LAUSR

OBJECTIVES Increased in-depth joint temperature measured by the rapid, easy-to-perform microwave radiometry (MWR) method may reflect inflammation, even in the absence of clinical signs. We hypothesized that MWR is useful for RA and spondyloarthritis patients' assessment. METHODS Clinical examination, joint ultrasound and/or MRI and MWR were performed in two independent patient-control cohorts (n = 243). RESULTS Among single RA joints MWR performed best in the knee using ultrasound as reference, with 75% sensitivity-73% specificity for grey-scale synovitis score ⩾2, and 80% sensitivity-82% specificity for power Doppler positivity. A stronger agreement was evident between increased knee relative temperature (Δt) and power Doppler positivity (82%) than with clinical examination (76%). In a different patient cohort with painful knees, a knee Δt ⩽0.2 predicted power Doppler positivity with 100% positive and negative predictive values. A thermo-score summing 10 Δt values of three large and seven small RA joints (elbow, knee, ankle, wrist, four hand and two foot joints of the clinically dominant arm or hand and leg or foot) correlated with ultrasound scores of synovitis/tenosynovitis (all P < 0.001) and the 28-joint Disease Activity Score (DAS28) (P = 0.004). The agreement of the thermo-score with ultrasound-defined joint inflammation (82%) was stronger than with DAS28 (64%). The thermo-score improved significantly after 90 days of treatment in patients with active RA at baseline (P = 0.004). Using MRI as reference, Δt of sacroiliac joints could discriminate between spondyloarthritis patients with or without sacroiliitis with 78% sensitivity-74% specificity. CONCLUSION In-depth increased MWR-derived joint temperature reflects both subclinical and clinically overt inflammation and may serve as a biomarker in arthritis.