Article Text

Download PDFPDF

Extended report
Efficacy and safety of belimumab in primary Sjögren's syndrome: results of the BELISS open-label phase II study
  1. Xavier Mariette1,
  2. Raphaèle Seror1,
  3. Luca Quartuccio2,
  4. Gabriel Baron3,
  5. Sara Salvin2,
  6. Martina Fabris2,4,
  7. Frederic Desmoulins1,
  8. Gaétane Nocturne1,
  9. Philippe Ravaud3,
  10. Salvatore De Vita2
  1. 1Université Paris-Sud, Assistance Publique-Hôpitaux de Paris (AP-HP), Hôpitaux Universitaires Paris-Sud, INSERM U1012, Le Kremlin Bicêtre, France
  2. 2Department of Medical and Biological Sciences, Clinic of Rheumatology, Udine University, Udine, Italy
  3. 3Faculty of Medicine, AP-HP, Hôtel Dieu Hospital, Centre of Clinical Epidemiology, INSERM, U738, University of Paris Descartes, Paris, France
  4. 4Institute of Clinical Pathology, University Hospital of Udine, Udine, Italy
  1. Correspondence to Dr Xavier Mariette, Service de Rhumatologie, Hôpital Bicêtre, 78 rue du Général Leclerc, Le Kremlin Bicêtre 94275, France; xavier.mariette{at}


Background Increased expression of B cell activating factor (BAFF or B lymphocyte stimulator) may explain the B cell activation characteristic of primary Sjögren's syndrome (pSS).

Objectives To evaluate the efficacy and safety of belimumab, targeting BAFF, in patients with pSS.

Methods Patients were included in this bi-centric prospective 1-year open-label trial if they fulfilled American European Consensus group criteria, were anti-Sjögren's syndrome A-positive and had current systemic complications or salivary gland enlargement, or early disease (<5 years), or biomarkers of B cell activation. They received belimumab, 10 mg/kg, at weeks 0, 2 and 4 and then every 4 weeks to week 24. The primary end-point, assessed at week 28, was improvement in two of five items: reduction in ≥30% in dryness score on a visual analogue scale (VAS), ≥30% in fatigue VAS score, ≥30% in VAS pain score, ≥30% in systemic activity VAS assessed by the physician and/or >25% improvement in any B cell activation biomarker values.

Results Among 30 patients included, the primary end-point was achieved in 18 (60%). The mean (SD) European League Against Rheumatism (EULAR) Sjögren's Syndrome Disease Activity Index decreased from 8.8 (7.4) to 6.3 (6.6) (p=0.0015) and EULAR) Sjögren's Syndrome Patient Reported Index from 6.4 (1.1) to 5.6 (2.0) (p=0.0174). The mean dryness, fatigue and pain VAS varied from 7.8 (1.8) to 6.2 (2.9) (p=0.0021), 6.9 (1.8) to 6.0 (2.2) (p=0.0606) and 4.6 (2.6) to 4.7 (2.4) (p=0.89), respectively. Salivary flow and Schirmer's test did not change.

Conclusions These encouraging results justify future randomised controlled trials of belimumab in a selected target population of pSS patients most likely to benefit from treatment.

Statistics from

Primary Sjögren's syndrome (pSS) is a systemic auto-immune disease that is characterised by mouth and eye dryness, fatigue and pain. Systemic manifestations occur in 20%–40% of patients.1 ,2 B lymphocytes play a crucial role in the pathogenesis of the disease by their capacity to secrete autoantibodies and cytokines or present auto-antigens. Evidence-based therapy for pSS is largely limited to symptomatic treatments that alleviate the dryness.3 Some efficacy of rituximab, an anti-CD20 monoclonal antibody, was suggested in retrospective open studies, in one registry4 and in one small controlled trial5 but was not confirmed in a recent larger controlled trial of 120 patients.6

