Association of HLA-DRB1 locus with treatment response to abatacept or TNF inhibitors in patients with seropositive rheumatoid arthritis.

The strongest genetic risk factor for rheumatoid arthritis (RA) has been known as HLA-DRB1 based on amino acid positions 11, 71, and 74. This study analyzed the association between specific HLA-DRB1 locus and treatment response to abatacept or TNF inhibitors (TNFi) in patients with seropositive RA. A total of 374 Korean RA patients were treated with abatacept (n = 110) or TNFi (n = 264). Associations between HLA-DRB1 and treatment response after 6 months were analyzed using multivariable logistic regression. Seropositive RA patients with HLA-DRB1 shared epitope (SE) had a favorable response to abatacept (OR = 3.67, P = 0.067) and an inversely associated response to TNFi (OR 0.57, P = 0.058) based on EULAR response criteria, but the difference was not statistically significant in comparison to those without SE. In analyses using amino acid positions of HLA-DRB1, a significant association was found between valine at amino acid position 11 of SE and good response to abatacept (OR = 6.46, P = 5.4 × 10-3). The VRA haplotype also showed a good response to abatacept (OR = 4.56, P = 0.013), but not to TNFi. Our results suggest that treatment response to abatacept or TNFi may differ depending on HLA-DRB1 locus in seropositive RA, providing valuable insights for selecting optimal therapy.

Transcriptomic network analysis reveals key drivers of response to anti-TNF biologics in patients with rheumatoid arthritis.

OBJECTIVE: Anti-TNF biologics have been widely used to ameliorate disease activity in patients with rheumatoid arthritis (RA). However, a large fraction of patients show a poor response to these agents. Moreover, no clinically applicable predictive biomarkers have been established. This study aimed to identify response-associated biomarkers using longitudinal transcriptomic data in two independent RA cohorts. METHODS: RNA sequencing data from peripheral blood cell samples of Korean and Caucasian RA cohorts before and after initial treatment with anti-TNF biologics were analyzed to assess treatment-induced expression changes that differed between highly reliable excellent and null responders. Weighted correlation network, immune cell composition, and key driver analyses were performed to understand response-associated transcriptomic networks and cell types and their correlation with disease activity indices. RESULTS: In total, 305 response-associated genes showed significantly different treatment-induced expression changes between excellent and null responders. Co-expression network construction and subsequent key driver analysis revealed that 41 response-associated genes played a crucial role as key drivers of transcriptomic alteration in four response-associated networks involved in various immune pathways: type I interferon signalling, myeloid leucocyte activation, B cell activation, and NK cell/lymphocyte-mediated cytotoxicity. Transcriptomic response scores that we developed to estimate the individual-level degree of expression changes in the response-associated key driver genes were significantly correlated with the changes in clinical indices in independent patients with moderate or ambiguous response outcomes. CONCLUSIONS: This study provides response-specific treatment-induced transcriptomic signatures by comparing the transcriptomic landscape between patients with excellent and null responses to anti-TNF drugs at both gene and network levels.

Pregnancy Outcomes Associated With Biologic Agent Exposure in Patients With Several Rheumatic Diseases and Inflammatory Bowel Diseases.

BACKGROUND: This study aimed to analyze pregnancy outcomes based on biologic agents use in women using the nationwide population-based database. METHODS: The study used the claims database to identify women of childbearing age with several rheumatic (rheumatoid arthritis, juvenile idiopathic arthritis, ankylosing spondylitis, psoriatic arthritis) and inflammatory bowel diseases (Crohn's disease and ulcerative colitis) who had pregnancy-related codes between January 2010 and December 2019. We analyzed live births and adverse pregnancy outcomes based on the previous use of biologics. We also stratified the patients according to duration of biologic agent exposure before pregnancy and the use of biologics during pregnancy to analyze the pregnancy outcomes by subgroups. RESULTS: We identified 4,787 patients with pregnancy events. Among them, 1,034 (21.6%) used biologics before pregnancy. Live birth rate was not different between the biologics group and biologics naïve group (75.0% vs. 75.2%). Multivariate analyses showed that biologics use was associated with higher risk of intrauterine growth retardation (odds ratio [OR], 1.780) and lower risk of gestational diabetes mellitus (OR, 0.776) compared with biologics naïve. Biologics use during pregnancy was associated with higher risk of preterm delivery (OR, 1.859), preeclampsia/eclampsia (OR, 1.762), intrauterine growth retardation (OR, 3.487), and cesarean section (OR, 1.831), but lower risk of fetal loss (OR, 0.274) compared with biologics naïve. CONCLUSIONS: Although there was no difference in live birth rate between the biologics group and biologics naïve group, biologics use seems to be associated with several adverse pregnancy outcomes, especially in patients with biologics during pregnancy. Therefore, patients with biologics during pregnancy need to be carefully observed for adverse pregnancy outcomes.

