C. Prati

Calprotectin and spondyloarthritis.

Spondyloarthritis is a chronic inflammatory disease with several phenotypes (including axial and peripheral), and possible extra-rheumatologic manifestations such as uveitis, psoriasis or inflammatory bowel disease (IBD). With an onset in young adults and a prevalence of approximately 1% of the adult population [1], the disease is associated with quality of life and functional impairments, leading to socioeconomic consequences. Several tools have been proposed for the evaluation of the disease, but the most currently used, particularly in axial disease, are the Bath Ankylosing Spondylitis Disease Index (BASDAI) and the Ankylosing Spondylitis Disease Activity Score (ASDAS), the latter of which incorporates the acute phase reactant C-reactive protein (CRP). CRP is the most widely used biomarker in current practice and clinical trials, but it has performance limitations [2]. Currently there is an unmet need for a simple reliable biomarker of disease activity and prognosis in spondyloarthritis [2]. Many candidate biomarkers may arise from the growing immunopathologic knowledge of the disease [3]. The pathogenesis of the disease involves genetic and environmental factors [1], leading to the activation of the IL-23/Th17 pathway [4]. The gut plays a major role in this scenario [5], representing a large immunological interface between environmental factors represented by gut microbiota modifications or dysbiosis, and the immune system, with data illustrating the relationship between dysbiosis, gut and IL-23/Th17 axis activation [5]. The presence of this gut involvement is also supported by the finding of ileocolonic histologic inflammation in approximately half of the patients with axial spondyloarthritis [1,5] and an association between chronic gut inflammation and inflammatory scores on sacroiliac MRI [5]. Inflammatory lesions found on MRI may secondarily evolve toward ossification, which is associated with functional consequences in this disease. Calprotectin is a member of the S100 protein family and is a heterodimeric complex of S100A8 and S100A9 proteins (calgranulin A and B, part of the Alarmins) [6]. Calprotectin is a calcium and zinc binding protein in the cytosol of monocytes, macrophages and neutrophil granulocytes. In local inflammation, activated cells migrate to the inflammatory site and release calprotectin locally; in the case of gut inflammation, calprotectin is released into the gut lumen and into the feces [7]. Fecal calprotectin is a biomarker that is frequently used in inflammatory bowel disease (IBD) [7], for the assessment of disease activity and response to treatment and in the prediction of disease relapse or postoperative recurrence [7]. Serum calprotectin may have similar relevance in IBD [8]. There is some evidence in the literature supporting the potential of calprotectin as a disease biomarker in spondyloarthritis. Hammer et al. [9] found in 52 cases of reactive arthritis followed over 104 weeks that plasma calprotectin was well correlated with CRP and disease activity and was the first biomarker to return to normal levels during recovery. Klingberg et al. [10] showed that fecal calprotectin was elevated in 68% of 205 ankylosing spondylitis patients and was not associated with gastrointestinal symptoms, whereas serum calprotectin levels remained low or normal in 98% of these cases. In this study, fecal calprotectin was higher in patients treated with nonsteroidal anti-inflammatory drugs (NSAID) and proton pump inhibitors, and lower in patients treated with a TNF blocker or with methotrexate. Another study evaluating 39 ankylosing spondylitis patients and 42 controls found elevated levels of fecal calprotectin in 41% of the patients and 10% of the controls [11], and no differences between fecal calprotectin-positive and negative patients in NSAID or anti-TNF use. In this study, an association was found between fecal calprotectin positivity and elevated levels of other biomarkers of gut inflammation, namely ASCA and Cbir1. Oktayoglu et al. [12] confirmed higher levels of serum calprotectin in 31 patients with ankylosing spondylitis in comparison to 45 healthy controls, but without correlation with disease activity (assessed by BASDAI and ASDAS), function, metrology, radiographic score or CRP. Increased levels of serum calprotectin were confirmed again by Turina et al. [13] in 37 patients with spondyloarthritis compared to 20 healthy controls. Moreover, they found a significant decrease in serum calprotectin after 2 weeks of treatment with the anti-TNF infliximab (n = 18), whereas there was no change with placebo (n = 19). A significant reduction in serum calprotectin levels was also demonstrated after 4 weeks of treatment with the TNF blocker etanercept in another group of 20 patients with

