Artigos - Autoimunidade.pdf
Documento PDF (5.5MB)

Exibir conteúdo Baixar

                    RevIewS
Autoinflammation and autoimmunity
across rheumatic and musculoskeletal
diseases
Zoltán Szekanecz 1 ✉, Iain B. McInnes
Szilvia Benkő5 and Gabriella Szűcs1

, Georg Schett

2

, Szilvia Szamosi1,

3,4

Abstract | Most rheumatic and musculoskeletal diseases (RMDs) can be placed along a spectrum
of disorders, with autoinflammatory diseases (including monogenic systemic autoinflammatory
diseases) and autoimmune diseases (such as systemic lupus erythematosus and antiphospholipid
syndrome) representing the two ends of this spectrum. However, although most autoinflammatory
diseases are characterized by the activation of innate immunity and inflammasomes and classical
autoimmunity typically involves adaptive immune responses, there is some overlap in the features
of autoimmunity and autoinflammation in RMDs. Indeed, some ‘mixed-​pattern’ diseases such
as spondyloarthritis and some forms of rheumatoid arthritis can also be delineated. A better
understanding of the pathogenic pathways of autoinflammation and autoimmunity in RMDs,
as well as the preferential cytokine patterns observed in these diseases, could help us to design
targeted treatment strategies.

1
Division of Rheumatology,
Faculty of Medicine,
University of Debrecen,
Debrecen, Hungary.
2
Institute of Infection,
Immunity and Inflammation,
University of Glasgow,
Glasgow, UK.
3
Department of Internal
Medicine 3, Friedrich
Alexander University
Erlangen-​Nuremberg
and Universitätsklinikum
Erlangen, Erlangen, Germany.
4
Deutsches Zentrum fur
Immuntherapie, Friedrich
Alexander University
Erlangen-​Nuremberg
and Universitätsklinikum
Erlangen, Erlangen, Germany.
5
Department of Physiology,
Faculty of Medicine,
University of Debrecen,
Debrecen, Hungary.

✉e-​mail: szekanecz.zoltan@
med.unideb.hu
https://doi.org/10.1038/
s41584-021-00652-9

When discussing rheumatic and musculoskeletal dis­
eases (RMDs), it is not always clear whether the disease
is strictly an autoimmune disease or is an autoinflamma­
tory disease with unchecked inflammation but without
autoimmunity1–4. Therefore, it is important to revisit the
classification used to describe RMDs1–4.
When considering whether a disease is an auto­
immune disease versus an autoinflammatory disease,
systemic lupus erythematosus (SLE) and mono­genic sys­
temic autoinflammatory diseases (SAIDs) can be
considered as prototypes of autoimmune and auto­
inflammatory diseases, respectively3,4. Autoimmune
diseases are characterized by the loss of immune
tolerance, the recognition of self-​antigens and the acti­
vation of T cells and B cells, followed by the production
of specific autoantibodies and the damage of multi­
ple organs owing to a dysregulated adaptive immune
response1,3,5. Autoinflammatory diseases are not directed
by specific antigens, and they harbour systemic chronic
inflammation without a break in immune tolerance or
the generation of specific autoantibodies4,6. External
environmental factors such as infections, tempera­
ture changes or mechanical stress can also lead to the
development of inflammation and provoke flare in cer­
tain genetic backgrounds, expanding the definition of
autoinflammation4,6.
RMDs are distributed along a spectrum based on
the involvement of autoimmunity and autoinflam­
mation in them (Fig. 1). Monogenic SAIDs are at the

Nature Reviews | RheumAtoloGy

autoinflammatory end of the spectrum, and SLE and
antiphospholipid syndrome (APS) are at the auto­immune
end. Rare monogenic autoimmune diseases such as
autoimmune polyendocrine syndrome 1, immune dys­
regulation, polyendocrinopathy, enteropathy, X-​linked
and autoimmune lymphoproliferative syndrome will
not be discussed in this Review as they are not classi­
cal RMDs7. Diseases related to autoimmunity that are
discussed here include SLE, rheumatoid arthritis (RA),
polyarticular juvenile idiopathic arthritis (pJIA), sys­
temic sclerosis (SSc), APS, primary Sjögren syndrome
(pSS), idiopathic inflammatory myopathies (IIMs),
mixed connective tissue disease and antineutrophil
cytoplasmic antibody (ANCA)-​associated vasculitis
(AAV)3,4,8–10 (Fig. 1). As discussed later, a mechanis­
tic immunological classification of RA has been pro­
posed based on the heterogeneity of disease subtypes8,9.
In addition to monogenic SAIDs, diseases related
to autoinflammation and discussed in this Review
include gout, spondyloarthritis (SpA), systemic juve­
nile idiopathic arthritis (sJIA), oligoarticular juvenile
idiopathic arthritis, adult-​onset Still disease (AOSD),
Behçet disease and Schnitzler syndrome3,4 (Fig. 1). As
described previously, most of these autoimmune and
autoinflammatory diseases can also be considered to
be ‘mixed-​pattern’ conditions4. Indeed, there is no strict
divide between autoimmune and autoinflammatory dis­
eases as some RMDs comprise elements of autoimmun­
ity and autoinflammation. In such mixed-​pattern RMDs,
volume 17 | October 2021 | 585

0123456789();:

Reviews
Key points
• Rheumatic and musculoskeletal diseases (RMDs) form a continuum between classical
autoimmune and autoinflammatory conditions.
• Classical autoinflammatory and autoimmune diseases are associated with the
activation of innate immunity and adaptive immune responses, respectively.
• There are some ‘mixed-​pattern’ disorders that carry the features of both autoimmune
and autoinflammatory conditions, and one disorder might have autoimmune and
autoinflammatory characteristics at different stages of disease development.
• The autoimmune, autoinflammatory or mixed phenotype of RMDs might help us to
develop and administer therapies targeted to specific disease phenotypes.

autoantibody-​mediated pathology has been observed
alongside activation of the innate immune system,
including of Toll-​like receptors (TLRs) and of the inflam­
masome. Moreover, immune cells and mediators char­
acteristic of both autoimmunity and autoinflammation
can be involved in these diseases1,3,5,11 (Fig. 1).
Indeed, in terms of immunity, autoimmune and auto­
inflammatory conditions can have an innate or adaptive
immunological background2,3 (Fig. 2). Innate immunity
delivers non-​specific cellular and humoral immune
responses and confers the first defensive responses
against pathogens. Innate immune responses are usu­
ally directed against pathogen-​associated molecular pat­
terns (PAMPs) or damage-​associated molecular patterns
(DAMPs). Several molecular systems, including TLRs,
NOD-​like receptors (NLRs), the caspase recruitment
domain (CARD) receptor family, proteins of the com­
plement system, cytoplasmic DNA-​sensing molecules
and inflammatory multimolecular complexes such as
inflammasomes, have evolved to permit diverse rec­
ognition and activation and effector function within
innate immunity. Immune cells activated during innate
immune responses include macrophages, natural killer
cells, neutrophils and mast cells (Fig. 2). In addition, other
cell types, such as epithelial and endothelial cells, are also
induced to express molecules recognizing DAMPs and
PAMPs and are classed as ‘innate responders’. Epithelial
barriers and their dysfunction, partially through alter­
ations in the microbiome, might also play a crucial role
in RMDs. The activation of innate immune responses
is primarily characteristic of autoinflammation and
the development of autoinflammatory diseases (Fig. 1).
Within the cytokine superfamilies, the IL-1 family,
TNF superfamily members, IL-6 and the type I inter­
ferons are particularly implicated in innate immune
responses1,4,12–14.
Adaptive immunity is teleologically younger than
innate immunity and exists only in vertebrates. As it
enables an immunological memory to form in response
to the first encounter with a pathogen, a prompt immune
response can develop after consecutive contacts with
the same external stimulus. Adaptive immunity is
pathogen-​specific and driven by T lymphocytes and
B lymphocytes, and long-​term defence can develop.
Temporal and spatial regulation of such a response, as
well as its attenuation, is needed to prevent tissue and
organ damage. The sustained activation of adaptive
immune responses and immunoregulatory defects
can lead to the development of classical autoimmune
diseases1–3,5 (Fig. 1).
586 | October 2021 | volume 17

During the past decade, multiple efforts have been
made to better understand the nature of autoimmun­
ity and autoinflammation1,4, including those using
genome-​wide association studies, mRNA sequencing,
molecular imaging and the study of tissue-​specific anti­
gen and gene expression patterns1,3,4. In this Review,
we first discuss the key features of diseases that are
predominantly autoimmune or predominantly auto­
inflammatory, before describing the overlap between
autoimmunity and autoinflammation in RMDs. We also
underscore mechanisms shared by autoimmunity and
autoinflammation, such as the involvement of patho­
genic pathways that are characteristic of autoinflamma­
tion in autoimmune conditions (and vice versa), and we
highlight how understanding these shared mechanisms
might enable us to enhance the efficacy of therapeutics
and realize the potential of personalized medicine in
rheumatology.

Major features of autoimmune RMDs
SLE, a prototype of systemic autoimmunity, produces
more than 100 autoantibody specificities and mani­
fests in various systemic organs (Fig. 1). SLE is based
on robust T cell and B cell activation and the forma­
tion of immune complexes, whereas cells and media­
tors that are characteristic of autoinflammation, such
as inflammasome activation and the production of
IL-1, do not seem to have a major role in this disease15.
Nonetheless, innate immunity still has an important
role in SLE. Indeed, single-​nucleotide polymorphisms
associated with SLE include those in the genes encod­
ing TLRs (TLR7 and TLR9), complement receptors
(C3, C4 and C1Q) and Fc receptors (FCGR2A and
FCGR3B), all of which are components of the innate
immune response (Table 1). The accumulation of ‘cel­
lular debris’ in tissues and blood in patients with SLE,
including as a result of secondary necrosis and the for­
mation of neutrophil extracellular traps (NETs), leads
to a breach in immune tolerance and the formation
of immune complexes, which triggers the release of
inflammatory mediators and organ damage15,16. This
cell debris-​induced breach in immune tolerance is
closely linked to dysfunction in complement receptors
and Fc receptors. Indeed, mutations in genes encod­
ing proteins of the complement system and the acti­
vation of a type I interferon (that is, IFNα and IFNβ)
signature, which is also a feature of an innate immune
response, are central features of SLE14,15,17. The com­
plement genes responsible for susceptibility to SLE are
C1Q, C2 and C4 (ref.15). Partial or complete deficiency
in C1, C2 or C4 disrupts early steps of the complement
cascade, resulting in inadequate clearance of immune
complexes. In addition, the Fc receptors FcγRIIIA and
FcγRIIIB have anti-​inflammatory activity as they clear
immune complexes, and mutations in genes encoding
these proteins impair this clearance function. In car­
riers of single-​nucleotide polymorphisms associated
with SLE, environmental factors that induce cell death,
such as ultraviolet light, are necessary for development
of the disease15,18–20. In SLE, extracellular DNA triggers
an IFN gene response associated with the production of
IFNα and IFNβ. DNA activates IFN genes (for example,
www.nature.com/nrrheum

0123456789();:

Reviews
IFNA) via the stimulator of interferon genes (STING)–
IRF3 pathway and TLR7 and TLR9 (refs15,19). Eventually,
the persistence of an interferon signature contributes to
disease progression15,18,21.
The importance of the type I interferon signature
and that of other risk alleles associated with components
of the innate immune response has also been described
in the predominantly autoimmune diseases SSc, IIMs
and pSS. For example, in SSc, the type I interferon
signature appears early in disease, before the onset of
fibrosis, and correlates with an increase in the expres­
sion of B cell-​activating factor (BAFF) mRNA (the pro­
tein product of which promotes B cell activation) and
an increase in collagen synthesis22,23. In the IIMs poly­
myositis and dermatomyositis, the expression of type I
interferon-​regulated genes has also been associated
with disease activity24. Furthermore, high expression
of interferon-​induced genes has been observed in the
skin of patients with dermatomyositis25. In pSS, clini­
cal symptoms, disease activity and B cell activation are
also associated with the type I interferon signature26,27.
Finally, certain subsets of RA presumably show a type I
interferon signature that promotes the production of
autoantibodies such as anti-​citrullinated protein anti­
body (ACPA), anti-​carbamylated protein (anti-​CarP)

Features of autoinflammatory RMDs
SAIDs comprise an expanding group of diseases, includ­
ing monogenic diseases caused by inborn errors (also
known as periodic fever syndromes) and adult-​onset
SAIDs such as AOSD, Schnitzler syndrome and idiopathic
recurrent autoimmune pericarditis (IRAP)33–36.
Monogenic autoinflammatory RMDs. In contrast to
autoimmune RMDs, monogenic SAIDs are exclu­
sively autoinflammatory conditions37 (Fig. 1; Table 1).
A common feature of these diseases, which include
both sporadic and monogenic inherited diseases with
an overactive innate immune system, is recurrent
febrile episodes in the absence of infectious agents. The
best described diseases in this group include familial
Mediterranean fever (FMF), periodic fever, aphthosis,
pharyngitis and adenitis syndrome, hyper-​IgD and peri­
odic fever syndrome (also known as mevalonate kinase
deficiency), TNF receptor-​associated periodic syndrome
(TRAPS), Blau syndrome and cryopyrin-​associated peri­
odic syndromes (CAPS). CAPS include three diseases
caused by mutations in NLRP3, the gene encoding the

Mixedpattern

Autoimmune
RMDs

Gout

Spondyloarthritis

Rheumatoid
arthritis

Systemic sclerosis

Schnitzler
syndrome

Oligoarticular
juvenile idiopathic
arthritis

Polyarticular
juvenile idiopathic
arthritis

Mixed connective
tissue disease

Systemic juvenile
idiopathic arthritis

ANCA-associated
vasculitis

Idiopathic
inflammatory
myopathies

Behçet disease

Primary Sjögren
syndrome

Systemic lupus erythematosus and
antiphospholipid syndrome

Monogenic systemic autoinflammatory
diseases

Autoinflammatory
RMDs

and rheumatoid factor17,28–30, and RA also carries other
autoinflammatory features (see below)8,31,32.

Adult-onset Still
disease

Innate
immunity

Adaptive
immunity

No sex
dominance

Female
dominance

Fig. 1 | Spectrum of autoinflammatory, mixed-pattern and autoimmune diseases. Prototypes of a classical
autoinflammatory disease are the group of monogenic systemic autoinflammatory diseases known as periodic fever
syndromes (pink). Prototypes of classical autoimmune disease are systemic lupus erythematosus and antiphospholipid
syndrome (blue). Diseases in the middle of the spectrum might be considered mixed-​pattern rheumatic and musculoskeletal
diseases (RMDs; mixed colour). Indicated by the spectra at the bottom of the figure, classical autoinflammatory conditions
are characterized by a predominance of innate immunity and have no sex dominance. By contrast, classical autoimmune
conditions are associated with more prominent adaptive immune responses and female dominance. ANCA, antineutrophil
cytoplasmic antibody.
Nature Reviews | RheumAtoloGy

volume 17 | October 2021 | 587
0123456789();:

Reviews
Innate immunity

Adaptive immunity

T cell
Macrophage Natural killer cell Dendritic cell
γδ T cell

Neutrophil

Cytokine/
chemokines

Complement

CD4+
T cell

CD8+
T cell

B cell

Memory cell

Natural killer T cell

Plasma cell

Basophil

Eosinophil

Mast cell

Autoinflammation

Antibodies

Mixed pattern

Autoimmunity

Fig. 2 | Cellular mediators of autoimmunity and autoinflammation. Cells of the innate immune system, including
macrophages, natural killer cells, dendritic cells, mast cells and different granulocyte subsets, and the complement system
promote autoinflammation. Cells of the adaptive immune system, including different T lymphocyte subsets, B cells and
plasma cells, as well as T memory cells and B memory cells, are primarily involved in the development of autoimmunity.
Natural killer T cells and γδ T cells are at the crossroads of autoinflammation and autoimmunity and promote the
development of mixed-​pattern immune-​mediated inflammatory diseases. Most of the cells involved in the development
of autoinflammation and autoimmunity produce cytokines and chemokines (as indicated by the blue circles), whereas
plasma cells release antibodies.

NLRP3 protein, namely familial cold autoinflammatory
syndrome, Muckle–Wells syndrome and chronic infan­
tile neurologic cutaneous and articular syndrome38,39.
The clinical features of these monogenic SAIDs have
been discussed elsewhere37–39. Most of these diseases are
caused by inborn errors, although some such as FMF,
TRAPS, CAPS, hyper-​IgD and periodic fever syndrome,
deficiency of adenosine deiminase 2 (ADA2), periodic
fever, aphthosis, pharyngitis and adenitis syndrome, and
type I interferonopathies can also have adult onset33,34.
Monogenic SAIDs are mostly associated with mutations
in MEFV, the gene encoding pyrin, NLRP3, or other
genes encoding proteins that regulate inflammation,
metabolism and body temperature (for example, NOD2;
also known as CARD15)37,39–41 (Table 1). Currently, our
understanding of monogenic SAIDs is moving from a
gene-​centric view towards a systems-​based view, and
various convergent pathways — such as pyrin and the
actin cytoskeleton, protein misfolding and cellular stress,
NF-​κB dysregulation and interferon activation — have
been associated with autoinflammation in SAIDs42.
Molecular pathways underlying autoinflammation.
Activation of the NLRP3 inflammasome and the IL-1β
pathway are key events in the pathogenesis of most
588 | October 2021 | volume 17

monogenic SAIDs and polygenic SAIDs (introduced
below)12,43,44. In the presence of a characteristic genetic
mutation, certain external environmental factors (for
example, infection, smoking or hormonal factors) can
cause uncontrolled activation of the inflammasome,
resulting in the development of a cytokine-​mediated sys­
temic inflammatory condition12,43,44. DAMPs and PAMPs
are involved in the initiation of inflammasome activa­
tion. Activation of the NLRP3 inflammasome is medi­
ated by the NLR family protein NLRP3 and leads to the
activation of caspase 1, which cleaves the cytokine pre­
cursors pro-​IL-1β and pro-​IL-18 to produce the biologi­
cally active forms of IL-1β and IL-18, respectively12,40,41,43.
In response to increased production of IL-1β and IL-18,
the endogenous cytokine antagonists IL-1 receptor
antagonist (IL-1Ra) and IL-18 binding protein (IL-18bp)
restore the balance of these cytokines in the body12,40,41,43.
Loss of function mutation in genes encoding cytokine
antagonists also leads to increased activation of IL-1α
and IL-1β (refs40,41).
Activation of NF-​κB signalling contributes to the
development of certain autoinflammatory diseases, and
NOD2, a NLR family protein in addition to NLRP3 that
recognizes bacterial dipeptides, is an important regu­
lator of NF-​κB signalling. NOD2 mutation has a role
www.nature.com/nrrheum

0123456789();:

Reviews
in the pathogenesis of Blau syndrome and in Crohn’s
disease40.
Monogenic SAIDs associated with IL-1β family acti­
vation include FMF, familial cold autoinflammatory
syndrome, chronic infantile neurologic cutaneous and
articular syndrome, hyper-​IgD and periodic fever syn­
drome, Muckle–Wells syndrome and pyogenic arthritis,
pyoderma gangrenosum and acne40,41. The different gene
mutations present in each disease result in activation of
the NLRP3 inflammasome and uncontrolled secretion
of IL-1β (refs40,41). In addition to IL-1β and IL-18, TNF
is also involved in the pathogenesis of some monogenic
autoinflammatory disorders40,41. Other pathogenetic
mechanisms that affect innate immunity and have been
implicated in the pathogenesis of SIADs include NF-​κB
activation, endoplasmic reticulum stress, mutations
in genes encoding endogenous cytokine antagonists,
dysregulation of actin filament formation (in actinop­
athies), enhanced expression of IFN (in interferonopa­
thies) or a reduction in the enzymatic activity of ADA2
(refs33,34). TRAPS, which is one of the most prevalent
monogenic SAIDs, is associated with heterozygous
variants in TNFRSF1A, the gene encoding TNF recep­
tor 1 (refs33,45,46). Possible pathogenic mechanisms of
TRAPS include enhanced NF-​κB and NLRP3 activa­
tion through increased endoplasmic reticulum stress,
defective autophagy or defective receptor shedding
leading to TNF-​induced cell death and, eventually,
autoinflammation33,45,46.
Polygenic autoinflammatory RMDs. Among polygenic
autoinflammatory conditions we will discuss sJIA and
gout, two well-​known prototypes. sJIA is a typical auto­
inflammatory disease associated with fever, rash, hepato­
splenomegaly and lymphadenopathy, especially in the
early, acute phase47. Genetic and epigenetic changes are
associated with this disease but, although mutations
have been described in several genes, unlike in periodic
Table 1 | Genes associated with common autoimmune and autoinflammatory
disorders
Classification

Disease

Associated genes

Autoimmune
diseases

Systemic lupus erythematosus

TLR7, TLR9, C3, C4, C1Q,
FCGR2A, FCGR3B, IFNA

Systemic sclerosis

IFN signature genes

Idiopathic inflammatory myopathy

IFN signature genes

Autoinflammatory Monogenic systemic
diseases
autoinflammatory diseases

NLRP3, NOD2, MEFV,
TNFRSF1A, MVK, TNFAIP3,
ADA2, TREX1, UBA1

Systemic juvenile idiopathic arthritisa IL1, IL1R, IL6, IL10, IL20,
IL8, MIF
Adult-​onset Still diseasea

MEFV, TNFRSF1A, NLRP3

Behçet disease

MEFV, TNFRSF1A, NLRP3,
HLAB51

Ankylosing spondylitis

HLAB27, ERAP1 (also
known as ARTS1)

Rheumatoid arthritis

HLADRB1, PTPN22,
NLRP3, MEFV, NOD2

a

Mixed-​pattern
diseasesa

This table is not comprehensive and shows only the most common diseases and their genetic
associations. aDiseases that can also be mixed-​pattern diseases.

Nature Reviews | RheumAtoloGy

fever syndromes, none of these mutations alone results
in sJIA47. Gene mutations characteristic of monogenic
diseases (for example, mutations in NLRP3, NOD2
and MEFV) are not observed in sJIA47. sJIA has, rather,
been associated with genes encoding pro-​inflammatory
cytokines (such as IL1, IL1R, IL6, IL10 and IL20) and
other mediators of inflammation (such as IL8 and MIF;
MIF encodes macrophage migration inhibitory fac­
tor)47 (Table 1). The proteins encoded by these genes are
involved in the innate immune response and, ultimately,
create an inflammatory microenvironment; the activa­
tion of effector T cells only occurs as a consequence
of autoinflammation3,47. In the more advanced stage of
sJIA, activation of the adaptive immune system and
joint tissue destruction can be observed, suggesting that
sJIA is associated with the activation of innate and (to a
lesser extent) adaptive immunity at different stages of the
disease48,49. Nonetheless, B cell-​mediated autoimmun­
ity is absent in sJIA. Important questions are how and
when spurious inflammation in sJIA switches to chronic
inflammation1,49, and whether this switch can be pre­
vented or delayed by early intervention with anti-​IL-1
or anti-​IL-6 strategies50.
Autoinflammation is also essential in the develop­
ment of gout and the central event of gouty inflam­
mation is the activation of white blood cells by
monosodium urate (MSU) crystals12,51,52. Cell mem­
brane damage by activated leukocytes and their medi­
ators results in the activation of pattern recognition
receptors, inducing a response against cellular debris
to try to minimize the damage. MSU crystals act as
DAMPs and are phagocytosed through TLR2 and TLR4
to form a phagolysosome. Phagolysosome formation is
followed by activation of the NLRP3 inflammasome,
which leads to the activation of caspase 1 and to the
release of IL-1β and IL-18 (refs12,51,52). The production
and release of the pro-​inflammatory cytokines IL-1,
IL-6 and TNF from cells of the innate immune system,
independent of inflammasome activation, initiate an
inflammatory cascade in which additional mediators of
inflammation, such as matrix metalloproteinases, pros­
taglandins, leukotrienes and reactive oxygen species,
also play a role12,51.
Although monogenic SAIDs, sJIA and gout are the
prototypes of autoinflammatory RMDs, AOSD, Behçet
disease, IRAP, synovitis, acne, pustulosis, hyperostosis,
osteitis syndrome and Schnitzler syndrome can also be
classified as adult-​onset SAIDs33–35,38 (Fig. 1). AOSD is an
acquired fever syndrome characterized by well-​defined
clinical (intermittent fever, typical rash and arthritis)
and laboratory (hyperferritinaemia, leucocytosis, neu­
trophilia and abnormal transaminase levels) features.
AOSD has been associated with an increased production
of cytokines, including of IL-1, IL-6, IL-18 and TNF53.
Activation of the NLRP3 inflammasome and patholog­
ical IL-1 signalling have also been observed in patients
with AOSD53. Mutations in MEFV and TNFRSF1A (the
gene encoding TNF receptor 1) have been described
in patients with AOSD, linking AOSD to monogenic
SAIDs54 (Table 1). Behçet disease is a systemic vascu­
litis affecting the small vessels, and most commonly
manifests as mucosal and genital ulcers and uveitis.
volume 17 | October 2021 | 589

0123456789();:

Reviews
In addition to other cytokines, the NLRP3–IL-1 sys­
tem is important in the development of Behçet disease,
meaning that this is a predominantly autoinflammatory
condition that can also have mixed-​pattern features (see
below)55–57. Again, mutations in MEFV and TNFRSF1A
are more common in this disease compared with other
autoinflammatory conditions4. Schnitzler syndrome is
also an acquired fever syndrome and is characterized
by chronic urticaria associated with monoclonal gam­
mopathy, recurrent fever, bone pain and arthralgia. It is
considered to be a neutrophil dermatosis with notable
involvement of neutrophils, cells that are involved in
innate immunity58. Hereditary factors are unlikely to
play a role in the pathogenesis of this disease based on
its late onset in patients33,36,59,60.

Mixed-​pattern RMDs
Diseases with features of both autoinflammatory and
autoimmune RMDs include SpA and some forms of RA.
These disorders have also been termed mixed-​pattern
RMDs4 (Fig. 1).
As well as ankylosing spondylitis (AS) and pso­
riatic arthritis (PsA), forms of SpA include entero­
pathic arthritis (also known as inflammatory bowel
disease-​associated arthritis), reactive arthritis and
undifferentiated SpA61,62. In contrast to classical auto­
immune diseases, SpA is associated with HLA-​B but
not with HLA-​DR, which is characteristic of polygenic
autoimmune diseases61,63–65. Moreover, unlike other auto­
immune diseases, there is no female dominance in SpA.
Furthermore, SpA has been associated with autoantibod­
ies; some patients with AS and PsA have autoantibodies
to mutated citrullinated vimentin, CarP, sclerostin, heat
shock proteins or β2-​microglobulin61,63–65. CD74 is the
invariable γ-​chain of MHC class II, and anti-​CD74 anti­
bodies are considered to be specific for SpA in European
but not Asian cohorts65. Among cytokines, in addition to
TNF, IL-17 and IL-23 seem to have a predominant role
in mixed-​pattern RMDs61,66. Associations of SpA with
mutations in ERAP1 (also known as ARTS1, encoding
endoplasmic reticulum aminopeptidase 1) and with
MHC class I suggest that T cells interact with cytokine
pathways, including the IL-23–IL-17 axis but not the
IL-1 pathway, in patients with this disease56,57,67 (Table 1).
In terms of autoinflammation, NLRP3 and caspase 1 are
upregulated in AS, suggesting that autoinflammation is
involved in the pathogenesis of this disease68. In short,
features of both autoimmunity (such as autoantibodies)
and autoinflammation (such as gender balance and nat­
ural immune responses to microbial pathogens) have
been identified in SpA61.
RA generally has autoimmune features in the
early phase of the disease but has a macrophage and
fibroblast-​d ominated pathogenesis in the chronic
phase. Thus, RA is an example of a condition in which
the phase of the disease relates to its autoimmune or
autoinflammatory nature4,10,30,69. Five patients with sero­
positive RA had HLA-​DRB1*01 and/or HLA-​DRB1*04
shared epitopes as well as mutations in NLRP3, MEFV or
NOD2 (ref.9) (Table 1). These patients showed features of
autoinflammation and responded to colchicine9. Based
on the findings of this study, the authors proposed
590 | October 2021 | volume 17

an immunology-​based reclassification of RA that includes
classical seropositive autoimmune RA, autoinflamma­
tory seronegative forms of RA and mixed forms of RA
that are seronegative8,9. This proposed reclassification
reflects the commonly accepted idea that RA is a syn­
drome based on different pathophysiologic events rather
than a single disease.
Juvenile idiopathic arthritis can also be a mixed-​
pattern disease with both autoinflammatory and auto­
immune features. For example, pJIA shares many of the
features of adult RA described above47,70. Also, although
sJIA is largely considered to be a SAID dominated by
innate immunity-​driven inflammation, in later stages it
can progress towards an adaptive immunity-​dependent
arthritis47–49.
Among diseases primarily considered to be auto­
inflammatory, AOSD and Behçet disease have also been
associated with adaptive immunity and T cell responses
and thus can also be considered mixed-​p attern
conditions4,56,57. AOSD can be systemic with predom­
inantly autoinflammatory features or have a chronic
articular pattern resembling classical RA, which could
have relevance for therapy. For example, different
pheno­types of AOSD respond to different biologics4,71.
Moreover, genetic analysis has confirmed that sJIA and
AOSD might form a continuum of a single disease.
Specifically, sJIA and AOSD can share common genes,
and the differentiation between these two diseases
is mainly based on the age of onset35. Behçet disease,
a primarily autoinflammatory condition, is also asso­
ciated with the MHC class I molecule HLA-​B51, nota­
ble T cell responses and the production of IL-23 and
IL-17 (refs56,57), highlighting that it also has features of
autoimmune conditions.
Finally, among monogenic SAIDs, haploinsufficiency
of A20 — which is caused by mutations in TNFAIP3,
the gene encoding the NF-​κB regulatory protein A20
(refs33,72) — is a good example of a condition with auto­
immune and autoinflammatory features that result from
the same pathogenetic pathways. This disease carries
characteristics of RA, gout, Behçet disease, AOSD, SLE,
periodic fever, aphthosis, pharyngitis and adenitis syn­
drome, as well as skin, ocular and gastrointestinal symp­
toms. Therefore, diagnosis and differential diagnosis of
haploinsufficiency of A20 is difficult72.
In conclusion, mixed-​pattern RMDs carry both clas­
sical autoimmune and autoinflammatory features and
are often associated with non-​rheumatic conditions1,3,4,8.

