Search

BackgroundType I interferons (IFN-Is) play a role in a broad range of rheumatic and musculoskeletal diseases (RMDs), and compelling evidence suggests that their measurement could have clinical value, although testing has not progressed into clinical settings.ObjectiveTo develop evidence-based points to consider (PtC) for the measurement and reporting of IFN-I assays in clinical research and to determine their potential clinical utility.MethodsEULAR standardised operating procedures were followed. A task force including rheumatologists, immunologists, translational scientists and a patient partner was formed. Two systematic reviews were conducted to address methodological and clinical questions. PtC were formulated based on the retrieved evidence and expert opinion. Level of evidence and agreement was determined.ResultsTwo overarching principles and 11 PtC were defined. The first set (PtC 1–4) concerned terminology, assay characteristics and reporting practices to enable more consistent reporting and facilitate translation and collaborations. The second set (PtC 5–11) addressed clinical applications for diagnosis and outcome assessments, including disease activity, prognosis and prediction of treatment response. The mean level of agreement was generally high, mainly in the first PtC set and for clinical applications in systemic lupus erythematosus. Harmonisation of assay methodology and clinical validation were key points for the research agenda.ConclusionsIFN-I assays have a high potential for implementation in the clinical management of RMDs. Uptake of these PtC will facilitate the progress of IFN-I assays into clinical practice and may be also of interest beyond rheumatology.

ObjectivesTo systematically review the literature for assay methods that aim to evaluate type I interferon (IFN-I) pathway activation and to harmonise-related terminology.MethodsThree databases were searched for reports of IFN-I and rheumatic musculoskeletal diseases. Information about the performance metrics of assays measuring IFN-I and measures of truth were extracted and summarised. A EULAR task force panel assessed feasibility and developed consensus terminology.ResultsOf 10 037 abstracts, 276 fulfilled eligibility criteria for data extraction. Some reported more than one technique to measure IFN-I pathway activation. Hence, 276 papers generated data on 412 methods. IFN-I pathway activation was measured using: qPCR (n=121), immunoassays (n=101), microarray (n=69), reporter cell assay (n=38), DNA methylation (n=14), flow cytometry (n=14), cytopathic effect assay (n=11), RNA sequencing (n=9), plaque reduction assay (n=8), Nanostring (n=5), bisulphite sequencing (n=3). Principles of each assay are summarised for content validity. Concurrent validity (correlation with other IFN assays) was presented for n=150/412 assays. Reliability data were variable and provided for 13 assays. Gene expression and immunoassays were considered most feasible. Consensus terminology to define different aspects of IFN-I research and practice was produced.ConclusionsDiverse methods have been reported as IFN-I assays and these differ in what elements or aspects of IFN-I pathway activation they measure and how. No ‘gold standard’ represents the entirety of the IFN pathway, some may not be specific for IFN-I. Data on reliability or comparing assays were limited, and feasibility is a challenge for many assays. Consensus terminology should improve consistency of reporting.

EULAR [SCI019]; UK National Institute for Health and Care Research (NIHR) Leeds Biomedical Research Centre, Rodríguez-Carrio J, Burska A, Conaghan PG, Dik WA, Biesen R, Eloranta ML, Cavalli G, Visser M, Boumpas DT, Bertsias G, Wahren-Herlenius M, Rehwinkel J, Frémond ML, Crow MK, Ronnblom L, Vital E, Versnel M

Gene expression profiles are associated with the clinical heterogeneity of SLE but are not well studied as biomarkers for therapy. Many clinical and demographic features influence treatment responses. We studied gene expression and response to rituximab in a multi-ethnic UK cohort refractory to standard therapy.Baseline expression of transcripts known to associate with clinical features of SLE was evaluated in whole blood by 96-probe Taqman® array in patients (n=213) with active SLE, prospectively enrolled in British Isles Lupus Assessment Group (BILAG) Biologics Registry. Autoantibodies were measured using immunoprecipitation and ELISA. Response to first cycle rituximab (n=110) was determined by BILAG-2004 criteria at 6 months.Interferon scores were lower in European ancestry patients than all other groups. The relationship between blood interferon scores and plasmablast, neutrophil, myeloid, inflammation and erythropoiesis-annotated scores differed between patients of European and non-European ancestries. Hierarchical clustering revealed 3 distinct non-European ancestry patient subsets with stratified response to rituximab which was not explained by sociodemographic and clinical variables. Response was lowest in an interferon-low, neutrophil-high cluster and highest in a cluster with high expression across all signatures (p0.001). Clusters within European ancestry patients did not predict response to rituximab but segregated patients by global disease activity and renal involvement. In both ancestral groups, interferon-high clusters associated with U1RNP-Sm antibodies.Ancestry appears central to the immunological and clinical heterogeneity in SLE. These results suggest that ancestry, disease activity and transcriptional signatures could each assist predict the effectiveness of B-cell depletion. This article is protected by copyright. All rights reserved.

ACR/EULAR-2010 classification criteria for rheumatoid arthritis (RA) rely heavily on the presence of anti-citrullinated peptide antibody (ACPA). The role of anti-carbamylated protein antibodies (anti-CarP) in this context is uncertain. We aimed to investigate the value of anti-CarP for RA classification in patients with early inflammatory arthritis.Patients (n=402) were recruited from an early arthritis clinic and followed for 24 months. Healthy controls (n=95) were included. An anti-CarP ELISA was performed (aU/mL). Statistical analysis used regression and AUC analysis.The criteria for RA were met by 195/402 patients at inclusion; 28 developed RA during follow-up and 179 had other diagnosis (non-RA). 97/195 (49%) RA patients were anti-CarP+ (median 250 uA/mL [IQR 25-762]). In the group that progressed to RA, 7/28 (25%) were positive (82 uA/mL [13-235]) compared to non-RA (p=0.001) with 13/179 (7%) positive (26 uA/mL [5-80]). Being anti-CarP+ alone was observed in 17 patients of whom 7 (41%) were RA. Levels/positivity were not associated with other parameters. Anti-CarP+ had an odds ratio (OR) 6.5 for predicting RA (OR=17.1 for ACPA+ and OR=2.5 for RF+). In ACPA- patients, anti-CarP+ was also predictive of RA (OR=2.39). Being ACPA+/anti-CarP+/RF+ had a high predictive value for RA (OR=29.9 sensitivity/specificity (sen/spe) 33%/99%, positive/negative predictive values (ppv/npv) 97%/54%), however, being ACPA+/anti-CarP+ was superior (OR=36.1 sen/spe=41%/99%, ppv/npv=98%/57%) while being ACPA+/RF+ was inferior (OR=11.9, sen/spe=54%/95%, ppv/npv=94%/62%).For RA classification, anti-CarP+ was less sensitive than ACPA, but more specific than RF. Anti-CarP+ may prove useful, classifying early arthritis patients, notably ACPA- patients.

ObjectivesTo explore the prognostic value of subclinical cardiovascular (CV) imaging measures and serum cardiac biomarkers in systemic sclerosis (SSc) for the development of CV outcomes of primary heart involvement (pHI).MethodsPatients with SSc with no clinical SSc-pHI and no history of heart disease underwent cardiovascular magnetic resonance (CMR) imaging, and measurement of serum high-sensitivity-troponin I (hs-TnI) and N-terminal-pro-brain natriuretic peptide (NT-proBNP). Follow-up clinical and CV outcome data were recorded. CV outcomes were defined as myocarditis, arrhythmia and/or echocardiographic functional impairment including systolic dysfunction and/or diastolic dysfunction.ResultsSeventy-four patients with a median (IQR) age of 57 (49, 63) years, 32% diffuse cutaneous SSc, 39% interstitial lung disease, 30% Scl70+ were followed up for median (IQR) 22 (15, 54) months. Ten patients developed CV outcomes, comprising one patient with myocarditis and systolic dysfunction and nine arrhythmias: three non-sustained ventricular tachycardia and six supraventricular arrhythmias. The probability of CV outcomes was considerably higher in those with NT-proBNP >125 pg/mL versus normal NT-proBNP (X2=4.47, p=0.035). Trend for poorer time-to-event was noted in those with higher extracellular volume (ECV; indicating diffuse fibrosis) and hs-TnI levels versus those with normal values (X2=2.659, p=0.103; X2=2.530, p=0.112, respectively). In a predictive model, NT-proBNP >125 pg/mL associated with CV outcomes (OR=5.335, p=0.040), with a trend observed for ECV >29% (OR=4.347, p=0.073).ConclusionThese data indicate standard serum cardiac biomarkers (notably NT-proBNP) and CMR indices of myocardial fibrosis associate with adverse CV outcomes in SSc. This forms the basis to develop a prognostic model in larger, longitudinal studies.

Background/Aims We developed two interferon-stimulated gene (ISG) expression scores (IFN-Score-A and IFN-Score-B) that predict clinical outcomes in SLE better than a conventional “interferon signature”. IFN-Score-A includes the ISGs usually present in an interferon signature. IFN-Score-B includes additional ISGs responsive to multiple IFN subtypes and was better at predicting development of SLE among ANA-positive individuals, development of RA among CCP-positive individuals, and subclinical synovitis in SLE. Here, we investigate whether IFN-Score-A and IFN-Score-B predict response to Rituximab (RTX) therapy. The British Isles Lupus Assessment Group Biologics Register (BILAG-BR) is a UK wide study enrolling patients undergoing treatment with RTX for active SLE despite previous treatment with cyclophosphamide or mycophenolate mofetil. MASTERPLANS is a consortium aimed at identifying therapeutic biomarkers in SLE. Methods We studied all BILAG-BR subjects undergoing their first cycle of RTX in whom a pre-treatment RNA sample was available. Disease activity was measured blind to biomarker status, using BILAG-2004. Response was defined as improvement in BILAG-2004 disease activity by at least one grade, with a maximum of one domain showing persistent BILAG-2004 grade B activity, and no new BILAG grade A or B disease flares at 6 months. Whole blood was collected into TEMPUS tubes and RNA extracted for measurement of IFN-Scores using a custom Taqman array as previously described. Multivariable logistic regression was used to test IFN-Scores and key baseline clinical covariates as predictors of response at 6 months. Results 147 patients were recruited, of whom there were 6 month BILAG data in 90. 59/90 (65.6%) were responders. In univariable and multivariable analysis, higher IFN-Score-B expression was significantly associated with clinical response (Table 1). Other characteristics typically associated with more severe SLE (younger age, African ancestry, and autoantibody repertoire) also independently predicted response. Conclusion High expression of interferon scores predicts better response to RTX. The novel IFN score-B was more predictive than a typical interferon signature. Future work will validate this biomarker in a replication cohort and integrate other with data on IFN biomarkers in other contexts. P169 Table 1:Baseline VariableNon respondersRespondersUnivariable OR (95% CI)P valueMultivariable OR (95% CI)P valueInterferon Score-A, median (IQR)-1.75 (-4.7,-1.0)-1.1 (-2.24,-0.1)1.21 (0.98,1.48)0.070.94 (0.58,1.53)0.81Interferon Score-B, median (IQR)-2.4 (-3.2,-1.6)-1.6 (-2.6,-1.11.56 (1.04,2.33)0.033.27 (1.39,7.68)0.007Age years, median (IQR)44 (38,52)38 (30,50)0.97 (0.94,1.00)0.060.92 (0.86,0.97)0.005African ancestry, n/N (%)8/31 (26)4/59 (7)0.21 (0.06,0.76)0.020.005 (0.000,0.22)0.006Numerical BILAG-2004, median (IQR)21 (20,25)18 (13,24)0.91 (0.86,0.93)0.010.80 (0.67,0.94)0.010Count of ANA subtypes, median (IQR)1 (1,3)2 (1,3)1.28 (0.84,1.95)0.252.38 (1.06,5.31)0.034Count of ANA subtypes is sum of positive Ro, La, Sm, RNP & dsDNA by local laboratories. Disclosure L.M. Carter: None. A. Alase: None. Z. Wigston: None. A. Burska: None. A. Psarras: None. Y. Yusof: None. J. Reynolds: None. M. Wittmann: Honoraria; Abbvie, Celgene, Janssen, L’Oreal, Novartis and Pfizer. I. Bruce: Consultancies; AstraZeneca, Eli Lilly, GlaxoSmithKline, ILTOO Pharma, MedImmune, Merck Serono. Member of speakers’ bureau; GlaxoSmithKline, UCB Pharma. Grants/research support; Genzyme Sanofi & GlaxoSmithKline. E.M. Vital: Honoraria; Roche, GSK and AstraZeneca. Grants/research support; Roche, GSK and AstraZeneca.

