Jun Won Park

Head-to-head comparison of udenafil vs amlodipine in the treatment of secondary Raynaud’s phenomenon: a double-blind, randomized, cross-over study

OBJECTIVEnRP is a reversible vasoconstriction of digital arteries that causes pain and skin discoloration. This study compared the efficacy of the new phosphodiesterase type 5 inhibitor udenafil with that of the calcium channel blocker amlodipine in the treatment of secondary RP.nnnMETHODSnA total of 29 patients with secondary RP associated with connective tissue diseases were enrolled in this double-blind, randomized, cross-over study. The patients were randomized to receive udenafil 100 mg/day or amlodipine 10 mg/day for 4 weeks. After a washout period they were crossed over to the other drug for another 4 weeks. The primary outcome was RP frequency before and after treatment. The secondary outcomes were RP condition scores, RP duration, number of digital ulcers, HAQ, physician global assessment and digital artery flow before and after treatment.nnnRESULTSnAmlodipine and udenafil both decreased the rate of RP attack significantly. The drugs did not differ in terms of RP frequency or any of the secondary outcomes except for digital blood flow; udenafil improved it significantly better than amlodipine (P = 0.021). Udenafil was well tolerated without serious adverse effects.nnnCONCLUSIONnUdenafil and amlodipine have comparable efficacy in improving RP attacks. In addition, udenafil improves the blood flow in digital arteries compared with amlodipine.nnnTRIAL REGISTRATIONnwww.clinicaltrials.gov, protocol number NCT01280266.

Genetic and non-genetic factors affecting the visual outcome of ocular Behcet’s disease

OBJECTIVEnTo examine the prognostic factors for visual outcome in Korean BD patients with uveitis.nnnMETHODSnSeventy-seven Korean BD patients with uveitis were enrolled. HLA-B and HLA-A genotypes were determined by PCR-based method. Visual acuity was measured by Snellen chart. Vision loss was graded into visual impairment (VI) defined as VA<20/40 for more than 6 months, loss of useful vision (LUV) as VA < 20/200, and near total blindness (NTB) as VA of light perception or worse.nnnRESULTSnVI was associated with a longer duration of uveitis, posterior uveitis, and cataract, LUV with male gender, a longer duration of uveitis, posterior uveitis, and cataract, and NTB with a longer duration of uveitis, cataract, and glaucoma. HLA-B*51 and HLA-A*26:01 did not show any association with VI, LUV, or NTB. However, HLA-B*51 carriers had earlier onset of uveitis and HLA-A*26:01 was strongly associated with posterior uveitis. In patients with posterior uveitis, VI was associated with a longer duration of uveitis and cataract, LUV with a longer duration, and NTB with HLA-B*51.nnnCONCLUSIONnLonger duration of uveitis, posterior uveitis, male gender, cataract, and glaucoma were found to be associated with poor visual outcome in BD-related uveitis. HLA-B*51 was associated with NTB in patients with posterior uveitis. HLA-A*26:01 showed no association with VI, LUV, or NTB, however, was strongly associated with posterior uveitis.

Clinical factors and treatment outcomes associated with failure in the detection of urate crystal in patients with acute gouty arthritis.

Background/Aims To investigate the rate of detection of monosodium urate (MSU) crystals in the synovial fluid (SF) of patients with acute gouty arthritis and factors associated with false-negative results. Methods A total of 179 patients with acute gouty arthritis who had undergone SF crystal examination were identified from the data warehouse of two university hospitals. Clinical and laboratory data were obtained from the medical records. Results The overall rate of detection of MSU crystals was 78.8%. In univariate analyses, the only significant differences between the variables of crystal-negative and crystal-positive patients were a lower C-reactive protein level (p = 0.040) and fewer patients undergoing emergent surgery in the crystal-positive group (p = 4.5 × 10-6). In logistic regression analyses, MSU crystal-negative results were significantly associated with the interval from arthritis onset to crystal examination (p = 0.042), and this was the most significant risk factor for arthroscopic surgery (p = 2.1 × 10-4). Seventeen patients who underwent arthroscopic surgery had a significantly longer hospital stay (p = 0.007) and a significant delay in gout treatment (p = 8.74 × 10-5). The distribution of crystal-negative patients differed significantly between the SF samples that were evaluated by both the laboratory medicine and the rheumatology departments (p = 1.2 × 10-14), and the κ value was 0.108. Conclusions Although several clinical features were associated with detection failure, SF MSU crystal identification was critically dependent on the observer. Considering the impact on the treatment outcomes, implementation of a quality control program is essential.

Impact of Dose Tapering of Tumor Necrosis Factor Inhibitor on Radiographic Progression in Ankylosing Spondylitis.