Increased expression of the cytokine B cell activating factor (BAFF), also called B lymphocyte stimulator (BLyS),7 may explain pathogenic B cell activation in several systemic autoimmune diseases and lymphomas. BAFF plays a crucial role in B cell maturation, plasma cell survival, antibody response promotion and immunoglobulin-class switch recombination.7 ,8 Its involvement in the pathogenesis of autoimmune diseases is demonstrated by BAFF transgenic mice showing autoimmune diseases mimicking systemic lupus erythematosus (SLE) and pSS and a rate of B cell lymphoma twice that of control mice.8 In humans, patients with SLE and pSS have elevated serum levels of BAFF,9 ,10 and some studies report correlations between serum levels of BAFF and autoantibodies.10–12 Likewise, high serum levels of BAFF have been found associated with pSS-associated lymphoproliferative complications13 ,14 and might be also implicated in the resistance to B cell-targeted therapies.15

Inhibition of BAFF by TACI-Fc local gene therapy has been found to reduce salivary inflammation in NOD mice, a mouse model of pSS.16 Belimumab, a monoclonal anti-BAFF (BLyS) antibody, was recently approved for treatment of SLE based on two phase III studies with positive results.17 ,18 Because pathophysiological studies suggested an involvement of BAFF in the pathogenesis of pSS and in SLE, we conducted the first open-label proof of concept study to evaluate the efficacy and safety of belimumab in pSS.

Patients and methods


Patients were included in two identical studies conducted at the same time in two European centres, Paris, France and Udine, Italy (clinical trial registration numbers: NCT01160666 and NCT01008982). Patients had to fulfil the American European Consensus group criteria for pSS,19 be positive for anti-Sjögren's syndrome A (SSA) or anti-Sjogren's syndrome B (SSB) antibodies and have at the time of inclusion at least one of the three following characteristics: systemic complications (defined as the presence of polysynovitis, vasculitis, auto-immune cytopenia or cutaneous, renal, pulmonary, neurological involvement) or persistent salivary gland enlargement, early disease (≤5 years from the beginning of symptoms), and presence of at least one biomarker of B cell activation (increase in IgG level or free light chains or β2-microglobulinemia, decrease in complement component 4 (C4) level, presence of cryoglobulinaemia or monoclonal component). The patients received belimumab, 10 mg/kg, at week (W) 0, 2 and 4 and then every 4 weeks to W24. Patients who responded to treatment at W28 continued with belimumab monthly through W48, with a final evaluation scheduled at W52 (4 weeks after the last dose).

We excluded patients who had received B cell-targeted therapy during the previous 12 months, cyclophosphamide during the previous 6 months or any biological therapy during the previous 3 months; had begun any new immunosuppressive/immunomodulatory agent or antimalarial during the previous 2 months; or had severe renal or haematological failure, a history of cancer (with the exception of pSS-associated lymphoma), hepatitis B or C, HIV, tuberculosis, aggressive current lymphoma, severe diabetes or any other chronic disease or evidence of infection. The following concomitant treatments were allowed if kept at a stable dose: prednisone (up to 10 mg/day), non-steroidal anti-inflammatory drugs, immunosuppressive/immunomodulatory or antimalarial agents, secretagogue therapy or tear substitutes.

The study protocols were approved by the institutional review boards or ethics committees of the individual study sites and registered as NCT01160666 and NCT01008982. The study was conducted according to the current regulations of the International Conference on Harmonisation guidelines and the principles of the Declaration of Helsinki. All patients provided written informed consent before participating in any protocol-specific procedures.


The primary end-point evaluated at W28 was improvement in two of the five following items: ≥30% reduction in dryness score on a visual analogue scale (VAS), ≥30% reduction in fatigue score on a VAS, ≥30% reduction in musculoskeletal pain score on a VAS, ≥30% reduction in systemic activity score on a VAS assessed by the physician, and/or ≥25% reduction in serum levels of any of the following B cell activation biomarkers (free light chains of immunoglobulin, β2-microglobulin, monoclonal component, cryoglobulin, IgG) or ≥25% increase in C4 level.

