Pathogenesis of Sjögren's disease

INTRODUCTION — 

Sjögren's disease (SjD) is a chronic, autoimmune multisystem inflammatory disorder characterized by lacrimal and salivary gland dysfunction and resulting in the unique combination of dry eyes and dry mouth. Additional sicca manifestations may also be present, including dryness of the airways, skin, and vagina. Systemic extraglandular features include arthritis, nephritis, cytopenia, pneumonitis, and vasculitis. Neurologic manifestations include peripheral neuropathy, myelopathy, and cognitive disturbances. There is an increased risk of lymphoma in comparison with other autoimmune disorders. (See "Clinical manifestations of Sjögren's disease: Exocrine gland disease" and "Clinical manifestations of Sjögren's disease: Extraglandular disease".)

The pathogenesis of SjD is reviewed here. The clinical manifestations, diagnosis, treatment, and prognosis of this disorder are discussed separately. (See "Clinical manifestations of Sjögren's disease: Exocrine gland disease" and "Clinical manifestations of Sjögren's disease: Extraglandular disease" and "Diagnosis and classification of Sjögren's disease" and "Overview of the management and prognosis of Sjögren's disease" and "Treatment of dry eye in Sjögren's disease: General principles and initial therapy" and "Treatment of dry mouth and other non-ocular sicca symptoms in Sjögren's disease".)

OVERVIEW — 

The pathogenesis of SjD is typically modeled as a multistep process triggered by an environmental factor, most likely viral, in a genetically predisposed individual. The initial events engage the innate immune system, but propagation and perpetuation of the autoimmune process require a continual interplay between the innate and adaptive immune systems [1].

The result is autoreactive B-cell stimulation, autoantibody production, and chronic inflammation of the salivary and lacrimal glands and other tissues. Extraglandular manifestations may arise from autoimmune exocrinopathy akin to that in the salivary glands (eg, interstitial nephritis, biliary cholangitis), immune-complex deposition (eg, cryoglobulinemic vasculitis), and extranodal lymphoproliferation (eg, lymphocytic interstitial pneumonitis). Germinal center formation and chronic stimulation of B cells in the target tissue may promote lymphomagenesis, again through a multistep process in a genetically susceptible individual [2].

RISK FACTORS AND ETIOPATHOGENESIS — 

Interactions between genetic and nongenetic factors are involved in disease susceptibility and pathogenesis.

Genetic factors — Many different regions of the genome, both within and outside of the major histocompatibility complex (MHC), confer susceptibility to Sjögren's disease (SjD) but differ between populations and studies [3-5]. (See 'HLA genes' below and 'Non-HLA genes' below.)

A familial tendency to develop SjD has been well-documented, along with an increased risk of various autoimmune disorders in relatives of patients with SjD [5,6]. A concordance rate for SjD in monozygotic twins has not been reported; however, it is estimated to be approximately 20 percent based on studies of other autoimmune diseases that overlap with SjD, including systemic lupus erythematosus (SLE) and rheumatoid arthritis [5,7,8]. Thus, a substantial role for epigenetic factors and the environment is likely in SjD pathogenesis.

Multiple polymorphisms have been identified by genome-wide association studies (GWAS) and other methodologies [5]. The GWAS have involved cohorts of SjD patients of European [3,9] and Han Chinese [4] descent and another consisting of Sjögren's International Collaborative Clinical Alliance (SICCA) registrants, primarily of either European or Asian (including Chinese and Japanese) descent [10]. Many of the implicated genes are associated with innate or adaptive immune responses. Surprisingly, majority of the identified single nucleotide polymorphisms (SNPs) were located in the non-protein-coding regions, which suggests a regulatory activity of these regions on gene expression.

HLA genes — SjD shows the most robust genetic association within MHC genes, including those in the human leukocyte antigen (HLA)-DR region. The most widely reported associations have been with the DR2 and DR3 alleles at the DRB1 locus in Caucasian populations [11]. Considerable heterogeneity of this HLA association is observed across different ethnic groups [12]. A strong association of the HLA region with anti-SSA (Ro52 and/or Ro60) antibodies in SjD patients of European ancestry was also observed in a large GWAS [9].

While 20 to 25 percent of the general White European and American population shows the extended haplotype of HLA-DR3, B8, DQ2, and C4 null allele, this haplotype is present in approximately 50 percent of White Northern European patients with SjD [13]. Considerable heterogeneity of this association is observed across different ethnic groups [10,11,14-18]. For example, in SjD patients from Taiwan, HLA-DR8 was the highest risk factor and was associated with anti-Ro antibodies [19].

