Torell et al.

Arthritis Research & Therapy (2024) 26:65 Arthritis Research & Therapy
https://doi.org/10.1186/s13075-024-03301-0

RESEARCH Open Access

Low CD4 + T cell count is related to specific

anti‑nuclear antibodies, IFNα protein positivity
and disease activity in systemic lupus
erythematosus pregnancy
Agnes Torell1*, Marit Stockfelt1,2, Kaj Blennow3,4,5,6, Henrik Zetterberg10,3,4,7,8,9, Tansim Akhter11,
Dag Leonard12, Lars Rönnblom12, Sofia Pihl13,14, Muna Saleh15, Christopher Sjöwall15, Helena Strevens16,
Andreas Jönsen17, Anders A. Bengtsson17, Estelle Trysberg2, Maria Majczuk Sennström18, Agneta Zickert19,
Elisabet Svenungsson19, Iva Gunnarsson19, Johan Bylund20, Bo Jacobsson21,22,23, Anna Rudin1 and
Anna‑Carin Lundell1

Abstract
Background Lymphopenia, autoantibodies and activation of the type I interferon (IFN) system are common features
in systemic lupus erythematosus (SLE). We speculate whether lymphocyte subset counts are affected by pregnancy
and if they relate to autoantibody profiles and/or IFNα protein in SLE pregnancy.
Methods Repeated blood samples were collected during pregnancy from 80 women with SLE and 51 healthy con‑
trols (HC). Late postpartum samples were obtained from 19 of the women with SLE. Counts of CD4 + and CD8 + T cells,
B cells and NK cells were measured by flow cytometry. Positivity for anti-nuclear antibodies (ANA) fine specificities
(double-stranded DNA [dsDNA], Smith [Sm], ribonucleoprotein [RNP], chromatin, Sjögren’s syndrome antigen A [SSA]
and B [SSB]) and anti-phospholipid antibodies (cardiolipin [CL] and β2 glycoprotein I [β2GPI]) was assessed with mul‑
tiplexed bead assay. IFNα protein concentration was quantified with Single molecule array (Simoa) immune assay.
Clinical data were retrieved from medical records.
Results Women with SLE had lower counts of all lymphocyte subsets compared to HC throughout pregnancy,
but counts did not differ during pregnancy compared to postpartum. Principal component analysis revealed that low
lymphocyte subset counts differentially related to autoantibody profiles, cluster one (anti-dsDNA/anti-Sm/anti-RNP/
anti-Sm/RNP/anti-chromatin), cluster two (anti-SSA/anti-SSB) and cluster three (anti-CL/anti-β2GPI), IFNα protein
levels and disease activity. CD4 + T cell counts were lower in women positive to all ANA fine specificities in cluster one
compared to those who were negative, and B cell numbers were lower in women positive for anti-dsDNA and anti-
Sm compared to negative women. Moreover, CD4 + T cell and B cell counts were lower in women with moderate/
high compared to no/low disease activity, and CD4 + T cell count was lower in IFNα protein positive relative to nega‑
tive women. Finally, CD4 + T cell count was unrelated to treatment.

*Correspondence:
Agnes Torell
agnes.torell.bjorkman@gu.se
Full list of author information is available at the end of the article

© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecom‑
mons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 2 of 15

Conclusion Lymphocyte subset counts are lower in SLE compared to healthy pregnancies, which seems to be a fea‑
ture of the disease per se and not affected by pregnancy. Our results also indicate that low lymphocyte subset counts
relate differentially to autoantibody profiles, IFNα protein levels and disease activity, which could be due to divergent
disease pathways.
Keywords Systemic lupus erythematosus (SLE), Lymphocyte count, Pregnancy, Interferon alpha, Autoantibodies

Background frequency of lymphopenia compared to the third

Systemic lupus erythematosus (SLE) is an autoim- group that included fewer anti-dsDNA-positive indi-
mune disease that afflicts mostly women, often in fer- viduals [18]. Another study revealed four groups where
tile ages [1]. An aberrant activation in many aspects of the first was dominated by anti-SSA/SSB positivity, the
the immune system has been documented for SLE, but second by anti-Sm/dsDNA/RNP positivity, the third
common denominators for most patients include T cell- by aPL positivity and the last was seronegative [17].
dependent autoantibody production and type I inter- Lymphopenia was more frequent and disease activ-
feron (IFN) overexpression [2–4]. The widely distributed ity higher in the seropositive groups compared to the
immune activation is reflected by a diversity in laboratory seronegative group [17]. It is still unknown if counts of
abnormalities (including lymphopenia, hypocomple- specific lymphocyte subsets including T cells, B cells
mentemia and presence of autoantibodies) and in clini- and NK cells relate to different disease-related autoan-
cal features (including arthritis, renal and dermatological tibody profiles, and if they relate to each other during
manifestations) [4]. Moreover, SLE pregnancy carries a pregnancy in SLE.
risk for disease flare and an increased risk of pregnancy Many patients with SLE present with increased
complications compared to the general population [5, 6]. expression of type I interferon (IFN)-regulated genes in
Lymphopenia is common in individuals with SLE, blood cells and in tissue, an IFN signature [2, 3], and a
occurring in about 40% of the patients [7–9], and low single cell RNA sequencing analysis of PBMC showed
absolute numbers of the lymphocyte subsets CD4 + T that increased expression of type I IFN regulated genes
cells, CD8 + T cells, B cells and NK cells have been in monocytes correlate with low naïve CD4 + T cells in
reported in patients with SLE compared to healthy SLE [21]. Individuals with SLE also have higher IFNα
individuals [10–12]. In SLE, lymphopenia is indepen- protein concentrations in blood compared to healthy
dently associated with organ damage accrual, neurolog- subjects as demonstrated by the use of an ultrasensitive
ical involvement and disease activity [8, 9, 13], but it is single molecule array (Simoa) digital enzyme-linked
unknown whether specific lymphocyte subset numbers immunosorbent assay (ELISA) [22]. In SLE, Simoa-
in blood are affected by pregnancy in SLE and if subset quantified IFNα protein levels strongly correlate with a
counts relate to disease activity during pregnancy. whole blood IFN-I gene score, and these methods iden-
Various autoantibodies are described in SLE [14], but tify associtations with SLE disease activity equally well
only a few are assessed routinely in the clinical setting. [23]. Using the digital ELISA technology, we reported
These include anti-nuclear antibodies (ANA) directed that IFNα protein positivity is present in a subgroup of
to double stranded DNA (dsDNA), Smith antigen (Sm), pregnant women with SLE, but the protein concentra-
ribonucleoprotein (RNP), Sjögren’s syndrome antigen tions are similar during pregnancy and in the late post-
A (SSA) and B (SSB) and anti-phospholipid antibodies partum period [24]. Additionally, we and others have
(aPL) directed to cardiolipin (CL) and β2-glycoprotein reported that pregnant and non-pregnant SLE patients
I (β2GPI) [15]. The heterogeneity of SLE has motivated positive for anti-SSA antibodies have increased IFNα
attempts to stratify patients into subgroups based on protein levels in blood [24, 25]. Yet, it is unknown if
disease-related autoantibody profiles in non-preg- lymphocyte subset counts relate to IFNα protein levels
nant patients with SLE [16–18]. A large international in pregnant women with SLE.
longitudinal study recently identified four serologi- The first aim of this study was to compare total CD4 + T
cal clusters that differed in clinical features but also cell, CD8 + T cell, B cell and NK cell counts prospectively
predicted long term events [16]. The latter and other throughout pregnancy in women with SLE relative to the
studies also report that positivity for most SLE- late postpartum period and to healthy pregnant women.
related autoantibodies decrease over time [16, 19, 20]. Secondly, we aimed to investigate whether the lympho-
In cross-sectional setting, one study separated SLE cyte subset counts were related to autoantibody profiles,
patients into three groups where two were dominated IFNα protein levels, disease activity and gestational age at
by anti-dsDNA-positive individuals who had a higher birth in SLE pregnancy.
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 3 of 15