Higher Genetic Risk Loads Confer More Diverse Manifestations and Higher Risk of Lupus Nephritis in Systemic Lupus Erythematosus.

OBJECTIVE: Systemic lupus erythematosus (SLE) is a highly heritable complex disorder with heterogeneous clinical manifestations. In this study, we aimed to identify the genetic risk load using clinical and serological manifestations in SLE patients. METHODS: We genotyped a total of 1,655 Korean patients with SLE (n = 1,243 as a discovery set and n = 412 as a replication set) using a customized genome-wide single-nucleotide polymorphism (SNP) array, KoreanChip. A weighted genetic risk score (wGRS) for an individual was calculated from 112 well-validated non-HLA SNPs and HLA haplotypes of SLE-risk loci. We analyzed associations between individual wGRS and clinical SLE subphenotypes and autoantibodies using multivariable linear or logistic regression adjusted by onset age, sex, and disease duration. RESULTS: Childhood-onset SLE (<16 years) conferred the highest genetic risk compared with adult-onset (16-50 years) or late-onset (>50 years) SLE (P = 6.8 × 10-6 ). High wGRS significantly increased associations with SLE manifestations, regardless of onset age, sex, and disease duration. Individual wGRS significantly correlated positively with more clinical American College of Rheumatology criteria (ß = 0.143, P = 1.8 × 10-6 ). Subphenotype analysis revealed significant associations between the highest and lowest wGRS quartile with risk of renal disorder (hazard ratio [HR] 1.74, P = 2.2 × 10-8 ) and anti-Sm antibody production (HR 1.85, P = 2.8 × 10-5 ). Higher wGRS markedly modulated the pathogenesis of proliferative and membranous lupus nephritis class III or IV (HR 1.98, P = 1.6 × 10-5 ) and class V (HR 2.79, P = 1.0 × 10-3 ), but especially lupus nephritis class V in anti-Sm-positive SLE (area under the curve 0.68, P = 1.8 × 10-4 ). CONCLUSION: Patients with SLE and high wGRS tended to have earlier age of SLE onset, higher anti-Sm antibody positivity, and more diverse clinical phenotypes. Genetic profiling may predict high risk for lupus nephritis and a diverse clinical course in SLE patients.

MUC5B promoter variant rs35705950, rare but significant susceptibility locus in rheumatoid arthritis-interstitial lung disease with usual interstitial pneumonia in Asian populations.

BACKGROUND: MUC5B variant rs35705950 is the common and most significant risk variant for rheumatoid arthritis-interstitial lung disease (RA-ILD) in Western populations. However, little is known about its significant association with RA-ILD in Asian populations. We here investigate the association of rs35705950 with Korean patients with RA-ILD. METHODS: In this cross-sectional study, we genotyped rs35705950 in 2444 patients with RA. Among them, 683 patients with RA who have chest CT were divided into RA-ILD and RA-noILD. RA-ILD was classified as usual interstitial pneumonia (UIP) and other than UIP. The associations of rs35705950 with RA-ILD and its subtype were analysed using multivariable regression adjusted for age at RA diagnosis. Meta-analysis of a previously reported Japanese dataset and Korean dataset obtained for this study was conducted. RESULTS: The minor allele (T) frequency of rs35705950 was 0.37%, 1.43% and 2.38% in 2444 patients with RA, 105 patients with RA-ILD and 63 patients with UIP, respectively. Genotypic association of rs35705950 with RA-ILD was insignificant (OR 2.49, 95% CI 0.64 to 9.69, p=0.187), but showed significant association with UIP (OR 4.90, 95% CI 1.23 to 19.59, p=0.024) compared with RA-noILD. In meta-analysis (123 UIP and 878 RA-noILD) combining our data with previously reported Japanese data, this variant was found to be significantly associated with UIP (OR 3.51, 95% CI 1.19 to 10.37, p=0.023). CONCLUSION: MUC5B variant rs35705950 is a rare but significant risk factor for Asian patients with RA-ILD with UIP, suggesting a sharing of the genetic background between Asian and Western populations.

