Safety of treatments for primary Sjögren’s syndrome

ABSTRACT

Introduction: Primary Sjögren’s syndrome (pSS) is a disabling auto-immune disease, affecting exocrine glands and several organs.

Areas covered: In this review we analyze the safety of therapies used in pSS. Symptomatic treatment is widely applied due to the good supportive effect and good safety profile. Systemic stimulation of tears and saliva can be successful in pSS. However, cumbersome adverse events can influence the tolerability of this therapy. Evidence for the effectiveness of synthetic DMARDs therapies in pSS is limited, while there is a risk of adverse events. Several studies on biologic DMARD treatment of pSS patients have shown promising efficacy and safety results.

Expert opinion: The safety of symptomatic treatment of pSS is very good. However, systemic therapy is necessary to achieve long-term relieve and prevention of organ-damage. Synthetic DMARDs have not shown much efficacy in earlier studies, and their benefits do not weigh up to the possible harms, while biologic DMARDs show promising results regarding efficacy and cause mostly mild adverse events. Many questions remain unanswered regarding safety of DMARDs in pSS. There is a need for well designed studies, in which safety should be evaluated in a uniform manner to be able to compare the results between studies.

KEYWORDS:

• Primary Sjögren’s syndrome
• drug safety
• topical treatment
• synthetic DMARDs
• biologic DMARDs

1. Introduction

Primary Sjögren’s syndrome (pSS) is a systemic autoimmune disease characterized by lymphocytic infiltration of the exocrine glands leading to, among others, sicca symptoms of the eyes and mouth. Several systemic and extraglandular manifestations can develop, including fatigue, arthritis, and involvement of organs such as the skin, lungs, and kidneys. Although the pathogenesis of pSS is not fully elucidated, T-cell-mediated B-cell hyperactivity is thought to play an important role, as reflected by the presence of autoantibodies, cryoglobulins, and hypergammaglobulinemia.[1] pSS is a disabling disease and has a large effect on health-related quality of life.[2] Besides symptomatic treatments that improve dryness, no effective treatments have yet been approved for use in pSS. However, treatment with biologic disease-modifying antirheumatic drugs (DMARDs) has shown promising outcomes.[3]

Evaluation of treatment outcomes in pSS has been challenging in the past due to the wide range of outcome measurements that were applied, making it difficult to compare studies. Early studies primarily focused on exocrine gland function (saliva, tears) as a primary outcome measurement, whereas later studies focused on fatigue and systemic symptoms. Furthermore, pSS has a very heterogeneous course. Most patients show a chronic progressive decrease in exocrine gland function, until very low or no saliva and tear production remains.[4] However, systemic symptoms can present in different patterns, as patients can show chronic involvement (e.g. polyneuropathy) and exacerbations (e.g. polyarthritis).[5,6] The recent development and validation of disease activity indices, the European League against Rheumatism Sjögren Syndrome Disease Activity Index (ESSDAI) and Patient Reported Index (ESSPRI), have enhanced clinical research as it is now possible to reliably measure changes in disease activity and patient-reported complaints.[7,8]

When assessing the efficacy of new treatments, and also when applying ESSDAI and ESSPRI, it is important to keep the balance between efficacy and adverse effects in mind. Therefore, the aim of this review is to summarize the safety of treatments currently applied in pSS and to identify in which areas knowledge is still lacking. This review will discuss symptomatic treatment, synthetic DMARDs, and biologic DMARDs, with a focus on the treatments that have shown promising results.

2. Safety of symptomatic therapies

Symptomatic treatment of patients with pSS is widely applied due to the noninvasive nature, good supportive effect, and good safety profile. Educating the patient with regard to lubricant use, preventive dental care, and general personal hygiene, avoiding windy or low-humidity environments and exposure to irritants such as dust and cigarette smoke is important. Attention should also be paid to several medical conditions and medications that can aggravate sicca symptoms.

2.1. Ocular manifestations

First-line treatment of ocular sicca symptoms consists of topical treatment with artificial tears, gels, and ointments.[9] If the effect of tear replacement is inadequate, topical immunomodulatory agents such as cyclosporine and corticosteroids, and systemic stimulation of tear production can be added to the treatment.

2.1.1. Artificial tears

Substitution therapy, such as eye drops, gels, and ointments, is the mainstay of the treatment of sicca symptoms. There are different types of artificial tears available on the market, based on hydroxypropyl methylcellulose, carboxymethylcellulose, hyaluronic acid, polyethylene glycol, or propylene glycol as well as gel/lipid formulations, ointments, and liposomal sprays. Ophthalmic gels and ointments may be used at night. Highly viscous drops, gels, and ointments have longer effect duration, but they may cause visual blurring. In general, if used appropriately, artificial tear substitutes have good safety and tolerability characteristics. The most common adverse event (AE) is a temporary burning sensation. Other AEs include eye redness, discharge, watery eyes, eye pain, foreign body sensation, itching, stinging, and blurred vision.[10]

Blepharitis may worsen by the use of artificial tears, especially those with high viscosity or those containing preservatives, which can also damage the corneal epithelium and disrupt the tear film.[11,12] The advantage of preservatives is that the drops are available in multidose administration bottles. However, patients can develop an adverse reaction to the preservative. The use of artificial tears containing preservatives should therefore be restricted to three times a day to prevent high concentrations of these substances. Preservative-free artificial teardrops should be used in the form of single-dose dispensers to prevent infection risk, which in turn increases the costs of these substitutes. The choice of artificial tears should be based on individual patient characteristics.