Targeted synthetic disease-modifying antirheumatic drugs in spondyloarthritis

984 (2015). 14 Paramarta JE, Turina MC, Noordenbos T et al. A proof-ofconcept study with the tyrosine kinase inhibitor nilotinib in spondyloarthritis. J. Transl. Med. 14(1), 308 (2016). • A well-documented study illustrating different efficacy of kinase inhibition according to the presentation of the

Flare in axial spondyloarthritis. The dark side of the outcome

Spondyloarthritis (SpA) is a chronic inflammatory rheumatic disease with many phenotypes,1 but the frame of the disease is still a matter of debate, particularly regarding the non-radiographic forms of axial SpA.2 ,3 The disease evolution may have several profiles, mainly related to the treatment strategy, balancing from periods of remission or low disease activity to flares of the disease. The recommended treatment strategies are supposed to be tailored to the disease activity, aiming to reach remission or low disease activity in a T2T strategy,4 with management of remission (reduction of dosage or increase in interval of administration), as well as treatment intensification in case of flares. Definition of a flare may be difficult, particularly in a multifaceted disease like SpA, and in fact there is currently no clear, universally recognised, definition of a flare in axial SpA. This is part of the research agenda of recent recommendations.5 Looking at a medical dictionary, flare is defined as a sudden intensification in disease. In first analysis, one may propose that flare represents a worsening of symptoms with an increase in disease activity. Taking this into consideration, many questions arise: should we consider relative variation or absolute value above a significant threshold? Should we consider only generalised increase in activity? Should we include extra rheumatic manifestations such as uveitis or inflammatory bowel disease (IBD)? Are synovitis and enthesitis also part of this evaluation? Should we include biological markers (but reliable biomarkers are lacking in …

SAT0606 Disease modifying anti rheumatic drugs in the treatment of sapho syndrome: systematic literature analysis

Background SAPHO (Synovitis Acne Pustulosis Hyperostosis Osteitis) Syndrome is a rare, heterogeneous clinical entity with cutaneous and osteoarticular expression. The therapeutic management is not codified and there is no validated evaluation tool for SAPHO syndrome. Objectives To perform a systematic analysis of the literature in order to evaluate the effects of DMARDs in SAPHO syndrome. Methods Bisphosphonates, conventional and targeted synthetic DMARDs, anti-TNF alpha, and other biologics have been subjected to advanced Pubmed research. Treatment was considered effective when the patient validated the response criteria defined in the study or if at least partial benefit was obtained for a minimum of three months. The different treatments were ranked according to their effectiveness rate in three interest groups and then grouped by therapeutic class to determine an overall response rate. These rates led to the calculation of an efficacy index weighted by the number of patients treated in the subgroup (molecule or therapeutic class) compared to the total number of patients in our study. Results Treatment efficacy was evaluable in 284 of the 292 patients analysed. The clinical presentation of cases was reported in 205 patients for osteoarticular involvement, and 193 for cutaneous involvement. The group of treatments that most often induces a therapeutic response (in more than 75% of cases) includes Ibandronate, Etanercept, Anakinra, Infliximab, Pamidronate and Adalimumab. Pamidronate, which represents the largest subpopulation in our study, has the highest weighted index of efficacy. Zoledronic acid, Lefulnomide, Ustekinumab and Methotrexate have less often induced a therapeutic effect (efficacy between 50% and 75%) and Sulfasalazine and Secukinumab have response rates of less than 30%. In total, bisphosphonates and TNF alpha antagonists have efficacy rates of 87.77% and 85% respectively. The weighted index, more than twice as high for bisphosphonates (42.96 versus 17.96), reflects the predominant use of these in SAPHO syndrome, with most often a beneficial effect. Conventional and synthetic targeted DMARDs and other biological treatments are less often effective in our study, with response rates of 47% and 58% respectively. The frequencies of each clinical manifestation were in agreement with the data of the literature. There was no clear clinical profile of a good responder to a particular treatment. Therapeutic class Number of patients Number of responders Efficacy rate (%) Weighted index Bisphosphonates 139 122 87,77 42,96 Conventional DMARds 68 32 47,06 11,27 Synthetic targeted drugs 0 NA NA NA Anti TNF alpha 60 51 85 17,96 Other biologics 17 10 58,82 TOTAL 284 215 75,82 NA Conclusions This work made it possible to rank the different DMARDs used in the SAPHO syndrome. Anti TNF alpha and Pamidronate are the treatments that seem to bring the higher benefit. Disclosure of Interest None declared

