CO:10:3 | Clinical response to JAK inhibitors or non-TNFi-targeted biologicals according to synovial tissue characteristics in patients with difficult-to-treat rheumatoid arthritis

Mariangela Salvato¹, Alessandro Giollo¹, Francesca Frizzera¹, Kiren Khalid¹, Lorenzo Di Luozzo¹, Maria Capita¹, Marianna Tamussin¹, Carlo Garaffoni², Marny Fedrigo³, Annalisa Angelini³, Ettore Silvagni², Andrea Doria¹. | ¹Rheumatology Unit, Department of Medicine - DIMED, University of Padova; ²Rheumatology Unit, Department of Medical Sciences, University of Ferrara and Azienda Ospedaliero-Universitaria S. Anna; ³Cardiovascular Pathology, Department of Cardiac, Thoracic, Vascular Sciences and Public Health, University of Padova, Italy

Background. Patients meeting the EULAR difficult-to-treat rheumatoid arthritis (D2TRA) definition are classically non-responders to TNF inhibitors (TNFi). As such, they can be treated with biological drugs with other mechanisms of action (OMA; tocilizumab, sarilumab, abatacept, rituximab) or Janus Kinase inhibitors (JAKi). The aim of the study was to investigate whether characteristics of synovial tissue may predict treatment response to OMA or JAKi in D2TRA patients.

Materials and Methods: The MATRIX-D2T was a prospective cohort, multi-centre study that included RA patients with moderate to high disease activity undergoing ultrasound-guided synovial tissue biopsy (UGSTB) at two institutions. In this post-hoc analysis, patients were eligible for inclusion if they fulfilled the EULAR D2TRA definition. The primary outcome was a change in CDAI after six months; the secondary outcome was low disease activity (LDA or CDAI<=10). Predictors were synovial grading according to Krenn Synovial Score (2-4, low grade; 5-9, high grade); synovial pathotypes (lympho-myeloid, diffuse-myeloid, pauci-immune fibroid); and the count of lymphoid aggregates. Outcome comparisons between treatment groups (JAKi vs. OMA) were assessed through repeated measures ANOVA corrected for ACPA titre.

Results. Among 132 patients undergoing USGTB during the study period, 26 patients met the study criteria and were included (mean (SD) age 58 (11) years, duration 15 (10) years, females 62%). Twelve patients initiated JAKi and 14 OMA; the two groups were comparable for baseline characteristics (Table 1). The median (IQR) number of failed b/tsDMARDs was 3 (3), and for csDMARDs, it was 2 (3). Median (IQR) KSS was 5 (7). After 6 months, the mean (SD) CDAI significantly decreased more in JAKi than OMA (mean difference -10.7, 95% CI -19.4 to -2.0, p=0.018). Compared to OMA, a stronger CDAI decrease with JAKi was shown in patients having high-grade synovitis only (Figure 1a-b), with no clear association with synovial pathotypes or the presence of lymphoid aggregates. Patients in LDA were numerically more frequent in JAKi recipients (n=6/12, 50%) compared to OMA (n=3/14, 21.4%; p=0.218; Figure 1c).

Conclusions. Our results suggest that UGSTB may help predict clinical response to JAKi or OMA in DT2RA patients. However, the sample size was small, and results must be confirmed in larger independent cohorts.