Innate immunity in autoimmune RMDs
Having discussed the main features of autoimmune,
autoinflammatory and mixed-​p attern RMDs, it is
important to consider the innate immune mechanisms
that most commonly occur in both autoinflammatory
and autoimmune diseases.
We have already discussed activation of the NLRP
inflammasome and the consequent production of IL-1β
and IL-18 in autoinflammation12,44. However, these fea­
tures have also been demonstrated in autoimmune and
mixed-​pattern conditions. NLRP3 activation and the
consequent production of cytokines, as well as relevant
genetic polymorphisms (for example, in NLRP3 and
www.nature.com/nrrheum

0123456789();:

Reviews
NOD2), have been associated with RA30,73–76, SpA77,78,
pJIA and oligoarticular juvenile idiopathic arthritis70.
NLRP3 is also activated, with inflammasome activa­
tion leading to tissue injury, in autoimmune RMDs
such as RA79,80, SLE76,81,82, SSc83,84, pSS85 and IIMs86.
TLR-​d ependent pathways and abnormal TLR sig­
nalling are also characteristic for SLE, RA and other
autoimmune RMDs82.
Type I interferon is upregulated in genetically based
interferonopathies, which are not always linked to auto­
immunity. STING is a DNA sensor, and a mutation in
the gene encoding this protein can lead to the induc­
tion of genes involved in IFNα and INFβ-​mediated
responses and thus, indirectly, the synthesis of numer­
ous pro-​inflammatory cytokines14,40,87. Rare examples of
these interferonopathies also include STING-​associated
vasculopathy with onset in infancy as well as Aicardi–
Goutiéres syndrome14,40,87. As discussed above, type I
interferon signatures play a key role in autoimmune
diseases such as SLE and can also be involved in RA
and SSc87.
NETs are web-​like structures of decondensed chro­
matin, histones and antimicrobial peptides that are
involved in the defence against pathogens58,88–90 and, pri­
marily, have a role in autoinflammatory conditions such
as gout91,92 or Schnitzler syndrome58. In gout, the forma­
tion of NETs might also be a counter-​regulatory mech­
anism aimed at resolving inflammation91,92. Specifically,
NETs can stop gout episodes by inducing neutrophil
death, encapsulating MSU crystals and inactivating
cytokines 91,92. However, neutrophil activation and
NET formation contribute to autoimmune-​mediated
inflammation in SLE90,93, RA90,92 and AAV90,92.
Prolonged innate immunity-​based inflammation
can induce adaptive immune responses, as described
above for sJIA48. However, this phenomenon can also
be observed in other RMDs. In monogenic SAIDs
and other autoinflammatory diseases, an acute
‘hyper-​inflammatory state’ leading to the resolution of
inflammation within days and a prolonged ‘autonomous
inflammatory state’ have been proposed to occur49,94.
In the latter state, prolonged IL-1β and IL-18 produc­
tion, in part in synergy with IL-6 and IL-23 activation,
can promote T cell differentiation, the induction of T
helper 17 cells (TH17 cells) and the production of IL-17
(refs 49,95) . Moreover, IL-18 can induce adaptive T H1
cells and B cells49. Thus, innate immunity is involved
in some autoimmune RMDs. Finally, a sustained
innate immune response can induce trained immunity
in autoimmune RMDs, which can contribute to the
activation of adaptive immune pathways49,96.

Comorbidities associated with RMDs
Comorbidities are associated with many RMDs and
determine their outcome. The most relevant comorbid­
ities are cardiopulmonary disease (including cardiovas­
cular disease, IRAP and interstitial lung disease (ILD)),
osteoporotic fractures, neuropsychiatric manifestations,
diabetes mellitus and malignancies97,98.
The inflammatory condition accelerated atheroscle­
rosis and the consequent cardiovascular disease can carry
both autoimmune and autoinflammatory features99–101.
Nature Reviews | RheumAtoloGy

The autoantibodies ACPA102,103 and anti-​carP104 might
be involved in the development of atherosclerosis in
RA. Citrullinated proteins have been detected in the
atherosclerotic plaque, suggesting a possible target for
ACPA in RA103. With respect to autoinflammation, in
one large study NLRP3 gene polymorphisms were not
associated with cardiovascular disease in RA105, whereas
in another cohort the presence of the NLRP3Q705K minor
allele doubled the risk of stroke (also known as transient
ischaemic attack) but did not increase the risk of myo­
cardial infarction in RA106. In patients without rheumatic
disease, NLRP3 and caspase 1 transcripts are abundantly
expressed in atherosclerotic plaques107. Polymorphisms
in CARD-​containing protein 8 were not associated with
any type of cardiovascular event in RA106. With respect
to pro-​inflammatory cytokines, inflammatory athero­
sclerosis associated with RMDs has been characterized
by the increased production of TNF and IL-6 (refs99,100).
In addition, both IL-1 and IL-18 are abundantly pro­
duced in the atherosclerotic plaques107,108, and IL-18
is a predictor of mortality in patients with cardiovas­
cular disease109. In patients with SLE, IL-18 produc­
tion has also been associated with kidney damage and
cardiovascular disease82.
The comorbidity IRAP should also be considered
when monitoring and treating RMDs. Recurrent per­
icarditis can occur in viral infections but can also be
associated with various autoimmune RMDs (for exam­
ple, SLE, SSc, IIMs, pSS and RA) and autoinflammatory
RMDs (for example, FMF, TRAPS and Behçet dis­
ease)110,111. IRAP can carry some autoimmune features
as it has been linked to the production of anti-​heart and
anti-​intercalated disk autoantibodies, as well as to auto­
reactive T cells110. However, IRAP has also been associ­
ated with notable NLRP3 activation, and cases resistant
to NSAIDs, corticosteroids and/or colchicine might
respond well to the inhibition of IL-1 (refs110,111). Based
on these observations, IRAP can also be considered an
autoinflammatory disease110–112.
ILD is mostly associated with autoimmune condi­
tions such as SSc or IIMs, and the presence of specific
autoantibodies, such as anti-​S cl70, anti-​PLββ-7 and
anti-​PL-12, correlates with an increased risk of devel­
oping ILD in these diseases113,114. By contrast, there is
limited information on the possible involvement of auto­
inflammation in ILD. One study investigated the role of
NLRP3 inflammasomes in patients with idiopathic pul­
monary fibrosis and in patients with RA and usual inter­
stitial pneumonia. IL-1β and IL-18 levels were elevated
in bronchoalveolar lavage fluid and bronchoalveolar
lavage fluid macrophage cultures from patients with RA
and usual interstitial pneumonia compared with healthy
individuals115. However, the role of autoinflammation in
ILD has not been confirmed.
A great number of autoimmune (for example, SLE),
autoinflammatory (for example, TRAPS and FMF) and
mixed-​pattern (for example, Behçet disease) diseases
also have neuropsychiatric comorbidities. Based on
the nature of these manifestations, these comorbidities
might not have the same pathogenesis; however, neuro­
inflammation could be a common link between these
disorders4,57,116,117.
volume 17 | October 2021 | 591

0123456789();:

Reviews
Finally, most RMDs have been associated with gen­
eralized bone loss leading to osteoporosis and fragility
fractures68,97,118. Proinflammatory cytokines, such as TNF,
IL-1, IL-6 and IL-17 (ref.118), as well as various DAMPs,
including purine metabolites and fatty acids, have been
implicated in inflammatory bone disorder68. Cytokines
and DAMPs both stimulate NLRP3 and NLRC4 inflam­
masomes, and NLRP3-​deficient mice are protected from
bone loss68. Thus, autoinflammation is implicated in
osteoporosis that occurs secondary to RMDs.

Treating RMDs across the spectrum
The pathogenesis of autoimmunity and autoinflam­
mation, especially the cytokine networks characteristic
of these conditions, might enable effective targeting
strategies43,66,119.

and PsA) might also effectively treat classical autoim­
mune diseases. By contrast, cytokine inhibitors such as
those that block IL-1 and TNF, which are effective in
autoinflammatory diseases and in diseases such as RA
with both autoinflammatory and autoimmune features,
show limited efficacy in these autoimmune diseases.
However, the IL-6 inhibitor tocilizumab gave promising
results in SSc149 and might be tried in the treatment of
other autoimmune diseases150,151.
TNF appears to be an excellent target in many
inflammatory diseases, such as RA, AS, PsA and pJIA66.
However, it might not be the optimal target in classical
autoimmune disorders, such as SLE, SSc, AAV or pSS66.

Treating mixed-​pattern diseases. JAK inhibitors have
been approved for treating RMDs with a mixed innate
and adaptive immune activation, such as RA and
Treating autoinflammatory diseases. Autoinflammation SpA, and preliminary data suggest that they show prom­
often responds well to recombinant IL-1RA (anakinra), ise for the treatment of patients with SLE, IIM, pSS, type I
anti-​IL-1β antibody (canakinumab) or recombinant interferonopathies, sJIA, AOSD, Behçet disease and
IL-1R fusion protein (rilonacept)119–121. Canakinumab monogenic SAIDs152. Mixed-​pattern diseases could also
has been registered for the treatment of CAPS, TRAPS, be treated with a combination of therapeutic strategies.
FMF, AOSD, sJIA and refractory gouty flares122,123. In For example, haploinsufficiency of A20, AOSD, Behçet
addition, rilonacept124,125 and anakinra126 are also effec­ disease or sJIA can be treated with TNF, IL-1 or IL-6
tive in treating monogenic SAIDs. Among the less inhibitors based on the dominance of autoinflammatory
common monogenic SAIDs, recombinant IL-18bp can versus autoimmune features in the patient66,71,72.
Finally, trials to inhibit common molecular mech­
be administered in NLRC4 inflammasome-​associated
diseases caused by the overproduction of IL-18 (ref.41). anisms of autoinflammation and autoimmunity, such
In autoinflammatory diseases associated with NF-​κB as inflammasomes or NETs, have been carried out89.
activation, such as TRAPS, IL-1 inhibitors are the first-​ Several inflammasome inhibitors that target components
choice treatment; however, TRAPS also responds well to of the NLRP3 cascade are under investigation for the
TNF inhibitor therapy as the TNF receptor activates the treatment of autoinflammatory conditions12,44,153. Among
NF-​κB pathway41. With respect to gout, IL-1 inhibitors currently used anti-​rheumatic drugs, antimalarials and
are effective in treating refractory flares, with most data JAK inhibitors also inhibit NETs89. Some inhibitors of the
available for canakinumab12,127, although rilonacept128 protein arginine deiminase enzyme involved in protein
and anakinra129,130 are also effective in treating gouty citrullination might also block NET formation89.
flares. For patients with sJIA, canakinumab 131,132,
the anti-​IL-6 receptor antibody tocilizumab 133 and Conclusions
anakinra134 are registered for treatment, and rilonacept135 Autoimmune and autoinflammatory RMDs can be
is also effective in treating this disease. Canakinumab136 considered to be a spectrum of disorders. Monogenic
and anakinra126,137 are effective in, and registered for, SAIDs, and SLE and APS, are likely to represent the
treating patients with AOSD. Rilonacept can be admin­ two ends of this spectrum of RMDs. Autoinflammatory
istered off-​label to patients with AOSD137, and TNF and diseases such as gout, sJIA, Behçet disease, AOSD or
IL-6 inhibitors are also effective in treating patients with Schnitzler syndrome are characterized by the activa­
AOSD32,138. IL-1 inhibitors, such as canakinumab and tion of innate immunity, whereas classical autoimmune
anakinra, also showed efficacy in treating patients with diseases such as SSc, IIM, pSS, mixed connective tissue
Behçet disease139. All IL-1 inhibitors are also effective in disease or seropositive RA are associated with adaptive
patients with Schnitzler syndrome36,140.
immune responses and the production of autoantibod­
ies. In addition to the fact that both autoinflammatory
Treating autoimmune diseases. In autoimmune diseases, and autoimmune diseases can carry some features of
T cells, B cells and their cytokines play a notable role the other disease type, there are mixed-​pattern diseases
in disease pathogenesis, and the B cell inhibitor rituxi­ that include SpA, AAV, pJIA, oligoarticular juvenile idio­
mab can be used off-​label for treating most autoimmune pathic arthritis and some forms of RA. The involvement
diseases, including SLE141, SSc142, dermatomyositis143 of characteristic pathogenic proteins or pathways, such
and pSS144. Belimumab, an anti-​BAFF antibody, has as of PAMPs, DAMPs, pattern recognition receptors,
been approved for the treatment of SLE145, and the complement or inflammasome activation in auto­
CTLA4–Ig fusion protein abatacept can also be admin­ inflammation, or of type I interferon signatures and the
istered to inhibit T cells in selected cases of SLE146, SSc147 production of autoantibodies in autoimmunity, along
and pSS148. It is also possible that cytokines that activate with preferential cytokine patterns, might help inform
TH17 cells (such as IL-17 and IL-23) and are used to the design of directed treatment strategies.
treat RMDs with a mixed innate (neutrophil activation)
and adaptive (T cell activation) background (such as AS Published online 2 August 2021
592 | October 2021 | volume 17

www.nature.com/nrrheum
0123456789();:

Reviews
1.

2.
3.

4.

5.
6.

7.

8.

9.

10.
11.

12.

13.

14.

15.
16.

17.

18.

19.

20.

21.
22.

23.
24.

25.

26.

27.

28.

Hedrich, C. M. & Tsokos, G. C. Bridging the gap
between autoinflammation and autoimmunity.
Clin. Immunol. 147, 151–154 (2013).
Abbas, A., Lichtman A. H., Pillai S. Cellular and
Molecular Immunology 9th edn (Elsevier, 2017).
Hedrich, C. M. Shaping the spectrum — from
autoinflammation to autoimmunity. Clin. Immunol.
165, 21–28 (2016).
McGonagle, D. & McDermott, M. F. A proposed
classification of the immunological diseases.
PLoS Med. 3, e297 (2006).
Davidson, A. & Diamond, B. Autoimmune diseases.
N. Engl. J. Med. 345, 340–350 (2001).
Masters, S. L. Broadening the definition of
autoinflammation. Semin. Immunopathol. 37,
311–312 (2015).
Michels, A. W. & Gottlieb, P. A. Autoimmune
polyglandular syndromes. Nat. Rev. Endocrinol. 6,
270–277 (2010).
McGonagle, D., Watad, A. & Savic, S. Mechanistic
immunological based classification of rheumatoid
arthritis. Autoimmun. Rev. 17, 1115–1123 (2018).
Savic, S. et al. Autoimmune-​autoinflammatory
rheumatoid arthritis overlaps: a rare but potentially
important subgroup of diseases. RMD Open 3,
e000550 (2017).
Smolen, J. S. et al. Rheumatoid arthritis. Nat. Rev.
Dis. Prim. 4, 18001 (2018).
Kuek, A., Hazleman, B. L. & Ostor, A. J. Immune-​
mediated inflammatory diseases (IMIDs) and biologic
therapy: a medical revolution. Postgrad. Med. J. 83,
251–260 (2007).
Szekanecz, Z., Szamosi, S., Kovacs, G. E., Kocsis, E.
& Benko, S. The NLRP3 inflammasome–interleukin 1
pathway as a therapeutic target in gout. Arch.
Biochem. Biophys. 670, 82–93 (2019).
Frizinsky, S. et al. The innate immune perspective
of autoimmune and autoinflammatory conditions.
Rheumatology 58, vi1–vi8 (2019).
Melki, I. & Fremond, M. L. Type I Interferonopathies:
from a novel concept to targeted therapeutics.
Curr. Rheumatol. Rep. 22, 32 (2020).
Tsokos, G. C. Systemic lupus erythematosus. N. Engl.
J. Med. 365, 2110–2121 (2011).
Cook, H. T. & Botto, M. Mechanisms of disease: the
complement system and the pathogenesis of systemic
lupus erythematosus. Nat. Clin. Pract. Rheumatol. 2,
330–337 (2006).
David, T., Ling, S. F. & Barton, A. Genetics of immune-​
mediated inflammatory diseases. Clin. Exp. Immunol.
193, 3–12 (2018).
Crow, M. K. Advances in understanding the role of
type I interferons in systemic lupus erythematosus.
Curr. Opin. Rheumatol. 26, 467–474 (2014).
Crispin, J. C., Hedrich, C. M. & Tsokos, G. C. Gene-​
function studies in systemic lupus erythematosus.
Nat. Rev. Rheumatol. 9, 476–484 (2013).
Brown, E. E., Edberg, J. C. & Kimberly, R. P. Fc receptor
genes and the systemic lupus erythematosus diathesis.
Autoimmunity 40, 567–581 (2007).
Trinchieri, G. Type I interferon: friend or foe?
J. Exp. Med. 207, 2053–2063 (2010).
Brkic, Z. et al. The interferon type I signature is
present in systemic sclerosis before overt fibrosis and
might contribute to its pathogenesis through high
BAFF gene expression and high collagen synthesis.
Ann. Rheum. Dis. 75, 1567–1573 (2016).
Wu, M. & Assassi, S. The role of type 1 interferon
in systemic sclerosis. Front. Immunol. 4, 266 (2013).
Greenberg, S. A. et al. Relationship between disease
activity and type 1 interferon- and other cytokine-​
inducible gene expression in blood in dermatomyositis
and polymyositis. Genes. Immun. 13, 207–213
(2012).
Wong, D. et al. Interferon and biologic signatures in
dermatomyositis skin: specificity and heterogeneity
across diseases. PLoS ONE 7, e29161 (2012).
Chiche, L. & Cornec, D. Mysterious uncoupled clinical
symptoms and interferon signature in Sjogren’s
syndrome: limitations of current approaches for
unravelling complexity? Rheumatology 59, 5–6
(2019).
Brkic, Z. et al. Prevalence of interferon type I signature
in CD14 monocytes of patients with Sjogren’s
syndrome and association with disease activity
and BAFF gene expression. Ann. Rheum. Dis. 72,
728–735 (2013).
Klareskog, L., Malmstrom, V., Lundberg, K.,
Padyukov, L. & Alfredsson, L. Smoking, citrullination
and genetic variability in the immunopathogenesis
of rheumatoid arthritis. Semin. Immunol. 23, 92–98
(2011).

29. Gordon, R. A., Grigoriev, G., Lee, A., Kalliolias, G. D.
& Ivashkiv, L. B. The interferon signature and STAT1
expression in rheumatoid arthritis synovial fluid
macrophages are induced by tumor necrosis
factor α and counter-​regulated by the synovial fluid
microenvironment. Arthritis Rheum. 64, 3119–3128
(2012).
30. McInnes, I. B. & O’Dell, J. R. State-​of-the-​art:
rheumatoid arthritis. Ann. Rheum. Dis. 69,
1898–1906 (2010).
31. Kahlenberg, J. M. & Kang, I. Advances in disease
mechanisms and translational technologies:
clinicopathologic significance of inflammasome
activation in autoimmune diseases. Arthritis
Rheumatol. 72, 386–395 (2020).
32. Kraetsch, H. G., Antoni, C., Kalden, J. R. & Manger, B.
Successful treatment of a small cohort of patients
with adult onset of Still’s disease with infliximab:
first experiences. Ann. Rheum. Dis. 60 (Suppl. 3),
iii55–iii57 (2001).
33. Betrains, A. et al. Systemic autoinflammatory
disease in adults. Autoimmun. Rev. 20, 102774
(2021).
34. Krainer, J., Siebenhandl, S. & Weinhausel, A. Systemic
autoinflammatory diseases. J. Autoimmun. 109,
102421 (2020).
35. Nirmala, N. et al. Gene-​expression analysis of adult-​
onset Still’s disease and systemic juvenile idiopathic
arthritis is consistent with a continuum of a single
disease entity. Pediatr. Rheumatol. Online J. 13, 50
(2015).
36. Rowczenio, D. M. et al. Molecular genetic
investigation, clinical features, and response to
treatment in 21 patients with Schnitzler syndrome.
Blood 131, 974–981 (2018).
37. Georgin-​Lavialle, S. et al. Systemic autoinflammatory
diseases: clinical state of the art. Best Pract. Res.
Clin. Rheumatol. 34, 101529 (2020).
38. Ter Haar, N. M. et al. Development of the
autoinflammatory disease damage index (ADDI).
Ann. Rheum. Dis. 76, 821–830 (2017).
39. ter Haar, N. M. et al. Recommendations for the
management of autoinflammatory diseases.
Ann. Rheum. Dis. 74, 1636–1644 (2015).
40. Martinon, F. & Aksentijevich, I. New players driving
inflammation in monogenic autoinflammatory
diseases. Nat. Rev. Rheumatol. 11, 11–20 (2015).
41. Holzinger, D., Kessel, C., Omenetti, A. & Gattorno, M.
From bench to bedside and back again: translational
research in autoinflammation. Nat. Rev. Rheumatol.
11, 573–585 (2015).
42. Savic, S., Caseley, E. A. & McDermott, M. F. Moving
towards a systems-​based classification of innate
immune-​mediated diseases. Nat. Rev. Rheumatol. 16,
222–237 (2020).
43. Dinarello, C. A. The IL-1 family of cytokines and
receptors in rheumatic diseases. Nat. Rev. Rheumatol.
15, 612–632 (2019).
44. Swanson, K. V., Deng, M. & Ting, J. P. The NLRP3
inflammasome: molecular activation and regulation
to therapeutics. Nat. Rev. Immunol. 19, 477–489
(2019).
45. Cudrici, C., Deuitch, N. & Aksentijevich, I. Revisiting
TNF receptor-​associated periodic syndrome (TRAPS):
current perspectives. Int. J. Mol. Sci. 21, 3263
(2020).
46. McDermott, M. F. et al. Germline mutations in the
extracellular domains of the 55 kDa TNF receptor,
TNFR1, define a family of dominantly inherited
autoinflammatory syndromes. Cell 97, 133–144
(1999).
47. Bruck, N., Schnabel, A. & Hedrich, C. M. Current
understanding of the pathophysiology of systemic
juvenile idiopathic arthritis (sJIA) and target-​directed
therapeutic approaches. Clin. Immunol. 159, 72–83
(2015).
48. Kessel, C., Hedrich, C. M. & Foell, D. Innately adaptive
or truly autoimmune: is there something unique about
systemic juvenile idiopathic arthritis? Arthritis
Rheumatol. 72, 210–219 (2020).
49. Ter Haar, N. M., Jansen, M. H. A., Frenkel, J. F. &
Vastert, S. J. How autoinflammation may turn into
autoimmune inflammation: insights from monogenetic
and complex IL-1 mediated auto-​inflammatory
diseases. Clin. Immunol. 219, 108538 (2020).
50. Nigrovic, P. A. Review: is there a window of
opportunity for treatment of systemic juvenile
idiopathic arthritis? Arthritis Rheumatol. 66,
1405–1413 (2014).
51. Rock, K. L., Kataoka, H. & Lai, J. J. Uric acid
as a danger signal in gout and its comorbidities.
Nat. Rev. Rheumatol. 9, 13–23 (2013).

Nature Reviews | RheumAtoloGy

52. Martinon, F., Petrilli, V., Mayor, A., Tardivel, A. &
Tschopp, J. Gout-​associated uric acid crystals activate
the NALP3 inflammasome. Nature 440, 237–241
(2006).
53. Gerfaud-​Valentin, M., Jamilloux, Y., Iwaz, J. & Seve, P.
Adult-​onset Still’s disease. Autoimmun. Rev. 13,
708–722 (2014).
54. Sighart, R. et al. Evidence for genetic overlap between
adult onset Still’s disease and hereditary periodic fever
syndromes. Rheumatol. Int. 38, 111–120 (2018).
55. Marshall, S. E. Behcet’s disease. Best Pract. Res.
Clin. Rheumatol. 18, 291–311 (2004).
56. McGonagle, D., Aydin, S. Z., Gul, A., Mahr, A. &
Direskeneli, H. ‘MHC-​I-opathy’-unified concept for
spondyloarthritis and Behcet disease. Nat. Rev.
Rheumatol. 11, 731–740 (2015).
57. Gul, A. Behcet’s disease as an autoinflammatory
disorder. Curr. Drug Targets Inflamm. Allergy 4,
81–83 (2005).
58. Bonnekoh, H. et al. Skin and systemic inflammation in
Schnitzler’s syndrome are associated with neutrophil
extracellular trap formation. Front. Immunol. 10, 546
(2019).
59. Simon, A. et al. Schnitzler’s syndrome: diagnosis,
treatment, and follow-​up. Allergy 68, 562–568
(2013).
60. Pathak, S. et al. Exploratory study of MYD88 L265P,
rare NLRP3 variants, and clonal hematopoiesis
prevalence in patients with Schnitzler syndrome.
Arthritis Rheumatol. 71, 2121–2125 (2019).
61. Generali, E., Bose, T., Selmi, C., Voncken, J. W. &
Damoiseaux, J. Nature versus nurture in the
spectrum of rheumatic diseases: classification of
spondyloarthritis as autoimmune or autoinflammatory.
Autoimmun. Rev. 17, 935–941 (2018).
62. Deodhar, A., Miossec, P. & Baraliakos, X.
Is undifferentiated spondyloarthritis a discrete entity?
A debate. Autoimmun. Rev. 17, 29–32 (2018).
63. Chimenti, M. S. et al. Auto-​reactions, autoimmunity
and psoriatic arthritis. Autoimmun. Rev. 14,
1142–1146 (2015).
64. Bodnar, N. et al. Anti-​mutated citrullinated vimentin
(anti-​MCV) and anti-65 kDa heat shock protein
(anti-hsp65): new biomarkers in ankylosing
spondylitis. Jt. Bone Spine 79, 63–66 (2012).
65. Liu, Y., Liao, X. & Shi, G. Autoantibodies in
spondyloarthritis, focusing on anti-​CD74 antibodies.
Front. Immunol. 10, 5 (2019).
66. Schett, G., Elewaut, D., McInnes, I. B., Dayer, J. M.
& Neurath, M. F. How cytokine networks fuel
inflammation: toward a cytokine-​based disease
taxonomy. Nat. Med. 19, 822–824 (2013).
67. Pazar, B. et al. Association of ARTS1 gene
polymorphisms with ankylosing spondylitis in
the Hungarian population: the rs27044 variant is
associated with HLA-​B*2705 subtype in Hungarian
patients with ankylosing spondylitis. J. Rheumatol. 37,
379–384 (2010).
68. Alippe, Y. & Mbalaviele, G. Omnipresence of
inflammasome activities in inflammatory bone
diseases. Semin. Immunopathol. 41, 607–618 (2019).
69. Szekanecz, Z. & Koch, A. E. Macrophages and their
products in rheumatoid arthritis. Curr. Opin.
Rheumatol. 19, 289–295 (2007).
70. Yang, C. A., Huang, S. T. & Chiang, B. L. Association
of NLRP3 and CARD8 genetic polymorphisms with
juvenile idiopathic arthritis in a Taiwanese population.
Scand. J. Rheumatol. 43, 146–152 (2014).
71. Maria, A. T. et al. Adult onset Still’s disease (AOSD)
in the era of biologic therapies: dichotomous view for
cytokine and clinical expressions. Autoimmun. Rev. 13,
1149–1159 (2014).
72. Zhou, Q. et al. Loss-​of-function mutations in TNFAIP3
leading to A20 haploinsufficiency cause an early-​onset
autoinflammatory disease. Nat. Genet. 48, 67–73
(2016).
73. Guo, C. et al. NLRP3 inflammasome activation
contributes to the pathogenesis of rheumatoid
arthritis. Clin. Exp. Immunol. 194, 231–243
(2018).
74. Mathews, R. J. et al. Evidence of NLRP3inflammasome activation in rheumatoid arthritis (RA);
genetic variants within the NLRP3-inflammasome
complex in relation to susceptibility to RA and
response to anti-​TNF treatment. Ann. Rheum. Dis. 73,
1202–1210 (2014).
75. Dong, X. et al. ACPAs promote IL-1β production
in rheumatoid arthritis by activating the NLRP3
inflammasome. Cell Mol. Immunol. 17, 261–271
(2019).
76. Kahlenberg, J. M. & Kaplan, M. J. The inflammasome
and lupus: another innate immune mechanism

volume 17 | October 2021 | 593
0123456789();:

Reviews
contributing to disease pathogenesis? Curr. Opin.
Rheumatol. 26, 475–481 (2014).
77. Kim, S. K., Cho, Y. J. & Choe, J. Y. NLRP3
inflammasomes and NLRP3 inflammasome-​derived
proinflammatory cytokines in peripheral blood
mononuclear cells of patients with ankylosing
spondylitis. Clin. Chim. Acta 486, 269–274 (2018).
78. Kastbom, A. et al. Genetic variants in CARD8 but not
in NLRP3 are associated with ankylosing spondylitis.
Scand. J. Rheumatol. 42, 465–468 (2013).
79. Strand, V. & Kavanaugh, A. F. The role of interleukin-1
in bone resorption in rheumatoid arthritis.
Rheumatology 43 (Suppl. 3), iii10–iii16 (2004).
80. Watt, I. & Cobby, M. Treatment of rheumatoid arthritis
patients with interleukin-1 receptor antagonist:
radiologic assessment. Semin. Arthritis Rheum. 30,
21–25 (2001).
81. Guo, C. et al. Pathogenesis of lupus nephritis: RIP3
dependent necroptosis and NLRP3 inflammasome
activation. J. Autoimmun. 103, 102286 (2019).
82. Deuteraiou, K., Kitas, G., Garyfallos, A. &
Dimitroulas, T. Novel insights into the role of
inflammasomes in autoimmune and metabolic
rheumatic diseases. Rheumatol. Int. 38, 1345–1354
(2018).
83. Henderson, J. & O’Reilly, S. Inflammasome lights up in
systemic sclerosis. Arthritis Res. Ther. 19, 205 (2017).
84. Martinez-​Godinez, M. A. et al. Expression of NLRP3
inflammasome, cytokines and vascular mediators in the
skin of systemic sclerosis patients. Isr. Med. Assoc. J.
17, 5–10 (2015).
85. Vakrakou, A. G. et al. Systemic activation of NLRP3
inflammasome in patients with severe primary
Sjogren’s syndrome fueled by inflammagenic DNA
accumulations. J. Autoimmun. 91, 23–33 (2018).
86. Yin, X., Han, G. C., Jiang, X. W., Shi, Q. & Pu, C. Q.
Increased expression of the NOD-​like receptor
family, pyrin domain containing 3 inflammasome
in dermatomyositis and polymyositis is a potential
contributor to their pathogenesis. Chin. Med. J. 129,
1047–1052 (2016).
87. Barrat, F. J., Crow, M. K. & Ivashkiv, L. B. Interferon
target-​gene expression and epigenomic signatures in
health and disease. Nat. Immunol. 20, 1574–1583
(2019).
88. Brinkmann, V. et al. Neutrophil extracellular traps kill
bacteria. Science 303, 1532–1535 (2004).
89. Goel, R. R. & Kaplan, M. J. Deadliest catch: neutrophil
extracellular traps in autoimmunity. Curr. Opin.
Rheumatol. 32, 64–70 (2020).
90. Mutua, V. & Gershwin, L. J. A review of neutrophil
extracellular traps (NETs) in disease: potential anti-​
NETs therapeutics. Clin. Rev. Allergy Immunol. https://
doi.org/10.1007/s12016-020-08804-7 (2020).
91. Schett, G., Schauer, C., Hoffmann, M. & Herrmann, M.
Why does the gout attack stop? A roadmap for the
immune pathogenesis of gout. RMD Open 1,
e000046 (2015).
92. Delgado-​Rizo, V. et al. Neutrophil extracellular traps
and its implications in inflammation: an overview.
Front. Immunol. 8, 81 (2017).
93. Villanueva, E. et al. Netting neutrophils induce
endothelial damage, infiltrate tissues, and expose
immunostimulatory molecules in systemic lupus
erythematosus. J. Immunol. 187, 538–552 (2011).
94. Gul, A. Dynamics of inflammatory response in
autoinflammatory disorders: autonomous and
hyperinflammatory states. Front. Immunol. 9, 2422
(2018).
95. Ikeda, S. et al. Excess IL-1 signaling enhances
the development of TH17 cells by downregulating
TGF-​β-induced Foxp3 expression. J. Immunol. 192,
1449–1458 (2014).
96. Netea, M. G. et al. Trained immunity: a program of
innate immune memory in health and disease. Science
352, aaf1098 (2016).
97. Dougados, M. et al. Prevalence of comorbidities
in rheumatoid arthritis and evaluation of their
monitoring: results of an international, cross-​sectional
study (COMORA). Ann. Rheum. Dis. 73, 62–68
(2014).
98. Radner, H., Yoshida, K., Smolen, J. S. & Solomon, D. H.
Multimorbidity and rheumatic conditions — enhancing
the concept of comorbidity. Nat. Rev. Rheumatol. 10,
252–256 (2014).
99. Szekanecz, Z. et al. Autoimmune atherosclerosis in
3D: how it develops, how to diagnose and what to do.
Autoimmun. Rev. 15, 756–769 (2016).
100. Szekanecz, Z., Kerekes, G., Kardos, Z., Baráth, Z.
& Tamási, L. Mechanisms of inflammatory
atherosclerosis in rheumatoid arthritis. Curr. Immunol.
Rev. 12, 35–46 (2016).