Materials And Methods Supplementary Table 2. All blood samples were kept at
Cohort ambient temperature until processed the day after, within
This Swedish multicenter study enrolled pregnant women 24 h after venipuncture at our laboratory in Gothenburg.
with SLE (n = 80) meeting the 1997 American College of Whole blood was used for flow cytometry analysis of
Rheumatology (ACR) and/or the 2012 Systemic Lupus total lymphocyte subset counts. Density centrifugation
International Collaborating Clinics (SLICC) classifica- of whole blood was performed to isolate plasma that was
tion criteria [26, 27] between November 2018 and June kept frozen (-80 °C) until further analysis.
2022 at Rheumatology clinics in: Gothenburg (Sahlgren-
ska University hospital, n = 24), Stockholm (Karolinska Autoantibody status during pregnancy in SLE
University Hospital, n = 38), Uppsala (Uppsala University Analysis of positivity for immunofluorescence (IF)-
Hospital, n = 3), Linköping (Linköping University Hos- ANA, for ANA fine specificities including antibodies to
pital, n = 6) and Lund (Skåne University Hospital, n = 9). dsDNA, Sm, RNP, SSA, SSB, chromatin and ribosomal
Healthy pregnant women (HC, n = 51) were enrolled at P protein, and for anti-phospholipid antibodies includ-
one antenatal clinic in Gothenburg (Regionhälsan, Goth- ing anti-CL and anti-β2GPI in plasma collected during
enburg) between October 2018 and December 2022. pregnancy was performed at the accredited laboratory of
Most pregnant women with SLE and HC were included Clinical Immunology, Sahlgrenska University Hospital.
at 10–12 weeks of gestation and followed in the second Most of these samples were collected in trimester three
(week 18–20) and third (week 32–34) trimester. Dis- and plasma was diluted 1:1 in PBS. IF-ANA was analyzed
ease activity was evaluated according to the SLE Disease using Hep-2 cells according to routine, and 66 out of 80
Activity Index 2000 (SLEDAI-2 K) [28], and measured (83%) women with SLE were positive. ANA fine specifici-

the BioPlex™ 2200 System (BioRad Laboratories, Hercu-

at least once between week 10 and week 34 and if the ties were analyzed using multiplexed bead technology by
disease activity was assessed more than once, the high-
est score was used in analyses. The number of pregnant les, USA). The cut-off for most ANA-specificities was 1.0
women with SLE for whom SLEDAI-2 K assessments AI (antibody index) except for anti-chromatin (1.5 AI),
were obtained from in each trimester is shown in Sup- anti-RNP (3.0 AI) and anti-dsDNA (10 IU/mL). Positive
plementary Table 1. Clinical data including disease dura- ANA-specificities were confirmed with another method
tion, medication, ever autoantibody positivity, gestational according to the manufacturers recommendation: Crith-
age at birth and giving birth to a small for gestational age idia luciliae test for anti-dsDNA (ImmunoConcept,

Alegria® (Orgentec Diagnostics, Mainz, Germany) for

(SGA) infant were retrieved from medical records. SGA Sacramento, CA), automated ELISA-based test system
(n = 13) was defined as birth weight less than the 10th
percentile for expected birth weight [29]. Exclusion crite- anti-SSA52 and line blot ANA Profile 5 IgG for all other
ria were inability to understand the study-related patient ANA-specificities (Euroimmun, Lübeck, Germany).

using the BioPlex™ 2200 multiplex immunoassay system

information and informed consent form, presence of Anti-CL and anti-β2GPI of IgG isotype were examined
other serious disease, including active cancer and other
rheumatic autoimmune diseases, or treatment with anti- and APLS reagents. Cut-off values for positivity were 20
BAFF or anti-CD20 antibodies within 12 months before GPL for aCL IgG and 20 AU/mL for anti-β2GPI IgG as
inclusion. Women who had miscarriage during trimester recommended by the manufacturer. Ever autoantibody
one were excluded. None of the women with SLE were status was obtained from medical records and included
treated with anifrolumab before or during their preg- positivity for anti-dsDNA/Sm/SSA/SSB, anti-CL/β2GPI
nancy. All participants have given their written informed and lupus anticoagulant (LAC). The method for analy-
consent and the Ethics board in Gothenburg (Dnr 404– sis of ever antibody positivity differed between the study
18) and The Swedish Ethical Review Authority (amend- sites and dsDNA was confirmed by either Crithidia luci-
ments Dnr 2020–05101 and Dnr 2022–01158-02) have liae test or ELISA. Positivity was determined according
approved the study. to cut-off levels at the local laboratories.

TruCount™ assay was used to analyze total number of

Sample collection Flow cytometry
Peripheral blood samples were collected in heparinized
tubes from pregnant women with SLE and HC in the first, lymphocytes, CD4 + T cells, CD8 + T cells, B cells and
second and third trimester. One additional blood sam- NK cells in whole blood. In brief, blood and antibodies

(Supplementary Table 3) were added to BD TruCount™

ple was collected late postpartum from 19 of the women against CD45, CD3, CD4, CD8, CD19, CD20 and CD56
with SLE (at least six months after delivery (median 10

blood cells were then lysed using BD FACS™ Lysing

[6,–36] months). Information about the number of blood tubes (BD Bioscience) and incubated for 15 min. Red
samples collected for each trimester is presented in
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 4 of 15

