Ultrasound based withdrawal of biologics in Rheumatoid Arthritis

Introduction: Up to 30% of Rheumatoid arthritis (RA) patients treated with synthetic and biologic disease modifying antirheumatic drugs (DMARDs) achieve clinical remission. Once remission is achieved, cessation of the biologic DMARD may be considered; however, up to 60% of patients flare after stopping the biologic treatment. The aim of this study is to investigate whether ultrasound, performed before discontinuation of the biologic DMARD, may be valuable to predict a flare of the disease. Hypotheses: The primary hypothesis of this prospective study is that residual inflammation in patients in clinical remission [according to the criteria proposed by the American College of Rheumatology (ACR) / European League Against Rheumatism (EULAR)], as determined by sonography before discontinuation of the biologic DMARD, predicts a flare of the disease at week 16. Secondary hypotheses are that sonography may be useful to predict flares at 24 and 52 weeks, that ultrasound detected inflammation is higher at time of active disease than at remission, that sonography verified disease activity precedes a clinical flare and that patients with high level of ultrasound verified inflammation re-achieve remission slower than patients with low level of inflammation. Methods: We will recruit 110 RA-patients with persistent clinical remission (=6 month, documented at =2 visits) according ACR/EULAR remission criteria. At baseline, the biologic treatment is stopped. Patients undergo a total of 9 study visits within 52 weeks, clinical laboratory and ultrasound examinations of 14 joints are performed at each visit. Patients with a flare of the disease are excluded from the active phase of the study and are treated according to current guidelines. We expect that 60% of patients will flare and that 40% of the relapsing patients have ultrasound verified inflammation at baseline compared to 10% of patients with persistent remission. Innovation: This is the first study investigating the value of ultrasound for the prediction of successful discontinuation of biological agents in an adequately powered trial of RA patients in persistent clinical remission. We need reliable predictors for the course of RA patients receiving biological agents to best tailor individual management plans. Successful discontinuation of a biological agent in a RA patient in clinical remission would reduce not only the individual risk of adverse events but also the costs for the social insurance system.

Up to 35% of patients with Rheumatoid Arthritis (RA) achieve remission by using synthetic and biologic DMARDs; however, there is considerable uncertainty regarding the management of patients once this clinical state has been achieved. The hypothesis of this study was that ultrasound verified subclinical inflammation [Power Doppler (PD) score =1 within 14 joints amenable to sonography] at the time of withdrawal of the biological agent predicted a disease flare between baseline and week 16 (primary hypothesis), week 24 and/or week 52. Additional hypotheses were that the PD score at time of relapse was higher than in remission and that the PD score at baseline was more likely to predict a relapse than the residual swollen joint count. Inclusion criteria were RA-patients with clinical remission according to ACR-EULAR remission criteria at time of inclusion and remission or low disease activity according to SDAI in the preceding 6 months. At baseline, biological DAMRD therapy was stopped and synthetic DMARDs were continued. Patients underwent 9 study visits within 52 weeks. Ultrasound examinations of 14 joints at hands and feet as well as clinical and laboratory assessments were conducted at each visit. Patients were considered to have a relapse if disease status changed from remission to active disease according to clinical scores. Out of the 110 patients stipulated in the protocol, 38 (34.5%) were included. Overall, there were 9, 10 and 13 relapses between baseline and weeks 16, 24 and 52, respectively. Relapses till week 16 (primary outcome) tended to be more common in patients with PD positive synovitis at baseline than in those without [9/30 (30.0%) vs. 0/7 (0%), p=0.160)]. Similar observations were made for weeks 24 and 52. PD scores were higher at the time of relapse as compared to the preceding visits [mean difference in the PD score 3.2 (4.5) points, p=0.034]. There were trends towards a higher mean baseline PD score in patients who had a relapse between baseline and week 16 as compared to those who remained in remission (5.2 5.8 vs. 2.3 3.0, p=0.079). Similar observations were made for relapses until weeks 24 and 52. No difference was observed comparing mean residual swollen joint counts at baseline between patients with and without a relapse. In summary, our results imply that in RA patients in strict clinical remission, PD assessment at baseline could help to identify patients who will relapse after the cessation of a bDMARD. Due to premature termination and insufficient power of the present trial, however, no definitive conclusion can be made.