2.1.2. Autologous serum

Autologous serum eye drops might be superior to artificial tear substitutes due to presence of a variety of biological factors, such as epidermal growth factor (EGF), vitamin A, transforming growth factor beta (TGF-β), fibronectin, substance P, insulin-like growth factor 1 (IGF-1), and nerve growth factor (NGF). Moreover, autologous serum eye drops have an osmolarity comparable to that of natural tears.[13]

Although there is some evidence for the effectiveness and safety of autologous serum eye drops in pSS patients,[14,15] large randomized controlled trials (RCTs) are warranted to provide sufficient evidence on the superiority of these drops over artificial tears.

The AEs of autologous serum eye drops are mild and include increased discomfort, slight epitheliopathy, bacterial conjunctivitis, and eyelid eczema.[14,16,17] Autologous serum eye drops should be prepared under a strict protocol and in sterile conditions.

2.1.3. Topical cyclosporine A

Topical cyclosporine A (tCsA) 0.05% ophthalmic emulsion is an immunomodulatory agent with the ability to downregulate T-cell proliferation, activity, and receptor signal transduction. tCsA has an anti-inflammatory effect due to decreased formation of proinflammatory cytokines. The latter effects of tCsA contribute to the stability of the tear film by interruption of the inflammatory cascade, inhibition of apoptosis, and stimulation of production of goblet cells in the corneal epithelium. Goblet cells produce mucin, which serves as an interface between the hydrophobic corneal epithelium and aqueous tear fluid.[18–21]

tCsA is recommended for the treatment of pSS patients with moderate-to-severe inflammation of the cornea.[22] Long-term use of tCsA is well tolerated in pSS patients.[23] A variety of AEs are reported, including burning and stinging symptoms, foreign body sensation, and blurring. These adverse effects resolve with cessation of treatment.[23,24] No systemic side effects were observed during tCsA treatment. Patients with ocular infections should discontinue tCsA use.[23] Taken together, tCsA is an important tool in the management of ocular manifestations in SS with a good tolerability, no systemic side effects, and overall good safety profile. Unfortunately, tCsA is not registered for use in pSS in several countries.

2.1.4. Topical glucocorticoids

Non-preserved glucocorticoid eye drops are used in pSS patients with moderate-to-severe disease. By reducing the levels of cytokines, such as interleukin-1 and interleukin-8, topical glucocorticoids suppress the inflammatory process. Furthermore, these eye drops reduce the activity level of matrix metalloproteinases.[25] Although the overall safety of topical glucocorticoids in clinical trials in pSS and keratoconjunctivitis sicca was considered satisfactory, prolonged use of topical glucocorticoids in pSS patients is restricted by their ability to induce glaucoma, cataract, decreased wound healing, increased risk of secondary infections, and epithelial defects.[26–29] Therefore, topical glucocorticoids are only recommended for short-term use when treatment with artificial tears is insufficient, and rapid reduction of inflammation should be achieved.

2.1.5. Topical NSAIDs

Use of nonsteroidal anti-inflammatory drug (NSAID) eye drops in pSS was evaluated in a couple of studies.[30–32] Inhibition of prostaglandins and the arachidonic acid cascade by topical NSAIDs can relieve ocular hyperalgesia. However, in patients with corneal problems, a common phenomenon in pSS patients, the use of topical NSAIDs is associated with corneal-scleral melts, perforation, and severe keratopathy.[33,34] Therefore, there is no place for topical NSAID eye drops in the treatment of pSS patients due to potential corneal complications and inferiority to topical corticosteroids.

2.2. Oral manifestations

The treatment of the sensation of a dry mouth (xerostomia) and salivary gland hypofunction (hyposalivation) in pSS patients should be performed according to the following instructions.[35] Stimulate the flow of saliva by gustatory and mechanical stimulation, or systemic stimulation. If the saliva cannot be adequately stimulated, decreased sicca symptoms can be achieved by coating the surfaces of the oral mucosa with artificial saliva. Preserve and protect the teeth and the oral soft tissues with topical fluorides. The mainstay of this therapy is to make it as simple and as safe as possible for the patients, e.g. limit salivary stimulation therapy to gustatory and mechanical stimulation as this is accompanied by fewer side effects than systemic stimulation therapy.

2.2.1. Gustatory and mechanical stimulation

The combination of chewing and taste, as provided by gums, candies, and mints, can be very effective in relieving symptoms for patients who have residual salivary gland function. Masticatory stimulatory techniques are the easiest to implement and have few AEs. Gums, candies and mints should be sugar free and preferably contain mild acids, with a low risk to harm teeth and oral mucosa, such als malic acid.

2.2.2. Saliva replacement therapy

Artificial saliva (saliva substitutes) is available for the treatment of moderate-to-severe dry mouth in patients with pSS. A variety of saliva substitutes are available, some are water-based and often short-working, others are gels that are preferably used when stimulation or frequent moistening is not applicable, e.g. at night. In this regard, it also has to be mentioned that many pSS patients use water to moisten their mouth, which can be used freely, but is a poor moistener of the oral mucosa.

