Real-world immune response to SARS-CoV-2 vaccination in Thai patients with systemic lupus erythematosus: a cross-sectional observational study

Abstract

Background

Immunogenicity to SARS-CoV-2 vaccination in patients with SLE varies by vaccine type and immune-modulating therapy. However, data in Southeast Asian populations, especially among Thai patients with SLE, remain limited.

Objective

To assess the levels of IgG response after the second dose of SARS-CoV-2 vaccination in Thai patients with SLE compared with healthy controls, and to explore factors associated with low immunogenicity to SARS-CoV-2 vaccine.

Methods

In this cross-sectional case–control study, adult Thai patients with SLE and age-matched and sex-matched healthy controls were enrolled following two SARS-CoV-2 vaccine doses under the Thai national immunisation programme. SARS-CoV-2 spike protein IgG was measured using electro-chemiluminescence immunoassay. Low immunogenicity was defined as IgG<15 U/mL.

Results

Among 92 patients with SLE and 41 controls, IgG levels were not significantly different (median: 221.3 vs 196.8 U/mL, p=0.41). The messenger RNA (mRNA) and viral vector vaccines yielded higher antibody levels than inactivated vaccines in patients with SLE. Factors such as active lupus nephritis and moderate-to-high dose corticosteroid use appeared to be associated with lower IgG responses, though not statistically significant.

Conclusions

Thai patients with SLE demonstrated an immune response comparable to that of healthy controls. A stronger immune response was observed in patients with SLE who received viral vector and mRNA vaccines, compared with those who received inactivated vaccines. Both vaccine type and disease-related factors may influence the magnitude of the immune response, emphasising the need for tailored vaccination strategies in this population.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Patients with SLE are known to have reduced immunogenicity to SARS-CoV-2 vaccines, especially those receiving immunosuppressive therapy. However, data from Southeast Asian populations, particularly with diverse vaccine regimens including inactivated vaccines, remain limited.

WHAT THIS STUDY ADDS

• Thai patients with SLE mount an IgG response on par with healthy controls after SARS-CoV-2 vaccination, with messenger RNA (mRNA) and viral vector vaccines eliciting higher responses than inactivated vaccines. Active lupus nephritis and higher corticosteroid doses were linked to lower immunogenicity, yet not statistically confirmed.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The findings support the importance of individualised vaccination strategies for patients with SLE, prioritising mRNA or viral vector vaccines and closer monitoring of those with active disease or immunosuppression. Real-world data from low- and middle-income settings can inform national policies and guide future research on booster timing and immune protection in immunocompromised populations.

Introduction

SLE is a chronic autoimmune disease characterised by dysregulated type I interferon (IFN) pathways, contributing to both disease pathogenesis and impaired antiviral immunity, including against SARS-CoV-2. In addition to type I IFN signalling, SLE is also associated with impaired germinal centre formation, autoreactive B-cell survival and dysfunctional T follicular helper cells, all of which compromise the development of high-affinity antibody responses.^{1 2}

The COVID-19 pandemic, caused by SARS-CoV-2 infection, has resulted in over 770 million confirmed cases and more than 7 million deaths worldwide. In Thailand, daily incidence rates have exceeded 10 000 cases at various points, with a high associated mortality rate.³ Factors contributing to COVID-19-related mortality include advanced age, obesity and pre-existing comorbidities.⁴

Patients with SLE are at increased risk of severe COVID-19 outcomes, including hospitalisation and death, compared with the general population.⁴,⁸ This heightened risk is attributed to both disease-related immune dysfunction and the use of immunosuppressive therapies.⁹,¹³

In response to the COVID-19 pandemic, various vaccines, including inactivated, viral vector and messenger RNA (mRNA) types, have been deployed globally. In Thailand, multiple SARS-CoV-2 vaccination regimens have been used, including inactivated SARS-CoV-2 vaccine (Sinovac (SV), Sinopharm (SP)), non-replicated viral vaccine (AstraZeneca, Janssen), mRNA vaccine (Pfizer/BioNTech, Moderna) and protein subunit vaccine (Covovax).

Reported humoral immunogenicity rates in immunocompetent adults were approximately 51–79% for inactivated vaccines (SV, SP), 63.1% for AstraZeneca and up to 90% for Pfizer/BioNTech.³ The immunogenicity and clinical efficacy of these vaccines, including reductions in transmission, hospitalisation, disease severity and mortality, are well characterised in the general population.¹⁴,¹⁸

However, their effectiveness in immunocompromised populations, including patients with SLE, remains less well understood. Studies from Western cohorts have demonstrated impaired humoral responses in patients with SLE, particularly among those receiving immunosuppressants.^{19 20} Nevertheless, data from Asian populations, especially those with access to diverse vaccine regimens like Thailand, remain limited.

This study aimed to evaluate the SARS-CoV-2 spike protein IgG response following the second vaccine dose, administered under the National Health Thailand programme, among Thai patients with SLE and to identify factors associated with low immunogenicity. Findings from this study may provide valuable information for healthcare planning and COVID-19 prevention strategies in patients with SLE.

