Ramos-Casals M, Tzioufas AG, Stone JH, Sisó A, Bosch X. Treatment of Primary Sjögren SyndromeA Systematic Review. JAMA. 2010;304(4):452–460. doi:10.1001/jama.2010.1014
Author Affiliations: Sjögren Syndrome Research Group (AGAUR), Josep Font Laboratory of Autoimmune Diseases, IDIBAPS, Department of Autoimmune Diseases (Dr Ramos-Casals), and Department of Internal Medicine, Instituto Clínico de Medicina y Dermatologia (Dr Bosch), Hospital Clínic, Barcelona, Spain; Department of Pathophysiology, School of Medicine, University of Athens, Athens, Greece (Dr Tzioufas); Division of Rheumatology, Allergy, and Immunology, Massachusetts General Hospital, Boston (Dr Stone); and Primary Care Research Group, IDIBAPS, Primary Care Center Les Corts, GESCLINIC, Barcelona (Dr Sisó).
Context A variety of topical and systemic drugs are available to treat primary Sjögren syndrome, although no evidence-based therapeutic guidelines are currently available.
Objective To summarize evidence on primary Sjögren syndrome drug therapy from randomized controlled trials.
Data Sources We searched MEDLINE and EMBASE for articles on drug therapy for primary Sjögren syndrome published between January 1, 1986, and April 30, 2010.
Study Selection Controlled trials of topical and systemic drugs including adult patients with primary Sjögren syndrome were selected as the primary information source.
Results The search strategy yielded 37 trials. A placebo-controlled trial found significant improvement in the Schirmer and corneal staining scores, blurred vision, and artificial tear use in patients treated with topical ocular 0.05% cyclosporine. Three placebo-controlled trials found that pilocarpine was associated with improvements in dry mouth (61%-70% vs 24%-31% in the placebo group) and dry eye (42%-53% vs 26%). Two placebo-controlled trials found that cevimeline was associated with improvement in dry mouth (66%-76% vs 35%-37% in the placebo group) and dry eye (39%-72% vs 24%-30%). Small trials (<20 patients) found no significant improvement in sicca outcomes for oral prednisone or hydroxychloroquine and limited benefits for immunosuppressive agents (azathioprine and cyclosporine). A large trial found limited benefits for oral interferon alfa-2a. Two placebo-controlled trials of infliximab and etanercept did not achieve the primary outcome (a composite visual analog scale measuring joint pain, fatigue, and dryness); neither did 2 small trials (<30 patients) testing rituximab, although significant results were observed in some secondary outcomes and improvement compared with baseline.
Conclusions In primary Sjögren syndrome, evidence from controlled trials suggests benefits for pilocarpine and cevimeline for sicca features and topical cyclosporine for moderate or severe dry eye. Anti–tumor necrosis factor agents have not shown clinical efficacy, and larger controlled trials are needed to establish the efficacy of rituximab.
Quiz Ref IDSjögren syndrome is a systemic autoimmune disease that presents with sicca symptomatology of mucosal surfaces,1 mainly dry mouth and dry eyes. There is often systemic involvement (extraglandular manifestations) and lymphoma is a recognized complication.2 Sjögren syndrome is one of the most prevalent autoimmune diseases (with an estimated 0.5 million to 3 million affected persons in the United States,3 primarily perimenopausal women). When sicca symptoms appear in a previously healthy person, this is classified as primary Sjögren syndrome.4
Standard management focuses on controlling sicca features using substitute topical agents, and extraglandular features are managed with glucocorticoids and immunosuppressive drugs.5 However, there are no evidence-based therapeutic guidelines for the management of primary Sjögren syndrome. Consequently, current therapeutic decisions are based on a mix of personal experience, expert opinion, and reported studies.
The purpose of this systematic review was to analyze and summarize the evidence on drug therapies (topical and systemic) for the main clinical manifestations of primary Sjögren syndrome (sicca symptoms and extraglandular involvement) in adults.
We searched MEDLINE and EMBASE using the MeSH term Sjögren's syndrome and subheading therapy with these restrictions: language (English), date (January 1, 1986, to April 30, 2010), study participants (humans), and age (adults).