Objective To investigate the impact of dose reduction of tumor necrosis factor inhibitor (TNFi) on radiographic progression in ankylosing spondylitis (AS). Methods One hundred and sixty-five patients treated with etanercept or adalimumab were selected from a consecutive single-center observational cohort based on the availability of radiographs at baseline and after two- and/or four-years of follow up. Radiographs were assessed by two blinded readers using the modified Stokes AS Spinal Score (mSASSS). Radiographic progression in patients treated with standard-dose TNFi (standard-dose group, n = 49) was compared with patients whose dosage was tapered during the treatment (tapering group, n = 116) using linear mixed models. Results Baseline characteristics between two groups were comparable except for higher BASDAI (7.1 vs. 6.3, p = 0.003) in the standard-dose group. At two years after the treatment, mean dose quotient (S.D.) of the tapering group was 0.59 (0.17). During follow up, rate of radiographic progression in overall patients was 0.90 mSASSS units/year. Radiographic progression over time between the two groups was similar at the entire group level. However, in the subgroup of patients with baseline syndesmophytes, progression occurred significantly faster in the tapering group after the adjustment for baseline status (1.23 vs. 1.72 mSASSS units/year, p = 0.023). Results were consistent when radiographic progression was assessed by the number of newly developed syndesmophytes (0.52 vs. 0.73/year, p = 0.047). Sensitivity analysis after multiple imputation of missing radiographs also showed similar results. Conclusion A dose tapering strategy of TNFi is associated with more rapid radiographic progression in AS patients who have syndesmophytes at baseline.

OP0077 Primary prophylaxis for pneumocystis pneumonia in patients with rheumatic disease and treated with prolonged, high-dose steroid: a retrospective cohort study with 12-year observation

Background Pneumocystis pneumonia (PCP) is a significant cause of morbidity and mortality in patients with rheumatic diseases, especially in the case of patients receiving high-dose steroid treatment. Objectives To investigate the efficacy and safety of PCP prophylaxis using trimethoprim/sulfamethoxazole (TMP-SMX) in patients with rheumatic disease receiving prolonged, high-dose steroids Methods This study includes 1522 cases of prolonged (≥4 weeks), high-dose (≥30mg/day prednisone or equivalent) steroid treatment from 1092 patients with any rheumatic diseases during a 12-year period in a single tertiary referral center in South Korea. Of them, prophylactic TMP-SMX was administered in 262 cases (prophylaxis group) with a mean (SD) duration of 229.0 (272.7) days whereas other 1260 cases received no prophylaxis (control group). Primary outcome was 1-year incidence of PCP between the two groups. Secondary outcomes included PCP-related mortality, ICU admission rate, all-cause in-hospital mortality and incidence of any adverse drug reactions (ADRs) of TMP-SMX. To minimize the baseline imbalance between the two groups, we introduced propensity-score matching and performed the same analysis in the post-matched population as a sensitivity analysis. Results Patients in the prophylaxis group were treated more often with secondary immunosuppressive drugs and had a higher proportion of patients with PCP high-risk diseases (ANCA-associated vasculitis and dermatomyositis) and lymphopenia at baseline. Overall, 30 cases of PCP occurred and resulted in death in 11 cases (36.7%). In the prophylaxis group, only one non-fatal case of PCP occurred. One-year PCP incidence was significantly lower in the prophylaxis group (adjusted HRs=0.096 [0.013–0.719]) (Figure). TMP-SMX also significantly reduced the PCP-related mortality (adjusted HR=0.101 [0.001–0.809]) whereas ICU admission rate and in-hospital mortality rate were not different between the two groups during the observation. This result was consistent in the sensitivity analysis where same analysis was performed in the post-matched population. Thirty-four cases of ADRs of TMP-SMX occurred, with an incidence rate (95% CI) of 24.2 (17.3–33.0) per 100 person-year. There were two cases of serious ADR (one pancytopenia and one Stevens Johnson syndrome) but they all recovered shortly after the discontinuation of TMP-SMX. The number needed to harm (NNH) of serious ADR was 109 whereas the number needed to treat (NNT) to prevent one case of PCP in the whole population was 52. Conclusions In patients with rheumatic disease receiving prolonged, high-dose steroid treatment, TMP-SMX prophylaxis significantly lower the incidence of PCP with favorable safety. References Stern A, Green H, Paul M, Vidal L, Leibovici L. Prophylaxis for Pneumocystis pneumonia (PCP) in non-HIV immunocompromised patients. The Cochrane database of systematic reviews. 2014(10):CD005590. Disclosure of Interest None declared

AB0317 Drug Survival of Biologic Agents in Rheumatoid Arthritis Using 10 Years of Data from The National Health Insurance in Korea: A Population-Based Study and Comparison with A Single-Institution Cohort

