Background The goals of treatment in Rheumatoid Arthritis (RA) are remission and low disease activity (LDA); therefore, it is important to identify reliable predictive factors of these targets so that treatment can be tailored.
Objectives We aimed at identifying positive and negative predictive factors of remission and LDA in RA patients treated in real practice with first line anti-TNF drugs.
Methods Retrospective analysis of DAS28 remission/LDA in 308 RA patients treated for two years with first line anti-TNF drugs from January 2008 to December 2014: [female n=251 (81.5%), age 55±13 years, disease duration 7.6±8.4 years, rheumatoid factor (RF, n=220 (71.4%)), anti-citrullinated antibodies (ACPA, n=214 (69.5%)), baseline C-reactive protein (CRP, 11.4±16.5 (mg/L)), baseline DMARDs (n=225 (73.1%)), baseline steroid (n=160 (51.9%)), Adalimumab (n=109 (35.4%)), Etanercept (n=177 (57.5%)), other anti-TNF (n=22 (7.1%))]. Predictive factors considered for achievement of remission and LDA were: gender, age at the time of anti-TNF treatment (cut-off >54 yrs old, n=139 (45.1%)), age at onset >65 yrs old (n=27 (8.8%)), early arthritis (n=99 (32.1%)), baseline negative CRP (n=150 (48.7%), RF/ACPA positivity, good/moderate EULAR response at 6 months (n=169 (54.9%)), presence of any comorbidity (n=210 (68.2%)), cardiovascular comorbidity (n=83 (26.9%)), metabolic syndrome (n=26 (8.4%)), infectious (n=49 (15.9%)), thyreopathies (n=64 (20.8%)), concomitant DMARDs/steroid treatment. Intention to treat, ROC-curve, univariate and multivariate analysis by logistic regression were performed. Selection of the variables to be included in the multivariate analysis was performed on the basis of their statistical significance in the univariate analysis.
Results In the univariate analysis positive predictors of remission and LDA at 6 months, 1 year and 2 years of anti-TNF treatments were identified as follows (Figure 1A): male gender, age at the time of anti-TNF treatment <54 yrs old, age at onset <65 yrs old, early arthritis, baseline negative CRP levels, good/moderate EULAR response at 6 months and concomitant DMARDs treatment. Negative predictors were presence of any comorbidity, cardiovascular comorbidity and metabolic syndrome. In the multivariate analysis the only significant positive predictor of remission and LDA at 6 months, 1 and 2 years of treatment was the negative CRP levels at baseline (Figure 1B).
Conclusions Negative CRP levels prior to commence anti-TNF treatments in RA patients appeared to be a reliable predictor of remission and LDA in clinical practice.
Disclosure of Interest None declared