Background Microwave Radiometry (MR) is a rapid, easy-to-perform, non-invasive method that detects in-depth tissue temperatures. In a proof-of-concept study we found that an increased knee joint temperature detected by MR reflects the presence of subclinical synovial inflammation in patients with rheumatoid arthritis.1 Objectives To test the hypothesis that MR-derived increased temperature of sacroiliac (SI) joints is a biomarker of local inflammation in patients with axial spondyloarthritis (SpA) in a cross-sectional study. Methods Sixty patients with SpA (32 with ankylosing spondylitis, 24 with psoriatic arthritis, 4 with enteropathic arthritis) underwent clinical and laboratory assessments, magnetic resonance, X-ray imaging, and MR measurements of SI joints. All MR measurements were performed by a physician who was blinded to the clinical evaluations. Patients were classified as having active sacroiliitis (pain and/or tenderness and/or bone marrow oedema on magnetic resonance, n=23), inactive sacroiliitis (n=19), whereas signs/symptoms of a present/past SI joint involvement were absent in the remaining 18 patients. Twenty five age-matched healthy individuals served as controls. Three MR measurements were performed along each SI joint and a number was calculated for each patient denoting the difference (ΔΤ) in temperature between the warmest point in either SI joint and a control point, which was the iliac bone. Results A lower ΔΤ, indicative of a warmer joint, was found in patients with either active or inactive sacroiliitis compared to patients without sacroiliitis [mean (SD) ΔΤ of 0.1 (0.5) vs. 0.6 (0.5), respectively, p=0.001] or healthy controls [mean (SD) ΔΤ of 0.60 (0.55), p=0.003]. A warmer SI joint was not significantly associated with the presence of local tenderness or pain, the patient visualised analogue pain scale for SI joints, the BASDAI score, the New York radiological grading of SI lesions, or C-reactive protein levels. Notably, all 20 patients with increased SI joint temperature (ΔΤ<0.05, which was the mean ΔΤ value minus one SD in healthy controls) had either symptoms/signs of sacroiliitis or imaging confirmed sacroiliitis even in the absence of clinical findings. Receiver operating analysis showed that a ΔΤ cut-off of 0.35 could discriminate with a sensitivity of 78% and specificity of 74% patients with sacroiliitis from those without, or from healthy controls (both p=0.001); sensitivity reached 100% for ΔΤ values equal or lower than 0.05. Conclusions MR-derived increased in-depth temperature indicative of local inflammation of SI joints may serve as an additional diagnostic marker in axial SpA Reference [1] Zampeli EI, et al. “Detection of subclinical synovial inflammation by microwave radiometry,”PLoS ONE2013May;8(5):e64606. Disclosure of Interest None declared