When prescribing a saliva substitute, it is important to instruct the patient properly on how to use the substitute to get the maximum effect from the therapy, as it is not an exception that use of artificial saliva is not well accepted in the long term by many patients, particularly when they have not been instructed on how to use it. Moreover, it is very useful to try another type of substitute in a patient when a particular substitute does not sufficiently relieve xerostomia; which substitute is effective in a particular patient is often related to the preference of a patient and is not easy to predict. The safety of saliva substitutes is very good, with a very small number of minor AEs reported.[36]

2.2.3. Topical fluorides

Topical fluorides in patients with salivary gland hypofunction are critical to control dental caries.[37] There are different fluoride therapies available, ranging from low-concentration, over-the-counter fluoride rinses, to more potent and highly concentrated prescription fluorides (e.g. 1.0% sodium fluoride). Oral health-care practitioners may also utilize fluoride varnishes. The dosage chosen and the frequency of application should be based on the severity of the salivary hypofunction and the rate of caries development.[37,38] In addition, particularly in patients with severe oral dryness, nonacidic gels and/or solutions should be used, as acidic sodium fluoride gels may induce a more rapid destruction of the teeth, and could cause sensitivity and pain in the gingival and oral mucosa. There is little information on the risk of AEs in the available studies. Known AEs of the use of fluorides are fluorosis, tooth staining/discoloration, oral allergic reactions, nausea, or vomiting.[39]

2.3. Systemic stimulation of tears and saliva

2.3.1. Pilocarpine

Pilocarpine is a cholinergic parasympathomimetic agonist with onset of action within 1 hour. It binds to muscarinic-M3 receptors of various exocrine glands to stimulate the secretion function.[40] Contraindications to the use of pilocarpine are uncontrolled asthma, untreated cardiovascular conditions, angle-closure glaucoma, and severe hepatic impairment. Precautions should be taken by patients with cholelithiasis or nephrolithiasis. The effect on saliva flow is dose-dependent and time-related, with a duration of effect of about 3–5 hours.

Several RCTs were conducted to evaluate the efficacy and safety of pilocarpine in SS patients.[41–44] Salivary flow rate and visual analogue scale (VAS) for dry mouth or dry eye were significantly improved in the pilocarpine groups compared to placebo. Pilocarpine showed improvement on the VAS for eye dryness and Rose Bengal test compared to artificial tears alone or punctual occlusion intervention. No serious AEs were reported. The most frequent AEs were sweating, increased urinary frequency, headache, flu syndrome, nausea, dyspepsia, rhinitis, and dizziness. Adverse effects occurred more often at higher doses. In these studies, 0–13% of patients receiving pilocarpine discontinued treatment due to AEs versus 0–10% of patients receiving placebo.[41–44] Recently, Kawakita and colleagues demonstrated that a lower dose of 2.5 mg pilocarpine three times a day is effective in patients with SS and can diminish the AEs.[45] Moreover, pilocarpine seems to be safe and effective in juvenile-onset SS.[46] Conclusively, pilocarpine can be successfully used in pSS patients, especially in those with sufficient residual salivary gland function. However, common and cumbersome AEs can influence the tolerability of this therapy.

2.3.2. Cevimeline

Cevimeline is a parasympathomimetic and muscarinic agonist that, just like pilocarpine, has a particular effect on M3 receptors. It stimulates saliva secretion, thereby alleviating dry mouth. Cevimeline has the same contraindication profile as pilocarpine.[47] Several RCTs confirm the effectiveness and favorable safety profile of cevimeline.[47–52] Cevimeline is not yet approved for use in Europe. However, the tolerability of cevimeline seems to be better compared to pilocarpine and is associated with lower discontinuation rates during the treatment.[53]

3. Safety of systemic immunosuppressive therapies

Several systemic immunosuppressive therapies have been studied in phase 2 and phase 3 trials with pSS patients.[3,54] Although no systemic treatments have yet been registered for use in pSS, the number of studies with systemic therapy in pSS is increasing. In the next section, the safety of several systemic therapies that have shown some effect in pSS will be discussed.

An important safety issue during immunosuppressive therapy in rheumatic diseases is the increased risk of serious infections.[55,56] The risk of infection in patients on systemic immunosuppressive therapy is, amongst others, influenced by comorbidity, use of other immunosuppressive medications, and age. In pSS, the presence of extraglandular manifestations such as interstitial lung and renal disease may further increase the risk of infection. During treatment with any systemic DMARD, physicians should be aware of this and monitor patients for the development of infections. Careful clinical and laboratory assessments need to be carried out to minimize the potential risk of adverse effects.

3.1. Synthetic DMARD therapies

Patients treated with synthetic DMARDs have a rather high risk of developing adverse reactions. The most common adverse effects of synthetic DMARDs are infections, bone marrow toxicity, gastrointestinal symptoms, and cardiovascular diseases (e.g. hypertension). Therefore, it should be assessed before onset of therapy whether the benefits of a therapy outweigh its possible adverse effects.

Most synthetic DMARDs have not been shown to be effective in patients with pSS in double-blind, randomized clinical trials. Methotrexate, leflunomide, and cyclosporine A have shown insufficient efficacy in clinical trials, and/or their use was accompanied by unacceptable rates of AEs.[57–61] As a consequence, most synthetic DMARDs are not used routinely for pSS. In case of severe or life-threatening organ involvement in pSS, however, synthetic DMARDS are frequently prescribed on an empiric basis or on the basis of a small case series.