Studies were eligible when the study population included adults with primary Sjögren syndrome and the intervention consisted of a drug therapy (nondrug therapeutic interventions were excluded) (eTable 1). Eligible studies had to be randomized controlled trials or prospective cohort studies (reviews, experimental studies, duplicate publications, retrospective cohort studies, case-control studies, and case reports were excluded). Studies also had to contain sufficient, clear information on the effect of the drug on clinical outcomes to be included. The control intervention could include placebo or standard therapy.
Two of us (M.R.C. and A.S.) read the titles and abstracts (if available) identified by the search and selected potentially eligible studies. Three of us (M.R.C., A.S., X.B.) fully reviewed the selected studies to determine criteria fulfillment. Disagreements between the 3 authors were discussed with the other 2 authors (A.G. and J.S.) until consensus was reached. Study authors were contacted when necessary (M.R.C.). Reference lists of relevant articles were also searched.
Three of us (M.R.C., A.S., X.B.) extracted the data independently. The data were entered into a database (M.R.C.), and the remaining 4 authors checked it.
The following variables were abstracted: first author, year of publication, population studied, number of patients, mean age and range, sex, subpopulations included (primary or associated Sjögren syndrome, sicca syndrome), classification criteria for primary Sjögren syndrome, study design and duration, drug, control population, primary and secondary outcomes, statistical comparisons, and adverse events.
Three of us (M.R.C., A.S., X.B.) assessed methodological quality of the selected trials according to the Cochrane Collaboration tool for assessing risk of bias,6 documenting the method of sequence generation, allocation concealment, blinding of participants, personnel and outcome assessors, incomplete outcome data, selective outcome reporting, and other sources of bias such as baseline imbalance or sources of bias related to the specific study design. Disagreements between the 3 authors were discussed with the other 2 authors (A.G. and J.S.). Assessment of study heterogeneity is detailed in eSupplement.
The search strategy yielded 37 controlled trials (5 with a crossover, washout design)7- 43 and 19 prospective cohort studies (Figure).44- 62
Sixteen trials included only primary Sjögren syndrome patients; 12 included patients with primary or associated Sjögren syndrome (7 detailed the types of patients, 263 of whom were primary and 136 associated); and 9 trials included patients with dry eye disease (including Sjögren syndrome), of which only 3 detailed how many Sjögren syndrome patients were included (389 of 1566 patients). In 5 trials,13,23,24,33,43 the epidemiological profile of participants was not typical of primary Sjögren syndrome (mean age, <50 years; female to male ratio, <8:1).
Six trials did not use a placebo,8,9,13- 15,30 and the drug tested was compared with standard therapy (artificial tears in 3 studies) or with other therapies (sucralfate, topical ocular diclofenac, punctal lachrymal occlusion). Therapies evaluated included eyedrops (topical nonsteroidal anti-inflammatory drugs [NSAIDs], glucocorticoids, or cyclosporine), oral sialogogues (pilocarpine or cevimeline), immunomodulatory and immunosuppressive drugs, and biologic therapies.
A variety of outcomes were defined (eTable 2). Primary and secondary outcomes were clearly specified in only 12 trials. The duration of included trials varied from 4 weeks to 1 year. When specific primary outcomes were not defined, we summarized the statistically significant results. A detailed evaluation of the methodological quality of trials is shown in eTable 3.
Eyedrops. One controlled trial9 compared 2 NSAIDs (0.1% diclofenac vs 0.1% indomethacin) and found a greater reduction in corneal sensitivity in the diclofenac group (43.5 mm vs 52.0 mm, P = .01). Two controlled trials and 1 prospective study evaluated different topical glucocorticoids. The first trial8 found that patients treated with fluorometholone had lower dry eye symptom severity scores in comparison with flurbiprofen (P = .03) and artificial tears (P = .03), and lower rose bengal and fluorescein staining scores in comparison with the flurbiprofen group (P = .046 and P = .02, respectively). The second trial7 found no significant differences between 0.5% loteprednol etabonate and placebo in the primary outcome of a combined corneal staining score (percent change at 4 weeks from baseline score, −5.9 vs 0.0, P > .05). The prospective study44 found significant improvement in ocular tests scores with respect to baseline after treatment with topical 1% methylprednisolone, although severe dry eye symptoms recurred in 21% of patients.
