New investigational drugs to treat Sjogren's syndrome: lessons learnt from immunology

ABSTRACT

Introduction

Sjögren’s syndrome is a heterogeneous autoimmune condition that impairs quality of life because of dryness, fatigue, pain, and systemic involvements. Current treatment largely depends on empirical evidence, with no effective therapy approved. Clinical trials on targeted drugs often fail to report efficacy due to common factors.

Areas covered

This review summarizes the pathogenesis and what caused the failure of new investigational drugs in clinical trials, highlighting solutions for more effective investigations, with greater consistency between research outcomes, clinical use, and patient needs.

Expert opinion

Unlinked pathobiology with symptoms resulted in misidentified targets and disappointing trials. Useful stratification tools are necessary for the heterogeneous SS patients. Composite endpoints or improvements in ESSDAI scores are needed, considering the high placebo response, and the unbalance between symptom burden and disease activity. Compared to classic biologics, targeted cell therapy will be a more promising field of investigation in the coming years.

KEYWORDS:

• Sjögren’s syndrome
• clinical trials
• biologics
• ESSDAI
• stratification

1. Introduction

Sjögren’s syndrome (SS) is a chronic, systemic autoimmune condition that affects the exocrine glands and other organs of the body. Over 90% of the patients are women, and the age of peak incidence is in the fifth to sixth decades [1]. The hallmarks of SS are ocular and oral dryness, fatigue, and pain, which appear in over 80% of the patients [2]. The disease is histologically revealed by the infiltration of inflammatory effectors into the lacrimal and salivary glands. The clinical characteristics of pSS include arthralgia, interstitial pneumonitis, interstitial nephritis, Raynaud phenomenon, and so on, and comorbidities include hypothyroidism and infections [3–5]. pSS patients had an elevated risk of overall cancer, which is not only contributed by non-Hodgkin lymphoma (NHL) but also by other hematological malignancies and solid tumors [6]. Being a chronic immunological condition, SS brings long-term malaise and requests years of regular healthcare service, which has brought significant burdens to patients physically, mentally, and economically [2,7–9].

Current treatment of SS relies mostly on symptomatic interventions (such as tear and saliva substitutes, gustatory and muscarinic agonists, analgesics, and fluoride prescriptions for dental caries prevention), steroids, immunosuppressants, and biological therapies when systemic complications are relatively severe [10,11]. However, the effects are often modest or yet to be identified and challenges remain unsolved in meeting the objective alleviation of disease and subjective improvements in the life quality in patients [12]. In recent decades, thanks to deeper insights into immunological pathogenesis and genetic signature of the disease, researchers and physicians were able to investigate more powerful and subtle biological targets with promising results in modifying the balance of immunity, such as B-cell targeted therapies, costimulatory inhibition, mesenchymal stem cell therapy, and even personal therapies [10]. Fast updates on the investigation brought great news as well as lessons.

2. Pathogenesis

Normally, a low level of auto-responsiveness exists in the human body, playing a crucial role in the normal function of the immune system. Only if auto-reactivity becomes overwhelming and a sustained immunological response is established toward self-antigens, ‘safeguards’ turn into ‘self-killers’ and severe damage occurs in host tissue. In healthy individuals, a complete succession of checkpoints in the body prevents autoimmunity from happening. Central tolerance removes most auto-reactive immature lymphocytes; peripheral tolerance prevents mature lymphocytes from being activated in the absence of ‘danger’ signals; and regulatory tolerance uses regulatory cells and inhibitory cytokines that restrain the immune reactions. In some triggering occasions, such as infection and apoptosis [13–15], a pro-inflammation environment is established, and ‘self-ignorant’ lymphocytes can be activated to cause harmful inflammations. Self-tolerance is the ‘brake’ and inflammatory signal is the ‘accelerator’. Broken brake and active accelerator bring an overbalance between yin and yang, eventually leading to autoimmune diseases including SS.