3.1.1. Hydroxychloroquine

Based on efficacy experience in systemic lupus erythematosus (SLE) patients, hydroxychloroquine (HCQ) is frequently used in pSS to treat skin involvement, e.g. purpura associated with hypergammaglobulinemia, myalgia, arthralgia, arthritis, and constitutional symptoms like fever and fatigue.[62] HCQ is also effective in prevention of cardiovascular events by reducing levels of total cholesterol, increasing high-density lipoprotein cholesterol, and improving the atherogenic index.[63] Moreover, a recent report has shown that HCQ impairs systemic interferon alpha (IFNα) production in pSS, which is hypothesized to play an important role in the pathogenesis of pSS.[64] It has been shown, however, that administration of HCQ did not resolve sicca signs and symptoms or extraglandular manifestations in pSS patients.[65,66] In both trials in pSS, serious AEs rarely occurred. Gottenberg et al. reported a similar frequency of serious AEs in the HCQ group and in the placebo group during the first 24 weeks. The most common adverse effects of HCQ are skin rash, hyperpigmentation of the skin, temporarily hair loss, and blurred vision.[62] Furthermore, bilateral bull’s-eye maculopathy is considered a serious adverse effect, resulting in loss of visual acuity, loss of peripheral vision, and loss of night vision.[67] Screening for bull’s-eye maculopathy should be done at the start of HCQ, after 5 years of treatment, and yearly thereafter. The risk of eye toxicity is low unless the patient suffers from impaired kidney function or is given HCQ at a high dose (dose of >6.5 mg/kg of ideal body weight). Furthermore, as HCQ is not retained in fatty tissues, obese patients can be seriously overdosed when HCQ is dosed on basis of the patients’ actual body weight instead of the ideal body weight. Another risk group is elderly patients with retinal and macular diseases. Rare adverse effects include cardiomyopathy, hearing disorders, and myopathy.[68–70] HCQ is safe to be used during pregnancy and lactation.[71,72] In conclusion, HCQ has a mild AE profile. The efficacy of HCQ in pSS patients needs further evaluation.

3.1.2. Glucocorticoids

Glucocorticoids are used in pSS patients with severe organ complications like renal involvement and myelitis. The effect of glucocorticoids on sicca symptoms and signs and extraglandular manifestations has not yet been proven in RCTs. Moreover, a prospective cohort study failed to show the effect of glucocorticoids on salivary function tests.[73]

Adverse effects of glucocorticoids are common, depending on the dose and duration of the therapy. Short-term adverse effects include steroid-induced diabetes mellitus, hypertension, peptic ulcers of the stomach, electrolyte disturbances, heart failure, and mood disorders. Long-term complications are susceptibility for infections, osteoporosis, steroid-induced myopathy, cataract, glaucoma, Cushing syndrome, thin skin, and central adiposity. The use of glucocorticoids should always be combined with prophylaxes for osteoporosis and peptic ulcers.

3.1.3. Cyclophosphamide

Cyclophosphamide is used in the treatment of severe and life-threatening conditions in pSS patients, e.g. severe renal involvement, vasculitis, mononeuritis multiplex, central nervous system involvement, and mucosa-associated lymphoid tissue (MALT) lymphoma.[54] However, treatment efficacy regarding these extraglandular manifestations has not been assessed in RCTs, and safety data is often not reported in the available pSS studies.

Hemorrhagic cystitis is more frequently seen in patients on oral cyclophosphamide than in patients on IV treatment due to higher cumulative dose.[74] Importantly, even in low doses (1–2 mg/kg body weight), administration of cyclophosphamide is accompanied by significant adverse effects, such as infections, pancytopenia, hair loss, sterility, hemorrhagic cystitis, urinary bladder cancer, development of lymphoma, and skin malignancies. Thus, frequent clinical and biochemical evaluations of patients treated with cyclophosphamide are mandatory. Mercaptoethane sulfonate has been added to cyclophosphamide treatment in patients with rheumatic diseases to prevent adverse effects, but conclusive evidence of its protective effect is lacking.[75]

3.1.4. Azathioprine

Potentially, azathioprine can be useful in pSS patients, analogous to SLE patients. A retrospective case series showed that azathioprine might be effective for the treatment of progressive pulmonary involvement in pSS patients.[76] An RCT in pSS patients showed no significant change in clinical, serological, or histological disease activity variables.[59] In this study, 6 of 25 patients receiving azathioprine withdrew because of AEs. Common adverse effects are leucopenia, abnormal liver biochemistry, and gastrointestinal symptoms.[74] Prolonged use of azathioprine has been associated with increased risk of skin cancer development.[77] Blood cell counts are recommended for every 2 weeks during the first 3 months of treatment and every 2–4 months thereafter.

3.1.5. Mycophenolate mofetil

In pSS, mycophenolate mofetil was evaluated in an open-label pilot trial with follow-up of 24 weeks. Authors reported improvement of subjective glandular and extraglandular manifestations as well as some laboratory parameters.[60] In addition, mycophenolate mofetil might be effective for the treatment of progressive interstitial lung disease in pSS patients.[78] No RCTs have yet been performed to confirm these findings.

Mycophenolate mofetil is associated with an increased risk of infections, gastrointestinal symptoms, bone marrow depression, metabolic changes (e.g. hyperlipidemia, hyperglycemia, hyperuricemia), and impairment of kidney and liver function.[74]

3.2. Biologic DMARD therapies

Several biologic DMARDs have shown promising outcomes in pSS.[3] Unfortunately, for most treatments, only short-term safety data from a limited number of patients are available. As some systemic treatments have been used extensively for other indications, such as rheumatoid arthritis (RA), psoriatic arthritis, and SLE, long-term safety data from worldwide registries are available for these indications. When applicable to pSS, these safety data will be discussed.