The genetic association of SjD with variants in the HLA region is restricted to SjD subgroup with anti-Ro/SSA antibodies and/or anti-SSB antibodies [9,12].

Non-HLA genes — Genes other than those within the HLA loci are also associated with an increased risk of disease [20]. The strongest associations on GWAS include IRF5 and TNIP1, both involved in innate immunity, and BLK, STAT4, IL12A, and CXCR5, which are involved in adaptive immunity (table 1). Among those of Han Chinese descent, GTF21, a gene that regulates immunoglobulin heavy chain transcription, and RBMS3, a proapoptotic and TGF-b downmodulating gene, were reported as risk factors for SjD [21]. Additional non-HLA loci that are associated with SjD at a genome-wide significance level include OAS1, NAB1, PTTG1-MIR146A, XKR6, MAPT-CRHR1, RPTOR-CHMP6-BAIAP2, TYK2 and SYNGR1, CD247, PRDM1-ATG5, and TNFAIP3 [9]. These genes have putative roles in toll-like receptor or interferon (IFN) signaling (eg, OAS1, TYK2, MAPT-CRHR1, NAB1), target tissue maintenance (eg, RPTOR-CHMP6-BAIAP2, XKR6), lymphocyte regulation (eg, CD247, PTTG1-MIR146A), and antigen presentation (eg, TNFAIP3, PRDM1-ATG5) [5].

MICA (MHC class I chain-related gene A) is a non-HLA gene encoded within the MHC complex. The MICA*008 allele was shown to be significantly associated with SjD in two independent cohorts from France and the United Kingdom [22].

The information on different gene-disease associations is curated by DisGeNET [23] and freely accessible for academic and not-for-profit researchers [24].

Epigenetic factors — Epigenetic factors, such as DNA methylation, histone acetylation, noncoding RNA (nc-RNA) transcripts, and gene recombination, may all play a role in the modulation of gene expression without affecting the actual DNA sequence. DNA methylation has been analyzed in the peripheral blood and target tissue of patients with SjD [25-30]. These studies have supported the role of methylation in regulating genes in SjD. In two large epigenome-wide association studies, type I IFN-regulated genes were the site of the most differentially methylated and robustly associated positions and regions [27,31]. Interestingly, many epigenetically modified regions have been previously identified as genetic risk factors for SjD [30,32,33]. Additional gene regulation in SjD occurs through the modulation of gene expression by microRNAs (miRNAs). Differential expression of miRNAs has been reported in the lymphocytes and salivary glands of SjD patients and controls, and these often target genes relevant to disease pathogenesis [33-36]. In summary, these studies demonstrate that complex interactions between genetic and epigenetic factors influence the development of SjD.

Sex — The striking predominance of SjD in females (10 to 15:1 female to male ratio) points to a role for sex hormones in the development of the disease, as it does for other systemic autoimmune diseases, such as SLE, rheumatoid arthritis, and systemic sclerosis [37,38]. The clinical onset of SjD in females is most often in the sixth decade of life, just after the onset of menopause. This is in sharp contrast to SLE, where the disease occurs most often in females during their reproductive-age years.

The characteristic autoantibodies of SjD are detectable for up to 18 years before clinical disease onset [39], so the autoimmune process may be initiated during reproductive-age years but not expressed clinically with sicca or other symptoms until estrogen levels drop sharply with the onset of menopause. In animal models, estrogen protects against lacrimal and salivary glandular inflammation while its withdrawal promotes apoptosis of salivary gland epithelial cells [40]. Females with SjD have lower cumulative estrogen exposure relative to non-SS females with sicca when measured by integration of age at menarche, menopause, parity, hysterectomy, and female hormone use [41]. Low serum concentrations of dehydroepiandrosterone and dihydrotestosterone have also been demonstrated [42], arguing for a protective role of androgens, similar to that of estrogen, in the pathogenesis of SjD.

Alternatively, the sex bias of SjD may be mediated independently of a sex hormone effect. Females with trisomy X (47, XXX) have normal sexual development and sex hormone levels, but they have a risk of SLE and SjD that is 2.5- and 2.9-fold higher, respectively, than in females with 46, XX, and 25- and 41-fold higher respectively than in males with 46, XY [43]. Thus, there is a sex chromosome dose effect in the predisposition of females to autoimmune rheumatic disease, which is independent of differences in sex hormone levels.

