Xavier Mariette

Objective. To investigate the frequency and risk factors of postoperative complications in RA patients treated with abatacept (ABA). Methods. The Orencia RA registry recruited 1012 patients receiving ABA for RA in routine care. Data from patients treated with ABA who underwent surgery were reviewed to describe the frequency of postoperative complications. Characteristics of patients and surgeries with and without complications were compared to identify factors associated with complications. Results. We identified 205 (20.3%) patients who underwent 263 surgeries, including 176 (66.9%) orthopaedic surgeries. Nineteen (7.2%) surgeries, in 19 patients (9.3%), entailed complications, including 7 delayed wound healing (2.7% of surgeries) and 6 surgical site infections (2.3% of surgeries). The median time between the last infusion of ABA and surgery was 5.9 weeks (range: 0.3–12.0 weeks), with no significant difference between patients with and without complications. The median corticosteroids daily dosage was higher in the group with complications [10.0 (6.25–15.0) vs 6.0 (5.0–10.0) mg/day, P = 0.042]. In multivariate analysis, only the duration of ABA treatment was significantly associated with postoperative complications [adjusted odds ratio (aOR) = 0.94 (95% CI: 0.89, 0.99) for each month of treatment], as were orthopaedic surgeries compared with other kinds of surgery [aOR = 4.45 (95% CI: 1.01, 20.2)]. Conclusion. In RA patients treated with ABA, the rate of surgical complications was low: 7.2% and higher in case of orthopaedic procedure and a more recent initiation of ABA. The median time between surgery and the last infusion of ABA was short and did not influence the rate of postoperative complications.

The 18th annual international Targeted Therapies meeting brought together over 100 leading scientists and clinicians from around the world in the field of rheumatology. During the meeting, breakout sessions were held consisting of 5 disease-specific groups each with 20-40 experts assigned to each group based on clinical or scientific expertise. Specific groups included: rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis/spondyloarthritis, systemic lupus erythematous, and other connective tissue diseases (e.g. Sjögren's, Behçet's, others). In each group, experts were asked to identify unmet needs in 3 categorical areas: basic/translational science, clinical science and therapeutic development, and clinical care. Needs were prioritised as primary or secondary. Overall, similar primary unmet needs were identified within each disease foci. Within translational science, these included the need for better understanding the heterogeneity within each disease, such that predictive tools for therapeutic response could be developed. Within clinical science and therapeutic trials, the ability to prevent progression to disease onset in those at risk, and the ability to cure disease were identified. A further unmet need was to develop new and accessible therapeutics, as well as to conduct strategic trials of currently approved therapies. Within the clinical care realm, improved co-morbidity management and patient-centered care were identified as unmet needs. Lastly, it was strongly felt there was a need to develop a scientific infrastructure for well-characterised, longitudinal cohorts married with biobanks and mechanisms to support data-sharing. This infrastructure could facilitate many of the unmet needs identified within each disease area.

To evaluate the efficacy of rituximab in central nervous system (CNS) manifestations of patients with primary Sjögren's syndrome (pSS). Prospective data from patients with pSS and CNS involvement included in the French AutoImmunity and Rituximab registry were analysed. All patients had diffuse white matter T2-weigted hypersignals. Neurological response was defined as improvement or disappearance of neurological signs. Eleven patients (mean age 55 years [38-77]) were treated with rituximab for their neurological involvement. The mean duration of pSS was 9 years (4-24). Mean baseline ESSDAI score was 17 (5-25). Neurological features were progressive multiple sclerosis-like manifestations (n=6), transverse myelitis (n=1), anxiety and depression disorder (n=1) and cognitive dysfunction (n=3). Mean Expanded Disability Status Score (EDSS) before rituximab was 4 (3-5.5). The mean follow-up was of 13 months (6-58). No neurological change occurred in all 6 patients with multiple sclerosis-like symptoms, in 2/3 patients with cognitive dysfunction or in the patient with anxiety-depression. One patient with depression and cognitive dysfunction disclosed subjective improvement. One patient with transverse myelitis, refractory to cyclophosphamide had an improvement of his walk perimeter (160 meters vs. 116). Mean EDSS score and ESSDAI remained stable. Rituximab does not seem to be effective in progressive multiple sclerosis-like manifestations of patients with pSS-related CNS involvement.