solution (BD Sciences). All samples were acquired in a to SLEDAI-2 K was 2 during pregnancy. Disease activ-
FACSVerse equipped with FACSuite Software (BD Bio- ity did not vary much between the trimesters (Supple-
sciences) and analyzed with FlowJo Software (TreeStar, mentary Fig. 1A-B). For all patients with moderate/high
Ashland, Oregon, USA). disease activity (SLEDAI-2 K ≥ 4), the components that
contributed to the score were due to the disease per se
IFNα protein quantification and not to pregnancy, and the most common features
The concentration of IFNα protein levels in plasma were increase in anti-dsDNA, low complement, and
diluted 1:1 in PBS was quantified with Single molecule arthritis. Historically, all except one were ever ANA-pos-
array (Simoa) digital ELISA on a HD-X Analyzer (Quan- itive by immunofluorescence and the majority were ever
terix, Billerica, MA). To prevent false positive results anti-dsDNA positive. In the cross-sectional analysis of
the Simoa assay contained an inhibitor for heterophilic autoantibody positivity during pregnancy, 71% were posi-
antibodies. If the concentration in a sample was below tive for at least one of the ANA fine specificities assessed.
the lower limit of quantification (70 fg/ml) its value was In line with previous findings [19, 20], the percentage
adjusted to 35 fg/ml. IFNα positivity was defined as pro- of positive women was mostly lower in cross-sectional
tein levels ≥ 136 fg/ml, representing three standard devia- compared to ever positive analysis: anti-dsDNA (36% vs.
tions above mean IFNα protein concentration among 84%), anti-Sm (15% vs 25%), anti-SSB (10% vs 14%), anti-
healthy blood donors [30]. CL (8% vs 16%) and anti-β2GPI (9% vs 20%). Moreover,
30% were positive for anti-chromatin, 23% for anti-Sm/
Statistical analysis RNP, 14% for anti-RNP, and 3% for anti-Ribosomal P dur-
Multivariate data analysis was performed using the ing pregnancy. Most women with SLE were treated with
SIMCA-P software (Sartorius Stedem Biotech, Goettin- hydroxychloroquine (93%) and acetylsalicylic acid (88%)
gen, Germany). Principal Component Analysis (PCA) during pregnancy, while one third or less were treated
was used to obtain an unsupervised descriptive over- with prednisone (33%), azathioprine (29%) and/or low
view of groupings and trends, associations, between total molecular weight heparin (24%).
number of CD4 + T cells, CD8 + T cells, B cells and NK
cells, IFNα protein levels, autoantibody positivity dur- Blood lymphocyte subset counts are not affected
ing pregnancy, disease activity, gestational age at birth by pregnancy in SLE but are lower compared to healthy
and SGA among pregnant women with SLE. Orthogonal controls
partial least squares (OPLS) analysis was performed to We first examined total numbers of circulating lympho-
investigate medication or gestational age at birth (Y-var- cytes and the lymphocyte subsets CD4 + T cells, CD8 + T
iables) in relation to total numbers of T cells, B cells and cells, B cells and NK cells in pregnant women with SLE
NK cells (X-variables) in pregnant women with SLE. In and HC. A representative gating strategy for the differ-
the PCA and OPLS models default settings were used; ent lymphocyte subsets in SLE and HC is presented in
data were centered and scaled to unit-variance to give Fig. 1A. Late postpartum samples from pregnant women
all variables equal weight. Model quality was based on with SLE were analyzed to determine if potential differ-
R2 and Q2 parameters that are presented in each figure. ences in lymphocyte subset numbers were an effect of
Univariate analyses were only performed for the strong- SLE combined with pregnancy or to SLE per se. Neither
est associations found in the PCA and OPLS models and total lymphocyte count, nor the subsets CD4 + T cells,
included Kruskal–Wallis followed by Dunn’s multiple CD8 + T cells, B cells and NK cells differed significantly
comparison test, Mann–Whitney U test and Spearman between trimesters or compared to late postpartum in
rank correlation test (GraphPad prism software, La Jolla, SLE (Fig. 1B-F). Similar results were observed when
CA, USA) as described in each respective figure legend. only including data of women from whom late postpar-
P-values of < 0.05 were considered statistically significant. tum samples were collected (Supplementary Fig. 2A-E).
However, total lymphocyte count and all subsets were
Results significantly lower in SLE compared to HC throughout
Clinical characteristics of pregnant women with SLE pregnancy (Fig. 1G-K, combined data from trimester
Demographic and clinical data of the cohort are shown one to three). The comparisons of cell counts between
in Table 1. The data presented, except for the cross- pregnant women with SLE and pregnant HC for each tri-
sectional analysis of autoantibody positivity, have been mester separately are shown in Supplementary Fig. 3A-E.
previously described for subsets of both patients and We also examined whether treatment related to lympho-
controls [24]. In brief, the age and percentage of nullipa- cyte subset counts in pregnant women with SLE. OPLS
rous women were similar in SLE and HC. For pregnant analysis indicated an inverse relation between azathio-
women with SLE, the median disease activity according prine treatment and numbers of NK cells and B cells in
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 5 of 15

Table 1 Characteristics of women with SLE and healthy controls

SLE (n = 80) Controls (n = 51)

Age (years), median (range)a 32 (23–43) 32 (18–41)

Nulliparous, n (%) 48 (60) 35 (69)
Disease duration at inclusion (years), median (range) 9 (0–26)
SLEDAI-2 K during pregnancy, median (range)b 2 (0–18)
ACR criteria, n (%) ever
Malar rash 31 (39)
Discoid rash 6 (8)
Photosensitivity 39 (49)
Oral ulcers 30 (38)
Arthritis 66 (83)
Serositis 17 (21)
Renal disorder 30 (38)
Neurological disorder 5 (6)
Hematological disorder 53 (66)
Immunological disorder 69 (86)
Anti-nuclear ­antibodyc 79 (99)
Antiphospholipid syndrome, n (%) 5 (6)
Autoantibodies, n (%) ever during pregnancy
ANA fine specificity
Any ANA fine specificity N/A 57 (71)
Anti-dsDNA 67 (84) 28 (36)
Anti-Sm 20 (25)b 12 (15)
Anti-SSA 23 (29) 24 (30)
Anti-SSB 11 (14) 8 (10)
Anti-Sm/RNP N/A 18 (23)
Anti-RNP N/A 11 (14)
Anti-Chromatin N/A 24 (30)
Anti-Ribosomal P N/A 2 (3)
Antiphospholipid antibodies
Anti-cardiolipin IgG 13 (16) 6 (8)
Anti-β2glycoprotein I IgG 16 (20)b 7 (9)
Lupus Anticoagulant 12 (15) N/A
Medication, n (%)d
Hydroxychloroquine or chloroquine phosphate 74 (93)
Acetylsalicylic acid 70 (88)
Low molecular weight heparin 19 (24)
Azathioprine 23 (29)
Prednisone 26 (33)
a
Age of mother at gestational week 12
b
Missing data from one patient
c
Positive by immunofluorescence microscopy (IF-ANA)
d
Early pregnancy

(See figure on next page.)

Fig. 1 Blood lymphocyte subset counts are not affected by pregnancy in SLE. (A) Representative flow cytometry plots from SLE and HC illustrating
the gating strategy for total number of lymphocytes, CD4 + T cells, CD8 + T cells, B cells and NK cells respectively. Total number of (B) lymphocytes,
(C) CD4 + T cells, (D) CD8 + T cells, (E) B cells and (F) NK cells in trimester one, two, three and late postpartum in women with SLE. Combined data
from trimester one, two and three comparing numbers of (G) total lymphocytes, (H) CD4 + T cells, (I) CD8 + T cells, (J) B cells and (K) NK cells
between SLE and HC pregnancy. **p < 0.01, ****p < 0.0001, (G-K) Mann–Whitney U test
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 6 of 15

Fig. 1 (See legend on previous page.)

Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 7 of 15

Fig. 2 Treatment-independent decrease of NK cells and B cells

in SLE pregnancies. OPLS loading column plots depicting counts
of lymphocyte subsets positively or negatively associated with (A)
azathioprine, (B) prednisone or (C) low molecular weight heparin.
(D) NK cell and (E) B cell counts in patients treated or not with
azathioprine and in HC. (F) B cell counts in patients treated
or not with prednisone and in HC. (G) B cell counts in patients treated
or not with low molecular weight heparin and in HC. *p < 0.05,
***p < 0.001, ****p < 0.0001, (D-E) Kruskal–Wallis followed by Dunn’s
multiple comparison test

all trimesters (Fig. 2A). Prednisone and low molecular

weight heparin related inversely to B cell counts in all
trimesters (Fig. 2B-C). These associations were cor-
roborated in univariate analysis. NK cell and B cell
counts were significantly lower in women treated com-
pared to not treated with azathioprine (Fig. 2D-E). B
cell counts were also lower in women treated compared
to not treated with prednisone or heparin, respectively
(Fig. 2F-G). Still, significantly lower numbers of both NK
cells and B cells were found in women without respec-
tive treatment compared to HC (Fig. 2D-G). Treatment
with hydroxychloroquine and acetylsalicylic acid were
unrelated to lymphocyte subset counts (Supplementary
Fig. 4A-B). In summary, pregnant women with SLE pre-
sent with lower number of lymphocyte subset counts in
blood relative to pregnant HC, which seem to be a feature
of SLE per se and in part by treatment with immunosup-
pressive drugs and heparin but not related to pregnancy.

