LAUSR

OBJECTIVES To compare the features of difficult-to-treat rheumatoid arthritis (D2TRA) patients using 2 different definitions according to the previous failure of targeted therapies. PATIENTS & METHODS We stratified consecutive RA patients treated at Cochin Hospital into two groups, a D2TRA group and a non-D2TRA group, according to two definitions of D2TRA. Both definitions defined D2TRA as RAs failing at least two targeted therapies, with a different mechanism of action for the EULAR-D2TRA or without prejudging the mechanism of action and for the Alternative D2TRA definition. RESULTS We included 320 consecutive RA patients. We identified 76 EULAR-D2TRA and 244 non-DTRA patients, and 120 Alternative D2TRA and 200 non-DTRA patients.Compared to non-D2TRA, D2TRA patients from both definitions were more likely to have lower socio-economic level, positive rheumatoid factor, interstitial lung disease, higher DAS28-CRP and were more likely to respond to rituximab and JAK inhibitors. Although EULAR and Alternative D2TRA displayed similar clinical and biological features, they were characterized by different therapeutic profiles. We observed fewer patients receiving methotrexate in the Alternative D2TRA group (53% vs 64%, p = 0.046).Patients with Alternative D2TRA not fulfilling the EULAR definition (n = 44) had all received two successive first-line TNF inhibitors, a monoclonal antibody and a soluble receptor, and were comparable to EULAR-D2TRAs with regards to all other characteristics. CONCLUSION Low socioeconomic status, diabetes, ILD, and absence of combination with methotrexate allow identification of D2TRA. In addition, the inclusion of patients failing two TNF inhibitors in the EULAR definition of D2TRA, as "early-D2TRA", would facilitate the rapid identification of D2TRA patients.