Background: Response to Rituximab (RTX) varies significantly between SLE patients. Ethnicity may play a role in these differences, and a possible relationship has been suggested between the clinical response to RTX and the presence of certain auto-antibodies (ab) (anti-ENA and anti-dsDNA ab) and C3 levels at baseline.

Objectives: The aim of this study was to identify biomarkers that could predict the response to RTX treatment in SLE patients.

Methods: This was a cross-sectional study of 121 SLE patients treated with RTX in UCLH between 2000 and 2016. Demographic, clinical and serological data were analysed. Disease activity was evaluated using the BILAG index. Patients were categorised as ‘Responders’ if all or some of the As and Bs from the BILAG score at the time the RTX was given were lost at 6 and at 12 months, and as ‘Non-Responders’ if none of the As and Bs were lost. Relapse after RTX treatment was defined as development of a new BILAG Grade A or B in any system. A uni and multivariate regression analysis were performed to identify predictive factors of response to RTX utilising a combination of clinical and biological markers.

Results: At 6 and at 12 months, 85% and 70% respectively of our patients had responded clinically to the RTX treatment. 24% of patients relapsed during the year after RTX. In the univariate analysis, constitutional symptoms at diagnosis (crude OR (95% C); 5.66 (1.53–20.88), p=0.009) and the absence of musculoskeletal disease at the time of RTX (0.27, (0.09–0.81), p=0.019) were related to response at 6 months. In the multivariate analysis, both remained significant, (adjusted OR (95% CI]): 5.33 (1.39–20.41), p=0.014 and 0.26 (0.08–0.81), p=0.021 respectively. With respect to the response at 12 months, in the univariate analysis the presence of arthritis as the main indication for RTX (3.16 (1.31–7.58), p=0.010), the absence of renal disease at diagnosis (0.36 (0.15–0.86), p=0.022) and of cardiorespiratory disease at the time of RTX (0.29 (0.09–0.89), p=0.031), less than one anti-ENA ab (0.28 (0.12–0.66), p=0.003), low levels of C3 at diagnosis (0.29 (0.09–0.89), p=0.031), increased anti-dsDNA ab levels (0.38 (0.17–0.89), p=0.025) and decreased C3 levels (0.27 (0.11–0.63), p=0.002) before RTX were related to the response. In the multivariate analysis, only the absence of more than one anti-ENA showed significance (0.30 (0.11–0.82), p=0.020). Having more than one anti-ENA was related to relapse (3.30 (1.36–8.05), p=0.009), while having arthritis as the main indication for RTX therapy was associated with a lower risk of flare (0.26 (0.10–0.64), p=0.004). On the multivariate analysis, having arthritis remained significant (0.29 (0.11–0.75), p=0.010).

Conclusions: There is a relation between the presence of more than one anti-ENA ab and a worse response to treatment at 12 months and a higher risk of flaring. Having arthritis at the time of RTX leads to a negative response at 6 months but a lower risk of flare before 1 year.

Disclosure of Interest: None declared

DOI: 10.1136/annrheumdis-2018-eular.2852

Oral presentation-3