Lymphocyte subset numbers inversely associate

with positivity for ANA specificities, IFNα protein levels
and disease activity in SLE pregnancy
We next performed PCA to investigate how CD4 + T cell,
CD8 + T cell, B cell and NK cell counts relate to IFNα
protein levels, autoantibody positivity during pregnancy,
as well as to SLEDAI-2 K, gestational age at birth and
SGA in SLE. We previously reported that plasma IFNα
protein positivity is present in a subgroup (36%) of preg-
nant women with SLE [24]. As shown in Fig. 3, lympho-
cyte subset counts were projected on the opposite side to
positivity for ANA specificities, IFNα protein levels and
SLEDAI-2 K indicating on an inverse association. ANA
positivity for anti-dsDNA/Sm/RNP/chromatin (clus-
ter one) separated from anti-SSA/SSB positivity (cluster
two), suggesting that these clusters relate differently to
lymphocyte subset numbers. IFNα protein levels associ-
ated with cluster two, while SLEDAI-2 K projected close
to anti-dsDNA in cluster one. Antiphospholipid antibody
positivity (anti-CL/β2GPI, cluster three) separated from
the other variables on the bottom right side of the plot.
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 8 of 15

Fig. 3 Lymphocyte subset counts inversely relate to autoantibody positivity, IFNα and disease activity in SLE pregnancy. Unsupervised principal
component analysis demonstrating the relationship between number of CD4 + T cells, CD8 + T cells, B cells, NK cells in trimester one to three,
cross sectional (cs) autoantibody positivity, IFNα protein levels, SLEDAI-2 K, gestational age at birth and small for gestational age (SGA) in pregnant
women with SLE

Lower CD4 + T cell and B cell counts in pregnant women anti-Sm compared to those who were negative (Fig. 4C).
with SLE positive for disease‑specific anti‑dsDNA NK cell numbers were lower in women positive for all
and anti‑Sm ANA in cluster one, except for anti-dsDNA, compared
Guided by the PCA analysis, we first investigated the to those who were negative (Fig. 4D). Lymphocyte sub-
association between lymphocyte subset counts and the set counts did not differ in women positive for anti-SSA
two ANA clusters. As shown in Fig. 4A, CD4 + T cell or anti-SSA/SSB compared to those who were nega-
counts were significantly lower in women positive to all tive (Supplementary Fig. 5A-D), and none of the women
ANA specificities in cluster one compared to those who were anti-SSB + /SSA-. Additionally, all lymphocyte sub-
were negative to respective ANA specificity. CD8 + T cell set counts were lower in women positive for three to five
count was lower among women positive for anti-Sm, anti- ANA compared to those who had one to two ANA and to
RNP and anti-chromatin compared to those who were those who were ANA negative (Fig. 4E-H). The CD4 + T
negative (Fig. 4B). B cell numbers were significantly lower cell counts were also lower in women who had one to two
in women positive for disease specific anti-dsDNA and ANA compared to those who were negative (Fig. 4E). Few

(See figure on next page.)

Fig. 4 Low lymphocyte subset numbers relate to a specific cluster of ANA positivity. (A) CD4 + T cell, (B) CD8 + T cell, (C) B cell and (D) NK
cell counts in women positive for anti-dsDNA, anti-Sm, anti-Sm/RNP, anti-RNP and anti-chromatin antibodies compared to women negative
for respective antibody. (E) CD4 + T cell, (F) CD8 + T cell, (G) B cell and (H) NK cell counts in women negative to ANA, positive to one to two ANA
or positive to three to five ANA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (A-D) Mann–Whitney U test, (E–H) Kruskal–Wallis followed by Dunn’s
multiple comparison test
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 9 of 15

Fig. 4 (See legend on previous page.)

Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 10 of 15

women were positive for aPL during pregnancy (anti-CL As expected, SLEDAI-2 K projected close to anti-dsDNA
n = 6 and anti-β2GPI n = 7), and there were no significant positivity in the PCA analysis, and disease activity was
differences in lymphocyte subset counts in women who higher in women positive for anti-dsDNA compared to
were aPL-positive compared to negative except for NK those who were negative (Supplementary Fig. 7A). SLE-
cells counts that were higher in women with compared DAI-2 K was unrelated to other ANA fine specificities in
to without aPL (Supplementary Fig. 5E-H). In summary, cluster one and to number of positive ANA (Supplemen-
lymphocyte subset counts are lower in women positive tary Fig. 7B-F). We have previously reported that higher
compared to negative for anti-dsDNA, anti-Sm, RNP, Sm/ proportions of low-density granulocytes in late SLE preg-
RNP and chromatin, while low counts are unrelated to nancy correlate to lower gestational age at birth [24], but
positivity for anti-SSA/SSB and aPL in SLE pregnancies. lymphocyte subset counts were unrelated to pregnancy
duration and to giving birth to an SGA infant in women
Lower CD4 + T cell and B cell counts in SLE pregnancies with SLE (Supplementary Fig. 8A-B and 9A-D). To con-
with moderate/high disease activity clude, pregnant women with SLE who have a moderate/
Next, we examined the association between lymphocyte high disease activity have lower circulating numbers of
subsets counts and disease activity in SLE pregnancies. CD4 + T cells and B cells compared to women with no/
Both CD4 + T cell and B cell counts were lower in women low disease activity, but low lymphocyte subset counts do
with moderate/high disease activity (SLEDAI-2 K ≥ 4) not predict shorter pregnancy duration in SLE.
compared to those with no/low disease activity, while
there was no difference in numbers of CD8 + T cells and Lower CD4 + T cell counts are related to higher IFNα
NK cells between the two groups (Fig. 5A-D). Similar protein levels in SLE pregnancies
results were obtained when lymphocyte subset counts for In PCA, there was an inverse association between lym-
each trimester were analyzed in relation to SLEDAI-2 K phocyte subset numbers and IFNα protein levels. In uni-
for the respective trimester (Supplementary Fig. 6A-C). variate analysis, CD4 + T cell counts inversely correlated