Objectives SSc primary heart involvement (SSc-pHI) is a significant cause of mortality. We aimed to characterize and identify predictors of subclinical SSc-pHI using cardiovascular MRI. Methods A total of 83 SSc patients with no history of cardiovascular disease or pulmonary arterial hypertension and 44 healthy controls (HCs) underwent 3 Tesla contrast-enhanced cardiovascular MRI, including T1 mapping and quantitative stress perfusion. High-sensitivity troponin I and N-terminal pro-brain natriuretic peptide were also measured. Results Cardiovascular MRI revealed a lower myocardial perfusion reserve in the SSc patients compared with HCs {median (interquartile range (IQR)] 1.9 (1.6–2.4) vs 3 (2–3.6), P < 0.001}. Late gadolinium enhancement, indicating focal fibrosis, was observed in 17/83 patients but in none of the HCs, with significantly higher extracellular volume (ECV), suggestive of diffuse fibrosis, in SSc vs HC [mean (s.d.) 31 (4) vs 25 (2), P Conclusion Subclinical SSc-pHI is characterized by myocardial microvasculopathy, diffuse and focal myocardial fibrosis but preserved myocardial contractile function. This subclinical phenotype of SSc-pHI was associated with high-sensitivity troponin I, N-terminal pro-brain natriuretic peptide, SSc disease severity and complicated peripheral vasculopathy. These data provide information regarding the underlying pathophysiological processes and provide a basis for identifying individuals at risk of SSc-pHI.

Objective Ankylosing spondylitis (AS) is a chronic inflammatory arthritis primarily affecting the spine and sacroiliac joints. TNF inhibitor (TNFi) drugs are recommended for patients not responding to NSAIDs; however, there is a significant need for biomarkers of response. IFN-regulated genes (IRGs) and other cytokines/chemokines are linked to autoimmune diseases and have been associated with treatment response. Our objective was to explore whether IRGs and cytokines/chemokines can be associated with response to TNFiagents in AS. Methods Peripheral blood mononuclear cells were obtained from 26 AS patients who were to receive a TNFi (I, n = 15) or placebo (P, n = 11) at week 0 and week 22. Response (R)/non-response (NR) was defined as reduction in ASDAS ≥ 1.2 points or reduction in sacroiliac/vertebral MRI lesions. The expression of 96 genes was quantified using TaqMan assays. Finally, ELISA was used to measure IL-6 in serum samples from another 38 AS patients. Results Analysis of gene expression in 26 baseline samples segregated patients into four groups defined by a signature of 15 genes (mainly IRGs). ASDAS response was associated with one group independently of treatment received. We then analysed response to the TNFi (n = 15) and identified a 12-gene signature associated with MRI response. A third IRG signature was also associated with a reduction in IRGs expression post-TNFi samples (n = 10 pairs). Finally, decreased circulating IL-6 was associated with BASDAI-R. Conclusion This pilot study suggests an association between IRG expression and response to TNFi in AS. These findings require validation in a larger cohort in order to construct predictive algorithms for patient stratification.

Background Rituximab is used for resistant SLE, but clinical response varies. We previously validated two interferon-stimulated-gene expression scores (IFN-Score-A and IFN-Score-B) that improved prediction of clinical outcomes in SLE. IFN-Score-A included most reported ISGs and predicted flares and glucocorticoid requirements. IFN-Score-B included ISGs that respond to multiple IFN subtypes and predicted development of SLE in At-Risk individuals. Diagnosis of SLE was associated with both scores, while only IFN-Score-B was elevated in RA. The British Society for Rheumatology Biologics Registry (BILAG-BR) collects samples for RTX-treated patients in the UK. MASTERPLANS ISA consortium to identify predictors of drug response. Methods Patients were recruited if they were starting a first cycle of RTX for active SLE (BILAG A or 2xBILAG B) despite previous cyclophosphamide or mycophenolate mofetil. Disease activity was measured using BILAG-2004. Clinical response was defined as improvement by > =1 grade in active BILAG-2004 systems with no worsening in other systems. Whole blood was collected into TEMPUS tubes and RNA extracted. IFN-Scores were measured using a custom Taqman array as previously described [El Sherbiny et al., 2018]. Multivariate logistic regression was used to test IFN-Scores and baseline clinical covariates as predictors of BILAG response at 6 months. Results Samples were available from 147 patients, of whom 84 had complete baseline and 6-month clinical data available and were included in this analysis. 40/84 (47.6%) patients had BILAG response at 6 months. In univariate and multivariate analysis, high IFN-Score-B expression was significantly associated with clinical response (see Table 1). Conclusion This preliminary analysis suggests that assessment of IFN activity has a role in predicting response to RTX. A novel IFN score (Score B) was more predictive than classic ISGs (Score A). These results add to a body of work showing that IFN-Score-B predicts clinically significant outcomes independently of overall IFN activity. Future work will analyse this biomarker in a larger cohort of patients and integrate with other putative clinical and biological predictors of response. Disclosures A. Alase None. Z. Wigston None. A. Burska None. M. Md Yusof None. J. Reynolds None. M. Wittmann None. I. Bruce None. E.M. Vital None.

BackgroundA reduction in serum lipids such as total cholesterol (TC) and low-density lipoprotein (LDL) has been associated with increased risk of cardiovascular events in active RA, paradoxical to the general population1. The effect of DMARDs and active inflammation on lipid profiles has been investigated in RA but detailed lipid profiling has been lacking in very early disease and across the spectrum of disease activity.ObjectivesIn a treatment naïve early RA trial cohort, we sought to compare circulating lipid profiles between patients treated with first line etanercept + methotrexate (ETN+MTX) versus methotrexate treat-to-target (MTX-TT) regime and between clinical remission and high disease activity.MethodsVEDERA trial (Very early Etanercept and Methotrexate versus Methotrexate with/without Delayed Etanercept in RA) randomised 120 treatment-naïve RA patients to either first-line ETN+MTX or MTX-TT regime with escalation to ETN+MTX if not in DAS28ESR remission at week 242. TC, triglycerides (TG), high density lipoprotein (HDL) and LDL were measured; apolipoproteins and atherogenic indices such as TC/HDL, atherogenic index of plasma (AIP) and apolipoprotein B/A-i ratio (aporatio) were calculated at baseline, weeks 12, 24 and 48. Linear mixed effects regression was used to test the effect of treatment on lipids and atherogenic indices in states of remission (DAS28-ESR ≤ 2.6) and high disease activity (DAS28-ESR > 5.1).ResultsBaseline clinical characteristics of individuals1 and lipid profiles including atherogenic indices were comparable between the two treatment groups (Table 1).Table 1.Baseline lipids (mmol/L) and atherogenic indices in VEDERA. Median with interquartile range and n/N (%) reported.Total, N = 120MTX_TT, N = 60ETN+MTX, N = 60TC4.69 (4.03,5.25)4.70 (4.04,5.42)4.69 (4.01,5.18)TG1.15 (0.90,1.41)1.18 (0.92,1.32)1.10 (0.89,1.55)HDL1.28 (1.04,1.53)1.29 (1.04,1.57)1.27 (1.05,1.48)LDL2.87 (2.22,3.27)2.87 (2.27,3.19)2.84 (2.01,3.29)TC/HDL3.53 (3.01,4.16)3.53 (3.00,4.14)3.58 (3.08,4.17)AIP-0.06 (-0.20,0.09)-0.06 (-0.22,0.08)-0.06 (-0.17,0.13)Aporatio0.63 (0.54,0.74)0.62 (0.54,0.74)0.64 (0.55,0.74)In clinical remission, a lowering of atherogenic indices and TC, TG, LDL levels as well as a rise in predicted HDL levels were observed. In high disease activity, both HDL and LDL were increased along with the atherogenic indices TC/HDL-C, AIP and aporatio. However, the predicted values at different weeks did not reach statistical significance (not shown).Treatment with MTX-TT and ETN+MTX had opposing effects on predicted HDL levels in remission and high-disease activity (Figure 1). In remission, MTX-TT treatment resulted in a predicted rise in HDL whilst with ETN+MTX a small reduction was observed (estimate 0.004, p 0.02). Similar trends were observed for HDL in high disease activity (p 0.5). In remission, both treatments resulted in a reduction in LDL (p 0.5), whilst in high disease activity ETN+MTX treatment resulted in a modest rise in LDL compared to MTX-TT (p 0.06). At weeks 24 and 48, significant differences were observed in LDL values between treatment groups in high disease activity (estimate 0.57, p 0.05 and estimate 1.13, p 0.04 respectively).ConclusionEffect of early ETN treatment on HDL, and to a lesser extent LDL, differs from MTX and is modulated by clinical disease activity. Further investigation is needed to understand the basis for these findings and the clinical implication of observed differences in LDL and HDL. These data may indicate direct effect of therapies on the qualitative and functional components of metabolic lipid pathway.References[1]Robertson J, Peters MJ, McInnes IB, et al. Changes in lipid levels with inflammation and therapy in RA: A maturing paradigm. Nat Rev Rheumatol. 2013;9(9):513–23.[2]Emery P, Horton S, Dumitru RB, et al. Pragmatic randomised controlled trial of very early etanercept and MTX versus MTX with delayed etanercept in RA: The VEDERA trial. Ann Rheum Dis. 2020;79(4):464–71.Disclosure of InterestsNone declared

BackgroundRA individuals are known to have excess CVD risk, equivalent to that of type 2 diabetes mellitus1. We recently reported vascular and myocardial abnormalities on cardiac magnetic resonance imaging (CMR) in an early RA cohort compared with control subjects, and improvement with DMARD treatment2. Non-invasive assays for high-sensitivity troponin and NT-pro-BNP have been evaluated in RA cohorts. No study has investigated these in early, treatment-naïve RA and in association with subclinical CMR findings.ObjectivesIn a treatment-naïve, early RA trial cohort, we examined serum cardiac markers of ventricular stretch (NT-pro-BNP) and myocardial injury (high-sensitivity troponin-I (hs-Tn-I)) and change with treatment, and whether these associated with cardiac MRI abnormalities.MethodsIn CADERA (a unique early RA RCT derived study), out of a total of 120 patients randomised to either first-line TNFi+MTX (n=60) or MTX and conventional synthetic DMARDs (n=60), 81 subjects with no previous history of CVD or diabetes and with only a maximum of 1 other traditional risk factor, underwent baseline and year 1 CMR. Corresponding levels of NT-pro-BNP (pg/ml) and hs-Tn-I (ng/L) were measured at baseline and year 1. CADERA reported CMR measures of vascular stiffness [aortic stiffness index (AoSI), and aortic distensibility (AD)]; tissue oedema/fibrosis [native T1 values (T1), myocardial extracellular volume (ECV), and late gadolinium enhancement (LGE)] and ventricular function 2. Bayesian linear and logistic regression models with mixed effects for overall data were used to test association between serum cardiac markers and CMR parameters.ResultsCADERA trial already reported baseline demographics and comorbidity data of patients. The most common risk factor was smoking history (55%) followed by hypertension (7%) and family history of ischaemic heart disease (5%)2. 7 patients had abnormal findings on LGE imaging at baseline and 4 at year 1.2From baseline to year 1, NT-pro-BNP remained within normal limits although a numerical improvement was noted at year 1(76.4 to 59.4, p = 0.038). Hs-Tn-I was within normal limits (At baseline, median hs-Tn-I was 2 (IQR 1.4 - 3.05) in patients with abnormal LGE findings compared to 1.60 (0.95 – 2.52) in those without (p = 0.3). At year 1, median hs-Tn-I was 2.70 (IQR 1.08 – 4.82) in patients with abnormal LGE findings versus 1.7 (IQR 1 – 2.5) in those without (p = 0.5). Table 1 outlines the association of serum cardiac markers with CMR parameters. At baseline, hs-Tn-I positively associated with LGE. This association was also demonstrated when considering overall data (baseline + week 48). Association of NT-pro-BNP with CMR was not significant.Table 1.Association of serum Hs-Tn-I with CMR parameters (posterior estimates with 95% credible intervals reported)CMR parameterHs-Tn-IOverallBaselineYear 1Measures of myocardial tissue oedema/fibrosisLGE0.56(0.1 to 1.11)0.6(0.04 to 1.27)0.54(-0.13 to 1.17)T1-4.42(-10.5 to 1.5)-0.38(-6.7 to 6.2)-6.02(-14.9 to 3.15)ECV0.002(-0.04 to 0.080.001(-0.05 to 0.07)0.001(-0.04 to 0.04)Measures of vascular stiffnessAoSI0.01(-0.23 to 0.25)0.05(-0.22 to 0.33)0.05(-0.33 to 0.42)AD-0.05(-0.23 to 0.13)-0.02(-0.18 to 0.14)-0.06(-0.37 to 0.25)ConclusionOur data suggest hs-Tn-I is within normal limits in a treatment-naïve, RA cohort but with numerical differences that associated with presence of focal fibrosis on CMR. Modest numbers of those with focal fibrosis may account for normal range values for hs-Tn-I.References[1]Peters MJL, Van Halm VP, Voskuyl AE, et al. Does rheumatoid arthritis equal diabetes mellitus as an independent risk factor for cardiovascular disease? A prospective study. Arthritis Care Res. 2009;61(11).[2]Plein S, Erhayiem B, Fent G, et al. Using cardiovascular magnetic resonance to define mechanisms of comorbidity and to measure the effect of biological therapy: the CADERA observational study. Effic Mech Eval. 2021;8(4):1–42.Disclosure of InterestsNone declared

