Treatment of primary Sjögren syndrome

Sjögren syndrome is a progressive autoimmune disease. The hallmark manifestation of Sjögren syndrome is sicca, which is caused by lymphocytic infiltrates that eventually destroy the lachrymal and salivary gland tissue¹. Parotid gland enlargement, although rare, is suggestive of Sjögren syndrome when bilateral or accompanied by other signs².

Sjögren syndrome can also cause systemic manifestations, which mainly involve the musculoskeletal system, nervous system, lungs, kidneys, skin and blood vessels^1,3. Severe quality-of-life impairments are seen consistently in patients with Sjögren syndrome, even in those patients who have no systemic manifestations. Primary SS (pSS) occurs alone, whereas secondary Sjögren syndrome occurs in combination with another autoimmune disorder. As a rule, secondary Sjögren syndrome is not very severe, and the accompanying disease is the main determinant of treatment decisions. Consequently, this Review focuses on pSS, although sicca is treated identically in both types of Sjögren syndrome.

At present, pSS is diagnosed using the 2002 ACR–EULAR criteria⁴ or 2012 ACR criteria⁵. The existence of two different criteria sets can raise difficulties, and an ACR–EULAR expert panel recently came to a consensus on a new criteria set. These new ACR–EULAR classification criteria, although still unpublished, were reported by Shiboski et al.⁶ at the International Symposium on Sjögren's Syndrome in Bergen in May 2015. They are intended to be used in patients with signs suggesting pSS, mainly sicca. Points are given based on the focus score , positive serology for antibodies to Sjögren syndrome-related antigen A (SSA; also known as Ro), ocular staining score (OSS) for one eye ≥5 (or van Bistjerveld score ≥4), Schirmer's test result ≤5 mm per 5 min and unstimulated whole salivary flow rate ≤0.1 ml per min. Nevertheless, when seeking to diagnose pSS, differential diagnosis is as important as identifying these disease criteria, which are classification and not diagnostic criteria².

The EULAR task force on pSS has created two new tools for assessing disease activity and patient-reported outcomes. The EULAR Sjögren's syndrome disease activity index (ESSDAI)⁷ is the sum of scores for each organ-specific domain (constitutional, lymphadenopathy, glandular, articular, cutaneous, respiratory, renal, muscular, peripheral nervous system, central nervous system (CNS), haematological and biological) multiplied by the activity level (range 0–3). The minimal clinically significant difference for the ESSDAI is 3 points⁸. The EULAR Sjögren's syndrome patient-reported index (ESSPRI)⁹ combines visual analogue scales (VASs) for sicca, pain and fatigue. The minimal clinically significant difference is 1 point⁸.

The ESSDAI and ESSPRI are increasingly being used as inclusion criteria and endpoints in clinical trials evaluating therapies for pSS. Predictors of extraglandular manifestations of pSS, including lymphoma development, are also considered as inclusion criteria. The main known predictors of extraglandular manifestations are parotidomegaly, purpura, polyadenopathy, vasculitis, anaemia, lymphopenia, hypocomplementaemia (low serum levels of complement protein C4), cryoglobulinaemia, elevated serum levels of Fms-related tyrosine kinsase 3 ligand (FLT3L), presence of germinal centres within the salivary glands, and the His159Tyr mutation of the B-cell-activating factor receptor (BAFFR; also known as TNFRSF13C)^{10,11,12,13,14}.

The objective of this Review is to summarize the available data on topical and systemic medications for pSS according to clinical signs and disease activity, and also to summarize previous and ongoing studies using biologic agents in pSS.

Experts suggest that patients with Sjögren syndrome should be managed by a multidisciplinary team including at least a rheumatologist, a dentist and an ophthalmologist¹⁵. The few available systematic literature reviews^16,17,18 have highlighted the paucity of rigorous clinical trials on which to base clear guidelines, which explains the current reliance on expert opinion^{15,16,17,18,19}. Tables 1, 2 summarize the principles of the treatment of pSS and the grade of recommendation according to the Oxford Centre for Evidence-Based Medicine Levels of Evidence (see Further information).

Conclusive evidence of efficacy exists only for pilocarpine and cevimeline to treat sicca, and topical ciclosporin to treat moderate or severe ocular dryness (Table 3). Few published randomized controlled trials (RCTs) have focused on extraglandular signs (Table 3).