Fig. 5 Lower CD4 + T cell and B cell numbers in pregnant women with moderate/high disease activity. (A) CD4 + T cell, (B) CD8 + T cell, (C) B
cell and (D) NK cell numbers in women with moderate/high disease activity (SLEDAI-2 K ≥ 4) compared to women with no/low disease activity.
**p < 0.01, ***p < 0.001, Mann–Whitney U test
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 11 of 15

with IFNα protein levels during SLE pregnancy (Fig. 6A), in non-pregnant patients relative to controls [34]. We
and CD4 + T cell counts were significantly lower in IFNα also show that none of the lymphocyte subset numbers
protein-positive compared to negative women (Fig. 6B). were affected by pregnancy in SLE. This contrasts with
Numbers of CD8 + T cells showed a weaker negative healthy pregnant women who have lower numbers of
correlation to IFNα protein levels and there was no dif- lymphocytes during pregnancy compared to the post-
ference in CD8 + T cell count between IFNα protein- partum period [35, 36]. The explanation for this discrep-
positive compared to negative women (Fig. 6C-D). B cell ancy could only be speculated upon, but disease-related
and NK cell counts were unrelated to IFNα protein levels homing of activated lymphocytes from the periphery to
(Supplementary Fig. 10A-B). As previously demonstrated inflamed tissues and organs and the presence of autoan-
by others [31–33], IFNα protein levels related to number tibodies with lymphocytotoxic activity may result in
of positive ANA specificities (Supplementary Fig. 10C). low lymphocyte counts that are not further affected by
Thus, pregnant women with SLE who are IFNα-positive pregnancy in SLE [37, 38]. Although the activation status
present with lower numbers of CD4 + T cells in blood of the different lymphocyte subsets was not examined
compared to those who are IFNα-negative. here, we have previously reported that pregnancy in SLE
results in increased activation of circulating granulo-
Discussion cytes compared to the late postpartum period [24].
To our knowledge, this is the first study to investigate Although SLE is a heterogenic disease with a vari-
if pregnancy affects blood lymphocyte subset counts ety in laboratory abnormalities and clinical features,
in SLE, and if T cell, B cell and NK cell counts associ- recent detailed serological profiling has identified sets
ate to disease-related autoantibody positivity and/or to of disease-related autoantibodies that commonly occur
IFNα protein concentrations during SLE pregnancy. together [16, 17, 39]. In accordance we here show for
In this longitudinal study we confirm that low total the first time in SLE pregnancy that autoantibody posi-
lymphocyte count is evident throughout pregnancy in tivity also separated into three clusters, the first domi-
women with SLE compared to HC, a well-known feature nated by positivity for anti-dsDNA/anti-Sm/anti-RNP/

Fig. 6 Lower CD4 + T cell counts in IFNα-positive pregnant women with SLE. (A) Correlation analysis of CD4 + T cell counts and IFNα protein
concentrations. (B) Comparison of CD4 + T cell counts in pregnant women with or without IFNα protein positivity (≥ 136 fg/ml). (C) Correlation
analysis of CD8 + T cell counts and IFNα protein concentrations. (D) Comparison of CD8 + T cell counts in pregnant women with or without IFNα
protein positivity. ****p < 0.0001, (A and C) Spearman rank correlation test, (B) Mann–Whitney U test
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 12 of 15

anti-chromatin, the second by anti-SSA/anti-SSB and pregnant women with SLE compared to those who were
the third by anti-CL/anti-β2GPI. Another novel finding negative. Our finding is in line with recent scRNA-seq
was that low numbers of particularly CD4 + T cells, but data showing that a reduction of naïve CD4 + T cells
also B cells, CD8 + T cells and NK cells, relate to posi- correlates with increased expression of type I IFN regu-
tivity for ANA specificities in cluster one, but not ANA lated genes in monocytes in non-pregnant individu-
positivity in cluster two or aPL positivity in cluster three. als with SLE [21]. Whether there is a direct effect of
Whether specific lymphocyte subset counts relate to IFNα on CD4 + T cell numbers in blood is unclear, but
autoantibody positivity profiles has not been examined administration of IFNα in healthy volunteers leads to a
in non-pregnant patients with SLE. The use of antibody drastic decrease of total lymphocyte numbers in blood
clustering to separate patients with SLE into endotypes [44, 45]. Mouse models suggest a partial mechanistic
may help to predict disease course and prognosis as explanation for this phenomenon by inhibited egress of
patients with distinct autoantibody profiles differ with CD4 + T cells, CD8 + T cells and CD19 + B cells from
regards to immunological variables, clinical manifesta- lymph nodes, as treatment with the IFNα inducer poly
tions, treatment, organ involvement and long-term dis- (I:C) retains lymphocytes in lymph nodes via regula-
ease activity [16–18, 40]. tion of CD69 and sphingosine 1-phosphate receptor-1
Lymphopenia is associated with SLE-specific anti- (S1P1) expression [46]. Still, the relationship between
dsDNA positivity, SLE-related autoantibody positiv- IFNα and low CD4 + T cell counts remains to be exam-
ity in general, and more severe/progressive disease ined further.
in non-pregnant subjects with SLE [9, 17, 18]. When Medication may affect numbers of lymphocyte subsets
we here divided total lymphocytes into subsets, only in blood. Indeed, pregnant women with SLE who were
CD4 + T cell and B cell counts were lower in pregnant treated with azathioprine had lower numbers of NK cells
women positive to anti-dsDNA compared to those who and B cells compared to women who were not treated.
were negative and counts of these subsets were also In accordance with this, azathioprine use is related to
lower in women with moderate/high (SLEDAI-2 K ≥ 4) reduced numbers and proportions of NK cells and B cells
compared to no/low disease activity. Accordingly, in non-pregnant subjects with SLE, inflammatory bowel
anti-dsDNA positivity, being part of the SLEDAI-2 K disease or ANCA-associated vasculitis [21, 47–50]. A pro-
score [28], was also related to higher disease activity. posed mechanism for azathioprine-related decrease in NK
Anti-Sm is another SLE-specific autoantibody, but its cells is caspase 3- and 9-induced apoptosis [49]. Addition-
pathologic significance is uncertain and there are con- ally, we also found lower B cell counts in women treated
flicting data regarding its association to disease activ- with prednisone or heparin compared to those who were
ity and clinical manifestations including lymphopenia not. For prednisone, similar results were reported from a
[41–43]. We found that all lymphocyte subset counts small cohort of patients with different autoimmune dis-
were lower in pregnant women who were positive for orders [51], and in a small cohort of healthy volunteers
anti-Sm relative to those who were negative, but anti- [52]. Importantly, we also found a treatment-independent
Sm positivity was unrelated to disease activity in our decrease in both NK cells and B cells in SLE compared to
cohort. A higher number of ANA fine specificities also HC pregnancies.
related to lower numbers of lymphocyte subset counts, A strength of the study is the inclusion of well-charac-
but not to disease activity. Still, a higher number of terized patients and controls from whom samples have
ANA specificities could indicate a more immunologi- been prospectively collected in parallel during pregnancy
cally active disease that leads to lymphocyte homing to and late postpartum. Others are that all flow cytometry
inflamed tissue and organs, which is not captured by analyses were performed on fresh blood in one labora-
the SLEDAI-2 K index. Whether there is a causal rela- tory on the same instrument and cross-sectional analysis
tionship between low lymphocyte subsets counts and of autoantibody positivity was analyzed with well-stand-
specific ANA positivity is not answered by the present ardized methods by staff in an accredited hospital labora-
data, but we add novel knowledge on how numbers of tory. Our study also has limitations. The cohort included
specific lymphocyte subsets in blood differ in relation few women with moderate or high SLE disease activ-
to antibody positivity profiles and disease activity in ity and therefore our results reflect a well-controlled
SLE pregnancies. cohort of pregnant women with SLE. Moreover, the study
Lower total lymphocyte counts have also been includes missing data, mainly due to missing blood sam-
reported in IFNα positive compared to negative non- ples from the first trimester, among pregnant women
pregnant patients with SLE [30]. We found that spe- with SLE.
cifically CD4 + T cell counts, but not counts of any To conclude, we report that pregnancy in women with
other lymphocyte subset, were lower in IFNα-positive SLE has no effect on blood lymphocyte subset counts
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 13 of 15