FRI0039 Endothelial dysfunction in rheumatoid arthritis: which effect of methotrexate? a study in adjuvant induced arthritis model

Background Rheumatoid arthritis (RA) is associated with increased cardiovascular (CV) risk [1] secondary to endothelial dysfunction (ED) [2]. There is accumulating evidence that methotrexate (MTX), first intention DMARD, reduces CV risk in RA [3], but the mechanisms involved are still unknown. Objectives The aim of this study was to determine the effect of MTX on endothelial function in arthritis and to investigate its effect on endothelial pathways. Methods Experiments were conducted in the adjuvant-induced arthritis (AIA) model in Lewis rat. At onset of arthritis, rats were treated by a sub-cutaneous injection of MTX (1 mg/kg/week) or phosphate buffer saline (vehicle) for 3 weeks. Arthritis score was daily monitored. At the end of treatment, thoracic aorta was harvested to measure the relaxation to acetylcholine on pre-constricted aortic rings in the presence or not of inhibitor of nitric oxide (NO) synthase (L-NAME), arginase (nor-NOHA), EDHF (Apamin/Charybdotoxin), or a superoxide dismutase analog (Tempol). The effect of norepinephrine (NE) and sodium nitroprusside (SNP) was studied on endothelium-denuded aortic rings. The effect of MTX on hind paw radiographic score, serum lipids and plasma pro-inflammatory cytokines (TNFα and IL-1β) levels was measured. Results As compared to Vehicle rats, MTX significantly reduced arthritis score (p<0.01) but did not change radiographic score. It reduced plasma cytokines levels (p=0.02) but not total cholesterol and triglycerides levels. MTX did not change Ach-induced relaxation as compared to Vehicle. As regards endothelial pathways, MTX increased vascular NOS activity (p<0.0001) and decreased superoxide anions production but did change neither COX-2 and arginase activities nor EDHF production. Vascular smooth muscle reactivity to NE and SNP was unchanged by the treatment. Conclusions Despite a reduction of clinical and biological inflammation, MTX did not improve endothelial function in AIA rats. The study of endothelial mechanisms highlights the role of COX and arginase as seminal targets for reducing ED in RA. This study suggests that other mechanisms than improvement of endothelial function are involved in the CV benefits of MTX in RA that remain to be elucidated. Our data also suggest that the adjunction of drugs targeting endothelial function to MTX in RA patients might improve their CV prognostic. References Agca R, Heslinga SC, Rollefstad S, et al. EULAR recommendations for cardiovascular disease risk management in patients with rheumatoid arthritis and other forms of inflammatory joint disorders: 2015/2016 update. Ann Rheum Dis 2016:annrheumdis-2016–209775. doi:10.1136/annrheumdis-2016–209775. Prati C, Demougeot C, Guillot X, et al. Endothelial dysfunction in joint disease. Joint Bone Spine Rev Rheum 2014;81:386–91. doi: 10.1016/j.jbspin.2014.01.014. Roubille C, Richer V, Starnino T, et al. The effects of tumour necrosis factor inhibitors, methotrexate, non-steroidal anti-inflammatory drugs and corticosteroids on cardiovascular events in rheumatoid arthritis, psoriasis and psoriatic arthritis: a systematic review and meta-analysis. Ann Rheum Dis 2015;74:480–9. doi: 10.1136/annrheumdis-2014–206624. Disclosure of Interest None declared

THU0558 Therapeutic response to prednisone according to the age in polymyalgia rheumatica: a controlled study