The X chromosome has the highest density of immunity-related genes, and transcriptional silencing of an entire X chromosome (X-chromosome inactivation) in each female cell, initiated during embryogenesis, equalizes the expression of X-linked genes between sexes. This inactivation process is mediated by allele-specific expression of the long noncoding RNA, termed XIST. Biallelic expression of X-linked immunity genes has been observed in the lymphocytes of female mammals, suggesting incomplete X-chromosome inactivation and providing an alternative mechanism for the sex bias in autoimmune diseases, including SjD [44,45]. XIST may act as a TLR7 ligand and scaffold for ribonucleoproteins and in so doing drive autoimmunity and its female predisposition [46-48].

Although the frequency of SjD development in males is significantly lower, some studies have suggested that male SjD patients have an increased risk of lymphoma development [49,50] and extraglandular manifestation [51], and several studies have reported a higher prevalence of interstitial lung disease in males [52]. Although precise mechanisms driving male resistance to SjD are not clear, robust regulation of autoimmunity might be in play. This point is highlighted by the observation that more males develop SjD following immune checkpoint inhibitor therapies targeting the programmed cell death protein 1 (PD-1)/PD-L1 pathway [53,54].

The potential role of viral infection — The first signs of SjD typically occur long before diagnosis, thereby impeding the study of its etiology. Many observations suggest a role for viruses in the pathogenesis of SjD, but no single virus has been implicated [55]. Evidence of ongoing or past viral infection can be detected in many patients, but no virus has been found at high levels in target tissues [56,57]. Observations supporting a viral etiology include:

●Certain viruses, particularly Epstein-Barr virus (EBV), the ubiquitous herpes-type virus that causes infectious mononucleosis, frequently infect the salivary glands. EBV is spread to noninfected individuals via the saliva; primary EBV infections progress to lifelong latent infection with periodic reactivation, and the site of latency for EBV is in the salivary gland. EBV can induce strong immune responses by T cells and activate B-cell production of autoantibodies. (See "Virology of Epstein-Barr virus" and "Clinical manifestations and treatment of Epstein-Barr virus infection" and "Infectious mononucleosis".)

●EBV can be identified in the ectopic lymphoid structures present in the salivary glands of some SjD patients (see 'The lymphocytic infiltrate and glandular pathology' below), but not in SjD glandular tissue lacking such structures. EBV-infected plasma cells within these structures produced antibodies to anti-Ro52 and anti-La/SSB [58]. In addition, SjD patients show a higher prevalence and titer of antibodies against EBV antigens [59]. In an analysis of a previously healthy female with primary EBV infection, anti-Ro52 and anti-Ro60 autoantibodies were detected seven days after primary infection and underwent immunoglobulin M (IgM) to IgG switching, suggesting that the EBV infection induced their formation [60]. These findings support the role of active EBV in supporting the local proliferation and differentiation of autoreactive B cells.

●At least three viruses (human T-lymphotropic virus [HTLV] type I, human immunodeficiency virus [HIV], and hepatitis C virus [HCV]) are associated with clinical syndromes that share many features of SjD [61,62]. Minor salivary gland biopsies from patients recovering from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) commonly have focal lymphocytic infiltrates similar to those seen in SjD [63].

●Hepatitis delta virus (HDV) was detected in the salivary glands of 50 percent of primary SjD patients and induces a primary SjD-like disease in mice [64,65].

●Retroelements are noncoding DNAs that constitute approximately half of the human genome and regulate gene expression. SjD patients show increased levels of retroviral long-interspersed nuclear element 1 [66]. These retroviral elements can activate innate immunity and induce excessive production of type I IFNs. Animal models have demonstrated a role for type I IFNs in SjD pathogenesis [67].

MECHANISMS OF IMMUNE-MEDIATED INJURY — 

Glandular dysfunction in Sjögren's disease (SjD) is generally presumed to result from autoimmune-induced inflammation and resultant damage and destruction of the tissue responsible for tear and saliva production [68]. Transcriptomic analyses of salivary gland tissue from SjD patients have demonstrated differentially expressed genes only in salivary gland tissue with inflammatory infiltrates [69]. However, the correlation between glandular dysfunction and the degree of glandular inflammation is weak [70]. In addition, glandular dysfunction can be induced in animal models of SjD before the appearance of inflammatory infiltrates [71]. In light of these observations, other mechanisms may contribute to the glandular dysfunction, including antibodies to the muscarinic receptor that may impair neural innervation of the gland (see 'Anti-muscarinic acetylcholine receptor antibodies' below) and direct effects of cytokines on neurotransmitter release or other secretory cell functions [72].