Background Since 2004, Korean national health insurance began to cover the cost of biologic agents in rheumatoid arthritis (RA) patients experiencing treatment failure by conventional disease modifying antirheumatic agents. However, real-world issues such as selecting or switching among biologic agents in clinical practice became more complicated as types of the agents increased. Objectives To evaluate 1) first choice among biologic agents, 2) drug retention rate and duration and 3) switching rate and duration before switching of each biologic agent in RA. Methods We used prospective cohort data by the National Health Insurance Service (NHIS) in Korea, which consisted of more than one million subjects representing age, sex and income status of whole Korean population followed from 2004 to 2013. All RA patients with any experience of biologic agents were checked by the code of each drug and KCD9 code of diagnosis including all types of RA. To supplement the results of the population based study, we also compared and validated with the independent cohort of biologic agent users in Seoul national university hospital (SNUH). Results One hundred and seventy three patients were found to experience any of TNFa inhibitors from Jan. 2004 to Dec. 2013. Etanercept was the most frequent choice as first TNFa inhibitor (43.4%), followed by adalimumab (35.3%), infliximab (20.2%), golimumab (0.6%) and abatacept (0.6%). Among TNFa inhibitors, retention rates of etanercept (73.1%, 62.7%) at month 12 and 24 were significantly higher (p=0.039, both) than adalimumab (57.7%, 44.2%) and infliximab (45.5%, 36.4%). In SNUH cohort, etanercept (79.8%, 67.0%) also had better retention rate than adalimumab (51.3%, 48.7%) and infliximab (58.6%, 34.5%) at month 12 and 24 respectively. Mean duration of retention was almost twice longer in etanercept (3.42 years) than adalimumab (1.83 years, p=0.004) and infliximab (1.71 years, p=0.011). Similarly, etanercept (3.24 years) prescribed longer than adalimumab (1.98 years) and infliximab (1.63 years) with statistical significance in SNUH cohort. Etanercept also had lower switching rate (20.9%) than adalimumab (32.1%) and infliximab (38.5%, p=0.040) and the tendency were similar in SNUH cohort. Preferred biologic agents after switching were adalimumab (27.4%) and rituximab (24.2%). Etanercept users (1.35 years) had longer duration of use before switching than infliximab (0.60 years, p=0.050) but no significant difference was shown comparing with adalimumab (0.73years, p=0.483). Conclusions As golimumab, abatacept, tocilizumab were available relatively later than existing TNFa inhibitors and rituximab, more follow up period would be required to evaluate their value in treating RA. This research can be a pilot study before analyzing nationwide population cohort encompassing whole population of Korea. Conclusively, etanercept has been prescribed longer in general and the rate of retention till 2 years was also better then adalimumab and infliximab for 10 years after national insurance coverage. Disclosure of Interest None declared