101. Agca, R. et al. EULAR recommendations for
cardiovascular disease risk management in patients
with rheumatoid arthritis and other forms of
inflammatory joint disorders: 2015/2016 update.
Ann. Rheum. Dis. 76, 17–28 (2017).
102. Geraldino-​Pardilla, L. et al. Association of anti-​
citrullinated peptide antibodies with coronary artery
calcification in rheumatoid arthritis. Arthritis Care Res.
69, 1276–1281 (2017).
103. Sokolove, J. et al. Brief report: citrullination within
the atherosclerotic plaque: a potential target for the
anti-​citrullinated protein antibody response in
rheumatoid arthritis. Arthritis Rheum. 65,
1719–1724 (2013).
104. Spinelli, F. R. et al. Association between antibodies to
carbamylated proteins and subclinical atherosclerosis
in rheumatoid arthritis patients. BMC Musculoskelet.
Disord. 18, 214 (2017).
105. Lopez-​Mejias, R. et al. Influence of elevated-​CRP
level-​related polymorphisms in non-​rheumatic
Caucasians on the risk of subclinical atherosclerosis
and cardiovascular disease in rheumatoid arthritis.
Sci. Rep. 6, 31979 (2016).
106. Kastbom, A., Arlestig, L. & Rantapaa-​Dahlqvist, S.
Genetic variants of the NLRP3 inflammasome are
associated with stroke in patients with rheumatoid
arthritis. J. Rheumatol. 42, 1740–1745 (2015).
107. Paramel Varghese, G. et al. NLRP3 inflammasome
expression and activation in human atherosclerosis.
J. Am. Heart Assoc. 5, e003031 (2016).
108. Ahmed, A. et al. Brief report: proatherogenic cytokine
microenvironment in the aortic adventitia of patients
with rheumatoid arthritis. Arthritis Rheumatol. 68,
1361–1366 (2016).
109. Blankenberg, S. et al. Interleukin-18 is a strong
predictor of cardiovascular death in stable and
unstable angina. Circulation 106, 24–30 (2002).
110. Cacoub, P. & Marques, C. Acute recurrent pericarditis:
from pathophysiology towards new treatment strategy.
Heart 106, 1046–1051 (2020).
111. Kontzias, A., Barkhodari, A. & Yao, Q. Pericarditis in
systemic rheumatologic diseases. Curr. Cardiol. Rep.
22, 142 (2020).
112. Cantarini, L. et al. Autoimmunity and
autoinflammation as the yin and yang of idiopathic
recurrent acute pericarditis. Autoimmun. Rev. 14,
90–97 (2015).
113. Mecoli, C. A. & Christopher-​Stine, L. Management
of interstitial lung disease in patients with myositis
specific autoantibodies. Curr. Rheumatol. Rep. 20, 27
(2018).
114. Castillo-​Tandazo, W., Gonzalez, J. & Flores-​Fortty, A.
Pathogenesis and therapeutics of interstitial lung
disease in systemic sclerosis. Curr. Rheumatol. Rev. 9,
105–112 (2013).
115. Lasithiotaki, I. et al. NLRP3 inflammasome expression
in idiopathic pulmonary fibrosis and rheumatoid lung.
Eur. Respir. J. 47, 910–918 (2016).
116. Duarte-​Delgado, N. P., Vasquez, G. &
Ortiz-​Reyes, B. L. Blood–brain barrier disruption and
neuroinflammation as pathophysiological mechanisms
of the diffuse manifestations of neuropsychiatric
systemic lupus erythematosus. Autoimmun. Rev. 18,
426–432 (2019).
117. Masson, C. et al. Adult Still’s disease: part I.
Manifestations and complications in sixty-​five
cases in France. Rev. Rhum. Engl. Ed. 62, 748–757
(1995).
118. Szentpetery, A. et al. Effects of targeted therapies
on the bone in arthritides. Autoimmun. Rev. 16,
313–320 (2017).
119. Dinarello, C. A. & van der Meer, J. W. Treating
inflammation by blocking interleukin-1 in humans.
Semin. Immunol. 25, 469–484 (2013).
120. Cantarini, L. et al. Interleukin-1: Ariadne’s thread
in autoinflammatory and autoimmune disorders.
Isr. Med. Assoc. J. 17, 93–97 (2015).
121. Schett, G., Dayer, J. M. & Manger, B. Interleukin-1
function and role in rheumatic disease. Nat. Rev.
Rheumatol. 12, 14–24 (2016).
122. European Commission. Ilaris alkalmazási előírás.
https://ec.europa.eu/health/documents/community-​
register/2016/20160801135455/anx_135455_hu.pdf
(2016).
123. Kuemmerle-​Deschner, J. B. et al. Canakinumab
(ACZ885, a fully human IgG1 anti-​IL-1β mAb) induces
sustained remission in pediatric patients with
cryopyrin-​associated periodic syndrome (CAPS).
Arthritis Res. Ther. 13, R34 (2011).
124. Hoffman, H. M. et al. Long-​term efficacy and
safety profile of rilonacept in the treatment of
cryopryin-​associated periodic syndromes: results

594 | October 2021 | volume 17

of a 72-week open-​label extension study. Clin. Ther.
34, 2091–2103 (2012).
125. Hoffman, H. M. et al. Efficacy and safety of rilonacept
(interleukin-1 Trap) in patients with cryopyrin-​
associated periodic syndromes: results from two
sequential placebo-​controlled studies. Arthritis
Rheum. 58, 2443–2452 (2008).
126. Kuemmerle-​Deschner, J. B. et al. Efficacy and safety
of anakinra therapy in pediatric and adult patients
with the autoinflammatory Muckle–Wells syndrome.
Arthritis Rheum. 63, 840–849 (2011).
127. Schlesinger, N. Canakinumab in gout. Expert Opin.
Biol. Ther. 12, 1265–1275 (2012).
128. Terkeltaub, R. et al. The interleukin 1 inhibitor
rilonacept in treatment of chronic gouty arthritis:
results of a placebo-​controlled, monosequence
crossover, non-​randomised, single-​blind pilot study.
Ann. Rheum. Dis. 68, 1613–1617 (2009).
129. So, A., De Smedt, T., Revaz, S. & Tschopp, J. A pilot
study of IL-1 inhibition by anakinra in acute gout.
Arthritis Res. Ther. 9, R28 (2007).
130. Vitale, A., Cantarini, L., Rigante, D., Bardelli, M. &
Galeazzi, M. Anakinra treatment in patients with gout
and type 2 diabetes. Clin. Rheumatol. 34, 981–984
(2015).
131. Ruperto, N. et al. Two randomized trials of
canakinumab in systemic juvenile idiopathic arthritis.
N. Engl. J. Med. 367, 2396–2406 (2012).
132. Ruperto, N. et al. A phase II, multicenter, open-​label
study evaluating dosing and preliminary safety and
efficacy of canakinumab in systemic juvenile idiopathic
arthritis with active systemic features. Arthritis Rheum.
64, 557–567 (2012).
133. Yokota, S. et al. Efficacy and safety of tocilizumab
in patients with systemic-​onset juvenile idiopathic
arthritis: a randomised, double-​blind, placebo-​
controlled, withdrawal phase III trial. Lancet 371,
998–1006 (2008).
134. Quartier, P. et al. A multicentre, randomised,
double-​blind, placebo-​controlled trial with the
interleukin-1 receptor antagonist anakinra in
patients with systemic-​onset juvenile idiopathic
arthritis (ANAJIS trial). Ann. Rheum. Dis. 70,
747–754 (2011).
135. Ilowite, N. T. et al. Randomized, double-​blind, placebo-​
controlled trial of the efficacy and safety of rilonacept
in the treatment of systemic juvenile idiopathic
arthritis. Arthritis Rheumatol. 66, 2570–2579
(2014).
136. Kedor, C. et al. Canakinumab for Treatment of Adult-​
Onset Still’s Disease to Achieve Reduction of Arthritic
Manifestation (CONSIDER): phase II, randomised,
double-​blind, placebo-​controlled, multicentre,
investigator-​initiated trial. Ann. Rheum. Dis. 79,
1090–1097 (2020).
137. Junge, G., Mason, J. & Feist, E. Adult onset Still’s
disease — the evidence that anti-​interleukin-1
treatment is effective and well-​tolerated
(a comprehensive literature review). Semin.
Arthritis Rheum. 47, 295–302 (2017).
138. Castaneda, S. et al. Tocilizumab for the treatment of
adult-​onset Still’s disease. Expert. Opin. Biol. Ther. 19,
273–286 (2019).
139. Vitale, A. et al. Interleukin-1 inhibition in Behcet’s
disease. Isr. Med. Assoc. J. 18, 171–176 (2016).
140. de Koning, H. D. et al. Sustained efficacy of the
monoclonal anti-​interleukin-1β antibody canakinumab
in a 9-month trial in Schnitzler’s syndrome.
Ann. Rheum. Dis. 72, 1634–1638 (2013).
141. Garcia-​Carrasco, M. et al. Use of rituximab in patients
with systemic lupus erythematosus: an update.
Autoimmun. Rev. 8, 343–348 (2009).
142. McQueen, F. M. & Solanki, K. Rituximab in
diffuse cutaneous systemic sclerosis: should we
be using it today? Rheumatology 54, 757–767
(2015).
143. Rios Fernandez, R., Callejas Rubio, J. L.,
Sanchez Cano, D., Saez Moreno, J. A. &
Ortego Centeno, N. Rituximab in the treatment of
dermatomyositis and other inflammatory myopathies.
A report of 4 cases and review of the literature.
Clin. Exp. Rheumatol. 27, 1009–1016 (2009).
144. Grigoriadou, S. et al. B cell depletion with rituximab
in the treatment of primary Sjogren’s syndrome: what
have we learnt? Clin. Exp. Rheumatol. 37 (Suppl 118),
217–224 (2019).
145. Navarra, S. V. et al. Efficacy and safety of belimumab
in patients with active systemic lupus erythematosus:
a randomised, placebo-​controlled, phase 3 trial.
Lancet 377, 721–731 (2011).
146. Pimentel-​Quiroz, V. R., Ugarte-​Gil, M. F. &
Alarcon, G. S. Abatacept for the treatment of systemic

www.nature.com/nrrheum
0123456789();:

Reviews
lupus erythematosus. Expert Opin. Invest. Drugs 25,
493–499 (2016).
147. Boleto, G., Allanore, Y. & Avouac, J. Targeting
costimulatory pathways in systemic sclerosis.
Front. Immunol. 9, 2998 (2018).
148. Machado, A. C. et al. Effectiveness and safety of
abatacept for the treatment of patients with primary
Sjogren’s syndrome. Clin. Rheumatol. 39, 243–248
(2019).
149. Khanna, D. et al. Safety and efficacy of subcutaneous
tocilizumab in systemic sclerosis: results from the
open-​label period of a phase II randomised controlled
trial (faSScinate). Ann. Rheum. Dis. 77, 212–220
(2018).
150. Alten, R. & Maleitzke, T. Tocilizumab: a novel
humanized anti-​interleukin 6 (IL-6) receptor antibody
for the treatment of patients with non-​RA systemic,
inflammatory rheumatic diseases. Ann. Med. 45,
357–363 (2013).

151. Illei, G. G. et al. Tocilizumab in systemic lupus
erythematosus: data on safety, preliminary efficacy,
and impact on circulating plasma cells from an open-​
label phase I dosage-​escalation study. Arthritis
Rheum. 62, 542–552 (2010).
152. Jamilloux, Y. et al. JAK inhibitors for the treatment of
autoimmune and inflammatory diseases. Autoimmun.
Rev. 18, 102390 (2019).
153. Chauhan, D., Vande Walle, L. & Lamkanfi, M.
Therapeutic modulation of inflammasome pathways.
Immunol. Rev. 297, 123–138 (2020).

Acknowledgements
This work was supported by the European Union Social Fund
TAMOP-4.2.4.A/2-11/1-2012-0001 ‘National Excellence
Program’ and the European Union GINOP-2.3.2-15-201600015 and GINOP-2.3.2-15-2016-00050 grants (to Z.S.). It
was also supported by the Hungarian National Scientific
Research Fund (NKFIH-​OTKA Grant No. K131844 to S.B.)

Nature Reviews | RheumAtoloGy

and the Faculty of Medicine of the University of Debrecen
(1G3DBKD0TUDF 247 to S.B.).

Author contributions
All authors contributed to all aspects of the article.

Competing interests
The authors declare no competing interests.

Peer review information
Nature Reviews Rheumatology thanks A. Doria, S. Savic and
the other, anonymous, reviewer(s) for their contribution to the
peer review of this work.

Publisher’s note
Springer Nature remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

© Springer Nature Limited 2021

volume 17 | October 2021 | 595
0123456789();:

nature reviews rheumatology

https://doi.org/10.1038/s41584-025-01291-0

Review article

Check for updates

Immune-cell profiling to guide
stratification and treatment of
patients with rheumatic diseases
Deepak A. Rao

Abstract

Sections

Methods for high-dimensional immune-cell profiling have advanced
dramatically in the past decade. Studies of tissue and blood samples
from patients with rheumatic diseases have revealed stereotyped
features of immune dysregulation in individual diseases and in subsets
of patients who share diagnosis of a heterogeneous disease. Translating
immunological patterns into clinically implementable, actionable
biomarkers has the potential to improve detection and quantification of
pathological immune activity and selection of appropriate treatments
for autoimmune rheumatic diseases. For example, cytometric features
can be used to distinguish the various forms of inflammatory arthritis,
stratify subsets of patients with rheumatoid arthritis or subsets of
patients with systemic lupus erythematosus and predict treatment
responses. Cellular immune profiling also enables the identification
of specific features of immune dysregulation in individuals with rare,
undiagnosed, inflammatory diseases. Several paths might lead to
translation of discoveries from broad immune profiling into clinical
tests to interrogate immune activation in people with rheumatic
diseases.

Introduction

Division of Rheumatology, Inflammation, Immunity, Brigham and Women’s Hospital and Harvard Medical School,
Boston, MA, USA.
e-mail: darao@bwh.harvard.edu

Nature Reviews Rheumatology

High-dimensional cellular
profiling
Distinct immune-cell features
across rheumatic diseases
Immune-cell heterogeneity
within a disease
Immune profiling of treatment
responses
Immune profiles of individuals
with undiagnosed disease
Translation into clinical
practice
Multi-modal immune profiling
Conclusion

Review article

Key points
• Profiling of immune cells in blood and tissue from patients with
rheumatic diseases has helped to define populations of activated
immune cells that are characteristically expanded in specific diseases,
highlighting both unique and shared features across diseases.
• Immune profiling of patients with SLE has identified specific axes
of immune dysregulation, including activation of type I IFN pathways,
proliferation of lymphocytes, expression of cytotoxic molecules on
T cells and upregulation of myeloid cell- and neutrophil-associated
signatures; these features vary across patients and help to delineate
subgroups of patients that differ in immune activity.
• Longitudinal evaluation of cellular profiles of patients receiving
treatments targeting rheumatic disease helps to associate
immunological features with treatment effects and predict response
to treatment.
• Incorporation of immune profiling into clinical evaluation of patients
with rheumatic diseases might enable improved patient stratification,
assessment of disease activity and prediction of treatment response.

Introduction
Individuals who seek rheumatological evaluation for a possible autoimmune rheumatic disease commonly describe that they feel ‘inflamed’
and often wonder whether their immune system is ‘overactive’. A major
challenge for the evaluating physician is to determine whether the
immune system has become pathologically activated, driving an
autoimmune or inflammatory condition, or whether symptoms are
caused by non-immune mechanisms. These decisions are impactful,
as they might dictate whether immunosuppressive therapies will be
used. Yet, currently available laboratory tests provide rheumatologists with a quite limited assessment of an individual’s immune status
at any given time.
Routine immunological tests include a complete blood count to
determine if the major blood-cell lineages (neutrophils, monocytes,
lymphocytes) are present at normal levels. Erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) tests provide measures
of systemic inflammation1. Total immunoglobulin levels determine
whether the immune system has made antibodies at normal levels,
and autoantibody tests demonstrate the presence of antibodies that
are, for example, specific to double-stranded DNA or antibodies to
citrullinated proteins. Tests of serum complement levels evaluate
whether the complement cascade has been activated, for example, by
antibody–antigen immune complexes2. However, this set of widely used
tests provides limited insight into the level of activity of the immune
system. In contrast to the range of functional tests to assess cardiac,
pulmonary, renal and hepatic function, we currently lack functional
tests to assess normal versus aberrant activity of the immune system.
Serum protein profiling, bulk RNA sequencing (RNA-seq) of whole
blood or specific cell populations and cytometric profiling — which is
the focus of this Review — have been used to interrogate the activity
of the immune system in a research setting. Over the past 10 years, the
advent of high-dimensional cellular-profiling technologies, including spectral flow cytometry, mass cytometry and single-cell RNA-seq
(scRNA-seq), have provided a powerful set of tools for analysing the
Nature Reviews Rheumatology

composition and functional states of immune cells in human blood
and tissues. Studies using these methods are providing a new set of
metrics for assessing normal and aberrant immune profiles, emphasizing abnormalities in the abundance or activation states of immune-cell
populations. These metrics have the potential to complement the current clinical evaluation of rheumatic diseases, including evaluation of
rheumatoid arthritis (RA) (Box 1).
This Review discusses insights from selected immune-profiling
studies that have highlighted specific features of immune-cell dysregulation across rheumatic diseases, across subgroups of patients with
the same rheumatic disease or in individuals with very unusual clinical
presentations. This discussion covers only a subset of the wide range
of immune-profiling studies that have been performed in past years,
with an emphasis on studies of RA and systemic lupus erythematosus
(SLE) that involve broad cytometric profiling approaches, including
approaches with single-cell resolution; studies using serum proteomics
or bulk transcriptomics are noted in some cases to provide context or
complement interpretation of cellular-profiling studies.

High-dimensional cellular profiling
Cytometric profiling with flow cytometry, mass cytometry or single-cell
transcriptomics enables the quantification of the various cell populations and activation states within a complex mixture of cells. Flow
cytometry methods have been well established for decades, using
antibodies tagged with fluorophores to quantify the expression
of cell-surface or intracellular proteins with single-cell resolution.
Advances in ‘spectral’ flow cytometry have dramatically improved the
ability to discriminate signals from different fluorophores, expanding
the number of proteins one can detect simultaneously to >30 (ref. 3).
Mass cytometry captures even more protein markers on individual cells
than flow cytometry, and it is based on a similar approach, although
using antibodies tagged not with fluorophores but with heavy metals4.
When attached to cells, heavy metals are quantified by a mass spectrometer at single-cell resolution, providing high-dimensional single-cell
analyses. scRNA-seq captures RNA from individual cells5, typically
using individual lipid droplets, yet the analysis is often conceptually
similar to other cytometry approaches, characterizing individual cell
identities based on expression of cellular markers. The addition of
oligo-DNA-tagged antibodies to the single-cell RNA-seq workflow enables the quantification of both transcription and cell-surface protein
expression6,7. Data from each of these analyses is often visualized in
two-dimensional UMAP plots, and clustering approaches have been
the go-to method of quantifying the abundance of the various cell
populations within a sample8,9. Several methods have been developed
to evaluate high-dimensional cellular profiling data to identify differences between two or more patient groups or to associate these data
with clinical or serological parameters10–12.

Distinct immune-cell features across
rheumatic diseases
Although standard laboratory tests currently used in clinical practice
provide limited insight into immune activation, they already indicate
some clear patterns of immune abnormalities across rheumatic diseases. The serum CRP level is characteristically elevated in giant cell
arteritis and polymyalgia rheumatica (PMR), reflecting an important
role for IL-6, which induces CRP expression, in these diseases13,14, yet CRP
levels are often normal during SLE flares15,16. By contrast, serum complement levels are reduced in active SLE but not in giant-cell arteritis
or PMR, reflecting the immune complex formation and complement

Review article

Box 1 | Features that reveal heterogeneity in patients with rheumatoid arthritis
Clinical heterogeneity

• Demographic characteristics
• Joint distribution
• Extra-articular disease manifestations
• Erosion extent150
• Imaging heterogeneity (in ultrasonography)151

Serological and transcriptomic heterogeneity

• Erythrocyte sedimentation rate and levels of C-reactive protein
• Seropositivity for autoantibodies such as anti-CCP and rheumatoid
factor
• Multi-analyte immunoassays
• Cytokine signatures
• Additional autoantibodies

Genetic heterogeneity

• HLA alleles152
• Non-HLA risk alleles153
• Polygenic risk scores154,155

Histological heterogeneity
• Krenn histological scores156
• Cellular density
• Immune-cell aggregates

cascade activation in SLE2. Early microarray studies highlighted a prominent upregulation of interferon (IFN)-stimulated genes in blood cells
of patients with SLE, far exceeding levels seen in patients with inflammatory arthritis17. This recognition of a prominent activation of a type I
IFN response in SLE fuelled the evaluation and ultimate approval of
type I IFN receptor (IFNAR) blockade with anifrolumab to treat SLE18.
SLE provides a benchmark for strong activation of a type I IFN response
against which other diseases can be compared19. Although type I IFN
signatures are not routinely measured, commercial tests that quantify
these signatures are becoming available20.
Moving beyond cytokine signatures, immune-cell profiling studies are now identifying some of the major axes of immune activation
that distinguish autoimmune rheumatic diseases or disease groups
with cellular resolution. Flow-cytometry profiling of blood cells from
almost 1,000 individuals, representing 11 autoimmune rheumatic
diseases, revealed characteristic patterns of activation in immune-cell
subsets across diseases. SLE and mixed connective tissue disease
(MCTD) showed shared patterns, as expected given their clinical and
serological similarities, whereas RA and spondyloarthropathies (SpA),
including psoriatic arthritis (PsA) and axial spondyloarthritis (axSpA),
shared a distinct set of immune features, perhaps reflecting the shared
responsiveness of RA and SpA to inhibitors of TNF21. SLE and MCTD
were associated with a particularly prominent expansion of activated
(HLA-DR+ CD38+) CD4+ and CD8+ T cells in circulation, consistent with
observations from other studies22–24. Additional studies have highlighted a T cell–B cell axis, involving expansion of both B cell-helper
T cells, such as T follicular helper (TFH) cells and T peripheral helper
(TPH) cells, as well as activated B cells, as a core immunological feature
of SLE25–31. This T cell–B cell axis stands out in blood immune profiles
of patients with SLE when these are compared with patients with other
Nature Reviews Rheumatology

• Pathotype: lympho-myeloid, diffuse-myeloid, or fibroid64
• Ectopic lymphoid structures

Cellular heterogeneity in the synovium39,113

• Cell-type abundance phenotypes: T cells + B cells (T + B);
T cells + myeloid cells (T + M); T cells + fibroblasts (T + F); myeloid cells
(M); fibroblasts (F); endothelial cells + fibroblasts + myeloid cells
(E + F + M)39
• T cell phenotypes: T peripheral helper and T follicular helper cells,
granzyme K+ T cells, granzyme B+ T cells, regulatory T cells39,42,50
• B cell infiltrates: age-associated B cells, plasma cells39
• Myeloid phenotypes: HBEGF+IL1B+, SLAMF7+, MERTK+
macrophages39,113,157, conventional type 2 dendritic cells, and
inflammatory dendritic type 3 cells114
• Fibroblast phenotypes: lining fibroblasts, sublining fibroblasts,
perivascular fibroblasts39,158,159

Cellular heterogeneity in the blood

• T cell phenotypes: T peripheral helper or T follicular helper cells,
effector-memory T cells that re-express the naive-cell marker
CD45RA, regulatory T cells, T helper 17 cells42,69
• B cell phenotypes: age-associated B cells, plasmablasts160
• Myeloid phenotypes: monocytes, dendritic cells161
• Pre-inflammatory mesenchymal cells108

rheumatic diseases; blood profiles of patients with RA do not show
the same extent of adaptive immune-cell dysregulation on average as
seen with patients with SLE25,32,33.
Combining cellular and transcriptomic profiling, the ImmunoNexUT consortium reported bulk RNA-seq transcriptomes of 28 sorted
immune-cell populations from the blood of 337 individuals with ten
rheumatic diseases34. With this broad approach, diseases segregated
into two major groups: immune profiles of SLE, MCTD, RA and systemic
sclerosis (SSc) segregated from those of the autoinflammatory conditions Behcet disease and adult-onset Still’s disease. Correspondingly,
IFN signatures were enriched in SLE and MCTD, as well as in some SSc,
idiopathic inflammatory myopathy and Sjögren disease, whereas IL-18
or IL-1 signatures were enriched in Behcet disease and adult-onset
Still’s disease34.

Rheumatoid arthritis versus spondyloarthritis
RA and PsA represent similar, but distinct, forms of inflammatory
arthritis, with distinguishable patterns of joint involvement, risk factors, demographics and genetics. Clinical trials have highlighted differences in the efficacy of various immunological therapies for these
conditions, with IL-17A blockade being more efficacious in SpA than
in RA, despite similar total levels of expression of IL-17A in synovial
samples35. By contrast, B cell depletion with rituximab is commonly
used to treat RA but has not shown clear efficacy in SpA36,37. These treatment response differences illustrate that immunological drivers differ
between these conditions; immune-cell profiling studies are now providing a clearer view of the cellular immunology that underlies these
therapeutic differences.
Immune cells in synovial tissue or synovial fluid have been evaluated by scRNA-seq and mass cytometry in both RA and PsA24,38–41.

Review article

TPH cells were markedly expanded in joints of patients with seropositive
RA but had comparatively lower abundance in patients with seronegative RA or SpA24,42,43. TPH cells are a subset of CD4 T helper cells specialized to provide help to B cells, much like TFH cells; TPH cell expansion in
seropositive RA is aligned with the roles of TPH cells in B cell recruitment
and stimulation44. By contrast, IL-17A+ CD8 T cells45,46 are enriched in
the joints of patients with PsA, and this finding is consistent with the
responsiveness of these patients to IL-17A blockade47,48. In addition,
profiling of T cells from synovial fluid of patients with axSpA showed
increased abundance of an integrin-expressing CD103+ CD49a+ CD8
T cell population that expressed both IL-17A and cytotoxic molecules49.
The distinct patterns of expanded TPH cells in seropositive RA versus
IL-17A+ CD8 T cells in SpA seem to align well with the differential efficacy
of B cell depletion versus IL-17A blockade in these two diseases. Nevertheless, other features, including accumulation of granzyme K (GZMK)+
CD8 T cells, granzyme B (GZMB)+ cytotoxic T cells, and regulatory
T (TReg) cells within joints, are shared between RA and SpA38,50. Further
studies are needed to associate lymphocyte features with differences
in synovial pathology between RA and PsA, including differences in the
patterns of vascular remodelling and immune–stromal interactions51–53.
Immune-cell profiling studies using blood samples have also indicated distinct circulating immune-cell patterns in RA and PsA. Consistent with results from synovial tissue and synovial fluid, TPH cells are
increased in the circulation of patients with seropositive RA, but not
seronegative RA or PsA42,43. Broad mass cytometry profiling comparing
peripheral blood mononuclear cells (PBMCs) from patients with RA
or PsA highlighted increased frequencies of terminally differentiated
(CD27− CD28−) effector CD8 T cells in seropositive RA but not in seronegative RA or PsA54. Interestingly, the blood-immune profiles of seronegative RA and PsA had no clear differences54,55. The large-scale effort
of the ‘Accelerating Medicines Partnership (AMP) on Autoimmune
and Immune-Mediated diseases’ (AMP-AIM) network, which includes
a comparison of blood and tissue immune profiles between RA and PsA,
will provide substantial power to define robust immunological differences distinguishing these diseases, also with spatial resolution within
tissues56. Although the presence or absence of autoantibodies provides
a foundational tool helping to distinguish clinically overlapping entities of seropositive RA, seronegative RA and SpA, one can imagine that
immunological assessment of TPH, TFH or TH17 cell pathways in patients
might help to further distinguish subsets of patients with undifferentiated arthritis or patients with seronegative RA to guide selection of an
RA- versus an SpA-aligned treatment framework.

Immune checkpoint inhibitor-induced arthritis
Cellular profiling of the active immune response in patient samples
has proven valuable in assessing a form of inflammatory arthritis that
has emerged with the advent of immunotherapies for the treatment
of cancers — immune checkpoint inhibitor (ICI)-induced arthritis.
ICI therapy using an antibody that blocks the inhibitory receptor
PD-1 induces a range of immune-related adverse events, including
ICI-induced inflammatory arthritis, which occurs in ~4% of treated
patients57,58. ICI-induced arthritis can involve RA-, PsA-, or PMR-like
manifestations, usually without generation of anti-cyclic citrullinated
peptide (anti-CCP) or rheumatoid factor autoantibodies58,59. Similar
to RA and PsA, ICI-induced arthritis involves an active, presumably
autoreactive, T cell response, yet the specific features of this response
differ starkly across the three conditions24,60. Mass cytometry-based
comparison of T cells from synovial fluid of patients with ICI-induced
arthritis, RA or PsA showed clear expansion of a population of CD38hi
Nature Reviews Rheumatology

CD8 T cells specifically in ICI-induced arthritis24. CD38hi CD8 T cells
were also expanded in the circulation of patients with ICI-induced
arthritis, and broadly among patients treated with ICIs, yet these cells
were not highly expanded in patients with RA or PsA24. Transcriptomic
comparisons of synovial fluid T cells demonstrated a higher type I IFN
response signature in T cells from patients with ICI-induced arthritis than synovial T cells from patients with RA or PsA, and in vitro
treatment of synovial fluid CD8 T cells from patients with RA or PsA
with type I IFN promoted acquisition of the CD38hi T cell phenotype
seen in patients with ICI-induced arthritis. The type I IFN signature in
ICI-induced arthritis samples provided an unexpected immunological
link between ICI-induced arthritis and SLE, a disease marked by high
type I IFN production that also features expanded CD38hi CD8 T cells19,23.
Defining such immunological benchmarks across diseases is likely
to provide a deeper understanding of why certain therapies work well
in one condition versus another and might help to identify therapies
that are likely to work in newly emerging conditions, such as those
induced by immunotherapies. Cross-disease comparisons integrating data across different forms of inflammatory arthritis, including
juvenile idiopathic arthritis and others61, should demonstrate the relative prominence of specific features of the active immune response in
inflamed joints, including the abundance of proliferating, exhausted
or stem cell-like lymphocytes, the expansion of TPH and TFH cells, the
presence of GZMK+ T cells versus GZMB+ T cells, and the frequencies of
TReg cells, infiltrating monocytes versus tissue-resident macrophages,
dendritic-cell (DC) populations, age-associated B cells (ABCs) and
plasmablasts, to amass a clearer taxonomy of inflammatory arthritides
according to features of immune activation62.

