B. Zengin

SAT0184 The effect of smoking on response to tumor necrosis factor-alpha inhibitor treatment in ankylosing spondylitis patients: results from the turkbio registry

Background: Although there is good evidence that smoking has a dose-dependent impact on structural damage progression in ankylosing spondylitis (AS) the evidence is poor for its impact on disease activity, physical mobility, life quality and treatment response. Objectives: We aimed to investigate the impact of smoking on disease acitivity, treatment adherence and treatment response in Turkish patients with AS treated with their first tumour necrosis factor-alpha inhibitor (TNFi) therapy in a real-life cohort. Methods: 561 patients fulfilling the modified New York criteria for AS and treated with their first TNFi therapy since 2011 from 8 centers in Turkey were included in the analysis. Treatment response was evaluated as achievement of “BASDAI50” or “ASDAS Clinically important improvement (CII)” at the 3-months’ and 6 months’ visits. Clinical and demographic parameters were compared between current/never and current/previous smoker groups. Demographic and descriptive data are presented by medians/interquartile ranges (IQRs). Groups were compared by non- parametric tests (x2, Kruskal Wallis and Mann Whitney tests). Kaplan Meier plots, Cox and logistic regression analyses were calculated for treatment adherence and treatment response. Results: Among 561 AS patients included in the study, 506 (90%) had known smoking status (37% current, 35% never, 17% previous smokers). The median follow-up time was 1.9 years (IQR 0.85–3.5) and disease duration was 3.1 years (0,6–7,7). At baseline, current smokers were younger (34, IQR 29–41) compared with never (38, IQR 30–46 p=0.007) and previous smokers (42, IQR 34–49 p<0,001). Current smokers had male predominance (n=148, 43.9%; n=85, 25.2%); lower erythrocyte sedimentation rate (28 mm/h (13–42); 34 mm/h, (20–49) and higher change in BASMI (40, IQR 10–57.5; 10, IQR 4–30) compared with never smokers (all p<0.005). HLA status, body mass index, CRP, baseline disease indexes (BASDAI, BASFI, BASMI, HAQ, ASDAS) and treatment response was not found to be different between current and never smoker patients in our population (table 1). In multivariate analysis, male (OR:1,98; 95% CI (1,39–2,82), p<0,01), HLA positive (OR:1,54; 95%CI (1,08–2,18), p=0,016) and active DMARD user (OR:1,84; (95%CI 1,12–3,01) p=0,015) patients had better treatment response and treatment adherence ((HR:1,93; 95% CI (1,36–2,73); HR:1,60; 95% CI (1,13–2,27); HR:1,80; 95% CI (1,10–2,95) all p<0,005) but smoking status were not significant (p>0,05). Conclusions: In this study of TNFi-treated AS patients in clinical practice, smoking was not found to be associated with disease activity, treatment response and treatment adherence. Disclosure of Interest: None declared

AB0835 Comparison of long term anti-tnf survival in patients with ankylosing spondylitis and non-radiographic axial spondyloarthritis; data from turkbio registry