The presence of a disease continuum in inflammatory arthritis (IA) is a recognised concept, with distinct stages from at-risk stage (presence of anti citrullinated-peptide autoantibody) to diagnosis of rheumatoid arthritis (RA), including therapy-induced remission. Despite T-cell dysregulation being a key feature of RA, there are few reports of T-cell phenotyping along the IA-continuum. We investigated the disturbances of naïve, regulatory and inflammation related cell (IRC) CD4+ T-cell subsets in 705 individuals across the IA-continuum, developing a simple risk-score (summing presence/absence of a risk-associated with a subset) to predict progression from one stage to the next. In 158 at-risk individuals, the 3 subsets had individual association with progression to IA and the risk-score was highly predictive (p

Objective The diagnosis of RA patients remains a challenge, especially in ACPA-negative disease. Novel T-cell subsets, particularly Th17 may be useful, although data on Th17 frequency using flow cytometry in RA are conflicting. We investigated whether a novel epigenetic qPCR assay for the quantification of Th17 could differentiate patients with RA from those with symptoms evolving towards an alternative diagnosis. Methods We used a qPCR assay measuring the extent of the methylation at a key position in the IL-17 and CD4 genes. Assays were performed on whole blood from 49 healthy controls (HC) and 165 early arthritis clinic patients. Flow cytometry was further used to detect the expression of CXCR4 on Th17 cells. Results In 75 inflammatory arthritis patients who progressed to RA, the qPCR assays showed significantly fewer Th17 cells compared with 90 patients who did not (P Conclusion The results of the epigenetic qPCR assay showed that low levels of Th17 cells were predictive of developing RA, particularly in the ACPA-negative patients. This could have value for insights into pathogenesis and management. The results suggest the recruitment of Th17 to the inflammatory disease site, consistent with high CXCR4 expression.

Background: Rituximab (RTX) is used for resistant SLE but clinical response varies. We previously validated two interferon-stimulated gene expression scores (IFN-Score-A and IFN-Score-B) that improved prediction of clinical outcomes in SLE. IFN-Score-A included most commonly reported ISGs and predicted flares and glucocorticoid requirements. IFN-Score-B included ISGs that respond to multiple IFN subtypes and predicted development of SLE in At-Risk individuals. Diagnosis of SLE was associated with both scores, while only IFN-Score-B was elevated in RA. The British Society for Rheumatology Biologics Registry (BILAG-BR) collects samples for RTX-treated patients in the UK. MASTERPLANS is a consortium to identify predictors of drug response. Objectives: To investigate whether IFN-Score-A and IFN-Score-B predict BILAG response to RTX at 6 months. Methods: This is a preliminary analysis of the first RTX-treated patients in the BILAG-BR with complete data. Patients were recruited if they were starting a first cycle of RTX for active SLE (BILAG A or 2xBILAG B) despite previous cyclophosphamide or mycophenolate mofetil. Disease activity was measured using BILAG-2004. Clinical response was defined as improvement by >=1 grade in active BILAG-2004 systems with no worsening in other systems. Whole blood was collected into TEMPUS tubes and RNA extracted. IFN-Scores were measured using a custom Taqman array as previously described [El Sherbiny et al., 2018]. Multivariate logistic regression was used to test IFN-Scores and baseline clinical covariates as predictors of BILAG response at 6 months. Results: Samples were available from 147 patients, of whom 84 had complete baseline and 6 month clinical data available and were included in this analysis. 40/84 (47.6%) patients had BILAG response at 6 months. In univariate and multivariate analysis, high IFN-Score-B expression was significantly associated with clinical response (see table 1). Conclusion: This preliminary analysis suggests that assessment of IFN activity has a role in predicting response to RTX. A novel IFN score (Score B) was more predictive than classic ISGs (Score A). These results add to a body of work showing that IFN-Score-B predicts clinically significant outcomes independently of overall IFN activity. Future work will analyse this biomarker in a larger cohort of patients and integrate with other putative clinical and biological predictors of response. Reference: [1] El-Sherbiny, Y. M.… E. M. Vital (2018). Sci. Rep. 8: 5793. Acknowledgement: We would like to thank the Medical Research Council, National Institute of Health Research, UK for funding the MASTERPLANS project. Disclosure of Interests: Adewonuola Alase: None declared, Zoe Wigston: None declared, Agata Burska: None declared, Elizabeth Hensor: None declared, Md Yuzaiful Md Yusof: None declared, John Reynolds: None declared, The Masterplans Consortium: None declared, Miriam Wittmann Consultant for: consultancy honoraria from Abbvie, Celgene, Janssen, L’Oreal, Novartis and Pfizer, Ian N. Bruce Grant/research support from: Genzyme Sanofi, GlaxoSmithKline, Consultant for: AstraZeneca, Eli Lilly, GlaxoSmithKline, ILTOO Pharma, MedImmune, Merck Serono, Speakers bureau: GlaxoSmithKline, UCB Pharma, Edward Vital Grant/research support from: He has received honoraria and research grant support from Roche, GSK and AstraZeneca.

Background Rituximab is used for resistant SLE but clinical response varies. Although biomarkers of time to relapse have been validated, there are few biomarkers to predict initial response. Interferon status may predict response to rituximab and anti-TNF in RA. We previously validated two interferon-stimulated gene expression scores (IFN-Score-A and IFN-Score-B) that improved prediction of clinical outcomes in SLE. IFN-Score-A included most commonly reported ISGs and predicted flares and glucocorticoid requirements. IFN-Score-B included ISGs that respond to multiple IFN subtypes and predicted development of SLE in At-Risk individuals. Diagnosis of SLE was associated with both scores, while only IFN-Score-B was elevated in RA. The British Society for Rheumatology Biologics Registry (BILAG-BR) collects data and samples for rituximab-treated patients in the UK. MASTERPLANS is an MRC-funded consortium to identify predictors of response. Methods This is a preliminary analysis of the first rituximab-treated patients in the BILAG-BR with complete data. Patients were recruited if they were starting a first cycle of rituximab for active SLE (BILAG A or 2xBILAG B) despite previous cyclophosphamide or mycophenolate mofetil. Disease activity was measured using BILAG-2004. Clinical response was defined as improvement by ≥1 grade in active BILAG-2004 systems with no worsening in other systems. Whole blood was collected into TEMPUS tubes and RNA extracted. IFN-Scores were measured using a custom Taqman array as previously described, normalised to PP1A [El Sherbiny et al. Sci Rep 2018]. Multivariate logistic regression was used to test IFN-Scores and baseline clinical covariates as predictors of BILAG response at 6 months. Results Samples were available from 147 patients, of whom 84 had complete baseline and 6 month clinical data available and were included in this analysis. 40/84 (47.6%) patients had BILAG response at 6 months. In univariate and multivariate analysis, high IFN-Score-B expression was significantly associated with clinical response (see table 1). Conclusions This preliminary analysis suggests that assessment of IFN activity has a role in prediction of response to rituximab. A novel IFN score (Score B) was more predictive than classic ISGs (Score A). These results add to a body of work showing that IFN-Score-B predicts clinically significant outcomes independently of overall IFN activity. Future work will analyse this biomarker in a larger cohort of patients and integrate with other putative clinical and biological predictors of response. Funding Source(s): Medical Research Council, National Institute of Health Research

Background:SLE is heterogeneous in clinical presentation, underlying immunology, and response to therapy. Patients with severe or resistant disease receive cyclophosphamide or rituximab. Although effective, late use of these therapies entails cumulative damage.Emerging predictors of response to rituximab include demographics, IFN-Scores and autoantibodies. For such predictors to change treatment strategy we need to understand their association with the effectiveness of other therapeutic options at earlier decision points.Prospective studies of treatment pathways are difficult to perform, requiring large populations followed for many years. We therefore investigated this question by analysing retrospective treatment pathways since diagnosis in patients whose biomarker status is subsequently known.Objectives:To use interferon scores and other markers to;[1]identify SLE patients more likely to need escalation of therapy to either rituximab or cyclophosphamide, and[2]determine if this occurred sooner than in others.Methods:In patients with established SLE, attending Leeds SLE clinic, we collected demographics, clinical characteristics and blood at a single timepoint. Two previously validated IFN-Scores[1,2] (IFN-Score A and IFN-Score B) were measured using Taqman. Autoantibodies were measured using immunoprecipitation. Treatment history covering all antimalarials, oral and intravenous immunosuppressants (agent, start and stop dates and reasons) since diagnosis was collected. Univariable and multivariable Cox proportional hazards models were used to test variables as predictors of the primary endpoint: time to cyclophosphamide/rituximab.Results:124 patients were included. Follow up since diagnosis was Median (IQR) 37.5 (27.4-52.3) years, total therapies per patient was 3 (1-4), therapies per year was 0.2 (0.1-0.3). 54/124 (44%) patients required cyclophosphamide/rituximab.Results of Cox regression are shown in the table 1. There was a significant association between high expression of IFN-Score B and requirement for cyclophosphamide/rituximab as well as trends for age in the first quartile (Table 1.CharacteristicCYC/RTX=Yes n(%)CYC/RTX =No n(%)Univariable Hazard ratio (95% CI)Univariable p valueMultivariable Hazard Ratio (95% CI)Multivariable p valueAge-Q1: under 2822 (71%)9 (29%)2.25 (0.99,5.12)0.0541.74 (0.95,3.18)0.074-Q2: 28 – 3812 (39%)19 (61%)1.03 (0.41,2.56)0.949Not includedNot included-Q3: 38 – 5212 (39%)19 (61%)1.52 (0.62,3.72)0.360Not includedNot included-Q4: over 528 (26%)23 (74%)referencereferenceNot includedNot includedEthnicity-Caucasian42 (48%)46 (52%)1.17 (0.58,2.35)0.660Not includedNot included-AfroCaribbean, South Asian, East Asian, Other10 (36%)18 (64%)referencereferenceIFN Score A, mean (SD)−2.59 (−2.67)−3.84 (−2.71)1.10 (0.98,1.24)0.113Not includedNot includedIFN Score B, mean (SD)−2.61 (−1.16)−3.53 (−1.76)1.31 (1.05,1.64)0.0171.23 (0.98,1.55)0.079Anti-Sm/U1RNP Ab positive17 (61%)11 (39%)1.50 (0.84,2.67)0.171Not includedNot includedAnti-Ro(60) Ab positive14 (45%)17 (55%)0.88 (0.47,1.65)0.686Not includedNot includedConclusion:Patients developing SLE before age 28 with raised interferon scores may be forecast to develop severe disease, fail oral therapies, and require rituximab. While this retrospective analysis may be confounded by a survivorship bias, our findings are consistent with other literature on these variables as predictors of severe disease. Future work will analyse more biomarkers in a larger and more varied patient population.References:[1]Md Yusof et al. Ann Rheum Dis 2018[2]El-Sherbiny et al. Sci Rep 2018Disclosure of Interests:Sabih-Ul Hassan: None declared, Khaled Mahmoud: None declared, Zoe Wigston: None declared, Agata Burska: None declared, Md Yuzaiful Md Yusof: None declared, Edward Vital Grant/research support from: Honoraria and research grant support from Roche, GSK and AstraZeneca.