Three placebo-controlled trials assessed topical cyclosporine in 1451 patients with moderate or severe dry eye disease (Table 1). The largest trial10 tested 2 doses (0.05% and 0.1%) and found significant improvement in Schirmer test scores for both groups (P < .007) but improvement in corneal staining scores only in the 0.05% group (P = .008; P = .06 in the 0.1% group); the 0.05% group showed a significant improvement in 1 of 7 symptoms of ocular discomfort (blurred vision) evaluated using a 5-point scale (P < .01) and a significant decrease in artificial tear use (P = .006). A 12-month extension of this trial using the 0.1% dose found no additional improvement in these outcomes.11 A controlled trial12 tested 4 doses (0.05%, 0.1%, 0.2%, and 0.4%) and found no linear dose-response results. The best results were obtained in the 0.1% and 0.05% groups.
Three controlled trials compared 0.05% cyclosporine with other therapies. Kim et al13 found significant improvement in subjective evaluation of dry eye symptoms, tear film break-up time, and Schirmer test scores in 150 patients treated either with 0.05% cyclosporine or with 0.05% retinyl palmitate in comparison with artificial tears. There were no differences between cyclosporine and retinyl palmitate. Sall et al14 found a better dry eye symptom composite (P = .02) and corneal staining scores (P = .005) for the combination of cyclosporine and glycol-based tears compared with cyclosporine and standard artificial tears in 60 patients. Roberts et al15 compared the use of 0.05% cyclosporine, punctal occlusion, and the 2 therapies combined in 30 patients; there were no differences between the cyclosporine/punctal occlusion combination and cyclosporine alone in Schirmer test (3.9 vs 3.0 mm over 3 minutes) and rose bengal staining scores (mean change, −1.0 vs 0.9) and less daily artificial tear use (3.9 vs 3.2 fewer uses per day).
A 6-month prospective study45 showed statistically significant improvements from baseline in subjective and objective measures of dry eye (P < .001) after treatment with topical 0.05% cyclosporine. The largest trial10 found a significantly higher percentage of total adverse events in the 0.1% group (relative risk [RR], 1.28; 95% confidence interval [CI], 1.08-1.51) but not in the 0.05% group (RR, 1.17; 95% CI, 0.98-1.40) compared with placebo.
A placebo-controlled trial16 evaluated 2 doses (1% and 2%) of topical ocular diquafosol, an agonist of the purinergic P2Y2 receptor, in 527 patients and found better corneal staining scores (mean, 0.81 and 0.83, respectively, vs 0.94 in placebo; P < .05) but not improved clearing of foreign body sensation.
Sialogogues.Three placebo-controlled trials evaluated oral pilocarpine in 673 patients (Table 2). The largest trial17 evaluated 2 doses (2.5 and 5 mg every 6 hours) and found a higher frequency of improvement in dry mouth (61% vs 31%, P < .001) and dry eye (42% vs 26%, P = .009) in the 5-mg group but not in the 2.5-mg group. Papas et al18 conducted a dose-escalating trial (from 5 to 7.5 mg every 6 hours) and also found a higher frequency of improvement in dry mouth (61% vs 31%, P < .001) and dry eye (53% vs 26%, P < .001), while Wu et al19 found a similar improvement in dry mouth (70% vs 24%, P = .003) using a dose of 5 mg every 6 hours. The largest trial17 found a higher frequency of sweating (RR, 2.24; 95% CI, 1.82-2.76) and increased urinary frequency (RR, 1.77; 95% CI, 1.38-2.28) compared with placebo. In the dose-escalating trial,19 23% of patients switched from 7.5 to 5 mg every 6 hours because of adverse effects. A prospective study46 used a lower dose (5 mg every 12 hours) and found greater improvement in subjective assessment of dry eye symptoms compared with artificial tears (P = .001) or punctal occlusion (P = .05).