2.1. Genetic predisposition

Evidence, such as the familial risk of pSS being higher among siblings of pSS patients [16], suggests that genetic predisposition plays a fundamental and intrinsic role in pSS. Extrinsic factors such as environmental interventions also have a retroactive effect on genetic features, resulting in significant differences in genetic expression as shown in emerging epigenetic studies [17]. In patients with positive anti-SSA/Ro and/or anti-SSB/La autoantibodies, both genetic and epigenetic variations of many loci have been found with genome-wide significance, corresponding to the clinical observation that autoantibody-positive patients are more likely to suffer from systemic conditions [18]. A review by Thorlacius et al. thoroughly listed and compared the variants or differences in GWAS and EWAS data between pSS patients and normal populations [17]. Evidence from genetic and epigenetic findings suggests that most variation signals occur in the TLR-interferon signaling pathway, whether it is a genetic variant or hypomethylation of an interferon-regulated gene. Examples include STAT4 in the JAK-STAT pathway [19], TNIP1 in the NF-κB pathway [20], CD247 in TCR signaling [19], and IRF5 regulating the expression of interferons [20]. Another hot spot of genetic variation associated with HLA loci indicates aberrance in the antigen presentation process and imperfect self-tolerance mechanisms. Examples include HLA-DQA1, HLA-DQB1, and HLA-DRA which present extracellular peptides by MHC class II molecules [20]. Other cytokine or chemokine receptors specific for lymphocyte function are also highlighted, such as BLK in BCR pathways [21], and CXCR5 responding to chemokine that mediates migration of B cells and T cells [22]. As a consequence of genetic alterations, the following pathological changes occur at the molecular or cellular level, which are described as immunological events chronologically.

2.2. Innate immunity

Innate immunity is initial and unspecific, leading to the massive production of cytokines (IFN). Innate immunity and an overactive IFN system are important features in the pathogenesis of SS [23]. Intrinsic susceptibility and extrinsic factors lead to overactivation of the axis of Toll-like receptors (TLRs)/cytosolic sensors – JAK-STAT pathway/NF-κB pathway – IFN, triggering continuous production of type I IFN and other cytokines in SS patients, establishing an inflammation-prone milieu. Intrinsically, the CGGGG insertion or deletion polymorphism in the IRF5 gene promotor causes a higher level of IRF5 mRNA, resulting in a high level of IFN production [24]. STAT4, a classical transcription factor of type II IFN, is also involved in the type I IFN pathway, and its polymorphism has an additive effect on IRF5 and is related to other IFN-induced genes [25]. Extrinsic triggers include viral infection and autoantigen exposure resulting from apoptosis. Hormonal factors especially estrogen deprivation also play an important role in triggering SS, which accounts for high incidence in menopausal women. Estrogen receptors are expressed on salivary gland epithelial cells (SGECs) [26]. Estrogen decrease induces RbAP48, leading to SGEC apoptosis and release of interleukin-18 (IL-18) and IFN-γ [27]. Interestingly, the role of SGECs in SS is not merely victims or bystanders, but promoters and mediators. SGECs act as atypical antigen-presenting cells by constitutively expressing TLRs, human leukocyte antigen (HLA) class I molecules, CD40 (a costimulatory protein), adhesion molecules, FAS receptors and ligands, and various cytokines as well as chemokines [5]. Cytokines activate more NKs, ILC3s, DCs, and macrophages, while chemokines attract more innate as well as adaptive lymphocytes to the scene, forming GC-like structures [28]. Infiltrated innate lymphocytes produce cytokines such as IL-7, IL-17, IL-21, IL-22, IL-23, and B cell activating factor (BAFF) that activate effector cells of the adaptive immune system discussed below [5]. The IFN signature is not only found at the glandular level but also at systemic level, such as peripheral blood mononuclear cells (PBMCs), isolated monocytes, plasmacytoid dendritic cells (pDCs), and B cells [23]. Notably, a feed-forward loop is established between innate and adaptive immunity. Type I IFN induces B cell activation and autoantibody production, which in turn leads to the formation of immune complexes and promotes escalated type I IFN production [23]. The inflammasome pathway is normally induced under stress conditions such as reactive oxygen species (ROS). Studies also provided evidence of SS pathogenesis via Nod-like receptor family protein 3 (NLRP3) inflammasome-mediated pyroptosis in exocrine glands [29]. Massive circulating cell-free DNA (cf-DNA) accumulation activates inflammasome [30], leading to activated caspase-1, causing elevated pro-inflammatory cytokines, and resulting in intensive production of pyroptosomes [31]. This process is further accelerated by type I IFN in SGECs [32].