Effect of resistance exercise on serum leptin levels in a prospective longitudinal study of women patients with rheumatoid arthritis.

BACKGROUND: Exercise has an anti-inflammatory effect and reduces fat mass. Leptin has been known to be proinflammatory adipokines mainly produced by adipocytes. However, few studies have investigated the association between exercise and changes in serum leptin levels of patients with RA. This study evaluated the effect of an individualized resistance exercise on inflammatory markers including leptin as well as muscle strength and exercise capacity in patients with rheumatoid arthritis (RA). METHODS: A total of 42 age- and sex-matched participants were assigned to a resistance exercise program (60 min, once a week for 12 weeks, and self-exercise twice a week) or to a control group. Muscle strength, exercise capacities, and inflammatory markers such as cytokines and adipokines were assessed at baseline and at 12 weeks follow-up. Longitudinal changes in muscle strength, exercise capacity, cytokines, and adipokines between groups were tested with repeated measures analysis of variance or using the generalized estimating equation, with adjustment for baseline disease activity score 28-C response protein as a covariate. RESULTS: A total of 37 of 42 female patients with RA completed this prospective intervention study. Grip strength improved significantly in the exercise group (P < 0.05), while no between-group changes were found. Quadriceps contraction power (P for group-time interaction = 0.035 for the right side and P for group-time interaction = 0.012 for the left side) and 6-minute walking distance (P for group-time interaction = 0.021) were all improved significantly in the exercise group compared with the control group. In addition, serum leptin levels were significantly decreased in the exercise group compared with the control group (P for group-time interaction = 5.22 × 10-5), but not the other cytokines or adipokines. The change in serum leptin levels correlated with the changes in fat mass (Rho = 0.491, P= 0.015) and visceral fat area (Rho = 0.501, P= 0.013). CONCLUSION: In addition to muscle strength and exercise capacity, the 12 weeks of individualized resistance exercise reduced serum leptin levels in keeping with body fat mass or visceral fat area, suggesting that serum leptin levels might be a surrogate marker of exercise in RA.

Use of Disease-modifying Antirheumatic Drugs After Cancer Diagnosis in Rheumatoid Arthritis Patients.

Objective: There is no recommendation for the use of disease-modifying antirheumatic drugs (DMARDs) in patients with rheumatoid arthritis (RA) who developed cancer. We examined changes in the DMARDs prescription patterns associated with cancer diagnosis in RA patients. Methods: We reviewed the medical records of 2,161 RA patients who visited rheumatology clinic between January 2008 and February 2017 and found 40 patients who developed cancer during RA treatment. In these patients, we examined DMARDs prescription patterns before and right after cancer diagnosis and at recent outpatient clinic visits. Results: Before cancer diagnosis, methotrexate (MTX)-combined conventional synthetic DMARDs (csDMARDs) were most commonly prescribed (22, 55.0%) and biological DMARDs (biologics) in nine patients (22.5%). For cancer treatment, 19 patients received chemotherapy (including adjuvant chemotherapy) and 21 patients had surgery only. Right after cancer diagnosis, changes in the DMARDs prescription patterns were similar in discontinuation (13, 32.5%), switching (14, 35.0%), and maintenance (13, 32.5%). DMARDs were discontinued more frequently in the chemotherapy group (9/19, 47.4%) than the surgery only group (4/2, 19.0%) (p<0.05). Among the 13 patients who discontinued DMARDs, nine (69.2%) resumed DMARDs after a median of 5.5 months (interquartile range [IQR] 2.9, 18.3) due to arthritis flare. At a median of 4.6 years (IQR 3.3, 6.7) after cancer diagnosis, 25 patients were evaluated at recent outpatient clinic visits. Four patients received no DMARD, three MTX monotherapies, 11 csDMARDs combination therapies, and seven biologics. Conclusion: A significant number of RA patients who developed cancer during RA treatment were still receiving DMARDs including biologics after cancer diagnosis.

Predictive Factors for Renal Response in Lupus Nephritis: A Single-center Prospective Cohort Study.

