E. Gremese

Long-term anti-TNF therapy and the risk of serious infections in a cohort of patients with rheumatoid arthritis: Comparison of adalimumab, etanercept and infliximab in the GISEA registry

OBJECTIVE To evaluate the risk of serious infections (SIs) in RA patients receiving anti-TNF therapy on the basis of the data included in the GISEA register. METHODS The study involved 2769 adult patients with long-standing RA (mean age 53.2±13.4 years; mean disease duration 9.0±8.3 years) enrolled in the GISEA register, who had been treated for at least 6 months with TNF inhibitors or had discontinued therapy due to SI: 837 (30%) treated with infliximab (IFN), 802 (29%) with adalimumab (ADA), and 1130 (41%) with etanercept (ETN). RESULTS 176 patients had experienced at least one of the 226 Sis during the 9 years of treatment with an anti-TNF agent, an overall incidence of 31.8/1000 patient-years (95% CI 25.2-38.3): 23.7/1000 patient-years (95% CI 13.1-34.2) on ADA; 12.8/1000 patient-years (95% CI 6.3-19.4) on ETN and 65.1/1000 patient-years (95% CI 48.4-81.8) on IFN. The risk was higher in the first than in the second year of treatment, but this difference was not statistically significant (p=0.08) (38.9% of the SIs were recorded in the first 12 months of treatment). The risk of SI was significantly different among the three treatment groups (p<0.0001). Multivariate models confirmed that the use of steroids (p<0.046), concomitant DMARD treatment during anti-TNF therapy (p=0.004), advanced age at the start of anti-TNF treatment (p<0.0001), and the use of IFN or ADA rather than ETN (respectively p<0.0001 and p=0.023) were strong and statistically significant predictors of infection. CONCLUSIONS Anti-TNF therapy is associated with a small but significant risk of SI that is associated with the concomitant use of steroids, advanced age at the start of anti-TNF treatment, and the type of anti-TNF agent.

SAT0066 Histological and ultrasound synovial predictors of clinical differentiation to defined arthritis in patients with seronegative undifferentiated peripheral inflammatory arthritis

Background Undifferentiated Peripheral Inflammatory Arthritis (UPIA) is a common diagnosis at the first clinical evaluation in rheumatological settings. However, the likelihood of developing a well-defined rheumatic disease in UPIA patients is still matter of debate. Objectives To examine the role of ultrasound (US) and histological parameters in the disease outcome of patients with seronegative UPIA. Methods Fourty-two patients with IgA/IgM-Rheumatoid Factor and anti-citrullinated peptide antibodies negative UPIA, naïve to any Disease-Modifying Anti-Rheumatic Drugs, underwent Gray Scale (GSUS) and power Doppler (PDUS) evaluation and US guided synovial tissue biopsy. Synovial expression of CD68, CD3, CD21, CD20 and CD31 was evaluated by immunohistochemistry. IL-6, VEGF-A and VEGF-D peripheral blood (PB) and synovial fluid (SF) levels were measured by ELISA. To exclude Reactive Arthritis, each patient underwent genital and throat swabs. Afterwards, each UPIA patient was treated with chloroquine 250 mg/daily and followed every 3 months for 1 year and classified as having UPIA, Rheumatoid Arthritis (RA), Spondyloarthritis (SpA) or Psoriatic Arthritis (PsA), respectively. Results During the follow-up 6 (14.3%) UPIA reached a defined diagnosis (2 RA, 2 SpA and 2 PsA, respectively). At baseline, UPIA who differentiated had higher GSUS (p=0.01) and PDUS scores (p=0.02) compared to patients who remained as UPIA within 1 year. At baseline, UPIA who differentiated towards defined arthritis had higher histological scores for lining and sublining CD68+ (p=0.005 and p=0.04 for lining and sublining, respectively), sublining CD3+ cells (p=0.002) and CD31+ vessels count (p<0.001) than patients who remained as UPIA. In addition, there were direct correlations between baseline GSUS and PDUS scores with lining CD68+ cells scores (p<0.001 for GSUS and p=0.02 for PDUS scores respectively), sublining CD68+ cells scores (p=0.02 for GSUS and p=0.03 for PDUS scores respectively), sublining CD3+ cells score (p=0.002 for GSUS and p=0.002 for PDUS scores respectively) and CD31+ vessels count (p<0.001 for GSUS and p=0.01 for PDUS scores respectively) in UPIA. Finally, the areas under the receiver operating characteristic (ROC) curves CD31+ vessels count (cut-off value: 24.3), GS score (cut-off value: 1.5) and PDUS score (cut-off value: 1.5) were calculated to assess the best cut-off points to identify the differentiation likelihood during the follow-up in UPIA patients. The logistic regression analysis, demonstrated that having baseline GSUS and PDUS scores ≥1.5 [OR:13.64 (95% CI: 0.98–242.59); p=0.05] and CD31+ vessels count ≥24.3 [OR:51.13 (95% CI: 3.15–829.16); p=0.01] were independent factors associated with the achievement of defined arthritis. Conclusions Histological and US assessment may help in the identification of patients with seronegative UPIA with high likelihood of clinical differentiation towards defined arthritis. Disclosure of Interest None declared