Glandular inflammation — Overexpression of interferon (IFN)-inducible genes in salivary glands and peripheral blood monocytes of SjD patients (termed the IFN signature) highlights the importance of the innate immune system in pathogenesis [73,74]. Damage to the salivary gland epithelial cells (eg, by an exogenous or endogenous viral trigger) is thought to induce apoptosis and migration of the SSA (also termed SS-A or Ro60) antigen from the nucleus in a complex with human small noncoding Y RNA (hYRNA) to a bleb on the cell surface [75]. The SSA molecule thus escapes normal apoptotic degradation. The presence of antibodies to the SSA-hYRNA complex promotes the uptake of this immune complex by local dendritic cells and B cells, access to intracytoplasmic toll-like receptors (TLRs), and stimulation of the IFN signatures characteristic of SjD [76-78].

Plasmacytoid dendritic cells, the main producers of type I IFN, may also be activated directly by viral or other environmental factors. Activation of the type I IFN system by the innate immune system promotes adaptive immune responses through T- and B-cell activation and induction of cytokine production.

A cycle of mutual stimulation of the innate and acquired immune systems leads to the perpetuation of glandular injury and dysfunction. Tissue injury occurs through the activation of these immune pathways by lymphocytes within the glandular tissues or extraglandular sites, leading to the release of cytokines, including IFN-gamma, interleukin 17 (IL-17), B-cell activating factor (BAFF; also known as B lymphocyte stimulator [BLyS]), and others, and the production of characteristic autoantibodies. Apoptosis of glandular cells and dysfunction of residual epithelial cells and tissues occur due to cytokines and metalloproteinases that interfere with salivary gland organization and function. Furthermore, premature senescence in salivary gland progenitor cells [79] and salivary gland stem cells [80] also contributes to the chronicity of glandular dysfunction. (See 'Autoantibodies' below and 'The lymphocytic infiltrate and glandular pathology' below.)

Transcriptomic analyses of peripheral blood and salivary gland tissue in large cohorts of SjD patients have defined disease subsets. The IFN gene signature is present in approximately 60 to 75 percent. It marks a subgroup that has a higher prevalence (or titer) of anti-Ro/SSA antibodies, hyperglobulinemia, leukopenia, greater glandular inflammation, and worse exocrine function [74,81-83]. The gene expression profiles allow hierarchical clustering based on not only the IFN signature but also pathways related to inflammation, cytokines/chemokines, hematopoiesis, and lymphocytes.

Considering the key role of type I IFNs in SjD pathogenesis, several clinical trials are exploring targeting the type I IFN pathway as a therapeutic strategy for SjD [1].

Extraglandular manifestations — Extraglandular manifestations of SjD can be classified based on their presumed pathogenesis, including autoimmune exocrinopathy, akin to that in the salivary glands (eg, interstitial nephritis, biliary cholangitis), immune-complex deposition (eg, cryoglobulinemic vasculitis), cell- or tissue-specific autoimmunity (eg, thrombocytopenia, ataxic sensory ganglionopathy, neuromyelitis optica), and lymphoproliferation (eg, lymphocytic interstitial pneumonitis) [84,85].

The pathogenesis of fatigue is most likely multifactorial [86]. An inverse correlation between fatigue and the level of proinflammatory cytokines has been observed [87,88]. These include IFN-gamma, tumor necrosis factor alpha (TNF-alpha), lymphotoxin alpha, and IFN-gamma inducible protein [88]. On the other hand, proteomic analyses of serum and blood in SjD patients with high versus low fatigue have identified distinct proteins upregulated in those with fatigue, including some involved in inflammation [89,90]. In a randomized phase II clinical trial of an RNase Fc fusion protein in primary SjD, there was an improvement in severe fatigue in the treatment arm compared with the placebo. This improvement was significantly correlated with increased expression of selected IFN-inducible genes [91].