Method

This study was a cross-sectional observational study conducted among patients with SLE aged 18 and above who attended the Rheumatology Clinic at Srinagarind Hospital, Khon Kaen University, Khon Kaen, Thailand, between February 2022 and May 2022. All participants had completed two doses of SARS-CoV-2 vaccination, regardless of vaccine type. Vaccine type was not assigned by investigators but was determined by national availability and individual preference during the public vaccination campaign. Patients were eligible if they had received their second vaccine dose within 28–90 days before IgG testing. We excluded (1) those with a previous history of SARS-CoV-2 infection at any time before the immunogenicity test either confirmed by PCR or rapid antigen test kit, (2) those who received a booster dose (third or fourth dose) of vaccination before the immunogenicity test and (3) those who denied or were unable to collect blood sample. All eligible patients had to sign informed consent before entry into the study. Healthy controls were selected from an existing cohort who underwent SARS-CoV-2 spike protein IgG testing using the same method. One control per case was matched by sex, age (±3 years) and vaccine regimen, where possible. Due to a smaller control pool (n=41) relative to patients with SLE (n=92), some controls were matched to more than one case. Appropriate statistical methods were applied to account for repeated control matching. Comorbidity, socioeconomic status and healthcare exposure data were not available in the control dataset.

Demographic data, including age, sex and SLE-related data, including duration of disease, disease activity, organ involvement, comorbidities, routine laboratory data and concomitant medication, were reviewed from medical records. Vaccination data, including the date of the first and second vaccination and the SARS-CoV-2 vaccine regimen, were collected, while anti-SARS-CoV-2 spike protein IgG levels were prospectively measured as part of the study protocol.

Laboratory methods

Specimens were collected using serum-separated tube gel, 5 mL of blood and analysed within 2 hours. Immunogenicity was assessed by quantitative test for SARS-CoV-2 spike protein IgG antibody by electrochemiluminescence immunoassay (ECLIA) technique. The process of sample test was included (1) step 1 (9 min): 12 µL of the patient sample was incubated with a mix of biotinylated and ruthenylated receptor-binding domain (RBD) antigen. Double antigen sandwich (DAGS) immune complexes were formed in the presence of corresponding antibodies, (2) step 2 (9 min): streptavidin-coated microparticles were added, then waiting for the DAGS complexes bonded to the solid phase via interaction of biotin and streptavidin and (3) step 3 (measurement): the reagent mixture was transferred to the measuring cell, where the microparticles were magnetically captured onto the surface of the electrode. Unbound substances were subsequently removed. Electrochemiluminescence was then induced by applying a voltage and measured with a photomultiplier. The signal yield increased with the antibody titre.

Operational definitions

SLE was diagnosed based on the American College of Rheumatology (ACR) 1997 criteria, the Systemic Lupus International Collaborating Clinics (SLICC) 2012 or the 2019 European Alliance of Associations for Rheumatology/ACR classification criteria for SLE.²¹,²³

The Thai SARS-CoV-2 vaccination programme was defined by the Thai Department of Disease Control and included all SARS-CoV-2 vaccination regimens in Thailand consisting of two doses of SARS-CoV-2 vaccines: (1) AstraZeneca-AstraZeneca, (2) SV-AstraZeneca or Pfizer/BioNTech, (3) Pfizer/BioNTech-Pfizer/BioNTech, (4) Moderna-Moderna and (5) other regimens (eg, SV-SV, SP-SP, AstraZeneca-Pfizer/BioNTech).²⁴

Disease activity was assessed using the Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K), a validated tool for clinical disease monitoring in SLE.²⁵ Clinical SLEDAI-2K, which excludes serological parameters, was used due to limited availability of anti-dsDNA and complement data, in accordance with real-world clinical practice.

Active disease was defined as a clinical SLEDAI-2K score ≥4, a cut-off consistent with previously published studies evaluating vaccine immunogenicity in SLE cohorts.^{19 26}

Obesity is defined by a body mass index ≥30 kg/m². Leucopenia is defined by a white blood cell count <3.0 ×10⁹/L, while thrombocytopenia is defined by a platelet count <100 ×10⁹/L. Autoimmune haemolytic anaemia is characterised by a haemolytic blood picture and a positive direct Coombs test. Lupus nephritis was defined as any presentation that met the Kidney Disease: Improving Global Outcomes (KDIGO) criteria on the basis of clinical documentation and corroborating laboratory data—even when histological confirmation was not available.²⁷ Active lupus nephritis is defined as lupus nephritis that needs immunosuppressants including glucocorticoids. Immunosuppressants included glucocorticoids (>7.5 mg/day of prednisolone), methotrexate, sulfasalazine, leflunomide, azathioprine, mycophenolate mofetil, ciclosporin, tacrolimus, cyclophosphamide, Janus kinase inhibitors and biological therapies (anti-cytokine therapy, B-cell depleting therapies). A prednisolone dose >7.5 mg/day was selected based on standardised glucocorticoid classification, which defines this range as ‘medium dose’, and reflects levels associated with clinically meaningful immunosuppressive effects and increased risk of adverse outcomes in SLE.²⁸,³⁰

Low immunogenicity to the SARS-CoV-2 vaccine was defined as an anti-SARS-CoV-2 spike protein IgG level of <15 U/mL, measured using the Elecsys Anti-SARS-CoV-2 S assay (ECLIA). This cut-off was selected based on prior data suggesting that levels ≥15 U/mL are associated with neutralising antibody activity.