Whereas sicca symptoms can be treated using topical medications, other manifestations of pSS require systemic treatment. In this section, we discuss systemic treatments for non-life-threatening manifestations according to clinical signs and disease activity. Life-threatening manifestations are discussed separately below.

Life-threatening manifestations are rare in pSS and usually related to cryoglobulinaemic vasculitis or lymphoma¹²⁴, as well as CNS involvement and ganglionopathies. Few studies have evaluated the efficacy of treatments for these manifestations in the context of pSS. Current practice relies mainly on case reports, data from similar diseases (vasculitis and SLE) and expert opinion. The treatments are organ-specific and consist mainly of immunosuppressant medications and high-dose glucocorticoids. In severe vasculitis or CNS involvement, methylprednisolone and cyclophosphamide pulses should be used, combined with intravenous immunoglobulins or plasma exchange in the most severe cases.

Lymphoma is one of the most serious complications of pSS, and its diagnosis is difficult. Lymphoma develops predominantly in the major salivary glands, but it can also arise in other mucosal tissues (for example, in lymphoid organs, the thyroid gland and the stomach). The standardized incidence ratio of lymphoma in patients with pSS is 4.9–9 in the most recent studies^125,126. In pSS, low-grade extranodal marginal zone B-cell non-Hodgkin lymphoma (also known as mucosa-associated lymphoid tissue (MALT) lymphoma) is the most common variant. Ultrasonography of the salivary glands can be helpful to guide a biopsy and to evaluate the size of the gland and location of the nodes. PET-CT should be used as a means of detecting disseminated lymphoma foci.

Both diagnosis of MALT lymphoma due to pSS and evaluation of the risk of its development are difficult, and are based on expert opinion and the exclusion of other aetiologies. The treatment of MALT lymphoma is difficult to standardize because this tumour is often indolent and the prognosis often good. Available strategies include surgery, radiotherapy, and chemotherapy and rituximab (alone or in combination). Close monitoring without treatment might deserve consideration in some cases¹²⁷.

When treatment is required in patients with pSS and MALT lymphoma, it should be tailored to each individual patient and based on the size of the involved gland, the number and location of the tumours, lymphadenopathy, international prognostic index, ESSDAI, and Helicobacter pylori status¹²⁸, although H. pylori eradication has not been evaluated in clinical trials as a therapeutic approach to pSS-related lymphoma.

Despite the fact that not enough data support its use, rituximab is sometimes prescribed for MALT lymphoma. The treatment of salivary gland MALT lymphoma relies on rituximab combined with alkylating agents (for example, bendamustine or fludarabine with or without cyclophosphamide), but some patients have been managed with rituximab alone. In an RCT in patients with indolent lymphoma, including 67 with non-gastric MALT lymphoma, the combination of rituximab plus bendamustine was superior to the combination of rituximab and CHOP (cyclophosphamide, doxorubicin, vincristine and prednisone) chemotherapy in terms of patient survival and toxicity¹²⁹.

Other lymphoma subtypes such as lymphoplasmacytoid and diffuse large B-cell lymphoma also occur at higher rates in patients with pSS¹²⁶. They are believed to arise from the transformation of previous low-grade lymphomas in about 10% of cases¹²⁶. The treatment of these aggressive lymphomas should be adapted to the histological grade and relies mainly on rituximab combined with the CHOP regimen¹³⁰.

As discussed above, the role for biologic agents in the treatment of pSS remains controversial. Although most open-label studies of biologics in pSS have reported efficacy^{61,62,131,132}, the primary efficacy endpoint was not met in seven of the eight published RCTs of biologics in pSS; the exception was a small single-centre study of rituximab, the results of which were not replicated in two larger RCTs (Table 3).

The most commonly used inclusion criteria for RCTs evaluating biologic agents for the treatment of pSS were the American–European Consensus Group Sjögren's syndrome classification criteria,⁴ which were published in 2002, 10 years before the ACR criteria. Some studies also required the presence of salivary-gland biopsy abnormalities, the presence of anti-SSA/Ro or anti-SSB/La autoantibodies, or a decreased salivary flow rate. Composite inclusion criteria were used in some of RCTs. All of the RCTs published to date (Table 3) were designed and/or conducted before the publication of the ESSDAI and ESSPRI in 2014.