Objective To examine blood B cell subsets in patients with rheumatoid arthritis (RA) prior to B cell depletion therapy and to assess their potential as predictors of clinical response to rituximab (RTX). Methods Blood B cell subsets were assessed by flow cytometry in 208 RA patients included in an RTX retreatment study (assessed prior to RTX treatment) and in 47 age‐matched controls. Expression of BAFF receptor (BAFF‐R) on B cells and serum B cell biomarkers was also measured. B cell subsets and BAFF‐R expression were compared between RA patient and control populations. Univariate and multivariate analyses were performed to identify baseline factors associated with a European League Against Rheumatism response 24 weeks after 1 cycle of RTX. Results Mean ± SD counts of both CD27− naive and CD27+ memory B cells were decreased in RA patients (188.6 ± 121.4/mm³) compared with controls (257.3 ± 154.1/mm³) (P = 0.001) and were partially restored in patients treated with methotrexate (MTX) plus anti–tumor necrosis factor compared with patients treated with MTX alone. Within the CD27+ memory B cells, the CD27+IgD− switched memory subtype was selectively decreased, irrespective of treatment. The frequency of CD27+ memory B cells correlated inversely with levels of several B cell activation biomarkers in RA. Serum BAFF level and BAFF‐R expression was comparable in RA patients and controls. A low baseline CD27+ memory B cell frequency was associated with a greater clinical response to RTX (odds ratio 0.97 [95% confidence interval 0.95–0.99], P = 0.0015). Conclusion In B cell depletion therapy–naive RA patients, a low frequency of CD27+ memory B cells correlated with levels of serum B cell activation biomarkers and may predict response to RTX. These results suggest that low memory B cell frequency may be indicative of a B cell–driven RA subtype that is more sensitive to B cell depletion therapy.

Objective: Sjögren's syndrome (SS)–related peripheral neuropathy is responsible for disability, but no treatment has been shown to improve its outcome. In some cases, intravenous immunoglobulin (IVIG) therapy has been associated with some benefit. In this study, we investigated the effectiveness of IVIG in SS‐related peripheral neuropathy.Methods: We assessed the efficacy and tolerance of IVIG in 19 patients with primary SS–related neuropathy without any necrotizing vasculitis in a retrospective national multicentric study. The evaluation of the response was assessed using the disability Modified Rankin Scale (MRS) and a global evaluation by the practitioner.Results: Eight patients (42%) exhibited a decrease of the MRS score corresponding to a clinical improvement, 10 patients (52%) exhibited a stable MRS score, and 1 patient (6%) showed an increase of MRS score. According to the global evaluation by the practitioner, 9 (47%) of the 19 patients were improved, 6 patients (31%) were stable, and 4 patients (21%) worsened. All the patients with sensorimotor (n = 5) or nonataxic sensory neuropathy (n = 4) were improved or stabilized. However, among the patients with ataxic neuropathy (n = 9), only 2 improved and 4 worsened. Ten of the 13 patients treated with corticosteroids could have had the prednisone dosage decreased from 15 mg/day (range 7.5–60) to 10 mg/day (range 5–20) with IVIG. Only 1 patient stopped the treatment after 1 dose because of a minor side effect and lack of initial efficacy.Conclusion: IVIG may be useful in the treatment of SS‐associated sensorimotor neuropathies or nonataxic sensory neuropathy without any necrotizing vasculitis. The benefit of such therapy in the SS‐related ataxic neuropathy seems less clear.

Objective: To examine whether serum B cell markers can predict response to rituximab, a B cell–depleting monoclonal antibody, in patients with refractory rheumatoid arthritis (RA).Methods: This rituximab re‐treatment dose study (SMART [eSsai MAbthera sur la dose de Re‐Traitement]) involved 208 patients with refractory RA. Serum markers of B cell activation (anti–cyclic citrullinated peptide [anti‐CCP] antibodies, rheumatoid factor [RF], serum IgG, IgA, and IgM levels, serum κ and λ free light chains, and serum BAFF) were assessed before the first rituximab cycle (1,000 mg on days 1 and 15). Univariate and multivariate analyses were performed to identify factors associated with a European League Against Rheumatism (EULAR) response at 24 weeks.Results: There were 149 responders (72%). Two baseline factors were associated with a EULAR response at 24 weeks in multivariate analysis: the presence of anti‐CCP antibodies or RF (odds ratio 3.5 [95% confidence interval 1.6–7.6]) and a serum IgG concentration above normal (odds ratio 2.11 [95% confidence interval 1.02–4.33]), with synergy between them (odds ratio 6.0 [95% confidence interval 2.2–16.2]).Conclusion: The presence of RF or anti‐CCP antibodies and elevated IgG are 2 simple biomarkers that can be used routinely before therapy to predict response to rituximab in patients with refractory RA.