Background Limited data are available on anti-TNF survival in non-radiographic axial spondyloarthritis (nr-axSpA) patients and their long-term survival in ankylosing spondylitis (AS). Objectives The aim of the study was to evaluate long term survival of the first anti-TNF drug treatment among patients with AS and nr-axSpA enrolled in the TURKBIO database and to compare the discontinuation rates for infliximab (INF), etanercept (ETN), and adalimumab (ADA) in each of the two groups. Methods All AS and nr-axSpA patients receiving biological therapies registered in the TURKBIO database between the dates of october 2011 and april 2017 were included in the study. AS diagnosis was made according to modified New York classification criteria and nr-axSpA according to ASAS AxSpA classification criteria. Demographic and clinical data, the date of starting to use biological drug, using frequency and dose of biological drugs, BASFI, BASDAI, BASMI, ASDAS scores, date and reason for discontinuing to use drug were collected. Baseline characteristics and drug survival rates were compared between AS and nr-axSpA patients. Drug survival was calculated by the Kaplan-Meier method and risk for discontinuation among treatment groups cpmpared by Long Rank test. Results A total of 924 patients were included in the study (AS, n=871 and nr-axSpA, n=53). More than half of the patients with AS were male (60.7% in AS vs 34.0% in nr-axSpA group, p<0.001).AS patients had longer symptom duration (104.90±79.06 vs 75.11±45.29 months, p<0.036) compared to nr-axSpA. Median levels of CRP and ESR were similar for nr-axSpA (CRP: 27.03±34.71, ESR: 30.50±25.77) and AS (CRP: 22.32±29.95, ESR: 35.40±22.91). The scores of BASFI, BASMI and ASDAS were found to be similar in both groups. Median BASDAI scores at first TNFi initiation were higher in patients with nr-axSpA than in patients with AS (58.65±18.21, 51.06±18.91, p=0.030). Cumulative drug survival rates did not show significant difference among INF (at 59. months:18,5%), ADA (at 71. months: 39,5%) and ETN (at 51. months: 24,2%) in nr-axSpA group (p=0,699) (figure 1). Similarly, drug survival rates at 78, 77, 78. months for 3 anti-TNF drugs had shown no difference in AS patients (INF (at 78. months: 38,1%), ADA (at 77. months: 52,4%), ETN (at 78. months: 39,0%)) (p=0,151) (Figure 2). Cumulative survival rates in AS patients (at 78. months:42,2%) were found to be significantly higher than that (at 71. months:28,2%) in nr-axSpA patients (p<0,001) (Figure 3).Abstract AB0835 – Figure 1 Drug survival rates anti-TNF in nr-axSpA. Abstract AB0835 – Figure 2 Drug survival rate by anti-TNF in AS. Abstract AB0835 – Figure 3 Overall drug survival on first anti-TNF in nr-axSpA and AS patients. Conclusions In contrast to the literature that revealed similar short term survival rates for anti-TNF drugs in patients with AS and nr-axSPA, we found higher survival rates in patients with AS compared to patients with nr-axSpa in this long-term observational study.A limitation of the study may be the low number of nr-axSpa patients using anti-TNF, related to the requirements of social insurance system. Disclosure of Interest None declared

Phenotype Differences in HLA-B27 Positive Versus Negative Patients with Ankylosing Spondylitis in a Population That Show Relatively Weaker Association with HLA-B27

For a searchable version of these abstracts, please visit www.acrabstracts.org.

THU0314 Radiologic activity is the major determinant for physicians while deciding active disease in takayasu arteritis

Background There are no valid follow-up parameters in the assessment of disease activity in Takayasu arteritis (TA). Objectives We investigated the impact of incorporation of vascular imaging into ITAS in the assessment of disease activity in TA. Methods 52 patients who fulfilled the ACR criteria were included in the study. PGA, Kerr et al.s criteria and ITAS2010/ITAS-A scores were evaluated in all patients in serial visits. All the patients were followed using 3–6 monthly B-mode/Doppler ultrasonography (USG) and 6–12 monthly magnetic resonance angiography (MRA). Radiological activity (Rad) was defined based on the presence of any of the 3 parameters including new vessel involvement by any technique (5 points),increase in vessel wall thickness on USG (3 points) and vessel wall edema on MRA (3 points).Then we incorporated these scores with ITAS-A to obtain a composite disease activity index (ITAS2010-A-Rad) (Table 1). Active disease was defined as ITAS-A-Rad >4 points. Results Total 410 visits of 52 TA patients (mean age 50.7 yrs, F: 92.3%, mean follow-up duration:6.4±2.9 yrs) were evaluated. Radiological assessment was done in 359 visits (by USG in 271 and by MRA in 190). Patients were categorized as having active disease in 194 visits (47.4%) according to PGA and 72 visits (17.5%) according to Kerr et al. criteria.The agreement between them was fair (66%, κ: 0.29). Radiological activity was determined in 105 out of 359 visits (29.2%). The total agreement between radiological activity and Kerr at al. criteria was 83% (κ: 0.58). It was found to be 76% (κ: 0.52) between radiological activity and PGA. Mean ITAS-A-Rad scores were found to be significantly higher in visits with active disease compared to visits with inactive disease according to both PGA and Kerr et al. criteria (Table 2). The ITAS-A-Rad was significantly correlated with all the other activity parameters including ITAS2010, ITAS-A, and APRs. There were 43 visits with new vessel involvement by any radiologic technique; all visits included patients with active disease based on both PGA and Kerr et al. criteria. Whereas in 50% of these visits, patients had normal CRP, and %49 had normal ESR. The agreement between ITAS2010 and PGA was fair (69%, κ: 0.38).When APR was added (ITAS-A), it did not improve (68%, κ: 0.34). But the agreement between ITAS-A-Rad and PGA (72%, κ: 0.50) and also Kerr et al. criteria (82%, κ: 0.56)was found to be moderate. Interestingly, when only USG (ITAS-A-USG) or only MRA (ITAS-A-MRA) was used, the agreement with PGA was remained unchanged (73%, κ: 0.45 and 76%, κ: 0.52, respectively). When responsiveness to change of ITAS-A-Rad score was evaluated by serial visits of patients, it was found that the mean value of the score was discriminative for activity according to PGA in 9 of 11 visits (Figure 1).Table 1. The definition of ITAS-A-Rad Score Clinical ITAS2010 0– Laboratory APR ESR 0 for ESR<20 1 for 21–39 2 for 40–59 3 for >60 mm/h 0–3 CRP 0 for CRP≤5 1 for 6–10 2 for 11–20 3 for >20mg/l Radiology Radiological activity New vessel involvement with any radiological method 5 B-mode Doppler USG Progression on vessel wall thickness 3 MRA Presence of vessel wall edema 3 Total ITAS-A-RAD Score ITAS-A-Rad Score >4 –> Activity. Conclusions The results of this study suggest that ITAS-A-Rad may be used to be a valuable foIlow-up parameter in the assessment of disease activity. Disclosure of Interest None declared