Background:IFN Scores are increasingly important as biomarkers to stratify RMDs and select patients for interferon targeted therapies. Several studies have previously demonstrated that different sets of interferon stimulated genes (ISGs) have different clinical associations. We validated a 2-score system of ISGs expression consisting of Score A (a set of ISGs included in most interferon signature assays) and Score B (a set of less commonly measured ISGs). Score B consistently demonstrated stronger clinical associations in studies of diagnosis, prediction of progression to SLE in At-Risk individuals, response to rituximab and imaging-proven synovitis. Previous literature suggested that subsets of ISGs were explained by different interferon subtypes.Objectives:To understand the determinants of IFN Score A and IFN Score B.Methods:PBMCs and whole blood Tempus from 4 different healthy doners were stimulated with various cytokines and IFNs at 6- and 24-hour time points. RNA from both sample types were extracted and the expression of 26 interferon stimulated genes were measured using TaqMan and normalised to house-keeping gene PPIA. IFN Scores A and B were calculated as previously described [1]. To compare relative increase in expression with each condition, we calculated delta Ct fold change relative to non-stimulated. To represent greater expression as numerically higher positive values, relative expression was calculated as (fold change*-100) +100. Independent T-test calculated the significant differences between each condition compared to IFN α stimulation.Results:Table 1 shows differences between each condition compared with IFN α at the 6-hour time point. In both sample types, all conditions excluding IFN β stimulation were shown to induce significantly lower expression of both scores compared to IFN α (p Table 1.ConditionIFN Score AIFN Score BMean relative expression vs. non-stimulatedP value(vs IFN α)Mean relative expression vs. non-stimulatedP value(vs IFN α)PBMCIFN α103.4-71.8-IFN β97.10.37364.70.352IFN γ37.529.50.002IFN κ16.63.1IFN λ24.811.1IL-19.74.4IL-64.54.1IL-10-15.28.0TNF α9.61.2Whole BloodIFN α116.7-74.1-IFN β116.60.99773.20.894IFN κ23.00.0017.90.003IFN λ23.00.0017.6Conclusion:We found a difference in the induction of ISG expression between subtypes of Type I interferon, as well as other interferons and cytokines. However, the relative expression of IFN Score A and IFN Score B is not easily explained by the subtypes of interferon. We have also previously shown that these scores are both similarly expressed comparing different cell subsets. The explanation for the coordinated expression of these ISGs is therefore unclear and future work will explore the scores with a combination of conditions.References:[1]El-Sherbiny, Y.M., et al. Scientific Reports, 2018. 8(1): p. 5793.Disclosure of Interests:Zoe Wigston: None declared, Agata Burska: None declared, Miriam Wittmann Consultant of: Abbvie, Celgene, Janssen, L’Oreal, Novartis and Pfizer, Edward Vital Speakers bureau: Becton Dickinson and GSK, Consultant of: AstraZeneca, GSK, Roche/Genentech, and Sandoz, Grant/research support from: AstraZeneca, Roche/Genentech

Background:The interferon (IFN) pathway is a complex system with multiple proteins and diverse downstream effects on gene and protein expression. IFNs have been implicated in multiple RMDs. Despite significant potential, IFN assays have not progressed into clinical practice.Objectives:To perform a SLR on IFN assays in RMDs and propose a consensus terminology.Methods:OvidMedline, Embase and Web of Science were searched for reports of IFN and RMDs up to October 2019. Information about the properties of assays measuring type I IFN and measures of truth were extracted and summarised. Terminology was agreed through an interactive consensus process with reference to the existing evidence.Results:10037 abstracts were identified. 275 fulfilled eligibility criteria, and were used for data extraction. Some used more than one technique to measure IFN-I pathway activation. Hence, 275 papers generated data on 393 methods. There was great heterogeneity in the methods used and presentation of results. IFN-I pathway activation was measured using: qPCR (n=121), immunoassays (n=101), microarray (n=69), reporter cell assay (n=38), DNA methylation (n=14), flow cytometry (n=14), cytopathic effect assay (n=11), RNA sequencing (n=9), Plaque reduction assay (n=8), Nanostring (n=5), bisulphite sequencing (n=3). All papers fulfilled Face Validity. Due to lack of gold standard for IFN-I pathway activation, evidence of criterion validity was variable. Concurrent validity was presented for n=150 assays. The terminology used to describe aspects of type I IFN pathway activation was not consistent, so a consensus terminology for IFN research (Table 1) was proposed by the taskforce.Table 1.Consensus terminologyTermAbbreviationDefinitionInterferonIFNProteins with anti-viral activity; IFNs are mediators of an anti-viral response. They belong to the Type I, Type II and Type III IFN families.Type I interferonIFN-IThe IFNs alpha, beta, omega, kappa, epsilon, secreted by any nucleated cell, and binding to the IFNAR, which is expressed on any nucleated cell.Type II interferonIFN-IIIFN gamma, mostly secreted by T cells, binding to the IFNGR, which is expressed on most leucocytes.Type III interferonIFN-IIIIFN lambda, which are structurally more similar to IL-10 but share downstream signalling and gene expression with IFN-I.Interferon-stimulated genesISGsGenes whose expression is known to be upregulated by any kind of IFN. Individual ISGs may not exclusively represent Type I IFN pathway activation.Type I Interferon pathway activationAny evidence for function of the components of the Type I IFN pathway. This includes: secretion of a Type I IFN protein, binding to the IFNAR, initiation of JAK/STAT signalling pathways, expression of IFN-stimulated genes, expression of IFN-stimulated proteins.Type I interferon pathway assayAn assay measuring one or more components of the Type I IFN pathway at a molecular or functional level.Interferon stimulated gene expression signatureA qualitative description of coordinated expression of a set of ISGs that is indicative of Type I IFN pathway activation.Interferon stimulated gene expression scoreA quantitative variable derived from expression of a defined set of ISGs that is indicative of Type I IFN pathway activation.Interferon stimulated protein scoreA variable derived from expression of a defined set of soluble biomarkers known to be upregulated by IFN, although not specific for Type I IFN.InterferonopathyMonogenic diseases in which there is constitutive Type I IFN pathway activation with a causal role in pathology. The clinical picture may resemble rheumatic musculoskeletal diseases. However, most diseases with IFN pathway activation are not Interferonopathies.Conclusion:Diverse methods have been reported as IFN assays and these differ in what elements of type IFN-I pathway activation they measure. The taskforce consensus terminology on type I IFN reporting should be considered for research and clinical applications.Disclosure of Interests:Agata Burska: None declared, Javier Rodriguez Carrio: None declared, Philip G Conaghan: None declared, Willem A Dik: None declared, Robert Biesen: None declared, Maija-leena Eloranta: None declared, Giulio Cavalli: None declared, Marianne Visser: None declared, Dimitrios Boumpas: None declared, George Bertsias: None declared, Marie Wahren-Herlenius: None declared, Jan Rehwinkel: None declared, Marie-Louise Frémond: None declared, Mary K. Crow Consultant of: AstraZeneca, Bristol Meyers Squibb, Lilly, Shannon Pharmaceuticals, Grant/research support from: Gilead, Lars Ronnblom Consultant of: AstraZeneca, Edward Vital Speakers bureau: GSK, Consultant of: AURINIA, SANDOZ, GSK, AstraZeneca, Roche, Modus, Grant/research support from: AstraZeneca, Marjan Versnel: None declared

Background:Study of pre-clinical autoimmunity can reveal immune changes that initiate or prevent clinical disease. We previously reported an At Risk cohort of ANA positive individuals referred to secondary care, of whom 16% progressed to SLE or Sjogren’s syndrome during follow up[1]. Recently, we also reported that even ANA positive individuals who do not develop clinical autoimmunity, have defective pDC function and non-haematopoietic interferon-production[2].Objectives:To describe the immunophenotype of ANA-positive progressors and non-progressors compared to healthy controls.Methods:We conducted a prospective observational longitudinal study of At-Risk individuals in Leeds (n=150) over 3 years with annual follow up. Those who met classification criteria for autoimmune connective tissue disease (AI-CTD) during follow up were termed progressors and others; non-progressors. Among the non-progressors, those with no clinical diagnostic criteria at any time point were called absolute non-progressors (ANP).At baseline we analysed two IFN-stimulated gene expression scores previously described[3] and flow cytometry performed for major circulating subsets. Association between progression group and biomarkers were tested using independent samples t-test.Results:3 year follow up data were available in 148/150 patients. Outcomes were: progressors: 30/148 (20%) [SLE=25; pSS=5] and non-progressors: 118/148 (80%). Within the non-progressors, 69/148 (47%) were ANP.A detailed analysis of baseline biomarkers is included in Table 1. Despite not developing clinically significant disease, non-progressors had significant alterations in several biomarkers. This was also evident in the ANP sub-group.Table 1.BiomarkerHealthy ControlsMean(95% CI)Non-progressorsMean(95% CI)Difference from healthy controlspAbsolute non-progressorsMean(95% CI)Difference from healthy controls pProgressorsMean (95% CI)Difference from healthy controls pIFN score A-6.27 (-6.88, -5.65)-5.22 (-5.64, -4.80)0.016-5.11 (-5.91, -4.32)0.120-4.14 (-5.11, -3.18)0.001IFN score B-5.43 (-6.00, -4.85)-5.30 (-5.62, -4.98)0.193-4.96 (-5.45, -4.47)0.949-4.27 (-4.90, -3.65)0.136% Mono-cytes14.80 (10.50, 19.11)15.46 (12.96, 17.97)0.18714.94 (11.42, 18.47)0.73810.36 (7.42, 13.30)0.227% CD4 T cells62.90 (57.75, 68.05)64.89 (60.86, 68.92)0.00168.72 (62.97, 74.47)0.26760.93 (53.64, 68.21)0.001% CD8 T cells32.18 (27.47, 36.90)28.60 (25.12, 32.09)0.13525.02 (20.28, 29.77)0.04431.51 (25.63, 37.40)0.875% Naïve B cells65.59 (56.56, 74.61)73.74 (70.02, 77.46)0.00176.86 (72.21, 81.51)0.12868.86 60.62, 75.84)0.173% Memory B cells29.56 (20.42, 38.72)19.72 (15.84, 23.61)23.13 (16.65, 29.61)0.05425.57 (17.18, 33.96)0.163% Plasma-blasts1.12 (0.81, 1.43)0.91 (0.58, 1.24)0.9641.01 (0.44, 1.58)0.4570.81 (0.53, 1.10)0.112% NK cells13.65 (9.65, 17.65)12.30 (8.90, 15.71)0.48012.35 (6.65, 18.05)0.78111.90 (8.65, 15.14)0.195Conclusion:Clinically benign ANA positivity is a complex immune state with many features seen in active SLE, including pDC exhaustion, keratinocyte interferon production and strong skin interferon score, increased blood interferon score, and disturbance of B cell subsets. These findings suggest that other factors are necessary for clinical disease, or that regulatory mechanisms stabilise autoimmunity. Further work in this cohort will address these questions.References:[1]Md Yusof et al. Ann Rheum Dis 2018[2]El-Sherbiny et al. Sci Rep 2018[3]Psarras et al. Nat Commun. 2020Disclosure of Interests:Sabih-Ul Hassan: None declared, Zoe Wigston: None declared, Agata Burska: None declared, Md Yuzaiful Md Yusof: None declared, Edward Vital Grant/research support from: Honoraria and research grant support from Roche, GSK and AstraZeneca.

Background:Interferon (IFN) is known to play a role in the pathogenesis of many autoimmune diseases, among them, rheumatoid arthritis (RA). A study showed that two interferon-stimulated gene expression scores (IFN-Score-A and IFN-Score-B) can be used to predict progression to connective tissue disease (CTD) in at-risk individuals, positive for anti-nuclear antibodies (ANA+) [1]. This validated score could potentially be applied to individuals at-risk of RA.Objectives:To investigate the role of type I IFN in patients at-risk of RA and assess its potential role as a biomarker to predict progression to inflammatory arthritis (IA).Methods:PBMC samples were taken from 36 at-risk individuals positive for anti-citrullinated protein antibodies (ACPA+), with a non-specific musculoskeletal complaint but no clinical synovitis and a normal ultrasound scan at baseline (BL). 17 of them developed IA later on, and had a second sample taken at the moment of progression. The other 19 did not progress and had a second sample taken after 1 year. Expression of IFN stimulated genes was assessed using TaqMan. T-test was used to compare the expression of the genes at the two time points of each individual. Binary logistic regression was used to assess BL predictors of progression. Multivariable analysis was adjusted for confounders such as C-reactive protein (CRP) rheumatoid factor (RF) and ACPA titre.Results:Table 1 shows the list of tested genes. There were no differences in the gene expression of progressors at BL and at the moment of IA diagnosis. Similarly, no differences were found between the non-progressors at BL and after 1 year. Comparing BL samples from progressors and non-progressors, there was a trend to higher expression of IFN-score-B in the progressors (Mann Whitney p=0.055), whereas IFN-score-A was similarly expressed in both groups. This is similar to the results found in a cohort of ANA+ individuals at risk of progressing to CTD [1]. Multivariable analysis showed that IFN-score-B and RF titre were predictive for IA progression with an odds ratio (OR) of 2.18 (p=0.048) and OR 1.01 (p=0.035).Table 1.Interferon stimulated genes and scores A and BGENEIFN scoreGENEIFN scoreCCL8BST2CEACAM1IFI16CXCL10IFIH1GBP1LAMP3IFI27NT5C3BIFI44PHF11IFI44LASERPING1BIFIT1SOCS1IRF7SP100ISG15STAT1RSAD2TAP1XAF1TRIM38UBE2L6UNC93B1Table 2.Baseline characteristics and predictors of progressionConclusion:These exploratory results suggest that IFN changes in IA progressors precede subclinical inflammation on ultrasound. A preliminary risk model shows that IFN-score-B could be useful to predict progression to IA in at-risk of RA individuals; however, these results require validation in a larger at-risk cohort.References:[1]Md Yusof MY. Ann. Rheum. Dis. 2018;77(10):1432-9.Disclosure of Interests:Leticia Garcia-Montoya: None declared, Zoe Wigston: None declared, Agata Burska: None declared, Kulveer Mankia: None declared, Edward Vital Grant/research support from: AstraZeneca, Roche/Genentech, and Sandoz, Consultant of: AstraZeneca, GSK, Roche/Genentech, and Sandoz, Speakers bureau: Becton Dickinson and GSK, Paul Emery Grant/research support from: AbbVie, Bristol-Myers Squibb, Merck Sharp & Dohme, Pfizer, Roche (all paid to employer), Consultant of: AbbVie (consultant, clinical trials, advisor), Bristol-Myers Squibb (consultant, clinical trials, advisor), Lilly (clinical trials, advisor), Merck Sharp & Dohme (consultant, clinical trials, advisor), Novartis (consultant, clinical trials, advisor), Pfizer (consultant, clinical trials, advisor), Roche (consultant, clinical trials, advisor), Samsung (clinical trials, advisor), Sandoz (clinical trials, advisor), UCB (consultant, clinical trials, advisor)