The lack of efficacy of TNF antagonists in pSS RCTs is unsurprising given the absence of a sound underlying pathophysiological rationale. By contrast, strong evidence supports a role for B cells in pSS, and open-label studies have consistently suggested benefits from B-cell depletion with rituximab^96,133. Rituximab significantly improved not only the primary outcome (sicca), but also fatigue and extraglandular manifestations (ESSDAI), in a study from Netherlands³⁹ and showed trends towards effects on fatigue and sicca in the TEARS RCT⁴¹. A post hoc analysis of the TEARS trial results suggested that selection of the primary endpoint can affect the interpretation of efficacy data and that subgroups of patients should be identified for future studies¹³⁴. However, the TRACTISS RCT, conducted in the UK and presented as a late-breaking abstract at the 2015 ACR meeting, found no beneficial effects of rituximab^63,64. Until the TRACTISS trial is published in full and its data analysed concomitantly with those from the TEARS trial, rituximab should probably not be used in pSS, except in patients with severe disease and no other treatment options.

Potential targets for future therapies can be identified based on knowledge about the immunopathological mechanisms involved in pSS; this section provides an overview of these mechanisms (see also Supplementary information S1 (table)). Lymphocytic infiltration is the histological hallmark of pSS. Mild focal infiltrates do not considerably impair the organization of the lachrymal or salivary glands, whereas severe diffuse lesions disrupt the architecture of the duct epithelium and gland parenchyma. T cells and B cells comprise the vast majority of the mononuclear cells infiltrating the salivary and lachrymal glands (>90% of infiltrating mononuclear cells); infiltrates in the salivary glands are composed of 25–75% T cells (of which 50–70% are CD4⁺) and 20–60% B cells¹²⁰. B cells predominate in advanced lesions¹³⁵. The abundance of macrophages and dendritic cells ranges from 0.3% to 15% and 0.2% to 9%, respectively¹³⁵.

Salivary-gland epithelial cells are also involved in the immunopathogenesis of pSS. Compared with salivary-gland epithelial cells from healthy controls, those from patients with pSS show reduced global DNA methylation, which could explain the aberrant transcription of many genes by these cells¹³⁶. This global DNA demethylation was associated with decreased expression of mRNA transcripts encoding the methylating enzyme DNMT1 and increased expression of its demethylating partner growth arrest and DNA damage-inducible protein GADD45α¹³⁶. Epithelial cell apoptosis induced by the infiltrating lymphocytes is considered a key factor contributing to the decreased production of exocrine secretions. Signalling of Fas ligand (FasL; also known as TNF ligand superfamily member 6) through Fas (also known as TNF receptor superfamily member 6) could explain the increased epithelial cell apoptosis mediated by T cells¹³⁷, and B cells can directly induce epithelial cell death through a pathway involving translocation of protein kinase Cδ (PKCδ) into the epithelial cell nucleus¹³⁸. In addition, several Toll-like receptors (TLRs) — including TLR2, TLR3, TLR4 and TLR7 — are expressed by epithelial cells in salivary-gland tissue^139,140. TLR signalling in salivary-gland epithelial cells upregulates their expression of MHC class I, intercellular adhesion molecule 1 (ICAM1; which interacts with CD11a), CD40, Fas-associated proteins, CD80, and CD86, thereby linking the innate and adaptive immune responses¹³⁹. Salivary-gland epithelial cells expressing CD86 possess distinctive binding properties, as indicated by preferential binding to the co-stimulatory molecule CD28 and reduced binding to cytotoxic T-lymphocyte antigen 4 (CTLA4)¹⁴¹.

Cytokine signalling also influences epithelial cell activation: IFNγ increases MHC class II expression by epithelial cells, enhancing antigen presentation to T cells¹⁴², and IFNγ and IL-1β induce CD40 expression by epithelial cells¹⁴³. Epithelial cells might present nuclear autoantigens such as the SSA/Ro and SSB/La ribonucleoproteins to T cells¹⁴⁴, and antigen presentation may be enhanced in the lachrymal glands in pSS owing to increased activity of cathepsin S¹⁴⁵, which promotes the migration of MHC class II complexes to the cell surface for presentation. Epithelial cells also produce chemokines such as CXC-chemokine 13 (CXCL13), CC-chemokine 19 (CCL19) and CCL21, which promote lymphocyte migration into the salivary glands¹⁴⁶.

