Background
Systemic lupus erythematosus (SLE) is an autoimmune disease that afflicts mostly women, often in fertile ages [
1]. An aberrant activation in many aspects of the immune system has been documented for SLE, but common denominators for most patients include T cell-dependent autoantibody production and type I interferon (IFN) overexpression [
2‐
4]. The widely distributed immune activation is reflected by a diversity in laboratory abnormalities (including lymphopenia, hypocomplementemia and presence of autoantibodies) and in clinical features (including arthritis, renal and dermatological manifestations) [
4]. Moreover, SLE pregnancy carries a risk for disease flare and an increased risk of pregnancy complications compared to the general population [
5,
6].
Lymphopenia is common in individuals with SLE, occurring in about 40% of the patients [
7‐
9], and low absolute numbers of the lymphocyte subsets CD4 + T cells, CD8 + T cells, B cells and NK cells have been reported in patients with SLE compared to healthy individuals [
10‐
12]. In SLE, lymphopenia is independently associated with organ damage accrual, neurological involvement and disease activity [
8,
9,
13], but it is unknown whether specific lymphocyte subset numbers in blood are affected by pregnancy in SLE and if subset counts relate to disease activity during pregnancy.
Various autoantibodies are described in SLE [
14], but only a few are assessed routinely in the clinical setting. These include anti-nuclear antibodies (ANA) directed to double stranded DNA (dsDNA), Smith antigen (Sm), ribonucleoprotein (RNP), Sjögren’s syndrome antigen A (SSA) and B (SSB) and anti-phospholipid antibodies (aPL) directed to cardiolipin (CL) and β2-glycoprotein I (β
2GPI) [
15]. The heterogeneity of SLE has motivated attempts to stratify patients into subgroups based on disease-related autoantibody profiles in non-pregnant patients with SLE [
16‐
18]. A large international longitudinal study recently identified four serological clusters that differed in clinical features but also predicted long term events [
16]. The latter and other studies also report that positivity for most SLE-related autoantibodies decrease over time [
16,
19,
20]. In cross-sectional setting, one study separated SLE patients into three groups where two were dominated by anti-dsDNA-positive individuals who had a higher frequency of lymphopenia compared to the third group that included fewer anti-dsDNA-positive individuals [
18]. Another study revealed four groups where the first was dominated by anti-SSA/SSB positivity, the second by anti-Sm/dsDNA/RNP positivity, the third by aPL positivity and the last was seronegative [
17]. Lymphopenia was more frequent and disease activity higher in the seropositive groups compared to the seronegative group [
17]. It is still unknown if counts of specific lymphocyte subsets including T cells, B cells and NK cells relate to different disease-related autoantibody profiles, and if they relate to each other during pregnancy in SLE.
Many patients with SLE present with increased expression of type I interferon (IFN)-regulated genes in blood cells and in tissue, an IFN signature [
2,
3], and a single cell RNA sequencing analysis of PBMC showed that increased expression of type I IFN regulated genes in monocytes correlate with low naïve CD4 + T cells in SLE [
21]. Individuals with SLE also have higher IFNα protein concentrations in blood compared to healthy subjects as demonstrated by the use of an ultrasensitive single molecule array (Simoa) digital enzyme-linked immunosorbent assay (ELISA) [
22]. In SLE, Simoa-quantified IFNα protein levels strongly correlate with a whole blood IFN-I gene score, and these methods identify associtations with SLE disease activity equally well [
23]. Using the digital ELISA technology, we reported that IFNα protein positivity is present in a subgroup of pregnant women with SLE, but the protein concentrations are similar during pregnancy and in the late postpartum period [
24]. Additionally, we and others have reported that pregnant and non-pregnant SLE patients positive for anti-SSA antibodies have increased IFNα protein levels in blood [
24,
25]. Yet, it is unknown if lymphocyte subset counts relate to IFNα protein levels in pregnant women with SLE.
The first aim of this study was to compare total CD4 + T cell, CD8 + T cell, B cell and NK cell counts prospectively throughout pregnancy in women with SLE relative to the late postpartum period and to healthy pregnant women. Secondly, we aimed to investigate whether the lymphocyte subset counts were related to autoantibody profiles, IFNα protein levels, disease activity and gestational age at birth in SLE pregnancy.