An important safety issue during biologic DMARD therapy is systemic infusion-related reactions and injection-site reactions. The presence of antidrug antibodies, raised against these biologics, often related to the nonhuman origin of the biologic DMARDs, may increase the risk of systemic reactions.[79]

The risk of serious infections seems to be higher during biological DMARD therapy than during synthetic DMARD therapy for RA.[80] Therefore, before biological DMARD treatment is started, patients have to be screened for latent or active infections, such as tuberculosis, HIV, and hepatitis B and C. In case of latent infections, adequate prophylactic therapy should be initiated before onset of treatment to prevent reactivation. Furthermore, influenza, pneumococcal, and hepatitis B vaccination should be considered before onset of therapy, and life attenuated vaccinations should be avoided during treatment, in accordance with the European League Against Rheumatism (EULAR) guidelines for vaccination in rheumatic diseases.[81]

Whether patients treated with biologic DMARDs have an increased risk for development of malignancies is still under discussion. This relationship is confounded by the increased risk of hematological malignancies due to chronic inflammation in rheumatic diseases. Compared to conventional synthetic DMARDs, only patients with RA on tumor necrosis factor (TNF) inhibitors have shown an increased risk of nonmelanoma skin cancer.[56] As there are no indications that rituximab is associated with the occurrence of cancers, the American College of Rheumatology (ACR) recommends rituximab treatment for RA patients with treated melanoma and lymphoproliferative malignancies as well as treated solid and nonmelanoma malignancies less than 5 years ago.[82] The long-term data available are insufficient to draw conclusions about an increased risk for malignancies during and after biologic immunosuppressive treatment in pSS. When patients with prior malignancies are treated with biologicals, rheumatologists should therefore be aware of the possibility of the recurrence of these malignancies.

3.2.1. Rituximab

Rituximab therapy (anti-CD20), counteracting the B-cell hyperactivity in pSS, is widely used in the treatment of pSS-related lymphoma, often in combination with cyclophosphamide and prednisone.[83,84] B-cell-depleting therapy is also regularly used off-label for pSS patients with severe extraglandular manifestations. In several populations with moderate-to-high systemic disease activity, rituximab has shown a beneficial effect on systemic disease activity and ESSDAI scores.[85–92] Unfortunately, the effect on ESSDAI was not confirmed in two recent large RCTs.[93,94] Regarding glandular manifestations, rituximab has been reported to improve salivary flow in patients with enough residual gland function.[85,86,95,96] Other studies reported stabilization of exocrine gland function during rituximab treatment, whereas salivary gland function deteriorated in placebo patients.[91,93,94,97] In most studies, patient-reported symptoms such as fatigue and dryness were improved by rituximab treatment,[86,89,91,95–99] but in recent RCTs this effect was smaller or did not differ significantly from the placebo group.[93,94] The differences between the results of these trials are likely explained by differences in the baseline characteristics of the study populations. Safety results of rituximab trials in pSS patients are summarized in Table 1.

Table 1. Safety results of rituximab treatment in pSS.

Author	Design	Follow-up (weeks)	Patients		Serious AEs*		Infectious serious AEs*		Infusion reaction*†		Serum sickness-like AE*		Discontinued due to AEs*
			T	P	T	P	T	P	T	P	T	P	T	P
Meijer [85]	RCT	48	20	10	1 (5)	0 (0)	0 (0)	0 (0)	4 (20)	0 (0)	1 (5)	0 (0)	1 (5)	0 (0)
Devauchelle [94]	RCT	24	63	57	13 (21)	8 (14)	2 (3)	5 (9)	15 (24)	12 (21)	0 (0)**	0 (0)	5 (8)	1 (2)
Dass [89]	RCT pilot	52	8	9	2 (25)	0 (0)	1 (13)	0 (0)	2 (25)	0 (0)	1 (13)	0 (0)	0 (0)	0 (0)
Gottenberg [87]	Registry	152***	78	0	10 (13)		3 (4)		4 (5)		1 (1)		6 (8)
Gottenberg [100]	Retrospective	36***	6	0	0 (0)		0 (0)		1 (17)		1 (17)		1 (17)
Seror [88]	Retrospective	63***	16	0	1 (6)		0 (0)		1 (6)		2 (13)		NR
Carubbi [86]	Open-label	120	19	22††	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)
Pijpe [95]	Open-label	12	15	0	3 (20)		0 (0)		2 (13)		3 (20)		3 (20)
Devauchelle [97]	Open-label	36	16	0	2 (12)		0 (0)		2 (12)		1 (6)		0 (0)
St Clair [98]	Open-label	52	12	0	2 (17)		0 (0)		NR		0 (0)		0 (0)

Only articles with complete safety information and populations that were not reported in other studies were included in this table. * Number of patients who experienced this type of event (percentage). ** 2 patients with treatment-related purpura were reported, but human anti-chimeric antibodies were not measured and the authors did not classify these events as serum sickness-like reactions. *** Median follow-up time. † Infusion reactions include nonserious as well as serious infusion reactions and exclude serum sickness-like reactions. †† Control group received one or more synthetic DMARDs instead of placebo. AE = adverse event. P = placebo group. NR = not reported. T = treatment group.