Objective: To identify the predictive factors for renal response in patients with lupus nephritis (LN). Methods: Patients and data were extracted from a prospective systemic lupus erythematosus cohort in Korea, in which clinical data were collected at 0, 3, 6, and 12 months after induction therapy. Treatment response of LN were evaluated as a complete response (CR), partial response (PR), or non-response (NR) at 3, 6, and 12 months, respectively. Predictive factors for CR at 6 months were evaluated using multivariable Poisson regression analysis. Results: A total of 75 patients with LN who underwent biopsy was enrolled. The mean age at diagnosis of LN was 28.9±9.7 years, and 68 (90.7%) were female. The frequencies of classes III, IV, III+V, IV+V, and V were 20.0%, 44.0%, 16.0%, 12.0%, and 8.0%, respectively. Compared to relapsed LN, new-onset LN showed a lower percentage of glomerulosclerosis (45.5% vs. 76.2%, p=0.013). The overall proportions of CR, PR, and NR at 6 and 12 months were 52.0%, 26.7%, 21.3% and 50.7%, 24.0%, 25.3%, respectively. In multivariate analysis, age at enrollment (odds ratio [OR]=1.02, p=0.022), relapsed LN (OR=0.71, p=0.037), anti-Ro antibody (OR=0.67, p=0.014), and class III LN (OR=1.48, p=0.001) were associated with CR at 6 months. Conclusion: In our prospective cohort, class III LN was a good predictive factor for CR at 6 months in patients with LN, whereas younger age, relapsed LN, and anti-Ro antibody were poor predictive factors.

Influenza infection as a trigger for systemic lupus erythematosus flares resulting in hospitalization.

In patients with systemic lupus erythematosus (SLE), there are concerns that infections may increase the risk of flares. We evaluated the association between influenza infection and SLE flares resulting in hospitalization. SLE flares resulting in hospitalization and influenza cases were ascertained from the Korean national healthcare insurance database (2014-2018). We used a self-controlled case series design. We defined the risk interval as the first 7 days after the influenza index date and the control interval was defined as all other times during the observation period of each year. We estimated the incidence rates of SLE flares resulting in hospitalization during the risk interval and control interval and compared them using a Poisson regression model. We identified 1624 influenza infections among the 1455 patients with SLE. Among those, there were 98 flares in 79 patients with SLE. The incidence ratio (IR) for flares during the risk interval as compared with the control interval was 25.75 (95% confidence interval 17.63-37.59). This significantly increased the IRs for flares during the risk interval in both women (IR 27.65) and men (IR 15.30), all age groups (IR 17.00-37.84), with and without immunosuppressive agent (IR 24.29 and 28.45, respectively), and with and without prior respiratory diseases (IR 21.86 and 26.82, respectively). We found significant association between influenza infection and SLE flares resulting in hospitalization. Influenza infection has to be considered as a risk factor for flares in all SLE patients regardless of age, sex, medications, and comorbidities.

Risk factors for cervical spine instability in patients with rheumatoid arthritis.

INTRODUCTION: Cervical spine (C-spine) instability is a unique and significant characteristic of rheumatoid arthritis (RA) because its occurrence is not rare and it can cause compressive cervical myelopathy, which may lead to serious neurologic sequelae. This study evaluated the prevalence and risk factors of C-spine instabilities in RA patients with a focus on anti-citrullinated protein antibody (ACPA) and biologic disease-modifying antirheumatic drug (DMARD) therapies. METHODS: The presence of C-spine instabilities in 1114 patients with RA was evaluated using C-spine radiographies according to the defined metrics. Multivariable logistic regression analyses were performed to identify independent predictors of C-spine instability. The initiation of biologic DMARDs was assessed via a Kaplan-Meier analysis and compared using log-rank tests. RESULTS: In total, 306 (27.5%) patients presented with C-spine instabilities. The most common type was atlantoaxial subluxation (AAS; n = 199 [17.9%]). Male sex, positivity for rheumatoid factor and ACPA, erosive change in the peripheral joints, and presence of osteoporosis were independently associated with C-spine instabilities (all P < 0.05). In particular, positivity for ACPA was the most powerful risk factor (odds ratio: 2.33 [95% confidence interval: 1.37, 3.96], P = 0.002), and it was closely associated with AAS. Patients with AAS were at a higher risk for early initiation of biologic DMARDs. CONCLUSIONS: Positivity for ACPA was a significant risk factor for C-spine instability, and AAS was remarkably correlated to the early initiation of biologic DMARDs, a surrogate index of poor long-term outcomes. Key Points • The presence of antibodies against citrullinated proteins was a strong risk factor for C-spine instability in patients with rheumatoid arthritis. • Atlantoaxial subluxation was significantly associated with early initiation of biologic DMARDs, a surrogate index of poor long-term outcome.