Achievement of sustained deep remission with adalimumab in a patient with both refractory ulcerative colitis and seronegative erosive rheumatoid arthritis

Inflammatory bowel disease (IBD) is commonly associated with peripheral inflammatory arthritis, and it has been estimated that as many as 12% of IBD patients report these manifestations. However, rheumatoid arthritis (RA) is rarely associated with ulcerative colitis (UC). Among all the biological agents available, nine have been currently approved for the treatment of RA. Conversely, only Infliximab and recently Adalimumab have been approved for UC. In particular, the efficacy of Adalimumab in UC has been demonstrated by both recent randomized controlled trials and real-life studies. Moreover, Adalimumab is a well-established treatment for RA. Herein, we describe a patient with RA and UC treated successfully with ADA.

OP0107 Adipocytokines imbalance is associated with vascular damage in systemic sclerosis

Background Adipocytokines are implicated in the development of fibrosis, vasculopathy and immune abnormalities through a variety of biological effects, but their role in systemic sclerosis (SSc) is not fully investigated. Chemerin is implicated in chemotaxis of immune cells, in promoting angiogenesis and it is involved in inflammation. Adiponectin (APN) has metabolic actions and anti-inflammatory properties, while Leptin (LEP) mediates actions in endothelial cells, such as angiogenesis, vasodilation, NO production and upregulates various mediators of vascular inflammation. Objectives In this study we investigated Chemerin, LEP and APN levels in SSc patients according to disease subtypes and clinical characteristics. Methods Chemerin, LEP and APN levels were evaluated in 100 SSc patients and in sex, age and BMI matched healthy controls. Clinical and demographical characteristics were available for all patients. Results Chemerin, APN and LEP levels were lower in SSc patients compared to healthy controls (Chemerin: 58.7±27.6 ng/ml vs 74.0±29.0 ng/ml, p=0.004; LEP:19.6±18.3 ng/ml vs 28.5±23.8 ng/ml, p=0.03, APN: 6.5±3.9 µg/ml vs 12.8±6.0 µg/ml, p<0.001) Chemerin levels were lower in patients with anti-topoisomerase antibodies (50.2±22.7 ng/ml) respect to patients with other autoantibodies (64.6±29.7 ng/ml), p=0.018. Regarding capillaroscopic damage, Chemerin levels were lower in patients with late pattern (44.8±18.9 ng/ml) compared to patients with early (64.3±28.5 ng/ml) and active pattern (71.7±29.9 ng/ml), p<0.001. APN levels inversely correlate with IL-6 levels (R=-0.4, p<0.001), while directly correlate with capillary density (R=0.3,p=0.03). Patients with avascular areas presented lower levels of APN (5.3±3.9 µg/ml) compared to patients without avascular areas (7.3±3.4 µg/ml),p=0.005. LEP levels directly correlate with vascular density on nailfold capillaroscopy (R=0.3, p=0.02), confirming the role of LEP in endothelial homeostasis. Furthermore, patients with avascular areas presented lower LEP levels (15.5±13.0 ng/ml) compared to patients without avascular areas (31.1±28.4 ng/ml), p=0.003. LEP levels were lower in patients with active digital ulcers (9.3±6.6 ng/ml), compared to patients without ulcers (9.3±6.6 ng/ml), p=0.01. The anti-inflammatory and endothelium protective role of APN emerged also when we considered the lung involvement: in fact patients with DLCO >50% presented higher levels of APN (7.0±3.9 µg/ml) compared to patients with DLCO <50% (5.8±3.8 µg/ml), p=0.05. Considering the cardiopulmonary involvement, LEP levels inversely correlate with PAPs on echocardiography (R=-0.24, p=0.02). Finally LEP levels inversely correlate with skin score (R=-0.3, p=0.009) and patients with early disease presented lower LEP levels (15.1±13.2 ng/ml) compared to patients with long lasting disease (29.9±28.7 ng/ml), p=0.006. Conclusions Our data suggest an imbalance of the levels of adipocytokines in SSc, their down-regulation in patients with a more aggressive pattern on nailfold videocapillaroscopy and organ damage, suggesting a possible role of Chemerin, LEP and APN in the impaired angiogenesis and in the development of vasculopathy of SSc patients. Disclosure of Interest None declared