Background Polymyalgia rheumatica (PMR) is an inflammatory rheumatic disorder which usually affect patients over 65 years old. Different poor prognostic factors are involved in prednisone response including rapid decrease of prednisone dose or female sex. To date, there is no data relating the impact of the age on therapeutic response in PMR. Objectives The aim of this study was to compare, in case of PMR, the response to prednisone in patients younger than 60 to patients over 65 years old. Methods This was a retrospective, monocentric study. We included patients suffering from PMR, meeting ACR 2012 criteria. Patients were classified into two groups, one group with patients less than 60 years, and one group with patients over 65 years. We registered demographic, clinical, biological, and imaging data as well as therapeutic response profile. The local inflammation was evaluated with PET scan, by studying each anatomical site usually affected by PMR. Then, the rate of inflammation was scored from 0 to 3, according to the intensity of uptake compared to liver. The treatment was standardized. The initial dose of prednisone was of 0.3mg/kg/j during the two first weeks, then, the dose was slowly decreased by 10% each month. The main endpoint was a steroid dependence defined by the recurrence of PMR symptoms and/or the increase of CRP at two times during the decrease of prednisone. Results We included 14 patients younger than 60 years old (average age 54+/- 0.8 years) and 28 patients older than 65 years old (average age 75.8 +/- 1.5 years). The population younger than 60 years was mainly male (60% VS 27%, p<0.05). Both groups were similar in terms of morning stiffness (2,1±0,4 VS 1,9±0,3 hours; p>0.05), disease duration (4,2±0,8 VS 4,1±0,6 months; p>0.05), leukocytes rate (8,3±1,37 VS 8±0,7 G/L; p>0.05) and percentage of antinuclear antibodies rate over 1/320 (20% VS 10%; p>0.05). However, regarding to local inflammation, the intensity of FDG uptake highlighted by the Pet scan was lower among young patients (score of 16,9±1,7 VS 26,5±3,0; p<0.05). Furthermore, we observed a significant difference concerning therapeutic response according to the age: 60% of the young patients developed a steroid dependence compared to20% in group of old patients (p<0.05). Moreover, the introduction of methotrexate was necessary for 35% of the young patients against 6.5% (p<0.05). Conclusions Our study is the first to highlight the age as a bad prognosis factor in case of PMR. This difference is independent to the systemic inflammation and surprisingly, local inflammation (assessed by the TEP score) is more important in elderly people. Young patients suffering from PMR are mostly men and are more dependent on steroids. Thus, methotrexate could be straightaway proposed, particularly in patient younger than 60 years old. Disclosure of Interest None declared

SAT0536 Utility of 18F-Fluoro-Dexoxyglucose Positron Emission Tomography for The Diagnosis of Polymyalgia Rheumatica: A Controlled Study

Objectives To compare18f-Fluoro-Dexoxyglucose Positron Emission Tomography (FDG-PET/CT) findings in patients with polymyalgia rheumatica (PMR) and controls without rheumatologic disease. Methods We have retrospectively included 50 patients with a diagnosis of PMR (EULAR/ACR criteria), who have had a FDG-PET/CT. For comparison, 53 patients who have had a FDG-PET/CT for initial staging or follow-up of neoplasm have been included as controls. 17 sites have been analyzed (2 shoulders, 2 acromio-clavicular and 2 sterno-clavicular joints, 2 greater trochanters, 2 hips, 2 ischial tuberosities, 2 iliopectinal bursitis, 2 pubic symphysis enthesis and only the interspinous space with the most FDG uptake). An FDG uptake score has been used (semi-quantitative scoring system, by Goerres and al. to evaluate visually FDG uptake (0 to 3 compared to liver uptake) into these sites. We calculated the FDG uptake score corresponding to the mean of intensity over all sites, and the number of sites with significant uptake (FDG uptake 2 or 3) for each patient. The cut offs for number of sites with high activity and for FDG uptake score were assessed using ROC curves and Odds Ratios. Results The two groups were comparable for the median patient age (69.3 years for PMR vs. 68.1 for controls). Significant differences between the two groups were found for FDG uptake score (1.12 vs 0.34, p<0.00001) and for number of sites with significant uptake: 6.36 sites vs 1.49 sites, p<0.00001. The presence of 3 or more sites with significant uptake was correlated with the diagnosis of PMR with 74% sensitivity, 79% specificity [OR =10.8]. For the FDG uptake score the cut off is 0.53 [Se 80%, Sp 77%; OR 13.6]. We found significant differences in all sites for FDG uptake score, and number of sites with significant uptake, particularly marked for shoulders, ischial tuberosities and interspinous bursitis (p<0.00001 for FDG uptake score). FDG uptake score (0–3) Number of sites with significant uptake (0–17) PMR group (N=50) 1.12 [%CI 0.90–1.33] 6.13 [%CI 4.99–7.26] Control group (N=53) 0.34 [%CI 0.26–0.41] 1.49 [%CI 1.38–1.59] p p<0.00001 p<0.00001 Conclusions We found significant uptake in all sites studied in PMR compared to controls. We propose the number of 3 sites with significant uptake and a FDG uptake score superior at 0.53 for the diagnosis of PMR. Disclosure of Interest None declared

