Introduction Quality and duration of rituximab therapy response in rheumatoid arthritis have not been totally explained. In the synovium, three studies have indicated a relationship between higher plasma cell numbers at baseline and worse, or shorter lasting responses [1–3]. In this study we measured synovial expression of genes involved in B-cell biology, cell survival, cell trafficking and inflammation.

Methodology Synovial arthroscopic biopsies were collected and snap frozen from patients at baseline and 26 weeks after rituximab. Immunohistochemistry was performed for CD3, CD19, CD20cy, CD138 and CD68 and scored for positive cells/mm² using digital image analysis. The remainder of the tissue was used for RNA extraction. A 48 gene custom Taqman array was designed, including genes for cell lineage markers (CD19, CD20, CD3, CD138 & CD68), BAFF-APRIL system, immunoglobulins, cytokines, chemokines and adhesion molecules or cell trafficking. Informative data are available from 32 biopsies pre-Rituximab treatment and 23 post-treatment. Unsupervised clustering was performed to all gene expression to all patients pre and post rituximab.

Results Clustering of gene expressions revealed 3 main groups pre-rituxmab synovial biopsies, a definite cluster of B cell/survival factors signature associated low expression other genes synovial biopsies was associated with a higher disease activity (DAS28) at baseline and after treatment compared to the other two groups pooled together. This resulted in a lower rate of EULAR Good response.

Comparing EULAR responders and non-responders, there were trends to higher baseline CD20 gene expression in non-responders (p = 0.079) and greater reduction in CXCL13 (p = 0.066) and MMP (p = 0.024) in responders.

Lower expression of ICAM (p = 0.021), FGF (p = 0.044), CD20 (p = 0.055) and p53 (p = 0.025) and higher expression of APRIL (p = 0.029) at baseline was associated with normalisation of CRP after therapy. Furthermore, these patients also showed a significantly greater reduction in expression of CD4, CD55, CD68, CXCL12, EGF, FGF, ICAM, PECAM, STAT5, TGF-beta, APRIL and BAFF (all p < 0.05).

Conclusions Results point to key differences in synovial gene expression in patients with clinical response to rituximab, obviously in relation with genes involved in cell trafficking and survival.

References 1. Teng et al, A&R (2008) 56(12).

2. Thurlings et al, ARD (2008) 67.

3. Vital et al, EULAR 2012.