Secondary end-points were change from baseline to W28 in the following items: each of the five items of the primary end-point, Schirmer's test, unstimulated salivary flow, European League Against Rheumatism (EULAR) Sjögren's Syndrome Disease Activity Index (ESSDAI),20 EULAR Sjögren's Syndrome Patient Reported Index (ESSPRI)21 and SF36 quality-of-life index. The day before evaluation, the patients were asked not to take their saliva or tear substitutes.

All the primary and secondary end-points at W24 were also secondary end-points at W52, the end of the study.

Levels of autoantibodies were monitored during the study, as was the focus score from biopsy of the labial salivary gland at baseline and at W28 in French patients and at W52 in Italian patients. BAFF was measured using ELISA (Quantikine kit, R&D Systems).

Statistical analysis

We did not calculate a sample size because of the pilot nature of the study. Categorical variables are described with frequencies and percentages. Quantitative variables are described with the mean (SD). Paired t test was used to assess differences between baseline and W28 quantitative assessments, and mean differences were expressed with 95% CI. In the case of distributions with extreme outliers, a Wilcoxon's signed rank was used. Paired proportions were compared using McNemar mid-p test for paired data. Data analysis involved use of SAS V.9.3 for Windows (SAS Institute, Cary, North Carolina, USA).

Role of the funding source

This study was sponsored by AP-HP (Assistance Publique-Hopitaux de Paris) and by the Azienda Ospedaliero Universitaria “Santa Maria della Misericordia”, Udine, Italy. It received an unrestricted research grant from Human Genome Science/GlaxoSmithKline. Human Genome Science/GlaxoSmithKline did not have any role in the design, conduction, interpretation and reporting of the study.


Patient characteristics

We included 31 patients, 16 in Paris and 15 in Udine. One patient (from Paris) was excluded from analyses because she had lupus-like severe glomerulonephritis (which was a non-inclusion criterion). She received only two belimumab infusions without any change in renal variables, and then the disease responded to cyclophosphamide. She had pSS with no other manifestations of lupus and no anti-DNA antibodies. Among the remaining 30 patients, 28 completed the 28-week infusions (figure 1). In one patient with lymphoma and cryoglobulinemic vasculitis, treatment was discontinued at W12 because of disease worsening, a new therapy was introduced 7 weeks later. One other patient stopped belimumab at W18 because of meningitis. Both patients could be evaluated at W28 for the primary end-point. The main baseline characteristics of patients are in table 1. All patients were women (mean age 49.5 (16.5) years). Mean disease duration was 5.7 (5.6) years. In all, 20 patients had systemic complications, 10 early disease and 22 increased value of at least one other biomarker of B cell activation. In all, 5 (16.7%) patients were treated with prednisone at a mean dose of 5.6 (2.5) mg/d.

Table 1

Baseline characteristics of patients

Figure 1

Flow of study participants. W, week.


At W28, 18 of 30 patients (60%) had achieved the primary end-point: reduced VAS dryness score in 11 (37%); VAS fatigue score in 7 (23%); VAS pain score in 7 (23%); and physician VAS systemic activity score in 13 (43%); and improved biological values, in 22 (73%) (figure 2). The percentage of responders was 6/10 (60.0%) for patients with early disease, 11/20 (55.0%) for those with systemic complications and 11/22 (50.0%) for those with biological activity. If the biological response was removed from the response criteria (improvement of two of the four remaining VAS), 12 of the 30 (40%) patients were considered responders. Also, 11 (84.6%) of the 13 patients with improved systemic activity according to the physician (VAS) had improvement of B cell biomarkers. Conversely, only 11 (50%) of the 22 patients with improved B cell biomarkers also improved their systemic activity (VAS).

Figure 2

Response to treatment for primary outcome (achieving two of five items at 28 weeks) and each component (dryness, fatigue and/or pain score; systemic activity assessed by the physician; B cell activation biomarkers).