Cryoglobulinemic vasculitis occurs in approximately 10 percent of SjD patients and is a strong risk factor for lymphoma development [92,93]. In the usual context of SjD-related cryoglobulinemia, a monoclonal pentameric immunoglobulin M (IgM) rheumatoid factor binds polyclonal IgG molecules, generating a macromolecular complex that forms insoluble aggregates with cooling below 30˚C. These aggregates are reversible with warming, both in vivo and in vitro, and thus explain cold-induced acrocyanosis in patients and the formation of cryoprecipitates in serum collected and separated at body temperature but then kept overnight in a refrigerator. The cryoprecipitable IgM-IgG immune complexes can lead to a small vessel vascular injury either through deposition with complement activation or the formation of insoluble aggregates [94]. The B-cell clones that make these pathogenic IgM rheumatoid factors are distinguished by a high frequency of somatic mutations in lymphoma and/or leukemia driver genes and in genes regulating V(D)J recombination [95,96]. The former lead to accelerated B-cell proliferation, and the latter account for the ability of the IgM antibodies to form insoluble aggregates in the cold. These observations highlight a shared pathogenetic mechanism between lymphoid malignancy and benign autoimmunity.

Autoantibodies — SjD is characterized by the presence of autoantibodies, most notably anti-Ro/SSA and anti-La/SSB in 60 to 80 percent of those affected (see "The anti-Ro/SSA and anti-La/SSB antigen-antibody systems"). Antinuclear antibodies (ANA) are present in 90 percent of patients, and high-titer rheumatoid factor is also frequent. (See 'Ro/SSA and La/SSB' below and "Diagnosis and classification of Sjögren's disease" and "Clinical manifestations of Sjögren's disease: Extraglandular disease", section on 'Autoantibodies'.)

Autoantibodies can precede the clinical onset of SjD by many years, as evidenced by the analysis of stored premorbid serum samples in SjD patients [39]. Additionally, mothers of children with neonatal lupus have high-titer anti-Ro/SSA and anti-La/SSB antibodies and are often asymptomatic. However, these asymptomatic women are at an increased risk for SjD. The fact that SjD develops in only a minority of these individuals indicates that autoantibodies alone are insufficient for the induction of disease [97].

Ro/SSA and La/SSB — Anti-Ro/SSA antibodies are predominantly of the IgG1 subclass and recognize two distinct proteins, the 52kD, Ro52 protein and the 60kD, Ro60 protein, encoded by different genes [98]. Both antigens are located primarily in the nucleus but are also expressed in the cytoplasm and on the cell surface.

The SSA Ro52 autoantigen belongs to the large family of the tripartite motif (TRIM)-containing family of proteins [99] and is also known as TRIM21. Functionally, Ro52 is an E3 ubiquitin ligase, and it plays a critical role in regulating innate immunity, particularly the type I IFN response [100]. Ro52 also acts as an intracellular Fc receptor and has been shown to bind the Fc portion of IgG antibodies complexed with viruses that have entered the host cell [101]. A pathogenic role of anti-Ro52 antibodies in the induction of salivary gland dysfunction has been demonstrated in experimental mouse model systems [102,103].

The 60kD Ro60 protein, also known as TROVE2, binds the small cytoplasmic RNA moieties termed hYRNA and is involved in the clearance of defective RNA transcripts [104]. The Ro60 protein, like Ro52, is also involved in regulating inflammatory gene expression by binding to endogenous Alu retroelements [105].

SjD patients have different anti-Ro/SSA antibody serologic profiles. The majority has both anti-Ro60 and anti-Ro52 antibodies and, in several studies, higher disease severity than those with either isolated anti-Ro60 or anti-Ro52 antibodies [106-108]. Anti-Ro52 antibody alone is the least specific of the anti-Ro/SSA antibody serologic profiles, being found in patients with myositis, interstitial lung disease, autoimmune hepatitis, and malignancy [108-111]. In one study, isolated anti-Ro52 antibody identified a subset of SjD with more frequent vasculitis, lung involvement, and cryoglobulinemia [112].

Anti-La/SSB antibodies are found in 50 percent of patients with SjD. These antibodies recognize a 47 kD phosphoprotein associated with all newly synthesized RNA polymerase III transcripts. Binding of the La/SSB protein to these small RNAs protects them from exonuclease digestion [113]. The gene encoding La/SSB is unusual because it has two promoter sites, encoding for two different size mRNAs, and raising the possibility of gene switching under disease conditions [114]. The pathogenic role of anti-La/SSB autoantibodies in SjD is not clear.