Immune-cell heterogeneity within a disease
Immune-cell heterogeneity in rheumatoid arthritis

In addition to highlighting differences across diseases, immune-cell
profiling is a valuable tool for dissecting immunological heterogeneity among patients who share a diagnosis. Patients with RA display
substantial variability in clinical course, the likelihood of developing erosions and response to treatments. Correspondingly, studies
of synovial tissues have highlighted differences in synovial immune
infiltrates among patients with RA, even when they share comparable
imaging and clinical features of synovitis63,64. Patients with seropositive
RA frequently show a ‘lympho-myeloid’ pattern of immune infiltration in the inflamed synovium, with aggregates of synovial B cells and
T cells that range from loose, disorganized clusters to well-organized
follicular structures65. In other patients with RA, the synovium either
shows a diffuse myeloid-cell infiltrate without lymphoid follicles or a
‘fibroid’ or ‘pauci-immune’ synovial pattern with few immune-cell infiltrates. Patients with a lympho-myeloid pathotype are the most likely to
develop erosions and joint damage progression, whereas patients with
a fibroid pathotype show the lowest disease activity, yet also the weakest response to DMARD treatment66. Detailed cellular analyses have
defined the composition of immune cells in synovial-tissue samples
across the various pathotypes. scRNA-seq of RA synovial biopsies delineated six ‘cell-type abundance phenotypes’ (CTAPs), representing six
types of synovial inflammation, that differ in the relative abundance of
each of the following cell types: fibroblasts; T cells and NK cells; B cells;
endothelial cells; and myeloid cells39 (Box 1). These CTAPs roughly
correspond to histological patterns, with the CTAP containing both
T cells and B cells (CTAP-TB) showing the highest histological scores of
synovitis (according to the histological score developed by Krenn) and
aggregate density. However, immunological information captured by

Review article

CTAPs largely seems to be orthogonal to clinical or serological assessments, suggesting that these tissue analyses will be complementary,
and not redundant, with current clinical assessment of RA.
Given the difficulties of sampling synovial tissue from patients
with RA, there has been substantial interest in identifying signatures
in blood that capture immune activity in the joints. Direct parallels
between synovial infiltrates and immune-cell phenotypes in blood
are challenging to identify, although some shared features of the
adaptive immune response have been demonstrated, such as shared
T cell receptors (TCRs) and, occasionally, shared T cell clone phenotypes in synovium and blood of patients with RA, PsA and ICI-induced
arthritis24,50,67,68. Analyses of paired blood and tissue samples from large
numbers of patients, such as those profiled in the AMP RA/SLE Network,
should help to clarify the extent to which features of immune cells
in blood can reflect specific immune processes occurring within
synovium.
Independently of synovial analyses, flow cytometry profiling of
blood cells from over 500 patients with RA has highlighted substantial variability in immune-cell profiles that were non-redundant with
clinical and serological phenotypes69. These blood immune profiles
were used to stratify patients into peripheral blood-cell abundance
phenotypes (PCAPs, analogous to synovial CTAPs). Patients with distinct PCAPs showed distinct patterns of cell abnormalities, including one group of patients with expanded activated CD4 T cells, CD8
T cells and plasmablasts (PCAP-TB), a separate group with increased
effector-memory T cells that re-express the naive-cell marker CD45RA
(TEMRA cells) or TEMRA and TH1 cells (PCAP-T1/T1T4), and two more patient
subgroups (PCAP-LD and PCAP-SD) that cytometrically resembled
healthy individuals69. The frequency of anti-CCP antibody or rheumatoid factor did not differ across these groups, yet patients in the
PCAP-TB group showed the highest disease activity and ESR, as well
as the least frequent use of methotrexate. Inclusion of additional
immune-cell subsets with an emerging role in disease pathogenesis,
including TPH cells, THA cells — a CXCR3mid cytotoxic CD4 T cell population expanded with age70 — and GZMK+ T cells50, might enhance the
utility of blood-cell profiling in RA. In addition, the identification and
quantification of immune-cell subsets are aided by high-resolution
scRNA-seq and mass cytometry analyses that precisely define the
phenotypes of activated cells in the circulation71,72. In total, cellular
profiling of blood and tissue samples from patients with RA is providing an additional set of informative variables with which to understand
immunopathology in individual patients (Box 1).

Immune-cell heterogeneity in systemic lupus erythematosus
Patients with SLE display stark variability in terms of organs affected,
disease severity and response to immunosuppressive therapy, potentially reflecting substantial immunological heterogeneity. Serum proteomics, gene-expression profiling and flow-cytometry analyses have
illustrated key features of immune activation in SLE that are consistently observed across cohorts. Expression of IFN-stimulated genes has
reproducibly been found to be increased across many SLE studies, with
the majority of patients showing a type I IFN signature17,19. scRNA-seq
profiling has further refined immune-cell populations with the highest expression of an IFN response signature in the blood, including
monocyte and lymphocyte subsets73,74, and has demonstrated a clear
IFN response signature across many tissues, including skin and kidney,
in SLE75–78. In both kidney and skin samples, a subset of T cells and B cells
shows a very high IFN signature, above the basally elevated IFN signature seen broadly in cells from patients with SLE compared with healthy
Nature Reviews Rheumatology

individuals75–77. What distinguishes the IFN signature-high cells from
other cells in the tissue remains unclear. It will be interesting to integrate these observations with emerging spatial transcriptomics data,
which suggest that cells with the highest IFN signatures are enriched
in the glomeruli in the kidneys of patients with lupus nephritis (LN)79.
In addition to the IFN signature, other immune signatures
extracted from whole-blood transcriptomic analyses have enabled
patient stratification into subgroups, particularly when analyses were
run on longitudinal samples. Longitudinal whole-blood profiling of
patients with childhood-onset SLE stratified patients into seven groups
that vary in transcriptomic signatures associated with erythropoiesis,
IFN response, myeloid cells and neutrophils, plasmablasts and lymphocytes. Among these patient subgroups, a plasmablast-associated
signature was strongly associated with disease activity over time80.
Studies using blood-transcriptomic profiling of adult patients with
SLE have stratified patients into 3–7 subgroups based on similar but
not identical features to those used for the stratification of paediatric
patients81–83. In the adult cohorts, increased expression of inflammation, myeloid/neutrophil and plasmablast transcriptomic signatures
have been associated with increased disease activity, as defined based
on SLE Disease Activity Index scores81–83.
The cellular resolution of cytometric profiling studies has
in some cases extended understanding of the immune pathways
previously implicated by bulk RNA-seq in SLE, for example, the
plasmablast-associated signature. Cytometric studies evaluating B cell
phenotypes in SLE have extended the understanding of the activated
B cell response, which includes expansion of both plasmablasts and
ABCs (also known as DN2 B cells), which are characterized by high
expression of CD11c and TBET and low CXCR5 and CD21 (refs. 29–31,84).
The expansion of CD21low ABCs is perhaps the most prominent cytometric abnormality among circulating B cells in patients with SLE
and is highest in patients with active disease, including patients
with LN25,29,30,71,85.
Cytometric profiling can also capture immunological features
that are difficult to detect in whole-blood-transcriptomic analyses, for
example, the abundance of specific T cell subsets or T cell functional
states. Flow cytometry-based profiling of PBMCs stratified patients
with SLE into three subgroups based on T cell profiles, with one group
marked by expanded TFH cells and activated TH1 cells (that probably
included TPH cells) and a second group marked by expanded TReg cells86.
Disease activity or duration did not differ across the three groups, yet
the TFH cell-associated group had the highest total immunoglobulin levels, consistent with amplified T cell–B cell interactions. Mass
cytometry-based profiling of T cells from patients with SLE highlighted
a prominent expansion of TPH cells in patients with LN, with the expansion of circulating TPH cells exceeding that of TFH cells25. Both TPH cells
and TFH cells have been identified as expanded in multiple cohorts of
patients with SLE and associated with the clinical and serological measures of disease activity26–28,71,87,88. TFH cell expansion seems to be clearer
among patients with shorter disease duration compared with those
with longer disease duration87. The abundance of TPH cells correlates
positively with that of ABCs in patients with SLE, probably reflecting
an ongoing extrafollicular response25,89,90.
The AMP RA/SLE network used mass cytometry of PBMCs to stratify patients with LN into three immunologically distinct subgroups71.
Among patients with biopsy-demonstrated class III, VI or V nephritis,
more than half of whom had established disease with prior treatment
for LN and prior kidney biopsies, cytometric profiling identified one
subgroup that was immunologically indistinguishable from healthy

Review article

individuals, a second subgroup had a very high type I IFN response
signature, and a third subgroup had an intermediate type I IFN
response signature but a distinctive expansion of GZMB+ T cells, suggesting activation of a ‘cytotoxic lymphocyte’ axis. Both the type I IFN
response-high and GZMB+ subgroups had expansion of proliferating
B cells and TPH cells, indicating a shared activation of a B cell–T cell axis.
These patient subgroups had distinct features in terms of both kidney
histopathology and clinical course; the GZMB+ subgroup had patients
with the highest disease activity in the kidney based on the histological
NIH activity index and the highest likelihood of a good renal response
to standard-of-care therapy at 1 year71. By contrast, the immunologically quiet subgroup showed the highest degree of chronic kidney
damage histologically, which perhaps reflects prior immunological
injury. The poor response to treatment in this subgroup suggests that
these patients might have chronic kidney disease without ongoing
immune activation and might not benefit from escalated immunosuppressive therapy. Notably, kidney biopsies shared the specific
features of blood-immune profiles; patients with a high proportion
of GZMB+ T cells in blood also had an increased proportion of GZMB+
CD8 T cells in kidney tissue, and patients with the highest type I IFN
signatures in blood also showed the highest IFN signatures in cells
from the kidney71. Further validation of these signatures and additional
prospective studies are needed to determine if a very high type I IFN
response signature enriches for patients most likely to respond to IFN
blockade, or if cytotoxic T cell activation is differentially susceptible
to the various SLE therapies.
Given the observations from studies on RA and SLE discussed
above, cytometric immune profiling has the potential to identify immunologically distinct subgroups of patients in other rheumatic diseases
as well. In PsA, blood-cell profiling by flow cytometry highlighted four
subgroups of patients through principal components analysis, with a
subgroup that was characterized by increased frequencies of TH17 cells,
memory TReg cells, DCs and monocytes being associated with increased
disease duration and activity91. Moreover, scRNA-seq analysis of blood
segregated patients with Sjögren disease into two major subgroups,
corresponding to the presence or absence of anti-SSA antibodies; a
strong type I IFN signature was associated with anti-SSA seropositivity92.
Integrating large datasets, especially scRNA-seq datasets, across
diseases might provide the ability to identify immunologically
similar patients across clinical-disease presentations.

Immune profiling of treatment responses
In both RA and SLE, the expanded armamentarium of immunosuppressive drugs poses new challenges for patients and physicians in selecting
which therapy is most likely to be beneficial for an individual patient. In
RA, at least five mechanistically distinct classes of biologic therapies are
available: TNF blockade; IL-6 blockade; JAK inhibition; B cell depletion;
and T cell costimulation blockade93. However, there is little guidance
on the decision about which therapy to use for an individual patient.
In SLE, the expanded range of treatment options, now including B cell
inhibition or depletion94, IFNAR blockade18 and calcineurin inhibition95,
similarly poses questions about which drug to use for which patient.
Longitudinal studies of pre- and post-treatment samples provide crucial
insights into the major pathways affected by each DMARD and potentially identify cellular features at baseline that are associated with a
good response to treatment. This review will not attempt to broadly
summarize the wide range of cellular treatment response biomarker
studies in RA and SLE, but will, rather, highlight specific examples of
promising approaches or consistently observed signals.
Nature Reviews Rheumatology

Profiling treatment responses in rheumatoid arthritis

Blood-cell-based profiling of treatment responses. Identifying
predictors of patient responses to DMARDs remains an area of active
research in RA. Tremendous effort has been focused on identifying
biomarkers of response to TNF inhibitors, but analyses of standard
laboratory markers, antibodies, serum proteins, whole-blood transcriptomes and cell phenotypes have not yet led to the identification of
any robust predictors of treatment responses96,97. The advent of broad
profiling methodologies has yielded some successes: whole-bloodtranscriptomic analyses combined with advanced computational
approaches have led to the commercial development of a test to predict
the likelihood of a non-response to TNF inhibitor therapy98,99.
The search for treatment response biomarkers has been importantly advanced by applying immune-profiling studies within the
context of clinical trials, especially those involving randomization.
This approach leverages the standardized clinical assessment of
disease activity within a trial infrastructure, and the randomization
minimizes concerns about confounding by indication and unaccounted bias. Such studies have highlighted a reproducible relationship between the frequency of circulating TFH cells in the blood and
response to abatacept, a drug that blocks T cell costimulation. In a
cohort of patients analysed by flow cytometry in the NORD-STAR
trial, which randomized patients with early RA to methotrexate plus
one of four biologics, cytometric quantification of 12 T cell populations demonstrated a specific association between elevated baseline
PD1+ TFH cell frequency and achieving remission following treatment
with abatacept100. Similarly, in a prospective observational study
of patients with RA and an inadequate response to methotrexate,
patients who achieved remission after treatment with abatacept
had higher frequencies of PD1+ TFH cells in the blood at baseline than
patients who did not achieve remission101. Consistently, elevated
frequencies of activated TFH and TFH cells in the blood of patients with
early type 1 diabetes were associated with a good clinical response to
abatacept102. Abatacept robustly reduces the frequency of circulating
TFH and TPH cells, supporting the biologic plausibility of the cellular
association with treatment response101,102.
Studies looking for cytometric features predictive of response
to rituximab have highlighted an association with circulating B cell
populations. In the SMART trial of rituximab in patients with RA, flow
cytometric analysis of B cells indicated that a low proportion of circulating CD27+ memory B cells was associated with a good response to treatment at 24 weeks by EULAR criteria103. Independently, the FIRST study,
which evaluated 154 patients with RA who were treated with rituximab
using flow cytometry, associated a high proportion of circulating
CD27− IgD− B cells with a good response to rituximab, especially when
considered in combination with rheumatoid factor positivity. Combined with additional studies104,105, these observations strongly associate features of B cell activation or B cell memory with likelihood of
response to rituximab. Irrespective of B cell phenotype, a randomized
study of 25 patients with RA associated the detection of residual circulating B cells after two doses of rituximab with significantly improved
response rates to a third dose of rituximab106.
Applying standardized profiling methods across patients treated
with various DMARDs has the potential to identify specific cellular
patterns that are associated with an improved response to a specific
treatment. Exploratory studies using flow-cytometry profiling, applied
longitudinally to over 500 patients with RA as described above, identified subgroups of patients with a differing likelihood of response
to the various DMARD classes69. Using PCAPs to stratify patients as

Review article

described above and prospective longitudinal evaluation indicated
that patients in the PCAP-TB group, who are marked by an active B cell
response, were the least likely to achieve remission overall after treatment with one of four biologic DMARDs interrogated: abatacept, JAK
inhibitors, TNF inhibitors and IL-6 inhibitors. By contrast, patients in
the PCAP-SD or PCAP-LD groups, who collectively showed relatively
few cellular changes compared with healthy individuals, were more
likely to achieve remission following treatment with JAK inhibitors
than were patients in other PCAPs.
To define operational links between PCAPs and treatment
assignments, the authors then assigned each of the four specific
DMARDs as associated or not associated with a good response for each
PCAP-based patient subgroup. Patients were then assigned a status of
‘expected’ or ‘non-expected’, reflecting whether the patient received
a DMARD associated with a good response in their identified PCAP.
Promisingly, in a validation cohort of 183 patients, patients with an
‘expected’ designation, indicating that the patient received a DMARD
expected to produce a good response in their identified PCAP, were
more likely to achieve remission than patients with a ‘non-expected’
treatment assignment (33% versus 18%). These treatment-response
associations need to be validated further, and a substantial challenge
remains to identify stratification parameters that can be reproduced
and adopted widely. Nonetheless, the impressive scale of the study and
the ability to reproduce signatures in a validation cohort provide hope
for extension of this approach. A broader immune profiling approach
that captures activated TFH and TPH cells or other cell populations
with a key role in RA might improve treatment assignment to specific
patient subgroups.
Using a clever, alternative strategy, the BioRRA study investigated
how circulating immune-cell profiles change during arthritic flares
that occur in patients with RA after withdrawal of synthetic DMARD
treatment107. The analyses associated expansion of activated T cell and
B cell populations, including PD1+ CD38+ CD8 T cells and PD1+ ICOS+
CD38+ CD4 T cells, with disease flares after DMARD withdrawal. This
finding suggests that synthetic DMARDs hold these T cell populations in check, such that treatment withdrawal allows for PD1+ CD38+
CD8 T cell and PD1+ ICOS+ CD38+ CD4 T cell activation and expansion.
Frequencies of these cell populations at baseline (pre-drug withdrawal) did not differ between patients who remained in remission
and patients who experienced disease flares after drug withdrawal;
thus, it is unclear whether such signals can help predict disease relapse
prior to drug withdrawal. Nonetheless, the above cellular correlates
might provide a valuable readout to confirm re-activation of the
disease-associated immune response, if symptoms emerge following
treatment cessation. Immune-cell profiles that are potentially associated with disease flares were also identified by a separate study using
frequent, serial assessment of whole-blood samples by RNA-seq; in this
study, disease flares were associated with preceding changes in B cell
signatures and a concurrent increase in rare, circulating mesenchymal cells potentially related to synovial fibroblasts, called PRIME cells,
during the flare108.
Synovial cell-based profiling of treatment responses. There is major
interest in understanding the associations between immunological
features in synovium and response to the various treatments. Results
from the pioneering R4RA trial provided encouraging initial observations, indicating that patients with a diffuse myeloid infiltrate were
more likely to respond to tocilizumab than to rituximab109. Extending these observations using CTAP designations further supported
Nature Reviews Rheumatology

the idea that patients with a fibroid (CTAP-F) phenotype, generally
lacking large lymphocyte or myeloid infiltrates, were the least likely
to respond to biologic treatment39,110. One tangible prediction in
connecting synovial infiltrates to treatment response would be that
patients with a B cell-enriched synovium are more likely to respond
to B cell depletion with rituximab than patients without B cells in
synovium; however, this has not been consistently observed in the
clinical trials that have assessed synovium109,111,112. Among synovial
myeloid cells, an increased proportion of MerTK+ tissue macrophages
is associated with a state of treatment-induced remission in RA, and
an increased proportion of MerTK+ tissue macrophages in synovium
at baseline is associated with maintenance of remission after TNF
inhibitor withdrawal113. Spatial transcriptomic analyses have further
associated synovial DC populations with disease activity and treatment response, reporting on a tolerogenic AXL+ cDC2 population
that is present in healthy synovium but absent in RA synovium, even
when remission is achieved, suggesting a lasting remodelling of the
DC populations due to synovitis114.
With these early observations guiding new study design and analysis approaches, there remains substantial enthusiasm that cellular
features within synovium will provide crucial insights into the variable
treatment responses of patients with RA. Detailed single-cell analyses,
including spatial transcriptomic analyses, comparing pre-treatment
and post-treatment samples, as reported in inflammatory bowel disease, should aid in identifying specific cell populations associated
with response and non-response to treatment115. Indicatively, spatial
transcriptomic analyses of longitudinal synovial tissue biopsies from
patients with RA demonstrated a COMPhi fibrogenic fibroblast population that is enriched in pre-treatment samples of patients who do
not achieve remission and that persists in synovium despite effective
reduction of immune cells by DMARD therapy56.

Profiling treatment responses in systemic
lupus erythematosus
Cellular or molecular signals that are associated with treatment effects
and treatment responses have been identified in several clinical trials in SLE. Correlative transcriptomic and serum-profiling studies of
patients treated with anifrolumab have illustrated a clear reduction in
IFN responses at both transcriptomic and proteomic levels in treated
patients in both clinical trials and observational studies116–118. Profiling
of blood samples from patients with SLE before and after treatment
with anifrolumab in the MUSE trial demonstrated that anifrolumab
alters several measures of immune activation in SLE, with a high IFN signature at baseline; anifrolumab treatment increased numbers of circulating neutrophils, platelets and lymphocytes, especially naïve CD4 and
CD8 T cells118. Anifrolumab treatment also reduces circulating levels of
several chemokines, including CXCL13, a potent B cell chemoattractant
produced by TPH and TFH cells118,119. Longitudinal scRNA-seq profiling of
blood samples from a small cohort of patients that received anifrolumab demonstrated that IFNAR blockade reduces the abundance of
circulating TPH cells, and concurrently expands a counter-regulated
population of IL-22-producing CD4 T cells (TH22 cells), which are
reduced in patients with active SLE119. This reduction in circulating
TPH cells following type I IFN blockade functionally links IFN signalling
to enhanced T cell–B cell interactions and B cell activation in SLE119.
Understanding the effects of type I IFN blockade on other components
of the pathological adaptive immune response in SLE is of major interest. Thus far, it has not been evident from available data that patients
with a low IFN signature have a substantially weaker clinical response to

Review article

Table 1 | Immune profiling of response to treatment
Type of treatment

Engagement of primary
target

Pre-treatment vs
post-treatment
comparisons

Predictors of treatment
response

Assessment of treatment duration

What cell populations
or pathways are most
altered by treatment?

What cellular features at
baseline (pre-treatment)
predict a good response
to treatment?

For how long can therapy continue to
suppress signs or markers of immune
dysregulation?

Synthetic DMARDS

Unclear

Biologic DMARDs

Inhibition of targeted pathway
(for example, TNF, IL-6, IFN)

CAR T cells or depleting
antibodies

Depletion of targeted cell
population (for example,
B cells, plasma cells, PD1hi
T cells)

BiTEs

Depletion of targeted cell
population
Extent and nature of T cell
activation

For how long does cell depletion last?
For how long do signs or markers of
immune dysregulation remain absent after
a single dose of the respective treatment?

BiTEs, bispecific T cell engagers; CAR T cells, chimeric antigen receptor T cells; IFN, interferon; PD1, programmed cell death protein 1.

anifrolumab than those with a prominent IFN signature at baseline120;
further immunological assessments might help to dissect whether
specific features of the IFN response, such as very prominent and distinctive IFN activation71 or expansion of IFN-associated immune-cell
populations, predict a better response to anifrolumab.
Longitudinal profiling of blood samples from patients treated
with B cell-directed therapies have also identified cellular correlates of treatment effect and response. Treatment with belimumab,
an FDA-approved agent for SLE that blocks B cell activating factor
(BAFF), reduced whole-blood-transcriptomic signatures associated
with B cells, as well as signatures associated with IFN and IL-6 signalling
and neutrophils, especially in responders121. Similarly, patients treated
with tabalumab, an IgG4 antibody that blocks BAFF, also demonstrated
a reduction in B cell-associated transcripts in whole-blood transcriptomics, consistent with a reduction in circulating B cell counts122. Transcriptomic analyses of sorted leukocyte subsets from blood collected
before and after treatment with belimumab demonstrated clear effects
of belimumab on the transcriptomic features of B cell subsets, with few
effects on transcriptomes of circulating T cell or myeloid cell subsets,
consistent with the direct effects of belimumab on B cell activation123.
Further, the number of differentially expressed genes, comparing
pre-treatment and post-treatment B cell subset transcriptomes, was
higher in good responders to belimumab treatment than in poor
responders. A separate study reported reductions in both CD19+ B cells
and activated PD1+ T cells after treatment with belimumab124. Interestingly, a longitudinal flow-cytometry assessment of T cell subsets
from the blood of patients treated with belimumab demonstrated an
increase in the ratio of TReg–TH17 cells following treatment, an effect that
was reproduced in an independent, broader mass-cytometry profiling
study125,126. These observations associate specific immune alterations
with BAFF blockade, with both direct effects on B cells and secondary
effects on T cells.
The use of molecular profiling in studies evaluating new therapeutic agents might also facilitate the identification of molecular
predictors of treatment response in SLE. In a phase II trial of obexelimab, a bifunctional antibody that binds CD19 and the inhibitory
receptor FcγRIIB, given after initial high-dose steroid treatment,
whole-blood transcriptomics were used to classify patients into
subgroups: patients with increased expression of lymphocyte modules and cell-proliferation modules but without high expression
of inflammation-associated modules were more likely to respond
Nature Reviews Rheumatology

to obexelimab than patients from other subgroups, as assessed by
maintenance of disease improvement127. In a phase IIb study with
iberdomide, a cereblon ligand that promotes degradation of the B cell
transcription factors Ikaros and Aiolos, which are important for lymphocyte development and function and which both have polymorphisms associated with SLE128, blood-cell profiling demonstrated that
treatment resulted in dose-dependent decreases in the number of
circulating B cells and memory B cells, as well as in plasmacytoid DCs
and myeloid DCs20. Concurrently, the number of TReg cells increased
in a dose-dependent fashion, paralleling an increase in circulating IL-2
levels. Transcriptomic analyses also highlighted clear reductions in
IFN response signature with treatment, and patients with the highest IFN response signature at baseline were the mostly likely to have
reduced disease activity after treatment, as assessed by SLE responder
index 4 (SRI4)20.
Following these examples, it is of substantial interest to define the
effects of the commonly used synthetic DMARDs, such as azathioprine
and mycophenolate, given their widespread use and their difficult-topredict effects on cellular immunology. The effects of these drugs
have not yet been revisited in detail using high-dimensional cellular
profiling approaches. Looking forward, understanding the broad
scope of immunological changes induced by cell-depletion strategies, such as CD19 CAR T cells and bispecific T cell engagers (BiTEs),
will be crucial129–131. Deep B cell depletion with these methods has the
potential to correct multiple immune abnormalities in patients with
SLE, including normalization of complement levels and reduction in
type I IFN response129,132, but the extent to which B cell depletion also
corrects T cell and myeloid-cell abnormalities in patients with SLE
remains to be defined. The extent and nature of CD8 and CD4 T cell
activation induced by BiTEs that target B cells can also be assessed
using broad immune-profiling approaches (Table 1).

Immune profiles of individuals with
undiagnosed disease
Cellular profiling studies typically utilize a grouped comparison analysis strategy, comparing patients with healthy individuals, pre-treatment
with post-treatment, or responders with non-responders. However, cellular profiling also has the potential to identify individual patients with
very abnormal features of immune activation compared with a reference population. A pilot study evaluating this approach was performed
on samples from 16 patients seen in the Undiagnosed Diseases Network

Review article

programme, an NIH-funded programme that focuses on patients with
very rare or unusual disease presentations133,134. These 16 patients,
who all have unusual disease presentations thought to be potentially
immune associated, underwent whole-exome or -genome sequencing
that did not reveal a clear monogenic cause of disease. This cohort
therefore underwent mass cytometry immune profiling of blood cells,
and immune profiles were assessed against ~140 reference datasets
that included healthy individuals, patients with RA and patients with
SLE134. Immune profiles from 5 of the 16 patients from the Undiagnosed
Diseases Network programme were identified as ‘outliers’ based on the
presence of at least one immune-cell population that was extremely
expanded compared with the overall cohort, but no outliers were identified among the reference datasets. Of these patients, one had a dramatic
expansion of CD25hi TReg cells, which comprised 50% of the circulating
CD4 T cells, one was identified as having B cell leukaemia, one had an
aberrant expansion of a gamma delta T cell population, and one had a
very abnormal myeloid-cell phenotype. This exploratory work suggests
that immune profiling can be used to identify specific immunological
abnormalities in individuals with rare or unusual disease presentations
and enable individualized treatment strategies.
Such an immune profiling approach can complement interrogation for rare monogenic causes of immune-mediated disease using
whole-genome or whole-exome sequencing135 or bulk RNA-seq and
scRNA-seq analyses reporting outlier gene- or splice-variant expression profiles136,137. In both cases, immune profiling has the potential
to delineate pathways of immune activation that are activated in the
context of a monogenic disease and potentially relevant for treatment.
In addition, deep analyses using scRNA-seq might be able to identify
immunological abnormalities missed by the cytometric approach, as
scRNA-seq captures cytokine response signatures more readily than
protein cytometry. As scRNA-seq analyses of PBMCs from healthy
donors, individuals at risk of disease (for example, individuals at risk
of RA138,139) or individuals with defined diseases, including SLE73, RA,
Sjögren disease92, SSc140 and others, are becoming increasingly available, they enable mapping of any individual scRNA-seq profile to these
reference datasets to identify aberrant cell populations, phenotypes,
or states in an individual.

Translation into clinical practice
Cellular profiling studies have yielded several robust features of
immune activation or dysregulation that capture clinically relevant
information. Assessment of such features in clinical practice might be
Normal values in
standard laboratory tests

• Complement C3 and C4
• Anti-dsDNA antibodies
• Urine protein-to-creatinine ratio
• Complete blood count

Patient A

Patient B

Cytometric assessment of pathological immune activation
A straightforward path to clinical implementation might involve distilling down the most informative features from high-dimensional profiling
studies and then developing targeted, cost-effective tests for these features. In SLE, a disease with prominent immune abnormalities in blood,
several informative features from transcriptomic and cytometric studies
can be captured in straightforward ways. An IFN signature can be captured by flow cytometric screening for the cell surface marker Siglec-1,
a protein strongly upregulated on monocytes by type I IFN141,142. The additional four features (proliferating lymphocytes, cytotoxic T cells, CD21low
B cells, low-density neutrophils) that stratified subgroups of patients
with LN as discussed above could be distilled down to simple parameters
that can be measured by flow cytometry71. Similarly, the major defining features of synovial CTAPs in RA can be captured by standard flow
cytometry39. Although, thus far, flow cytometry has little regular use
in patients with rheumatic diseases, save for quantifying CD19+ B cells
in patients treated with anti-CD20 antibodies, this method is routinely
used in oncology to aid in the search for haematological malignancies143.
Further, in clinical immunology, flow cytometry is used routinely to
quantify lymphocyte subsets in patients with suspected immunodeficiencies, and also to detect features of immune dysregulation in these
diseases, such as expansions of activated B cells and T cells144.
The development of a flow cytometric test to quantify TFH cell
frequency in children with immune dysregulation provides a valuable
example of how these tests can be implemented clinically for evaluation
of immune activity145. Building on established flow-cytometry protocols, an assay to quantify PD1+ CXCR5+ TFH cells was developed with
robust reproducibility across instruments and sample storage times of
up to 24 h. Interrogation of cohorts of healthy individuals and individuals with relevant diseases using this assay defined normal ranges and
indicated a sensitivity of 88% and specificity of 94% in discriminating
autoimmune disease from autoinflammatory conditions145. With an
estimated cost of <$200 per test, this approach provides a practical,
feasible strategy for detection of features of immune activation that
are currently missed by routine tests such as ESR, CRP, complement
Visualization of SLE-associated
axes of immune activation

Diverse immune
activation profiles
Patient A Patient B Patient C
Cytotoxicity
IFN signature
Proliferating B and T cells
Myeloid cells

Cytotoxicity

Assessment of ANA+ women with
arthralgias, fatigue and rashes

complementary to and non-redundant with serological tests. Translation of findings from cellular profiling studies into clinical practice
could follow multiple paths, but two paths will be considered here: the
implementation of flow cytometry-based assessment of pathological
immune activation and the introduction of scRNA-seq analysis in a
clinical setting.