Novel classification of axial spondyloarthritis to predict radiographic progression using machine learning.

OBJECTIVES: Prediction and determination of drug efficacy for radiographic progression is limited by the heterogeneity inherent in axial spondyloarthritis (axSpA). We investigated whether unbiased clustering analysis of phenotypic data can lead to coherent subgroups of axSpA patients with a distinct risk of radiographic progression. METHODS: A group of 412 patients with axSpA was clustered in an unbiased way using a agglomerative hierarchical clustering method, based on their phenotype mapping. We used a generalised linear model, naïve Bayes, Decision Trees, K-Nearest-Neighbors, and Support Vector Machines to construct a consensus classification method. Radiographic progression over 2 years was assessed using the modified Stoke Ankylosing Spondylitis Spine Score (mSASSS). RESULTS: axSpA patients were classified into three distinct subgroups with distinct clinical characteristics. Sex, smoking, HLA-B27, baseline mSASSS, uveitis, and peripheral arthritis were the key features that were found to stratifying the phenogroups. The three phenogroups showed distinct differences in radiographic progression rate (p<0.05) and the proportion of progressors (p<0.001). Phenogroup 2, consisting of male smokers, had the worst radiographic progression, while phenogroup 3, exclusively suffering from uveitis, showed the least radiographic progression. The axSpA phenogroup classification, including its ability to stratify risk, was successfully replicated in an independent validation group. CONCLUSIONS: Phenotype mapping results in a clinically relevant classification of axSpA that is applicable for risk stratification. Novel coupling between phenotypic features and radiographic progression can provide a glimpse into the mechanisms underlying divergent and shared features of axSpA.

Pregnancy rates and perinatal outcomes in women with systemic lupus erythematosus: data from the Korean national health claims database.

INTRODUCTION/OBJECTIVES: The pregnancy rate in systemic lupus erythematosus (SLE) is not fully understood and comparisons of adverse pregnancy outcomes (APOs) with SLE versus the general population are limited. This study aimed to estimate the pregnancy rate and APOs in Korean SLE compared to those without SLE. METHOD: Pregnant women were identified using the ICD-10 codes for delivery and abortion in the Korean national health claims database (2013-2015). APOs were classified as fetal loss, intrauterine growth retardation (IUGR), pre-eclampsia/eclampsia, and gestational diabetes. Annual incidence rates (IRs) of pregnancy and APOs were calculated in women with SLE and the general population without SLE and the two groups were compared using age-adjusted incidence rate ratios (IRRs). Age-stratified IRRs were further analyzed. RESULTS: The annual IRs of pregnancy in SLE were 29.54-30.70 per 1000 persons. The IRRs were lower in women with SLE than in the general population: 0.68 (0.61-0.76), 0.66 (0.60-0.74), and 0.74 (0.66-0.82) in each respective year. The IRRs of fetal loss, IUGR, and pre-eclampsia/eclampsia were 1.30 (1.14-1.49), 4.65 (3.55-6.09), and 3.43 (2.70-4.36), respectively. However, the IRR of gestational diabetes in SLE did not significantly differ from that of women without SLE. Among the APOs, fetal loss, IUGR, and pre-eclampsia/eclampsia showed decreasing tendencies as age increased. CONCLUSIONS: Pregnancy rates in SLE were approximately 30% lower than those in the general population. Except for gestational diabetes, fetal loss, IUGR, and pre-eclampsia/eclampsia were higher in SLE and showed a decreasing tendency with age. Key Points • This population-based cohort study showed that pregnancy rates in SLE were approximately 30% lower than those in the general population. • SLE had a 1.3-fold higher rate of fetal loss, more than 4-fold higher IUGR rate, and more than 3-fold pre-eclampsia or eclampsia rate compared with the general population. • Adverse pregnancy outcomes in SLE showed a decreasing tendency with age.

Association of influenza infection with hospitalisation-related systemic lupus erythematosus flares: a time series analysis.