Materials And Methods
Cohort
This Swedish multicenter study enrolled pregnant women with SLE (
n = 80) meeting the 1997 American College of Rheumatology (ACR) and/or the 2012 Systemic Lupus International Collaborating Clinics (SLICC) classification criteria [
26,
27] between November 2018 and June 2022 at Rheumatology clinics in: Gothenburg (Sahlgrenska University hospital,
n = 24), Stockholm (Karolinska University Hospital,
n = 38), Uppsala (Uppsala University Hospital,
n = 3), Linköping (Linköping University Hospital,
n = 6) and Lund (Skåne University Hospital,
n = 9). Healthy pregnant women (HC,
n = 51) were enrolled at one antenatal clinic in Gothenburg (Regionhälsan, Gothenburg) between October 2018 and December 2022. Most pregnant women with SLE and HC were included at 10–12 weeks of gestation and followed in the second (week 18–20) and third (week 32–34) trimester. Disease activity was evaluated according to the SLE Disease Activity Index 2000 (SLEDAI-2 K) [
28], and measured at least once between week 10 and week 34 and if the disease activity was assessed more than once, the highest score was used in analyses. The number of pregnant women with SLE for whom SLEDAI-2 K assessments were obtained from in each trimester is shown in Supplementary Table
1. Clinical data including disease duration, medication, ever autoantibody positivity, gestational age at birth and giving birth to a small for gestational age (SGA) infant were retrieved from medical records. SGA (
n = 13) was defined as birth weight less than the 10th percentile for expected birth weight [
29]. Exclusion criteria were inability to understand the study-related patient information and informed consent form, presence of other serious disease, including active cancer and other rheumatic autoimmune diseases, or treatment with anti-BAFF or anti-CD20 antibodies within 12 months before inclusion. Women who had miscarriage during trimester one were excluded. None of the women with SLE were treated with anifrolumab before or during their pregnancy. All participants have given their written informed consent and the Ethics board in Gothenburg (Dnr 404–18) and The Swedish Ethical Review Authority (amendments Dnr 2020–05101 and Dnr 2022–01158-02) have approved the study.
Sample collection
Peripheral blood samples were collected in heparinized tubes from pregnant women with SLE and HC in the first, second and third trimester. One additional blood sample was collected late postpartum from 19 of the women with SLE (at least six months after delivery (median 10 [6,–36] months). Information about the number of blood samples collected for each trimester is presented in Supplementary Table
2. All blood samples were kept at ambient temperature until processed the day after, within 24 h after venipuncture at our laboratory in Gothenburg. Whole blood was used for flow cytometry analysis of total lymphocyte subset counts. Density centrifugation of whole blood was performed to isolate plasma that was kept frozen (-80 °C) until further analysis.
Autoantibody status during pregnancy in SLE
Analysis of positivity for immunofluorescence (IF)-ANA, for ANA fine specificities including antibodies to dsDNA, Sm, RNP, SSA, SSB, chromatin and ribosomal P protein, and for anti-phospholipid antibodies including anti-CL and anti-β2GPI in plasma collected during pregnancy was performed at the accredited laboratory of Clinical Immunology, Sahlgrenska University Hospital. Most of these samples were collected in trimester three and plasma was diluted 1:1 in PBS. IF-ANA was analyzed using Hep-2 cells according to routine, and 66 out of 80 (83%) women with SLE were positive. ANA fine specificities were analyzed using multiplexed bead technology by the BioPlex™ 2200 System (BioRad Laboratories, Hercules, USA). The cut-off for most ANA-specificities was 1.0 AI (antibody index) except for anti-chromatin (1.5 AI), anti-RNP (3.0 AI) and anti-dsDNA (10 IU/mL). Positive ANA-specificities were confirmed with another method according to the manufacturers recommendation: Crithidia luciliae test for anti-dsDNA (ImmunoConcept, Sacramento, CA), automated ELISA-based test system Alegria® (Orgentec Diagnostics, Mainz, Germany) for anti-SSA52 and line blot ANA Profile 5 IgG for all other ANA-specificities (Euroimmun, Lübeck, Germany). Anti-CL and anti-β2GPI of IgG isotype were examined using the BioPlex™ 2200 multiplex immunoassay system and APLS reagents. Cut-off values for positivity were 20 GPL for aCL IgG and 20 AU/mL for anti-β2GPI IgG as recommended by the manufacturer. Ever autoantibody status was obtained from medical records and included positivity for anti-dsDNA/Sm/SSA/SSB, anti-CL/β2GPI and lupus anticoagulant (LAC). The method for analysis of ever antibody positivity differed between the study sites and dsDNA was confirmed by either Crithidia luciliae test or ELISA. Positivity was determined according to cut-off levels at the local laboratories.