but SLE pregnancies are featured by a treatment-inde- Funding

Open access funding provided by University of Gothenburg. This study was
pendent decrease in blood lymphocyte counts compared supported by grants from the Swedish Research Council; the Swedish state
to healthy pregnant women. Moreover, low counts of under the agreement between the Swedish government and the county
specific lymphocyte subsets relate differentially to dis- councils, the ALF-agreement; foundations of The King Gustaf V’s 80th Birthday
found; the Swedish Rheumatism Association; the Gothenburg Society of
ease-related autoantibody positivity, IFNα protein levels Medicine; foundations of Ulla and Roland Gustafsson, Nanna Svartz, Rune and
and disease activity in SLE pregnancy. Still, further stud- Ulla Amlöv, Hjalmar Svensson, IngaBritt and Arne Lundberg, Ingegerd Johans‑
ies are needed to decipher the immunological charac- son and Emil and Wera Cornell; The Swedish Society of Medicine.
teristics of SLE phenotypes based of antibody profiles in
more detail, and to investigate if specific subgroups are Declarations
related to an increased risk for pregnancy complications. Ethical approval and consent to participate
The study complies with the Declaration of Helsinki and was approved by
the appropriate medical ethical committee, the regional ethics commit‑
Abbreviations tee of Gothenburg (Dnr 404–18 and amendment Dnr 2020–05101 and Dnr
ACR​ American college of rheumatology 2022–01158-02). All participants signed an informed consent form.
ANA Antinuclear antibodies
aPL Antiphospholipid andibodies Consent for publication
CL Cardiolipin Not applicable.
dsDNA Double-stranded DNA
ELISA Enzyme-linked immunosorbent assay Competing interests
HC Healthy controls HZ has served at scientific advisory boards and/or as a consultant for Abbvie,
IFN Interferon Acumen, Alector, Alzinova, ALZPath, Annexon, Apellis, Artery Therapeutics,
LAC Lupus anticoagulant AZTherapies, CogRx, Denali, Eisai, Nervgen, Novo Nordisk, Optoceutics, Pas‑
OPLS Orthogonal partial least square sage Bio, Pinteon Therapeutics, Prothena, Red Abbey Labs, reMYND, Roche,
PCA Principal component analysis Samumed, Siemens Healthineers, Triplet Therapeutics, and Wave, has given
RNP Ribonucleoprotein lectures in symposia sponsored by Cellectricon, Fujirebio, Alzecure, Biogen,
Simoa Single molecule array and Roche, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB
SLE Systemic lupus erythematosus (BBS), which is a part of the GU Ventures Incubator Program (outside submit‑
SLEDAI-2 K SLE disease activity index 2000 ted work). KB has served as a consultant and at advisory boards for Acumen,
SLICC Systemic lupus international collaborating clinics ALZPath, BioArctic, Biogen, Eisai, Lilly, Moleac Pte. Ltd, Novartis, Ono Pharma,
Sm Smith Prothena, Roche Diagnostics, and Siemens Healthineers; has served at data
SSA Sjögren’s syndrome antigen A monitoring committees for Julius Clinical and Novartis; has given lectures,
SSB Sjögren’s syndrome antigen B produced educational materials and participated in educational programs
β2GPI β2 Glycoprotein I 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
Supplementary Information in this paper. LR has collaboration with Astra Zeneca, Bristol Myers Squibb
The online version contains supplementary material available at https://​doi.​ (BMS), AMPEL Biosolutions, UCB and Bayer. ES has received a research grant
org/​10.​1186/​s13075-​024-​03301-0. from Merck and lecture honoraria from Janssen. The other authors declare no
conflict of interest.
Supplementary Material 1.
Author details
Supplementary Material 2. 1
Department of Rheumatology and Inflammation Research, Institute of Medi‑
cine, Sahlgrenska Academy at the University of Gothenburg, Guldhedsgatan
10A, 405 30 Gothenburg, Sweden. 2 Rheumatology, Sahlgrenska University
Acknowledgements
Hospital, Gothenburg, Sweden. 3 Department of Psychiatry and Neuro‑
We thank the study nurses Anita Nihlberg, Maria Andersson, Sonia Möller,
chemistry, Institute of Neuroscience and Physiology, Sahlgrenska Academy
Hans Kling, Eva Malmqvist, Rezvan Kiani, Marianne Petersson, Anna-Lena
at the University of Gothenburg, Mölndal, Sweden. 4 Clinical Neurochemistry
Åblad, Elisabeth Kling and Lena Pålsson at the different rheumatology clinics,
Laboratory, Sahlgrenska University Hospital, Mölndal, Sweden. 5 Paris Brain
the midwifes Charlotta Jansson and Charlotte Andersson and the assistant
Institute, ICM, Pitié-Salpêtrière Hospital, Sorbonne University, Paris, France.
nurse Anette Svensk at the Antenatal clinic in Gothenburg as well as all 6
Neurodegenerative Disorder Research Center, Division of Life Sciences
personnel at the delivery units. For assistance with the collection of clinical
and Medicine and Department of Neurology, Institute On Aging and Brain
data, we thank Kristina Karlsson, Martina Wahlberg and Ylva Folkesson. We
Disorders, University of Science and Technology of China and First Affiliated
thank Gunilla Larsson for proficient laboratory assistance. We also want to
Hospital of USTC, Hefei, People’s Republic of China. 7 Department of Neuro‑
thank Rille Pullerits, Cecilia Larsson and Robert Breiðfjörð at the laboratory of
degenerative Disease, UCL Institute of Neurology, Queen Square, London,
Clinical Immunology, Sahlgrenska University Hospital, for excellent technical
UK. 8 UK Dementia Research Institute at UCL, London, UK. 9 Hong Kong Center
assistance. At last, we wish to thank all the pregnant women who participated
for Neurodegenerative Diseases, Clear Water Bay, Hong Kong, China. 10 Win‑
in the study.
sconsin Alzheimer’s Disease Research Center, School of Medicine and Public
Health, University of Wisconsin, University of Wisconsin-Madison, Madison,
Authors’ contributions
WI, USA. 11 Department of Women’s and Children’s Health, Section of Obstet‑
Conception and design: AT, MS, AR and A-CL. Experimental procedures/
rics and Gynecology, Uppsala University, Uppsala, Sweden. 12 Department
design and interpretation of the data: AT, MS, KB, HZ, JB, AR and A-CL. Patient
of Medical Sciences, Rheumatology, Uppsala University, Uppsala, Sweden.
and healthy control recruitment, collection of clinical data: MS, TA, DL, LR, SP, 13
Department of Obstetrics and Gynecology, Linköping University Hospi‑
MS, CS, HS, AJ, AB, ET, MMS, AZ, ES, IG, BJ and AR. Initial manuscript drafting:
tal, Linköping, Sweden. 14 Department of Biomedical and Clinical Sciences,
AT and A-CL. All authors critically revised the manuscript for important intel‑
Division of Children’s and Women’s Health, Linköping University, Linköping,
lectual content, and all authors approved the final manuscript.
Sweden. 15 Division of Inflammation and Infection, Department of Biomedical
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 14 of 15