Transcriptome analyses of salivary-gland tissue and mononuclear cells from patients with pSS have shown overexpression of type I IFN-induced genes¹⁴⁷. Phosphatidylinositol 4,5-bisphosphate 3-kinase catalytic subunit-δ (PI3Kδ) is activated by TLR stimulation and is required for type I IFN production¹⁴⁸. One of the most relevant IFN-induced genes encodes BAFF and its overexpression is a hallmark of pSS¹⁴⁹. Compared with normal levels of BAFF expression, BAFF overexpression could enable autoreactive B cells to survive increased levels of autoantigen-triggered death signalling and thus promote autoantibody production. The contribution of BAFF to the pathogenesis of pSS has been established in Baff-transgenic mice that develop several clinical features of Sjögren syndrome including salivary-gland inflammation¹⁵⁰. BAFF levels are elevated in the serum and salivary glands of patients with pSS, and are associated with increased production of autoantibodies such as anti-SSA/Ro and anti-SSB/La antibodies and rheumatoid factor^151,152. BAFF might also have a role in the pathogenesis of lymphomas associated with pSS^153,154. Germinal centre-like structures are found in salivary-gland tissue from 25% of patients with pSS, and could be involved in disease pathogenesis in these patients¹⁰.

In combination with lymphotoxin-β (LTβ), LTα enhances ectopic lymphoid neogenesis¹⁵⁵. Interestingly, terminal differentiation of B cells to plasma cells and memory B cells occurs within germinal centres under the supervision of T follicular helper (T_FH) cells¹⁵⁶. IL-21 production by dendritic cells contributes to maintaining the terminal differentiation of T_FH cells¹⁵⁷. In addition, cytokines produced by T helper 2 (T_H2) cells are central in maintaining B-cell function. T_H2 cells produce a vast array of cytokines including IL-4, IL-5, IL-6 and IL-13. Other cytokines are also implicated in pSS; for example, high levels of IL-17 have been found in serum and saliva samples, and IL-17-producing T cells and epithelial cells are present within inflammatory lesions of patients with Sjögren syndrome^{158,159,160,161}. Research into the possible role of IL-17 in the immunopathogenesis of Sjögren syndrome is increasing. IL-17 might drive the differentiation of stromal cells into CXCL12-expressing cells, which enable follicle formation even in the absence of follicular dendritic cells and promote the formation of ectopic germinal centres in pSS¹⁶².

Although B-cell overactivity is evident in pSS, a role is emerging for a new category of B cells known as B regulatory (B_REG) cells, which can blunt the development of autoimmune disorders¹⁶³. The CD40–CD40 ligand (CD40L) interaction between B cells and T cells is critical for the acquisition of B_REG-cell function upon T_H1-cell differentiation through the production of IL-10, IL-35 and transforming growth factor-β (TGFβ) or indoleamine 2,3-dioxygenase^164,165,166. Various chronic inflammatory environments that can occur in pSS have been reported to induce B_REG cell populations, and B_REG cells seem to be efficient at controlling T-cell proliferation and T_H1-cell differentiation in patients with pSS^164,167. Consequently, the depletion of B_REG cells could explain the limited efficacy of global B-cell depletion in pSS⁴¹.

Further research into the role of B_REG cells and other mechanisms involved in the immunopathogenesis of pSS is needed; however, as mentioned, our current knowledge highlights several potential therapeutic targets, some of which are being explored in clinical trials (see the next section). Other potential targets identified from the above discussion include CD80, CD86, cathepsin S, CD28, type I interferon, IFNγ, proinflammatory cytokines and chemokines.

The emergence of biologic agents has expanded the therapeutic armamentarium available for pSS, and many pharmaceutical firms are showing interest in developing new drugs for pSS. Several ongoing RCTs are evaluating new drugs that target molecules implicated in the pathogenesis of pSS, including IL-6, CTLA4, CD40, BAFF, B7-related protein (also known as ICOSL), CD11a, lymphotoxin-β receptor (LTβR) and PI3Kδ (Table 4).

In 2016, we have a well-accepted but non-validated platform on which to build treatment strategies for pSS. At present, the role for biologic therapies is small; even the most extensively investigated biologic drug, rituximab, has failed to demonstrate superiority over placebo in RCTs (Table 3). However, some tools are now available for evaluating new treatments in pSS, and many potential treatment targets have been identified. Patients with pSS who fail to respond to conventional treatment, as well as those patients who respond well to conventional treatment but seek further symptomatic relief, should be invited to participate in ongoing RCTs.