SjD patients in whom anti-SSA autoantibodies are not detected differ substantially from those with these antibodies, having a minimal IFN gene signature and no significant association with HLA alleles [9,12,74]. These findings point to a different pathogenesis. At a clinical level, these patients have clinical and laboratory features distinct from the anti-SSA positive group, with lower frequencies of leucopenia, rheumatoid factor, hyperglobulinemia, and low C4 levels [115] but higher frequency of arthritis [116], sensory small fiber neuropathy [117], and greater pain severity [118].

Anti-muscarinic acetylcholine receptor antibodies — Antibodies to acetylcholine receptors of salivary glands might account for decreased secretion from histologically normal glands. It is uncertain whether such antibodies in SjD are primary or secondary phenomena [119-121]. An increased prevalence of these antibodies in SjD is not a universal finding and is dependent in part on the assay methodology [122]. In a large study of 361 Korean subjects that used a whole cell assay to detect antibody binding to a conformational epitope of the muscarinic-type-3 (M3R), SjD patients could be reliably distinguished from healthy controls and patients with non-SjD sicca or rheumatoid arthritis [123]. However, there was less discrimination between SjD patients and those with systemic lupus erythematosus (SLE). Among the 156 primary SjD patients, levels of anti-M3R antibodies correlated positively with ocular staining scores and negatively with salivary function measured by sialometry and scintigraphy. In a study assessing antibody binding to linear peptides corresponding to five different extracellular loops of M3R, there was no difference between healthy controls and 29 connective tissue disease patients with anti-Ro/SSA reactivity, including 19 with SjD [122]. In animal models, these anti-M3 acetylcholine receptor antibodies decrease glandular secretion in some but not all studies [120,124].

Other specificities — Several studies over the years have reported autoantibodies targeting many nuclear, cytoplasmic, membrane, and secreted proteins in SjD [125,126]. While some of these autoantibodies are associated with higher disease severity, the pathogenic mechanisms remain unclear [127]. Surprisingly, many of these autoantibodies are not clinically relevant and might indicate a hyperactive B-cell response and abrogation of tolerance to these proteins in SjD patients.

The lymphocytic infiltrate and glandular pathology — The primary pathologic lesion of SjD is lymphocytic infiltration of the salivary and lacrimal glands. The infiltrates consist of focal aggregates of lymphocytes, centered around the ducts (image 1). The cellular composition of these infiltrates depends on their severity. T cells, primarily CD4+, predominate in milder infiltrates, which are smaller and respect the architecture of the gland. B cells become more predominant in larger and denser infiltrates associated with acinar destruction and loss of tissue architecture [128]. Migration of the lymphocytes to sites in the glands occurs due to a series of events, including a response to chemokines, adhesion to specific vascular adhesion molecules, and entry into the gland where they interact with dendritic cells and epithelial cells [129].

B lymphocytes within the periductal infiltrates of the glands may invade the epithelium of striated and excretory ducts, leading to ductal epithelial hyperplasia and eventual loss of the ductal lumen (picture 1) [130]. This process defines a lymphoepithelial lesion and is highly specific for SjD. It can be identified in approximately 50 percent of major and minor salivary gland biopsies. Lymphoepithelial lesions are a diagnostic feature of mucosa-associated lymphoid tissue (MALT) lymphoma. The intraepithelial B cells within these lesions express Fc receptor-like protein 4 (FcRL4), a surface antigen that nearly all MALT lymphomas express. Thus, these cells may constitute a precursor pool for MALT lymphoma B cells. (See "Clinical manifestations, pathologic features, and diagnosis of extranodal marginal zone lymphoma of mucosa associated lymphoid tissue (MALT)", section on 'Clinical features'.).  

B cells expressing myeloid cell markers such as CB11b and CD11c have been associated with aging and autoimmunity and are known as age-associated B cells or atypical B cells [131]. The conditions required for the activation and proliferation of these cells are notably different than those in conventional B cells [132]. Notably, they are also expanded in human SjD [133-135] and animal models of SjD [136] and are known to generate autoantibodies. It should also be noted that these cells are refractory to B-cell depletion therapies [137], which might complicate therapies aiming to ablate B cells in SjD patients.

