Materials and methods. A cohort study based on a retrospective analysis of medical documentation data from 8985 hospitalizations at the round-the-clock inpatient facility (from 04.2018 to 03.2024) and 25071 hospitalizations at the day hospital (from 01.2014 to 03.2024) was conducted. The study included 1090 patients with a confirmed diagnosis of RA. The final study included 564 patients. Patients were divided into 2 subgroups: DtT RA (n=133 (12.2%)) – with the ineffectiveness/intolerance of at least 2 different disease-modifying antirheumatic drugs (DMARDs)/biologic DMARDS (bDMARDs) (according to the European Alliance of Associations for Rheumatology criteria, 2021); non-refractory RA (non-DtT; n=431 (39.5%)) – treated with ≥1 DMARD/bDMARD and achieving low disease activity (LDA) in RA (DAS28-CRP/ESR (Disease Activity Score 28 with detection of C-reactive protein/erythrocyte sedimentation rate) >2.6–3.2). An assessment and comparative analysis of clinical-laboratory and instrumental data in the RF (rheumatoid factor) and non-RF groups were conducted. Differences were considered statistically significant at p<0.05.

Results and discussion. DtT compared to non-DtT were significantly younger (median (Mе) age – 54 and 58 years, respectively), had a longer disease duration (Mе – 186 and 116 months, respectively), shorter lifespan (Me – 662 and 823 months, respectively), more often seropositive for DtT (82.0% and 74.5%, respectively) and anti-citrullinated protein antibodies (82.3% and 69.8%, respectively). The onset of RA in DtT compared to non-DtT also occurred at a younger age (37 and 46 years, respectively). DtT patients compared to non-DtT statistically significantly more frequently developed erosive RA form (88.0% and 70.7%, respectively), reached the fourth radiographic (erosive) stage (38.3% and 20.9%, respectively), had rheumatoid nodules (27.1% and 14.4%, respectively), a history of the withdrawal of medications due to adverse events (49.2% and 35.9%, respectively), less frequently achieved remission (Rem) and LDA in RA (Rem: 19.6% and 60.3%, respectively; LDA: 36.5% and 95.8%, respectively). Additionally, in the DtT group compared to the non-DtT group, a significantly later initiation of methotrexate (MT) after diagnosis (104 and 65 months, respectively), biologic therapy (80.0 and 38.0 months, respectively) was noted. Etanercept (17.4% and 8.3%, respectively), infliximab (17.4% and 1.9%, respectively), and abatacept (16.7% and 7.0%, respectively) were more frequently used as first-line treatments in the DtT group. Tumor necrosis factor α inhibitors predominated on the first two lines (52.8% versus 24.4%, respectively), baricitinib was never used, and IL-6 inhibitors were significantly less frequently used (19.6% versus 40.6% in the non-DtT group), rituximab (6.8% versus 19.6%), with more frequent switches to another drug class (72.4% and 58.6%).

Conclusion. The conducted analysis identified several significant predictors of treatment resistance in RA: young age at disease onset, positivity for ACPA, erosive nature of arthritis, presence of rheumatoid nodules, and history of the withdrawal of medications due to adverse events. These factors can be utilized for the establishment of risk groups. The prognostic model developed based on the identified predictors may be useful for determining a risk group for treatment resistance. Additionally, several factors related to therapeutic strategies may influence the development of treatment resistance in RA: delayed initiation of DMARD therapy, late commencement of biological therapy, and insufficient use of rituximab, abatacept, interleukin 6 inhibitors, and Janus kinase (JAK) inhibitors in the early lines of therapy for bDMARD /JAK inhibitors.