OBJECTIVES: This study aimed to investigate whether the influenza annual outbreak in Korea is related to hospitalisation-related flares in systemic lupus erythematosus (SLE) patients. METHODS: The weekly frequency of hospitalisation-related SLE flares (2012-2015) was collected from the Korean National Health Insurance claim database. The weekly laboratory-confirmed detection rate of influenza infection was obtained from the Korea Centers for Disease Control and Prevention database. A generalised linear model was used to examine the relative risks (RRs) of hospitalisation-related SLE flares associated with influenza infection, after adjusting for time trends and meteorological data. RESULTS: A total of 2,223 hospitalisation-related SLE flares were analysed. An interquartile range (24.5%) increase in influenza infection was associated with a 14.0% increase in hospitalisation-related SLE flares (RR, 1.14; 95% confidence interval [CI]: 1.04-1.25; p=0.006). In addition, influenza infections at lag 0-1 (over 2 weeks including concurrent and 1 previous week) and lag 0-2 (over 3 weeks including concurrent and 2 previous weeks) were associated with increase in hospitalisation-related SLE flares (RR, 1.14; 95% confidence interval [CI]: 1.03-1.26; p=0.014 and RR, 1.13; 95% CI: 1.02-1.26; p=0.023). Significant associations were especially observed in women (RR, 1.15; 95% CI: 1.15-1.16; p=0.006) and immunosuppressant (RR, 1.26; 95% CI: 1.26-1.27; p<0.001) or glucocorticoid recipients (RR, 1.17, 95% CI: 1.16-1.17; p=0.004). CONCLUSIONS: This study shows a significant association between seasonal influenza infection and flares in SLE patients, which suggests influenza can be a novel environmental risk factor for SLE flares.

Twelve Weeks of Strengthening Exercise for Patients with Rheumatoid Arthritis: A Prospective Intervention Study.

Rheumatoid arthritis (RA) patients may benefit from exercise for several reasons. However, whole-limb strengthening exercises for such patients remain poorly studied. We hypothesized that systemic strength training that includes the upper and lower extremities would improve strength per se and enhance the quality of life. Here, we investigated the effects of 12 weeks of upper- and lower-limb strengthening exercise on the strength and quality of life of RA patients using the International Classification of Functioning, Disability, and Health model. This was a prospective, interventional controlled trial. Forty female RA patients were recruited and assigned to two groups not based on willingness to exercise, with 20 patients in the exercise group and 20 in the control group. All patients in the exercise group received once-weekly training sessions of 60 min over 12 weeks. All participants were assessed before and after the 12-week intervention period. We measured the hand grip strength and isometric quadriceps contraction, the cross-sectional area of the rectus femoris (CSA-RF) (via ultrasonography), and performed the 30 s sit-to-stand test and the 6 min walk test (6MWT). We derived the Borg scale score after the 6MWT and assessed the extent of social participation and quality of life using a Korean version of the 36-Item Short Form Health Survey (SF-36). A total of 35 subjects completed the experiment (18 in the exercise group, 17 in the control group). After the 12-week intervention period, the lower-limb strength and the CSA-RF were significantly increased in the exercise group. The activity level did not change significantly in either group. The exercise group exhibited significant improvements in the SF-36 mental health domain scores. Thus, strengthening exercise is useful for patients with RA.

Predictive value of semi-quantitative index from F-18-fluoride PET/CT for treatment response in patients with ankylosing spondylitis.

PURPOSE: We aimed to investigate whether the lesion-to-background ratio (LBR) of the whole spinal column in ankylosing spondylitis (AS) patients based on baseline F-18-fluoride positron emission tomography/computed tomography (PET/CT) is associated with treatment response. METHODS: Twenty-eight male patients with a wide range of disease activity were included in this study. Abnormal bone metabolic lesions on baseline F-18-fluoride PET/CT were visually assessed in three areas: anterior discovertebral units (DVUs) and posterior joints (cervical, thoracic, and lumbar facet joints, costovertebral joints, and costotransverse joints) of vertebrae and sacroiliac joints of the sacrum. The highest maximum standardized uptake value (SUVmax) of lesions within each area was measured. The LBR as a semi-quantitative index was defined as SUVmax of the lesion divided by mean SUV of the L5 vertebral body. Clinical disease activity was assessed using the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), and treatment response was evaluated based on follow-up BASDAI score. RESULTS: The median follow-up period was 4.5 years (range, 0.9-5.5). The BASDAI score significantly decreased from median 6.6 (range, 1.0-9.5) at baseline to 2.0 (range, 0.6-8.1) at follow-up (P <  0.001), and the median change in BASDAI score was -3.4 (range, -8.2-4.4). The LBR of posterior joints in the spinal column had a significant positive correlation with follow-up BASDAI score (r = 0.394, P =  0.042). CONCLUSION: In patients with AS, the LBR of posterior joints of the spinal column on baseline PET/CT could be considered one of the potential surrogate markers for predicting treatment response.