Flow cytometry
TruCount™ assay was used to analyze total number of lymphocytes, CD4 + T cells, CD8 + T cells, B cells and NK cells in whole blood. In brief, blood and antibodies against CD45, CD3, CD4, CD8, CD19, CD20 and CD56 (Supplementary Table
3) were added to BD TruCount™ tubes (BD Bioscience) and incubated for 15 min. Red blood cells were then lysed using BD FACS™ Lysing solution (BD Sciences). All samples were acquired in a FACSVerse equipped with FACSuite Software (BD Biosciences) and analyzed with FlowJo Software (TreeStar, Ashland, Oregon, USA).
IFNα protein quantification
The concentration of IFNα protein levels in plasma diluted 1:1 in PBS was quantified with Single molecule array (Simoa) digital ELISA on a HD-X Analyzer (Quanterix, Billerica, MA). To prevent false positive results the Simoa assay contained an inhibitor for heterophilic antibodies. If the concentration in a sample was below the lower limit of quantification (70 fg/ml) its value was adjusted to 35 fg/ml. IFNα positivity was defined as protein levels ≥ 136 fg/ml, representing three standard deviations above mean IFNα protein concentration among healthy blood donors [
30].
Statistical analysis
Multivariate data analysis was performed using the SIMCA-P software (Sartorius Stedem Biotech, Goettingen, Germany). Principal Component Analysis (PCA) was used to obtain an unsupervised descriptive overview of groupings and trends, associations, between total number of CD4 + T cells, CD8 + T cells, B cells and NK cells, IFNα protein levels, autoantibody positivity during pregnancy, disease activity, gestational age at birth and SGA among pregnant women with SLE. Orthogonal partial least squares (OPLS) analysis was performed to investigate medication or gestational age at birth (Y-variables) in relation to total numbers of T cells, B cells and NK cells (X-variables) in pregnant women with SLE. In the PCA and OPLS models default settings were used; data were centered and scaled to unit-variance to give all variables equal weight. Model quality was based on R2 and Q2 parameters that are presented in each figure. Univariate analyses were only performed for the strongest associations found in the PCA and OPLS models and included Kruskal–Wallis followed by Dunn’s multiple comparison test, Mann–Whitney U test and Spearman rank correlation test (GraphPad prism software, La Jolla, CA, USA) as described in each respective figure legend. P-values of < 0.05 were considered statistically significant.
Discussion
To our knowledge, this is the first study to investigate if pregnancy affects blood lymphocyte subset counts in SLE, and if T cell, B cell and NK cell counts associate to disease-related autoantibody positivity and/or to IFNα protein concentrations during SLE pregnancy. In this longitudinal study we confirm that low total lymphocyte count is evident throughout pregnancy in women with SLE compared to HC, a well-known feature in non-pregnant patients relative to controls [
34]. We also show that none of the lymphocyte subset numbers were affected by pregnancy in SLE. This contrasts with healthy pregnant women who have lower numbers of lymphocytes during pregnancy compared to the postpartum period [
35,
36]. The explanation for this discrepancy could only be speculated upon, but disease-related homing of activated lymphocytes from the periphery to inflamed tissues and organs and the presence of autoantibodies with lymphocytotoxic activity may result in low lymphocyte counts that are not further affected by pregnancy in SLE [
37,
38]. Although the activation status of the different lymphocyte subsets was not examined here, we have previously reported that pregnancy in SLE results in increased activation of circulating granulocytes compared to the late postpartum period [
24].