Non-detectable immunogenicity was defined as an anti-SARS-CoV-2 spike protein IgG level <0.8 U/mL, which is the manufacturer’s threshold for a non-reactive result.³¹,³³

Sample size

The sample size calculation was based on the study of Izmirly et al¹⁹ We applied the median and SD (with assumed SD as one-fourth of the summation of the IQR) of immunoglobulin G against SARS-CoV-2 in SLE and healthy group of 235.2 (SD 151.8) (IQR 75.9–531.4) U/mL and 435.7 (259.4) (IQR 269.0–768.6) U/mL, respectively. Therefore, at least 18 patients in each group and each vaccination regimen were included for 80% power and 0.05 significance to detect a difference. To account for the diversity of vaccine regimens available in Thailand and improve real-world representativeness, we included 92 patients with SLE across all available vaccine regimens in this study. Although the initial study plan aimed to recruit at least 18 participants per regimen, the real-world distribution of vaccines in Thailand during the study period was governed by national supply and patient choice rather than investigator control. Consequently, participants received a diverse range of vaccine combinations, and several subgroups contained fewer patients than anticipated. To preserve external validity, we therefore included all consecutive eligible patients with SLE (n=92) irrespective of regimen frequency.

Statistical analysis

Demographic data were presented as a percentage or proportion for categorical data and mean±SD or median with IQR for continuous data. The level of SARS-CoV-2 spike IgG antibody level was presented as mean±SD or median with IQR as appropriate. The prevalence of low immunogenicity to SARS-CoV-2 spike IgG antibody was calculated with its 95% CI. OR with a 95% CI was applied to identify the factors associated with low immunogenicity to the SARS-CoV-2 vaccine. Variables with p values<0.10 were entered into a multiple logistic regression model. P values <0.05 were statistically significant. All statistical analyses were performed by using Stata V.16.0 (StataCorp, College Station, Texas, USA).

Results 

We included 92 patients with SLE (mean age 38.6 years; 95.7% female) and 41 age-matched and sex-matched healthy controls. The median IgG levels were 221.3 U/mL in SLE and 196.8 U/mL in controls (p=0.41); 22.8% of patients with SLE had low immunogenicity (table 1).

Table 1. Demographic data.

Data	Healthy control N=41	Patients with SLE N=92	P value
Female; n (%)	41 (100)	88 (95.7)	0.31
Age (years); mean±SD	34.6±12.6	38.6±13.5	0.07
Vaccination regimen; n (%)
SV-SV	0	1 (1.1)
SV-SP	0	2 (2.2)
SV-PZ	1 (2.4)	3 (3.3)
SV-AZ	0	14 (15.2)
SP-SV	0	2 (2.2)
SP-SP	9 (22.0)	9 (9.8)
PZ-PZ	11 (26.8)	29 (31.5)
PZ-AZ	0	2 (2.2)
MD-MD	0	9 (9.8)
AZ-PZ	14 (34.1)	15 (16.3)
AZ-AZ	6 (14.6)	6 (6.5)
Age at onset (years); mean±SD		28.0±13.4
Disease duration (years); mean±SD		10.6±6.9
BMI (kg/m2); mean±SD		23.3±4.6
Comorbidity; n (%)
Diabetes		9 (9.8)
Hypertension		38 (41.3)
Dyslipidaemia		33 (35.9)
Chronic kidney disease (eGFR≤60 mL/min/1.73 m²))		11 (12.0)
Obesity		16 (17.6)
Cancer		4 (4.4)
Laboratory
Haemoglobin (g/L); median (IQR)		120.0 (108.0–131.0)
White blood cell (×10⁹/L); mean±SD		6.8±3.3
Platelet (×10⁹/L); mean±SD		261.5±77.0
Creatinine (mg/dL); median (IQR)		0.7 (0.6–0.9)
Albumin (g/dL); mean±SD		4.1±0.5
Globulin (g/dL); mean±SD		3.4±0.7
ALT (IU/L); mean±SD		20.9±17.2
AST (IU/L); mean±SD		24.2±12.6
ALP (IU/L); mean±SD		87.7±59.4
ESR (mm/hour); mean±SD		48.6±29.7
CRP (mg/dL); median (IQR)		3.5 (0.6–36.3)
SLE involvement; n (%)
Fever; BT≥38°C (excluding infection)		1 (1.1)
Cutaneous
Discoid lupus erythematosus		5 (5.4)
Cutaneous vasculitis		2 (2.2)
Arthritis		2 (2.2)
Myositis		0
Serosal		0
Haematological		15 (16.3)
Neuropsychiatric		0
Renal		22 (23.9)
Biopsy-confirmed lupus nephritis		18 (19.6)
Class II		3 (3.3)
Class III		1 (1.1)
Class IV		11 (12.0)
Class V		3 (3.3)
APS; n (%)		9 (9.8)
Asymptomatic APS		2 (2.2)
Symptomatic APS		7 (7.6)
Concomitant medications
Prednisolone; n (%)		82 (89.1)
≤7.5 mg/day		65 (70.7)
>7.5 mg/day		27 (29.3)
Dose (mg/day); median (IQR)		5.0 (3.5–10.0)
Cyclophosphamide; n (%)		1 (1.1)
Dose (mg/month); median (IQR)		500.0 (500.0–500.0)
Mycophenolate mofetil; n (%)		29 (33.7)
Dose (mg/day); mean±SD		1321.6±634.0
Ciclosporin A; n (%)		6 (6.5)
Dose (mg/day); median (IQR)		175.0 (100.0–200.0)
Azathioprine; n (%)		5 (5.4)
Dose (mg/day); median (IQR)		50.0 (18.75–87.5)
Hydroxychloroquine/chloroquine; n (%)		74 (80.4)
Dose (mg/day); mean±SD		156.6±70.4
Outcome
SARS-CoV-2 IgG level (U/mL); median (IQR)	196.8 (15.5–462.6)	221.3 (43.1–1113.5)

ALP, alkaline phosphatase; ALT, alanine transaminase; APS, antiphospholipid syndrome; AST, aspartate transaminase; AZ, AstraZeneca; BMI, body mass index; BT, body temperature; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; ESR, erythrocyte sedimentation rate; MD, Moderna; PZ, Pfizer; SP, Sinopharm; SV, Sinovac.