Salivary gland epithelial cells can play an active role in initiating and maintaining glandular inflammation [68]. Their potential pivotal role in SjD immunopathogenesis has led to the concept of "autoimmune epithelitis" [138] as a unifying feature of the disease. Salivary gland epithelial cells, activated by type I IFN or a viral infection, can affect the following: surface expression of major histocompatibility complex (MHC) class II molecules, including human leukocyte antigen (HLA)-DR, and costimulatory factors, including CD80, CD86, and CD40, empowering them to interact with T cells; release of cytokines such as BAFF [139,140], interleukin 1 (IL-1), IL-6, tumor necrosis factor alpha (TNF-alpha), and IL-22 [141], which are crucial to both innate and adaptive immune responses; promotion of lymphocytic and dendritic cell infiltration by the production of CXCL13 and other chemokines [142]; and mediation of the release of intracellular antigens (eg, Ro/SSA-La/SSB) through apoptosis [143] and the release of exosomes, thereby driving the generation of autoreactive B cells.

In approximately 30 to 40 percent of SjD patients, the glandular periductal lymphoid aggregates develop a structure highly similar to typical secondary lymphoid organs, with B-cell follicles surrounded by T-cell rich areas, high endothelial venules, and networks of follicular dendritic cells [144,145]. These ectopic tertiary lymphoid structures promote antigen-driven selection of B-cell clones via affinity maturation and provide a conducive microenvironment for antibody production in the target tissue (picture 2) [146]. The development of these ectopic lymphoid structures is dependent on the expression of lymphotoxin-beta and lymphoid chemokines (eg, CXCL13, CCL19, and CCL21) by T, B, dendritic, and stromal cells within the infiltrate [144,147,148].

Pathogenetic roles have also been attributed to T helper 1 (Th1) cells, natural killer (NK)-like cells, and Th17 cells that produce IFN-gamma. Considerable attention has been given to the role of a subset of CD4+ T cells, termed the follicular helper cells (Tfh), in the pathogenesis of SjD [149]. The Tfh cells are characterized by the expression of CXCR5, PD1, ICOS, and Bcl-6 [150]. These cells are the major producer of the cytokine IL-21, provide help to B cells, and are involved in forming germinal centers. SjD patients have elevated levels of circulating Tfh, and these cells are readily detected in salivary gland biopsies [149].

Cytokines in the glandular tissue — Multiple cytokines have been identified in the salivary gland tissue and represent potential targets for therapy.

The cytokine milieu of salivary glands from SjD patients is primarily characterized by a Th1/Th17 profile, with the production of proinflammatory IL-2, IL-10, and IFN-gamma by infiltrating CD4+ cells and IL-17 by infiltrating Th17 cells. In addition, the pro-inflammatory cytokines, IL-1, TNF-alpha, and IL-6, can be secreted by activated salivary gland epithelial cells [151,152].

BAFF is considered a key cytokine in SjD; it is induced by type I and type II IFN and promotes the activation, proliferation, and survival of B cells. BAFF levels are elevated in the salivary glands and serum of SjD patients, and the latter levels correlate with those of anti-SSA/Ro antibodies, anti-SSB/La antibodies, and rheumatoid factor [153]. BAFF is produced by monocytes, macrophages, dendritic cells, salivary gland epithelial cells, and B and T cells. Thus, BAFF could be an essential link between the activation of the innate immune system and the development of autoimmunity through the adaptive immune system.

Lymphomagenesis — Chronic B-cell stimulation and other factors may result, through a series of steps, in malignant B-cell transformation in some patients with SjD. Patients with SjD have an increased risk of lymphoma, with estimates ranging from 5- to 44-fold, compared with age-matched controls [154]. This increased risk is higher than that observed in other systemic autoimmune diseases, including rheumatoid arthritis, SLE, and Crohn disease [155]. These lymphomas are most frequently extranodal marginal zone non-Hodgkin lymphomas (NHL) of "mucosal-associated lymphoid tissue" (MALT). Higher-grade diffuse B-cell lymphomas and T-cell lymphomas are much less frequent in SjD [156,157]. The MALT lymphomas often develop in mucosal locations where SjD is active, such as salivary glands or the gastrointestinal tract (MALT) [157]; or in the lung, where bronchial-associated lymphoid tissue (BALT) lymphomas can be seen [158,159]. (See "Clinical manifestations, pathologic features, and diagnosis of extranodal marginal zone lymphoma of mucosa associated lymphoid tissue (MALT)".)