and Clinical Sciences, Linköping University, Linköping, Sweden. 16 Department 16. Choi MY, Chen I, Clarke AE, Fritzler MJ, Buhler KA, Urowitz M, et al.
of Obstetrics and Gynecology, Institute of Clinical Sciences, Skåne University Machine learning identifies clusters of longitudinal autoantibody profiles
Hospital, Lund, Sweden. 17 Department of Clinical Sciences Lund, Rheumatol‑ predictive of systemic lupus erythematosus disease outcomes. Ann
ogy, Lund University, Skåne University Hospital, Lund, Sweden. 18 Department Rheum Dis. 2023;82(7):927–36.
of Womens and Childrens Health, Division for Obstetrics and Gynecology, 17. Diaz-Gallo LM, Oke V, Lundström E, Elvin K, Ling WuY, Eketjäll S, et al.
Karolinska University Hospital, Karolinska Institute, Stockholm, Sweden. 19 Divi‑ Four Systemic Lupus Erythematosus Subgroups, Defined by Autoan‑
sion of Rheumatology, Department of Medicine Solna, Karolinska Institute, tibodies Status, Differ Regarding HLA-DRB1 Genotype Associations
Karolinska University Hospital, Stockholm, Sweden. 20 Department of Oral and Immunological and Clinical Manifestations. ACR Open Rheumatol.
Microbiology and Immunology, Institute of Odontology, Sahlgrenska Acad‑ 2022;4(1):27–39.
emy at the University of Gothenburg, Gothenburg, Sweden. 21 Department 18. To CH, Petri M. Is antibody clustering predictive of clinical subsets
of Obstetrics and Gynecology, Sahlgrenska University Hospital, Gothenburg, and damage in systemic lupus erythematosus? Arthritis Rheum.
Sweden. 22 Department of Obstetrics and Gynecology, Sahlgrenska Academy 2005;52(12):4003–10.
at the University of Gothenburg, Gothenburg, Sweden. 23 Department 19. Frodlund M, Walhelm T, Dahle C, Sjöwall C. Longitudinal Analysis of
of Genetics and Bioinformatics, Division of Health Data and Digitalisation, Anti-cardiolipin and Anti-β2-glycoprotein-I Antibodies in Recent-Onset
Institute of Public Health, Oslo, Norway. Systemic Lupus Erythematosus: A Prospective Study in Swedish Patients.
Front Med (Lausanne). 2021;8:646846.
Received: 31 October 2023 Accepted: 1 March 2024 20. Frodlund M, Wetterö J, Dahle C, Dahlström Ö, Skogh T, Rönnelid J, et al.
Longitudinal anti-nuclear antibody (ANA) seroconversion in systemic
lupus erythematosus: a prospective study of Swedish cases with recent-
onset disease. Clin Exp Immunol. 2020;199(3):245–54.
21 Perez RK, Gordon MG, Subramaniam M, Kim MC, Hartoularos GC, Targ S,
References et al. Single-cell RNA-seq reveals cell type-specific molecular and genetic
1. Tsokos GC. Autoimmunity and organ damage in systemic lupus erythe‑ associations to lupus. Science. 2022;376(6589):eabf1970.
matosus. Nat Immunol. 2020;21(6):605–14. 22. Mathian A, Mouries-Martin S, Dorgham K, Devilliers H, Barnabei L, Ben
2. Baechler EC, Batliwalla FM, Karypis G, Gaffney PM, Ortmann WA, Espe Salah E, et al. Monitoring Disease Activity in Systemic Lupus Erythema‑
KJ, et al. Interferon-inducible gene expression signature in peripheral tosus With Single-Molecule Array Digital Enzyme-Linked Immunosorb‑
blood cells of patients with severe lupus. Proc Natl Acad Sci U S A. ent Assay Quantification of Serum Interferon-α. Arthritis Rheumatol.
2003;100(5):2610–5. 2019;71(5):756–65.
3. Rönnblom L, Leonard D. Interferon pathway in SLE: one key to unlocking 23. Chasset F, Mathian A, Dorgham K, Ribi C, Trendelenburg M, Huynh-Do
the mystery of the disease. Lupus Sci Med. 2019;6(1): e000270. U, et al. Serum interferon-α levels and IFN type I-stimulated genes score
4. Kaul A, Gordon C, Crow MK, Touma Z, Urowitz MB, van Vollenhoven R, perform equally to assess systemic lupus erythematosus disease activity.
et al. Systemic lupus erythematosus. Nat Rev Dis Primers. 2016;2(1):16039. Ann Rheum Dis. 2022;81(6):901–3.
5. Tedeschi SK, Massarotti E, Guan H, Fine A, Bermas BL, Costenbader KH. 24. Torell A, Stockfelt M, Larsson G, Blennow K, Zetterberg H, Leonard D, et al.
Specific systemic lupus erythematosus disease manifestations in the six Low-density granulocytes are related to shorter pregnancy duration but
months prior to conception are associated with similar disease manifes‑ not to interferon alpha protein blood levels in systemic lupus erythema‑
tations during pregnancy. Lupus. 2015;24(12):1283–92. tosus. Arthritis Res Ther. 2023;25(1):107.
6. Buyon JP, Kim MY, Guerra MM, Laskin CA, Petri M, Lockshin MD, et al. Pre‑ 25. Oke V, Gunnarsson I, Dorschner J, Eketjäll S, Zickert A, Niewold TB, et al.
dictors of Pregnancy Outcomes in Patients With Lupus: A Cohort Study. High levels of circulating interferons type I, type II and type III associate
Ann Intern Med. 2015;163(3):153–63. with distinct clinical features of active systemic lupus erythematosus.
7. Rivero SJ, Díaz-Jouanen E, Alarcón-Segovia D. Lymphopenia in systemic Arthritis Res Ther. 2019;21(1):107.
lupus erythematosus. Clinical, diagnostic, and prognostic significance. 26. Hochberg MC. Updating the American College of Rheumatology revised
Arthritis Rheum. 1978;21(3):295–305. criteria for the classification of systemic lupus erythematosus. Arthritis
8. Vilá LM, Alarcón GS, McGwin G Jr, Bastian HM, Fessler BJ, Reveille JD. Rheum. 1997;40(9):1725.
Systemic lupus erythematosus in a multiethnic US cohort, XXXVII: 27. Petri M, Orbai AM, Alarcón GS, Gordon C, Merrill JT, Fortin PR, et al. Deriva‑
association of lymphopenia with clinical manifestations, serologic tion and validation of the Systemic Lupus International Collaborating
abnormalities, disease activity, and damage accrual. Arthritis Rheum. Clinics classification criteria for systemic lupus erythematosus. Arthritis
2006;55(5):799–806. Rheum. 2012;64(8):2677–86.
9. Yavuz S, Cansu DU, Nikolopoulos D, Crisafulli F, Antunes AM, Adam‑ 28. Gladman DD, Ibañez D, Urowitz MB. Systemic lupus erythematosus
ichou C, et al. Lymphopenia as a risk factor for neurologic involvement disease activity index 2000. J Rheumatol. 2002;29(2):288–91.
and organ damage accrual in patients with systemic lupus erythe‑ 29. Marsál K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine
matosus: A multi-center observational study. Semin Arthritis Rheum. growth curves based on ultrasonically estimated foetal weights. Acta
2020;50(6):1387–93. Paediatr. 1996;85(7):843–8.
10. Erkeller-Yüsel F, Hulstaart F, Hannet I, Isenberg D, Lydyard P. Lymphocyte 30. Mathian A, Mouries-Martin S, Dorgham K, Devilliers H, Yssel H, Garrido
subsets in a large cohort of patients with systemic lupus erythematosus. Castillo L, et al. Ultrasensitive serum interferon-α quantification during
Lupus. 1993;2(4):227–31. SLE remission identifies patients at risk for relapse. Ann Rheum Dis.
11. Wouters CH, Diegenant C, Ceuppens JL, Degreef H, Stevens EA. The circu‑ 2019;78(12):1669–76.
lating lymphocyte profiles in patients with discoid lupus erythematosus 31. Hubbard EL, Pisetsky DS, Lipsky PE. Anti-RNP antibodies are associated
and systemic lupus erythematosus suggest a pathogenetic relationship. with the interferon gene signature but not decreased complement levels
Br J Dermatol. 2004;150(4):693–700. in SLE. Ann Rheum Dis. 2022;81(5):632–43.
12. Carvajal Alegria G, Gazeau P, Hillion S, Daïen CI, Cornec DYK. Could Lym‑ 32. Reynolds JA, Briggs TA, Rice GI, Darmalinggam S, Bondet V, Bruce E,
phocyte Profiling be Useful to Diagnose Systemic Autoimmune Diseases? et al. Type I interferon in patients with systemic autoimmune rheumatic
Clin Rev Allergy Immunol. 2017;53(2):219–36. disease is associated with haematological abnormalities and specific
13. Kandane-Rathnayake R, Louthrenoo W, Golder V, Luo SF, Wu YJ, Lateef autoantibody profiles. Arthritis Res Ther. 2019;21(1):147.
A, et al. Independent associations of lymphopenia and neutropenia in 33. Rodero MP, Decalf J, Bondet V, Hunt D, Rice GI, Werneke S, et al. Detection
patients with systemic lupus erythematosus: a longitudinal, multinational of interferon alpha protein reveals differential levels and cellular sources
study. Rheumatology (Oxford). 2021;60(11):5185–93. in disease. J Exp Med. 2017;214(5):1547–55.
14. Yaniv G, Twig G, Shor DB, Furer A, Sherer Y, Mozes O, et al. A volcanic explo‑ 34. Han BK, Wysham KD, Cain KC, Tyden H, Bengtsson AA, Lood C. Neutrophil
sion of autoantibodies in systemic lupus erythematosus: a diversity of 180 and lymphocyte counts are associated with different immunopatho‑
different antibodies found in SLE patients. Autoimmun Rev. 2015;14(1):75–9. logical mechanisms in systemic lupus erythematosus. Lupus Sci Med.
15. Gómez-Bañuelos E, Fava A, Andrade F. An update on autoantibodies in 2020;7(1):e000382.
systemic lupus erythematosus. Curr Opin Rheumatol. 2023;35(2):61–7.
Torell et al. Arthritis Research & Therapy (2024) 26:65 Page 15 of 15