Although SLE is a heterogenic disease with a variety in laboratory abnormalities and clinical features, recent detailed serological profiling has identified sets of disease-related autoantibodies that commonly occur together [
16,
17,
39]. In accordance we here show for the first time in SLE pregnancy that autoantibody positivity also separated into three clusters, the first dominated by positivity for anti-dsDNA/anti-Sm/anti-RNP/anti-chromatin, the second by anti-SSA/anti-SSB and the third by anti-CL/anti-β
2GPI. Another novel finding was that low numbers of particularly CD4 + T cells, but also B cells, CD8 + T cells and NK cells, relate to positivity for ANA specificities in cluster one, but not ANA positivity in cluster two or aPL positivity in cluster three. Whether specific lymphocyte subset counts relate to autoantibody positivity profiles has not been examined in non-pregnant patients with SLE. The use of antibody clustering to separate patients with SLE into endotypes may help to predict disease course and prognosis as patients with distinct autoantibody profiles differ with regards to immunological variables, clinical manifestations, treatment, organ involvement and long-term disease activity [
16‐
18,
40].
Lymphopenia is associated with SLE-specific anti-dsDNA positivity, SLE-related autoantibody positivity in general, and more severe/progressive disease in non-pregnant subjects with SLE [
9,
17,
18]. When we here divided total lymphocytes into subsets, only CD4 + T cell and B cell counts were lower in pregnant women positive to anti-dsDNA compared to those who were negative and counts of these subsets were also lower in women with moderate/high (SLEDAI-2 K ≥ 4) compared to no/low disease activity. Accordingly, anti-dsDNA positivity, being part of the SLEDAI-2 K score [
28], was also related to higher disease activity. Anti-Sm is another SLE-specific autoantibody, but its pathologic significance is uncertain and there are conflicting data regarding its association to disease activity and clinical manifestations including lymphopenia [
41‐
43]. We found that all lymphocyte subset counts were lower in pregnant women who were positive for anti-Sm relative to those who were negative, but anti-Sm positivity was unrelated to disease activity in our cohort. A higher number of ANA fine specificities also related to lower numbers of lymphocyte subset counts, but not to disease activity. Still, a higher number of ANA specificities could indicate a more immunologically active disease that leads to lymphocyte homing to inflamed tissue and organs, which is not captured by the SLEDAI-2 K index. Whether there is a causal relationship between low lymphocyte subsets counts and specific ANA positivity is not answered by the present data, but we add novel knowledge on how numbers of specific lymphocyte subsets in blood differ in relation to antibody positivity profiles and disease activity in SLE pregnancies.
Lower total lymphocyte counts have also been reported in IFNα positive compared to negative non-pregnant patients with SLE [
30]. We found that specifically CD4 + T cell counts, but not counts of any other lymphocyte subset, were lower in IFNα-positive pregnant women with SLE compared to those who were negative. Our finding is in line with recent scRNA-seq data showing that a reduction of naïve CD4 + T cells correlates with increased expression of type I IFN regulated genes in monocytes in non-pregnant individuals with SLE [
21]. Whether there is a direct effect of IFNα on CD4 + T cell numbers in blood is unclear, but administration of IFNα in healthy volunteers leads to a drastic decrease of total lymphocyte numbers in blood [
44,
45]. Mouse models suggest a partial mechanistic explanation for this phenomenon by inhibited egress of CD4 + T cells, CD8 + T cells and CD19 + B cells from lymph nodes, as treatment with the IFNα inducer poly (I:C) retains lymphocytes in lymph nodes via regulation of CD69 and sphingosine 1-phosphate receptor-1 (S1P1) expression [
46]. Still, the relationship between IFNα and low CD4 + T cell counts remains to be examined further.