In a Cochrane review of the safety profile of biologics during treatment of several diseases, rituximab showed the lowest odds for serious infections compared to control treatment.[101] The infectious side effects of rituximab also seem to be mild in pSS, as in RCTs of rituximab treatment in pSS infection rates were comparable between treatment and placebo groups.[85,94] These results were confirmed by a prospective registry of rituximab treatment in systemic autoimmune diseases (AIR registry), in which 78 pSS patients were included with a median follow-up of 34.9 months. According to the AIR registry, the rate of serious infections during rituximab treatment was lower in pSS than in SLE (1.3/100 patient years versus 6.6/100 patient years, respectively).[87,102]

Despite pretreatment with IV corticosteroids, antihistamines, and paracetamol, and co-treatment with oral corticosteroids in some studies, infusion reactions such as fever, rigors, and urticaria occur more commonly during rituximab treatment than during placebo treatment, occurring in 8–25% of patients.[85,89,94,95,97] Specifically, Devauchelle et al. reported a higher incidence of respiratory disorders within 24 hours of injection with rituximab compared to placebo.[94] Gottenberg et al. reported serious infusion reactions in 6.4% of pSS patients in the AIR registry.[87] One study did not report any infusion reactions during long-term treatment with rituximab in 19 patients.[86] However, as no safety analysis plan was included in the methods of this study, it is unclear how infusion reactions were defined.

Patients with active pSS seem to develop serum sickness-like reactions after rituximab treatment more often than patients with other rheumatologic diseases.[88,89,95,100] Serum sickness-like reactions were generally characterized by fever, purpura, arthralgia, myalgia, and sometimes low complement levels and proteinuria. Serum sickness-like reactions were often associated with the development of human anti-chimeric antibodies (HACAs), which support the diagnosis of true serum sickness.[88,95] However, the presence of HACAs was not always reported and in some cases a delayed infusion reaction may have been falsely interpreted as serum sickness. The presence of hypergammaglobulinemia in pSS might explain the higher prevalence of serum sickness in pSS, as it might increase the chance of immune complex deposition. In more recent studies, the prevalence of serum sickness-like reactions was low, probably due to adequate pretreatment with high-dose corticosteroids.[85–87,94]

Progressive multifocal leukoencephalopathy (PML) due to JC virus replication in the brain is a rare but life-threatening condition associated with rituximab treatment.[103] Rheumatologists should be aware of the risk of developing PML during B-cell-depleting treatment of pSS patients, as a case of PML has been reported in a pSS patient.[104]

In RA, vaccination response is decreased by B-cell-depleting therapy, which will probably also be the case during rituximab therapy in pSS.[105] In addition, an open-label study that evaluated the safety of rituximab in 12 patients with pSS reported an exaggerated adverse reaction to pneumococcal vaccination in three out of eight patients who received this vaccination.[98] This might be an extra argument to administer pneumococcal and influenza vaccination prior to onset of rituximab treatment in pSS, although these results have not been confirmed by larger studies of vaccination in pSS.

In conclusion, rituximab has shown promising results regarding efficacy in populations with moderate-to-high systemic disease activity. Rituximab is generally safe in pSS when adequate co-treatment is given and monitoring for infusion reactions takes place. Although patients should be monitored for development of infusion reactions and serum sickness-like disease, these adverse reactions are usually fully reversible. Long-term safety effects of rituximab in pSS are still unclear and should be recorded in prospective registries.

3.2.2. Epratuzumab

Epratuzumab (anti-CD22) targets B-cells. An advantage of epratuzumab above rituximab is that it is fully humanized. An open-label study of epratuzumab in 16 pSS patients showed a beneficial clinical response in 67% of the patients, which was defined as a 20% improvement in two out of four domains: Schirmer score, unstimulated salivary flow, VAS fatigue, and erythrocyte sedimentation rate and/or serum immunoglobulin G (IgG).[106] Two patients discontinued treatment due to infusion reaction and one serious infectious AE occurred. Three patients developed a low level of anti-epratuzumab antibodies, but these were not associated with infusion reactions.

In phase 2 trials of epratuzumab in SLE patients, rates of (serious) AEs and infusion reactions were similar between treatment and placebo arms.[107] In other words, epratuzumab has a good safety profile in SLE at a dose of 360 mg/m², but the tolerability of epratuzumab in pSS has to be confirmed in larger RCTs.

3.2.3. Belimumab

B-cell activating factor (BAFF) blockade by belimumab, a human monoclonal antibody, inhibits survival of autoreactive B-cells and could therefore be beneficial in diseases characterized by B-cell hyperactivity. In SLE, a significant and greater proportion of responders according to the SLE Responder Index in belimumab plus standard therapy versus placebo plus standard therapy has been reported.[108,109] Therefore, belimumab is already registered for use in SLE. The safety of belimumab in SLE is favorable, with low rates of infection, malignancy, and infusion reactions.[108,109] A 7-year follow-up study of belimumab treatment in SLE did not report any additional safety concerns and showed a stable or decreasing rate of AEs and infections during follow-up.[110]

One open-label trial has been performed in 30 pSS patients, showing a beneficial response in 60% of the patients.[111] Although a significant improvement was shown in ESSDAI and ESSPRI scores, the minimal clinically important improvement was not reached for both indices.[8] One serious infection led to discontinuation of treatment, and in another patient breast cancer was diagnosed 3 months after the last infusion. No infusion reactions were reported in the pSS open-label trial. The percentage of patients experiencing AEs was lower in the open-label pSS trial than in SLE trials (54% of pSS patients versus 93% of SLE patients), which might be due to the larger percentage of SLE patients receiving co-treatment with immunosuppressants.[111,112] A 52-week extension study of belimumab treatment in 19 pSS patients did not show any additional serious AEs or infusion reactions.[113]

In summary, belimumab is effective and well tolerated in SLE. The efficacy and safety of belimumab in pSS, although favorable in the open-label trial, should be further investigated in RCTs.