Among patients with SLE, those receiving Moderna-Moderna had the highest median IgG level (2480.0 U/mL), while SP-SP had the lowest level (5.40 U/mL) (table 2).

Table 2. Levels of SARS-CoV-2 spike IgG in patients with SLE, categorised by vaccine regimen and grouped by platform.

Platform	Regimen (1st-2nd dose)	n	Level of antibody (U/mL); median (IQR)
Inactivated (homologous)	SV-SV	1	318.8 (318.8–318.8)
	SP-SP	9	5.4 (0.4–42.4)
Inactivated → other	SV-SP	2	191.1 (7.5–374.6)
	SV-PZ	3	221.9 (0.3–6873.0)
	SV-AZ	14	231.6 (13.2–517.1)
	SP-SV	2	42.0 (0.3–83.6)
Viral vector (homologous)	AZ-AZ	6	104.1 (92.9–707.4)
mRNA (homologous)	PZ-PZ	29	220.7 (66.9–681.7)
	MDMD	9	2480.0 (1102.0–3169.0)
Heterologous (mRNA ↔ vector)	PZ-AZ	2	978.4 (761.8–1195.0)
	AZ-PZ	15	768.6 (83.0–1617.0)

AZ, AstraZeneca; MD, Moderna; mRNA, messenger RNA; PZ, Pfizer; SP, Sinopharm; SV, Sinovac.

Although IgG levels in patients with SLE were numerically lower than those in healthy controls, the difference was not statistically significant (table 3).

Table 3. The comparison of IgG level against spike protein of SARS-CoV-2 between patients with SLE and healthy controls (matched regimens).

Regimen	SLE (n)	IgG level against spike protein of SARS-CoV-2 in patients with SLE (U/mL); median (IQR)	Controls (n)	IgG level against spike protein of SARS-CoV-2 in healthy control (U/mL); median (IQR)	P value
SP-SP	9	5.4 (0.4–42.4)	9	15.6 (15.6–141.1)	0.07
PZ-PZ	29	372.5 (66.9–1375.0)	11	462.6 (0.4–462.6)	0.46
AZ-PZ	15	556.4 (83.0–1301.0)	14	1950.0 (9.2–3311.0)	0.41
AZ-AZ	6	104.1 (92.9–707.4)	6	344.4 (196.8–377.6)	0.59

In the SP-SP control group, 3 of 9 participants (33.3%) had identical IgG levels of 15.6 U/mL, while in the PZ-PZ group, 7 of 11 (63.6%) had IgG levels of 462.6 U/mL. These values may indicate clustering near the lower or upper range of reported assay results.

AZ, AstraZeneca; PZ, Pfizer; SP, Sinopharm.

In univariate analysis, patients with low immunogenicity were more likely to have active lupus nephritis (p=0.02), higher serum creatinine (p=0.003) and lower globulin levels (p=0.048) compared with those with normal immunogenicity (table 4). Patients with active SLE (SLEDAI≥4) also had a numerically higher rate of low immune response compared with those with inactive disease (30.6% vs 17.9%), with an OR of 2.02 (p=0.20).

Table 4. The clinical differences between patients with SLE who had normal and low immunogenicity against SARS-CoV-2.