2.3. Adaptive immunity

Adaptive immunity is antigen-specific, triggered by effectors of innate immunity. B cell hyperactivity is a hallmark of SS pathogenesis. B cell activation is completed through the IFN – BAFF – B axis. As discussed above, innate immunity produces type I and II IFN, inducing BAFF secretion by and release of interleukin-18 (IL-18) and IFN-γmonocytes, DCs, and even T cells, B cells, and SGECs in SS patients [33–36]. BAFF serves as the link between innate and adaptive immunity. Its level is reported to correlate with the anti-SSA/SSB and rheumatoid factor levels [37], suggesting an important role in SS pathogenesis. T cells play an indirect yet important role that should not be overlooked in SS pathogenesis. Antigen – CD4+ T cell – B cell axis is the key mechanism involved in ectopic germinal center (GC) formation in SG and B cell lymphomagenesis [38].

Activation of T cells depends on the combination of three signals: MHC with linear antigen (pMHC) serves as the first signal, acting as the ‘car engine’. Co-stimulatory molecule stimulation such as CD28 serves as the second signal, acting as the ‘car wheels’. Both ‘engine’ and ‘wheels’ are necessary to drive a successful activation. Different cytokines serve as the third signal, acting as the ‘steering wheel’, controlling the direction of CD4+ T cell differentiation into different effector subsets, such as Th1, Th2, Th17, Follicular-helper T-cells (Tfh), and regulatory T-cells (Treg) [39]. Th1 induced by IL-12 facilitates cell-mediated immunity, while Th2 induced by IL-4 facilitates humoral immunity. Th17 induced by TGFβ and IL-6 polarizes naive T cells by producing IL-17 and IL-22 and mediates inflammation by producing IL-6 and TNFα [40]. It is deduced from observations that Th1 and Th17 might initiate SS, while Th2 and Tfh cause SS progression via GC formation [41]. Tfh and peripheral-helper T-cells (Tph) secretes IL-21 which drives B-cell differentiation to plasma cells, resulting in uncontrolled autoantibody production and ectopic GC formation. IL-21 production, which is found to be elevated in SS patients, mainly results from stimulation of inducible costimulatory (ICOS) expressed on Tfh and Tph. The immunosuppressive counterpart of Tfh is T follicular regulatory cells (Tfr), whose level is also elevated in SS patients due to the compensation effect [42]. Although B cells play a devastating role in SS pathogenesis, there are regulative B cells that also display a suppressive effect on inflammation. They produce the inhibitory cytokine IL-10 and restrain the Tfh response in pSS [43]. Treg usually controls inflammation via mechanisms such as secreting regulatory cytokines such as IL-10 and TGF-β, competing for IL-2 with T cells by high expression of IL2 receptor, killing cytotoxic T lymphocytes (CTLs) and NK, and expressing CTLA-4 inhibiting DC maturation. Treg level is also found to be elevated in SS patients due to compensation [44]. Therefore, it seems that the loss of self-tolerance in SS is not due to defective regulatory tolerance [45].

Costimulatory cross-talk pathways between APCs are also worth attention. APCs in the SS context include DCs, SGECs, T cells, and B cells. There are three families of costimulatory molecules involved: B7 family (CD28, CTLA-4, ICOS, ICOSL, and CD80/CD86), TNF/TNFR family (CD40 and CD40L), cell-adhesion molecules (LFA-1 and ICAM-1) [46]. CD28 interacts with CD80/CD86, acting as the second signal essential in T-cell activation. CD28 also interacts with CTLA-4, a negative regulator of T-cell activation [47]. ICOS is specifically involved in T-cell-dependent B-cell activation [48]. CD40-CD40L interaction between T and B cells leads to T-cell-dependent B-cell activation [49]. T cells and B cells proliferate and migrate to form ectopic GC-like structures, where B cells undergo somatic hypermutation and antigen-driven selection of B cell receptors (BCRs), contributing to ongoing activation and lymphoma occurrence [50]. GC formation is highly associated with CXCR5, a receptor of chemokine CXCL13 that mediates B cell and Tfh cell migration [51].

Besides immune abnormality, neuroendocrine abnormality is also involved in the pathogenesis of SS [52]. It was reported that the function of the hypothalamic–pituitary–adrenal (HPA) axis is decreased in SS patients [53], leading to lower basal adrenocorticotropic hormone and cortisol levels, promoting fatigue and depression [54]. The involvement of the neuroendocrine system explains why some SS patients having limited histopathological features on biopsies report severe dry eye syndrome [55]. All these immunological and endocrinological events play crucial roles in the pathogenesis of SS and serve as current or potential targets under investigation.