3.2.4. Abatacept

pSS is considered to be a B-cell hyperactivity-mediated disease, but co-stimulation by T-cells is needed for inducing and maintaining B-cell activation. Thus, blockade of T-cell-mediated B-cell hyperactivity is an interesting approach that has to be considered in pSS treatment too. Abatacept is a fully human fusion molecule of the Fc region of IgG with cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), which blocks co-stimulation of B-cells by T-cells. Abatacept is currently registered for use in RA and juvenile idiopathic arthritis (JIA) and has a low odds ratio for serious AEs, serious infections, and withdrawals due to AEs compared to other biologics.[101] Due to the molecular structure of abatacept, its immunogenicity is very low.[114,115] Subcutaneous abatacept has a comparable safety profile as IV abatacept in RA patients.[115]

Two small open-label trials of IV abatacept treatment in pSS have shown that abatacept decreases ESSDAI, ESSPRI, IgG, and rheumatoid factor, improves fatigue and decreases glandular inflammation. In the study by Meiners et al., [116] mild-to-moderate infusion reactions occurred in 40% of the patients in <1 hour of infusion (mostly hypotension and dizziness), and infections occurred in 66% of patients (mostly mild upper respiratory tract infections). Adler et al. did not include a safety analysis plan in their methods, but did report that 27% of the patients experienced an AE, of which one was infectious (diverticulitis).[117]

In conclusion, infectious AEs and mild infusion reactions are relatively common during abatacept treatment in pSS. However, no serious AEs or discontinuations were reported and, in general, abatacept has a good safety profile. Subcutaneous abatacept is currently being studied in an RCT of 88 pSS patients (NCT02067910).

3.2.5. Anakinra

Anakinra is a recombinant IL-1 receptor antagonist preventing activity of IL-1α and IL-1β. Anakinra is used to treat RA, gout, and JIA, among others. In RA, anakinra does not cause a significantly higher number of withdrawals, deaths, AEs, or infections compared to placebo groups, but it does cause a higher prevalence of injection-site reactions.[118] An advantage of anakinra with regard to safety is the short half-life of 6 hours, which allows for prompt discontinuation in the case of AEs.

Because anakinra has shown a beneficial effect on fatigue in RA, the effect of anakinra in pSS on fatigue was studied in an RCT of 26 patients.[119] Anakinra indeed reduced fatigue, but the study was underpowered for the primary outcome measurement and treatment duration was only 4 weeks. As expected, anakinra caused injection-site reactions in a large proportion of pSS patients (54%).[119,120] Due to the small sample size and short follow-up, no definitive conclusions can be made regarding the efficacy and safety of anakinra in pSS. Further study is needed.

3.2.6. IVIG

Intravenous immunoglobulin G (IVIG) is applied as substitution therapy in immunodeficiency syndromes, and as immunomodulating therapy in idiopathic thrombocytopenic purpura and several neurological auto-inflammatory disorders. Furthermore, it has been shown that IVIG decreases disease activity in SLE.[121] Possible severe side effects of IVIG treatment include renal failure, tromboembolic events, and aseptic meningitis, which may be prevented by a slow infusion rate and prehydration.[122] A retrospective study in 19 pSS patients with peripheral neuropathy reported a beneficial effect of monthly courses of IVIG on SS-associated sensorimotor neuropathy and non-ataxic sensory neuropathy.[123] Tolerance of IVIG was good. During the median treatment duration of 7 months, only one withdrawal due to an AE (nausea) was reported and no serious AEs occurred. Case reports have also suggested efficacy of IVIG in SS-associated thrombocytopenia, central nervous system involvement, and congenital heart block, but no safety analysis was included.[124–126] In summary, IVIG is a promising treatment for certain systemic manifestations of pSS and seems to be well tolerated, but should be further evaluated in prospective trials.

3.2.7. Baminercept

Baminercept is an inhibitor of the lymphotoxin-β pathway. Although baminercept treatment did reduce the IFN signature in RA patients, baminercept treatment showed disappointing clinical results.[127] As the lymphotoxin pathway might play an important role in lymphoid tissue organization and chronic inflammation in pSS, baminercept was recently retested in an RCT in 52 pSS patients.[128] Unfortunately, baminercept again did not improve exocrine gland function, fatigue, pain, or sicca symptoms. The ESSDAI score was slightly improved in the baminercept group, but the mean ESSDAI change from baseline of 1.6 points is below the minimally clinical important improvement of 3 points.[8] Furthermore, transaminase abnormalities occurred more often in the baminercept group, and seven serious AEs occurred, including two patients with grade 3 hepatic injury. Therefore, the benefits of baminercept treatment in pSS do not seem to outweigh possible safety concerns.

3.2.8. Anti-TNF

Two anti-TNF biologic DMARDs have been studied in pSS. In a pilot RCT by Sankar et al., 14 pSS patients were treated with etanercept and 14 patients with placebo.[129] The effect of infliximab on pSS was studied in an RCT of 103 patients.[130] Unfortunately, both trials did not show a significant difference between the study drug and placebo treatment. The number of treatment discontinuations due to AEs was somewhat higher in the anti-TNF groups than in the placebo groups. Interestingly, in the study by Mariette et al., serum immunoglobulin levels were significantly increased during infliximab treatment.[130] Impaired control of the IFNα pathway and subsequent BAFF overexpression by TNF inhibition might explain the inefficacy of this group of biologic DMARDs in pSS.[131]

4. Expert opinion

Symptomatic treatment, preventive measures, and patient’s education are of great importance in the management of pSS patients in daily practice and were, for decades, the only treatment modality to reduce SS-related complaints. However, although symptomatic treatment is safe and has little AEs, commonly only short-term symptomatic relief is achieved and this treatment does not protect patients from persistent disease activity or organ damage. Therefore, there is a need for registration of systemic therapies for the treatment of pSS.