Data	Normal immunogenicity N=71	Low immunogenicity N=21	P value*
Demographics
Female; n (%)	68 (95.8)	20 (95.2)	0.99
Age (years); mean±SD	38.0±13.1	40.4±14.9	0.48
Age at onset (years); mean±SD	28.0±13.3	27.9±13.9	0.97
Disease duration (years); mean±SD	10.0±6.3	12.6±8.4	0.14
BMI (kg/m2); mean±SD	23.2±4.5	23.7±4.9	0.67
Comorbidity; n (%)
Diabetes	8 (11.3)	1 (4.8)	0.68
Hypertension	27 (38.0)	11 (52.4)	0.24
Dyslipidaemia	25 (35.2)	8 (38.1)	0.81
Chronic kidney disease (eGFR≤60 mL/min/1.73 m²)	6 (8.5)	5 (23.8)	0.06
Obesity	13 (18.3)	3 (14.3)	0.99
Cancer	3 (4.2)	1 (4.8)	0.99
Laboratory
Haemoglobin (g/L); median (IQR)	120.0 (112.0–131.0)	121.0 (102.0–130.0)	0.80
White blood cell (×10⁹/L); mean±SD	6.7±3.4	6.9±2.9	0.88
Platelet (×10⁹/L); mean±SD	259.8±81.2	267.1±62.8	0.71
Creatinine (mg/dL); median (IQR)	0.7 (0.6–0.8)	0.8 (0.6–1.2)	0.003*
Albumin (g/dL); mean±SD	4.1±0.4	3.9±0.5	0.06
Globulin (g/dL); mean±SD	3.4±0.6	2.9±0.9	0.048*
ALT (IU/L); mean±SD	21.8±18.4	18.0±12.1	0.38
AST (IU/L); mean±SD	24.7±13.8	22.2±7.3	0.43
ALP (IU/L); mean±SD	90.8±65.1	76.4±30.1	0.34
ESR (mm/hour); mean±SD	43.1±21.2	74.7±30.1	0.10
CRP (mg/dL); median (IQR)	2.7 (0.4–44.5)	15.3 (2.5–28.1)	0.83
Complement 3 level: n (%)	N=17	N=4	0.29
Low (<90)	6 (35.3)	1 (25)
Normal†	11 (64.7)	3 (75)
Complement 4 level: n (%)			0.58
Low (<10)	5 (29.4)	2 (50)
Normal†	12 (70.6)	2 (50)
UPCI (mg/mg); median (IQR)	0.2 (0.1–0.4)	0.3 (0.1–1.2)	0.11
APS; n (%)
Asymptomatic APS	0 (0)	2 (1.5)	0.99
Symptomatic APS	6 (8.5)	1 (4.8)	0.99
Disease activity (SLEDAI-2K)			0.20
Inactive (<4); n (%)	46 (64.8)	10 (47.6)
Active (≥4); n (%)	25 (35.2)	11 (52.4)
SLE involvement; n (%)
Active lupus nephritis	13 (18.3)	9 (42.9)	0.02*
Autoimmune haemolytic anaemia	5 (7.0)	3 (14.3)	0.38
Thrombocytopenia	1 (1.4)	0	0.99
Leucopenia	4 (5.6)	2 (9.5)	0.62
Fever; BT≥38°C (excluding infection)	1 (1.4)	0	0.99
Arthritis	2 (2.8)	0	0.99
Myositis	0	0	1.00
Cutaneous vasculitis	1 (1.4)	1 (4.8)	1.00
Discoid lupus	4 (5.6)	1 (4.8)	0.99
Alopecia	0	0	1.00
Oral/nasal ulcer	0	0	1.00
Cardiopulmonary involvement	0	0	1.00
Gastrointestinal involvement	0	0	1.00
Neuropsychiatric involvement	0	0	1.00
Concomitant medications
Prednisolone dose (mg/day); median (IQR)	5.0 (2.5–10.0)	10.0 (5.0–10.0)	0.12
Cyclophosphamide dose (mg/month); median (IQR)	500.0 (500.0–500.0)	–
Mycophenolate mofetil dose (mg/day); mean±SD	1336.0±709.7	1295.5±497.7	0.36
Ciclosporin A dose (mg/day); median (IQR)	200.0 (100.0–200.0)	100.0 (100.0–100.0)	0.75
Azathioprine dose (mg/day); median (IQR)	37.5 (15.0–72.0)	125.0 (50.0–200.0)	0.13
Hydroxychloroquine/chloroquine dose (mg/day); mean±SD	160.9±66.8	143.0±81.4	0.36
Outcome
SARS-CoV-2 IgG level (U/mL); median (IQR)	375 (126.3–1418.0)	0.6 (0.3–7.5)	0.41

^*Statistically significant.

^†Reference ranges: C3=90–180 mg/dL; C4=10–40 mg/dL.

ALP, alkaline phosphatase; ALT, alanine transaminase; APS, antiphospholipid syndrome; AST, aspartate transaminase; BMI, body mass index; BT, body temperature; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; ESR, erythrocyte sedimentation rate; SLEDAI-2K, the Systemic Lupus Erythematosus Disease Activity Index 2000; UPCI, Urine Protein Creatinine Index.

In univariate analysis, active lupus nephritis (crude OR 3.35, p=0.024) and prednisolone >7.5 mg/day (crude OR 3.78, p=0.023) was significantly associated with low immunogenicity, while mycophenolate mofetil showed a borderline association (crude OR 2.81, p=0.071). However, none remained significant in multivariable analysis (all adjusted ORs (aORs) p>0.05). Although aORs for most risk factors were >1 and <1 for globulin (aOR 0.59), the wide CIs crossing 1.0 suggest statistical uncertainty, possibly due to confounding or limited power in subgroup analyses (table 5).

Table 5. Factors associated with low immunogenicity against SARS-CoV-2 by multivariable analysis.

Factors	Crude OR (95% CI)	P value	Adjusted OR (95% CI)	P value
Every 1 mg/dL increase in creatinine level	2.76 (0.87 to 8.69)	0.084	1.92 (0.63 to 5.90)	0.250
Every 1 g/dL increase in globulin level	0.34 (0.05 to 2.26)	0.267	0.59 (0.11 to 3.06)	0.531
Active lupus nephritis	3.35 (1.17 to 9.59)	0.024	1.48 (0.21 to 10.39)	0.693
Prednisolone dose >7.5 mg/day	3.78 (1.20 to 11.87)	0.023	2.53 (0.45 to 14.38)	0.294
Mycophenolate mofetil	2.81 (0.91 to 8.59)	0.071	2.56 (0.48 to 13.84)	0.274

Discussion

This is the first study that evaluated SARS-CoV-2 IgG levels categorising them by the vaccine regimens in SLE compared with healthy controls after receiving two doses of available vaccine regimens in Thailand.

Our study included 92 patients with SLE and 41 healthy controls who were tested for anti-SARS-CoV-2 IgG levels after the second dose of SARS-CoV-2 vaccination following the National Health Thailand programme. We aimed to assess whether Thai patients with SLE have a comparable immune response using the IgG level against SARS-CoV-2 as an indicator of the immune response to healthy controls, and to explore whether vaccine regimen or disease-related factors influenced the immunogenicity profile.