Background:Interferon (IFN) signature is associated with disease activity and flare in SLE. We previously described two independent IFN gene expression scores; IFN Score A (the most commonly measured ISGs) and IFN Score B (less commonly measured ISGs which may also respond to IFN-II or other immune mediators)[1]. Many more clinical outcomes are associated with IFN Score B than with a “classic” interferon signature. These include progression of At-Risk individuals to SLE, response to rituximab, and differentiation of IFN signature in RA and SLE.In previous work, the relationship of IFN Signatures with arthritis was less clear than for other SLE features. This may be related to the local regulatory effects of IFN-beta in the synovium, contrasting with the pro-inflammatory effects of other interferons. Another reason may be the proven imprecision of clinical examination as a measure of MSK inflammation in SLE.USEFUL was a multicentre longitudinal study including serial ultrasound assessment of SLE patients with inflammatory MSK pain receiving treatment with glucocorticoids (GC).Objectives:To determine whether IFN scores A and B are associated with imaging-proven synovitis in SLE and measure the responsiveness of IFN scores to GC treatment.Methods:133 SLE patients were recruited into the USEFUL study if the referring physician deemed they had inflammatory pain warranting treatment. Participants received depomedrone 120mg IM then were assessed at 0, 2 and 6 weeks using clinical instruments and ultrasound (US). OMERACT US criteria were used to categorise patients as active (GS2 or PD1 in at least one joint or tendon), active in both joints and tendons, or non-active (no GS1 and PD0 or better in all joints).Expression of 26 interferon stimulated genes, normalised to PP1A was measured in whole blood collected in TEMPUS tubes using a custom Taqman array. IFN scores A and B were calculated as previously described[1]. Missing data was imputed using expectation-maximisation method. Parametric tests were applied with post hoc Tukey to compare scores between groups.Results:At baseline, there was no significant difference in IFN Score A between ultrasound groups (F = 1.045, p = 0.355). In contrast, IFN Score B differed significantly between ultrasound groups (F = 4.168, p = 0.018). The greatest difference was between active ultrasound for both joints and tendons (n=22) and non-active ultrasound (n=53) (difference = 0.75, 95% CI 0.13, 1.37, p=0.013).There was no significant change from baseline in IFN Score A at week 2 (mean difference 0.08, 95% -0.14, 0.31, p = 0.45) or week 6 (mean difference -0.03, 95% -0.25, 0.19, p = 0.79). Similarly, there was no significant change in IFN Score B at week 2 (mean difference -0.01, 95% -0.18, 0.17, p = 0.93) or week 6 (mean difference -0.07, 95% -0.21, 0.08, p = 0.36).Conclusion:Previous studies were unable to demonstrate an association between a typical interferon signature and arthritis in SLE. Our study includes a homogenous patient population and therapy, objective measure of synovitis, and a more detailed assessment of IFN Status. We found that imaging-proven synovitis is associated with increased expression of a specific subset of ISGs (IFN score B), but not a the more typical interferon signature genes (IFN Score A).This increases the body of evidence for the value of IFN score B in predicting clinical outcomes. GC treatment did not affect systemic IFN signature scores at follow up. Future analysis will explore the role of IFN Scores in predicting clinical responses to therapy in this study.References:[1]El-Sherbiny, Y.M., et al. Scientific Reports, 2018.8(1): p. 5793.Disclosure of Interests:Zoe Wigston: None declared, Agata Burska: None declared, Adewonuola Alase: None declared, Khaled Mahmoud: None declared, Edward Vital Grant/research support from: AstraZeneca, Roche/Genentech, and Sandoz, Consultant of: AstraZeneca, GSK, Roche/Genentech, and Sandoz, Speakers bureau: Becton Dickinson and GSK

Objectives In a cross-sectional study, we evaluated the prevalence of ‘multi-dimensional remission’ (MDR) and its component parameters, assessed using objective measures in a cohort of RA patients in treatment-induced DAS28-remission, and their relationship with patient-reported outcome measures. We sought to confirm the feasibility and face validity of the MDR construct, providing a platform for future longitudinal studies in which its clinical utility might be further established. Methods 605 patients were selected from an inflammatory arthritis register using DAS28(CRP) Results Overall, only 53% (321/605) of the patients achieved clinical parameters, failures being mainly due to raised CRP (52%), TJC (28)>1 (37%) or SJC (28)>1 (16%). 211/364 (58%) of patients achieved ultrasound remission and 193/297 (65%) patients showed T-cell remission. Complete data were available for 231 patients. MDR was observed in only 35% and was associated with the best (lower) PRO scores (all P ⩽ 0.05 vs non-MDR) when compared with the other definitions of remission assessed. The MDR rate was similar in early and established RA patients on b-DMARDs; however, it was lower in established RA patients who received multiple cs-DMARDs (P = 0.011). Conclusions In this study, MDR, which may represent a state closer to normality, was found to occur in about a third of DAS28-remission patients and was associated with better patient-reported outcome measures. MDR could be a novel optimal treatment target, notably from a patient’s perspective. The relevance of these findings needs further assessment.

OBJECTIVES: To screen for significant arrhythmias with an implantable loop recorder (ILR) in patients with SSc and no known cardiovascular disease, and identify associated disease phenotype, blood and cardiovascular magnetic resonance (CMR) biomarkers.METHODS: Twenty patients with SSc with no history of primary SSc heart disease, traditional cardiovascular disease, diabetes or maximum one traditional cardiovascular risk factor underwent clinical assessment, contrast-enhanced CMR and ILR insertion.RESULTS: ILR data were available for 19 patients: 63% female, mean (s.d.) age of 53 (12) years, 32% diffuse SSc. Eight patients had significant arrhythmias over 3 years: one complete heart block, two non-sustained ventricular tachycardia [all three dcSSc, two anti-topoisomerase antibodies (Scl70) positive, three interstitial lung disease and two previous digital ulceration] and five atrial arrhythmias of which four were with limited SSc. These required interventions with one permanent pacemaker implantation, four anti-arrhythmic pharmacotherapy, one anticoagulation.Patients with significant arrhythmia had higher baseline high-sensitivity troponin I and N-terminal pro-brain natriuretic peptide [mean difference (95% CI) 117 (-11, 245) and 92 (-30, 215) ng/l, respectively], and CMR-extracellular volume [mean (s.d.) 32 (2) vs 29 (4)%]. Late gadolinium enhancement was observed in five patients, only one with significant arrhythmia.CONCLUSION: This first ILR study identified potentially life-threatening arrhythmias in asymptomatic SSc patients attributable to a primary SSc heart disease. Disease phenotype, CMR-extracellular volume (indicating diffuse fibrosis) and cardiac biomarkers may identify at-risk patients that would benefit from ILR screening. Future studies can inform a risk model and provide insights into SSc-associated arrhythmia pathogenesis.

Objectives Anticitrullinated protein antibody (ACPA)+ individuals with non-specific musculoskeletal symptoms are at risk of inflammatory arthritis (IA). This study aims to demonstrate the predictive value of T cell subset quantification for progression towards IA and compare it with previously identified clinical predictors of progression. Methods 103 ACPA+ individuals without clinical synovitis were observed 3-monthly for 12 months and then as clinically indicated. The end point was the development of IA. Naive, regulatory T cells (Treg) and inflammation related cells (IRCs) were quantified by flow cytometry. Areas under the ROC curve (AUC) were calculated. Adjusted logistic regressions and Cox proportional hazards models for time to progression to IA were constructed. Results Compared with healthy controls (age adjusted where appropriate), ACPA+ individuals demonstrated reduced naive (22.1% of subjects) and Treg (35.8%) frequencies and elevated IRC (29.5%). Of the 103 subjects, 48(46.6%) progressed. Individually, T cell subsets were weakly predictive (AUC between 0.63 and 0.66), although the presence of 2 T cell abnormalities had high specificity. Three models were compared: model-1 used T cell subsets only, model-2 used previously published clinical parameters, model-3 combined clinical data and T cell data. Model-3 performed the best (AUC 0.79 (95% CI 0.70 to 0.89)) compared with model-1 (0.75 (0.65 to 0.86)) and particularly with model-2 (0.62 (0.54 to 0.76)) demonstrating the added value of T cell subsets. Time to progression differed significantly between high-risk, moderate-risk and low-risk groups from model-3 (p=0.001, median 15.4 months, 25.8 months and 63.4 months, respectively). Conclusions T cell subset dysregulation in ACPA+ individuals predates the onset of IA, predicts the risk and faster progression to IA, with added value over previously published clinical predictors of progression.

SummaryObjectivesImmune age-related abnormalities may synergise with osteoarthritis (OA) pathology. We explored whether abnormalities in the blood immune cell composition are present in OA, beyond defects typically associated with ageing.DesignBlood was collected from 121 healthy controls (HC) and 114 OA patients. Synovial biopsies were obtained from another 52 OA patients. Flow cytometry was used to establish the frequencies of lineage subsets, naïve, memory and regulatory T and B-cells, cells with an abnormal phenotype related to inflammation (IRC) and memory-like CD8+ T-cells. Multivariate analysis of covariance (MANCOVA) was used to determine whether the relative subset frequencies differed between HC and OA, controlling for age.ResultsExpected histology and T/B-cell infiltration were observed.Following age adjusted analysis, we confirmed the lack of age association in HC for CD4+, B, NK and NKT cells but a negative trend for CD8+ T-cells. In OA, CD4+ T-cell and B-cell frequency were lower compared to HC while CD8+ T-cell frequencies were higher. CD8+ memory-like cells were more likely to be found in OA (odds ratio = 15). Increased CD8+ IRC frequencies were also present in OA. The relationship between age and CD4+ or CD8+ naïve T-cells in HC were changed in OA while the age relationships with memory cells were lost. The increase in CD4+ Treg with age was also lost in OA. B-cells showed limited evidence of disturbance.ConclusionsImmune dysfunction may be present in OA beyond what appears related to ageing; this requires further investigation.

The therapeutic goal for patients with rheumatoid arthritis (RA) is clinical remission. This is best achieved by early diagnosis and appropriate therapeutic intervention. RA is associated with dysregulation of T-cell subsets (naïve, regulatory (Treg) and inflammation-related cells (IRC)) early in the disease. Our aim was to test the hypothesis that T-cell subset quantification can predict the achievement of clinical remission with early treatment in RA.T-cell subsets were quantified in 108 drug-naïve, early RA patients commencing methotrexate (MTX) or MTX+antitumor necrosis factor (anti-TNF) and in 105 healthy controls (HC). The primary outcome assessed was remission (DAS282.6). A pilot study used frozen cells (38 patients and 35 HCs, see online supplementary material) and was validated with fresh blood (70 patients and 70 HCs).Immune dysregulation in early RA was confirmed with an association between age and reduced naïve cells compared with HCs (p=0.006), a lower age-adjusted Treg and higher IRC frequency (p=0.001). Anticitrullinated peptide antibody (ACPA) positivity was associated with lower naïve (p=0.031) and Treg frequencies (p=0.039). In 50 patients treated with MTX, ACPA/age-adjusted analysis demonstrated that higher naïve cell frequency (relative to HC) was associated with remission (OR 5.90 (1.66 to 20.98), p=0.006, sensitivity/specificity 62%/79%, Positive Predictive Value (PPV)/Negative Predictive Value (NPV) 66%/76%). Remission with MTX+anti-TNF (n=20) was not found to be associated with naïve cell frequency, and for patients with reduced naïve cells the remission rate increased from 24% (MTX) to 42% (MTX+anti-TNF).Baseline T-cell subset analysis has a value in predicting early RA remission with first therapy with MTX. Immunological analysis could be used in conjunction with clinical/serological features to predict response to MTX and help select the most appropriate therapy at disease presentation.