There is limited evidence in the available literature for effectiveness of systemic conventional DMARD therapy in pSS patients. Furthermore, the knowledge about AEs and drug toxicity of conventional DMARDs in pSS is limited and often based on expert opinion. Toxicity of conventional DMARDs in pSS patients does not appear to be different compared to patients with other autoimmune diseases for which these drugs are used. Based on the weak evidence of efficacy and high rate of AEs of conventional DMARDs, these agents should not be used in pSS routinely. Importantly, RCTs on conventional DMARDs were performed in small groups, with heterogeneous patient populations, without uniform end points and adequate measuring instruments. There is a great need for well-designed and large RCTs to assess the value of conventional DMARDs in pSS with regard to glandular and extraglandular manifestations, patient-reported and systemic disease activity, and organ-specific outcomes. Safety issues should take an important place in the analysis of these RCTs.

Recent years have shown a fast development regarding biologic systemic therapies for pSS. Several studies on biologic DMARD treatment of pSS patients have shown promising efficacy and safety results, especially in populations with high baseline systemic disease activity. In summary, the most frequent AEs during biologic therapy in pSS patients are infusion and injection reactions and infections, as in other rheumatologic diseases. The risk of infection does not seem to be high in pSS; in fact, it is often comparable to the infection risk in patients on placebo; patients with pSS have an inherent increased risk for developing infections and thus appropriate preventative actions should be taken. Infusion reactions occur most often when chimeric DMARDs are used, such as rituximab, necessitating pre- and co-treatment with corticosteroids and antihistamines.

As most biologic DMARDs have only been evaluated in phase 2 trials so far, only limited safety data are available, specifically for pSS, making it necessary to deduce information from safety analyses of biologic therapy for other indications. However, patients with other rheumatologic diseases more often use concomitant immunosuppressive therapy than patients with pSS, which may influence the risk of infection and other AEs. Furthermore, patients with pSS might respond differently to certain medications than patients with other rheumatic diseases. An example of this is the increased risk of serum-sickness like reactions seen during some studies of rituximab therapy for pSS.[88,89,95] Therefore, safety analyses should be performed specifically in pSS patients.

This review of the literature taught us that many questions regarding drug safety in pSS remain unanswered. Larger numbers of treated patients with a longer follow-up are needed to investigate serious AEs with a low prevalence. Further study is needed to determine the specific dosage of systemic drugs for pSS in which the benefits/harms ratio is maximal. Furthermore, for most drugs, the long-term safety is still largely unknown. To answer these questions, larger RCTs are needed and off-label treatment of pSS patients should be registered in cohort studies.

New outcome parameters such as the ESSDAI and ESSPRI, with a stronger sensitivity to change, have made it easier to monitor the efficacy of DMARDs in pSS treatment. Importantly, these indices have made it possible to compare the efficacy of different therapies. Unfortunately, for safety analyses, a multitude of methods is still used in various trials, which makes it difficult to directly compare the results of different studies and drugs. For an example, authors might use different definitions of AEs, but do not always report which definition they used. Furthermore, it should be clear if AEs are investigated only by open-ended questioning or by specific screening for certain AEs (e.g. by asking about it or by laboratory analysis). In addition, there are large differences in the data that authors choose to report. Some authors only report serious AEs, or only AEs of a certain type, while others provide a complete list of coded AEs. Therefore, we urge authors of future therapeutic studies in pSS to follow the extension of the Consolidated Standards of Reporting Trials (CONSORT) statement, which describes guidelines for the reporting of harms in randomized trials.[132]

In conclusion, the available data suggest that the safety of symptomatic treatment is very good and symptomatic treatment should be applied in all pSS patients. However, systemic therapy is necessary to achieve long-term relief and for the prevention of organ damage and exocrine gland dysfunction. Therefore, further evaluation of the effectiveness and AEs of systemic DMARDs in pSS is important. Conventional DMARDs have not shown much efficacy in earlier studies, and their benefits do not seem to weigh up to the risk of AEs. However, methodological problems may have influenced these results. As biologic DMARDs cause mostly mild AEs in pSS, and show promising results regarding efficacy, the benefits of biologic DMARD therapy seem to outweigh possible safety concerns.

Article highlights

• pSS is a disabling autoimmune disease, affecting exocrine glands and several organs, leading to impaired health-related quality of life. Currently, no effective systemic treatments have been registered for pSS.

• Although symptomatic treatment is tolerated well, only short-term relief is achieved and this treatment does not protect patients from persistent disease activity or organ damage.

• Synthetic DMARD therapies have not shown sufficient efficacy in previous trials. Although the safety of synthetic DMARDs does not seem to be different in pSS compared to patients with other autoimmune diseases, the beneficial effect does not outweigh the possible harms.

• Biologic DMARDs have shown promising results in populations with high systemic activity, and seem to have a good safety profile in patients with pSS. However, larger trials and registries are needed to confirm these results.

• When evaluating new therapies, safety evaluation should take place according to the Consolidated Standards of Reporting Trials (CONSORT) statement.

This box summarizes key points contained in the article.

Declaration of Interest

A Vissink has received unrestricted grants from Roche and Bristol-Meyers Squibb. H Bootsma has received unrestricted grants from Roche and Bristol-Meyers Squibb and served on the advisory board of Union Chimique Belge and GlaxoSmithKline. E Brouwer received unrestricted grants from Pfizer, MSD and Abbott. The authors have no other 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 apart from those disclosed