AB0597 Adaptation of Ucla Scleroderma Clinical Trial Consortium Gastrointestinal Tract 2.0 Questionnaire into Turkish

Background Nearly 90% of patients with scleroderma (SSc) have gastrointestinal tract (GIT) involvement in variable severities and is a challenging process for clinicians. The University of California Scleroderma Clinical Trial Consortium Gastrointestinal Tract 2.0 (UCLA SCTC GIT 2.0) is a questionnaire including 34 items, 7 multi-item scales: reflux, distention/bloating, diarrhea, fecal soilage, constipation, emotional well-being and social functioning. By these parameters, a total GIT score is calculated (1). This scale translated in German, Italian, French, Polish, Spanish, Swedish, Dutch before, they are available in http://www.uclascleroderma.researchcore.org/website (1–3). There is no Turkish version of this scale yet. Objectives Our aim is to make translation, cultural adaptation of the UCLA SCTC GIT 2.0 into Turkish, and assess reliability of the scale in patients speaking Turkish. Methods UCLA SCTC GIT 2.0 scale was translated into Turkish according to international guidelines and applied to 97 SSc patients. The questionnaire repeated in 29 patients after an interval of 15 days for determining reliability. For internal consistency, Cronbachs alpha was calculated, reliability coefficient if item deleted and test-retest reliability also determined. External consistency was measured by comparing with the Short Form (SF)-36 by Spearmans correlation analysis (rho: ≤0.29 weak, 0.30–0.49 middle, ≥0.50 strong). Results 97 scleroderma patients were included in this study (female:87.6%, mean age:55.4±11.4). Internal consistency Cronbachs alpha was calculated as 0.89, reliability coefficient if item deleted was 0.89–0.90. External consistency of UCLA SCTC GIT 2.0 was measured by comparing with the SF-36, correlation was meaningful in medium level (Table 1,2).Table 1. Descriptive statistics and internal consistency statistics UCLA SCTC GIT 2.0 Scale n Mean score (SD) Minimum score Maximum score Cronbach alpha Floor effect % Ceiling effect % Reflux 97 0.64 (0.54) 0.0 2.6 0.83 17.5 0.0 Distension 97 1.02 (0.75) 0.0 3.0 0.58 7.2 1.0 Soilage 97 0.30 (0.72) 0.0 3.0 0.68 82.5 3.1 Diarrhea 97 0.28 (0.47) 0.0 1.5 0.36 69.1 0.0 Social Functioning 97 0.17 (0.32) 0.0 1.3 0.47 67.0 0.0 Emotional Wellbeing 97 0.30 (0.43) 0.0 2.2 0.73 41.2 0.0 Constipation 97 0.63 (0.69) 0.0 2.5 0.56 34.0 0.0 Total GIT score 97 0.45 (0.37) 0.0 1.6 0.82 3.1 0.0 All scales are scored from 0.00 (better HRQOL) to 3.00 (worse HRQOL) except the diarrhea and constipation (range from 0.00–2.00 and 0.00–2.50, respectively). The UCLA GIT 2.0 provides a total score of GIT severity and calculated by summation of all scales (except constipation) and ranges from 0.00 to 2.83. Conclusions UCLA SCTC GIT 2.0 scale had strong internal consistency, good reliability and acceptable validity when adapted into Turkish. Turkish-speaking patients with scleroderma, this scale will be useful to assess GIT symptoms. The basic constraint of our study was, not using image procedures for objective GIT involvement evidences. References Khanna D,Reliability and validity of the UCLA SCTC GIT Instrument.Arthritis Rheum,2009. Bae S,Development and validation of French version of the UCLA SCTC GIT Instrument.Clin Exp Rheumatol, 2011. Meijs J,Translation,cross-cultural adaptation,and validation of the UCLA SCTC GIT 2.0 into Dutch. Clin Exp Rheumatol,2014. Disclosure of Interest None declared