The mean dryness, fatigue and pain VAS scores changed from 7.8 (1.8) to 6.2 (2.9) (mean difference −1.6, 95% CI (−2.6 to −0.6), p=0.0021), 6.9 (1.8) to 6.0 (2.2) (mean difference −0.9, 95% CI (−1.8 to 0.1), p=0.0606) and 4.6 (2.6) to 4.7 (2.4) (mean difference 0.1, 95% CI (−1.0 to 1.1), p=0.89), respectively (figure 3). The mean ESSPRI score decreased from 6.4 (1.1) to 5.6 (2.0) (mean difference −0.8, 95% CI (−1.5 to −0.2), p=0.0174) (figure 3).

Figure 3

Change in clinical parameters. Longitudinal evolution of dryness score on a visual analogue scale (VAS), pain score on a VAS, fatigue score on a VAS, EULAR Sjögren's Syndrome Patient Reported Index (ESSPRI), systemic activity score on a VAS assessed by the physician and EULAR Sjögren's Syndrome Disease Activity Index (ESSDAI).

The mean ESSDAI score decreased from 8.8 (7.4) to 6.3 (6.6) (mean difference −2.5, 95% CI (−4.0 to −1.0), p=0.0015). Interstitial pneumonitis developed in one patient during treatment. Before inclusion this patient was positive for anti-SSA and anticentromere antibodies, she had mild Raynaud’s phenomenon but had no other symptom of systemic sclerosis and no interstitial pneumonitis (on lung CT scan). During follow-up, she exhibited an evolving systemic sclerosis (interstitial pneumonitis and scleroderma skin involvement). The pneumonitis appeared after pneumococcal pneumonia and she actually had secondary SS. On excluding this patient from the analysis, the mean ESSDAI decreased from 8.7 (7.5) to 5.7 (5.9) (mean difference −3.0, 95% CI (−4.1 to 1.9), p<0.0001). The decrease in ESSDAI was ≥4 points for 11 of 30 patients (36.7%), ≥3 points for 15 (50.0%) and ≥2 points for 19 (63.3%).

The following domains mainly contributed to the ESSDAI score at baseline: glandular, biological, lymphadenopathy, articular, haematological and pulmonary. The number (%) of patients active in the different ESSDAI domains at baseline versus W28 were as follows: glandular, 15 (50.0%) versus 8 (26.7%) (p=0.0078); biologic, 27 (90.0%) versus 20 (66.7%) (p=0.0078); lymphadenopathy, 9 (30.0%) versus 5 (16.7%) (p=0.0625); articular, 9 (30.0%) versus 4 (13.3%) (p=0.0313); haematological, 5 (16.7%) versus 3 (10.0%) (p=0.25); and pulmonary, 4 (13.3%) versus 5 (16.7%) (p=0.50). At W28, 10/13 (76.9%) patients with non-malignant parotid swelling (confirmed by biopsy, whenever possible) showed improvement in the glandular domain, but parotid swelling did not improve in 2/2 patients with parotid low-grade lymphoma (stage IE).

Seven patients had been previously treated with rituximab. Two had response to rituximab (including one with lymphoma), one also responded to belimumab. Five had no response with rituximab; three of them responded to belimumab.

We found no change in unstimulated whole salivary flow (from 0.6 (1.2) to 0.7 (1.3) mL/15 min (mean difference 0.1, 95% CI (−0.1 to 0.4), p=0.27) or Schirmer's test (from 3.7 (6.6) to 4.2 (8.1) mm (mean difference 0.5, 95% CI (−1.1 to 2.1), p=0.51), even in the subgroup of recent onset patients (data not shown).

The mean SF36 physical health and mental health component scores changed from 41.1 (11.6) to 41.8 (11.3) (mean difference 0.7, 95% CI (−3.1 to 4.4), p=0.71) and from 44.7 (14.8) to 45.0 (13.7) (mean difference 0.3, 95% CI (−4.4 to 5.0), p=0.90), respectively.

