Introduction:To assess the association between growth differentiation factor-15 (GDF15) and radiographic features including bone marrow edema and bone erosion in Spondyloarthritis (SpA).Methods:Patients with SpA (n = 120) receiving treatment in the Guangdong General Hospital, China, between August 2012 and December 2016 were retrospectively included. Serum of patients and healthy controls (n = 30) were collected and GDF15 levels were measured using ELISA. Inflammation was assessed by C-reactive protein (CRP), and magnetic resonance imaging (MRI) of the sacroiliac joint using Spondyloarthritis Research Consortium of Canada score and a method of dichotomy to assess fat metaplasia, bone erosion, and ankylosis. Radiographs of the pelvis were scored using the modified New York (mNY) score.Results:Serum GDF15 levels were higher in SpA patients compared to controls (503.52 ± 222.92 vs. 190.86 ± 104.18 pg/mL, P < .0001). Patients who suffered from bone erosion on MRI had higher levels of GDF15 (525.72 [186.33, 801.62]vs. 428.06 [255.15, 670.98] pg/mL, P = .0375). There was a positive correlation between serum GDF15 and CRP (r = 0.5442, P < .0001). Moreover, GDF15 levels were related to CRP levels (r = 0.5658, P < .0001) in those X-ray scores were III, according to 1984mNY criteria. Receiver operating characteristic (ROC) analysis showed that GDF15 levels above 501.98pg/mL could predict presence of bone erosion on MRI.Conclusion:The present study suggested that serum GDF15 levels are higher in SpA patients than in healthy controls. The GDF15 level was correlated with CRP and may be a surrogate biomarker in bone erosion.

Aim Inflammatory joint diseases (IJDs) are chronic arthritis, but frequently present with co‐morbidities of other organs and systems, which is known as extra‐articular manifestations (EAMs). It is still unclear which clinical characteristics or bio‐markers can predict the development of EAMs. The aim of this study was to estimate the proportion of EAMs in southern Chinese patients with IJDs and to explore the risk factors. Methods This was a retrospective cohort study of a total 1135 IJDs patients, including 788 rheumatoid arthritis (RA) patients, 307 ankylosing spondylitis (AS) patients and 40 psoriatic arthritis (PsA) patients. Demographic data, disease characteristics, laboratory blood tests, medical imaging, and the presence of EAMs were recorded. Results We found 459 (40.44%) patients presented with EAMs: 30.84% had cardiovascular involvement, 7.67% had pulmonary involvement, 5.29% had osteoporosis/low bone mineral density, 2.29% had ocular, 0.79% had gastrointestinal and 0.26% had renal involvements. Multivariate logistic regression showed older age (odds ratio [OR] 1.06, P < .001) and higher anti‐cyclic citrullinated peptide antibody (anti‐CCP) levels (OR 1.003, P = .019) were independent risks of EAMs in RA patients. In the AS group, older age (OR 1.07, P < .001) and higher disease activity (OR 3.24‐7.42, both P < .05), were independent risks of EAMs. In the PsA group, longer disease duration (OR 1.01, P = .036) and higher disease activity (OR 1.15, P = .004) were univariate associated factors. Conclusion These results suggested the high prevalence of EAMs, and it is important to regularly screen for EAMs, as they influence treatment decisions and impact on patients’ quality of life.