Myeloid
cells

Patient C

Proliferating
B and T cells

Patient A

Patient B
IFN signature

Patient C

Fig. 1 | Focused immune assessments to identify immune dysregulation
in patients with suspected systemic lupus erythematosus. Evaluation of
individuals who are seropositive for anti-nuclear antibodies (ANAs) for possible
systemic lupus erythematosus (SLE) with standard laboratory tests, followed by
immune profiling for specific SLE-associated features of immune dysregulation.

Nature Reviews Rheumatology

In individuals with normal results of standard laboratory tests, immune profiling
might reveal SLE-associated immune activity to aid in the diagnosis of SLE.
Individual patient profiles can be visualized on axes of immune dysregulation71.
C3, complement C3; C4, complement C4; dsDNA, double-stranded DNA.

Review article

Single-cell RNA sequencing as a clinical tool

factors and autoantibodies. Although broad adoption of such tests
will require standardization of cytometry markers and analysis methods across laboratories, potential value seems clear. For example,
for individuals who present at rheumatology clinics with a positive
anti-nuclear antibody (ANA), arthralgias, rashes and fatigue but have
otherwise normal laboratory tests, a flow-cytometric quantification
of circulating TFH cells, TPH cells, ABCs and plasmablasts might help to
distinguish between SLE-associated pathological immune activity and
immunological quiescence (Fig. 1).

Broad profiling approaches such as scRNA-seq of blood samples are
likely to be translated into clinical practice in the next decade, following
the example of whole-genome sequencing. The rapid advancement of
clinical genome sequencing was aided by technological advances that
made DNA sequencing feasible at reasonable costs, as well as by the
establishment of a reference genome. In this context, immune profiling
has struggled with myriad variations in cytometric definitions for the
quantification of cell populations, complicating comparison of results

a
Patients with rheumatic diseases

Healthy individuals

Rheumatic disease immune profile
Healthy immune profile
Dendritic cells

Resting T cells

B cells
Dendritic cells
Monocytes

Resting
T cells

Disease-associated
cellular features

Monocytes
Activated T cells
B cells
Activated
T cells

B cells

RA
SLE
SjD
SSc
PsA
1 2 3 4 5 6 7 8 9...

Cell features or clusters

Single-cell RNA-seq

Abnormal cell features

GZMK
CD21lowB
TPH
CD38
GZMK
IL-18
pDCs
LDG
Mono
IFN
Tc17
TH17

b

Treatment predictor

Disease classifier
RA
SLE
SjD
SSc
PsA

Fig. 2 | Broad immune profiling to identify immune abnormalities in
rheumatic diseases. a, Broad immune profiling by single-cell RNA sequencing
(RNA-seq) is able to define cell types (for example, monocytes, dendritic cells,
B cells, T cells) or cell-activation states (for example, among resting T cells or
activated T cells) that are characteristically altered in specific diseases and to
generate healthy- and rheumatic-disease-associated reference datasets. These
rheumatic-disease-associated cell types or cell states can be considered globally
in a multi-dimensional fashion and then delineated further as specific cellular
parameters. b, The immune profile of an individual sample is mapped against

Nature Reviews Rheumatology

reference profiles to identify cellular features that differ from the healthy
control-associated reference. Comparison with rheumatic disease-associated
references matches individual profiles of undiagnosed individuals to the most
fitting rheumatic-disease reference profile. A combination of these analyses
has the potential to identify treatments that are the most suitable to modulate
the pathologically activated pathways. LDG, low density granulocyte; Mono,
monocyte; pDCs, plasmacytoid dendritic cells; PsA, psoriatic arthritis; RA,
rheumatoid arthritis; SjD, Sjögren disease; SLE, systemic lupus erythematosus;
SSc, systemic sclerosis.

Review article

across studies and samples. The widespread use of droplet-based
scRNA-seq now provides an opportunity to establish generalizable
health- and disease-associated reference datasets. Moreover, advancing
computational approaches will enable integration of scRNA-seq results
from many diverse datasets, enabling cross-disease comparisons across
studies, despite technical and methodological differences146,147.
Leveraging a common language for scRNA-seq-based immune
profiling will enable the mapping of scRNA-seq profiles of individual patient samples against a reference database of health- and
disease-defined scRNA-seq profiles (Fig. 2), indicatively, screening
for a strong type I IFN signature, as seen in patients with SLE, an expansion of ABCs, plasmablasts, TPH and TFH cells, as seen in SLE, ongoing
T cell–B cell interactions, expansion of TH17 cells, activation of IL-1β or
TNF pathways in myeloid cells or abnormal TReg cell profiles. Unbiased
approaches should be able to define SLE-like, RA-like, PsA-like, and
other disease-associated immune profiles, enabling an immunophenotypic definition of immune health or disease-like status with any
sample. Comparison with other states of immune activation, such
as protective anti-viral and antibacterial responses as well as vaccine
responses, will also be valuable.
Currently, technical and logistical challenges remain to be overcome for this kind of approach. RNA transcriptomes change with
incubation time, such that improved methods are needed that limit
artefactual changes in transcriptomic profiles after sample acquisition that may obscure biological signals. Costs of scRNA-seq remain
considerable (typically >US$1,000 per sample), slowing the generation of foundational datasets that can demonstrate the utility of such
immune profiling, yet newer scRNA-seq profiling methods using probe
capture are substantially reducing sequencing costs and broadening
the ability to analyse fixed samples148. Incorporation of scRNA-seq
profiling into the protocols of ongoing industry-sponsored clinical trials, as has been done previously using whole-blood RNA-seq, would be
immensely valuable to generate urgently needed scRNA-seq biomarkers of treatment effect and treatment response. As with more focused
assays, standardization of methods across laboratories, and agreement
on standard reference datasets, will be required to implement these
approaches broadly.

Multi-modal immune profiling
A set of cellular profiling assays has the potential to complement
other modalities that assess immune or inflammatory features, such
as serum proteomic profiling or metabolomic profiling. In some cases,
the different modalities might converge on the same fundamental
observations; for example, a type I IFN signature can be detected by
bulk RNA-seq, PCR, serum proteomic, or cytometric assays in SLE; in
this case, the simplest and most cost-effective method should be used.
Some modalities might, however, measure a given pathway more efficiently than others; for example, the enzyme-linked immunosorbent
assay may be most suitable for detecting a circulating cytokine, mass
spectrometry for a key metabolite and cytometry for a relevant cell
population. Given the rapid advances in tissue-biopsy profiling with
high-dimensional imaging and spatial transcriptomics, specific features of tissue architecture or cell infiltrates, or features of stromal or
parenchymal cells, might also provide unique, non-redundant measures of disease-relevant immunopathology. Key informative inputs
from any of these modalities can be incorporated as components of a
broad assessment of immune dysregulation in patients, adding to the
current assessments of CRP, complement components and autoantibodies. Machine-learning approaches that incorporate both molecular
Nature Reviews Rheumatology

and clinical data also have the potential to establish robust diagnostic
markers, as in a study that improved identification of patients with PsA
using this approach149.

Conclusion
In total, the rapidly expanding universe of immune-profiling data on
blood and tissue samples from patients with rheumatic diseases is
providing an increasingly well-defined set of parameters of immune
dysregulation that is typical for these diseases, highlighting similarities
and differences across diseases and among patients sharing a diagnosis.
Immune profiling has so far highlighted several straightforward parameters of immune dysregulation that are ready for clinical implementation. In the near future, broad tests that assess the current activity level
of the immune system, with an ability to detect pathological immune
activation or deviation from homeostasis, might become as available
as blood tests currently used to interrogate the functioning of other
organs, such as the kidney and the liver. These methods have the potential to dramatically improve assessment of immune-mediated disease
and guide therapeutic decisions for patients with rheumatic diseases.

Published online: xx xx xxxx
References
1.

Bray, C. et al. Erythrocyte sedimentation rate and C-reactive protein measurements and
their relevance in clinical medicine. WMJ 115, 317–321 (2016).
2.
Weinstein, A., Alexander, R. V. & Zack, D. J. A review of complement activation in SLE.
Curr. Rheumatol. Rep. 23, 16 (2021).
3.
Sanjabi, S. & Lear, S. New cytometry tools for immune monitoring during cancer
immunotherapy. Cytom. B Clin. Cytom. 100, 10–18 (2021).
4. Hartmann, F. J. & Bendall, S. C. Immune monitoring using mass cytometry and related
high-dimensional imaging approaches. Nat. Rev. Rheumatol. 16, 87–99 (2020).
5.
Stubbington, M. J. T., Rozenblatt-Rosen, O., Regev, A. & Teichmann, S. A. Single-cell
transcriptomics to explore the immune system in health and disease. Science 358,
58–63 (2017).
6. Stoeckius, M. et al. Simultaneous epitope and transcriptome measurement in
single cells. Nat. Methods 14, 865–868 (2017).
7.
Peterson, V. M. et al. Multiplexed quantification of proteins and transcripts in single cells.
Nat. Biotechnol. 35, 936–939 (2017).
8.
Becht, E. et al. Dimensionality reduction for visualizing single-cell data using UMAP.
Nat. Biotechnol. https://doi.org/10.1038/nbt.4314 (2018).
9.
Zhang, S., Li, X., Lin, J., Lin, Q. & Wong, K. C. Review of single-cell RNA-seq data clustering
for cell-type identification and characterization. RNA 29, 517–530 (2023).
10. Fonseka, C. Y. et al. Mixed-effects association of single cells identifies an expanded
effector CD4+ T cell subset in rheumatoid arthritis. Sci. Transl. Med. 10, eaaq0305
(2018).
11. Reshef, Y. A. et al. Co-varying neighborhood analysis identifies cell populations
associated with phenotypes of interest from single-cell transcriptomics. Nat. Biotechnol.
40, 355–363 (2022).
12. Dann, E., Henderson, N. C., Teichmann, S. A., Morgan, M. D. & Marioni, J. C. Differential
abundance testing on single-cell data using k-nearest neighbor graphs. Nat. Biotechnol.
40, 245–253 (2022).
13. Stone, J. H. et al. Trial of tocilizumab in giant-cell arteritis. N. Engl. J. Med. 377, 317–328
(2017).
14. Devauchelle-Pensec, V. et al. Effect of tocilizumab on disease activity in patients with
active polymyalgia rheumatica receiving glucocorticoid therapy: a randomized clinical
trial. JAMA 328, 1053–1062 (2022).
15. Dima, A., Opris, D., Jurcut, C. & Baicus, C. Is there still a place for erythrocyte
sedimentation rate and C-reactive protein in systemic lupus erythematosus? Lupus 25,
1173–1179 (2016).
16. Aringer, M. Inflammatory markers in systemic lupus erythematosus. J. Autoimmun. 110,
102374 (2020).
17. Bennett, L. et al. Interferon and granulopoiesis signatures in systemic lupus
erythematosus blood. J. Exp. Med. 197, 711–723 (2003).
18. Morand, E. F. et al. Trial of anifrolumab in active systemic lupus erythematosus.
N. Engl. J. Med. 382, 211–221 (2020).
19. Cooles, F. A. H. & Isaacs, J. D. The interferon gene signature as a clinically relevant
biomarker in autoimmune rheumatic disease. Lancet Rheumatol. 4, e61–e72 (2022).
20. Lipsky, P. E. et al. Biological impact of iberdomide in patients with active systemic lupus
erythematosus. Ann. Rheum. Dis. 81, 1136–1142 (2022).
21. Tanaka, H. et al. Extracting immunological and clinical heterogeneity across autoimmune
rheumatic diseases by cohort-wide immunophenotyping. Ann. Rheum. Dis. 83, 242–252
(2024).

Review article

22.
23.

24.
25.
26.

27.
28.

29.
30.

31.

32.
33.

34.
35.
36.
37.
38.

39.
40.

41.

42.
43.

44.
45.
46.

47.
48.

49.

50.
51.

52.
53.

Burns, M. et al. Dysregulated CD38 expression on peripheral blood immune cell subsets
in SLE. Int. J. Mol. Sci. 22, 2424 (2021).
Katsuyama, E. et al. The CD38/NAD/SIRTUIN1/EZH2 axis mitigates cytotoxic CD8 T cell
function and identifies patients with SLE prone to infections. Cell Rep. 30, 112–123.e4
(2020).
Wang, R. et al. Clonally expanded CD38hi cytotoxic CD8 T cells define the T cell infiltrate
in checkpoint inhibitor-associated arthritis. Sci. Immunol. 8, eadd1591 (2023).
Bocharnikov, A. V. et al. PD-1hi CXCR5− T peripheral helper cells promote B cells responses
in lupus via MAF and IL-21. JCI Insight 4, e130062 (2019).
He, J. et al. Circulating precursor CCR7loPD-1hi CXCR5+ CD4+ T cells indicate Tfh cell
activity and promote antibody responses upon antigen reexposure. Immunity 39,
770–781 (2013).
Lin, J., Yu, Y., Ma, J., Ren, C. & Chen, W. PD-1+CXCR5−CD4+T cells are correlated with the
severity of systemic lupus erythematosus. Rheumatology 58, 2188–2192 (2019).
Makiyama, A. et al. Expanded circulating peripheral helper T cells in systemic lupus
erythematosus: association with disease activity and B cell differentiation. Rheumatology
58, 1861–1869 (2019).
Wang, S. et al. IL-21 drives expansion and plasma cell differentiation of autoreactive
CD11chiT-bet+ B cells in SLE. Nat. Commun. 9, 1758 (2018).
Jenks, S. A. et al. Distinct effector B cells induced by unregulated toll-like receptor 7
contribute to pathogenic responses in systemic lupus erythematosus. Immunity 49,
725–739.e6 (2018).
Tipton, C. M. et al. Diversity, cellular origin and autoreactivity of antibody-secreting cell
population expansions in acute systemic lupus erythematosus. Nat. Immunol. 16,
755–765 (2015).
Marks, K. E. & Rao, D. A. T peripheral helper cells in autoimmune diseases. Immunol. Rev.
307, 191–202 (2022).
Balog, J. et al. Comparative single-cell multiplex immunophenotyping of therapy-naive
patients with rheumatoid arthritis, systemic sclerosis, and systemic lupus erythematosus
shed light on disease-specific composition of the peripheral immune system.
Front. Immunol. 15, 1376933 (2024).
Ota, M. et al. Dynamic landscape of immune cell-specific gene regulation in
immune-mediated diseases. Cell 184, 3006–3021.e17 (2021).
Chen, S. et al. Interleukin 17A and IL-17F expression and functional responses in rheumatoid
arthritis and peripheral spondyloarthritis. J. Rheumatol. 47, 1606–1613 (2020).
Edwards, J. C. et al. Efficacy of B-cell-targeted therapy with rituximab in patients with
rheumatoid arthritis. N. Engl. J. Med. 350, 2572–2581 (2004).
Mease, P. J. Is there a role for rituximab in the treatment of spondyloarthritis and psoriatic
arthritis? J. Rheumatol. 39, 2235–2237 (2012).
Penkava, F. et al. Single-cell sequencing reveals clonal expansions of pro-inflammatory
synovial CD8 T cells expressing tissue-homing receptors in psoriatic arthritis. Nat. Commun.
11, 4767 (2020).
Zhang, F. et al. Deconstruction of rheumatoid arthritis synovium defines inflammatory
subtypes. Nature 623, 616–624 (2023).
Zhang, F. et al. Defining inflammatory cell states in rheumatoid arthritis joint synovial
tissues by integrating single-cell transcriptomics and mass cytometry. Nat. Immunol. 20,
928–942 (2019).
Yager, N. et al. Ex vivo mass cytometry analysis reveals a profound myeloid
proinflammatory signature in psoriatic arthritis synovial fluid. Ann. Rheum. Dis. 80,
1559–1567 (2021).
Rao, D. A. et al. Pathologically expanded peripheral T helper cell subset drives B cells in
rheumatoid arthritis. Nature 542, 110–114 (2017).
Fortea-Gordo, P. et al. Two populations of circulating PD-1hiCD4 T cells with distinct B cell
helping capacity are elevated in early rheumatoid arthritis. Rheumatology 58, 1662–1673
(2019).
Sowerby, J. M. & Rao, D. A. T cell-B cell interactions in human autoimmune diseases.
Curr. Opin. Immunol. 93, 102539 (2025).
Povoleri, G. A. M. et al. Psoriatic and rheumatoid arthritis joints differ in the composition
of CD8+ tissue-resident memory T cell subsets. Cell Rep. 42, 112514 (2023).
Steel, K. J. A. et al. Polyfunctional, proinflammatory, tissue-resident memory phenotype and
function of synovial interleukin-17A+CD8+ T cells in psoriatic arthritis. Arthritis Rheumatol.
72, 435–447 (2020).
Mease, P. J. et al. Secukinumab inhibition of interleukin-17A in patients with psoriatic
arthritis. N. Engl. J. Med. 373, 1329–1339 (2015).
McInnes, I. B. et al. Secukinumab, a human anti-interleukin-17A monoclonal
antibody, in patients with psoriatic arthritis (FUTURE 2): a randomised, double-blind,
placebo-controlled, phase 3 trial. Lancet 386, 1137–1146 (2015).
Qaiyum, Z., Gracey, E., Yao, Y. & Inman, R. D. Integrin and transcriptomic profiles identify
a distinctive synovial CD8+ T cell subpopulation in spondyloarthritis. Ann. Rheum. Dis.
78, 1566–1575 (2019).
Jonsson, A. H. et al. Granzyme K+ CD8 T cells form a core population in inflamed human
tissue. Sci. Transl. Med. 14, eabo0686 (2022).
Reece, R. J., Canete, J. D., Parsons, W. J., Emery, P. & Veale, D. J. Distinct vascular patterns
of early synovitis in psoriatic, reactive, and rheumatoid arthritis. Arthritis Rheum. 42,
1481–1484 (1999).
Floudas, A. et al. Distinct stromal and immune cell interactions shape the pathogenesis
of rheumatoid and psoriatic arthritis. Ann. Rheum. Dis. 81, 1224–1242 (2022).
Abji, F. et al. Proteinase-mediated macrophage signaling in psoriatic arthritis. Front. Immunol.
11, 629726 (2020).

Nature Reviews Rheumatology

54. Fragoulis, G. E. et al. Distinct innate and adaptive immunity phenotypic profile at the
circulating single-cell level in psoriatic arthritis. Clin. Immunol. 253, 109679 (2023).
55. Li, B. et al. Differential immunological profiles in seronegative versus seropositive
rheumatoid arthritis: Th17/Treg dysregulation and IL-4. Front. Immunol. 15, 1447213
(2024).
56. Bhamidipati, K. et al. Spatial patterning of fibroblast TGFβ signaling underlies treatment
resistance in rheumatoid arthritis. Preprint at bioRxiv, https://doi.org/10.1101/
2025.03.14.642821 (2025).
57. Naidoo, J. et al. Toxicities of the anti-PD-1 and anti-PD-L1 immune checkpoint antibodies.
Ann. Oncol. 26, 2375–2391 (2015).
58. Ghosh, N. & Bass, A. R. Rheumatic complications of immune checkpoint inhibitors.
Rheum. Dis. Clin. North. Am. 48, 411–428 (2022).
59. Ghosh, N. et al. Identification of outcome domains in immune checkpoint
inhibitor-induced inflammatory arthritis and polymyalgia rheumatica: a scoping review
by the OMERACT irAE working group. Semin. Arthritis Rheum. 58, 152110 (2023).
60. Kim, S. T. et al. Distinct molecular and immune hallmarks of inflammatory arthritis induced
by immune checkpoint inhibitors for cancer therapy. Nat. Commun. 13, 1970 (2022).
61. Maschmeyer, P. et al. Antigen-driven PD-1+ TOX+ BHLHE40+ and PD-1+ TOX+ EOMES+
T lymphocytes regulate juvenile idiopathic arthritis in situ. Eur. J. Immunol. 51, 915–929
(2021).
62. Barturen, G., Beretta, L., Cervera, R., Van Vollenhoven, R. & Alarcón-Riquelme, M. E. Moving
towards a molecular taxonomy of autoimmune rheumatic diseases. Nat. Rev. Rheumatol. 14,
75–93 (2018).
63. Pitzalis, C., Kelly, S. & Humby, F. New learnings on the pathophysiology of RA from
synovial biopsies. Curr. Opin. Rheumatol. 25, 334–344 (2013).
64. Lewis, M. J. et al. Molecular portraits of early rheumatoid arthritis identify clinical and
treatment response phenotypes. Cell Rep. 28, 2455–2470.e5 (2019).
65. Pitzalis, C., Jones, G. W., Bombardieri, M. & Jones, S. A. Ectopic lymphoid-like structures
in infection, cancer and autoimmunity. Nat. Rev. Immunol. 14, 447–462 (2014).
66. Humby, F. et al. Synovial cellular and molecular signatures stratify clinical response
to csDMARD therapy and predict radiographic progression in early rheumatoid arthritis
patients. Ann. Rheum. Dis. 78, 761–772 (2019).
67. Dunlap, G. et al. Clonal associations between lymphocyte subsets and functional states
in rheumatoid arthritis synovium. Nat. Commun. 15, 4991 (2024).
68. Durham, L. E. et al. Substantive similarities between synovial fluid and synovial tissue
t cells in inflammatory arthritis via single-cell RNA and T cell receptor sequencing.
Arthritis Rheumatol. 76, 1594–1601 (2024).
69. Kubo, S. et al. Peripheral blood immunophenotypic diversity in patients with rheumatoid
arthritis and its impact on therapeutic responsiveness. Ann. Rheum. Dis. 84, 210–220
(2025).
70. Goto, M. et al. Age-associated CD4+ T cells with B cell-promoting functions are regulated
by ZEB2 in autoimmunity. Sci. Immunol. 9, eadk1643 (2024).
71. Horisberger, A. et al. Blood immunophenotyping identifies distinct kidney histopathology
and outcomes in patients with lupus nephritis. J. Clin. Invest. https://doi.org/10.1172/
jci181034 (2025).
72. Sasaki, T. et al. Clonal relationships between Tph and Tfh cells in patients with SLE and in
murine lupus. Preprint at bioRxiv, https://doi.org/10.1101/2025.01.27.635189 (2025).
73. Perez, R. K. et al. Single-cell RNA-seq reveals cell type-specific molecular and genetic
associations to lupus. Science 376, eabf1970 (2022).
74. Nehar-Belaid, D. et al. Mapping systemic lupus erythematosus heterogeneity at the
single-cell level. Nat. Immunol. 21, 1094–1106 (2020).
75. Billi, A. C. et al. Nonlesional lupus skin contributes to inflammatory education of
myeloid cells and primes for cutaneous inflammation. Sci. Transl. Med. 14, eabn2263
(2022).
76. Dunlap, G. S. et al. Single-cell transcriptomics reveals distinct effector profiles of
infiltrating T cells in lupus skin and kidney. JCI Insight 7, e156341 (2022).
77. Arazi, A. et al. The immune cell landscape in kidneys of patients with lupus nephritis.
Nat. Immunol. 20, 902–914 (2019).
78. Der, E. et al. Tubular cell and keratinocyte single-cell transcriptomics applied to lupus
nephritis reveal type I IFN and fibrosis relevant pathways. Nat. Immunol. 20, 915–927
(2019).
79. Danaher, P. et al. Childhood-onset lupus nephritis is characterized by complex
interactions between kidney stroma and infiltrating immune cells. Sci. Transl. Med. 16,
eadl1666 (2024).
80. Banchereau, R. et al. Personalized immunomonitoring uncovers molecular networks that
stratify lupus patients. Cell 165, 551–565 (2016).
81. Guthridge, J. M. et al. Adults with systemic lupus exhibit distinct molecular phenotypes
in a cross-sectional study. eClinicalMedicine 20, 100291 (2020).
82. Hubbard, E. L. et al. Analysis of transcriptomic features reveals molecular endotypes
of SLE with clinical implications. Genome Med. 15, 84 (2023).
83. Garantziotis, P. et al. Molecular taxonomy of systemic lupus erythematosus through
data-driven patient stratification: molecular endotypes and cluster-tailored drugs.
Front. Immunol. 13, 860726 (2022).
84. Rincon-Arevalo, H. et al. Deep phenotyping of CD11c+ B cells in systemic autoimmunity
and controls. Front. Immunol. 12, 635615 (2021).
85. Wehr, C. et al. A new CD21low B cell population in the peripheral blood of patients with
SLE. Clin. Immunol. 113, 161–171 (2004).
86. Kubo, S. et al. Peripheral immunophenotyping identifies three subgroups based on T cell
heterogeneity in lupus patients. Arthritis Rheumatol. 69, 2029–2037 (2017).

Review article

87.

Sasaki, T. et al. Longitudinal immune cell profiling in early systemic lupus erythematosus.
Arthritis Rheumatol. 74, 1808–1821 (2022).
88. Choi, J. Y. et al. Circulating follicular helper-like T cells in systemic lupus erythematosus:
association with disease activity. Arthritis Rheumatol. 67, 988–999 (2015).
89. Jenks, S. A., Cashman, K. S., Woodruff, M. C., Lee, F. E. & Sanz, I. Extrafollicular responses
in humans and SLE. Immunol. Rev. 288, 136–148 (2019).
90. Rao, D. A. T cells that help B cells in chronically inflamed tissues. Front. Immunol. 9, 1924
(2018).
91. Skougaard, M. et al. Four emerging immune cellular blood phenotypes associated with
disease duration and activity established in psoriatic arthritis. Arthritis Res. Ther. 24, 262
(2022).
92. Urbanski, G. et al. Single-cell RNA sequencing of peripheral blood defines two
immunological subtypes of Sjögren’s disease. Preprint at bioRxiv, 2025.2002.2027.640483,
https://doi.org/10.1101/2025.02.27.640483 (2025).
93. Aletaha, D. & Smolen, J. S. Diagnosis and management of rheumatoid arthritis: a review.
JAMA 320, 1360–1372 (2018).
94. Navarra, S. V. et al. Efficacy and safety of belimumab in patients with active systemic lupus
erythematosus: a randomised, placebo-controlled, phase 3 trial. Lancet 377, 721–731 (2011).
95. Rovin, B. H. et al. Efficacy and safety of voclosporin versus placebo for lupus nephritis
(AURORA 1): a double-blind, randomised, multicentre, placebo-controlled, phase 3 trial.
Lancet 397, 2070–2080 (2021).
96. Mulhearn, B., Barton, A. & Viatte, S. Using the immunophenotype to predict response to
biologic drugs in rheumatoid arthritis. J. Pers. Med. 9, 46 (2019).
97. Wientjes, M. H. M., den Broeder, A. A., Welsing, P. M. J., Verhoef, L. M. & van den Bemt,
B. J. F. Prediction of response to anti-TNF treatment using laboratory biomarkers in
patients with rheumatoid arthritis: a systematic review. RMD Open. 8, e002570 (2022).
98. Cohen, S. et al. A molecular signature response classifier to predict inadequate response
to tumor necrosis factor-α inhibitors: the NETWORK-004 prospective observational
study. Rheumatol. Ther. 8, 1159–1176 (2021).
99. Bergman, M. J. et al. Clinical utility and cost savings in predicting inadequate response
to anti-TNF therapies in rheumatoid arthritis. Rheumatol. Ther. 7, 775–792 (2020).
100. Aldridge, J. et al. Blood PD-1+TFh and CTLA-4+CD4+ T cells predict remission after
CTLA-4Ig treatment in early rheumatoid arthritis. Rheumatology 61, 1233–1242 (2022).
101. Monjo-Henry, I. et al. Circulating Tfh cells are differentially modified by abatacept or
TNF blockers and predict treatment response in rheumatoid arthritis. Rheumatology 64,
517–525 (2025).
102. Edner, N. M. et al. Follicular helper T cell profiles predict response to costimulation
blockade in type 1 diabetes. Nat. Immunol. 21, 1244–1255 (2020).
103. Sellam, J. et al. Blood memory B cells are disturbed and predict the response to
rituximab in patients with rheumatoid arthritis. Arthritis Rheum. 63, 3692–3701 (2011).
104. Möller, B. et al. Class-switched B cells display response to therapeutic B-cell depletion in
rheumatoid arthritis. Arthritis Res. Ther. 11, R62 (2009).
105. Roll, P., Dörner, T. & Tony, H. P. Anti-CD20 therapy in patients with rheumatoid arthritis:
predictors of response and B cell subset regeneration after repeated treatment.
Arthritis Rheum. 58, 1566–1575 (2008).
106. Vital, E. M., Dass, S., Buch, M. H., Rawstron, A. C. & Emery, P. An extra dose of rituximab
improves clinical response in rheumatoid arthritis patients with initial incomplete B cell
depletion: a randomised controlled trial. Ann. Rheum. Dis. 74, 1195–1201 (2015).
107. Baker, K. F. et al. Single-cell insights into immune dysregulation in rheumatoid arthritis
flare versus drug-free remission. Nat. Commun. 15, 1063 (2024).
108. Orange, D. E. et al. RNA identification of PRIME cells predicting rheumatoid arthritis
flares. N. Engl. J. Med. 383, 218–228 (2020).
109. Rivellese, F. et al. Rituximab versus tocilizumab in rheumatoid arthritis: synovial
biopsy-based biomarker analysis of the phase 4 R4RA randomized trial. Nat. Med. 28,
1256–1268 (2022).
110. Nerviani, A. et al. A pauci-immune synovial pathotype predicts inadequate response
to TNFα-blockade in rheumatoid arthritis patients. Front. Immunol. 11, 845 (2020).
111. Rivellese, F. et al. Stratification of biological therapies by pathobiology in biologic-naive
patients with rheumatoid arthritis (STRAP and STRAP-EU): two parallel, open-label,
biopsy-driven, randomised trials. Lancet Rheumatol. 5, e648–e659 (2023).
112. Humby, F. et al. Rituximab versus tocilizumab in anti-TNF inadequate responder patients
with rheumatoid arthritis (R4RA): 16-week outcomes of a stratified, biopsy-driven,
multicentre, open-label, phase 4 randomised controlled trial. Lancet 397, 305–317 (2021).
113. Alivernini, S. et al. Distinct synovial tissue macrophage subsets regulate inflammation
and remission in rheumatoid arthritis. Nat. Med. 26, 1295–1306 (2020).
114. MacDonald, L. et al. Synovial tissue myeloid dendritic cell subsets exhibit distinct
tissue-niche localization and function in health and rheumatoid arthritis. Immunity 57,
2843–2862.e12 (2024).
115. Thomas, T. et al. A longitudinal single-cell atlas of anti-tumour necrosis factor treatment
in inflammatory bowel disease. Nat. Immunol. 25, 2152–2165 (2024).
116. Carter, L. M., Wigston, Z., Laws, P. & Vital, E. M. Rapid efficacy of anifrolumab across
multiple subtypes of recalcitrant cutaneous lupus erythematosus parallels changes in
discrete subsets of blood transcriptomic and cellular biomarkers. Br. J. Dermatol. 189,
210–218 (2023).
117. Baker, T. et al. Type I interferon blockade with anifrolumab in patients with systemic lupus
erythematosus modulates key immunopathological pathways in a gene expression and
proteomic analysis of two phase 3 trials. Ann. Rheum. Dis. 83, 1018–1027 (2024).
118. Casey, K. A. et al. Type I interferon receptor blockade with anifrolumab corrects innate
and adaptive immune perturbations of SLE. Lupus Sci. Med. 5, e000286 (2018).

