AB0904 PERSISTENCE AND REASONS FOR DISCONTINUATION OF DENOSUMAB IN PATIENTS WITH RHEUMATOID ARTHRITIS

Background: Denosumab, a monoclonal antibody of receptor activator of NF-κB ligand promotes a strong action for bone mineral density (BMD) gain. This agent is often used for patient with rheumatoid arthritis (RA) because of its strong anti-osteoclastogenesis action, with that joint structural damage is induced. However, factors what affects BMD gain for patient with RA is still unclear. Objectives: Factors that may affect BMD gain for patient with RA is evaluated statistically. Methods: Patients with RA to whom denosumab is administrated consecutively three shots or more were picked up. BMD in lumbar spine (LS) and femoral neck (FN) measured with dual-energy X-ray absorptiometry was routinely measured at the initial administration (BL). BMDs were measured every six months when another denosumab is administrated. Change of BMD for each bone was calculated. Patient’s age at BL, at onset of RA, disease duration, sex, anti-cyclic citrullinated polypeptide antibodies (ACPA), whether denosumab is naive, body mass index (BMI) at BL were harvested. BMD in each bone, serum tartrate resistant acid phosphatase 5b (TRACP5b), total type one procollagen-N-peptide (P1NP), calcium (Ca), creatinine (Cr), cystatin C (CysC), estimated glomerular filtration ratio based on CysC (eGFR), serum Cr-to-CysC ratio (Cr/CysC), and Barthel Index, were measured at BL and every six months thereafter. Relationship between BMD gain from BL to second administration and such like factors at BL were evaluated with linear regression analysis at first with univariate model and then multivariate model with factors that demonstrated statistical significance within 5%. Binary logistic regression analysis for these factors was also performed according to BMD gain. These procedures were performed as a same manner regarding with BMD gain from BL to third administration. Results: A total of 397 patients with 43 males (10.4%) and 354 females (89.6%) were recruited. Average age was 81.3 and average disease duration of RA was 6.9years. 227 patients (57.4%) was denosumab naive, and prior to BL, 170 patients were already administered with alendronate in 26, risedronate in 26, minodronate in 23, ibandronate in 12, raloxifene in 39, bazedoxifene in 7, teriparatide in 36. BMD gain in LS from BL to the second administration demonstrated significant correlation with age and TRACP5b at BL with univariate model, and only aging correlated significant negative correlation with BMD gain with multivariate model. In binary logistic regression analysis, aging demonstrated no significant regression with BMD gain. From BL to third administration, BMD gain also demonstrated significant correlation with aging, but no correlation with TRACP5b, but Cr/CysC at BL. These two factors also demonstrated significant correlation with BMD gain in LS, in these aging demonstrated negative and Cr/CysC demonstrated positive correlation. Cr/CysC demonstrated significant regression with BMD gain in LS from BL to the third with binary logistic regression analysis. BMD gain in FN from BL to the second demonstrated significant correlation with age, BMD in FN at BL, and ACPA with univariate model, and all of the three demonstrated significant correlation with BMD gain with multivariate model. However, no factors demonstrated significant regression with BMD gain with binary logistic regression analysis. From BL to third administration, BMD gain in FN demonstrated significant correlation with aging and BMD in FN at BL. However, BMD in FN at BL demonstrated the only factor to correlate with BMD gain in FN. BMD in FN at BL demonstrated significant regression with BMD gain in FN from BL to the third with binary logistic regression analysis. Conclusion: These results suggested that BMD gain in LS and FN was affected by different factors. These results may be helpful reference in choosing denosumab against osteoporosis in RA patient. Disclosure of Interests: None declared