35. Kraus TA, Engel SM, Sperling RS, Kellerman L, Lo Y, Wallenstein S, et al.
Characterizing the pregnancy immune phenotype: results of the viral
immunity and pregnancy (VIP) study. J Clin Immunol. 2012;32(2):300–11.
36. Watanabe M, Iwatani Y, Kaneda T, Hidaka Y, Mitsuda N, Morimoto Y, et al.
Changes in T, B, and NK lymphocyte subsets during and after normal
pregnancy. Am J Reprod Immunol. 1997;37(5):368–77.
37. Shoenfeld Y, Zamir R, Joshua H, Lavie G, Pinkhas J. Human monoclonal
anti-DNA antibodies react as lymphocytotoxic antibodies. Eur J Immunol.
1985;15(10):1024–8.
38. Hagberg N, Theorell J, Hjorton K, Spee P, Eloranta ML, Bryceson YT, et al.
Functional anti-CD94/NKG2A and anti-CD94/NKG2C autoantibodies
in patients with systemic lupus erythematosus. Arthritis Rheumatol.
2015;67(4):1000–11.
39. Pacheco Y, Barahona-Correa J, Monsalve DM, Acosta-Ampudia Y, Rojas
M, Rodríguez Y, et al. Cytokine and autoantibody clusters interaction in
systemic lupus erythematosus. J Transl Med. 2017;15(1):239.
40. Artim-Esen B, Çene E, Şahinkaya Y, Ertan S, Pehlivan Ö, Kamali S, et al.
Cluster analysis of autoantibodies in 852 patients with systemic lupus
erythematosus from a single center. J Rheumatol. 2014;41(7):1304–10.
41. Ahn SS, Jung SM, Yoo J, Lee SW, Song JJ, Park YB. Anti-Smith antibody
is associated with disease activity in patients with new-onset systemic
lupus erythematosus. Rheumatol Int. 2019;39(11):1937–44.
42. Arroyo-Ávila M, Santiago-Casas Y, McGwin G Jr, Cantor RS, Petri M,
Ramsey-Goldman R, et al. Clinical associations of anti-Smith anti‑
bodies in PROFILE: a multi-ethnic lupus cohort. Clin Rheumatol.
2015;34(7):1217–23.
43. Correa-Rodríguez M, Pocovi-Gerardino G, Callejas-Rubio JL, Ríos-Fernán‑
dez R, Martín-Amada M, María-Gracia C-C, et al. Clinical and serological
associations of autoantibodies in patients with systemic lupus erythema‑
tosus. J Investig Med. 2021;69(8):1417–25.
44. Aulitzky WE, Tilg H, Vogel W, Aulitzky W, Berger M, Gastl G, et al. Acute
hematologic effects of interferon alpha, interferon gamma, tumor necro‑
sis factor alpha and interleukin 2. Ann Hematol. 1991;62(1):25–31.
45. Corssmit EP, Heijligenberg R, Hack CE, Endert E, Sauerwein HP, Romijn JA.
Effects of interferon-alpha (IFN-alpha) administration on leucocytes in
healthy humans. Clin Exp Immunol. 1997;107(2):359–63.
46. Shiow LR, Rosen DB, Brdicková N, Xu Y, An J, Lanier LL, et al. CD69 acts
downstream of interferon-alpha/beta to inhibit S1P1 and lymphocyte
egress from lymphoid organs. Nature. 2006;440(7083):540–4.
47. Green MR, Kennell AS, Larche MJ, Seifert MH, Isenberg DA, Salaman
MR. Natural killer cell activity in families of patients with systemic lupus
erythematosus: demonstration of a killing defect in patients. Clin Exp
Immunol. 2005;141(1):165–73.
48. Lord JD, Shows DM. Thiopurine use associated with reduced B and
natural killer cells in inflammatory bowel disease. World J Gastroenterol.
2017;23(18):3240–51.
49. Yusung S, McGovern D, Lin L, Hommes D, Lagishetty V, Braun J. NK cells
are biologic and biochemical targets of 6-mercaptopurine in Crohn’s
disease patients. Clin Immunol. 2017;175:82–90.
50. Merkt W, Salzer U, Thiel J, Jandova I, Bergner R, Venhoff AC, et al. Blood
CD3-(CD56 or 16)+ natural killer cell distributions are heterogeneous in
healthy adults and suppressed by azathioprine in patients with ANCA-
associated vasculitides. BMC Immunol. 2021;22(1):26.
51. Cooper DA, Petts V, Luckhurst E, Penny R. The effect of acute and
prolonged administration of prednisolone and ACTH on lymphocyte
subpopulations. Clin Exp Immunol. 1977;28(3):467–73.
52. Yu DT, Clements PJ, Paulus HE, Peter JB, Levy J, Barnett EV. Human
lymphocyte subpopulations. Effect of corticosteroids J Clin Invest.
1974;53(2):565–71.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub‑
lished maps and institutional affiliations.