The treatment induced significant changes in B cell biomarker values (figure 4): serum levels of IgG from 21.2 (9.5) to 18.2 (7.3) g/L (mean difference −3.0, 95% CI (−5.4 to −0.5), p=0.0201), IgA from 3.7 (1.9) to 3.3 (1.8) g/L (mean difference −0.4 95% CI (−0.8 to −0.1), p=0.0465) and IgM from 1.6 (0.8) to 1.3 (0.6) (mean difference −0.4, 95% CI (−0.5 to −0.2), p<0.0001); and levels of κ free light chain from 33.2 (24.7) to 25.6 (23.4) mg/L (mean difference −7.6, 95% CI (−11.1 to −4.1), p=0.0002) and λ free light chain from 28.3 (16.6) to 20.9 (12.2) mg/L (mean difference −7.5, 95% CI (−11.3 to −3.6), p=0.0007). Finally, the mean rheumatoid factor titre decreased from 146.2 (170.3) to 106.7 (102.3) IU (p=0.0003). The mean number of B cells significantly decreased from 187.2 (146.5)/mm3 to 65.1 (44.5)/mm3 (mean difference -122.1, 95% CI (−172.4 to −71.8), p<0.0001). Cryoglobulinaemia disappeared in the three positive patients. No initially positive patient showed negativity for anti-SSA or anti-SSB antibodies during follow-up; in one patient, anti-SSB antibody appeared during treatment. The focus score of the lymphoid labial salivary gland (LSG) infiltrate, evaluated in 15 patients at W28, decreased from 1.9 (1.5) at W0 to 1.7 (2.3) at W28 (mean difference −0.2, 95% CI (−1.2 to 0.7), p=0.57).

Figure 4

Change in biological parameters. Longitudinal evolution of levels of serum IgG (g/L), IgA (g/L), IgM (g/L), κ free light chain (mg/L) and λ free light chain (mg/L) levels, and rheumatoid factor titres (international unit (IU)).

Mean baseline BAFF level was 1383 (1369) pg/mL, and did not differ between responders and non-responders according to the primary end-point (p=0.57), or to other criteria of response such as a decrease of three points or more of ESSDAI (p=0.44).


Twenty patients presented 37 adverse events (table 2). No infusion reactions occurred. Only one severe adverse event was noted in the 28-week period: pneumococcus meningitis developed in one patient after six infusions. She was admitted in intensive care unit and treated with intravenous antibiotics. This patient, who was responder to belimumab, completely recovered from the meningitis without any sequellae. Another patient had infectious pneumonia not needing hospitalisation or intravenous antibiotics. Finally, as indicated above, progressive scleroderma developed in one patient during follow-up. A patient developed breast cancer outside the study period 3 months after the last infusion.

Table 2

Adverse events of interest


In this first open-label study of belimumab in 30 patients with pSS, with the chosen criteria for response, 60% of patients were responders, and systemic activity measured by the ESSDAI was significantly decreased. Compared with systemic and glandular involvement, patient reported outcomes improved modestly. Overall, only dryness significantly improved but not fatigue and pain. Non-malignant parotid swelling was reduced in 10/13 patients. Finally, most B cell biomarker values were significantly improved in this population of patients who were all positive for anti-SSA antibodies. Despite the occurrence of one serious pneumococcus infection, overall safety appeared good.

The treatment of pSS remains a challenge. Hydroxychloroquine is frequently proposed, but its use is based on observational studies22 and one cross-over study that showed improvement of only biological values.23 Recently, a randomised controlled study failed to show achievement of its primary end-point. Classical immunosuppressive or immunomodulatory agents and antitumour necrosis factor treatment failed to show any effect.24 ,25 Recently, the importance of the role of B cells in the disease suggested great hope with rituximab treatment, but the results remain controversial.4–6 In light of the role of BAFF in the disease and the efficacy of belimumab in SLE, the use of belimumab in pSS appears appealing.

