Prediction of response to disease modifying antirheumatic drugs in rheumatoid arthritis.

Aim: To investigate potential predictors of response to conventional DMARDs in RA.

Methods: Study design - 6-month follow-up prospective study.

Participants: RA patients with active disease. INTERVENTION AND FOLLOW-UP: Introduction of one DMARD. Response to treatment evaluated at 6 months (ACR20 criteria).

Analysis: Potential predictors of response, patients' demographics, disease activity, percentages of PBMC subsets expressing P-gp, serum IL-1β, IL-6, IL-8, IL-10, IL-12, TNF-α levels, were evaluated using univariate and multivariate logistic regression analysis. ROC curve analyses were performed in order to obtain thresholds allowing the prediction of response.

Results: Forty-two patients (mean age = 57 ± 13 years, mean disease duration = 5.4 ± 7.2 years) were included. MTX was given to 30. The response to therapy was predicted by the baseline serum level of TNF-α (mean = 30.2 pg/ml ± 18 in non-responders vs. 11.9 pg/ml ± 11.2 in responders). The threshold, which predicted with the best accuracy the response to treatment, was 20.1 pg/ml (sensitivity, specificity, positive and negative predictive values of 75, 78.9, 83.3, and 69.2%, respectively; AUC = 80.3%, 95% CI = 62.8-97.7%). Similar results were obtained in the subgroups of patients treated with MTX and patients with early RA of less than 3 years duration.

Conclusion: In the present work, the serum concentration of TNF-α was related to further response to DMARDs. Other works are needed for confirmation and to assess whether such biomarker could be used to predict the response to DMARDs at the individual level.