Nature Reviews Rheumatology

119. Law, C. et al. Interferon subverts an AHR-JUN axis to promote CXCL13+ T cells in lupus.
Nature 631, 857–866 (2024).
120. Vital, E. M. et al. Anifrolumab efficacy and safety by type I interferon gene signature and
clinical subgroups in patients with SLE: post hoc analysis of pooled data from two phase
III trials. Ann. Rheum. Dis. 81, 951–961 (2022).
121. Moysidou, G. S. et al. Molecular basis for the disease-modifying effects of belimumab
in systemic lupus erythematosus and molecular predictors of early response: blood
transcriptome analysis implicates the innate immunity and DNA damage response
pathways. Ann. Rheum. Dis. 84, 262–273 (2025).
122. Hoffman, R. W. et al. Gene expression and pharmacodynamic changes in 1,760 systemic
lupus erythematosus patients from two phase III trials of BAFF blockade with tabalumab.
Arthritis Rheumatol. 69, 643–654 (2017).
123. Nakano, M. et al. Distinct transcriptome architectures underlying lupus establishment
and exacerbation. Cell 185, 3375–3389.e21 (2022).
124. Sun, W. et al. Heterogeneity of peripheral immune cell landscape in systemic lupus
erythematosus patients after belimumab treatment. Clin. Exp. Rheumatol. 43, 1259–1276
(2025).
125. Maeda, S. et al. High-dimensional analysis of T-cell profiling variations following belimumab
treatment in systemic lupus erythematosus. Lupus Sci. Med. 10, e000976 (2023).
126. Prete, M. et al. Belimumab restores Treg/Th17 balance in patients with refractory
systemic lupus erythematosus. Lupus 27, 1926–1935 (2018).
127. Merrill, J. T. et al. Obexelimab in systemic lupus erythematosus with exploration of
response based on gene pathway co-expression patterns: a double-blind, randomized,
placebo-controlled, phase 2 trial. Arthritis Rheumatol. 75, 2185–2194 (2023).
128. Schafer, P. H. et al. Cereblon modulator iberdomide induces degradation of the
transcription factors Ikaros and Aiolos: immunomodulation in healthy volunteers and
relevance to systemic lupus erythematosus. Ann. Rheum. Dis. 77, 1516–1523 (2018).
129. Müller, F. et al. CD19 CAR T-cell therapy in autoimmune disease — a case series with
follow-up. N. Engl. J. Med. 390, 687–700 (2024).
130. Hagen, M. et al. BCMA-targeted T-cell-engager therapy for autoimmune disease.
N. Engl. J. Med. 391, 867–869 (2024).
131. Alexander, T., Krönke, J., Cheng, Q., Keller, U. & Krönke, G. Teclistamab-induced remission
in refractory systemic lupus erythematosus. N. Engl. J. Med. 391, 864–866 (2024).
132. Wilhelm, A. et al. Selective CAR T cell-mediated B cell depletion suppresses IFN
signature in SLE. JCI Insight 9, e179433 (2024).
133. Ramoni, R. B. et al. The Undiagnosed Diseases Network: accelerating discovery about
health and disease. Am. J. Hum. Genet. 100, 185–192 (2017).
134. Mueller, A. A. et al. High-dimensional immunophenotyping reveals immune cell aberrations
in patients with undiagnosed inflammatory and autoimmune diseases. J. Clin. Invest. 133,
e169619 (2023).
135. Boycott, K. M. et al. International cooperation to enable the diagnosis of all rare genetic
diseases. Am. J. Hum. Genet. 100, 695–705 (2017).
136. Frésard, L. et al. Identification of rare-disease genes using blood transcriptome
sequencing and large control cohorts. Nat. Med. 25, 911–919 (2019).
137. Montgomery, S. B., Bernstein, J. A. & Wheeler, M. T. Toward transcriptomics as a primary
tool for rare disease investigation. Cold Spring Harb. Mol. Case Stud. 8, a006198 (2022).
138. Liu, C. et al. Multi-lineage transcriptional and cell communication signatures define
pathways in individuals at-risk for developing rheumatoid arthritis that initiate and
perpetuate disease. Preprint at bioRxiv, https://doi.org/10.1101/2025.02.08.619913 (2025).
139. Inamo, J. et al. Deep immunophenotyping reveals circulating activated lymphocytes in
individuals at risk for rheumatoid arthritis. J. Clin. Investig. 135, e185217 (2025).
140. Sasaki, T. et al. A CD57+ cytotoxic CD8 T cell subset associated with fibrotic lung disease
in systemic sclerosis. Preprint at bioRxiv, 2025.2001.2027.635121, https://doi.org/
10.1101/2025.01.27.635121 (2025).
141. Biesen, R. et al. Sialic acid-binding Ig-like lectin 1 expression in inflammatory and resident
monocytes is a potential biomarker for monitoring disease activity and success of
therapy in systemic lupus erythematosus. Arthritis Rheum. 58, 1136–1145 (2008).
142. Rose, T. et al. IFNα and its response proteins, IP-10 and SIGLEC-1, are biomarkers of
disease activity in systemic lupus erythematosus. Ann. Rheum. Dis. 72, 1639–1645
(2013).
143. Fang, H., Wang, S. A., Medeiros, L. J. & Wang, W. Application of flow cytometry
immunophenotypic analysis for the diagnosis of mature B-cell lymphomas/leukemias.
Hum. Pathol. 156, 105711 (2025).
144. Nguyen, A. A. & Platt, C. D. Flow cytometry-based immune phenotyping of T and
B lymphocytes in the evaluation of immunodeficiency and immune dysregulation.
Immunol. Allergy Clin. North. Am. 45, 189–203 (2025).
145. LaBere, B. et al. Clinical utility of measuring CD4+ T follicular cells in patients with
immune dysregulation. J. Autoimmun. 140, 103088 (2023).
146. Kotliar, D. et al. Reproducible single cell annotation of programs underlying T-cell
subsets, activation states, and functions. Preprint at bioRxiv, https://doi.org/10.1101/
2024.05.03.592310 (2024).
147. Kang, J. B. et al. Efficient and precise single-cell reference atlas mapping with Symphony.
Nat. Commun. 12, 5890 (2021).
148. De Simone, M. et al. Comparative analysis of commercial single-cell RNA sequencing
technologies. Preprint at bioRxiv, 2024.2006.2018.599579, https://doi.org/10.1101/
2024.06.18.599579 (2024).
149. Liu, J. et al. Combined single cell transcriptome and surface epitope profiling identifies
potential biomarkers of psoriatic arthritis and facilitates diagnosis via machine learning.
Front. Immunol. 13, 835760 (2022).

Review article

150. Scott, D. L., Smith, C. & Kingsley, G. Joint damage and disability in rheumatoid arthritis:
an updated systematic review. Clin. Exp. Rheumatol. 21, S20–S27 (2003).
151. Wakefield, R. J. et al. After treat-to-target: can a targeted ultrasound initiative improve
RA outcomes? Ann. Rheum. Dis. 71, 799–803 (2012).
152. Raychaudhuri, S. et al. Five amino acids in three HLA proteins explain most of the association
between MHC and seropositive rheumatoid arthritis. Nat. Genet. 44, 291–296 (2012).
153. Ishigaki, K. et al. Multi-ancestry genome-wide association analyses identify novel genetic
mechanisms in rheumatoid arthritis. Nat. Genet. 54, 1640–1651 (2022).
154. McDermott, G. C. et al. Polygenic risk scores for rheumatoid arthritis and idiopathic
pulmonary fibrosis and associations with RA, interstitial lung abnormalities, and quantitative
interstitial abnormalities among smokers. Semin. Arthritis Rheum. 72, 152708 (2025).
155. Yarwood, A. et al. A weighted genetic risk score using all known susceptibility variants to
estimate rheumatoid arthritis risk. Ann. Rheum. Dis. 74, 170–176 (2015).
156. Krenn, V. et al. Grading of chronic synovitis — a histopathological grading system for
molecular and diagnostic pathology. Pathol. Res. Pract. 198, 317–325 (2002).
157. Kuo, D. et al. HBEGF+ macrophages in rheumatoid arthritis induce fibroblast invasiveness.
Sci. Transl. Med. 11, eaau8587 (2019).
158. Croft, A. P. et al. Distinct fibroblast subsets drive inflammation and damage in arthritis.
Nature 570, 246–251 (2019).
159. Wei, K. et al. Notch signalling drives synovial fibroblast identity and arthritis pathology.
Nature 582, 259–264 (2020).
160. Bao, W., Xie, M. & Ye, Y. Age-associated B cells indicate disease activity in rheumatoid
arthritis. Cell Immunol. 377, 104533 (2022).
161. Cooles, F. A. H. et al. Phenotypic and transcriptomic analysis of peripheral blood
plasmacytoid and conventional dendritic cells in early drug naïve rheumatoid arthritis.
Front. Immunol. 9, 755 (2018).

Nature Reviews Rheumatology

Acknowledgements

D.A.R. has been supported in part by the Burroughs Wellcome Fund Career Award in Medical
Sciences and the Doris Duke Clinical Scientist Development Award. I thank Kevin Wei, Craig
Platt, Kathryne Marks and Daimon Simmons for helpful discussions.

Competing interests

D.A.R. reports grant support from Janssen, Merck and Bristol Myers Squibb outside of the
current report, and reports personal fees from AstraZeneca, Merck, AbbVie, Biogen, Simcere,
Epana, HiFiBio and Bristol Myers Squibb. He is co-inventor on a patent using TPH cells as a
biomarker of autoimmune diseases.

Additional information

Peer review information Nature Reviews Rheumatology thanks Elena Hsieh, Yoshiya Tanaka,
George Kalliolias and the other, anonymous, reviewer(s) for their contribution to the peer
review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this
article under a publishing agreement with the author(s) or other rightsholder(s); author
self-archiving of the accepted manuscript version of this article is solely governed by the
terms of such publishing agreement and applicable law.
© Springer Nature Limited 2025

nature reviews rheumatology

https://doi.org/10.1038/s41584-024-01179-5

Review article

Check for updates

The essential roles of memory
B cells in the pathogenesis of
systemic lupus erythematosus
Thomas Dörner

1

& Peter E. Lipsky

2

Abstract

Sections

Emerging evidence indicates that memory B cells are dysfunctional
in systemic lupus erythematosus (SLE). They are hyporesponsive to
signalling through the B cell receptor (BCR) but retain responsiveness
to Toll-like receptor (TLR) and type I interferon signalling, as well as to
T cell-mediated activation via CD40–CD154. Chronic exposure
to immune complexes of ribonucleoprotein (RNP)-specific
autoantibodies and TLR-engaging or BCR-engaging cargo is likely to
contribute to this partially anergic phenotype. TLR7 or TLR8 signalling
and the resulting production of type I interferon, as well as the sustained
activation by bystander T cells, fuel a positive feedforward loop in
memory B cells that can evade negative selection and permit preferential
expansion of anti-RNP autoantibodies. Clinical trials of autologous
stem cell transplantation or of B cell-targeted monoclonal antibodies and
chimeric antigen receptor (CAR) T cells have correlated replenishment
of the memory B cell population with relapse of SLE. Moreover, the BCR
hyporesponsiveness of memory B cells might explain the failure of
non-depleting B cell-targeting approaches in SLE, including BTK inhibitors
and anti-CD22 monoclonal antibodies. Thus, targeting of dysfunctional
memory B cells might prove effective in SLE, while also avoiding the
adverse events of broad-spectrum targeting of B cell and plasma cell
subsets that are not directly involved in disease pathogenesis.

Introduction
SLE predisposition and
pathogenic B cells
B cell abnormalities in
SLE initiation
Converging pathogenic
pathways
Implications for B cell
therapies
Limitations and future
research directions
Conclusions

Department Medicine/Rheumatology and Clinical Immunology, Charite Universitätsmedizin Berlin & Deutsches
Rheumaforschungszentrum (DRFZ), Berlin, Germany. 2AMPEL BioSolutions, Charlottesville, VA, USA.
e-mail: thomas.doerner@charite.de

1

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

770

Review article

Key points

B cell maturation and B cell activation require highly sophisticated
orchestration of molecular processes in the bone marrow and the
periphery, respectively. B cell maturation in the bone marrow involves
sequential rearrangements of the genes that encode the heavy and
light chains of the B cell receptor (BCR). In the periphery, naive B cell
activation occurs in secondary lymphoid tissues (spleen, lymph
nodes and Peyer’s patches) following the stimulation of innate
immune cells and T cells by exogenous stimuli. BCR binding often
results in antigen internalization and presentation using MHC class II,
and the initial steps of B cell activation are supported by signals from
appropriately stimulated T follicular helper cells (TFH cells) or T peripheral helper cells (TPH cells)1. T cell–B cell collaboration occurs at the
T cell to B cell interface of secondary lymphoid organs and requires
physical interactions between CD154 (also known as CD40L) and
CD40, followed by signalling via the cytokines IL-21, type I interferons, type II interferon (also known as IFNγ) and TGFβ, and later on by
IL-6, BAFF and APRIL, which are all important for the survival of B cells
and plasma cells.
Effectively activated B cells enter germinal centres within B cell
follicles and undergo rapid clonal expansion, immunoglobulin heavy
chain class switching, affinity maturation, somatic hypermutation and
differentiation into both memory B cells and plasma cells. Vaccination
studies in healthy volunteers have provided a basis for understanding
B cell activation during primary and secondary immune responses2,3.
Vaccination is known to lead to the maturation of IgM followed by

IgG responses, and selection of high-affinity clones, finally ending
with the resolution of the immune response. A successful primary
immune response to exogenous antigens requires the tight regulation of naive B cell activation, expansion, somatic hypermutation,
differentiation and selection by cytokines, and interactions between
T cells, B cells and follicular dendritic cells (FDCs) that take place within
the dark and light zones of germinal centres4. The germinal centre light
zone is an important site for peripheral negative selection of autoreactive B cell clones. Upon secondary vaccine challenges, memory B cells
are immediately reactivated and naive B cells are recruited to germinal
centres and extrafollicular sites, but these responses resolve within
2–4 weeks. Memory B cells circulate widely in secondary lymphoid
organs and tissues, whereas plasma cells largely home to the bone marrow, where they mature into short-lived or long-lived plasma cells that
produce antibodies over various periods of time5. This highly regulated
T cell-dependent B cell activation is essential for immune protection
against infections but is simultaneously prone to lymphomagenesis
or tolerance breakdown and allergy or autoimmunity. In systemic
lupus erythematosus (SLE), many aspects of this process are deranged,
resulting in persistent activation and differentiation of effector cells
without appropriate resolution.
Although signalling through the BCR is involved in several steps of
B cell activation, maturation and negative selection, B cells are additionally regulated by Toll-like receptor (TLR) signalling and CD40 activation
by the T helper cell surface molecule CD154. In the context of SLE, the
intracellular TLR7 and TLR8 molecules respond to nuclear antigens
(such as single-stranded RNA (ssRNA) or U-rich ssRNA) by triggering
the production of type I interferon, whereas activated TFH and TPH cells
provide B cell survival and differentiation signals6–8 (Fig. 1).
SLE is an autoimmune disease characterized by profound perturbations of B cell activation and differentiation, which together
result in the production of a variety of pathogenic autoantibodies9.
Numerous abnormalities of B cell function have been documented
in patients with SLE10, and B cell-directed therapies have shown efficacy in some of these patients11. Early research in SLE mainly delineated phenotypic differences, including an expanded memory B cell
population12,13 and associated the persistence of subsets of atypical
memory B cells, such as CD27−IgD− (double-negative (DN) B cells)14
and CD11c+ age-associated B cells7,15 (ABCs) that do not rapidly contract as seen following virus infections16, with disease activity. Later
studies found substantial functional and spatial abnormalities in
B cell activation and differentiation in SLE6,17,18. The initial activation
of antigen-naive B cells appears to be dysregulated in SLE, potentially
abetted by ineffective negative selection and, therefore, resulting in the
enrichment for autoantigen reactivity19. Autoreactive memory B cell
subsets have been found to differentiate at sites that do not support
negative selection processes, such as within extrafollicular sites or in
tissues18,20,21. Importantly, memory B cells from patients with SLE have
shown reduced BCR responsiveness in functional studies, reflecting
a profound post-activation anergy22,23 (Fig. 1). This finding strikingly
contradicts textbook knowledge referring to ‘hyperactive B cells in SLE’,
which is often used as a rationale for B cell targeting. In this context, the
study of SLE-associated B cells has revived interest in BCR-independent
B cell activation pathways, namely TLR7 and CD40 activation, providing the basis for an ‘SLE pathogenesis at the crossroad of innate and
adaptive immunity’ hypothesis.
Based on these developments, we suggest that memory B cells
are essential for the induction and persistence of SLE, and that effective SLE therapies might require specific depletion or suppression of

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

771

• In systemic lupus erythematosus (SLE), memory B cells are
hyporesponsive to B cell receptor (BCR) stimulation but can be
activated upon engagement of Toll-like receptors (TLRs) and
interaction with T cells (mainly via the CD40–CD40L axis). Both innate
and adaptive immune signalling by B cells (‘bridging’) contribute to
SLE pathology, possibly via a pathogenic positive feedforward loop.
• This feedforward loop is accentuated by anti-ribonucleoprotein
(anti-RNP) autoantibodies sequestering RNP antigens, which, when
internalized via the BCR, stimulate TLR7 and TLR8 signalling and type I
interferon production.
• Incomplete X chromosomal inactivation of TLR7, TLR8 and CD40L
might further contribute to such a positive feedforward loop, thereby
potentially explaining the female sex bias in SLE.
• Clinical outcomes of B cell depletion in SLE, via anti-CD20 or anti-CD19
or autologous stem cell transplantation, have clearly associated relapse
with memory B cell repletion, independently of the recurrence of naive
B cells or autoantibodies.
• The safety and efficacy of CD19-targeted and BCMA-targeted
chimeric antigen receptor (CAR) T cells, or bispecific T cell engagers in
SLE, and their impact on tissue-resident memory B cells remain to be
elucidated.
• BCR signalling inhibition approaches did not result in sufficient
efficacy potentially owing to an incomplete impact on memory B cells.

Introduction

Review article

B cell signalling in SLE

B cell signalling in infection

CD154
CD40

CD79A
CD79B

BCR

BCR
PI3K

SYK LYN
TRAF2/6
TRAF3

AKT1

BTK

CD19

CD40

PI3K

SYK LYN
TRAF2/6
TRAF3

CD154

CD19

AKT1

BTK

mTOR
NIK

PKCβ

IKK

IKK

NF-κΒ

NF-κΒ

mTOR

MYD88
IRAK4
TRAF3 IRAK1 IRAK2
IRAK3
IKK
NF-κΒ

TLR8
TLR7
Endosome

NIK

PKCβ

IKK

IKK

NF-κΒ

NF-κΒ

MYD88
IRAK4
TRAF3 IRAK1 IRAK2
IRAK3
IKK

IKKα

NF-κΒ

Type I IFN
and ISRE
IRF7

Nucleus

Fig. 1 | Signalling abnormalities in B cells, especially memory B cells,
in systemic lupus erythematosus. a, In B cells from healthy individuals,
stimulation of the B cell receptor (BCR) induces phosphorylation of the
downstream signalling cascade via the kinases SYK, LYN and BTK, finally resulting
in adequate activation of transcription factor NF-κB. In addition, CD19 signalling
leads to the activation of the PI3K and AKT and the downstream initiation
of mTOR signalling. Ligation of CD40 by CD154 (also known as CD40L) leads to
engagement of TNFR-associated factors (TRAFs) and the downstream activation
of NF-κΒ-inducing kinase (NIK) and IκΒ kinase (IKK), leading to the activation of
NF-κΒ. In addition, the binding of endosomal single-stranded RNA molecules to

this B cell population. We argue that current B cell-directed therapies
are only active insofar as they affect the memory B cell population
and that B cell-directed therapies will only achieve improved and
persistent efficacy if they ensure eradication or re-education of this
B cell population. Moreover, we suggest that targeting other B cell
subsets, such as plasma cells, might confer some transient benefit by
the reduction of autoreactive plasma cells fuelling the formation of
pathogenic immune complexes, but disease recrudescence is inevitable if SLE-associated memory B cells persist and reinitiate disease
immunopathogenesis.
In this Review, we highlight preconditions and early events leading
to the emergence of autoreactive B cells and SLE initiation, as well as
the ensuing mechanisms of memory B cell formation and reactivation,
and discuss how abnormalities in memory B cell responses, including
BCR hyporesponsiveness combined with sustained TLR signalling and
activation via bystander T cells might explain the continuous breach
of peripheral tolerance, the preferential autoantibody repertoire,
and female sex bias in SLE. We next propose that a positive feedforward loop involving both innate (TLR7 or TLR8) and adaptive (CD40)
immune signalling underlies the memory B cell reactivation and autoreactive plasma cell differentiation in SLE, and discuss how disruption
of this feedback loop might correlate with efficacy of B cell-targeted
therapeutic strategies.
Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

Endosome

IKKα
IRF7

IRF7

NF-κΒ

TLR8
TLR7

Type I IFN
and ISRE
NF-κΒ

IRF7

Nucleus

Toll-like receptor 7 (TLR7) or TLR8 leads to engagement of the adaptor molecule
MYD88 and downstream recruitment of IL-1 receptor-associated kinases (IRAKs),
leading to the downstream activation of NF-κΒ and interferon regulatory factor 7
(IRF7). b, In patients with systemic lupus erythematosus (SLE), memory B cells
have a hyporesponsive BCR with inadequate phosphorylation of the BCR
signalling cascade but accessible TLR7 and TLR8 signalling initiated within
endosomes and activated by single-stranded RNAs, as well as a functional
CD40 activation pathway activated by the CD154 molecule on T helper cells.
IFN, interferon; ISRE, IFN-stimulated response element.

SLE predisposition and pathogenic B cells
Sex bias and other genetic factors

Several factors known to predispose to SLE, such as female sex or
certain HLA class II haplotypes, appear to be important contributors to abnormal B cell function in this disease. In particular, SLE
displays a striking female sex bias, with approximately 90% of
the people affected being women24. Various hypotheses, including the hormone–hormone receptor signalling hypothesis, have
been proposed to explain this phenomenon. A compelling explanation of the female sex bias in SLE relates to gene dosage of X
chromosome-encoded immune molecules. Notably, molecules with
a key role in innate and adaptive immune activation are encoded
on the X chromosome, including TLR7, TLR8, interleukin-receptor
associated kinase-1 (IRAK1), Bruton’s tyrosine kinase (BTK), CD154,
as well as X-inactive specific transcript (Xist) (Supplementary Fig. 1).
Incomplete inactivation of X chromosomal loci in women might result
in a gene dose-dependent increase in relevant immune functions
promoting female sex bias for type I interferon-associated rheumatic
autoimmune inflammatory diseases (RAIDs)25. The same mechanisms
might also confer enhanced protection against virus infections in
women, which might explain their lower morbidity and mortality26,27
following infections. Thus, improved virus protection appears to be
associated with an enhanced risk of autoimmunity in females, and this
772

Review article

risk seems to be at least in part related to certain immune molecules
encoded on the X chromosome.
Accordingly, early associations of systemic autoimmunity with
TLR7 duplications28 and incompletely silenced X chromosome genes in
Klinefelter syndrome (XXY)29 were followed by recognition of the association of the X chromosome-encoded long non-coding RNA Xist with
the female sex bias of RAIDs30. Xist encodes a long non-coding RNA that
is universally expressed and regulates X chromosome inactivation31.
Immune genes that escape X chromosomal silencing28, such as TLR7
(refs. 32,33), CD154, CXCR3, BTK, TASL, interleukin-2 receptor (IL2R) or
IRAK1, have been associated with the expansion of atypical memory
B cells in humans with autoimmunity and in mouse models of autoimmunity and ageing. In addition, interaction of Xist with numerous
RNA-binding proteins30 appears to increase the immunogenicity of
the latter and has also been associated with the expansion of atypical
memory B cells and ABCs30.
Another important genetic precondition of adaptive immunity is related to HLA class II haplotypes. The DR2 alleles DRB1*1501/
DQB1*0602 and the DR3 alleles DRB1*0301/DQB1*0201 were found to
be present in nearly two-thirds of 780 patients with SLE and their family
members. DR2 haplotypes have been associated with autoantibodies to
the nuclear antigen Sm, whereas DR3 genotypes have been associated
with SSA/Ro-specific and SSB/La-specific autoantibodies. Thus, HLA
class II DR2 and DR3 haplotypes are key elements involved in specific
autoantibody production and susceptibility to SLE34.

Early emergence of autoantibodies
In SLE, current data clearly support the view that abnormal B cell activation and the associated breach of immune tolerance give rise to
autoantibodies years before SLE manifestations35,36. There is considerable evidence that T cells are involved in establishing the autoimmune
B cell and plasma cell repertoire in SLE, resulting in the emergence of
typical autoantibodies years before autoimmune disease onset (Supplementary Fig. 2). In individuals producing typical autoantibodies
without overt disease, increased levels of type II interferon (IFNγ) are
an early abnormality36, and these findings combined with the known
contribution of IFNγ to T cell responses suggest that initial production
of autoantibodies is largely controlled by T cells7,34 (Supplementary
Fig. 2). Upon occurrence, most anti-ribonucleoprotein (anti-RNP) antibodies are strongly associated with the type I interferon signature37,
but the plethora of SLE-related autoantibodies makes it very difficult to
understand the role of each of them in driving disease. With the exception of anti-Sm antibodies, other anti-RNP autoantibodies do not occur
exclusively in SLE. Thus, an early break of immune tolerance precedes
the onset of SLE, but the induction mechanisms of autoimmunity in
these individuals remain largely to be delineated.

B cell abnormalities in SLE initiation

Skewed B cell repertoires have been noted in SLE19,38, suggesting abnormalities in both B cell differentiation within germinal centres or at extrafollicular sites and B cell survival within lymphoid organs and peripheral
tissues. Resulting perturbances in B cell function as a result of chronic
BCR cell stimulation or innate immune cell-derived and T cell-derived
cytokines have provided important new insights and clearly support a
role for B cells, and especially for memory B cells, in SLE pathogenesis.

hypermutation of immunoglobulin heavy chain genes, the detailed contribution of T cells to B cell responses during established SLE remains
to be resolved. Patients with SLE have reduced numbers of regulatory
T cells (Treg cells) and increased numbers of T helper 17 cells (TH17 cells),
associated with an imbalance between IL-2 and IL-17 levels25. In addition,
tissue-resident memory T cells (TRM cells) that produce the chemokine
CXCL13 and are able to attract B cells into germinal centres and potentially other tissues have been implicated in SLE pathogenesis8. CXCL13+
TPH and TFH cells appear to be expanded and related to persistent type I
IFN-driven T cell abnormalities. Even though these cells may be important in stimulating B cell responses, the effectiveness of B cell-directed
therapies suggests they have no other essential functions in disease
pathogenesis39–41.

Perturbed naive B cell function and repertoire
A CD19+CD20+CD5+CD38+CD10+CD9+IgD+CD27− population of pre-naive
B cells that are enriched in autoantibody specificities has been shown to
be expanded in SLE42. As with other early B cell subsets, this population
responds suboptimally to BCR activation, but can be activated via CD40
engagement42. Importantly, the lack of complete culling of autoreactivity in this pre-naive B cell population suggests that, when activated
by bystander help, or by TLR7 engagement, pre-naive B cells are likely
to differentiate into memory B cells and plasma cells or plasmablasts
with enrichment in autoantibody specificities43.
Conventional CD27− naive B cells from patients with SLE also
respond suboptimally to BCR engagement42. This phenotype resembles the BCR anergic status of B cells seen in chronic viral infections44,
suggesting that chronic in vivo activation, particularly when combined
with prolonged exposure to type I interferon, might induce partial
anergy already in the premature immune repertoire. Indeed, elevated
STAT1 expression in B cells and T cells from patients with SLE is a typical sign of long-term exposure to type I interferon signalling45. This
persistent type I interferon environment might also be an important
factor for maintaining BCR hyporesponsiveness22.
In summary, naive B cells with a hyporesponsive BCR occur in
patients with SLE, potentially as a result of chronic exposure to autoantigens, type I interferons or both. The observations of naive B cell
repletion and reduced type I interferon expression in responders following B cell depletion are consistent with the assumption that type I
interferon contributes to the abnormal BCR responsiveness of naive
B cells in SLE44. Interestingly, signalling via the largely intact CD40
(T cell) and TLR activation pathways can overcome the anergic status
of conventional CD27− B cells from patients with SLE. Thus, a unique
set of abnormalities can skew the ability of both pre-naive and naive
B cells to be properly activated and regulated in SLE22. Suboptimal BCR
signalling might also underlie the compromised immune protection
against infections in SLE. Whether any of these abnormalities can be targeted uniquely by novel therapeutics is currently not known, although
the possibility that reducing the exposure to type I interferon and/or
engagement by TLR7 and TLR8 ligands offer potential approaches for
future evaluations.