SAT0451 The HLA Class II Profile in Korean Patients with Inflammatory Myositis

Background Polymyositis (PM) and dermatomyositis (DM) is known to be associated with certain HLA alleles in several ethnic groups including Caucasians, African-Americans, and the Japanese. Objectives To investigate the profile of HLA class II alleles in Korean patients with PM and DM Methods Sequencing based genotyping for HLA-DRB1 and -DPB1 was performed in 140 Korean patients with DM (n=104) or PM (n=36) and in 207 healthy controls. Associations of each HLA-DR or -DP allele with myositis or clinical subsets of myositis were examined. Results HLA-DRB1*14:54 (odds ratio [95% confidence interval] 1.05 [1.01-1.09]), -DRB1*15:02 (2.44 [1.03-5.81]), -DRB1*16:02 (1.04 [1.00-1.07], and -DPB1*09:01 (2.44 [1.03-5.81]) were found to be susceptibility alleles while HLA-DRB1*14:01 (0.93 [0.90-0.97]) and -DRB1*15:01 (0.38 [0.18-0.79]) to be protective alleles for inflammatory myositis (Table 1). In DM, HLA-DRB1*15:02 (3.14 [1.30-7.62]) and -DPB1*09:01 (2.87 [1.17-7.05]) carriers were more frequent while HLA-DRB1*14:01 (0.93 [0.90-0.97]), -DRB1*15:01 (0.30 [0.12-0.74]), and -DPB1*04:02 (0.44 [0.20-0.94]) carriers were less frequent than controls. In PM, HLA-DRB1*14:03 (6.52 [1.78-23.81]), and -DRB1*14:54 (1.13 [1.00-1.23]) carriers were more frequently observed. When PM and DM were compared, HLA-DRB1*14:03 carriers were more frequently observed in PM (5.43 [1.23-24.02]). When the association between each allele and clinical subsets was examined, patients with interstitial lung disease (ILD) carried HLA-DRβ1*04:03 (6.34 [1.35-29.76]), -DRβ1*07:01 (0.22 [0.08-0.64]), -DRβ1*09:01 (0.16 [0.05-0.60]) -DRβ1*12:02 (2.91 [1.16-8.01]) DRβ1*15:01 (0.10 [0.01-0.80]), -DPβ1*14:01 (8.90 [1.08-73.21]), and -DPβ1*17:01 (0.11 [0.01-0.92]) more frequently than patients without. DM patients with ILD showed higher frequency of HLA-DRβ1*04:05 than DM patients without ILD (4.68 [1.24-17.74]). Other associations included dysphagia with HLA-DRβ1*08:02 (6.73 [1.51-30.00]), -DRβ1*14:54 (5.19 [1.09-24.57]), and -DPβ1*03:01 (5.19 [1.09-24.57]), arthralgia with HLA-DRβ1*03:01 (4.79 [1.21-18.98]) and -DPβ1*02:01 (2.30 [1.10-4.82]), and mechanics hand with HLA-DRB1*08:03 (3.25 [1.12-9.45]) and -DRB1*12:02 (5.04 [1.66-15.29]). Regarding skin rashes, heliotrope rash was associated with HLA-DRβ1*09:01 (2.98 [1.05-8.46]) and -DPβ1*05:01 (0.41 [0.18-0.87]), Gottrons papules with HLA-DRβ1*14:05 (8.77 [1.07-72.17]), -DPβ1*02:01 (2.29 [1.11-4.73]), -DPB1*02:02 (0.086 [0.01-0.67]), and -DPB1*05:01 (0.41 [0.20-0.85]), and shawl sign with HLA-DRB1*14:05 (4.67 [1.15-18.99]) and -DPB1*14:01 (4.67 [1.15-18.99]). Cancer development was associated with HLA-DRB1*12:01 (7.25 [1.65-31.89]) and -DRB1*13:02 (0.81 [0.74-0.88]). No HLA-DRB1 or -DPB1 allele was found to be associated with anti-Jo1 antibody. Conclusions Korean patients with inflammatory myositis showed a unique profile of HLA-DR and -DP alleles compared with other ethnic groups. Alleles associated with disease susceptibility were different between PM and DM. Clinical subsets, particularly ILD, showed different immunogenetic backgrounds. Disclosure of Interest None declared

THU0211 Low Dose Etanercept Treatment for Maintenance of Clinical Remission in Ankylosing Spondylitis: Retrospective Cohort Study

Background Dose reduction of etanercept after clinical remission in patients with ankylosing spondylitis (AS) is not uncommon practice. However, efficacy and optimal schedule of dose tapering is rarely evaluated. Objectives To investigate the efficacy and safety of low dose etanercept treatment (25mg or less/week) after clinical remission of AS in real world. Methods In this study, 134 AS patients who treated with etanercept for at least 12 months and achieved clinical remission (BASDAI <4 and normal C-reactive protein level) between 2004 and 2013 were enrolled. Dose reduction was performed in 100 patients (low dose group) after achieving clinical remission based on physicians medical decision. Other 34 patients (standard dose group) maintained the dosage of 50mg/week until discontinuation. Drug survival (time to drug discontinuation due to inefficacy or adverse event) and incidence of significant adverse events were compared between two groups. In patients with low dose group, clinical factors and time to dose reduction associated with longer drug survival were estimated. Results Clinical and demographic features were comparable between two groups at the time of starting etanercept except for age (43.4 years in low dose group vs. 52.0 years in standard dose group, p=0.001). In low dose group, median time to dose reduction was 19.5 (10.1-39.6) weeks. During 536.8 person-years (PYs) of follow up, 27 patients stopped etanercept (21/440.9PYs in low dose group and 6/95.9PYs in standard dose group). Crude drug survival of low dose group was not significantly different from that of standard group (98.0% vs. 93.5% at 2 years, 92.1% vs. 89.4% at 3 years and 84.4% vs. 76.2% at 4 years). This finding was consistent after adjustment of clinical factors including age, gender, disease duration, initial BASDAI, concomitant MTX and previous TNF-blocker use (adjusted HR=0.552, 95% C.I. 0.208-1.465) (Figure). Incidence of adverse events which led to discontinuation of etanercept also showed no difference between two groups (40.82/1000PYs in low dose group vs. 52.16/1000PYs in standard dose group, p=0.783). In the subgroup analysis with low dose group, dose reduction after more than 24 weeks of standard dose treatment was significantly associated with longer drug survival after adjustment of other clinical factors (adjusted HR 0.273, 95% C.I. 0.085-0.875). Conclusions In patients with AS who achieved clinical remission, low dose etanercept treatment showed comparable long-term efficacy and safety in real world. Among various strategies of dose reduction, more than 24 weeks of standard dose treatment before adjustment was associated with longer drug survival. Disclosure of Interest None declared