OBJECTIVES: To determine the change in established biomarkers of cardiovascular (CV) risk, namely, total cholesterol/high-density lipoprotein cholesterol ratio (TC/HDL-C), N-terminal pro-brain natriuretic peptide (NT-proBNP) and insulin resistance (IR) in patients with early RA treated with two different treat-to-target strategies.METHODS: Fasting glucose, lipids, insulin and NT-proBNP were measured at baseline, weeks 26 and 78 in 79 DMARD-naïve RA patients, free of CV disease, as part of a double-blind randomized controlled trial of MTX with either infliximab (IFX) or methylprednisolone as induction therapy. Homeostasis model assessment-estimated IR (HOMA-IR) (glucose*insulin/405) was used to measure IR. Multiple imputation was employed, and linear regression analyses were adjusted for baseline values.RESULTS: Changes in DAS44-CRP did not differ between the treatment arms at weeks 26 and 78. Mean TC/HDL-C, HOMA-IR and NT-proBNP improved in both groups at weeks 26 and 78, although change in NT-proBNP was not statistically significant at week 78. Changes in TC/HDL-C and NT-proBNP were similar between treatment arms, but HOMA-IR values in the IFX + MTX arm were 42% lower than those treated with MTX + methylprednisolone at week 78 (P = 0.003); the difference remained significant after adjustment for baseline BMI, ACPA positivity, smoking status and intramuscular glucocorticoid use (P = 0.007).CONCLUSION: When implementing a treat-to-target approach, treatment of early RA was associated with improvement in TC/HDL-C, HOMA-IR and NT-proBNP, and a greater long-term improvement in HOMA-IR was seen in those treated with IFX.TRIAL REGISTRATION: EU Clinical Trials Register, http://www.clinicaltrialsregister.eu, Eudract-2005-005013-37; ISRTCNregisrty, http://www.isrctn.com, ISRCTN48638981.

Introduction Interleukin-7 (IL-7) is a cytokine essential for T-cell lymphopoiesis, survival and polarization with an emerging role in autoimmunity. We previously demonstrated reduced levels of circulating IL-7 in rheumatoid arthritis (RA), although high amounts are expressed in joints, suggesting differences between systemic and synovial effects. We observed healthy levels of IL-7 in 48% of RA patients in clinical remission (CR) and aimed to investigate the consequences of IL-7 deficiency on T-cell responses. Methods We used RA patients with active disease and in CR presenting various levels of IL-7, to investigate its modulatory effects on T cells by analysing responses to phyto-haemagglutinin (PHA), expression of polarization or survival factors, or suppression by regulatory T cells (Tregs). Results IL-7 levels were normal (>10 pg/ml) in 48% of RA patients in CR. Amongst 63 CR patients followed up for 18 months, lack of IL-7 recovery was observed in 13 out of 15 (86%) patients experiencing relapse but only 11 out of 48 (23%) of those who did not (P = 0.0002). Binary regressions showed high significance for below normal IL-7 levels for self-reported maternal family history of arthritis (odds ratio (OR): 7.66, P = 0.006) and a trend for smoking (OR: 3.33, P = 0.068) with no further demographic or clinical associations. Serum IL-7 correlated with restored CD4+T-cell response to PHA (rho = 0.879); this was not related to an increase in T-cell proliferation capacity or expression of survival factors B-cell lymphoma 2 (BCL2) and BCL2-associated protein X (BAX). Expression of Th1 polarization factor (TBET) was also dependent on exposure to IL-7 in vivo (rho = 0.600). In contrast CD25highTregs’ response to PHA was not affected by in vivo IL-7, but their suppression capabilities were related to circulating IL-7 (rho = 0.589). Co-stimulation with IL-7 (mimicking the joint environment) increased responsiveness of CD4+T-cells to PHA, lowering the ability of CD25highTregs to suppress them. Conclusions Our data demonstrate that IL-7 has a critical role in modulating T-cell function in vivo, possibly explaining opposing effects observed systemically and in the joint. Lack of IL-7 recovery in CR by maintaining a suppressed immune system may be a determinant factor in the occurrence of relapse. Electronic supplementary material The online version of this article (doi:10.1186/s13075-014-0511-3) contains supplementary material, which is available to authorized users.

RA is a complex disease that develops as a series of events often referred to as disease continuum. RA would benefit from novel biomarker development for diagnosis where new biomarkers are still needed (even if progresses have been made with the inclusion of ACPA into the ACR/EULAR 2010 diagnostic criteria) and for prognostic notably in at risk of evolution patients with autoantibody-positive arthralgia. Risk biomarkers for rapid evolution or cardiovascular complications are also highly desirable. Monitoring biomarkers would be useful in predicting relapse. Finally, predictive biomarkers for therapy outcome would allow tailoring therapy to the individual. Increasing numbers of cytokines have been involved in RA pathology. Many have the potential as biomarkers in RA especially as their clinical utility is already established in other diseases and could be easily transferable to rheumatology. We will review the current knowledge’s relation to cytokine used as biomarker in RA. However, given the complexity and heterogeneous nature of RA, it is unlikely that a single cytokine may provide sufficient discrimination; therefore multiple biomarker signatures may represent more realistic approach for the future of personalised medicine in RA.

Background Evidence suggests cardiovascular disease (CVD) risk in patients (pt, pts) with Inflammatory arthritis (IA), Rheumatoid arthritis (RA) is equivalent to that of type II diabetes mellitus 1 . The accepted guidance is lifestyle factor modifications should be addressed in IA 2 , but little is known of the impact of CVD risk reduction pt education programmes delivered to pts with IA within a Rheumatology service setting. Objectives To describe the utility of a comprehensive CVD risk reduction pt education programme in pts with newly diagnosed IA following a treat to target standard therapeutic regime. Methods 387 pts with IA (RA and Undifferentiated Inflammatory Arthritis: UIA) were assessed at baseline (BL) for CV risk factors (RFs) (see Table 1) Pts were; RA; pts (1987 or 2010 ACR criteria) (n=237) and UIA; pts not meeting ACR RA criteria (n=150). A standardized co-morbidity pt education programme, which included CVD risk reduction, was delivered throughout the first 12 months (M). The impact on modifiable CVD RFs was assessed at 12M. Results 108 RA, 34 UIA pts had complete datasets (see Table 1). Of those smoked at BL, 15% reported being ex-smokers at 12M and the number smoked had decreased in 44%. 10% who reported being ex-smokers at BL altered their status to never having smoked by 12M and their reported number previously smoked reduced by 43% at 12M. Alcohol use reduced in 39% and increased in 24% (overall decrease, p=0.033). At BL, 56% of pts reported doing no moderate exercise at all; at 12M around a third of the pts had either increased or decreased the amount of exercise (no overall change). Disease activity generally decreased (p Conclusions A significant reduction in alcohol units and improvement in smoking habits were seen. However, despite significant improvements in IA disease activity and mobility, there was no improvement in exercise from the low level reported at BL. The need to address modifiable CVD RFs in IA is widely accepted by Rheumatologists but effecting lifestyle change is challenging. Behavioural change in exercise is complex, particularly in chronic disease. Further studies are required to optimise the approach in IA. References Ann Rheum Dis 2009; 68:1395–1400 ARD 2010; 69:325–331 Disclosure of Interest None declared

Background An increased risk of heart failure in RA exists 1 . Aetiology is unclear with some evidence that cardiomyopathy may occur early in RA, from pathology distinct to atherosclerosis 3 . Large well characterised CMR-RA studies to investigate pathogenesis are few 2 . Objectives To evaluate sub-clinical CVD in an asymptomatic-CVD established RA cohort compared to healthy controls (HC), using reference standard CMR-measured outcomes & novel CMR-measures of carotid artery. Methods 76 ACR1987 RA patients (pts) with dis.>5yrs, no CVD/diabetes, assessed for CV traditional risk factor (TRFs)/RA profile, pulse wave velocity (PWV), non-contrast 3T CMR (heart & carotids) reported by CMR-cardiologists, compared to 26 HC. Carotid measures inc. mean (MWT) & maximum (MxWT) wall thickness, wall (WVol) & luminal (LVol) volume & WVol-index (WVol-I, WVol/(WVol+LVol)). UVA variables (var.)=TRFs, BMI, waist/hip ratio (WHR), HOMA-IR, NTproBNP, dis. duration (ddur), 3vDAS28, ACPA, HAQ-DI, joint surgery hx, biologic use & PWV. MVA var.=age/sex/known associated (ass.) var./UVA r>0.3. Results Mean (SD) age (yrs) of RA pts; 60 (9.2), 74% female; HC 52 (11.4), 54% female. Median (IQR) ddur 16.5 (10.7, 25.7)yrs, 81% ACPA+ve, DAS28CRP 2.59 (1.30, 3.33), 67% on biologic. There were no significant differences in lipids/glucose/HOMA-IR between RA/HC. Trend for higher NT-proBNP in RA seen. In MVA (inc. age/sex), increasing ddur ass. with NTproBNP (0.5% rise in NTproBNP per 1% increase in ddur, p0.034). RA pts (after adj. for age/sex/TRFs) had reduced LVmass/BSA, LVEF, native T1 values & increased strain (Mid-S9) vs HC; no difference seen for LVEDV, LVESV, SV, LVmass/EDV, torsion or arterial stiffness (PWV/distensibility). No significant differences in carotid measures between RA/HC seen. In RA pts, 10yr JBS2 CV risk score correlated with WVol (r0.377 p=0.004) & LVol (r0.441 p=0.001). TRFs ass. with cardiac & carotid measures in RA. Conclusions CMR evidence for reduced LV mass in RA, suggesting pathology other than atherosclerosis as the cause of heart failure; perhaps microvascular dysfunction. TRFs, not RA features, are important determinants of CMR measures of subclinical CVD in RA. References ArthRheum 2005;52:412–20 EULAR 2015 OP0163 A&R 2010:62v4 Disclosure of Interest None declared

Background and Objectives Ankylosing spondylitis (AS) is a debilitating chronic inflammatory condition characterised by sacroiliac/axial joint inflammation and peripheral/extra-articular manifestations in some cases. Anti-TNF treatment has revolutionised disease control, however, not all patients respond, creating a significant need for biomarkers to tailor biological treatments to appropriate patients. Interferon (IFN)-related genes (IRGs) and gene expression signatures have been associated with a number of inflammatory autoimmune arthritides, including AS. Therefore, the aims of this study were to explore a large panel of IRGs to select possible candidate for further work developing a signature. Materials and Methods 12 AS patients treated with infliximab (4 infusions of 6 mg/kg over 22 weeks) were selected from our tissue bank. Response was defined as a decrease in Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) of > 2 points or > 50% at 6 months post treatment. cDNA was obtained from baseline peripheral blood mononuclear cells and was analysed for the expression of 75 IRGs using custom TaqMan Microfluidics Cards. Differences in gene expression between response groups was then evaluated using Mann-Whitney U tests, box- plots and unsupervised hierarchical clustering. Results Overall 6 patients responded to treatment (R) and 6 did not (NR). Although the study was exploratory and not powered for statistical significance, between the two groups, a trend (P Conclusions In this pilot study, we demonstrate the potential for IRGs to be associated with response to anti-TNF treatment in AS. Furthermore, certain genes may be unique to AS differentiating it from other autoimmune inflammatory diseases. However, due to small sample numbers, statistical significance could not be reliably determined. Selected genes will now be validated in a larger cohort to determine reproducibility and statistical significance.