Abnormal memory B cell function

Although there are essential requirements for T cell collaboration
in the induction of immunoglobulin class switching and somatic

Functionally, memory B cells from healthy individuals have greatly
reduced requirements to differentiate into plasma cells compared
with naive B cells. This characteristic is important for rapid immune
protection against pathogens and for mounting adequate vaccine
responses. Memory B cells rapidly differentiate into plasma cells in
response to IL-21 and BAFF signalling, even in the absence of T cells

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

773

T cell help

Review article

or BCR engagement, and this rapid response ensures survival from
infection46,47.
In RAIDs, there is considerable evidence of the pathogenic relevance of the memory B cell compartment. Initially, phenotypic abnormalities of peripheral B cell subsets in SLE38,48 and other RAIDs were
identified, with expanded conventional memory B cells, atypical DN1,
DN2 and DN3 memory B cells or CD11c+ ABC subsets, and with peripheral plasmacytosis18,49. Functional memory B cell abnormalities have
been identified in both conventional CD27−IgD+ and atypical CD27−IgD−
B cells17,22 (Fig. 1), along with a unique dependence on glycolysis for both
survival and function50. These findings make it plausible that memory
B cells are biologically different in patients with SLE than in healthy
controls, and function as independent drivers of disease (Box 1).
In SLE, conventional and atypical memory B cells can be extensively reactivated outside germinal centres, especially within extrafollicular sites and autoimmune tissues8,18,51. Light zone-like structures
containing FDCs are usually not detectable in extrafollicular sites and
in tissues affected by autoimmune activity, indicating that peripheral
negative selection of autoreactive clones might be defective in extrafollicular or in situ autoimmune responses (Fig. 2). In this context,

Box 1 | Immune activation during
established autoimmune disease does
not follow principles of conventional
secondary activation
Spatial abnormalities: In systemic lupus erythematosus (SLE),
memory B cell reactivation is largely confined to extrafollicular and
autoimmune tissue reactivation lacking germinal centre regulation
and negative selection. During secondary vaccine challenges,
memory B cell reactivation occurs preferentially in secondary
lymphoid organs with germinal centres and at extrafollicular sites.
Functional abnormalities: In SLE, memory B cells have an anergic
phenotype, with B cell receptor (BCR) hyporesponsiveness as
a result of increased protein tyrosine phosphatase activity that
controls BCR signalling. Toll-like receptor (TLR) signalling and CD40
signalling remain intact. TLR7 and TLR8 signalling overrides the
response to BCR activation, and autoantibody specificities are fixed
as B cells do not undergo affinity maturation. Autoantibodies are,
thus, often specific to TLR ligands, and although they are usually
mutated and of high avidity, maturation of BCR binding during
the disease does not occur. Moreover, cytokine (mainly type I
interferon) signalling is constitutive, lacking sequential adaptation
of the immune response via cytokine switch. The emerging
autoreactive memory B cell and plasma cell compartments are not
contracting over time. By contrast, after vaccination, BCR affinity
is finely tuned, and orchestrated cytokine sequences impact on
immunoglobulin switch and somatic hypermutation to increase
the immunoglobulin repertoire exposed to negative selection by
follicular dendritic cells in the light zones of the germinal centres.
Subsequently, the memory compartments become quiescent and
contract.
Metabolic abnormalities: In SLE, B cells predominantly undergo
glycolysis, and this is not observed in control B cells.

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

autoreactive clones can emerge in SLE and thereby differ from conventional memory B cell reactivation as observed for secondary vaccine
responses (Box 1). In summary, the reactivation of memory B cells and
induction of the atypical memory B cell subsets DN1, DN2 and DN3 in
SLE are largely confined to extrafollicular and in situ tissue activation
contributing to the critical expansion of the memory compartment
and insufficient mechanisms to control autoreactivity.
The disbalance between anergic BCR and responsive TLR7 signalling in anti-RNP+ memory B cells in SLE might also contribute to the
selective expansion of memory B cells outside the germinal centres.
There is clear evidence of memory B cell activation at extrafollicular
sites38,52,53, such as proliferative lymphoid nodules (PLNs) within the
spleen, which, in the case of immune thrombocytopenia51, are distant
from germinal centres or in some circumstances nearby atrophic germinal centres, or within affected tissues17, such as the tubulointerstitial tissue in lupus nephritis54. PLNs within the spleen or kidney lack
important features of germinal centres, including the polarized FDCs,
and maintain autoantigen presentation even during steady state. As
such, reactivation of memory B cells in the absence of the regulatory
influences of FDCs might perpetuate ongoing autoimmune responses.
Of particular note, TPH cells have been associated with the expansion of
DN1, DN2, DN3, and CD11c+ ABC atypical memory B cells within autoimmune tissues1,8. However, it still remains to be clarified whether TPH cells
simply co-localize with B cells within affected tissues or also provide
help to these B cells.

Converging pathogenic pathways
Type I interferon and B cell abnormalities co-occur in patients with
SLE8,45 or some other RAIDs, including in patients with extraglandular
Sjögren syndrome55 and a subset of patients with rheumatoid arthritis
(RA)56, which suggests that these two types of dysregulation might
be interrelated during both the initiation and the maintenance of
autoimmune disease (crossroad hypothesis).
More specifically, emerging evidence indicates that perturbed
memory B cell activity resulting in the production of anti-RNP
autoantibodies and the upregulation of the type I interferon pathway converge to potentiate autoimmune disease43. From a reverse
translational perspective, the downregulation of type I interferon
signalling upon selective B cell depletion treatments and the association of disease recurrence with memory B cell repletion are supportive of this model (for details see below). Memory B cells appear
to be the major focal point of the convergence with dysregulation
of the type 1 interferon pathway (Fig. 2). Although the rudiments of
memory B cell generation are similar in autoimmunity and in response
to exogenous antigens, some aspects are distinct (Box 1). In particular, protective responses result in orchestrated downregulation of
immune activation, including the downregulation of type I interferon
signalling, when the infection is resolved44. In individuals predisposed
to autoimmunity, this resolution of immune activation leads to chronic
autoimmunity and a continuous immune stimulation that potentially
overrides peripheral negative selection of autoreactive B cell clones6,17.
The resulting memory B cells have autoreactive specificities and a
dysregulated signalling programme manifested as BCR hyporesponsiveness and enhanced responsiveness to bystander T cell help and
TLR7 and TLR8 engagement. As a result, these memory B cells can
respond to TPH cells at extrafollicular foci or in autoimmune tissues and
differentiate into autoantibody-secreting plasmablasts or plasma cells
that produce autoantibodies, including anti-RNP autoantibodies.
Notably, they are not exposed to censoring by FDCs4.
774

Review article

RNA

TLR7/8 inhibition
IRAK4 inhibition

RBP

Hyporesponsive
BCR

anti-CD40 or anti-CD154

CD154 CD40

Plasma cell
TLR7

TFH, TRM,
or TPH cell
Anti-BAFF
Anti-BAFF/APRIL
BAFF/APRIL
Myeloid/antigenpresenting cell

TCR
Peptide-MHC

Autoreactive anti-RNP
PC subsets

NF-κB

MYD88

IRF7

Memory
B cell
Anti-RNP

Memory B cell and
plasma cell survival

Damaged cells
Type I IFN

Anti-IFNAR (anifrolumab)

DNA
RNA
DAMP or
PAMP

pDC
FcR blockade
HMGB1

Tissue damage

Anti-RNP-antigen
immune complexes

Fig. 2 | A positive feedforward loop links abnormal memory B cell
signalling, autoantibody production and the type I interferon signature
of SLE. During initiation of systemic lupus erythematosus (SLE), naive B cells
and their corresponding autoreactive plasma cells are primed to recognize
ribonucleoprotein (RNP) autoantigens at germinal centres years before disease
onset. Genetic factors such as certain HLA class II haplotypes predispose towards
the activation of B cells that recognize RNP antigens, leading to the generation of
autoreactive memory B cells and CD19− long-lived plasma cells that reside in the
bone marrow, where they produce anti-RNP autoantibodies. Following disease
initiation, B cell receptor (BCR) signalling is anergic or dysfunctional and the BCR
can only function to internalize the cognate autoantigens. Following antigen
internalization, memory B cells are reactivated via Toll-like receptor 7 (TLR7)
signalling, and this process defines the spectrum of autoantibodies with largely
TLR7 specificities. Notably, memory B cell reactivation occurs outside germinal
centres with support from T peripheral helper (TPH) cells or tissue-resident
memory T (TRM) cells which reside within areas that do not support effective
peripheral negative selection, such as the extrafollicular areas of lymphoid
tissues or the autoimmunity-affected tissues, where follicular dendritic cells

(FDCs) are absent. The resulting autoantibodies form immune complexes that
fuel the activation of plasmacytoid dendritic cells (pDCs), which produce type I
interferons (IFN) and support the production of BAFF by myeloid cells. The
feedforward loop comprises a unique but crucial interaction between the two
cellular subsets, memory B cells (probably including the atypical memory B cell
subsets DN1, DN2 and DN3) and autoreactive plasma cells as well as the two key
cytokines type I interferon and BAFF. These elements along with TLR7 and TLR8
ligands and CD40 bystander stimulation define the destiny of anti-RNP memory
B cells and are all important drivers of SLE pathology. Non-depleting strategies
such as anti-CD40 or anti-CD154 monoclonal antibodies, TLR7 and TLR8
inhibitors, interleukin-receptor associated kinase 1 (IRAK4) inhibitors, blockade
of Fc receptors (FcR), interferon α receptor (IFNAR)-specific monoclonal
antibody anifrolumab, or anti-BAFF and anti-APRIL monoclonal antibodies,
that are able to interfere with the proposed model are indicated within yellow
fields. DAMP, damage-associated molecular pattern; PAMP, pathogen-associated
molecular pattern; PC, plasma cell; RBP, RNA-binding protein; TCR, T cell
receptor; TFH, T follicular helper.

Anti-RNP antibodies

how intracellular and intranuclear autoantigens drive autoantibody
production even when their localization excludes their direct binding
by cognate autoantibodies. This concept might explain why anti-RNP
autoantibodies dominate across diseases with distinct clinical phenotypes that only share a common type I interferon signature (SLE, Sjögren
syndrome, RA subsets, mixed connective tissue disease).

Anti-RNP autoantibodies build immune complexes with RNA binding
proteins57, and these immune complexes can feedback on memory
B cells via engagement of TLR7 and TLR8. In addition, the immune
complexes along with type I interferon itself enhance production of
BAFF, further propagating memory B cell stimulation58. This evolving
pattern of events, with memory B cells acting at the crossroad, might
account for the propagation of autoimmunity and provides many targets for intervention. Even when the hyporesponsive BCR (Fig. 1) does
not transmit an appropriate intracellular activation signal, we hypothesize that it might function to internalize autoantigens22, including the
TLR7 and TLR8 ligands bound to anti-RNP immune complexes. Based
on their characteristics, the BCR might identify certain RNP structures,
whereas TLR7 and TLR8 signalling is activated strictly by nucleic acids.
The crossroad hypothesis, as such, might also provide a clue concerning
Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

A positive feedforward loop for B cell activation
After development of autoantibodies in individuals without overt disease who subsequently develop SLE, an increase of type II interferon has
been noted36, consistent with the hypothesis that T cell involvement
potentially instructs B cells at this stage. Subsequently, and shortly
before disease presentation, the levels of type I interferon were found to
increase, indicating a potential contribution of type I interferon during
the stage immediately preceding the onset of clinical manifestations36.
775

Review article

These findings suggest that the pathway towards clinical disease
involves several steps: initial B cell activation is followed by effector
T cell activation and, next, by an innate immune response that heralds
the onset of clinical manifestations.
During established SLE, we propose that a positive proinflammatory feedforward loop (Fig. 2), interconnects abnormal
­memory B cell reactivation with increased type I interferon and BAFF
levels as well as with the high titres of anti-RNP autoantibodies that are
produced by plasma cells and form immune complexes, which both
stimulate memory B cells and activate plasmacytoid dendritic cells to
produce more type I interferon. The observation of decreased type I
interferon signatures after selective B cell depletion or reduced autoantibody levels39,59 is consistent with such a unique feedforward loop
connecting the B cell and plasma cell compartments.
Collectively, there is clear evidence of a central role of memory
B cells and their associated autoreactive plasma cells in SLE pathogenesis. The two B cell compartments are crucial elements in the positive
feedforward loop leading to chronic production of type I interferon
not observed in protective immunity.

Implications for B cell therapies
Translational insights about the impact of B cell therapies on autoantibody production, immune complex formation and type I interferon
production have substantiated the crucial role of memory B cells in
RAIDs. B cell-targeting therapies have permitted such insights (Fig. 3).
In support of our proposed model that incorporates suboptimal BCR
signalling, the targeting of molecules involved in BCR-mediated activation (BTK inhibitors (BTKi)60,61, anti-CD22 (ref. 62) and anti-CD19
combined with Fc receptor inhibition63) has failed in SLE trials. By contrast, B cell depletion strategies (anti-CD20 (ref. 64), CD19-­targeting
chimeric antigen receptor (CAR) T cells39, anti-CD52 (ref. 65) and
autologous stem cell transplantation (ASCT)66) that can each deplete
memory B cells with variable efficiencies have led to varying degrees
of naive B cell repopulation and reduced autoantibody levels, as well
as decreased type I interferon expression (Fig. 3).

in reduced autoantibody titres and immune complex formation, but
also impairs the differentiation of atypical memory B cells and causes
displacement of tissue-resident memory B cells. In further support of
our model, a meta-analysis of all registered belimumab trials found
improved responses to belimumab in patients who had elevated
­baseline levels of BAFF protein or medium-to-high BAFF and type I IFN
mRNA levels75. These findings further support a relationship between
BAFF and type I interferon levels in SLE.
TACI (TNFRSF13B) Fc fusion proteins that block both BAFF and
APRIL (also known as TNFsf13), including atacicept76 and telitacicept77,
have been studied in SLE. BAFF and APRIL both provide survival signals
to plasma cells but their differential impact on certain plasma cell subsets remains to be delineated. Telitacicept showed remarkable efficacy
in a phase II clinical trial in SLE77 and is currently undergoing further
development for potential registration. An important signature of both
TACI Fc fusion proteins is their impact on all immunoglobulin isotypes
produced by plasma cells (IgM and IgG, but also IgA). Importantly, the
differential impact of BAFF and APRIL on memory B cell generation and
differentiation into plasma cells has not been fully studied, although
BAFF has been shown to have an important role in this process24. How
APRIL and BAFF distinctly impact serum IgM, IgA and IgG levels requires
further delineation. In this context, selective anti-APRIL blockade by the
monoclonal antibody sibeprenlimab has been tested in IgA nephropathy and has provided additional evidence that APRIL is important for
IgA production. A phase II study showed a dose-dependent effect on
proteinuria78 in this otherwise difficult to treat nephropathy. In IgA
nephropathy, autoantibodies target the galactose-deficient hinge
region of IgA1, which leads to the formation of pathogenic immune
complexes79. Sibeprenlimab treatment decreased the levels of all IgA,
including of galactose-deficient IgA1. The role of IgA autoantibodies in
SLE remains uncertain but circulating IgA+ plasmablasts, probably of
mucosal origin, are found to be increased in SLE80. Therefore, studies
of selective APRIL blockade might hold value in SLE.

Insights from direct B cell targeting

Belimumab is an approved monoclonal antibody that blocks BAFF, a
cytokine of the TNF superfamily67 Belimumab impacts atypical memory
B cells by diminishing bystander help required for their activation68.
Interestingly, ABCs express high levels of the BAFF receptor (BAFFR)7
and their numbers are increased in SLE. Treatment with belimumab
resulted in remarkably reduced numbers of atypical memory B cells69,
decreased expression of activation markers by DN B cells70 and diminished autoantibody production71,72. Moreover, contraction of two memory B cell clusters — surface IgA-positive memory B cells, which are likely
to be activated within extrafollicular sites or tissues, and CD11c+CD21−
ABCs68 — was associated with improved responses to belimumab. Belimumab has also been noted to affect certain plasma cell subsets and
reduce IgM and, to a lesser extent, IgG and IgA production72. An initial
increase in peripheral memory B cells is well documented for the first
weeks of treatment with belimumab69 followed by a subsequent decline
in memory B cell numbers below baseline. In addition, patients treated
with belimumab develop deactivated, non-proliferative recirculating
memory B cells with features of disrupted lymphocyte trafficking73,
possibly representing displaced tissue resident memory B cells. Finally,
patients treated with belimumab for up to 312 weeks (6 years) had
remarkable decreases in all B cell subsets74. Thus, belimumab not only
interferes with survival of certain antibody-producing cells resulting

The dysfunctional signalling status of memory B cells in SLE with the
anergic BCR but retained TLR and CD40 responsiveness is a central
element of the proposed positive feedforward loop and appears to
be supported by the outcomes of strategies blocking BCR signalling
by targeting downstream kinases, such as BTK or SYK. Lack of efficacy
in SLE has been noted for the BTKi fenebrutinib61 and evobrutinib60
(Fig. 3), supporting the hypothesis that BCR signalling is defective in
memory B cells in SLE and that its inhibition is unlikely to result in clinical benefit12. In addition, the lack of clinical benefit of the non-depleting
CD22-blocking antibody epratuzumab that also interferes with BCR
signalling81 is consistent with the above hypothesis22,82. Further, the
non-cytolytic monoclonal antibody obexelimab63 that binds CD19,
which is part of the BCR complex, and Fcγ receptor IIb (FcγRIIb) and
is able to inhibit shared signalling pathways downstream of the BCR
in B cells also did not show efficacy in a phase II trial. The aggregate
of these data supports the view that BCR signalling is unnecessary in
chronic SLE and that targeting BCR signalling is unlikely to be successful in SLE. By contrast, BTKi molecules have shown efficacy in
relapsing–remitting multiple sclerosis83 and Sjögren syndrome84,
although inhibition of BCR signalling has shown limited value in
RA85. The differences between overall B cell depletion responses and
responses to BTKi represent unique opportunities to disentangle subtle functional B cell abnormalities in the various autoimmune and
inflammatory diseases.

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

776

Insights from B cell cytokine targeting

Review article

BCR signalling inhibition

Targeting bystander T cell help

Lack of efficiency for BTKi
(fenebrutinib and evobrutinib)
or the non-depleting anti-CD22
(epratuzumab) or the anti-CD19
(obexelimab)

Anti-CD40 (iscalimab) and
anti-CD154 (dapirolizumab,
dazodalibep, frexalimab/
SAR441344) under clinical
evaluation in SLE

CD19

CD20

IgD BCR
CD79
CD40

Naive B cell
CD19+ CD20+ IgD+

BAFF- or APRIL-targeting
strategies

B cell depletion strategies

Bone marrow
Lymphoid tissues
Blood

• Re-occurrence of naive
B cells does not
correlate with relapse
after ASCT, anti-CD20,
anti-CD52 or anti-CD19
CAR T cell therapies
(unknown for TCE/BiTE)

CD19

CD20

IgM BCR

CD27

CD38
CD79

CD79

TACI

BAFFR

CD19

CD38

CD20
CD27

CD19low

CD27
BCMA

CD38

CD27

BCMA

TACI

BAFFR

CD40

BCMA

Memory B cell
CD19+ CD20+ CD27+ IgM+ IgG+
IgA+

Plasmablast
CD19+ CD20+ CD27+ CD38+ Ki67+
(CD19low/CD27high)

Plasma cell
CD19low CD20+ CD38+ BCMA+

Plasma cell
CD19– CD20+ CD38+
BCMA+

Germinal centres
Extrafollicular sites
Inflamed autoimmune
tissue(s)

Bone marrow
(Lymphoid tissue)
(Blood)
(Inflamed autoimmune tissue)

Bone marrow
Lymphoid tissues
Inflamed autoimmune
tissue

Bone marrow

• Re-appearance of memory B
cells correlates with relapse
after ASCT and rituximab
• Better outcomes of
obinutuzumab (2nd
generation) as compared to
rituximab might correlate with
deeper depletion of memory B
cells in extrafollicular sites and
tissues
• Anti-CD52 (alemtuzumab)
correlated with repletion of
memory lymphocytes

• Anti-BAFFR (ianalumab)
associated with decreased
IgM, IgG and IgA
• Anti-CD20 impacts on IgM
followed by IgG levels but not
IgA
• Anti-CD19 CAR T cells
associated with decreased
autoantibody titers and IgM,
IgG and IgA

• Anti-CD19 CAR T cells or
TCE reduce autoantibody
titres but not/less on
protective IgG
• Anti-BCMA CAR T cells or
TCE reduce all
immunoglobulin classes

• Targeted by BCMA only

• Belimumab associated with
decreased numbers and
activation of atypical
memory B cells
• Belimumab reduces
anti-dsDNA production,
reduces IgM, followed by IgG
and IgA
• Belimumab associated with
displacement of tissue-resident
memory B cells and blocks
subsequently naive B cell
differentiation

• Anti-BAFF/APRIL (atacicept
and telitacicept) results in
substantial decline of IgM,
followed by IgG and IgA titres
• Anti-APRIL (sibeprenlimab)
associated with selective
decrease of IgA titres in IgA
nephropathy

• Affecting only a few B cell
subsets but with broad
impact on immunoglobulin
titres

• Affecting only a few B cell
subsets but with broad
impact on
immunoglobulin titres

Fig. 3 | Selective targets of B cell depletion. The development of B cells is
accompanied by changes in surface molecule expression, some of which have
been applied in clinical interventions. Anti-CD20, anti-CD19, anti-CD38 and antiBAFF receptor (BAFFR) targeting results in differential B cell depletion including
CD19low plasma cells (which preferentially reside in the bone marrow). A notable
distinction between CD19 and CD20 targeting relates to the broader coverage
of CD19 targeting from pro B cells through CD19low bone marrow plasma cells.
It remains to be delineated how deep depletion of tissue-resident B cells, including
atypical double-negative memory B cell subsets as well as germinal centreresident B cells in lymphoid and target tissues can be achieved. Here, potential
differences between anti-CD20 and anti-CD19 (chimeric antigen receptor (CAR)
T cells with lymphodepletion/bispecific T cell engagers (BiTEs)) might provide
advantages. CD19 targeting is also considered to differentially affect the bone

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

marrow plasma cell compartment, leaving only CD19− bone marrow plasma
cells untouched and remaining unaffected by CD20 targeting. Targeting BAFF
or BAFFR as well as APRIL (by a direct antibody) or targeting of BAFF and APRIL
by atacicept and telitacicept, respectively, are expected to differentially impact
on certain bone marrow plasma cell subsets. In this context, anti-CD38 as well as
anti-BCMA targeting is considered to completely deplete bone marrow plasma
cells. The aggregate of these different depletion possibilities of B cell lineage cells
will not only provide new treatments for individual patients but also very
detailed insights into B cell lineage development in health and autoimmunity.
Note: During development, CD38 expression extends until the pre-naive stage.
ASCT, autologous stem cell transplantation; BCR, B cell receptor; BTKi, Bruton’s
tyrosine kinase inhibitors; dsDNA, double-stranded DNA; SLE, systemic lupus
erythematosus; TCE, T cell engager.

777

Review article

Another approach to interfere with B cell activation and differentiation is modulation of T cell–B cell interaction by targeting checkpoint molecules. This is usually a non-depleting strategy and targets
intracellular signalling pathways that are distinct from BCR signalling
but provide co-stimulation for proper activation of naive B cells4.
Here the CD40–CD154 pathway is crucial for T cell-dependent
B cell activation and is required for many events in germinal centre
reactions (B cell differentiation, immunoglobulin class switching,
somatic hypermutation) and the reactivation and proliferation of
established memory B cells at extrafollicular sites and within tissues.
Importantly, CD40 signalling is intact in memory B cells in SLE and
might mediate bystander T cell help to activate memory B cells in an
antigen-independent manner. Blockade of CD154 by a humanized
monoclonal antibody (BG9588, 5c8) has been shown to normalize
peripheral B cell abnormalities in lupus nephritis86. Currently, monoclonal antibodies targeting CD40 (iscalimab and others) or CD154
(dapirolizumab, dazodalibep and frexalimab/SAR441344)87 are in
clinical development. A central aim of targeting checkpoint molecules such as CD40 in SLE would be to prevent bystander T cell help

from rescuing memory B cells from BCR hyporesponsiveness and
promoting memory B cell reactivation.
Years of experience with depleting anti-CD20 therapies (rituximab,
ocrelizumab and ofatumumab) with regulatory approval for various
indications, and the concept of depleting a broad spectrum of B cell
lineage cells from pre-B cell to memory B cells (Fig. 3) have provided
evidence that rejuvenation of the B cell system is feasible with sufficient
efficacy and safety11,64. Safety was a key concern during the initial phase
of anti-CD20 therapy development88. Subsequently, strategies were
developedtotargeteitherabroaderrangeofBcellsubsets(CD19-targeted39
or CD19 and BCMA co-targeted CAR T cells89 and bispecific T cell engagers (BiTEs)39,90,91) or selective plasma cell populations (therapies targeting BCMA or CD38 (ref. 92)) (Box 2). The underlying studies will provide
unprecedented insights into the distinct contributions of B cell subsets
including the autoantibody-producing plasma cell subset in individual
patients. A prediction based on the feedforward model is that strategies
that deplete or reprogramme memory B cells might be necessary to
obtain long-term remission, whereas those that decrease plasma cell
numbers and autoantibody titres might be associated with transient or
incomplete responses because of the continuous tick-over of memory
B cells to autoantibody-producing plasma cells.
Evidence obtained from studies of B cell lineage depletion as a
result of ASCT has implications regarding the requirement for extensive reduction of extrafollicular and tissue-resident memory B cells
(‘deep depletion’) in SLE66,93,94. For example, following ASCT, an intervention that affects the entire adaptive immune system but has been
associated with remission for over 5 years in about 70% of patients with
SLE95, the reappearance of naive T cells and B cells was not associated
with disease recurrence41,96. Instead, it was the emergence of memory
T cells and B cells that was linked with relapse. Similarly, following B cell
and T cell co-targeting with anti-CD52 (alemtuzumab65), disease recurrence was related to memory lymphocyte repletion, although success
was limited by overall toxicity owing to long-term lymphopenia and
neutropenia with increased infection risk97. Collectively, strategies
simultaneously depleting T cells and B cells have been associated with
substantial risks of infections and secondary autoimmunity97,98 but
have also provided support to the hypothesis linking a lasting clinical
response to the extensive depletion of memory B cells.
The introduction of selective B cell depletion strategies permitted
additional mechanistic insights. However, anti-CD20 strategies with
rituximab or ocrelizumab were only marginally effective in SLE. The
reasons are manifold but might in part relate to incomplete depletion
of tissue-resident memory B cells or extrafollicular B cells. In this context, second-generation anti-CD20 (obinutuzumab) with enhanced
antibody-dependent cellular cytotoxicity (ADCC) and binding to a
different CD20 epitope showed clinical efficacy in lupus nephritis
and an association with peripheral B cell depletion99. An alternative
explanation of limited anti-CD20 activity was the insufficient targeting of memory T cells, in particular TRM cells in target tissues. In this
regard, incomplete depletion of memory B cells might require that the
T cells driving their tick-over to autoantibody-producing plasma cells
are fully depleted to achieve clinical benefit. New strategies with the
potential to achieve complete memory B cell depletion might obviate
the requirement for T cell depletion. One main finding in long-term
anti-CD20 therapies of SLE was that sustained clinical responses were
noted when the re-populating B cells were dominated by naive B cells100.
Alternative B cell depletion strategies comprise anti-BAFFR targeting with ianalumab101, a monoclonal antibody with the capacity to both
block BAFF binding and deplete BAFFR-positive cells, resulting in more

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

778

Box 2 | Potential approaches to target
autoreactive plasma cells
Several innovative approaches have been proposed to target
autoreactive plasma cells and thereby reduce the source of high
autoantibody levels that otherwise result from abnormal memory
B cell activation in systemic lupus erythematosus (SLE). Such
approaches mostly involve the selective inhibition of plasma cell
differentiation or plasma cell survival and are listed below.
1. Targeting of plasma cell differentiation via the blockade of the
transcription factors IRF4, PRDM1 or XBP1 (ref. 109).
2. Proteasome inhibition to interfere with protein turnover
and, thereby, affect plasma cell survival. Bortezomib has
been used to inhibit overall proteasome activity110, whereas
zetomipzomib has been used for the selective inhibition of the
immunoproteasome111. Zetomipzomib acts very selectively only
on certain subunits of the immunoproteasome complex and is
currently under clinical development for the treatment of lupus
nephritis.
3. Plasma cell depletion with anti-CD38 (ref. 59) monoclonal
antibodies92,112.
4. Targeting of BCMA on the cell surface of plasma cells. BCMA
has been effectively targeted in a mouse model of SLE50, and a
bispecific BCMA-targeted and CD19-targeted chimeric antigen
receptor (CAR) T cell therapy is currently being tested in a phase I
clinical trial89. In addition, an anti-BCMA and anti-CD3 bispecific
antibody or with the BCMA-targeted bispecific T cell engager
(BiTE) teclistamab have shown good responses in patients with
SLE40. As BiTEs are off-the-shelf therapeutics, they might hold
advantages over CAR T cell therapeutics for clinical applications,
as they circumvent the need and risks of lymphodepletion and
the challenges of preparing a genetically engineered cellular
therapeutic. BCMA-targeted CAR T cells, which were initially
developed to eradicate myeloma clones in patients with multiple
myeloma, have shown promising responses in SLE89 and other
autoimmune diseases (neuromyelitis optica spectrum disorder113,
anti-SRP necrotizing myopathy114,115 and myasthenia gravis).

Review article

effective tissue depletion of B cells and partially targeting bone marrow plasma cells (BMPCs). Notably, the expression of BAFFR is higher
on atypical memory B cells and on some plasmablasts or plasma cells
than on other B cell subsets7. Reduction of all immunoglobulin isotypes
(IgM, IgG and in particular IgA) with ianalumab treatment102 could be a
surrogate marker indicating that the treatment impacts at least certain
BMPCs that are not depleted with anti-CD20 treatment.
Recently, CD19-based B cell targeting has been the subject of
regained interest. Despite the limited effectiveness of the anti-CD19
monoclonal antibody in SLE103, recent experience with CD19-targeted
CAR T cells39 has shown remarkable responses in patients with autoimmunity, including eight patients with SLE who achieved DORIS remission without continuous immunosuppressive treatment. Following
treatment with CD19-targeting CAR T cells, patients achieved long-term
remission, although B cell depletion lasted only 112 ± 47 days39. The
correlation between naive B cell dominance after treatment and clinical response is consistent with observations in patients responding
to ASCT41. These data further validate the finding that memory B cells
are key drivers of RAIDs and also link memory B cell depletion with a
diminished type I interferon signature, which is consistent with the
proposed feedforward loop hypothesis. CD19-targeted CAR T cells
also selectively decreased autoantibody versus protective antibody
titres, confirming that the two distinct plasma cell populations (CD19−
and CD19low cells)104 might be clinically relevant. The data imply that
deeper tissue depletion of B cells by migrating CAR T cells might
account for the higher efficacy compared with anti-CD20 antibody
treatments. With regard to targeting BMPCs, bispecific CD19-targeted
and BCMA-targeted CAR T cells also showed promising efficacy and
tolerability in 13 patients with SLE89. This strategy is able to completely
deplete BMPCs, in contrast to CD19 targeting, which is considered
to deplete only a CD19low BMPC subfraction104. The differential role
of the conditioning regimen, appropriate CAR T targets and optimal
patient populations remain to be delineated when compared with
monoclonal antibody therapies.
The aggregate of currently available data suggests that depletion
of the entire memory B cell compartment is key to success in treating
SLE. The key learning may be the association between clinical response
and successful depletion of the memory B cell compartment. Importantly, naive B cells are apparently less important in disease pathogenesis, as active disease does not occur when naive B cells return. To
what extent memory B cell depletion is a precondition for blocking
differentiation of plasma cells and autoantibody production remains
to be determined in further studies.

Although non-depleting interventions might be supported by
a mechanistic rationale and by experiences with belimumab and
anifrolumab, responses to these interventions require time. Most
intriguing and timely responses were found to B cell depletion strategies that result in persistent memory B cell depletion and subsequent
repopulation by naive B cells.