SAT0052 High chance of comprehensive disease control (CDC) in very early and normal weight rheumatoid arthritis patients treated according to the treat to target strategy

Background The ultimate goal for Rheumatoid Arthritis (RA) management is the simultaneous achievement of all clinical, functional and structural efficacy, i.e. comprehensive disease control (CDC) [1]. Objectives To evaluate the effective chance and the consequences of CDC achievement in real world practice of Early Arthritis Clinic (EAC). Methods A total of 349 early rheumatoid arthritis (ERA) patients with a disease duration of less than 12 months were enrolled in the study. ERA patients fulfilled the 2010 ACR criteria for RA and were followed according to the treat-to-target strategy. Subjects with symptom duration less than 3 months were defined as having “very early RA” (VERA). The mean follow-up (FU) was 38.2±32.8 months. At baseline, and every three months, the ACR/EULAR core data set variables were recorded. At baseline and every year hand and foot radiographs were examined according to modified Total Sharp score (mTSS). At each visit, clinical improvement and remission were evaluated according to EULAR criteria. The achievement of CDC (28-joint Disease Activity Score using C reactive protein <2.6, Health Assessment Questionnaire <0.5 and change from baseline in mTSS ≤0.5) was assessed every year of follow-up. Results At the twelfth month of FU 148 (42.4%) ERA patients achieved CDC, while at the time of last FU 228 (65.3%) subjects reached this target. Patients achieving CDC at the 12th month of FU were younger (p=0.05), in higher percentage male (p=0.004), and with a normal weight (body mass index, BMI <25) (p=0.003) and had a shorter disease duration, comprising a greater number of VERA (p=0.01), compared to subjects not achieving disease control. There were no differences concerning autoantibody positivity and presence of erosions at onset between the two analyzed cohorts. Adjusting the analysis for age, the variables that arose as independent predictors of CDC at the 12th month of FU were a disease duration less than 3 months [OR (95% CI): 1.97 (1.23–3.14)] and a normal BMI [OR (95% CI): 2.05 (1.32–3.21)]. In our cohort, 105 (30.1%) ERA patients were treated with biological disease modifying anti-rheumatic drugs (bDMARDs) over time. Biotechnological therapy was less frequently started by subjects in CDC, both after 12 months (p=0.003) and at the time of last FU (p<0.0001). At the multivariate analysis, not achieving CDC at the 12th month of FU [OR (95% CI): 2.69 (1.59–4.57)] and a BMI ≥25 [OR (95% CI): 2.05 (1.23–3.42)] were the variables significantly associated to bDMARD therapy over time. Conclusions The simultaneous achievement of symptom control, inhibition of radiographic progression and normalization of function, is a feasible target in real word EAC. Having a VERA and a normal weight are associated to a high chance of “deep” remission. References Emery P, et al. Ann Rheum Dis 2015; 74: 2165–2174. Disclosure of Interest None declared

OP0239 Histological features of joint and colonic inflammation in inflammatory bowel disease patients treated with anti-tnf