Four placebo-controlled trials evaluated oral cevimeline in 376 patients (Table 2). The largest trial20 evaluated 2 doses (15 and 30 mg every 8 hours) and found a higher frequency of improvement in dry mouth (45% in the 15-mg group and 66% in the 30-mg group, vs 37% in the placebo group; P = .006 and P < .001, respectively), dry eye (31% and 39% vs 24%, P > .05 and P = .04, respectively), and overall dryness (32% and 66% vs 36%, P > .05 and P < .001, respectively). The second trial21 evaluated 2 doses (30 and 60 mg every 8 hours) and found a higher frequency of improvement in dry mouth (76% in the 30-mg group and 67% in the 60-mg group compared with 35% of patients in the placebo group, P = .004 and P = .02, respectively), dry eye (72% and 52% vs 30%, P = .007 and P = .10), and overall dryness (P = .004 and P = .03). The third trial22 found significant differences in subjective sicca symptoms in the 20-mg group but not in the 30-mg group, while the fourth trial, which tested 30 mg of cevimeline every 8 hours using a crossover design,23 found no significant results.
The 2 latter trials included 104 patients compared with 272 included in the 2 larger trials. The 2 largest trials showed a significantly higher frequency of nausea (RR, 1.68; 95% CI, 1.18-2.40; P = .02) and sweating (RR, 2.16; 95% CI, 1.65-2.82) in the 30-mg group compared with placebo,20 and a higher frequency of nausea (RR, 2.77; 95% CI, 1.79-4.28), sweating (RR, 3.00; 95% CI, 1.70-5.28), and rigors (RR, 1.92; 95% CI, 1.28-2.87) in the 60-mg group in comparison with placebo.21
A prospective study47 identified greater baseline stimulated whole saliva (P < .001), lesser lymphocytic infiltration of salivary gland biopsy (P = .003), and less-advanced sialography stage (P = .004) as independent predictors of stimulated whole saliva after 4 weeks of cevimeline.
Immunomodulatory or Immunosuppressive Drugs. One controlled trial24 compared oral prednisone (30 mg per day) with piroxicam (20 mg per day) and placebo (8 patients each group). There was no difference between groups in salivary flow rate (SFR), Schirmer test, rose bengal staining score, or histopathological focus score. A prospective study48 found that glucocorticoids did not influence progressive SFR worsening in 60 patients with primary Sjögren syndrome. However, another prospective study in 20 patients found that oral prednisolone increased SFR.49
In 1993, Kruize et al25 conducted a 2-year crossover trial using 400 mg of hydroxychloroquine per day in 19 patients and found no significant differences in hydroxychloroquine vs placebo for sicca symptoms, parotid enlargement, fatigue, myalgia, and arthralgia and no significant differences in ocular tests. A prospective study of hydroxychloroquine50 in 14 patients found no effects on sicca symptoms and fatigue but significant differences in laboratory tests, including erythrosedimentation rate (27.2 vs 43.1 mm per hour at baseline, P < .05), C-reactive protein (2.1 vs 3.4 mg/dL, P < .05), and IgG levels (1560 vs 2350 mg/dL, P < .05). No retinal toxicity or severe adverse events were reported in either study.
Two placebo-controlled trials evaluated azathioprine26 and oral cyclosporine27 in 13 and 20 patients, respectively. The first trial26 found no significant differences in any outcome. The second trial found a higher rate of improvement in xerostomia in patients treated with cyclosporine (80% vs 20%, P < .01) without significant differences in the Schirmer test score (5.2 vs 5.0 mm) and SFR (1.14 vs 1.06 mL per 5 minutes). Three prospective studies evaluated the use of methotrexate,51 leflunomide,52 and mycophenolic acid53 and found limited improvements in sicca symptoms. A common finding in all the studies was the high rate of adverse events (31% for azathioprine, 41% for methotrexate, 60% for cyclosporine, 63% for mycophenolic acid, and 100% for leflunomide). One controlled trial of thalidomide was stopped early because of an excess of adverse events.28
Three controlled studies evaluated oral interferon alfa-2a (150 IU daily). A small, controlled trial (12 patients) suggested a beneficial effect on unstimulated SFR and ocular or oral dryness,29 while a single-blinded, sucralfate-controlled trial30 found a significant time-dependent increased production of whole saliva at 3 months but not at 6 months. In contrast, a large placebo-controlled trial including 497 patients31 found significant improvement in only 1 of 28 outcomes evaluated (unstimulated whole saliva, P = .01) and a higher percentage of adverse events (40% vs 25% in the placebo group, P < .001).