Machine learning-based prediction of radiographic progression in patients with axial spondyloarthritis.

INTRODUCTION: Machine learning is applied to characterize the risk and predict outcomes in multi-dimensional data. The prediction of radiographic progression in axial spondyloarthritis (axSpA) remains limited. Hence, we tested the feasibility of supervised machine learning algorithms to predict radiographic progression in axSpA. METHODS: This is a retrospective and hospital-based study. Clinical and laboratory data obtained from two independent axSpA groups were used as training and testing datasets. Radiographic progression over 2 years was assessed using the modified Stoke Ankylosing Spondylitis Spine Score (mSASSS) and mSASSS worsening by ≥ two units was defined as progression. Seven machine learning models with different algorithms were fitted, and their performance for the testing dataset was assessed using receiver-operating characteristic (ROC) and precision-recall (PR) curve. RESULTS: The training and testing groups had equivalent characteristics, and radiographic progression was identified in 25.3% and 23.7%, respectively. The generalized linear model (GLM) and support vector machine (SVM) were the top two best-performing models with an average area-under-curve (AUC) of ROC of over 0.78. SVM had the higher AUC of PR compared with GLM (0.56 versus 0.51). Balanced accuracy was over 65% in all models. mSASSS was the most informative variable, followed by the presence of syndesmophyte(s) at the baseline and sacroiliac joint grades. CONCLUSIONS: Clinical and radiographic data-driven predictive models showed reasonable performance in the prediction of radiographic progression in axSpA. Further modelling with larger and more detailed data could provide an excellent opportunity for the clinical translation of the predictive models to the management of high-risk patients.Key Points• Clinical and radiographic data-driven predictive models showed reasonable performance in the prediction of radiographic progression in axSpA.• Further modelling with larger and more detailed data could provide an excellent opportunity for the clinical translation of the predictive models to the management of high-risk patients.

Association of Serum Uric Acid with Cardiovascular Disease Risk Scores in Koreans.

As the prevalence of gout and hyperuricemia increases, the comorbidities of gout and hyperuricemia have become a public health burden. In particular, risks of cardiovascular disease (CVD)-related complications are increasing. However, a few guidelines exist for the management of hyperuricemia. This cross-sectional study aimed to investigate the association of serum uric acid with CVD risk in the general population of Korean adults. We examined cross-sectional data from the first and second years of the seventh Korea National Health and Nutrition Examination Survey 2016-2017. Among 16,277 participants, 8781 were analyzed. We estimated the CVD risk using a 10-year CVD risk score prediction formula. There was a significant association of serum uric acid with 10-year CVD risk scores after adjusting for physical activity, body mass index, serum creatinine, and alcohol consumption in both sexes (p < 0.001). In the fitted fractional polynomial model, an approximate U-shaped association between serum uric acid levels and 10-year CVD risk scores was found in men. At the serum uric acid level of 6.9 mg/dL, the CVD risk was lowest. An approximate J-shaped association between serum uric acid levels and 10-year CVD risk scores was found in women. Our study showed that hyperuricemia was associated with an increased CVD risk. Hypouricemia was also associated with an increased CVD risk in men. We, therefore, recommend proper management of uric acid levels in the general population to reduce CVD risks.

Respiratory viral infections and the risk of rheumatoid arthritis.