The inclusion criteria in the Belimumab in Sjogren's syndrome (BELISS) trial slightly differed from that in other randomised controlled trials of pSS6 ,22 because the presence of anti-SSA antibodies was mandatory in our trial. Anti-SSA antibodies are present in two-thirds of patients with pSS, so focusing on these patients to evaluate a B cell-targeted therapy seems logical. Interestingly, the phase II study of SLE patients with and without autoantibodies failed to show an effect of belimumab compared with placebo,26 whereas the phase III trial including only patients with autoantibodies was successful.17 ,18 Moreover, in lupus, the difference in efficacy between placebo and belimumab was even greater for patients with than without anti-DNA antibodies.27 In addition, we included only patients with systemic complications, very early disease or increased biomarker values, which may represent patients more likely to show improvement with a B cell-targeted therapy.28

Our results support the possible efficacy of belimumab, with improvement in some symptoms (dryness) but mainly in some objective clinical signs (eg, parotid swelling), systemic activity and values of some biomarkers (Ig levels, rheumatoid factor titre and cryoglobulinaemia).

Thus, in our study, the improvement in patient’s symptoms was quite limited. We found no change in unstimulated whole salivary flow or Schirmer’s test even in the subgroup of recent onset patients (data not shown). The most interesting results are the reduction in objective symptoms such as parotid swelling and in values of B cell activation biomarkers, which demonstrates that the drug is able to decrease B cell activation, a characteristic of this autoimmune disease. In light of the diversity of systemic complications, precisely determining which patients can be improved with belimumab is difficult. However, clearly, the drug appeared to be efficient for non-malignant parotid enlargement; effectively, 10/13 patients showed improvement in this condition. No improvement in parotid swelling was seen in the two patients with low grade parotid lymphoma; however, belimumab was not prescribed for this reason in these two cases. Interestingly, 85% (11/16) of the patients with systemic improvement of at least 30% according to physician VAS also had improvement of B cell biomarkers. Thus, improvement of B cell biomarkers could be a surrogate marker predictive of clinical improvement.

Obviously, the main limitation of the study is its open design and, thus, the improvement of symptoms must be interpreted with caution. The same three patient reported outcomes (ie, dryness, fatigue and pain VAS scores) were used in the Trial of Remicade in Primary Sjogren's Syndrome (TRIPSS) randomised controlled study evaluating the effect of infliximab in pSS. Overall, 20% (6/30) of our BELISS patients experienced a 30% decrease in at least two of these three VAS scores, the primary end-point of TRIPPS, which was also achieved by 26% of patients receiving placebo treatment in the latter study.24

The safety of the drug was in line with what is known in lupus patients, except for one case of severe pneumococcal meningitis. This patient did not have other risk factors of pneumococcal infection. After this episode, we proposed antipneumococcal vaccination to all patients included in the trial. Data analysis at W52, the end of the study, is in course and should allow more extensive conclusions on long-term efficacy and safety of treatment.

In conclusion, the results of this first open-label phase II study of belimumab in pSS patients are encouraging and justify further randomised controlled trials with the drug in a selected target population of pSS patients most likely to benefit from treatment.


We thank Mohcine Benbijja, Nadine Vasseur-Porchet, Mickael Randrianandrasana, Unité de Recherche Clinique Paris-Sud (Sofiane Chekki) and Miriam Isola, Institute of Statistics, University of Udine, who participated to the management of the patients and the database.



  • Handling editor Tore K Kvien

  • XM and RS contributed equally.

  • Contributors Conception and design: XM and SDV. Acquisition of data: all authors. Analysis and interpretation of data: GB, RS and XM. Drafting the article or revising it critically for important intellectual content: all authors. Final approval of the version published: all authors.

  • Funding This study was sponsored by AP-HP (Assistance Publique-Hopitaux de Paris) and by the Azienda Ospedaliero Universitaria “Santa Maria della Misericordia”, Udine, Italy. It received an unrestricted research grant from Human Genome Science/GlaxoSmithKline.

  • Competing interests MX and RS received honoraria from GSK.

  • Patient consent Obtained.

  • Ethics approval The study protocol was approved by the institutional review boards or ethics committees of the individual study sites.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement All data of major outcomes are available on request.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.