Biologic Agents. A placebo-controlled trial32 evaluated infliximab in 103 patients and found no significant differences in the primary outcome (≥30% improvement in 2 of 3 visual analog scales [VAS] measuring joint pain, fatigue, and dryness) (17% vs 20%, P = .62). No significant differences were found for the secondary outcomes (Table 3). In contrast, a previous prospective study in 16 patients found significant improvements in subjective and objective sicca measures.54
A placebo-controlled trial34 evaluated etanercept in 28 patients and found no significant differences in the primary outcome (≥20% improvement in the values on 2 of 3 domains: oral, ocular, and laboratory) (36% vs 21%, P = .20). No significant differences were found for the secondary outcomes. Similar negative results were found in a prospective study in 15 patients.58
Two placebo-controlled trials evaluated the use of rituximab (two 1000-mg doses 15 days apart) (Table 3). The first 48-week trial33 included 30 patients with stimulated SFR at 0.15 mL or greater per minute and achieved the primary outcome (improvement of stimulated SFR) at 12 weeks (0.87 vs 0.28 mL/minute in the placebo group, P = .04) but not at the end of the study. Only the VAS score for dry eye significantly improved at 48 weeks (46 vs 76, P < .05) while other secondary outcomes improved at different study time points but not at 48 weeks. The second trial35 including 17 patients with a baseline fatigue VAS score greater than 50 did not achieve the primary outcome (≥20% improvement in VAS fatigue score with respect to placebo) (87% vs 56%, P = .36). The 2 trials included a premedication dose of 100-mg methylprednisolone and a short course (<15 days) of high-dose oral prednisone in both the rituximab and placebo groups; concurrent medication such as immunosuppressive agents was only permitted in the second trial.35 Two prospective studies found significant improvements in sicca and general symptoms compared with baseline values.56,57
A prospective study in 16 patients treated with epratuzumab found significant improvements in fatigue and subjective patient and physician assessments.55
Miscellanea. A placebo-controlled trial tested oral N-acetylcysteine in 26 patients and found significant improvement in the van Bijsterweld score and dry eye and mouth symptoms compared with baseline, although specific results were not detailed.36
Several drugs have been tested in placebo-controlled trials (rebamipide, doxycycline, dehydroepiandrosterone, and lamivudine)37- 43 and prospective studies (nizatidine, zidovudine, mizoribine, and D-penicillamine)59- 62 without significant benefits.
This systematic review reveals a very low level of evidence for the majority of drugs currently used in primary Sjögren syndrome. In addition, less than 10% of the studies were designed to compare different drugs in the same clinical scenario. However, some recommendations on drug therapy can be proposed for the management of the main symptoms.
Quiz Ref IDFrequent use of preservative-free tear substitutes are recommended, while ocular lubricating ointments are usually reserved for nocturnal use.2,4,5 Controlled trials support the use of topical 0.05% cyclosporine twice daily for patients with moderate to severe dry eye disease, although the largest trials did not include predominantly patients with primary Sjögren syndrome.10,12 However, in one study, no further improvements were reported after 6 additional months of treatment.11Quiz Ref IDPatients with severe refractory ocular dryness may require the addition of topical NSAIDs or glucocorticoids, although these should only be prescribed by ophthalmologists for the minimum time necessary, because of adverse events associated with long-term use.63
Quiz Ref IDSaliva replacement products and sugar-free chewing gums may be effective for mild to moderate dry mouth. Alcohol and smoking should be avoided and thorough oral hygiene is essential.2,4,5 For patients with residual salivary gland function, oral pilocarpine and cevimeline are the treatment of choice. However, the efficacy of the 2 drugs has not been compared. The doses that best balance efficacy and adverse effects are 5 mg every 6 hours for pilocarpine and 30 mg every 8 hours for cevimeline. In patients with contraindications or intolerance to muscarinic agonists, N-acetylcysteine may be an alternative.
