POS0511 USING HIP STRUCTURAL ANALYSIS MEASUREMENTS TO PREDICT FRACTURE IN RHEUMATOID ARTHRITIS

Abstract

Background:Rheumatoid arthritis (RA) is a chronic inflammatory autoimmune disease and increases the risk of developing osteoporosis. Incidence of fracture is higher in this group of patients compared to the general population and can lead to increased morbidity (1). Bone strength of the proximal femur is not only linked to bone mineral density; it also depends on the geometric properties of the bone mass (2). Hip structural analysis (HSA) is a technique used to assess hip bone structure that takes geometric measurements of the femur from dual-energy X-ray absorptiometry (DEXA) images (3).Objectives:To determine whether HSA measurements help predict fracture in patients with RA.Methods:Data were collected from June 2004 to August 2017 from RA patients who underwent a DEXA scan at a District General Hospital. This included hip axis length (HAL), cross-sectional area (CSA), cross-sectional moment of inertia (CSMI), distance from centre of femoral head to centre of femoral neck (D1) and to inter-trochanteric line (D2), mean femoral neck diameter (D3), shaft angle (A) neck/shaft angle (Θ) and proximal femur strength index (SI) and distance from centre of mass of femoral neck to superior neck margin (Y). Fracture was predicted by a series of binomial logistic regression models, adjusted for sex, age and bone mineral density (BMD). Odds ratios with 95% confidence intervals and area under the receiver operating characteristic curve (AUC) were calculated.Results:2077 patients with RA were identified, 1632 were female and the mean age was 66.7. HAL, D1, D2, D3, A, Θ and Y were not significant predictors of fracture in regression models; odds ratios are included in table 1. CSA, CSMI and SI predicted fracture risk. The AUC for CSA, CSMI and SI regression models were 0.632, 0.609 and 0.625 respectively.Table 1.Odds ratios of fracture for different HSA parameters in RA patientsHSA ParameterOdds Ratio (95% Confidence Interval)HAL1.01410 (0.99958 - 1.02883)CSMI0.99994 (0.99990 - 0.99998)CSA0.98523 (0.98065 - 0.98982)D11.01683 (0.98925 - 1.04518)D21.01286 (0.99886 - 1.02705)D31.00664 (0.96958 - 1.04511)Y1.04580 (0.98633 - 1.10886)A1.00898 (0.98878 - 1.02959)Θ1.00276 (0.98672 - 1.01906)SI0.56769 (0.43400 - 0.74258)Figure 1.Receiver operating characteristic curves for CSA (red), CSMI (green) and SI (blue). AUC for CSA was 0.632, CSMI-0.609 and SI-0.625.Conclusion:These data suggest that CSA, CSMI and SI help predict the fracture risk in patients with RA. HAL, D1, D2, D3, A, Θ and Y do not predict risk of fracture. The CSA regression model was the strongest predictor of fracture. HSA measurements can therefore help predict risk of fracture in conjunction with other factors. Limitations of the study are that it was retrospective and only studied patients who had a DEXA scan.