Background Definitive biomarkers have not been identified to predict flares in patients with RA in clinical remission.
Objectives To determine clinical, serological and ultrasonografic variables predicting disease relapse at 12 months of follow-up in a cohort of RA patients in clinical remission.
Methods We included consecutively RA patients in clinical remission defined as DAS28-ESR <2.6 for >6months evaluated by two independent rheumatologist from our Arthritis Clinic. Complete clinical and biological assessment, determination of serum levels of angiogenic and proinflamatory cytokines, and ultrasound scans of both hands were performed at baseline and after 12 months of follow-up. Risk factors odds ratio (OR) related with 12 months flare were adjusted with multivariate logistic models.
Results Sixty patients with RA in remission were included; 78.3% were female. Median age was 53 years (range: 31–74), weight was 67 kg (47–108). Disease and remission mean duration were 111.9 ± 86.4 and 37 months, respectively. 71% of patients were RF positive and 81% were ACPA positive. Patients taking low-dose prednisone, csDMARDs and bDMARDs were 16 (26%), 47 (78%) and 28 (46.7%), respectively. At baseline, 66% had power-doppler (PD) signal and 48% also had synovial hyperplasia grade ≥2, so fulfilling the criteria previously defined of ultrasound defined active sinovitis (UdAS). Twenty-six (43%) patients had disease flare during the follow-up.
In the univariate analysis, disease flares were more frequent in patients with PD signal (p=0.056) or UdAS (p=0.117), and significantly less frequent in patients with mild-moderate intake of alcohol. However, in the multivariate analysis, only higher ESR levels (OR: 1.1, 95%CI: 1–1.1), higher weight (OR: 3.1, 95%CI: 1.4–7) and low-dose prednisone treatment (OR: 1.8, 95%CI: 1.2–2.8) were independent risk factors at baseline for disease flares. This model has a sensitivity of 76.9% (95%CI: 56.4–91), a specificity of 76.5% (95%CI: 58.8–89.3), a predictive positive value of 71.40% (95%CI: 52.4–88.4), a predictive negative value of 81.3% (95%CI: 62.7–91.7) and a good predictive power (ROC: 0.849, IC 95%: 74.4–95.4%). We did not find differences between angiogenic or proinflammatory cytokine serum levels and not differences in RAMRIS between patients relapsing or maintaining the remission at baseline.
Conclusions This prospective study suggests that higher ESR levels, higher weight and low-dose prednisone treatment are relevant independent risk factors for development of disease flares in patients with RA in clinical remission. However, imaging data (Ultrasound or MRI) or angiogenic/cytokine serum levels were no predictors of flare in this study.
Disclosure of Interest None declared