Background New onset of joint inflammation in patients under anti-TNF-alpha for inflammatory bowel disease (IBD) has been previously described. However, histological characterization of synovial and bowel compartments has not been reported so far. Objectives Aim of the study was to evaluate the histological characteristics of paired synovial (ST) and colonic tissues in IBD patients under TNF-alpha blockers. Methods Consecutive IBD patients without history of co-existing joint involvement who developed peripheral arthritis under TNF-alpha blockers, were prospectively enrolled. Each patient underwent rheumatological evaluation and ultrasound (US) assessment (using Gray scale for synovial hyperthrophy and Power Doppler Signal) of the affected joints. Each patient underwent US guided ST biopsy of the knee, following a standardized procedure1 and colonoscopy with mucosal biopsies. Each ST and colonic paired sample was stained through immunohistochemistry (IHC) for CD68, CD21, CD20, CD3 and CD1172. H&E staining was performed for Paneth cells identification. Clinical and immunological parameters [Anti-citrullinated peptides antibodies (ACPA), IgM-Rheumatoid Factor (RF) and IgA-RF respectively] were collected for each patient. Results 10 patients with IBD [46.0±9.7 years old, 13.2±9.9 years of disease duration, 2.5±1.6 years of TNF-alpha blockers exposure, 6 with Crohns Disease and 4 with Ulcerative Colitis respectively] were studied. All patients were negative for ACPA, IgM-RF or IgA-RF and 4 patients were under Methorexate therapy. 5 (50.0%) patients showed endoscopic and histologically proven inflammation of colonic mucosa. Moreover, IHC revealed that 6 (60.0%) patients had diffuse and 4 (40.0%) had follicular synovitis, respectively. In particular, there was a direct correlation between CD68+, CD21+, CD3+, CD20+ and CD117+ cells distribution in paired ST and gut tissues in the whole cohort (p<0.05). No significant differences in terms of disease duration (p=0.48), TNF-alpha blockers exposure time (p=0.29), ESR (p=0.26) and CRP (p=0.91) values were found comparing patients with follicular and diffuse synovitis respectively. Conclusions Our findings suggest that patients with IBD may develop histologically proven synovitis during TNF-alpha treatment, showing similar histological features in terms of CD68+, CD21+, CD20+, CD3+ and CD117+ cells between synovial and colonic compartments. Molecular mechanisms triggered by TNF-alpha blockers leading to joint inflammation have to be clarified. References van de Sande MJT et al. Ann Rheum Dis 2011. Alivernini S. et al. Nat Communications 2016. Disclosure of Interest None declared

Is Citrullination Required for the Presence of Restricted Clonotypes Reacting With Type II Collagen? Comment on the Article by Chemin et al.

tory phenotype in response to carbamylated collagen, it is possible that T cells restricted to DRb1*04:01 are either deleted in the thymus or are tolerized to this collagen epitope, and carbamylation of this peptide alone may not be sufficient to overcome that tolerance. Based on the data presented in our recent article, in which we demonstrated differences in binding of citrullinated versus native vimentin and a-enolase (which correlate with disease susceptibility), we would be interested to see the effects of peripheral blood mononuclear cell stimulation with these peptides in carbamylated form.

SAT0155 Signature of Inflammatory Genes in B-Cells and Double Negative Memory Phenotype as Biomarkers of Response To Tocilizumab