3. Investigational drugs: a retrospect of failures

3.1. Targeting innate immunity

3.1.1. IFNs

Upregulation of type I and II IFN-regulated genes is involved in SS pathogenesis as discussed above. Reducing the upstream IFN expression is a plausible solution. RSLV-132 is an RNase fused to the Fc domain of IgG1, aiming to degrade the mRNA of IFNs in circulation. Fortunately, a phase II RCT proved its effect in alleviating severe fatigue in patients with pSS. But to the researcher’s astonishment, the expression of selected IFN-inducible genes was significantly increased, rather than reduced [56]. Interestingly, it has been reported and validated that fatigue in SS is associated with lower levels of proinflammatory cytokines, not higher (however, cytokine level in fatigue SS patients is still higher than healthy controls) [57,58]. It is proposed that after the initial proinflammatory response, the constant immune challenge leads to inappropriate activation of the anti-inflammatory negative feedback loop, and an exaggerated immune regulatory response reduces inflammatory markers, and plays a role in persistent fatigue [57]. In this RNase trial, increased activation of the IFN pathway may be a compensatory negative feedback mechanism to overcome the disease, since increased activation of the pathway is correlated with improved symptoms in primary SS [56]. Or, RNase has simply removed the circulating microRNAs [56].

Other targets of intervention controlling the IFN pathway include kinases. A phase II trial on filgotinib (JAK1 inhibitor), lanraplenib (SYK inhibitor), and tirabrutinib (BTK inhibitor) on SS patients failed to show significant efficacy in both ESSDAI and ESSPRI [59]. There are several common limitations behind this. First, there is a lack of stratification. For example, in the subgroup with high ESSDAI scores at baseline, filgotinib showed a significant effect in improving ESSDAI compared with placebo controls [59]. A pilot study on baricitinib (JAK1 and JAK2 inhibitor) treating SS showed the desired effect with the criteria that the baseline ESSDAI score is no lower than 5 [60]. Second, large decreases in ESSDAI in all treatment groups including placebo, as a result of heterogenous SS natural history. Third, the inconsistency between the classical parameters of disease activity and clinical response, as described above.

3.1.2. TNFs

TNFα is the initiator of downstream inflammatory signaling pathways, and its level is found to be elevated in SS patients [61]. TNF inhibitors have also been proven successful in treating many autoimmune diseases [62]. In viewing the ‘success’ of infliximab (a monoclonal antibody against TNFα) on SS patients, researchers put in efforts and resources to bet on the efficacy of other TNFα inhibitors, such as etanercept (a soluble TNF receptor) in treating SS. However, the results turned out to be rather disappointing [63,64]. Even subsequent trials on infliximab itself could not repeat its former glory [65]. In retrospect, the ‘successful’ infliximab trial was a small open-label pilot study [63] (and relevant publications of the infliximab trial were retracted in 2013 [66]). Further investigation shows that TNFα level was neither associated with the presence of systemic markers nor with the severity of pathogenic lesions. Counterintuitively, after etanercept administration, the level of circulating TNFα even significantly increased [67], with IFN-α and BAFF also elevated, as observed both in vivo and in vitro [68]. These trials provide us with several lessons. First, TNFα is not, at least, the pivotal factor in SS pathogenesis. Modifying its level is too indirect and unpredictable to provide the efficacy we need. Second, the immune system is subtle and highly self-regulative, a slight move in one part may affect the situation as a whole. Inhibiting TNF results in negative feedback such as the downregulation of suppressive IL-10, and subsequently the upregulation of relevant inflammatory cytokines, leading to the imbalance of pro- and anti-inflammatory factors. Third, the pharmacological uniqueness of a specific drug needs to be thoroughly understood. In this case, the soluble TNFR itself not only neutralizes TNFα but also prolongs its half-life [69], which accounts for the raised circulating TNFα level [67].