Medication may affect numbers of lymphocyte subsets in blood. Indeed, pregnant women with SLE who were treated with azathioprine had lower numbers of NK cells and B cells compared to women who were not treated. In accordance with this, azathioprine use is related to reduced numbers and proportions of NK cells and B cells in non-pregnant subjects with SLE, inflammatory bowel disease or ANCA-associated vasculitis [
21,
47‐
50]. A proposed mechanism for azathioprine-related decrease in NK cells is caspase 3- and 9-induced apoptosis [
49]. Additionally, we also found lower B cell counts in women treated with prednisone or heparin compared to those who were not. For prednisone, similar results were reported from a small cohort of patients with different autoimmune disorders [
51], and in a small cohort of healthy volunteers [
52]. Importantly, we also found a treatment-independent decrease in both NK cells and B cells in SLE compared to HC pregnancies.
A strength of the study is the inclusion of well-characterized patients and controls from whom samples have been prospectively collected in parallel during pregnancy and late postpartum. Others are that all flow cytometry analyses were performed on fresh blood in one laboratory on the same instrument and cross-sectional analysis of autoantibody positivity was analyzed with well-standardized methods by staff in an accredited hospital laboratory. Our study also has limitations. The cohort included few women with moderate or high SLE disease activity and therefore our results reflect a well-controlled cohort of pregnant women with SLE. Moreover, the study includes missing data, mainly due to missing blood samples from the first trimester, among pregnant women with SLE.
To conclude, we report that pregnancy in women with SLE has no effect on blood lymphocyte subset counts but SLE pregnancies are featured by a treatment-independent decrease in blood lymphocyte counts compared to healthy pregnant women. Moreover, low counts of specific lymphocyte subsets relate differentially to disease-related autoantibody positivity, IFNα protein levels and disease activity in SLE pregnancy. Still, further studies are needed to decipher the immunological characteristics of SLE phenotypes based of antibody profiles in more detail, and to investigate if specific subgroups are related to an increased risk for pregnancy complications.
Acknowledgements
We thank the study nurses Anita Nihlberg, Maria Andersson, Sonia Möller, Hans Kling, Eva Malmqvist, Rezvan Kiani, Marianne Petersson, Anna-Lena Åblad, Elisabeth Kling and Lena Pålsson at the different rheumatology clinics, the midwifes Charlotta Jansson and Charlotte Andersson and the assistant nurse Anette Svensk at the Antenatal clinic in Gothenburg as well as all personnel at the delivery units. For assistance with the collection of clinical data, we thank Kristina Karlsson, Martina Wahlberg and Ylva Folkesson. We thank Gunilla Larsson for proficient laboratory assistance. We also want to thank Rille Pullerits, Cecilia Larsson and Robert Breiðfjörð at the laboratory of Clinical Immunology, Sahlgrenska University Hospital, for excellent technical assistance. At last, we wish to thank all the pregnant women who participated in the study.
Declarations
Competing interests
HZ has served at scientific advisory boards and/or as a consultant for Abbvie, Acumen, Alector, Alzinova, ALZPath, Annexon, Apellis, Artery Therapeutics, AZTherapies, CogRx, Denali, Eisai, Nervgen, Novo Nordisk, Optoceutics, Passage Bio, Pinteon Therapeutics, Prothena, Red Abbey Labs, reMYND, Roche, Samumed, Siemens Healthineers, Triplet Therapeutics, and Wave, has given lectures in symposia sponsored by Cellectricon, Fujirebio, Alzecure, Biogen, and Roche, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program (outside submitted work). KB has served as a consultant and at advisory boards for Acumen, ALZPath, BioArctic, Biogen, Eisai, Lilly, Moleac Pte. Ltd, Novartis, Ono Pharma, Prothena, Roche Diagnostics, and Siemens Healthineers; has served at data monitoring committees for Julius Clinical and Novartis; has given lectures, produced educational materials and participated in educational programs for AC Immune, Biogen, Celdara Medical, Eisai and Roche Diagnostics; and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program, outside the work presented in this paper. LR has collaboration with Astra Zeneca, Bristol Myers Squibb (BMS), AMPEL Biosolutions, UCB and Bayer. ES has received a research grant from Merck and lecture honoraria from Janssen. The other authors declare no conflict of interest.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.