FRI0229 Evaluation of Patient Acceptable Symptom State in Patients with Axial Spondylarthritis; Similar Thresholds for Radiographic and Non-Radiographic Subgroups

Background The Patient Acceptable Symptom State (PASS), a single-question outcome, has been defined as an absolute level of patient well-being. A few studies have assessed PASS in patients with ankylosing spondylitis (AS), but it is not known whether the results of those studies apply also to the group of non-radiographic (nr) axial spondylarthritis (axSpA) Objectives To estimate the PASS values for disease activity and several patient reported outcomes both in the whole group of axSpA and in the two subgroups of AxSpA. Methods This single-center cross-sectional analysis included patients fulfilling the ASAS criteria for axSpA, who have been registered in our local database. All patients responded to the global yes/no question for PASS. A variety of other outcome measures in regard with global scales, disease activity, functional status, health status and quality of life were collected at the same time. The thresholds at which patients rated themselves in PASS for disease activity (BASDAI and ASDAS) and for each of the assessed patient self-reported outcome measures were estimated using the 75th centile (25th centile for SF 36)estimation and receiver operating characteristic (ROC) analyses in the whole group, as well as in each subgroup of axSpA. Contributors which can affect PASS were evaluated with logistic regression analysis. Results The analysis was based on 356 axSpA patients (261 AS, 95 nr-axSpA) with a mean age of 42.2±12.0 years and mean disease duration of 14.7±10.8 years. Of the patients with axSpA, 271 (%76.1) considered themselves in PASS (76.6% in AS, 74.7% in nr-axSpA). PASS thresholds for disease activity and all other assessed outcome measures were shown in table and there were not significant difference between AS and nr-axSpA group. PASS cut-off points for BASDAI, BASFI and HAQ identified by the 75th percentile method were slightly higher than those determined by the ROC analysis, but similar for the rest of the outcome measures. The patients with an acceptable status had significantly lower mean disease activity scores and good results with the all outcome measures. PASS had no relationship with age, sex, disease duration and education (years) in logistic regression analysis. Of the axSpA patients with BASDAI (≥4), 61.4% and those with ASDAS (>3.5), 50% rated themselves in PASS, whereas 5.5% of the patients with a BASDAI score <2, and 4.5% of those with ASDAS <1.2 were not in PASS. Conclusions PASS thresholds for disease activity and outcome measures were similar to the figures previously reported in some studies with no apparent difference between patients with AS and nr-axSpA. However, more than half the Turkish axSpA patients considered themselves in PASS, which needs to be evaluated in further studies. Disclosure of Interest None declared

AB0766 Comparing Iphone Compass Application with Inclinometer and Universal Goniometer for the Assessment of Cervical Rotation in Patients with Ankylosing Spondylitis