3.1.3. Inflammasomes and IL-1

Despite the well-known role as energy ‘currency,’ ATP also serves as a signaling molecule which is important in the inflammasome pathway of SS pathogenesis. Extracellular ATP is the natural agonist of purinergic receptors including the P2×7receptor (P2X7R), which is expressed in hematopoietic cells, neurons, glial cells, osteoblasts, endothelial cells, and epithelial cells [70]. Sustained activation of P2X7R can lead to ROS production, which plays a key role in inducing the assembly of NLRP3 inflammasome [71], resulting in the production of mature IL-1β and IL-18 [72]. Targeting this pathway, P2×7antagonist A438079 successfully reduced salivary gland inflammation and improved saliva flow in mouse SS models [73]. An RCT on SS using Anakinra (an IL-1 receptor antagonist) reported that the Fatigue Severity Scale did not differ from the control group at 4 weeks. However, when assessed with visual analog scales (VASs) in retrospect, significantly more patients in the Anakinra group had a fatigue reduction of more than 50% [74].

3.1.4. IL-6

IL-6 is crucial for both B-cell activation and T-cell polarization [75]. A multicenter RCT reported that tocilizumab (an IL-6 receptor inhibitor) did not reach its primary outcome in SS, and the immunological impact of systemic IL-6 inhibition was unexpectedly low [76]. However, a recently published study stratified patients into different subtypes and found increased expression of IL-6 and IL-1a in the subtype reporting the highest depression, anxiety, and fatigue, alongside tocilizumab trial reanalysis showing a reduction in the fatigue score within this subtype, suggesting potential merit in revisiting trials of monoclonal antibodies against these cytokines with stratification tools and revised endpoints [77].

3.2. Targeting adaptive immunity

3.2.1. Antigen presentation and costimulation

MHC class II mediated antigen presentation is dependent on protease cathepsin S in cutting its invariant chain. Cathepsin S inhibition prevents autoantigen presentation and holds promise in developing selective immunosuppressants since it will not directly target T cells but will specifically target CD4+ T cells, causing fewer adverse effects of immunosuppression [78]. Efficacy such as decreased glandular lymphocytic infiltration, decreased autoantibody production, and recovery of autoimmune manifestations is proven in SS mouse models [79]. However, a recent study using RO5459072 (a cathepsin S inhibitor) failed to prove its clinical efficacy [80], due to unexpectedly high placebo-response as described above. Despite this, a decrease in B and CD8+ T cells was observed in the RO5459072 group compared to the placebo-controlled group.

Another signal required during the antigen presentation process is the costimulatory signal. CD28 on the T cell surface binds stimulatory CD80/86 or inhibitory CTLA-4 on the APC surface. Abatacept is a humanized CTLA-4 fused with IgG. Two large RCTs on abatacept failed to meet the primary endpoint, since equally large decreases in ESSDAI were seen in placebo and active-treatment groups [81,82]. After using a novel composite endpoint for assessing treatment efficacy (the Composite of Relevant Endpoints for Sjögren’s Syndrome, CRESS), significant clinical effects compared to placebo could be detected in several otherwise ‘failed’ trials [83]. Also, despite the 24-week abatacept pilot study not showing efficacy on ocular and oral dryness, the 24-month pilot study showed a significant increase in saliva gland function [84]. This brings another lesson that a longer trial period may be necessary for certain drugs to perform efficacy [85].

Prezalumab (or AMG557, MEDI5872) is a human anti-ICOSL antibody. A phase IIa RCT study on prezalumab failed to show a significant difference in change in ESSDAI scores from the placebo group [86]. However, IgM-, IgG-, and IgA-RF levels all decreased significantly with active treatment, but not with placebo, suggesting that the dose used in this study achieved biological efficacy, which is confirmed in salivary gland biopsies [46].

Efalizumab (marketed as Raptiva) is a monoclonal antibody against the alpha-subunit of LFA-1. Based on the finding of an association between the use of Raptiva and an increased risk of progressive multifocal leukoencephalopathy (PML), a rare and usually fatal disease of the central nervous system seen in the TEARS trial, the study was stopped [87]. As a result, a pilot study on efalizumab is prematurely terminated, as shown in ClinicalTrials.gov (ID: NCT00344448). Another immunosuppressant also targeting adhesion molecules, natalizumab (marketed as Tysabri), has been reported to cause PML as well [88]. The underlying mechanism might be inhibiting CTL entrance to CNS, resulting in an inadequate cell-mediated immune response against John Cunningham polyomavirus (JCV) [89]. This provides a clue to the association between drug mechanisms and the severe adverse effects, which might enable finer predictions on drug safety.

