This prospective study aimed to determine whether the comprehensive ultrasonographic assessment of synovial inflammation predicts relapse after discontinuation of treatment with a biologic agent in patients with rheumatoid arthritis (RA) in clinical remission.


RA patients in clinical remission (Disease Activity Score in 28 joints [DAS28] <2.6) receiving treatment with a biologic agent who agreed to discontinue the treatment were recruited. Patients underwent a comprehensive ultrasound scan on 134 synovial sites in 40 joints and were prospectively followed up for 6 months. Physicians who evaluated the patients during the study period were blinded to the baseline ultrasound findings.


Forty-two patients receiving either a tumor necrosis factor antagonist or tocilizumab were enrolled. Using the optimal cutoff values determined by receiver operating characteristic curve analysis, relapse rates were significantly higher in patients whose total ultrasound scores at discontinuation were high than in those whose total ultrasound scores were low (P < 0.001 for both total gray-scale and power Doppler scores), whereas the difference between high and low DAS28 was not statistically significant (P = 0.158 by log rank test). Positive and negative predictive values were 80.0% and 73.3% for the total gray-scale score and 88.9% and 74.2% for the total power Doppler score, respectively.


In RA patients in clinical remission receiving treatment with a biologic agent, residual synovial inflammation determined by comprehensive ultrasound assessment predicted relapse within a short term after discontinuation of the treatment. Our data provide a rationale and groundwork to conduct a large-scale study for establishment of ultrasound-based strategies to optimize the period of treatment with a biologic agent.