In this study, we found that the immunogenicity of the SARS-CoV-2 vaccine between 28 and 90 days after the second dose of COVID-19 vaccination in patients with SLE was effective, with no statistically significant difference compared with healthy controls. Only 21 out of 92 patients with SLE (22.8%) had low immunogenicity, defined by the IgG antibody level of <15 U/mL using the ECLIA technique. There is currently no universally accepted cut-off for defining low immunogenicity to SARS-CoV-2 vaccination. We selected a threshold of <15 U/mL based on internal validation data from the Elecsys Anti-SARS-CoV-2 S assay, which showed that antibody levels ≥15 U/mL have a high positive predictive value (about 100%) for neutralising activity against both live-virus and surrogate assays. Moreover, prior studies, including those by Khoury et al and Gilbert et al, have demonstrated that antibody levels <100–150 U/mL are associated with reduced protection.^{34 35} Our cut-off of <15 U/mL represents a conservative estimate identifying patients with minimal protective immunity. Similar thresholds have been used in other studies to define low responders.³⁶ This finding is consistent with a previous study by Mormile et al, which evaluated anti-trimeric spike protein-specific IgG antibodies to SARS-CoV-2 after 21–28 days following the second dose of the Pfizer vaccine in 41 patients with SLE compared with healthy controls. Mormile et al found a positive serological response in 37 out of 41 patients (90.24%), with mean antibody levels of 3273.40±3679.57 in patients with SLE and 3443.97±2753.72 in healthy controls, showing no significant difference (p=0.825).³⁷

However, results from other patient with SLE cohorts have shown different findings. A recent study by Petri et al analysed the SARS-CoV-2 spike protein IgG levels on day 14 after receiving two doses of Moderna or Pfizer vaccinations in 334 patients with SLE compared with 1887 healthcare workers. The authors reported that patients with SLE had lower levels than healthcare workers (7.4±3.2 vs 8.9±1.4, p<0.0001).²⁰ A similar result was also reported by Izmirly et al who studied IgG antibodies to the SARS-CoV-2 spike RBD at 2 weeks after receiving two doses of Moderna or Pfizer vaccinations in 90 patients with SLE compared with 20 healthy controls and found that 26 from 90 patients with SLE (28.8%) had a low response to the COVID-19 vaccine, defined by the IgG level <100 units/mL, which was associated with receiving any immunosuppressive agents other than antimalarials prior to vaccination.¹⁹

In addition to comparing the immune response between patients with SLE and healthy controls, we also investigated the IgG level against SARS-CoV-2, categorising them based on the vaccine regimens used in Thailand due to the diversity of available SARS-CoV-2 vaccine regimens. This allowed us to assess whether different vaccine regimens resulted in varying levels of IgG antibodies in both patients with SLE and healthy individuals. The most common regimen of SARS-CoV-2 vaccine among our patients with SLE was Pfizer and Pfizer.

When analysing the IgG level against SARS-CoV-2 by categorising them based on the vaccine regimens, our study found that mRNA-based vaccines (Moderna, Pfizer) produced a higher immune response compared with non-replicated viral vector vaccine (AstraZeneca) and inactivated viral vaccines (SP, SV), which is consistent with previous studies. So and colleagues investigated the level of SARS-CoV-2-specific neutralising antibody at 28 days after receiving two doses of the SV or Pfizer vaccine in 65 patients with SLE compared with 50 healthy controls. The results showed that 92% of patients with SLE had positive neutralising antibodies, while 100% of healthy controls were seropositive (p=0.058). Furthermore, compared with controls, the neutralising antibody levels were significantly lower (64.4±22.9% vs 83.0±18.0%, p<0.001) in patients with SLE. In subgroup analyses according to the type of vaccine received, the level of neutralising antibody was significantly lower (mean 46.6% vs 77.0%, p<0.001) in the SV group compared with the Pfizer group.³⁸

In our study, the regimen with the highest immune response, based on the median of SARS-CoV-2 spike protein IgG level, was Moderna-Moderna. Importantly, vaccine regimens in this study were not assigned by investigators but reflected real-world vaccine availability during Thailand’s national vaccination campaign, thereby enhancing the generalisability of our findings. However, as with any observational study, unmeasured confounding cannot be entirely excluded. Conversely, the lowest median IgG levels were observed in patients with SLEs who received the SP and SP regimen. This finding aligns with a study by Zamani et al, which evaluated the efficacy of the SP vaccine by determining the level of SARS-CoV-2 neutralising antibodies between 1 and 3 months after the second vaccination in autoimmune inflammatory rheumatic diseases (AIIRD), including patients with SLE (26 out of 100), compared with 100 healthy controls. The study reported significantly lower SARS-CoV-2 neutralising antibody levels in AIIRD compared with healthy controls, with values of 60.5% vs 72.8% and 17.58±20.2 µ/mL vs 22.87±20.9 µ/mL, respectively (p<0.01).³⁹ In addition, Yuki and colleagues studied anti-SARS-CoV-2 S1/S2 IgG antibodies at 6 weeks after the second dose of the SV vaccine in 215 patients with SLE compared with 53 healthy controls. They found that patients with SLEs had a lower immune response when receiving two doses of SV compared with healthy controls, with a median of anti-SARS-CoV-2 S1/S2 IgG antibody level of 29.6 (24.8–35.4) versus 77.0 (64.5–91.8), respectively (p<0.001).²⁶ The markedly lower IgG response observed among recipients of the inactivated vaccine (SP–SP) was consistent with prior studies showing inferior immunogenicity compared with mRNA platforms.^{40 41} Notably, the approximately 10-fold gap seen in our SLE cohort is larger than previously reported, a difference that may reflect host-related factors such as active lupus nephritis, higher corticosteroid exposure and concurrent immunosuppressants, which can synergistically attenuate vaccine-induced antibody production. Although our overall findings remain reliable, some regimens (eg, SV–SV and SP–SV) had very small sample sizes (≤3 participants), which limits the precision of antibody-response estimates; therefore, comparisons involving these regimens should be considered exploratory.