3.2.2. B cell activation

Since B cell hyperactivity was the most cardinal feature of SS pathogenesis, targeting B cell activation seems like the most promising strategy. Rituximab is a monoclonal antibody against CD20, a cell surface antigen expressed on B cells. The main problem is the lack of consistency between trials in the efficacy of rituximab [90]. The large randomized controlled TEARS trial reported failure in SS patients because the proportion of patients achieving the primary endpoint was not significantly different between the rituximab and placebo groups [91]. However, several secondary endpoints were improved with statistically significant differences. The TRACTISS trial also reports negative results [92] that are positive in other trials [93]. The inconsistency might result from the inclusion of heterogeneous cohorts in the former trial, and the allowance of using background immunosuppressants during the process, which is also seen in the TEARS trial [90]. In retrospect, the failure partly resulted from the inability of the ESSDAI primary endpoint to evaluate the actual effect of rituximab [94]. A composite Sjögren’s Syndrome Responder Index (SSRI)-30 is thus developed with validation using rituximab and infliximab with clinical observation [95]. Half of the rituximab group showed notable improvements when assessed either with SSRI-30 [95] or ultrasonography [96]. Another study reanalyzed both TEARS and TRACTISS trials with the recently developed Composite of Relevant Endpoints for Sjögren’s Syndrome (CRESS [83]) and Sjögren’s Tool for Assessing Response (STAR [97]). It demonstrated a significantly higher response rate in the rituximab-treated arm compared to placebo [98]. It was also observed that BAFF plays a pivotal role in rituximab efficacy, with high baseline levels of BAFF associated with no clinical response to rituximab [94], which provides inspiration for personalized therapy of SS. The trials of Rituximab have shown that composite endpoints can be a potential solution to the issues that arise in the current outcome measurement system. The drug has also been proven to yield positive results in selected groups of patients, which highlights the need for more effective stratification tools in the case of SS patients. As for rituximab-refractory pSS, case reports suggest that daratumumab (monoclonal antibody targeting CD38) may provide efficacy [99], but relevant clinical trials are few.

Belimumab, a monoclonal antibody targeting BAFF, is administered in a phase II RCT and performs significant benefits, especially parotid swelling, lymphadenopathy, cryoglobulinemia, and articular symptoms, but with very modest improvement in fatigue, pain, and salivary and lacrimal secretion [93]. The relative specificity of symptoms to drugs (and reversely, drugs to symptoms) suggests that the selection of therapies can be symptom-directed since symptoms are derived from different mechanisms in pathogenesis, targeted by a specific class of drugs. For instance, fatigue is mainly mediated by type I IFN pathways [57], and B cell-related symptoms are mainly mediated by BAFF and other related pathways.

Besides, belimumab has given a good example of using its complementary effect to neutralize the shortcomings of others. Since BAFF can mediate the resistance to anti-CD20 tissue B-cell depletion [100], anti-CD20 therapy is effective on parotid lymphoma and cryoglobulinemia when given soon after belimumab [101]. Belimumab is also useful to prevent the increase in BAFF levels after anti-CD20 therapy; otherwise, it would lead to the repopulation of autoimmune B cells after treatment [102]. A phase II RCT on sequential use of belimumab and rituximab reported satisfying effects [103]. This effect can also be realized via a single drug with multiple targets. Ianalumab (VAY736) is a monoclonal antibody against the BAFF receptor, which has two modes of action: direct lysis of B cells by antibody-dependent cytotoxicity, and BAFF receptor blockade. This therapy hit two birds with one stone and was proven to have significant efficacy in decreasing ESSDAI in a phase IIb RCT [104]. Besides belimumab, new investigational drugs targeting BAFF have also emerged with promising results. Telitacicept, acting on both BAFF and proliferation-inducing ligand (APRIL) [105], showed significant benefits in improving ESSDAI score and disease activity in a phase II RCT [106]. Due to its limited trial size and high withdrawal, confirmation of the efficacy of telitacicept should be carried out further by phase III trials.

Tirabrutinib and filgotinib are drugs affecting the BCR signaling pathway, targeting BTK and JAK/STAT pathways, respectively. Their efficacy is already discussed in the ‘targeting innate immunity’ section.