The presence of ectopic lymphoid structures in minor salivary gland biopsies is associated with an increased risk of lymphoma [160,161]. Such structures were much more common in biopsies from SjD patients who later presented with an NHL compared with patients without NHL (86 versus 22 percent) [160]. Ectopic lymphoid structures are sites of antigen-driven B-cell stimulation and clonal expansion, Ig class switching, and somatic hypermutation, potentially engendering lymphoma development.

Chronic stimulation of autoimmune B cells may be associated with malignant transformation, through a series of steps involving the development of a clonal population and eventually uncontrolled clonal proliferation. Salivary gland MALT lymphomas frequently express B-cell antigen receptors with rheumatoid factor activity and bind IgG with high affinity [162,163]. Locally produced IgG autoantibodies directed against the ribonucleoproteins SSA/Ro52, SSA/Ro60, and SSB/La form immune complexes particularly suited for dual-ligand stimulation of B cells with rheumatoid factor B-cell and TLR-7 receptors [164]. The proliferation of these autoreactive B-cells is partly driven by BAFF, serum levels of which correlate with disease activity and the degree of B-cell activation.

Malignant transformation has been associated with specific genetic polymorphisms. A20 (encoded by gene TNFAIP3) is a regulator of NF-kappaB activation and is downregulated in SjD [20]. Further, a polymorphism of this gene has been found in a high percentage of SjD patients. Mutations and downregulation of A20 have been associated with increased germinal center (GC) formation and MALT lymphomas [165]. Polymorphisms of CXCR5, involved in the organization of GC structures, are associated with SjD and NHL [166].

INFORMATION FOR PATIENTS — 

UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Beyond the Basics topics (see "Patient education: Sjögren's disease (Beyond the Basics)")

SUMMARY

●Risk factors – Both genetic and nongenetic factors are involved in disease susceptibility and the disease process. Molecular genetic analyses suggest essential roles for human leukocyte antigen (HLA)-DR molecules and genes encoding innate and acquired immunity elements. The strongest associations are with genes in the HLA-DR region, but considerable heterogeneity exists across different ethnic groups. Epigenetic factors may play a role in the modulation of gene expression. (See 'Risk factors and etiopathogenesis' above and 'Genetic factors' above and 'HLA genes' above and 'Non-HLA genes' above.)

●Pathogenesis – A cycle of mutual stimulation of the innate and acquired immune systems perpetuates glandular injury and dysfunction. Tissue injury occurs through the activation of these immune pathways by lymphocytes within the glandular tissues or extraglandular sites, leading to the release of cytokines, including interferon (IFN)-gamma, interleukin 17 (IL-17), IL-21, and B-cell activating factor. Salivary and lacrimal gland epithelial cells and the local vascular adhesive molecules play essential early roles. (See 'Mechanisms of immune-mediated injury' above.)

●Potential role of viral infection – Many observations suggest a role for viruses in the pathogenesis of Sjögren's disease (SjD), but no single virus has been implicated. Evidence of ongoing or past viral infection can be detected in many patients, but no virus has been found at high levels in target tissues. An explanation might be that the viral trigger occurred years before the development of SjD. (See 'The potential role of viral infection' above.)

●Autoantibodies – SjD is characterized by the presence of specific autoantibodies, which by some criteria are required for the diagnosis; these include anti-Ro/SSA and anti-La/SSB antibodies. Antinuclear antibodies (ANA) are present in 90 percent of patients, and high-titer rheumatoid factor is also frequent. Other autoantibodies may also be seen, including antibodies to acetylcholine receptors of salivary glands. (See 'Autoantibodies' above and 'Ro/SSA and La/SSB' above.)

●Pathology – The principal pathologic lesion of SjD is a lymphocytic infiltrate. The salivary and lacrimal glands are the most frequently affected tissues, but these infiltrates are common to all affected organs, including extraglandular sites, and can result in glandular and systemic extraglandular features. The infiltrates consist of focal aggregates of lymphocytes, beginning around the ducts and spreading to involve the entire lobule. Migration of the lymphocytes occurs to the gland in response to chemokines, adhesion to specific vascular adhesion molecules, and entry into the glandular cells, where they interact with dendritic cells and epithelial cells. (See 'The lymphocytic infiltrate and glandular pathology' above.)

ACKNOWLEDGMENT — 

The UpToDate editorial staff acknowledges Paul Creamer, MD, who contributed to an earlier version of this topic review.