Limitations and future research directions
Considering the complexity of innate and adaptive immune signals
that converge at memory B cells according to the crossroad hypothesis,

Glossary
Age-associated B cells

T follicular helper cells

(ABCs). B cells that increase in number

(TFH cells). TFH cells are

as a result of ageing, viral infections,

antigen-experienced CD4+ T cells

immunodeficiency and autoimmune

expressing PD1 and typically producing

diseases (rheumatoid arthritis and

IL-21, able to initiate and maintain

systemic lupus erythematosus). ABCs

germinal centre formation within

are identified by CD11c expression.

secondary lymphoid organs.

Atypical memory B cells

T peripheral helper cells

A term largely applied to CD27−IgD−

(TPH cells). Unlike T follicular

B cells that lack expression of CD27,

helper cells (TFH cells), which interact with

a marker of memory B cells, but

B cells within lymphoid organs, TPH cells

otherwise have features of B cell

provide help to B cells, and especially to

memory.

memory B cells, within inflamed tissues,

Follicular dendritic cells
(FDCs). Cells of mesenchymal origin

supporting plasma cell differentiation.
Distinct features of TPH cells, as compared
with TFH cells, are the expression of

that are found in the germinal centre

CXCR5, which is associated with TPH cell

light zone of primary and secondary

localization within inflamed tissues, and

lymphoid tissue. FDCs capture

a low BCL6 to BLIMP1 ratio. TPH cells

and present antigens to support

depend on various cytokines for their

B cell activation and, along with

survival within tissues, such as IL-6, type I

CD40–CD40L-based B cell–T cell

interferon and IL-12 or IL-23.

interactions, ensure negative selection
of autoimmune B cells.

Germinal centres

Tissue-resident memory
T cells
(TRM cells). CD4+ memory T cells that

Transiently formed structures within

express BCL6 and are crucially involved

the B cell zone (follicles) in secondary

in the development of autoimmune

Strategies targeting innate signalling pathways

lymphoid organs that harbour a dark

B cell and CD8+ T cell memory

In order to interfere with the consequences of the feedforward feedback loop that is reinforced by the autoantibody-initiated immune
complexes and type I interferon signalling, selective blockade of
TLR7 and TLR8 activation105 or IRAK-4 inhibition106 represent innovative interventions to block these innate immune activation pathways (Fig. 2), and are being assessed in early clinical studies. It will
be of interest to study whether blockade of these pathways results
in clinical efficacy and how the resulting data shed new or confirmatory light on the SLE model outlined above. As most of the anergic
lymphocytes in SLE undergo metabolic adaptations with increased
glycolysis, (semi)selective metabolic approaches by itaconate, metformin or 2-deoxy-d-glucose might also show potential in reverting
abnormal B cell activation or status, although no clinical studies have
yet been announced.

zone where immunoglobulin class

responses. TRM cells can permit the

switching and somatic hypermutation

activation of B cells at extrafollicular or

are taking place and a light zone

tissue sites and thus escape censoring

where BCR/immunoglobulin selection

by germinal centres.

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

occurs based on T cell and follicular
dendritic cell interactions.

Heavy and light chains of the
B cell receptor

TLR7 and TLR8
Members of the Toll-like receptor
family and innate receptors DAMPs
(damage-associated molecular pattern

Antibody molecules are composed of

molecules) able to recognize GU-rich

two immunoglobulin heavy chains and

single-stranded RNA (ssRNA) (TLR7) or

two immunoglobulin light chain proteins,

U-rich ssRNA (TLR8) in endosomes and

the variable regions of which define their

to initiate B cell activation in the contexts

binding specificity.

of viral and autoimmune responses.

779

Review article

questions arise about the crucial role of T cell–B cell interactions in
SLE beyond the impact of CD40–CD154. The extent to which continuous activation of T cells, including TRM cells, in affected tissues might
contribute8 to the proposed positive feedforward loop remains unclear.
Therefore, the impact of memory TPH and TFH cell subsets requires further research, especially given that they appear to resist selected CD20
and CD19 depletion.
Memory B cell activation might be differentially fuelled by signalling downstream of TLR7 versus CD40 across individual patients with
SLE, and biomarker profiling might inform treatment selection, favouring either TLR7 and TLR8 inhibition or CD40–CD40L blockade107,108.
Patient stratification will be of utmost importance given that various
treatments become available in the clinic.
Finally, clinical research in SLE permits unique opportunities
for translational and reverse translational insights into immunology.
Based on the initial identification of antinuclear autoantibodies as a
diagnostic marker, we have entered a period of a better understanding
of the underlying mechanisms and developing targeted therapies. In
this context, selective memory B cell depletion in SLE is currently not
feasible because of the lack of distinct surface markers, but future
research may identify such opportunities.

Conclusions
An aggregate of data from basic, translational and clinical research
emphasizes the crucial role of memory B cells in SLE, a disease that
involves pathogenetic pathways at the cellular crossroads of innate and
adaptive immunity. The preferential expansion of autoreactive memory
B cells with suboptimal BCR signalling but intact responsiveness to
TLR ligands or bystander T cells engaging CD40 appears to explain
several SLE characteristics. First, this hypothesis suggests that the
female predominance in SLE might be associated with the incomplete
silencing of the X chromosomal genes TLR7 or CD40L. Second, the
impairment of protective immune responses and associated susceptibility to infections in patients with SLE appears to be independent of
medication but related to the decreased BCR responsiveness. Third,
the perpetuated production of anti-RNP antibodies by autoreactive
plasma cells might be explained based on the hyporesponsive BCR and
accessible TLR7 and CD40 signalling as combined with the preferential
internalization and processing of cognate TLR7 ligands (RNPs) and the
expansion of bystander T cells. The key aberrations in memory B cell
biology might also explain the ineffectiveness of certain strategies
inhibiting BCR signalling pathways in SLE. Nevertheless, the aggregate of ongoing studies applying various depleting and non-depleting
B cell interventions will provide unprecedented insights into mechanisms of autoimmunity and will also contribute to basic knowledge
of humoral immunity while potentially identifying novel targets for
effective treatment.

Published online: 7 November 2024
References
1.

2.

3.
4.

Akiyama, M., Alshehri, W., Yoshimoto, K. & Kaneko, Y. T follicular helper cells and
T peripheral helper cells in rheumatic and musculoskeletal diseases. Ann. Rheum. Dis. 82,
1371–1381 (2023).
Giesecke, C. et al. Simultaneous presence of non- and highly mutated keyhole limpet
hemocyanin (KLH)-specific plasmablasts early after primary KLH immunization
suggests cross-reactive memory B cell activation. J. Immunol. 200, 3981–3992
(2018).
Giesecke, C. et al. Tissue distribution and dependence of responsiveness of human
antigen-specific memory B cells. J. Immunol. 192, 3091–3100 (2014).
Dorner, T., Giesecke, C. & Lipsky, P. E. Mechanisms of B cell autoimmunity in SLE.
Arthritis Res. Ther. 13, 243 (2011).

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

5.
6.

7.
8.
9.
10.
11.

12.
13.

14.

15.

16.

17.
18.
19.
20.
21.

22.
23.

24.
25.
26.

27.

28.
29.

30.
31.
32.
33.
34.

35.
36.

37.

Hiepe, F. & Radbruch, A. Plasma cells as an innovative target in autoimmune disease with
renal manifestations. Nat. Rev. Nephrol. 12, 232–240 (2016).
Jenks, S. A. et al. Distinct effector B cells induced by unregulated toll-like receptor 7
contribute to pathogenic responses in systemic lupus erythematosus. Immunity 52, 203
(2020).
Cancro, M. P. Age-associated B cells. Annu. Rev. Immunol. 38, 315–340 (2020).
Law, C. et al. Interferon subverts an AHR-JUN axis to promote CXCL13+ T cells in lupus.
Nature 631, 857–866 (2024).
Tsokos, G. C. Systemic lupus erythematosus. N. Engl. J. Med. 365, 2110–2121 (2011).
Jenks, S. A., Cashman, K. S., Woodruff, M. C., Lee, F. E. & Sanz, I. Extrafollicular responses
in humans and SLE. Immunol. Rev. 288, 136–148 (2019).
Reddy, V., Jayne, D., Close, D. & Isenberg, D. B-cell depletion in SLE: clinical and
trial experience with rituximab and ocrelizumab and implications for study design.
Arthritis Res. Ther. 15, S2 (2013).
Odendahl, M. et al. Disturbed peripheral B lymphocyte homeostasis in systemic lupus
erythematosus. J. Immunol. 165, 5970–5979 (2000).
Jacobi, A. M. et al. Correlation between circulating CD27high plasma cells and disease
activity in patients with systemic lupus erythematosus. Arthritis Rheum. 48, 1332–1342
(2003).
Jenks, S. A. et al. Distinct effector B cells induced by unregulated toll-like receptor 7
contribute to pathogenic responses in systemic lupus erythematosus. Immunity 49,
725–739.e726 (2018).
Rubtsova, K., Rubtsov, A. V., Cancro, M. P. & Marrack, P. Age-associated B cells: a T-betdependent effector with roles in protective and pathogenic immunity. J. Immunol. 195,
1933–1937 (2015).
Rincon-Arevalo, H. et al. Impaired humoral immunity to SARS-CoV-2 BNT162b2 vaccine in
kidney transplant recipients and dialysis patients. Sci. Immunol. https://doi.org/10.1126/
sciimmunol.abj1031 (2021).
Szelinski, F. et al. Plasmablast-like phenotype among antigen-experienced CXCR5-CD19low
B cells in systemic lupus erythematosus. Arthritis Rheumatol. 74, 1556–1568 (2022).
Caielli, S., Wan, Z. & Pascual, V. Systemic lupus erythematosus pathogenesis: interferon
and beyond. Annu. Rev. Immunol. 41, 533–560 (2023).
Dorner, T. & Lipsky, P. E. Beyond pan-B-cell-directed therapy – new avenues and insights
into the pathogenesis of SLE. Nat. Rev. Rheumatol. 12, 645–657 (2016).
Ferreira-Gomes, M. et al. Recruitment of plasma cells from IL-21-dependent and
IL-21-independent immune reactions to the bone marrow. Nat. Commun. 15, 4182 (2024).
Hutloff, A. et al. Involvement of inducible costimulator in the exaggerated memory
B cell and plasma cell generation in systemic lupus erythematosus. Arthritis Rheum. 50,
3211–3220 (2004).
Weissenberg, S. Y. et al. Identification and characterization of post-activated B cells in
systemic autoimmune diseases. Front. Immunol. 10, 2136 (2019).
Schrezenmeier, E. et al. Postactivated B cells in systemic lupus erythematosus: update
on translational aspects and therapeutic considerations. Curr. Opin. Rheumatol. 31,
175–184 (2019).
Scofield, R. H. & Lewis, V. M. Is X chromosome inactivation a cause or effect of SLE?
Nat. Rev. Rheumatol. 20, 599–600 (2024).
Wahren-Herlenius, M. & Dorner, T. Immunopathogenic mechanisms of systemic
autoimmune disease. Lancet 382, 819–831 (2013).
Gemmati, D. et al. COVID-19 and individual genetic susceptibility/receptivity: role
of ACE1/ACE2 genes, immunity, inflammation and coagulation. Might the double
X-chromosome in females be protective against SARS-CoV-2 compared to the single
X-chromosome in males? Int. J. Mol. Sci. https://doi.org/10.3390/ijms21103474 (2020).
Scully, E. P., Haverfield, J., Ursin, R. L., Tannenbaum, C. & Klein, S. L. Considering how
biological sex impacts immune responses and COVID-19 outcomes. Nat. Rev. Immunol.
20, 442–447 (2020).
Moisini, I. et al. The Yaa locus and IFN-α fine-tune germinal center B cell selection in
murine systemic lupus erythematosus. J. Immunol. 189, 4305–4312 (2012).
Scofield, R. H. et al. Klinefelter’s syndrome (47,XXY) in male systemic lupus
erythematosus patients: support for the notion of a gene-dose effect from the
X chromosome. Arthritis Rheum. 58, 2511–2517 (2008).
Dou, D. R. et al. Xist ribonucleoproteins promote female sex-biased autoimmunity. Cell
187, 733–749.e16 (2024).
Souyris, M. et al. TLR7 escapes X chromosome inactivation in immune cells.
Sci. Immunol. https://doi.org/10.1126/sciimmunol.aap8855 (2018).
Brown, G. J. et al. TLR7 gain-of-function genetic variation causes human lupus. Nature
605, 349–356 (2022).
Huret, C. et al. Altered X-chromosome inactivation predisposes to autoimmunity.
Sci. Adv. 10, eadn6537 (2024).
Graham, R. R. et al. Specific combinations of HLA-DR2 and DR3 class II haplotypes
contribute graded risk for disease susceptibility and autoantibodies in human SLE.
Eur. J. Hum. Genet. 15, 823–830 (2007).
Arbuckle, M. R. et al. Development of autoantibodies before the clinical onset of
systemic lupus erythematosus. N. Engl. J. Med. 349, 1526–1533 (2003).
Munroe, M. E. et al. Altered type II interferon precedes autoantibody accrual and
elevated type I interferon activity prior to systemic lupus erythematosus classification.
Ann. Rheum. Dis. 75, 2014–2021 (2016).
Hubbard, E. L., Pisetsky, D. S. & Lipsky, P. E. Anti-RNP antibodies are associated with the
interferon gene signature but not decreased complement levels in SLE. Ann. Rheum. Dis.
81, 632–643 (2022).

780

Review article

38. Tipton, C. M. et al. Diversity, cellular origin and autoreactivity of antibody-secreting cell
population expansions in acute systemic lupus erythematosus. Nat. Immunol. 16,
755–765 (2015).
39. Muller, F. et al. CD19 CAR T-cell therapy in autoimmune disease – a case series with
follow-up. N. Engl. J. Med. 390, 687–700 (2024).
40. Alexander, T., Kronke, J., Cheng, Q., Keller, U. & Kronke, G. Teclistamab-induced remission
in refractory systemic lupus erythematosus. N. Engl. J. Med. 391, 864–866 (2024).
41. Alexander, T. et al. Depletion of autoreactive immunologic memory followed by
autologous hematopoietic stem cell transplantation in patients with refractory SLE
induces long-term remission through de novo generation of a juvenile and tolerant
immune system. Blood 113, 214–223 (2009).
42. Lee, J., Kuchen, S., Fischer, R., Chang, S. & Lipsky, P. E. Identification and characterization
of a human CD5+ pre-naive B cell population. J. Immunol. 182, 4116–4126 (2009).
43. Wen, L. et al. Toll-like receptors 7 and 9 regulate the proliferation and differentiation of
B cells in systemic lupus erythematosus. Front. Immunol. 14, 1093208 (2023).
44. Crouse, J., Kalinke, U. & Oxenius, A. Regulation of antiviral T cell responses by type I
interferons. Nat. Rev. Immunol. 15, 231–242 (2015).
45. Aue, A. et al. Elevated STAT1 expression but not phosphorylation in lupus B cells
correlates with disease activity and increased plasmablast susceptibility. Rheumatology
59, 3435–3442 (2020).
46. Wang, S. et al. IL-21 drives expansion and plasma cell differentiation of autoreactive
CD11chiT-bet+ B cells in SLE. Nat. Commun. 9, 1758 (2018).
47. Ettinger, R. et al. IL-21 and BAFF/BLyS synergize in stimulating plasma cell differentiation
from a unique population of human splenic memory B cells. J. Immunol. 178, 2872–2882
(2007).
48. Dorner, T., Jacobi, A. M., Lee, J. & Lipsky, P. E. Abnormalities of B cell subsets in patients
with systemic lupus erythematosus. J. Immunol. Methods 363, 187–197 (2011).
49. Cossarizza, A. et al. Guidelines for the use of flow cytometry and cell sorting in
immunological studies (third edition). Eur. J. Immunol. 51, 2708–3145 (2021).
50. Wilson, J. J. et al. Glucose oxidation-dependent survival of activated B cells provides
a putative novel therapeutic target for lupus treatment. iScience 26, 107487 (2023).
51. Daridon, C. et al. Splenic proliferative lymphoid nodules distinct from germinal centers
are sites of autoantigen stimulation in immune thrombocytopenia. Blood 120, 5021–5031
(2012).
52. Getahun, A., Beavers, N. A., Larson, S. R., Shlomchik, M. J. & Cambier, J. C. Continuous
inhibitory signaling by both SHP-1 and SHIP-1 pathways is required to maintain
unresponsiveness of anergic B cells. J. Exp. Med. 213, 751–769 (2016).
53. Nickerson, K. M., Cullen, J. L., Kashgarian, M. & Shlomchik, M. J. Exacerbated
autoimmunity in the absence of TLR9 in MRL.Faslpr mice depends on Ifnar1. J. Immunol.
190, 3889–3894 (2013).
54. Dorraji, S. E. et al. Kidney tertiary lymphoid structures in lupus nephritis develop into large
interconnected networks and resemble lymph nodes in gene signature. Am. J. Pathol.
190, 2203–2225 (2020).
55. Rose, T. et al. SIGLEC1 is a biomarker of disease activity and indicates extraglandular
manifestation in primary Sjögren’s syndrome. RMD Open. 2, e000292 (2016).
56. Iwasaki, T. et al. Dynamics of type I and type II interferon signature determines
responsiveness to anti-TNF therapy in rheumatoid arthritis. Front. Immunol. 13, 901437
(2022).
57. Ronnblom, L. & Eloranta, M. L. The interferon signature in autoimmune diseases.
Curr. Opin. Rheumatol. 25, 248–253 (2013).
58. Jego, G., Pascual, V., Palucka, A. K. & Banchereau, J. Dendritic cells control B cell growth
and differentiation. Curr. Dir. Autoimmun. 8, 124–139 (2005).
59. Ostendorf, L. et al. Targeting CD38 with daratumumab in refractory systemic lupus
erythematosus. N. Engl. J. Med. 383, 1149–1155 (2020).
60. Wallace, D. J. et al. Efficacy and safety of the Bruton’s tyrosine kinase inhibitor evobrutinib
in systemic lupus erythematosus: results of a phase II, randomized, double-blind,
placebo-controlled dose-ranging trial. ACR Open. Rheumatol. 5, 38–48 (2023).
61. Isenberg, D. et al. Efficacy, safety, and pharmacodynamic effects of the Bruton’s tyrosine
kinase inhibitor fenebrutinib (GDC-0853) in systemic lupus erythematosus: results of
a phase II, randomized, double-blind, placebo-controlled trial. Arthritis Rheumatol. 73,
1835–1846 (2021).
62. Clowse, M. E. et al. Efficacy and safety of epratuzumab in moderately to severely active
systemic lupus erythematosus: results from two phase III randomized, double-blind,
placebo-controlled trials. Arthritis Rheumatol. 69, 362–375 (2017).
63. Merrill, J. T. et al. Obexelimab in systemic lupus erythematosus with exploration of
response based on gene pathway co-expression patterns: a double-blind, randomized,
placebo-controlled, phase 2 trial. Arthritis Rheumatol. 75, 2185–2194 (2023).
64. Ryden-Aulin, M. et al. Off-label use of rituximab for systemic lupus erythematosus in
Europe. Lupus Sci. Med. 3, e000163 (2016).
65. Keith, M. P., Pitchford, C. & Bernstein, W. B. Treatment of hemophagocytic
lymphohistiocytosis with alemtuzumab in systemic lupus erythematosus.
J. Clin. Rheumatol. 18, 134–137 (2012).
66. Burt, R. K. et al. Five year follow-up after autologous peripheral blood hematopoietic
stem cell transplantation for refractory, chronic, corticosteroid-dependent systemic lupus
erythematosus: effect of conditioning regimen on outcome. Bone Marrow Transplant. 53,
692–700 (2018).
67. Stohl, W. et al. Belimumab reduces autoantibodies, normalizes low complement levels,
and reduces select B cell populations in patients with systemic lupus erythematosus.
Arthritis Rheum. 64, 2328–2337 (2012).

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

68. Ramskold, D. et al. B cell alterations during BAFF inhibition with belimumab in SLE.
EBioMedicine 40, 517–527 (2019).
69. Jacobi, A. M. et al. Effect of long-term belimumab treatment on B cells in systemic lupus
erythematosus: extension of a phase II, double-blind, placebo-controlled, dose-ranging
study. Arthritis Rheum. 62, 201–210 (2010).
70. Wang, J. et al. The rapid inhibition of B-cell activation markers by belimumab was associated
with disease control in systemic lupus erythematosus patients. Front. Pharmacol. 14,
1080730 (2023).
71. Furie, R. et al. A phase III, randomized, placebo-controlled study of belimumab,
a monoclonal antibody that inhibits B lymphocyte stimulator, in patients with systemic
lupus erythematosus. Arthritis Rheum. 63, 3918–3930 (2011).
72. Navarra, S. V. et al. Efficacy and safety of belimumab in patients with active systemic
lupus erythematosus: a randomised, placebo-controlled, phase 3 trial. Lancet 377,
721–731 (2011).
73. Arends, E. J. et al. Disruption of memory B-cell trafficking by belimumab in patients with
systemic lupus erythematosus. Rheumatology 63, 2387–2398 (2024).
74. Struemper, H. et al. Reductions in circulating B cell subsets and immunoglobulin G levels
with long-term belimumab treatment in patients with SLE. Lupus Sci. Med. https://doi.org/
10.1136/lupus-2021-000499 (2022).
75. Wilkinson, C. et al. The role of baseline BLyS levels and type 1 interferon-inducible gene
signature status in determining belimumab response in systemic lupus erythematosus:
a post hoc meta-analysis. Arthritis Res. Ther. 22, 102 (2020).
76. Merrill, J. T. et al. Efficacy and safety of atacicept in patients with systemic lupus
erythematosus: results of a twenty-four-week, multicenter, randomized, double-blind,
placebo-controlled, parallel-arm, phase IIb study. Arthritis Rheumatol. 70, 266–276 (2018).
77. Wu, D. et al. Telitacicept in patients with active systemic lupus erythematosus: results
of a phase 2b, randomised, double-blind, placebo-controlled trial. Ann. Rheum. Dis. 83,
475–487 (2024).
78. Mathur, M. et al. A phase 2 trial of sibeprenlimab in patients with IgA nephropathy.
N. Engl. J. Med. 390, 20–31 (2024).
79. Myette, J. R. et al. A proliferation inducing ligand (APRIL) targeted antibody is a safe and
effective treatment of murine IgA nephropathy. Kidney Int. 96, 104–116 (2019).
80. Mei, H. E. et al. Plasmablasts with a mucosal phenotype contribute to plasmacytosis in
systemic lupus erythematosus. Arthritis Rheumatol. 69, 2018–2028 (2017).
81. Sieger, N. et al. CD22 ligation inhibits downstream B cell receptor signaling and Ca2+ flux
upon activation. Arthritis Rheum. 65, 770–779 (2013).
82. Lumb, S. et al. Engagement of CD22 on B cells with the monoclonal antibody
epratuzumab stimulates the phosphorylation of upstream inhibitory signals of the B cell
receptor. J. Cell Commun. Signal. 10, 143–151 (2016).
83. Montalban, X. et al. Placebo-controlled trial of an oral BTK inhibitor in multiple sclerosis.
N. Engl. J. Med. 380, 2406–2417 (2019).
84. Dorner, T. et al. Efficacy and safety of remibrutinib, a selective potent oral BTK inhibitor, in
Sjögren’s syndrome: results from a randomised, double-blind, placebo-controlled phase 2
trial. Ann. Rheum. Dis. 83, 360–371 (2024).
85. Chan, P. et al. Population pharmacokinetics, efficacy exposure-response analysis,
and model-based meta-analysis of fenebrutinib in subjects with rheumatoid arthritis
[corrected]. Pharm. Res. 37, 25 (2020).
86. Grammer, A. C. et al. Abnormal germinal center reactions in systemic lupus
erythematosus demonstrated by blockade of CD154-CD40 interactions. J. Clin. Invest.
112, 1506–1520 (2003).
87. Stefanski, A. L. & Dorner, T. Immune checkpoints and the multiple faces of B cells in
systemic lupus erythematosus. Curr. Opin. Rheumatol. 33, 592–597 (2021).
88. Calabrese, L. H. & Molloy, E. S. Progressive multifocal leucoencephalopathy in the
rheumatic diseases: assessing the risks of biological immunosuppressive therapies.
Ann. Rheum. Dis. 67, iii64–ii65 (2008).
89. Wang, W. et al. BCMA-CD19 compound CAR T cells for systemic lupus erythematosus:
a phase 1 open-label clinical trial. Ann. Rheum. Dis. 83, 1304–1314 (2024).
90. Bucci, L. et al. Bispecific T cell engager therapy for refractory rheumatoid arthritis.
Nat. Med. 30, 1593–1601 (2024).
91. Subklewe, M. et al. Application of blinatumomab, a bispecific anti-CD3/CD19 T-cell engager,
in treating severe systemic sclerosis: a case study. Eur. J. Cancer 204, 114071 (2024).
92. Alexander, T. et al. Sustained responses after anti-CD38 treatment with daratumumab
in two patients with refractory systemic lupus erythematosus. Ann. Rheum. Dis. 82,
1497–1499 (2023).
93. Alexander, T. & Hiepe, F. Autologous haematopoietic stem cell transplantation for
systemic lupus erythematosus: time ready for a paradigm shift? Clin. Exp. Rheumatol. 35,
359–361 (2017).
94. Goklemez, S. et al. Long-term follow-up after lymphodepleting autologous haematopoietic
cell transplantation for treatment-resistant systemic lupus erythematosus. Rheumatology
61, 3317–3328 (2022).
95. Farge, D. et al. Autologous hematopoietic stem cell transplantation for autoimmune
diseases: an observational study on 12 years’ experience from the European Group
for Blood and Marrow Transplantation Working Party on Autoimmune Diseases.
Haematologica 95, 284–292 (2010).
96. Thiel, A. et al. Direct assessment of thymic reactivation after autologous stem cell
transplantation. Acta Haematol. 119, 22–27 (2008).
97. Baker, D., Herrod, S. S., Alvarez-Gonzalez, C., Giovannoni, G. & Schmierer, K. Interpreting
lymphocyte reconstitution data from the pivotal phase 3 trials of alemtuzumab.
JAMA Neurol. 74, 961–969 (2017).

781

Review article

98. Loh, Y. et al. Development of a secondary autoimmune disorder after hematopoietic
stem cell transplantation for autoimmune diseases: role of conditioning regimen used.
Blood 109, 2643–2648 (2007).
99. Furie, R. A. et al. B-cell depletion with obinutuzumab for the treatment of proliferative
lupus nephritis: a randomised, double-blind, placebo-controlled trial. Ann. Rheum. Dis.
81, 100–107 (2022).
100. Cambridge, G. et al. B cell depletion therapy in systemic lupus erythematosus:
relationships among serum B lymphocyte stimulator levels, autoantibody profile and
clinical response. Ann. Rheum. Dis. 67, 1011–1016 (2008).
101. Dorner, T. et al. Treatment of primary Sjögren’s syndrome with ianalumab
(VAY736) targeting B cells by BAFF receptor blockade coupled with enhanced,
antibody-dependent cellular cytotoxicity. Ann. Rheum. Dis. 78, 641–647 (2019).
102. Bowman, S. J. et al. Safety and efficacy of subcutaneous ianalumab (VAY736) in patients
with primary Sjögren’s syndrome: a randomised, double-blind, placebo-controlled,
phase 2b dose-finding trial. Lancet 399, 161–171 (2022).
103. Mei, H. E., Schmidt, S. & Dorner, T. Rationale of anti-CD19 immunotherapy: an option to
target autoreactive plasma cells in autoimmunity. Arthritis Res. Ther. 14, S1 (2012).
104. Mei, H. E. et al. A unique population of IgG-expressing plasma cells lacking CD19 is
enriched in human bone marrow. Blood 125, 1739–1748 (2015).
105. Deshmukh, A. et al. Preclinical evidence for the glucocorticoid-sparing potential of a
dual toll-like receptor 7/8 inhibitor in autoimmune diseases. J. Pharmacol. Exp. Ther. 388,
751–764 (2024).
106. Winkler, A. et al. The interleukin-1 receptor-associated kinase 4 inhibitor PF-06650833
blocks inflammation in preclinical models of rheumatic disease and in humans enrolled
in a randomized clinical trial. Arthritis Rheumatol. 73, 2206–2218 (2021).
107. Ramirez-Valle, F., Maranville, J. C., Roy, S. & Plenge, R. M. Sequential immunotherapy:
towards cures for autoimmunity. Nat. Rev. Drug Discov. 23, 501–524 (2024).
108. Hubbard, E. L. et al. Analysis of transcriptomic features reveals molecular endotypes of
SLE with clinical implications. Genome Med. 15, 84 (2023).
109. Le, T. A. et al. Efficient CRISPR-Cas9-mediated mutagenesis in primary human B cells for
identifying plasma cell regulators. Mol. Ther. Nucleic Acids 30, 621–632 (2022).
110. Neubert, K. et al. The proteasome inhibitor bortezomib depletes plasma cells
and protects mice with lupus-like disease from nephritis. Nat. Med. 14, 748–755
(2008).
111. Muchamuel, T. et al. Zetomipzomib (KZR-616) attenuates lupus in mice via modulation
of innate and adaptive immune responses. Front. Immunol. 14, 1043680 (2023).
112. McDonnell, S. R. P. et al. Mezagitamab in systemic lupus erythematosus: clinical and
mechanistic findings of CD38 inhibition in an autoimmune disease. Lupus Sci. Med.
https://doi.org/10.1136/lupus-2023-001112 (2024).

Nature Reviews Rheumatology | Volume 20 | December 2024 | 770–782

113. Qin, C. et al. Anti-BCMA CAR T-cell therapy CT103A in relapsed or refractory AQP4-IgG
seropositive neuromyelitis optica spectrum disorders: phase 1 trial interim results.
Signal. Transduct. Target. Ther. 8, 5 (2023).
114. Qin, C. et al. Single-cell analysis of refractory anti-SRP necrotizing myopathy treated with
anti-BCMA CAR-T cell therapy. Proc. Natl Acad. Sci. USA 121, e2315990121 (2024).
115. Woodruff, M. C. et al. Extrafollicular B cell responses correlate with neutralizing
antibodies and morbidity in COVID-19. Nat. Immunol. 21, 1506–1516 (2020).

Acknowledgements

The DRFZ is funded by the Leibniz Society and the Senate of Berlin. The authors thank
J. C. Ritter for his graphical support in the preparation of Fig. 1.

Author contributions

The authors contributed equally to all aspects of the article, including data research,
discussion of content and writing the article. All authors reviewed and edited the manuscript
before approval and submission.

Competing interests

T.D. declares honoraria for scientific advice from Abelzeta, BMS, Janssen, Novartis, Roche/GNE
and UCB, and fees for clinical studies (paid to the university) by BMS, Novartis, Eli Lilly & Company,
Janssen and Roche. P.E.L. is co-founder of AMPEL Biosolutions and an adviser to Abelzeta.

Additional information

Supplementary information The online version contains supplementary material available at
https://doi.org/10.1038/s41584-024-01179-5.
Peer review information Nature Reviews Rheumatology thanks Ioannis Parodis, Gregg Silverman
and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this
article under a publishing agreement with the author(s) or other rightsholder(s); author
self-archiving of the accepted manuscript version of this article is solely governed by the
terms of such publishing agreement and applicable law.
© Springer Nature Limited 2024

782