3.2.3. T cell activation

Since B cell hyperactivation is initiated and reinforced by CD4+ T cells, targeting T cell activation is also a promising strategy. IL-2 is the key cytokine in the homeostasis of CD4+ T cells, especially in the development, survival, and proliferation of Tregs when given at low dosage [107]. Evidence suggests short-term (5 days) and low-dose IL-2 helps decrease the Th17/Treg ratio to normal levels but failed to perform improvement in disease activity during short-term observation (6 days), despite a significant reduction in glucocorticoid and disease-modifying antirheumatic drugs (DMARDs) usage was observed in long term [108]. However, in a later small RCT study, longer-term low-dose LI-2 therapy (12 weeks) significantly improved ESSDAI scores, alleviation of main symptoms, and restored immune homeostasis in pSS patients, with no severe adverse effects (observed at week 24) [109]. This indicates that a longer and larger sample trial period may be necessary for success.

3.2.4. Germinal center formation

Lymphotoxin (LT) α1β2 is a proinflammatory mediator signaling through LTbR. The preclinical study suggested that anti-LTbR antibody decreased the level of B cell infiltrations into lacrimal glands and impeded high endothelial venules (HEV) formation, therefore preventing further lymphocyte migration [110], which may inhibit GC formation. A phase II RCT on baminercept, a monoclonal anti-LTbR antibody, failed to significantly improve glandular conditions in SS patients [111]. However, strong biological effects were observed as reduced plasma level of chemokine CXCL13, which is consistent with preclinical and other clinical studies [112,113]. The most likely cause of investigational inefficacy is the low ESSDAI scores at entry, which hampered the ability to detect true efficacy hidden behind the reduction of disease activity [111].

4. Conclusion

A successful trial requires a drug with intrinsic efficacy, a suitable group of patients, follow-up time, and proper criteria for efficacy determination. Drugs lacking intrinsic efficacy mostly result from misidentifying nonpathogenic parameters, severe adverse effects of infection, and opaque immunological mechanisms, especially its obscure role in the pathophysiology of a certain symptom [74]. Besides the intrinsic inefficacy of a drug, some trials with negative results underestimated the real efficacy of a therapy. The underlying factors include patients being overly heterogeneous, time duration being too short, and unsuitable endpoints.

There are multiple issues worth noting. First, since no consensus is made upon stratifying SS into different subtypes, patients who are not supposed to respond to a drug are still included in the trial, resulting in excessive false-negative results [59,76]. Thus, it is fairly important to test a drug in a specific subgroup of SS patients who are presumed to respond to the drug, rather than in heterogeneous patients. Second, it is difficult for ESSDAI to reveal drug efficacy on many occasions. For instance, when the baseline score is not high enough [59,113], or patients have limited systemic activity but severe symptom burden [114]. The high placebo response using the ESSDAI score as a scale also leads to the inability of a trial to manifest the true efficacy due to variation in SS disease activity [59,63,64,66,80–82,91], which calls for a unified and improved composite clinical endpoint. Also, different drugs have an intrinsic predisposition to alleviate different subgroups of symptoms [56,58,74], rather than broad-spectrum targeting, which indicates that further investigations are needed to clarify the correlation between certain symptoms and the pathogenetic pathway.

5. Expert opinion

Up to now, the classic paradigm for drug investigation in SS is analyzing the differences of many ‘−omic’ (genomic, epigenomic, transcriptomic, proteomic, etc.) patterns between SS patients and control, figuring out the immunological effect of those shifted molecules, selecting one promising variant to either upregulate or downregulate it, assessing efficacy in animal trials and subsequently clinical trials. However, it is not meticulous enough, since multiple immune pathways are mobilized in SS, but not all are necessarily linked with disease advancement. For example, although elevated in SS patients, TNF does not represent a key mediator of disease and thus may not be an appropriate intervention target [63–69]. Among the variants, there are ‘starters,’ ‘followers,’ and ‘compensators,’ which play very different roles in the pathogenesis of SS. The trials targeting the ‘compensators’ and ‘followers’ are not likely to get the expected results [74]. Therefore, it is fairly important to clarify the underlying pathways of the variations and identify the true determinants of shifted parameters in the pathway, which will save large amounts of resources and time on inefficient trials. Until now, much is still to be elucidated in terms of the pathophysiology behind many debilitating symptoms of SS such as fatigue and pain. Fatigue does not correlate with inflammatory cytokines [57] or BAFF levels [115]. Linking pathobiology with symptoms is a currently needed task [77].

