Current protocols based on early and intensive treatment with csDMARDs in rheumatoid arthritis (RA) have allowed the achievement of remission in a considerable proportion of the patients and opened the perspective, in selected cases, of a drug-free monitoring scheme. Treatment discontinuation can lead, however, to possible recurrence of joint inflammation and clinical flare. Understanding the dynamics acting upstream these events remains a fundamental research task with direct clinical and

more »

... -biologic implications. Objectives: To delineate the clinical, serological and ultrasonographic changes associated to a drug-free flare in patients discontinuing csDMARD after achievement of stable remission. Co-primary objective was to compare, through a retrospective analysis in the same patients, these changes with early features of the pathology at onset, before treatment introduction. Methods: 92 RA patients in stable DAS28 remission following a DAS-steered treatment strategy with MTX were recruited in our Centre and introduced to a drug-free monitoring scheme according to the following inclusion criteria: a) treatment introduced within 12 months from symptoms' onset, b) at least 24 months of continuative treatment, c) DAS28 <2.6 for at least 6 months in the absence of glucocorticoids. After discontinuation, all patients were follow-up at three months intervals across 24 months through complete clinical, ultrasonographic (power Doppler ultrasound -PDUSin hands-feet and tendons) and serological analyses. Treatment was re-introduced upon occurrence of moderate disease activity (DAS28 !3.2) in a single occasion. Results: A total drug-free follow-up of 1398 person-months was analysed with a median (IQR) of 15 6-24 months. Thirty-eight patients (27/38 in ACR/EULAR Boolean remission, 16/38 with PD score=0 at withdrawal visit) required treatment reintroduction after a median (IQR) time from discontinuation of 6 3-9 months (range 3-18). DAS28 variations at re-treatment showed a mean (SD) increase of 2.26 (1.03), reflecting significant differences in all DAS components (p<0.001 for ESR, tender joint count, swollen joint count and GH). Clinical activity in flaring subjects was paralleled by average changes in synovial US, with increased PD scores in hands joints (median [IQR]: 3.5 [0.5-7] vs 1 [0-2], p<0.001), feet (1 [0-4] vs 0 [0-0.5], p=0.002) and tendons (0 [0-2] vs 0 [0-0], p=0.002), determining ex-novo PD positivity in 85.7% of PD negative patients (p<0.001). Despite stringent remission achieved at the time of discontinuation, no significant differences were observed between disease onset and drug-free flare in DAS28 (p=0.26), patient global assessment of disease activity (p=0.53) and synovial US scores (p=0.61 for grey scale, p=0.31 for PD) with recurrence of similar patterns of clinical joint involvement. Conclusions: Drug-free clinical flare can occur over a wide temporal window, in the absence of detectable signs of inflammation at the time of treatment discontinuation. It can associate with ex-novo recurrence of US pathologic changes at joint and tendon level, reproducing some of the quantitative/qualitative features of disease onset.