Similar to previous reports by Petri et al and Yuki et al,^{20 26} our study found no statistically significant difference in vaccine response between patients with active and inactive SLE. While this difference did not reach statistical significance, the observed trend toward reduced immunogenicity in patients with active disease may suggest a potential association between disease activity and impaired vaccine response. Further studies with larger cohorts are warranted to confirm this relationship.

In our multivariate analysis, no factors were significantly associated with low immunogenicity against SARS-CoV-2. However, some factors showed adjusted ORs suggestive of potential associations, including prednisolone dose >7.5 mg/day (aOR 2.53, 95% CI 0.45 to 14.38), use of mycophenolate mofetil (aOR 2.56, 95% CI 0.48 to 13.84) and active lupus nephritis (aOR 1.48, 95% CI 0.21 to 10.39). Although not statistically significant, these associations are consistent with prior studies reporting reduced vaccine responses in patients with active lupus nephritis or those receiving moderate to high-dose corticosteroids or mycophenolate mofetil. The wide CIs likely reflect limited statistical power, particularly within these subgroups. Because of the clinical correlation among variables such as active lupus nephritis, prednisolone use and mycophenolate mofetil, residual collinearity may have weakened the associations observed in the multivariable model. Tang et al found that patients with AIIRD on prednisolone had a lower seroconversion rate following COVID-19 vaccination.⁴² Similarly, Izmirly et al reported that patients receiving an average of 7.2 mg/day of prednisolone had a lower immune response to the COVID-19 vaccine in patients with SLE.¹⁹ Although our study did not detect a statistically significant association, the adjusted OR observed for prednisolone >7.5 mg/day aligns with these prior findings.

Regarding mycophenolate mofetil, our study did not find a statistically significant association with lower immune response, which contrasts with findings by Ruddy et al, who investigated the IgG antibody level against the RBD of the SARS-CoV-2 spike protein (anti-RBD) after the second dose of mRNA vaccine in 404 patients with rheumatic and musculoskeletal diseases, including 87 with SLE. They found that the anti-RBD level was lower in participants on mycophenolate.⁴³ Similarly, Petri et al found that doses of mycophenolate >1000 mg/day were associated with reduced SARS-CoV-2 IgG levels.²⁰ Additionally, Izmirly et al reported that mycophenolate mofetil was associated with lower immunogenicity to the COVID-19 vaccine in patients with SLE. A potential explanation for this discrepancy is the lower average mycophenolate mofetil dose in our study population (mean 1321.6±634.0 mg/day) compared with that reported by Izmirly et al (1967±731.1 mg/day). These findings suggest that dosage intensity may play a key role in modulating vaccine responses.¹⁹

Active lupus nephritis was also associated with reduced immunogenicity against the SARS-CoV-2 vaccine (aOR 1.48, 95% CI 0.21 to 10.39), although this was not statistically significant. Notably, this finding differs from the study by Izmirly et al, which did not identify a significant association between lupus nephritis and vaccine response. However, their analysis included patients with any history of nephritis, while our study focused specifically on those with active nephritis at the time of vaccination, which may partly explain the discrepancy.¹⁹ This is consistent with the immunopathogenic basis of lupus nephritis, in which high disease activity and immunosuppressive treatment may impair humoral immune responses. Our study also suggested a possible, though non-significant, association between elevated serum creatinine and reduced immunogenicity (aOR 1.92, 95% CI 0.63 to 5.90). This association may reflect the severity of underlying disease activity, particularly in patients with active lupus nephritis, rather than renal dysfunction per se. Interestingly, our results differ from those of Trakarnvanich et al, who reported comparable immunogenicity following COVID-19 vaccination between patients with chronic kidney disease (CKD) and healthy controls. Their study assessed anti-RBD IgG levels 12 weeks after the second dose of mostly AstraZeneca vaccine in 55 patients with CKD (median creatinine 1.67 mg/dL) versus 22 healthy controls (median creatinine 0.75 mg/dL), with seroconversion observed in 100% of patients with CKD and 92.31% of controls. Their findings suggest that renal dysfunction alone may not impair vaccine response, highlighting the possible role of active autoimmune disease in modulating immunogenicity. Moreover, unlike CKD, lupus nephritis involves systemic immune dysregulation, which could contribute to suboptimal vaccine responses.⁴⁴

Additionally, our study revealed that serum globulin levels had an aOR <1 (aOR 0.59, 95% CI 0.11 to 3.06), suggesting a possible association between higher globulin levels and higher humoral responses. This may be biologically plausible, as serum globulins reflect immune status and contain components such as immunoglobulins (γ-globulins), complement (mainly β-globulins) and acute-phase proteins (primarily α-globulins) that contribute to antibody-mediated protection.⁴⁵

SLE pathophysiology involving dysregulation of type I IFN, abnormal B-cell responses and T-cell exhaustion may contribute to altered vaccine-induced immunity. In particular, impaired germinal centre formation and the presence of autoreactive B cells could compromise the development of high-affinity antibodies. These immunological features may explain why certain subsets of patients with SLE, such as those with active nephritis or under high-dose immunosuppression, demonstrated suboptimal IgG responses in our study.