Background Cervical rotation reflects restriction of mobility in axial disease in ankylosing spondylitis (AS) and it can be assessed in several approaches based on the use of either an inclinometer, a goniometer or a tape measure. New generations of smartphones are equipped with a gyroscope and an accelerometer which in combination with a smartphones operating system or specific software applications can be used for various inclinometric functions. Objectives The aim of the study was to assess the reliability and validity of using iphone built in compass application, as compared to using inclinometer and universal goniometer in the assessment of cervical rotation patients with AS. Methods The study sample included 30 AS patients (8 females, 22 males) with a mean age of 46.8 (±12.2). BASMI scores were obtained from patient charts. The mean BASMI score of AS patients was 37.6 (±30.6). Two examiners measured cervical rotation of each patient using iPhone4 compass application and also inclinometer and universal goniometer, twice with each method. A cap with a velcro patch on top and an iphone case with a Velcro patch on the bottom were used to stabilize the iphones position during measurements. Intra-rater and inter-rater reliability were examined with intra-class correlation coefficients (ICC). The agreements between the methods facing one another, were assessed by Bland-Altman method. Results We observed an excellent intra and inter-rater reliability in the whole study sample for all three methods (Table 1 and Table 2). Bland-Altman analysis showed a good agreement between the iphone and inclinometer methods with a mean difference (bias) of -5.6 for examiner 1 (95% CI -7.6 to -3.6) and -6.3 for examiner 2 (95% CI -8.8 to -3.8). Upper and lower limits of agreement were 4.9 (95% CI 1.4 to 8.3) and -16.2 (95% CI −19.6 to -12.7) for examiner, and 6.6 (95% CI 2.3 to 10.9) and -19.3 (95% CI -23.5 to -15.0) for examiner 2. Mean differences according to Bland-Altman analysis between the iphone and universal goniometer measurements were 2.3 for examiner 1 (95%CI -0.4 to 5.2) and 4.1 for examiner 2; between the universal goniometer and inclinometer were -8.0 for examiner 1 (95% CI -11.2 to -4.8) and -10.4 for examiner 2. Conclusions IPhone compass application is a simple and accessible way of measuring cervical rotation in patients with AS. Measurements acquired with iphone show excellent intra and inter-rater reliability and a good agreement with inclinometer and universal goniometer, with slightly lower values than obtained with inclinometer, and slightly higher values than with universal goniometer. Disclosure of Interest None declared

AB0803 Obesity is Associated with Psoriatic Arthritis and Contributes to the Increased Use of Biologics

Background An increased prevalence of obesity and metabolic syndrome (Mets) has been reported in psoriatic arthritis (PsA) as compared with the general population suggesting an association between the inflammation and atherosclerotic risk factors. Takayasu arteritis (TA) is a systemic inflammatory disease which affects mostly the aorta and its main branches. Increased atherosclerosis has also been reported in Takayasu patients. Objectives In this study, we aimed to investigate the prevalances of obesity and Mets in patients with PsA and to compare the results with those in patients with TA. Methods This cross-sectional study included patients with PsA and age matched control patients with TA who were followed in the Rheumatology out-patient clinic at Dokuz Eylul University. Patients with body mass index (BMI; kg/m2) between 25.0 and 29.99 were classified as overweight and those with ≥30 as obese. The NCEP-ACT III criteria were used to identify subjects with MetS. Disease activity was assessed in patients with PsA using the Composite Psoriatic Disease Activity Index (CPDAI), BASDAI and DAS28. HAQ and DLQI were used for the health assessment and BASFI for the functional status. Results There were 117 PsA patients (78 female; mean age: 47.2±12.0 years; disease duration: 8.0±8.4 years) who fulfilled the CASPAR criteria and 32 TA patients (29 female; mean age: 43.5±11.3 years; disease duration: 8.3±8.7 years) who fulfilled the ACR 1990 classification criteria for TA.The prevalence of obesity and overweight was found to be higher in PsA compared with TA (29.9% versus 18.8% and 40.2% versus 28.0%, respectively; p=0.03). The prevalence of MetS was 28.2% in PsA and 18.8% in TA (p0.05).In the further analysis of obese, overweight patients and patients with normal BMI, no differences were found regarding disease activity, functional status, health assessment, disease duration, and glucocorticoid and DMARD use (p0.05). However there was a higher frequency of biologic use in the obese and overweight patients compared with those with normal BMI (42.9% and 23.4%, respectively vs 11.4%, p0.05) Conclusions This study demonstrates a higher prevalence of obesity in PsA compared with TA. The increased use of biologics in obese patients suggests the association of obesity with worse disease severity in PsA. Disclosure of Interest None declared