FRI0409 Subclinical Atherosclerosis in Ankylosing Spondylitis. Does It Really Exist and Which Are The Effects of Treatments? A Systematic Review

Objectives Accelerated atherosclerosis and increased cardiovascular morbidity and mortality have been associated with ankylosing spondylitis (AS). Noninvasive angiological methods have been developped to evaluate endothelial and vascular dysfunction which is correlated with future development of atherosclerosis. The objectives were to determinate the presence or not of a subclinical vascular dysfunction in AS and if treatments could have an effect on it. A systematic review was done. Methods Studies evaluating subclinical atherosclerosis and vascular function in AS were identified using Pubmed, (Ovid, EMBASE). Search terms included “ankylosing spondylitis” AND (endothelial OR vascular OR intima media thickness (IMT) OR Flow mediated dilatation (FMD) OR pulse wave velocity (PWV) OR atherosclerosis). This identified 353 results after limiting to French and English. The final selection identified 29 articles. Results Overall, 1529 AS patients were included representing 38 studies: 8 studies about endothelial function, 198 AS patients and 130 healthy control (HC); 20 studies about carotid IMT, 900 AS and 644 HC; 10 studies about arterial rigidity, 431 AS and 285 HC. In cross-sectional studies, 4/6 indicated endothelial dysfunction in AS versus HC, 9/18 indicated increased cIMT and 3/5 increased arterial rigidity. About ED, 3 open label studies noted positive effect of TNFα blockers and spironolactone on FMD and rosuvastatin improved FMD in a placebo controlled study after 24 weeks. TNFα blockers seems not to improve neither cIMT (2 studies) nor arterial rigidity (5 studies). Exercise alone has improved arterial rigidity in 15 patients after 12 weeks. Conclusions Whereas early and accelerated atherosclerosis is patent in AS, presence of subclinical atherosclerotic lesions is controversial in the literature especially concerning cIMT and arterial stiffness. This is reinforced by the lack of TNF blockers efficacy. Conversely it seems that endothelial dysfunction is present and reversible after treatment with TNF blockers, statin and spironolactone. These results are consistent to treat AS patients with early effective treatment to prevent the risk of CV morbidity and mortality. Disclosure of Interest None declared

AB0271 Effects of Glucocorticoids on Endothelial Function: Focus on Rheumatoid Arthritis