Our findings emphasise the need for further studies evaluating both humoral and cellular immunity in this population. Additionally, stratified vaccine strategies, potentially including booster timing or mRNA-based platforms, should be considered for patients with SLE with high disease activity or receiving intensive immunosuppression.

To place our findings within the broader literature, we added a concise comparison of prior studies of SARS-CoV-2 vaccination in SLE (table 6). Whereas most prior studies report reduced immunogenicity in patients with SLE compared with controls, especially among patients on immunosuppressants, our Thai cohort showed a relatively comparable response, although the difference did not reach statistical significance. Differences in vaccine type, population characteristics and timing may explain these discrepant findings.

Table 6. Comparison of studies of SARS-CoV-2 vaccination in SLE.

Study	Country	Vaccine(s)	Study population	Key findings
This study	Thailand	Inactivated, viral vector, mRNA	92 SLE vs 41 controls	Comparable response to controls; mRNA>vector>inactivated. Lower response with active nephritis, renal impairment, higher dose steroids and MMF (non-significant).
So et al, 202238	Hong Kong	Inactivated, mRNA	65 SLE vs 50 controls	Lower response in SLE; mRNA>inactivated. MMF and steroids reduced immunogenicity.
Ruddy et al, 202143	USA	mRNA	91 SLE (from 404 RMD)	Lower response in SLE. MMF, steroids and rituximab reduced immunogenicity.
Izmirly et al, 202219	USA	mRNA, viral vector	90 SLE vs 20 controls	Lower response in SLE. MMF and steroids reduced immunogenicity.
Petri et al, 202320	USA	mRNA	334 SLE vs 1887 controls	Lower response in SLE. MMF, tacrolimus and belimumab reduced immunogenicity.
Mormile et al, 202237	Italy	mRNA	41 SLE vs 29 hereditary angioedema	No significant difference between SLE and controls.
Yuki et al, 202226	Brazil	Inactivated	215 SLE vs 53 controls	Lower response in SLE. mRNA induced stronger responses. MMF and steroids reduced immunogenicity.
Zamani et al, 202339	Iran	Inactivated	26 SLE (from 100 AIIRD) vs 100 controls	Lower response in SLE. MMF and high-dose steroids associated with reduced immunogenicity.

AIIRD, autoimmune inflammatory rheumatic diseases; MMF, mycophenolate mofetil; mRNA, messenger RNA; RMD, rheumatic and musculoskeletal diseases.

Our study has some limitations. First, it was conducted at a single centre in Thailand, which may restrict generalisability. Second, baseline anti-SARS-CoV-2 antibodies were not measured, so prior asymptomatic infection cannot be excluded. Third, we lacked post-vaccination clinical outcome data, precluding correlation of antibody levels with protection. Fourth, real-world vaccine distribution resulted in small sample sizes for certain regimens (≤3 participants), reducing power for between-regimen comparisons and for multivariable analyses of patients with active lupus nephritis or higher doses of corticosteroid use. Additionally, detailed phenotype data such as anti-dsDNA antibody levels (available in only 3 of 92 patients) and lupus nephritis histology (available in only 18 of 92 patients) were not systematically collected. In our setting, kidney biopsy is typically reserved for suspected severe, refractory or relapsing cases due to cost and resource constraints, which may explain the predominance of proliferative lupus nephritis (Class IV) among biopsied patients. This limits the representativeness of biopsy-based analyses. Furthermore, comorbidity data, healthcare exposure and socioeconomic status were not available in the matched healthy control group, potentially influencing immune response and introducing unmeasured confounding. These limitations restrict the comprehensiveness of disease-related immunogenicity analysis and should be considered when extrapolating our findings to broader SLE populations.

Despite these limitations, our study has several notable strengths. First, we enrolled a relatively large cohort of Thai patients with SLE and included multiple vaccine regimens reflective of real-world practices, thereby increasing the generalisability and clinical relevance of the findings. Second, we employed age-matched and sex-matched healthy controls with appropriate statistical adjustment for repeated matching, which helps reduce confounding and strengthen internal validity. Collectively, these findings can help clinicians identify SLE subgroups who may require closer post-vaccination monitoring of SARS-CoV-2 immunogenicity.

Overall, the results suggest that most Thai patients with SLE have a comparable IgG response to healthy controls, despite immunosuppressive therapy. Notably, variability in IgG levels was observed across vaccine types, with mRNA-based vaccines eliciting stronger responses than inactivated vaccines. Although no statistically significant predictors of poor immunogenicity were identified, patients with active lupus nephritis or receiving higher dose corticosteroids appeared to show lower IgG responses and may potentially benefit from tailored strategies such as early booster administration. These results may also inform national health policy, including longitudinal monitoring after booster doses and incorporation of cellular immunity testing to better evaluate vaccine-induced protection in immunocompromised populations.

Conclusions

Thai patients with SLE demonstrated a comparable IgG response to healthy controls following SARS-CoV-2 vaccination. mRNA-based vaccine regimens showed a higher immune response compared with viral vector and inactivated vaccines. These findings support individualised vaccine strategies in patients with SLE to optimise protection against COVID-19.

Data availability statement

Data are available upon reasonable request.