A paradigm of difficult-to-treat rheumatoid arthritis: subtypes and early identification.

Objectives: Multiple failures to biologic or targeted specific disease-modifying anti-rheumatic drugs (b/tsDMARDs) that lead to difficult-to-treat rheumatoid arthritis (D2TRA) may be the result of multi-drug inefficacy or reflect treatment problems related to adverse events, comorbidities, and/or poor adherence. We aimed to characterise a cohort of D2TRA patients in clinical practice, to analyse the differences between D2TRA due to inefficacy versus D2TRA from other causes, and to compare them with non-D2TRA.

Methods: The D2TRA group included patients who were receiving ≥2b/tsDMARDs due to inefficacy (D2TRA-inef cacy) or because of adverse events, poor adherence, contraindications, comorbidities, drug-intolerance, etc. (D2TRA-other). Patients who achieved low disease activity or remission with the rst bDMARD were classified as non-D2TRA patients. For all patients, demographic, clinical characteristics and laboratory parameters were assessed prior to starting the rst b/tsDMARD. Descriptive analysis was performed and bivariate logistic regression models were assembled.

Results: In total, 253 patients were included: 131 non-D2TRA and 122 D2TRA [86 (70.5%) D2TRA-inefficacy and 36 (29.5%) D2TRA-other]. Comparison of the two groups of D2TRA patients: no differences in gender, age at start of b/tsDMARD or age at RA diagnosis were found; this was also true of socioeconomic status, frequency of anxiety-depression and other comorbidities. Patients categorised as D2TRA-other had less extra-articular manifestations than D2TRA-inef cacy, as well as lower values of DAS28 at the start of the rst b/tsDMARD. Comparisons of Non-D2TRA patients versus D2TRA-other resulted in the following observations: no differences in sociodemographic characteristics were evident nor were there any differences in terms of disease activity.

Conclusions: Patients with D2TRA-other are indistinguishable from non-D2TRA patients at baseline, indicating the former cohort does not appear to have any predictive value during the early stages of b/tsDMARD treatment, unlike what occurs in patients with D2TRA-inefficacy.