Background/Purpose: Biologic DMARD (biologics) therapy for rheumatoid arthritis (RA) strongly suppresses joint destruction regardless of its efficacy for disease activity. On the contrary power Doppler (PD) signal detected by ultrasonography (US) is said to be the most potent predictive factor for subsequent radiologic progression. This study aimed to clarify whether PD signal predicts joint destruction of RA patients under biologics therapy in daily practice.

Methods: RA patients who began and continued to receive biologics for more than six months were included. Clinical, laboratory, and US examinations were conducted sequentially from baseline (1^st) to the last observation (last). Bilateral wrists and all of the MCP and PIP joints were examined by PDUS and the PD signals were graded from 0 to 3 in each joint. The total PD score was defined as the sum of scores of individual joint, and mean score of several assessments during observational period was also calculated. Structural damage of hands at baseline and at the last was measured by using modified Sharp scoring method for hand X-ray (TSS). Patients having the change in TSS (delta TSS) exceeded 0.5 U per year were defined as showing radiologic progression.

Results: Objectives were 100 RA patients (female 85%, age 59.1±13.4 y.o., disease duration 8.0±8.2 years, 1^st DAS28 4.84±1.43). Sixty-three patients continued the same drug (anti-TNF 31, tocilizumab 26, abatacept 6) whereas 37 patients switched biologics. During continuing biologics therapy for 25.3±16.8 months, structural damage progressed in 51% of the patients, who were classified into progressive group, including 18 patients with achieving clinical remission. The progressive group contained more patients who switched agents, and showed higher 1^st DAS28 and higher last DAS28 as well as higher last total PD score, compared to the other group. Furthermore, yearly radiologic progression (delta TSS) weakly related to 1^st, last, and mean total PD score in addition to 1^st DAS28, 1^st CRP, and 1^st and last MMP-3. Multiple regression analysis using stepwise method revealed that both 1^st CRP and mean total PD score were independently associated with joint destruction. For joint-based analysis of total 2200 joints, radiologic progression were seen in 155 joints (7.0%), of which wrists were mainly affected joints, and the existence of PD signal in a joint at any time during observational period was the potent risk for subsequent destruction of the joint (Table).

Conclusion: To practice ‘Treat to Target’ for achieving radiologic remission in real world, monitoring individual joint by PDUS will be helpful in carrying out optimal intervention even under biologics treatment.