P246 Predicting non-response to biologic therapy amongst patients with axSpA: results from the British Society for Rheumatology Biologics Register

Abstract

Abstract Background Biologic therapies have transformed treatment for axial spondyloarthritis (axSpA). However, although studies report overall benefits, these are average effects. There remains a subset of patients in whom response is not achieved. Here, we aimed to identify characteristics of patients who may need additional therapeutic approaches to optimise outcome. Methods The British Society for Rheumatology Biologics Register for Ankylosing Spondylitis (BSRBR-AS) is a prospective cohort of axSpA patients recruited from 83 centres across Great Britain. All patients were biologic-naïve at recruitment, however those in the “biologic” cohort commenced a biologic therapy shortly thereafter, or during follow-up. Clinical data was collected from medical records, and socio-economic/patient reported outcomes via questionnaires. Response was assessed at first follow-up, between 10 weeks and 9 months from therapy commencement, and defined in four ways: ASAS20 and ASAS40 criteria, ≥1.1 reduction in ASDAS, and achieving moderate/inactive ASDAS (<2.1). Factors associated with non-response were assessed by logistic regression and parsimonious models identified using stepwise methods. The ability to predict non-response was assessed by positive predictive value (PPV). Results 335 biologic participants provided information at a median follow-up of 14 weeks (inter-quartile range (IQR) 12-17). Median age was 47 years (IQR 36-56), 69% were male and 61% met AS modified New York criteria. The proportion meeting response varied by criteria: ASAS20 52%, ASAS40 33%, ASDAS reduction 47% and ASDAS <2.1 35%. Socio-economic circumstances predicted non-response, specifically (in all models) work status and (in some models) fewer years of education (Table 1). Poorer mental health and high number of co-morbidities was associated with non-response across multiple (but not all) outcomes, while body mass index, enthesitis and gender were included in models for a single outcome. Disease-specific factors were largely not associated with non-response. All models demonstrated a good level of fit and were effective at predicting non-response (PPV 65%-77%). Conclusion We have identified factors which predict non-response to biologic therapy, some of which may be modifiable and others which identify patients who are unlikely to benefit from biologic therapy alone. In such patients additional/alternative treatment strategies should be considered to maximise the benefits which others gain from biologic therapy. Disclosures L.E. Dean None. E. Pathan Other; E.P. has recieved salary funding from Jansen (2019) and Merck (2018). G.T. Jones None. G.J. Macfarlane None.