THU0368 Clinical and imaging characteristics of patients with axial spondyloarthritis reporting flare symptoms immediately prior to next dose in routine tnfi therapy

Background Tumour necrosis factor inhibitors (TNFi) are highly efficacious in axial spondyloarthritis (axSpA) with significant clinical responses mirrored in resolution of MRI determined bone marrow oedema (BMO) lesions representative of active inflammation in the sacroiliac joints (SIJ) and spine. However, symptom flare suggesting loss of response is common with many patients reporting deterioration days or weeks prior to the next dose. We hypothesised that intermittent suppression of inflammation with longer acting TNFis such as adalimumab or infliximab may be associated with a greater likelihood of recurrence of MRI determined BMO lesions than shorter acting agents such as etanercept. Objectives To explore the relationship between symptom flare immediately prior to next treatment dose and recurrence of MRI determined BMO lesions with different TNFi in axSpA. Methods Proof-of-concept, single centre study. Eligible participants were adults with axSpA established on adalimumab, etanercept or infliximab describing loss of response immediately prior to next treatment dose. Loss of response was defined as subjective “flare” or “wearing off” of drug effect before the expected duration of treatment effect. Participants attended at three points: Baseline: first day of treatment cycle (day of drug dose); Endpoint 1: 3–4 days after dose; and Endpoint 2: within 48 hours of next dose. ASDAS-CRP and whole spine and SIJ MR imaging utilising 3T MRI scanner were performed at each visit. Images were scored according to the semi-quantitative Leeds MRI scoring system by two observers blinded to participant identity and date of scan. Results 38 participants (16 adalimumab; 12 etanercept; 10 infliximab) with a total 113 MRI scans were analysed. 71% (n=27) male; 60% (n=23) HLA-B27+ with no differences between the groups; 73.7% (n=28) fulfilled mNYC for AS; the remainder were classified as nr-axSpA. 58% (n=22) had at least 1 Grade 1 BMO lesion at baseline with lesions more commonly seen with longer acting drugs (5/12 in etanercept group; 9/16 in adalimumab group; 8/10 in infliximab group) and 11 (50%) had at least 1 Grade ≥2 lesion (Table 2). There was a trend towards number and severity of BMO lesions fluctuating through the treatment cycle (Figure 1) mimicking subjective loss of response, also reflected on the ASDAS-CRP (Table 1) but this did not reach statistical significance. Conclusions This small proof-of-concept study shows subtle fluctuations on MRI changes of BMO and ASDAS-CRP corresponding to time of subjective “flare” before next scheduled TNFi dose in subjects with axSpA. Although these changes may be more common with the longer acting TNFi our data are not confirmatory. Larger studies are required to explore this concept further as these observations are potentially relevant for disease progression since cycles of inflammation and its subsequent suppression could theoretically increase the risk of new bone formation. Acknowledgements This study was sponsored by Pfizer. Disclosure of Interest R. Ansell Grant/research support from: Pfizer, H. Mathieson: None declared, M. Merashli: None declared, N. Busquets-Perez: None declared, D. McGonagle Grant/research support from: Celgene, Janssen & Pfizer, Consultant for: AbbVie, Celgene, Janssen, Novartis, UCB, Speakers bureau: Abbvie, Celgene, Janssen, UCB., H. Marzo-Ortega Grant/research support from: Janssen & Pfizer., Consultant for: AbbVie, Celgene, Janssen, Novartis, UCB. S, Speakers bureau: Abbvie, Celgene, Janssen, UCB.