Semi-quantitative analysis of rheumatoid finger joint synovitis using power Doppler ultrasonography: when to perform follow-up study after treatment consisting mainly of antitumor necrosis factor alpha agent

Abstract

To determine the timing for follow-up study of power Doppler ultrasonography (PDUS) by evaluating the response of finger joint synovitis in patients with rheumatoid arthritis (RA) to treatment including infliximab, an antitumor necrosis factor alpha agent. Bilateral second/third metacarpo-phalangeal (MCP) joints and second proximal inter-phalangeal (PIP) joints (total of six joints) in 21 patients (18 women and three men; median age 53 years) with chronic active RA were assessed by PDUS before and after 2 weeks, 6 weeks, 14 weeks, 30 weeks, 38 weeks, 46 weeks, and 54 weeks of infliximab infusion. Pulse Doppler settings were standardized for each patient and optimized for the detection of synovial blood flow by adjustment of color gain, pulse repetition, and flow optimization. Power Doppler signal was graded for each joint [joint grade for power Doppler (JGPD) signals], and the sum of the grades of six joints was defined as the PDUS index [joint index for power Doppler signals (JIPD)] at each visit. PDUS and clinical parameters [28-joint disease activity score (DAS28), health assessment questionnaire, and C-reactive protein (CRP) level] were independently assessed and compared with baseline values. The American College of Rheumatology (ACR) core set responders and non-responders at week 54 were compared for clinical parameters and PDUS index at each visit. Fourteen patients completed the planned treatment for 1 year, while six patients dropped out for various reasons and one died suddenly. PDUS was performed a total of 146 times on 467 joints. DAS28 was assessed 127 times. Both DAS28 and JIPD had decreased at the follow-up. Comparative analysis between DAS28 and PDUS was available 125 times. The transverse correlation between the PDUS index and DAS28 was not significant throughout the follow-up period. When responders and non-responders were discriminated at week 54, a logistic regression model for the binary endpoint of responder vs non-responder, with PDUS index as explanatory variable at time point 0, and follow-up revealed statistical significance from week 38 and on. PDUS reflected infliximab's effect on pannus vascular signals; this effect was observed as early as 2 weeks after treatment had begun. Also, the responders to treatment at 54 weeks tended to have fewer JIPD than non-responders in the follow-up period. PDUS may be performed at week 38 or later to foresee the response to the treatment at week 54.