Background/Purpose: Rheumatoid arthritis (RA) care has moved towards early diagnosis and treatment to improve long-term patient outcomes. Our objective was to assess prognostic factors for sustained remission and absence of clinical inflammation in an early RA population treated according to a strict tight-control algorithm with semi-personalized treatment adjustments.

Methods: RA patients who fulfilled the ACR/EULAR 2010 classification criteria, with <2 years from first swollen joint and who were DMARD naive with indication for DMARD treatment were included in the ARCTIC trial. Patients were followed by a tight control regimen with defined treatment target (DAS <1.6, SJC44=0 + no power Doppler signal in half the patients) and response evaluation. Rapid escalation of medication was allowed if negative prognostic factors for joint destruction were present (seropositivity with baseline radiographic erosions or MRI bone marrow edema). Data collections included clinical examination, laboratory tests, patient reported outcomes, ultrasound examination and radiographs. Multivariate logistic regression was used to assess the associations between baseline variables and different definitions of sustained remission between 16 and 24 months.  

Results: Mean [SD] disease duration of the 222 patients was 7.2 [5.4] months, mean original DAS 3.5 [1.2], 72% were RF and 82% ACPA positive, and 16% escalated treatment more rapidly to biologic therapy. Sustained (16 to 24 months) SDAI and ACR/EULAR remission were reached by 33% and 24% respectively. Tender joints assessed by Ritchie Articular Index was a consistent independent negative predictor of reaching sustained SDAI remission, ACR/EULAR remission, no swollen joints in any of the 44 joints evaluated, and a composite outcome of no swollen joints + radiographic progression≤0.5 per year + DAS remission between 16-24 months (table). Patients with high scores on the Ritchie Articular Index had low odds ratios for sustained remission and sustained absence of swollen joints when adjusting for baseline swollen joint count (figure).    

Conclusion: In this early RA treat-to-target study, tender joints assessed by Ritchie Articular Index at baseline is a consistent negative predictor of reaching sustained remission. Our findings support that semi-personalized targeted treatment might modify the impact of known predictors.