Background Tocilizumab (TCZ) is an effective treatment in Rheumatoid Arthritis (RA) inducing modifications of B cell subsets in vivo.1–3 Objectives To define whether the study of circulating B cell subsets could represent a possible tool to understand the pharmacological action of TCZ, helping the clinician to identify the best responders to IL6R blockade. Methods 70 RA patients (age 52.3±15.2 years, disease duration 8.5±10.1 years, 74.3% female) not responder to previous cDMARDs (n=39) and/or bDMARDS (n=31) were enrolled in the study, of which 19 (27.1%) with early RA (ERA) and 51 (72.9%) with a long-standing disease (LSRA). All patients were treated with TCZ at a dose of 8 mg/kg every 4 weeks. At each visit remission was evaluated according to ACR/EULAR criteria4 and peripheral blood (PB) derived B cells were collected for the analysis of B-cell subsets distribution by FACS using IgD-CD27 classification5. Among them, PB-derived B cells from 13 bDMARDs naïve RA patients were analyzed for specific genes expression through RT-PCR. Results In RA cohort, 16.7% and 19.3% of TCZ-treated RA patients reached ACR/EULAR remission (SDAI≤3.3) at 6th and 12th months FU. At TCZ baseline, a higher percentage of post-switched memory B-cells (p=0.005) together with plasmablasts (p<0.001) and a lower percentage of pre-switched memory B cells (p=0.01) were observed in RA patients compared to Healthy Controls (HC, n=45) mainly in bDMARDs naïve patients. After TCZ treatment, we showed an increase of pre-switched memory and a reduction of post-switched memory and double negative (DN) B cells in responders, more evident in patients reaching remission. In particular, patients reaching SDAI remission at 6 and 12 months of TCZ-treatment had a higher reduction of DN B cell subsets already after 3 months reaching a comparable value as in HC (p<0.05). At the multivariate analysis, a disease duration less than 12 months [OR (95%CI): 18.0 (1.6–203.9)], a moderate disease activity (SDAI≤26) at baseline [OR (95%CI): 20 (1.7–250)] and a higher reduction of DN B cells at 3 months of FU [OR (95%CI): 1.05 (1.01–1.08)] arose as significant independent predictors of ACR/EULAR remission (SDAI≤3.3) at 6th months of TCZ treatment. Finally, TCZ-treatment seems to influence B-cell genes expression inducing an up-regulation of proinflammatory molecules (TNF-alpha, IL6) genes, together with the related STAT3, SYK and ZAP-70 genes, in the earlier treatment phases (3 and 6 months respectively) and the subsequent down-regulation of their receptors (TNF-RII and IL6R). Interestingly, the expression of JAK1 decreased significantly at the 12th month, whereas that of JAK3 did not change at the different time points. Conclusions In our cohort of TCZ-treated RA patients, a significant reduction of DN memory B cells at the 3th month of FU has emerged as an early biomarker of remission afterwards. At the same time, profound effects at the JAK1 gene level occurred in B cells after TCZ treatment, even if an increase of Syk and ZAP-70 genes expression was observed, suggesting that B cells are not fully deactivated even after 6–12 months of treatment. References Roll P et al. Arthritis Rheum 2011; Muhammad K et al. Ann Rheum Dis 2011; Mahmood Z et al. Arth Res Ther 2015; Felson DT et al. Arthritis Rheum 2011; Sanz I et al. Semin Immunol 2008. Disclosure of Interest None declared

AB0904 Anti-Mullerian Hormone in A Cohort of Young Adult Women with Juvenile Idiopathic Arthritis

Background Juvenile idiopathic arthritis (JIA) represents one of more common chronic disease of the childhood, affects young people before sixteen and persists into their reproductive years. These patients sometimes experience significant long-term morbidity while continuing immunosuppressive treatment. Anti-Mullerian hormone (AMH) is produced in the ovary by granulose cells of early developing follicles and in fertility clinics is used to estimate ovarian reserve which constitutes the number of primordial follicles. Objectives To evaluate AMH serum levels in a cohort of young adult women affected from JIA with respect to healthy control comparable for age, to understand whether the disease may compromise ovarian reserve and to assess the influence of previous exposure to disease-modifying antirheumatic drug (DMARDs) and of other disease parameters on serum AMH concentration. Methods 29 young women with a diagnosis of JIA, aged 18 to 26 years and having regular menses, and 20 healthy women comparable for age were recruited. AMH blood levels was assayed according to a 2-stage enzyme-linked immunosorbent assay (ELISA) technique using a commercially available kit (AMH Gen II ELISA; Beckman Coulter). Assessment of age, body mass index (BMI), disease duration, duration of previous treatments, 44 swollen and tender joint count, disease activity score on 44 joints (DAS), health assessment questionnaire (HAQ) were performed at the time of blood sample. Results JIA patients had a mean age of 21.6±2.8 years, a disease duration of 12.0±6.1, a BMI of 24.2±5.1, a DAS of 1.25±0.65, and 6 (20.7%) were smokers. No significant differences were found between JIA and healthy subjects in AMH serum levels (6.1±2.3 vs 6.3±2.4 ng/ml, respectively, p=0.77). Considering the JIA cohort, 20 patients (69.0%) were treated with methotrexate (MTX) for a mean period of 4.2±3.6 years (range 1-13) and 15 (51.7%) with anti-TNF drugs for 4.9±1.9 years (range 1-8). 12 (41.3%) JIA women were treated with both MTX and anti-TNF. No correlations were found between AMH serum levels and patients age (p=0.93), disease duration (p=0.83) and duration of therapy with MTX (p=0.35) or anti-TNF (p=0.47). Dividing JIA patients according to MTX use, no differences were observed between MTX users and nonusers patients in AMH levels (6.0±2.0 vs 6.4±3.0 ng/ml, respectively, p=0.48), age, disease duration and other clinical characteristics. Similarly, AMH levels were comparable also between anti-TNF users and non users JIA subjects (p=0.11). Conclusions In our JIA group of young adult JIA women, ovarian reserve seems not be changed by the presence of the disease, the long disease duration and the use of immunosuppressive drugs. These findings could be important for adult JIA patients. Disclosure of Interest None declared DOI 10.1136/annrheumdis-2014-eular.5009