BACKGROUND: We aimed to investigate the effects of ambient respiratory viral infections in the general population on rheumatoid arthritis (RA) development. METHODS: Data of weekly incident RA (2012-2013) were obtained from the Korean National Health Insurance claims database, and those of weekly observations on eight respiratory viral infections were obtained from the Korea Centers for Disease Control and Prevention database. We estimated the percentage change in incident RA associated with ambient mean respiratory viral infections using a generalized linear model, after adjusting for time trend, air pollution, and meteorological data. RESULTS: A total of 24,117 cases of incident RA (mean age 54.7 years, 18,688 [77.5%] women) were analyzed. Ambient respiratory viral infections in the population were associated with a higher number of incident RA over time, and its effect peaked 6 or 7 weeks after exposure. Among the 8 viruses, parainfluenza virus (4.8% for 1% respiratory viral infection increase, 95% CI 1.6 to 8.1, P = .003), coronavirus (9.2%, 3.9 to 14.8, P < .001), and metapneumovirus (44%, 2.0 to 103.4, P = .038) were associated with increased number of incident RA. The impact of these respiratory viral infections remained significant in women (3.8%, 12.1%, and 67.4%, respectively, P < .05) and in older patients (10.7%, 14.6%, and 118.2%, respectively, P < .05). CONCLUSIONS: Ambient respiratory viral infections in the population were associated with an increased number of incident RA, especially in women and older patients, suggesting that respiratory viral infections can be a novel environmental risk factor for the development of RA.

Relationship between oxidative balance score and quality of life in patients with osteoarthritis: Data from the Korea National Health and Nutrition Examination Survey (2014-2015).

Osteoarthritis (OA) has a multifactorial etiology that includes oxidative stress. Oxidative balance score (OBS) is a well-known indicator of oxidative stress. However, the association between OBS and OA has not been assessed. Thus, this study aimed to investigate the associations of OBS with OA and quality of life (QOL) in patients with OA.By using data from the Korea National Health and Nutrition Examination Survey VI, patients previously diagnosed and/or treated by a physician were considered to have OA regardless of the affected joints. The control group was defined as participants without any form of chronic arthritis. OBS was calculated by combining 10 pro-oxidant and antioxidant factors through a baseline nutritional and lifestyle assessment. Higher OBS scores indicated a predominance of antioxidant exposure. Multivariable logistic regression was used to estimate the adjusted odds ratios (ORs) for OA, and the EuroQoL five-dimensional questionnaire (EQ5D) was used in patients with OA after adjusting for demographic factors and comorbidities.Among the 14,930 participants, 296 patients with OA, and 1,309 controls were included in the analysis. In the age- and sex-adjusted model, the OR of the total OBS for OA was significant. In the full model adjusted for age, sex, education, income, and comorbidities, the total OBS for OA was not significant. Only the non-dietary pro-oxidant OBS had a significant inverse association with OA. The patients with OA who had a high EQ5D score had a higher total OBS than those with a low EQ5D score. The OR of the total OBS for a high EQ5D score was 1.14 in the multivariable logistic regression model. As we analyzed the OBS as a categorical variable (reference = Q1), the ORs of the Q2, Q3, and Q4 (highest) total OBS were 1.43, 2.71, and 2.22, respectively.In the fully adjusted model, the total OBS was not associated with OA. However, a positive association was observed between the total OBS and QOL in the patients with OA, indicating that antioxidative status was associated with better QOL in patients with OA.

Association of cumulative anti-cyclic citrullinated protein antibodies with radiographic progression in patients with rheumatoid arthritis.

INTRODUCTION: Antibody against cyclic citrullinated protein (ACPA) is counted as one of the most important biomarkers in diagnosis, classification, and prognosis of rheumatoid arthritis (RA). We examined the evolution of ACPA during disease course and assess predictive value of time-weighted cumulative ACPA titer on radiographic progression in RA patients. METHOD: A group of 734 patients with RA was followed longitudinally over 2 years, with annual measurements of ACPA. The cumulative titers of ACPA were calculated using the trapezoidal rule and were divided into three categories: negative, low-to-moderate, and high. Radiographs of the hands were scored with the modified Sharp score (SHS). Multivariable logistic regression models were performed to identify independent predictors over follow-up for individual patients with different combinations of risk factors. The effect size was computed by Cohen's d method. RESULTS: The patients with radiographic progression had a higher SHS at baseline; and smoking status, diabetes, RF positivity, and use of biologic DMARDs were independently associated with radiographic progression (all P < 0.05). As for ACPA, reversion happened more commonly in men and was associated with younger onset age and lower titer at baseline, but it had no direct relevance to radiographic outcome. In multivariable regression analysis, only high cumulative or baseline titer of ACPA had a predictive power for rapid radiographic progression (all P < 0.05), and cumulative ACPA titer was superior in terms of statistical significance (Cohen's d, 0.637 versus 0.583). CONCLUSIONS: High cumulative ACPA titer was independently associated with accelerated radiographic progression, especially with initiation of joint damage.