The heterogeneity of SS patients suggests the urge for stratification. It will provide benefits for not only trial design but also clinical management and therapeutic development. The current classification criteria [116] are not fine enough to select targeted patients for trials [117]. The Newcastle Sjogren’s Stratification Tool (NSST) based on symptoms provides a promising solution [118]. A categorization of patients was conducted, classifying them into four subgroups based on their symptoms: pain, fatigue, dryness, and depression. The stratification based on patient-reported symptoms unveiled distinct pathobiological endotypes that exhibited distinct reactions to biological therapies. Future stratification tools may be based on serological, histopathological, and/or functional characteristics depending on the specific therapy objective. Stratification of patients with SS and SLE according to two shared immune cell signatures was also proposed [119], shedding light on common pathophysiological features shared by autoimmune conditions. Besides transverse stratification, a longitudinal subsection based on chronological development is also important in clinical assessment. For instance, the greatest potential advantage of rituximab may be observed in cases of early, active disease and patients displaying extra-glandular manifestations [114].

A revision of endpoints is suggested to avoid false-negative results. Using the current ESSDAI, disease activity can be very difficult to differentiate from damage, and it is also prone to result in an unnecessarily high placebo response. Given the large number of ongoing trials that rely on ESSDAI for performance evaluation, reassessing the use of this score as a primary outcome criteria is of vital importance [120]. Furthermore, maintaining the repeatability and accuracy of ratings in every domain continues to be difficult. There is a lot of room for subjectivity in the current scoring and weighting, with significant individual variation [80]. To develop ideal assessment tools, composite clinical endpoints are proposed and validated, such as CRESS and STAR [83,94,95,97,98]. Another study determined disease activity levels, PASS and MCII of ESSDAI and ESSPRI, which is also practical in inclusion and assessing efficacy [121].

Since artificial intelligence-driven drug development entered the scene, it will facilitate homogeneous clustering of SS subtypes, inferencing causality of perturbed immunological pathways and design, optimize, and evaluate drug candidates [122]. Human work might focus on the bridging between clinical, genomic, and pathophysiological outcomes, elevating consistency between them. Studies toward connecting molecular patterns to symptoms and differentiating between the SS subtypes have emerged. The differentially expressed proteins were interrelated in network analysis, supporting the concept that distinct molecular networks underlie the clinical subtypes of SS [77]. Symptom-based drug selection based on SS subtyping is likely to be the future paradigm.

In the coming years, cell therapy is anticipated to surpass classical biologics as a more promising means of combating autoimmune diseases. Therapeutic cells such as CAR T-cells have very long survival periods in the tissues and clear cellular targets, contrasting very short half-lives and vague effects of traditional drugs. Therefore, it enables a deep and sustained abrogation of auto-reactive cells and allows a permanent cessation of immunosuppressive drug administration in patients [123]. Some stem cell therapy and immune cell therapy are under investigation. However, stem cell therapy may not be an ideal method in the near future, since its effect on patients is very broad (therefore its application in so many scenes) and the mechanism underlying its therapeutic effect also remains a mystery. Autologous hematopoietic stem cell transplantation (auto-HSCT) is also restricted because of significant safety concerns [124]. In contrast, targeted immune cell therapy has clear targets and the therapeutic mechanism is direct and more predictable. Currently, most relevant investigations are focused on CD19 CAR T-cells, which resulted in fast and sustained eradication of autoimmune B cells [123]. More targets and more cell types can be developed into cell therapy against autoimmune diseases in the future, such as TCR-T, TIL, NKs, and DCs. Tumor therapies can provide inspiration for the treatment of autoimmune diseases, and therapeutic methods effective on SLE and other autoimmune conditions can provide inspiration for SS treatment. Despite the promising future of cell therapy in SS, the main task of investigation is and will always be the ultimate clarification of SS pathobiology. After all, there is no silver bullet if we have no idea where to shoot.

Article highlights

• The inclusion of overly diverse patients highlights the need for more effective tools that can stratify patients based on their specific characteristics.

• To measure the true effectiveness of a drug treating SS, either more composite endpoints need to be used or there must be notable enhancements in ESSDAI.

• Studies should prioritize the right targets within well-defined subgroups of patients.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was funded by the Chinese National Key Technology R&D Program, Ministry of Science and Technology [2017YFC0907601, 2017YFC0907605], CAMS Innovation Fund for Medical Sciences (CIFMS) [2021-I2M-1-005], National High-Level Hospital Clinical Research Funding [2022-PUMCH-B-013].