No clear benefit from hydroxychloroquine for general symptoms (muscle and joint pain, fatigue) was reported by controlled and prospective studies (all had small sample sizes); its use is only supported by retrospective studies.64,65 The off-label use of biologic agents to treat general symptoms only, even when severe, is not warranted at present.
There is limited evidence on the use of glucocorticoids and immunosuppressive agents, since controlled and prospective studies were small and were specifically designed to evaluate sicca features. Rituximab has shown improvement in some extraglandular features in a recent controlled trial (vasculitis) and 3 uncontrolled studies (vasculitis, neuropathy, glomerulonephritis, and arthritis).66- 68 However, while awaiting the results of larger trials, rituximab may be considered as a rescue therapy in patients refractory to standard treatment.69
Treatment of severe, life-threatening involvement has rarely been detailed and at present, there are only a few retrospective studies and isolated case reports. Quiz Ref IDHowever, this scanty evidence, taken together with expert review, suggests that methylprednisolone and cyclophosphamide pulses, possibly together with plasma exchanges, should be used in patients with rapidly progressing extraglandular features (glomerulonephritis, neuropathy, interstitial lung disease, or myelitis) or with severe systemic vasculitis.2,4,5 Rituximab is increasingly used in life-threatening situations and cases of B-cell lymphoma.69
In the last 3 decades, therapeutic approaches in primary Sjögren syndrome have been based on the use of substitute agents for sicca features and glucocorticoids and immunosuppressive agents for extraglandular involvement. The emergence of new immunosuppressive agents and biologic therapies has increased the therapeutic armamentarium available in the most severe situations, but their use is limited by the lack of specific licensing.
This systematic review highlights the limited evidence available for the drugs most frequently used in primary Sjögren syndrome and the difficulties of offering solid therapeutic recommendations. There is even less scientific evidence on the treatment of patients who do not respond to first-line therapies. International efforts are required to collect and characterize large multicenter cohorts of patients70 and develop consensual end points for homogeneous evaluation of the main outcomes.71 B-cell targeted agents seem to be the most promising future therapy, as suggested by recent preliminary studies,72,73 especially rituximab, which has been used in more than 100 reported cases. However, 2 recent controlled trials did not achieve the primary outcome. Agents that block BAFF (B cell–activating factor of the tumor necrosis factor family) may also be a promising therapy.74 Advances in knowledge of the molecular mechanisms involved in the etiopathogenesis of primary Sjögren syndrome may allow the development of more effective, highly selective therapies without the adverse effects often associated with standard, less-selective drugs.
Corresponding Author: Manuel Ramos-Casals, MD, PhD, Department of Autoimmune Diseases, Hospital Clínic, C/Villarroel, 170, 08036 Barcelona, Spain (email@example.com).
Author Contributions: Dr Ramos-Casals had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Ramos-Casals, Tzioufas, Stone.
Acquisition of data: Sisó, Bosch.
Analysis and interpretation of data: Sisó, Bosch.
Drafting of the manuscript: Ramos-Casals, Tzioufas, Sisó, Bosch.
Critical revision of the manuscript for important intellectual content: Tzioufas, Stone, Bosch.
Statistical analysis: Ramos-Casals, Stone, Sisó.
Obtained funding: Ramos-Casals.
Administrative, technical, or material support: Ramos-Casals.
Study supervision: Ramos-Casals, Stone, Sisó, Bosch.
Financial Disclosures: None reported.
Funding/Support: This study was supported by grants from La Marató de TV3 (071810) and Fondo de Investigaciones Sanitarias (080103).
Role of the Sponsor: The funding sources had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.
Additional Contributions: David Buss, BA, an independent medical writer, provided editorial assistance. His compensation was funded by grants from La Marató de TV3 and Fondo de Investigaciones Sanitarias.