AB0004 Collagen-specific tcr repertoire usage in rheumatoid arthritis and cytokine secretion.

Background We have previously reported the characterization of the TCR-beta chain repertoire of T cells specific for Human Collagen II p261-273 in DR4+ RA patients and DR4+healthy controls (1). Objectives We observed that it is possible to identify a limited number of TCR-beta chain recurrently used in patients during the acute presentation of the disease, as opposed to the same patients after disease amelioration, versus healthy control subjects. We could also show that a part of this repertoire is spontaneously enriched in the Synovial fluid. Methods We examined the usage of the two more commonly used TCR-beta chains (BV11 and BV13) in 85 consecutive samples from RA patients of various HLA DR haplotype (25 DR4+, 13 DR1+, 11 DR7+ and 37 of other haplotypes; one patient was DR4+ DR1+) at acute presentation (47) or remission (38) of disease. Peripheral Blood Mononuclear Cells (PBMC) were isolated, cultured in the absence or presence of collagen peptide and examined by immunoscope, as described. In addition, we examined the ability of collagen-specific individual T cells of 3 DR4+ patients to secrete IL-17 and IL-13 upon stimulation in vitro with the peptide, in the absence or presence of bacteria-derived products, following the method recently described (2). Results Confirming the previously reported observations, 40 (6/15) and 50 (7/15)% of RA patients in acute disease display the presence of BV11+ or BV13+ T cells. Pooling patients positive for either or both TCRs, the frequency of positive samples increased only to 8/16. The frequency of the same T cells in the collagen-specific repertoire residual after induction of remission decreases to 1/11 and 2/10, respectively. Two more groups of RA patients displayed collagen-specific T cells using similar BV11 and BV13 beta chains, namely DR1+ patients (2/9 in each case) and DR7+ patients (3/6 in each case). In the case of DR7+ patients, however, the frequency of positive samples did not decrease following remission of symptoms. Three out of 18 patients with other haplotypes showed usage of BV11+ cells and 1/19 used BV13+ cells. We examined the ability of a total of 17 individual T cells from 3 DR4+patients (1 in acute disease and 2 in remission) to secrete IL-17 and IL-13. When PBMC were stimulated in the presence of peptide only, just 1 T cells was able to secrete IL-17 and 1 IL-13, both observed in the patient in acute disease. However, when stimulation with the peptide was associated with the presence of bacterial products, 3 more individual T cells became able to secrete IL-17. Conclusions We confirm that BV11 and BV13+ cells are bystander of the acute presentation of the disease in DR4+ patients. The relative high frequency of usage of these clonotipic TCRs in DR1+ and DR7+ patients is possibly due to the similarities in presentation of the antigen between these DR alleles. The secretion of the highly pro-inflammatory cytokine IL-17 is modulated by endogenous or exogenous factors (possibly interacting with PRRs) which may thereby influence the severity of local damage. References Ria F et al, Arth Res Ther 2008, 10:R135. Nicolò C et al, J Immunol 2010, 184:222-235. Disclosure of Interest None Declared