B‐cell‐activating factor (BAFF) plays a key role in promoting activation of autoimmune B cells. This cytokine may be expressed in and secreted by salivary gland epithelial cells (SGEC) after stimulation with type I IFN or viral or synthetic dsRNA. Because this BAFF expression depends only in part on endosomal TLR and type I IFN, we investigated whether other dsRNA sensors could be implicated in BAFF expression. Using human SGEC, we confirmed the partial dependence of BAFF expression on TLR‐3 by replicating the partial inhibition of BAFF expression observed upon endosomal inhibition using TLR‐3 or Toll/IL‐1R domain‐containing protein inducing IFN‐β silencing mRNA, but not with TLR‐7 silencing mRNA. Melanoma differentiation‐associated gene 5 silencing mRNA had no effect on BAFF expression, but retinoic acid‐inducible gene I silencing mRNA had a slight effect observed following infection with dsRNA reovirus‐1. Inhibition of RNA‐activated protein kinase (PKR) by 2‐aminopurine completely abolished both BAFF mRNA and protein production after reovirus‐1 infection and poly(I:C) stimulation through NF‐κB and p38 MAPK pathways, with the latter implicated only after poly(I:C) stimulation. Thus, PKR is the dsRNA sensor implicated in BAFF induction in SGEC after dsRNA stimulation. In autoimmune diseases, PKR may be an interesting target for preventing BAFF following the induction of innate immunity.

We investigated B‐cell activating factor of the tumour necrosis factor family (BAFF) level in peripheral blood mononuclear cells (PBMCs), monocytes and T cells from patients with primary Sjögren’s syndrome (pSS) and controls both ex vivo and in vitro after cytokine stimulation. PBMCs, monocytes and T cells were isolated from 15 patients with pSS and 17 controls. Cells were cultured alone or with interferon (IFN)α, IFNγ and interleukin 10 (IL‐10). T cells were stimulated with phytohaemagglutin and anti‐CD3. BAFF protein was assessed by enzyme‐linked immunosorbent assay. Ex vivo, no difference was observed in BAFF mRNA level in PBMCs and monocytes from patients and controls. Blood monocytes were the main cell type secreting BAFF both in patients and controls. In vitro, after IFNα stimulation, BAFF mRNA level was significantly higher in cells from patients than from controls (63.8 versus 20.7, P = 0.03). T cells from patients secreted a higher level of BAFF protein than those from healthy donor cells (17.4 versus 2.9 pg/ml, respectively, P = 0.04) but at a lower level than that from monocytes. Stimulation of T cells did not change BAFF secretion level. The induction of Th17 cells showed no increased BAFF expression. In conclusion, similar to epithelial cells, blood monocytes in patients with pSS show increased production of BAFF under IFNα, which confirms the involvement of IFNα in pSS. BAFF expression is also increased in blood T cells of such patients, independently of T‐cell stimulation.

Introduction: Latent tuberculosis infection (LTBI) is detected with the tuberculin skin test (TST) before anti-TNF therapy. We aimed to investigate in vitro blood assays with TB-specific antigens (CFP-10, ESAT-6), in immune-mediated inflammatory diseases (IMID) for LTBI screening. Patients and methods: Sixty-eight IMID patients with (n = 35) or without (n = 33) LTBI according to clinico-radiographic findings or TST results (10 mm cutoff value) underwent cell proliferation assessed by thymidine incorporation and PKH-26 dilution assays, and IFNγ-release enzyme-linked immunosorbent spot (ELISPOT) assays with TB-specific antigens. Results: In vitro blood assays gave higher positive results in patients with LTBI than without (p0.3) and lower results than those with LTBI (pConclusion: Anti-TB blood assays are beneficial for LTBI diagnosis in IMID. Compared with TST, they show a better sensitivity, as seen by positive results in 5 patients with certain LTBI and negative TST, and better specificity, as seen by negative results in most patients with intermediate TST as the only criteria of LTBI. In the absence of clinico-radiographic findings for LTBI, blood assays could replace TST for antibiotherapy decision before anti-TNF.