the Omicron variant during the study timeframe may have also contributed, considering its ability to evade the current vaccines-induced immune protection.^3,7,27,28

Most studies have concluded that immunosuppression impairs the efficacy and immune response to the vaccine, although the exact role of each type of therapy is not fully understood.³¹ Overall, patients under bDMARD seem to develop lower titers of antibodies, although use of other immunosuppressive drugs, such as steroids and mycophenolate mofetil, also reduces immune response.^1,6,17,19,32 It is relatively consensual that anti-CD20 therapies, particularly rituximab, are associated with weakened humoral response,^{1,2,5,10,12,18,19,29} but anti-tumor necrosis factor (TNF) agents have also been implicated.^2,5,6,38 Limited evidence has shown that infliximab produces an attenuated response to the vaccine with antibody seroconversion rates after primary vaccination of around 85–90%.^5,6 In our study, use of bDMARD resulted in higher and earlier COVID-19 development after the booster, although not statistically significant. Furthermore, this group presented statistically significantly lower median levels of post-anti-S IgG and anti-S IgG variation, as well as lower percentage of T-cell reactivity, which is in line with what has been previously described. However, only infliximab revealed an independent association with COVID-19 development, with its use conferring a risk nearly 8 times higher. Thus, it can be argued that the bDMARD-associated lower observed effectiveness could have been driven by the infliximab influence. Nevertheless, this finding is difficult to contextualize in this study, considering the low number of patients in each group and the lack of individualization of mono versus combination therapy.

Considering the evidence reporting slightly better efficacy of mRNA in comparison with vector-based vaccines,^4,5 we hypothesized that the type of primary vaccination could influence clinical and immune response to the booster vaccine. Our results showed that mRNA primary vaccination was associated with a greater pre-booster, post-booster anti-S IgG and antibody variation. However, this did not reflect in clinical or cellular outcomes. These data suggest that type of primary vaccination may play a significant role in the post-booster humoral response, but that both a homologous and heterologous scheme may be effective regarding clinical outcomes, as long as the third dose is a mRNA vaccine.

Finally, other potential risk factors for low vaccine response and COVID-19 development after the vaccine are far from being well established, and available literature is still scarce. The influence of age in the response to the vaccine is still unclear, with conflicting findings on seroconversion.^{1,18,19,27,39} In this study, age was inversely, although weakly, associated with breakthrough infection, which can, in fact, be most likely linked to a higher degree of exposure, due to typically less social distancing, than to an actual lower age-associated risk.⁴⁰ Comorbidities may also play a role in increasing risk and worse outcomes of SARS-CoV-2 infection in the general population^41–46 and, more specifically, in SARD patients,^19,25,40,47 although data available is limited and considerably heterogeneous. In our study, only chronic liver disease showed potential association with disease development, but current evidence is also contradictory.^45,46,48,49 However, this result should be carefully interpreted considering the small sample size in each group and the heterogeneity of the exact underlying cause for liver disease. In addition, none of the specific SARD diagnostic groups nor other treatment-related characteristics correlated with COVID-19 development, which should also be interpreted in light of a considerable patient heterogeneity and low patient number in each sub-group.

This study has some limitations that should be addressed in future trials. The main limitations are lack of a healthy control group and inclusion of a heterogenous population with different autoimmune diseases and treatment regimes, both of which handicap the effectiveness and immunogenicity comparative analysis. In addition, patient drop-out and lack of compliance reduced the post-booster immunological assessment sample number and testing time-frame homogeneity, which underpowers the statistical analysis.

Conclusions

This study is in line with the current literature that a SARS-CoV-2 booster vaccine is effective in eliciting strong humoral responses, resulting in low COVID-19 severity and rates of hospitalization in immunosuppressed SARD patients, even in patients under bDMARD. Therefore, it suggests the importance of a booster vaccine administration in these patients, particularly considering the several SARS-CoV-2 variants in circulation. Identification of risk factors for breakthrough infection, such as age, comorbidities, and, particularly, the use of bDMARDs is also paramount for selecting patients in need of a more rigorous monitoring, although available data is still scarce and contradictory. Larger cohorts with a more homogeneous population are needed to fully evaluate the effectiveness and immunogenicity of the SARS-CoV-2 booster vaccine in this population, as well as to assess the impact of specific immunosuppressive therapies, in order to provide more robust evidence to guide possible periodic booster administration in this vulnerable population.

Informed consent

Written informed consent was obtained from all participants. This study was approved by the Ethics Committee of Unidade Local de Saúde de Coimbra.

Declaration of generative AI and AI-assisted technologies in the writing process

No AI was used for the writing of this manuscript.

Funding

No specific funding was used for the development of this research.

Conflict of interest

None to disclose.