Synovial Markers in Refractory Rheumatoid Arthritis: Distinguishing Non-Inflammatory and Persistent Inflammatory Phenotypes

Background: Refractory rheumatoid arthritis (RA) is an emerging clinical condition with two proposed patterns: Non-inflammatory refractory RA (NIRRA) and persistent inflammatory RA (PIRRA). NIRRA is associated with higher comorbidities, obesity, and painful syndromes, while PIRRA is characterized by musculoskeletal ultrasound (MSUS)-confirmed active inflammation or elevated inflammatory markers. However, MSUS and inflammatory markers do not clearly define NIRRA or PIRRA, and the association with underlying synovial pathology remains unclear. Aims: This study aimed to compare synovial inflammation, immunohistochemistry (IHC), and the main synovial pathotypes (lymphoid, diffuse myeloid, and fibroid/pauci-immune) in patients with NIRRA, PIRRA, and non-refractory RA. Materials and methods: This prospective, observational cohort study included patients with established RA unresponsive to standard therapy, conducted between January 23 and January 10, 2024. All patients had active disease (clinical disease activity index > 10). MSUS-guided synovial biopsies were performed following EULAR-OMERACT guidelines. Joint inflammation was graded (0 to 3) using OMERACT, and the Krenn Synovial Score (KSS) was calculated. Patients' perceptions were assessed using patient-reported outcomemeasures, including SF-36, HAQ-DI, and FACIT-F. RISULTATI Of the 47 patients who underwent US-guided synovial biopsies, 38 were analyzed (68.6% females, median age 59 years). The median disease duration was 5 years. PIRRA was diagnosed in 5 (13.2%) patients, NIRRA in 10 (26.3%), and non-refractory RA in 23 (60.5%). Compared to PIRRA, NIRRA patients had lower MSUS scores, fewer tender joints, lower infl ammatory markers, and lower physician global assessmentscores. NIRRA patients had lower SF-36 scores in most domains, while PIRRA patients had the best emotional domain scores, even surpassing non-refractory patients. FACIT-F and HAQ-DI scores did not differ significantly among the groups. Body weight and comorbidities were similar except for fibromyalgia and depression, more common in NIRRA and non-refractory patients than in PIRRA. Synovialhistology showed the lowest KSS in NIRRA patients. Synovial hyperplasia, density, and infiltrates were lowest in NIRRA, while density was highest in PIRRA. PIRRA patients had the highest CD20+ and CD138+ scores, while NIRRA had lower scores for all IHC markers. The fibroid/pauci-immune pathotype was most common in NIRRA patients (50%) compared to 0% in PIRRA and 17.4% in non-refractory patients. The lymphoid pathotype was more frequent in PIRRA (60%) and non-refractory (61%) patients. Conclusions: NIRRA patients exhibited a more prevalent fibroid/pauci-immune pathotype, while PIRRA patients showed a lymphoid pathotype. These findings suggest a mechanistic basis for categorizing refractory RA into two distinct groups.