Introduction It has been recently established that the quantification of naive CD4 + T-cell frequency at baseline can predict the outcome of MTX treatment (1). The purpose of this study was to investigate whether T-cell subsets varied during treatment with MTX and MTX+anti-TNF. Methods Fresh blood samples were collected from 116 patients with early RA that had received no previous DMARDs: a first group received MTX (n = 67) and the second MTX+anti-TNF (n = 49). Flow cytometry was conducted at 3 monthly intervals using standard protocols. The frequency of naive and IRC subsets were recorded as% of CD4 + T-cells. Naive T-cells decrease with age but IRC are more closely related to levels of inflammation (2). The deviation from the expected age-corrected frequency was calculated for naive cells and used to analyse data. Results At baseline, there was no difference in the average frequency of naive CD4 + T-cells and IRC between treatment groups. The average frequency of naive CD4 + T-cells increased slowly from week 26 to week 48 in the MTX group plateaued at ˜2.8 fold. In contrast, in the MTX+anti-TNF group, naive cell frequency increases steadily from baseline up to 6.5 fold at 48 weeks. IRC marginally increased in the MTX group (1.25 fold at week 48) but were immediately reduced by half (0.46 fold) in the MTX+anti-TNF group and stayed low up to week 48 (p Conclusion Results suggest an association between better inflammation control resulting in lower IRC frequency in the MTX+anti-TNF group as well as a steady increase in naive cells frequency observed with more aggressive treatment. Considering the predictive value of naive cell frequency at baseline (1), continued naive cells deterioration in the MTX group is an important finding suggesting lack of immune function recovery compare to the MTX+anti-TNF. References F Ponchel, et al . ARD. 2014;73:2047-2053 F Ponchel, et al. Blood. 2002;100:4550–4556

Background A considerable interest has risen about the interplay between immune suppression and polarisation particularly with Th17 cells. Although Th2 polarisation seems intact in rheumatoid arthritis (RA), Th1 has long been known to be perturbed with lack of appropriate epigenetic commitment. Defective polarisation could be a mechanism by which Th1/Th17 cells preferentially develop. There are conflicting results about Th17 cell frequency in the blood of RA patients due to different phenotypes being used to define Th17 by flow-cytometry. Here, we investigated whether a novel epigenetic assay for Th17 cell frequency quantification could be used to differentiate patients with early RA from patients with early inflammatory arthritis symptoms evolving towards an alternative diagnosis. Materials and methods We used an epigenetic Th17 qPCR assay that quantifies Th17 cells based on a cell-specific methylation pattern at a key position in the IL-17 gene/promoter (Th17 qPCR; ). CD4+ T-cell frequencies were measured using the same epigenetic assay format based on a CD4+ T-cell-specific biomarker and results were used to normalise Th17 frequency to CD4+ T-cells. The Th17/CD4+ T-cell epigenetic assays were performed in whole blood extracted from healthy controls (HC, n=45), stable undifferentiated arthritis (UA) over 12–24 months (n = 43), UA patients progressing to RA (n = 22) and newly diagnosed RA patients (n = 45), other inflammatory arthritis (n = 30) and non-persisting symptoms (n = 16), as well as palindromic arthritis (n = 15). Results There was no difference between groups for the frequencies of CD4+ T-cells. Data suggest that Th17/CD4+ T-cells are significantly less frequent in the blood of early, DMARDs naive RA patients compared to HC (p < 0.0001) and patients with long lasting UA (p < 0.0001). While HC and UA are not different (p = 0.513), patients starting with UA and progressing to RA (within 6–12 months) are different from both HC (p < 0.00001) and stable UA (p = 0.001), but not from RA (p = 0.475). This finding can be explained by Th17 cell recruitment to the inflammatory disease site from the circulation as recently reported1,2,3 thus resulting in lower Th17/CD4+ T-cell frequencies in the blood. A power calculation based on these results suggested the need to test a further 30 patients from an early arthritis clinic evolving towards alternative diagnoses (AS, PsA, gout, reactive), and 16 patients with non-persistent symptoms. 15 patients with palindromic arthritis were also included. The measurement showed that Th17 frequencies in RA and other forms of inflammatory arthritis are less significantly different (p = 0.01), although frequencies between RA and non-inflammatory diseases are (p < 0.00001). Palindromic arthritis was not distinguishable from RA. Conclusions While there are conflicting results with respect to Th17 cell frequency in blood, we demonstrated using an epigenetic qPCR assay that RA patients show robust differences in their blood Th17/CD4+ levels compared to the group comprising HC, UA or non-persistent diseases. There was less difference between RA and other forms of inflammatory arthritis including palindromic arthritis. Our results suggest that together with Treg, Th17 may leave the blood in early RA and be recruited to the site of inflammation. A certain degree of interplay and plasticity between Treg/Th17 cells may explain why Treg are not capable of suppressing responses at the disease site if Th17 are also present. References 1. Chalan P, et al . PLoS One; 2013;8(11):e79370 2. Arroyo-Villa I, et al . PLoS One; 2012;7(8):e42189 3. Hirota K, et al . J Ex Med; 2007;204(12):2803–2812

Background and objectives Despite the well-established value of currently used diagnostic criteria for RA there is a constant demand for novel biomarkers to further improve early diagnosis notably in ACPA-negative patients. Inteleukin7 (IL-7) has been reported as a candidate diagnostic biomarker based on ACR-1987 criteria for RA, within 6 months of symptom development. 1 The aim of the present study was to validate these results using EULAR-2010 criteria. Materials and methods 255 patients attending early arthritis clinic were recruited. Demographic (age, gender, symptom duration) and clinical data (joint counts swollen and tender (SJC, TJC), CRP, DAS28, autoantibodies (RF and ACPA), SE) were collected. A commercial ELISA (R&D Systems, Abingdon, UK) was used to measure IL-7. Univariate and regression analysis were used to model outcome. Results At follow-up over 24 months, 118 patients had a confirmed diagnosis of RA (EULAR-2010) while 137 were classified as non-RA (persistent UA, other diagnostic or non-inflammatory). The RA diagnosis was associated with ACPA, RF and SE positivity (p Conclusions This study validates previous results using the EULAR-2010 criteria for diagnosing RA (in UK population of patients), however the predictability associated with lower IL-7 is less significant than in the initial study using ACR-1987 criteria (in French cohort). IL-7 remains a potential biomarker for ACPA-negative RA although further validation with larger number of ACPA-negative RA is needed. Further replication studies are required to translate these results into clinical applicability. Reference Goeb VP, Aegerter R, Parmar P, et al . Progression to rheumatoid arthritis in early inflammatory arthritis is associated with low IL-7 serum levels. Ann Rheum Dis. 2013;72:1032–6

Objectives The current optimal therapeutic approach in early rheumatoid arthritis (RA) is to start Methotrexate (MTX) to target inflammation and induce remission. Prediction of MTX therapy response remains a key clinical need to enable the identification of patients who would benefit from an alternative, more aggressive treatment strategy. T-cell subset analysis has been shown to have biomarker value in predicting MTX remission in early RA. 1 Given the clinical importance of these findings, our aim was to validate the original data 1 using T-cell subset quantification this time performed solemnly by routine hospital services, in order to replicate these original findings and validate naive T-cell frequencies as biomarker of the the achievement of clinical remission with MTX treatment in early RA. Methods T-cell subsets were quantified in 70 drug-naive early RA patients, commencing MTX. The primary outcome was remission (DAS28 Results Remission was achieved in 37/70 patients. In univariate analysis, the achievement of remission was strongly associated with age-corrected naive CD4 + T-cell frequency and current not-smoking (p Conclusion This study confirms the ability of baseline naive T-cell subset analysis in predicting early RA remission with first-line MTX therapy, suggesting clinical utility for the selection of the most appropriate therapy at disease presentation. All data being acquired by hospital routine services, they should be transferable widely towards personalised medicine. Reference Ponchel F, Goeb V, Parmar R, et al . An immunological biomarker to predict MTX response in early RA. Ann Rheum Dis 2014;73(11):2047–53

Background Immuno-senescence and inflammageing are features of the ageing immune system. Such age-related abnormalities may be synergising with structural defects of the musculoskeletal system augmenting the development of OA. Our aim was to establish if abnormalities of blood immune cell composition were associated with OA, beyond defects already associated with ageing. Methods Blood was collected from 120 healthy controls (age 18–69) to establish variations associated with age and 110 OA patients (age 49–85). Synovial tissue biopsies were obtained for 52 OA patients. We examined loss or acquisition of age-related changes in blood immune cells composition using flow-cytometry to establish frequencies of CD4/CD8, B, NK-cells. T and B-cells were analysed further for phenotypes including an inflammation related cell (IRC) subset. Immunohistochemistry was used to analyse immune cell infiltration in OA synovial tissue. Results OA synovial tissue had a normal appearance in 13/52 samples. In 28/52 samples diffuse infiltration of T and B-cells could be observed while formation of small aggregates was found in 10 cases. The presence of an ectopic germinal centre-like structure was only seen in one sample. Lineage analysis showed no change in relation with age for NK, CD4, and B-cells, weak decline in CD8 ( rho = -0.300, p = 0.019) and increase in NKT ( rho = 0.315, p = 0.012). Phenotyping T/B-cells revealed clear age-related changes with reduction of naive CD4 + T-cell (rho = 0.817, p (rho = -0.518, p = 0.002). In OA, NK cells become positively related to age ( rho = 0.350, p rho = -0.318, p = 0.001, rho = -0.260,p = 0.006 respectively). The CD8 relationship was lost. Naive CD4 cells appeared to be particularly affected with increased frequencies in 32/110 patients. Treg were reduced in 41/105 patients irrespective of age. IRC were only increased in 16 patients. B-cell subsets were also affected resulting in loss of age-relationships observed in health. We observed that longer disease duration was associated with loss of CD4 + T cells ( rho = -0.297, p = 0.025). IRC frequency was also associated with disease duration ( rho = 0.248, p = 0.058) and were also more frequent in patients affected in the hip (p = 0.002) or foot (p = 0.023). We observed no further association with B-cell subsets/phenotypes. Conclusion This analysis of the immune cell composition of the blood of OA patients suggests that immune dysfunction is present in OA beyond what is directly related to ageing. The relationship to severity of OA inflammation requires further investigation. The therapeutic implications of this warrant further study.

Background Lack of accurate early diagnostic testing of Rheumatoid Arthritis (RA) remains the limiting factor for early treatment to achieve remission. We hypothesised that molecular pathways implicated in RA pathology in the joint can inform a biomarker discovery programme based on flow-cytometry phenotyping of blood cells. Based on published work, we selected 74 subset/phenotypes and used high-dimensional dataset analysis to established their potential value in discriminating patients from an early arthritis clinic with Methods 46 patients were enrolled. 6–8 colour flowcytometry was performed using standard protocols. We recorded% of positive cells when 2 populations could be distinguished, and levels of expression (MFI) for single populations. A random forest ordered predictors from these 74 subset/phenotypes according to importance. Then a classification tree was fitted based on the most important factors and finally the variables identified from the classification tree were fitted in a logistic regression. Results 23 of the 46 patients had RA, 11 non-persistent inflammation and 12 other rheumatic diseases (AS, SPA, gout, OA and reactive). 47 individual phenotypes were analysed and 36 additional ones using a combination of 2 markers. No difference in lineage representation was found; significant difference were observed for 11/47 subsets (P Due to the low number of patients, multivariable analysis was limited. We ran un-supervised cluster analysis which separated 2 groups (RA/non-RA) quite efficiently with 5-RA and 6-non-RA misclassified. A 3-node classification tree using 47 phenotypes classified 20/23 RA patients correctly. The 3 phenotypes showing best discrimination power were IL-6R+NKT-cells, naive CD4 + T-cells and CCR6+ monocytes. A logistic regression confirmed that these 3 phenotypes were highly significant phenotypes (P + T-cells phenotypes, 1 CD8 + T-cells, 2 B-cells, 1 monocyte and 1 NKT-cells have potential value for discriminating RA from non-RA patients based on accuracy. Conclusion Despite the small sample size, this analysis demonstrated the value of hypothesis driven marker analysis, with 12 of the 47 phenotypes predicted to have potential showing significant differences. By triangulating different analysis approaches, the robustness of the findings was improved giving confidence in the identification of relevant biomarkers.