Background Rheumatoid arthritis (RA) is the most common systemic autoimmune disease characterized by reduced life expectancy due to accelerated atherogenesis. Endothelial dysfunction (ED) precedes and initiates atherosclerosis, and its correction is associated with reduced CV risk. ED is present in RA and appears as a seminal target for reducing CV risk in RA. Despite the dramatic changes that occurred in the treatment approach in RA, majority of patients still use glucocorticoids (GCs). The potential harmful CV effects of GCs are “well-known” in the general population. In inflammatory diseases such as RA, the effects of GCs on CV risk is still a matter of debate. On the one hand, GC might increase CV risk by increasing CV risk factors but on the other hand, GC might reduce CV risk by decreasing the systemic and/or vascular inflammation. Objectives To provide the available data regarding the effects of GC on endothelial function, with a special focus on RA. Methods Systematic review of the literature on the PubMed database with the following keywords: “Glucocorticoids”, “rheumatoid arthritis”, “endothelial function” and “vascular function”. Results In the vascular system, GC receptors are expressed by intact arteries, cultured vascular smooth muscle cells (VSMC) and endothelial cells. In physiological conditions, experimental data provide evidence of harmful effects of GCs on endothelial function, due to decreased Nitric oxide synthase activity/expression, decreased production of endothelium-derived hyperpolarizing factor (EDHF) and prostacyclin, the three main vasorelaxant endothelium-derived factors, associated with increased production of endothelin-1, an endothelium-derived vasoconstrictor. These experimental data were not unanimously confirmed by the few studies in healthy controls, which reported negative or not effect of GCs on endothelial function. Effect of GCs is totally different in inflammatory conditions in which GC reduced endothelial dysfunction and slowed progression of atherosclerotic plaques in animal models. Only 6 clinical studies were performed in RA. The available data suggested that low dose GC (5 mg/d) did not change endothelial function while a high dose for a short period (7 days) improved endothelial function in RA patient. No harmful effect of GC on endothelial function were reported in RA. Conclusions Data from experimental models highlighted the Janus face of GCs since they can exert deleterious effects on endothelial function when applied in “normal” cells/animals although they seem to be protective under inflammatory conditions. In RA patients, data are scarce and highly controversial due to methodological weaknesses. However, contrary to common believe that GCs deteriorate CV function, no effect and even positive effect of GCs on endothelial function have been reported. As more RA patients are starting GCs early in their disease course, there is an urgent need to perform clinical trials specifically-designed to define the best GCs strategy to prevent CV risk in RA. Disclosure of Interest None declared

AB0089 Non-Steroidal Anti-Inflammatory Drugs and Endothelial Function in Ra: Harmful or Beneficial? A Study in Rat Adjuvant-Induced Arthritis

Background Rheumatoid Arthritis (RA) is associated with increased cardiovascular risk (CVR) explained in part by accelerated atherosclerosis as a consequence of endothelial dysfunction. Non-steroidal anti-inflammatory drugs (NSAIDs) are widely prescribed in RA patients in case of acute pain. Surprisingly, despite the commonly-held belief that NSAIDs worsen CVR, data regarding their effect on endothelial function in RA are lacking. Objectives The present study investigated the effect of naproxen (a non-selective COX inhibitor), diclofenac (a preferential COX-2 inhibitor) and celecoxib (a selective COX-2 inhibitor) on vascular function in arthritic rats and identified the underlying mechanisms. Methods Adjuvant-induced arthritis (AIA) was induced in 6 weeks-old male Lewis rats by injection of Mycobacterium butyricum in adjuvant at the basis of the tail. At the onset of arthritis, rats were daily treated (i.p.) with naproxen (10 mg/kg/d), diclofenac (5mg/kg twice a day), celecoxib (3 mg/kg/d) or saline (Vehicle) for 21 days. Arthritis scores were daily monitored. At the end of treatment, thoracic aortas were harvested to measure the relaxation to acetylcholine (Ach) on pre-constricted aortic rings in the presence or not of inhibitor of nitric oxide (NO) synthase (L-NAME), arginase (nor-NOHA), EDHF (Apamin/Charybdotoxin), or a superoxide dismutase analog (Tempol). The effects of a NO donor (sodium nitroprussiate, SNP) and norepinephrine (NE) were studied on endothelium-denuded aortic rings. Results All NSAIDs significantly reduced to the same extent arthritis score in AIA (p<0.05). By contrast, differences exist between NSAIDs regarding endothelial function. Both naproxen and diclofenac significantly (p<0.05) improved Ach-induced relaxation whereas celecoxib did not change the response to Ach. These differences on endothelial function were not explained by differences in the response of vascular smooth muscle to NE or SNP which remained unchanged after the 3 treatments. Of interest, the positive effects of naproxen and diclofenac likely relied on different mechanisms. Whilst naproxen tends to improve NOS activity and decrease superoxide anion production, diclofenac decreased arginase activity, superoxide anion production and improved EDHF production. Conclusions Our study demonstrates for the first time that naproxen and diclofenac have positive effects on endothelial function in case of arthritis whereas celecoxib had no effect, despite a similar reduction of inflammatory symptoms. From a clinical perspective, these data highlight the heterogeneity of effects of NSAIDs as regards CVR in RA and encourage designing clinical trials to confirm our experimental data and help guiding the better choice of NSAIDs for the global management of RA patients. Disclosure of Interest None declared