Background Synovial inflammation is known to play a central role in the development of bony erosions and articular cartilage degradation in RA. We aimed to determine if circulating metalloproteinase-3(MMP-3), hyaluronic acid (HA), serum amyloid A(SAA), interleukin 6(IL-6) and calprotectin (S100A8/A9) would outperform CRP as surrogate biomarkers for synovitis detection. Methods Patients were recruited to a randomised controlled trial; 1 baseline samples and clinical data were included in this analysis. Patients fulfilled 1987 ACR RA classification criteria, had 3–12 month symptom duration, active disease (DAS44 > 2.4) and were DMARD naive. In a subset, grey scale (GS) and power Doppler (PD) ultrasound (US) semi-quantitative scores (0–3) were assigned to wrists, MCPs and PIPs 2 and 3 and MTPs 1–5 bilaterally. Both counts (joints scoring GS > 1, PD > 0, simultaneously GS > 1 and PD > 0) and score totals were created. To reflect synovitis burden, joint measures were weighted by multiplying each joint’s score by a relative area weight, 2 before summating. Samples were tested for MMP-3, HA, SAA, IL-6, S100A8/A9 using research-use-only multiplex platform IMPACT (Immunological Multi-Parameter Chip Technology) (Roche Professional Diagnostics, Germany), 3 CRP was measured in the routine lab. We calculated bootstrapped confidence intervals (CI) for the differences between CRP and the other markers in the strength of Kendall’s tau-a rank correlation with the joint assessments using Stata 13.1. Results Fifty-nine patients were included: mean (SD) age 52.7 (13.8) years, 71% female, median (IQR) disease duration 1.0 (0.7, 1.6) months. Marker values [median (IQR)] were CRP 16.4 (8.2, 52.0); MMP3 58.9 (42.1, 119.3); HA 36.5 (20.5, 70.2); SAA 36.5 (6.3, 249.0); IL-6 46.3 (21.4, 69.1); S100 A8/A9 12.1 (8.6, 15.2). Associations were stronger when joint weights were used, particularly for CRP and MMP3 (data not shown), but remained weak-to-moderate; the strongest were between weighted clinical joint counts and CRP. The association with SJC28 was substantively stronger for CRP (tau = 0.46) than for any of the other markers and was significantly stronger than for HA (tau = 0.24; difference 0.04–0.44). Similar values were obtained using 28- or 44-joint clinical counts. CRP, MMP3 and IL-6 were associated with US to a comparable degree (tau circa 0.30); associations with US for SAA and S100A8/A9 were slightly weaker than CRP, although differences were not significant. Conclusion Weighting for joint area improved correlation with circulating inflammatory markers. None of the markers outperformed CRP as a surrogate biomarker of synovitis across the full range of observed values. We intend to investigate their utility in patients with low/normal CRP. Funding: Roche Professional Diagnostics provided free of charge access to the IMPACT platform and IMPACT reagents. This work was also supported by grants from Arthritis Research UK and the NIHR. References Nam JL, Villeneuve E, Hensor EMA, et al . Ann Rheum Dis 2014; 73 :75. Lansbury J, Haut DD. Am J Med Sci 1956; 232 :150. Claudon A, Vergnaud P, Valverde C, et al . Clin Chem 2008; 54 :1554.

Background and Objectives ACPA+ individuals with non-specific musculoskeletal symptoms are at high risk of developing rheumatoid arthritis (RA). We previously demonstrated dys-regulation of T-cell subsets with loss of naive and regulatory T-cells (Treg) in early disease. The aim of the current study is to demonstrate the predictive value of T-cell subset analysis for progression towards symptom onset in ACPA+ individuals. Materials and Methods 84 ACPA+ individuals without clinical synovitis at recruitment were followed. 95 healthy controls (HC) provided a reference group. At baseline T-cell subset analyses were performed using 6-colour flowcytometry for naive T-cells (CD4+ CD45RB + CD45RA+ CD62L+), Treg (CD4+ CD25 high Foxp3 + CD127 low ) and inflammation related cells (IRC: CD4+ CD45RB + CD45RA+ CD62L-). The relationship between naive cell frequency and age was established in HC and used to age-correct values in ACPA+ . ROC curve analysis was used to identify 2 T-cell cut-offs predicting progression to IA at any time; one which maximised the Youden index (sensitivity + specificity-1), and one which prioritised specificity over sensitivity. Results 42/84 (50%) of patients developed clinical synovitis within a median follow-up of 6.0 months (range 1 week-46 months). For age-corrected naive T-cells area under the ROC curve (AUC) was 0.67 (95% CI 0.55, 0.79; n = 84, p = 0.007), for IRC 0.70 (0.59, 0.81; n = 81, p = 0.002) and for Treg 0.67 (0.53, 0.80; n = 65, p = 0.021). For each of the three subsets, the Youden index cut-off correctly classified >65% of patients (Table 1). Cut-offs prioritising specificity were identified which did not greatly reduce overall classification success. The confidence intervals for these estimates remain wide and our sample size may still be limited for running such analysis. Abstract A1.33 Table 1 Sensitivity and specificity of T-cell subset frequencies for progression to IA, using two different cut-off values for each subset; one where the Youden index was maximised and another that prioritised specificity over sensitivity. Conclusions T-cell dys-regulation in ACPA+ individuals with non-specific musculoskeletal pain may be useful in predicting progression to inflammatory arthritis. Multivariable modelling in larger cohorts is needed to quantify the utility of T-cell subsets in predicting progression to IA.

Background Biomarkers for early diagnosis of RA are still needed despite the improvement brought about by the use of the ACR/EUALR-2010 criteria which include the ACPA biomarker. B-cell subset dys-regulation has not been studied in detail in the early RA disease continuum. The aim of this pilot study is to determine whether B-cell subset phenotyping investigating differentiation subset and cell surface expression for TLRs and chemokine receptors can discriminate between patients who clearly are RA from those with other (inflammatory) rheumatic diseases in an early arthritis clinic. Methods 46 patients with Results Using newly developed 2010-RA diagnosis criteria 23 of the 46 patients were diagnosed with RA. Not surprisingly, ACPA positivity was associated with RA (P + CD27+CD38-) were increased in RA patients (P = 0.028). Furthermore, putative regulatory B-cells (CD19 + CD38+CD24+) were reduced (P = 0.003). More B-cells expressing TLR4 (P = 0.037) were associated with RA but less B-cells expressed CXCR4 (P = 0.012) and CCR6 (P = 0.002). Clustering analysis was used to further investigate pattern of expression of specific groups of markers or phenotypes. Clustering showed 3 groups of patients, one enriched for RA and the other two with mixed RA and non-RA patients (P = 0.088). ACPA positivity was associated with the first group (P = 0.024). Clustering also highlighted IL-6R as a possible marker associated with subgroup-1 (p = 0.001) and ACPA positivity. Conclusion There are clear B-cell related immunological differences in patients with early RA. We will therefore need to evaluate these biomarkers in early inflammatory patients with a high risk of progression towards RA and establish their value in larger number of patients.

Background Patients with rheumatoid arthritis (RA) are at higher risk of cardiovascular (CV) disease. Emerging evidence suggests that this accelerated risk is present in the early stages of RA. Objectives Our objective was to describe the CV profile of those with a new diagnosis of inflammatory arthritis (IA). Methods 339 patients (pts) with IA enrolled into the Inflammatory Arthritis disease CONtinuum (IACON) registry in Leeds were assessed for known CV disease (CVD) and CV risk factors (RFs) including; hypertension, diabetes, smoking, family history of CVD, blood pressure, body mass index, waist/hip circumference and lipid profile. In a sub-group of those meeting ACR RA criteria (n=92); insulin resistance (homeostasis model assessment of insulin resistance (HOMA-IR) and quantitative insulin sensitivity check index (QUICKI)) and NT-proBNP were measured. Pts were split into 3 groups; RA1; pts meeting 1987 ACR RA criteria (n=143, 42%), RA2; pts meeting 2010 ACR RA criteria (n=55, 16%) and UIA; those not meeting either ACR criteria for RA i.e. undifferentiated IA (n=141, 42%). Descriptive statistics and analysis for correlation, chi-squared and ANOVA tests were performed. Results Positivity for RF (p No differences in the prevalence of traditional CV RFs, overt CVD or metabolic syndrome (according to WHO and IDF criteria) were seen. Hypertension was prevalent across all groups (28, 27 and 25% respectively). There was a trend to significance in the prevalence of hypercholesterolaemia (26, 13 and 17% respectively, p=0.067). Similar proportions of pts within the groups were receiving anti-hypertensive, anti-platelet and lipid-lowering (LLT) therapies. Total cholesterol (TC) was lower in RA1 vs. RA2 (p=0.041) but this lost significance when results were adjusted for CRP. There was no significant difference in TC/HDL ratio. In RA1 and RA2, TC and LDL negatively correlated with CRP (p=0.003, rho=-0.402 and p=0.041, rho=-0.28 respectively), and TC/HDL ratio and HDL showed no association with CRP. Pts ACPA+ve were more likely to be smokers (27 vs. 19% ACPA–ve, p=0.016), have a past history of hypercholesterolaemia (25 vs. 15% ACPA-ve, p=0.029) and be receiving LLT (21 vs. 12% ACPA-ve, p=0.042). However, there was no significant difference in the lipid profile, even with adjustment for CRP and pts receiving LLT excluded. There was no significant difference in HOMA-IR, QUICKI and NT-proBNP between RA1 and RA2 or according to ACPA status. Conclusions Within this cohort of early IA CV RFs are prevalent, correlating with inflammation and not just a specific phenotype of RA. This supports the hypothesis that inflammation plays a key role in the known accelerated CV risk in RA, and effective treatment across the continuum is important. There may be an association between ACPA and lipid metabolism. Long-term follow-up is required to evaluate potential change in risk profile with effective disease modifying treatment. Disclosure of Interest L.-A. Bissell: None Declared, A. Burska: None Declared, S. Horton: None Declared, H. Donica: None Declared, J. Freeston: None Declared, M. Buch Consultant for: Abbot, Pfizer, Roche, Chugai, BMS, UCB, P. Emery Consultant for: Abbot, Pfizer, Roche, Chugai, BMS, UCB, MSD, J. Andrews: None Declared

Background Patients with rheumatoid arthritis (RA) have an elevated risk of cardiovascular disease (CVD), predominantly ischaemic heart disease (IHD). Systemic inflammation is thought to be a key contributor to this risk and it is hoped early therapeutic suppression of inflammation will reduce cardiovascular (CV) events in the long term. Insulin resistance, associated with metabolic syndrome, N-terminal pro-brain natriuretic peptide (NT-proBNP) and atherogenic lipid profiles have been proposed as potential surrogate measures of atherosclerosis in RA. Each has been shown to correlate with systemic inflammation, but their utility in RA above measurement of the acute phase response remains unknown. Objectives To determine whether suppression of inflammation in early RA influences surrogate measures of atherosclerosis, including the homeostasis model assessment of insulin resistance (HOMA-IR), NT-proBNP and dyslipidaemia. Methods CRP, fasting glucose and lipids, and in a sub-group, NT-proBNP and fasting insulin, were measured at baseline and week 26 in DMARD naive RA patients, with 3-12 months symptoms, recruited into the Infliximab as Induction therapy for Early rheumatoid Arthritis (IDEA) multicentre double-blind randomised controlled trial. Patients met the 1987 ACR criteria. Treatment was randomised to IFX +MTX or MTX with 250mg IV methylprednisolone as induction therapy. 120mg IM methylprednisolone was given if DAS>2.4 at 3 time points within the first 26 weeks. HOMA-IR was calculated at each time point. Descriptive statistics, Wilcoxon and paired t-test, and Spearman’s correlation were performed. Results Data from 112 patients (69% female, mean age 53 years, 55% RF and 70% anti-CCP positive) were analysed. At baseline, 9% were on a statin and 14% an anti-hypertensive agent. 4 patients had known IHD and were excluded from the subsequent analysis. The median (range) CRP was 13mg/l (0-30) with CRP normal ( 1 . In the subgroup of 79 patients with data on NT-proBNP and fasting insulin, 17% and 38% of patients had raised levels of NT-proBNP and HOMA-IR, respectively. Comparing week 26 to baseline, there was a significant fall in CRP (p Conclusions In this study, suppression of inflammation was associated with improvement in insulin resistance and increased HDL, but no significant change in NT-proBNP, TG or LDL following 26 weeks intensive RA treatment. Further analyses of biomarkers at week 78 and drug specific effects are underway. Long term follow up is being performed to determine if those patients with persistently abnormal CV biomarkers develop subclinical or overt CVD in the future despite intensive early treatment for RA. References JAMA. 2001;285(19):2486-97 Disclosure of Interest L.-A. Bissell: None Declared, S. Mackie: None Declared, L. Kozera: None Declared, J. Nam: None Declared, A. Burska: None Declared, E. Hensor: None Declared, H. Keen: None Declared, E. Villeneuve: None Declared, H. Donica: None Declared, P. Conaghan: None Declared, J. Andrews: None Declared, P. Emery: None Declared, A. Morgan Grant/Research support from: Supported in part by a research grant from Investigator-Initiated Studies Program of Merck Sharp & Dohme Limited. The opinions expressed in this paper are those of the authors and do not necessarily represent those of Merck Sharp & Dohme Limited