Background Chronic inflammatory activity in juvenile idiopathic arthritis (JIA) affects the bone, possibly resulting in growth abnormalities and decreased bone mineral density (BMD). Monitoring the bone status of JIA patients is important for the prevention of fractures and osteoporosis at later age. Dual-energy X-ray absorptiometry (DXA) is most widely used to determine BMD. BoneXpert is a new, feasible and reproducible method for automatic determination of cortical BMD on hand radiographs. Moreover radiation exposure is low and in low-risk peripheral areas.
Objectives The aim of this study was to compare BoneXpert and DXA in the assessment of BMD in JIA patients.
Methods Thirty-five JIA patients with available DXA and hand radiograph within a five month time period were included from a tertiary hospital of the Dutch Arthritis and Biologicals in Children register. Outcome measures for BMD were Bone Health Index (BHI) from BoneXpert and BMD total body, BMD lumbar spine and Bone Mineral Apparent Density (BMAD) from DXA. For all outcome measures Z-scores, reflecting the standard deviations compared to a healthy population, were calculated. Correlations between BMD measurements by DXA and BoneXpert were assessed with Pearson correlation coefficients.
Results The patients in this study had significantly lower mean BMD compared to the healthy population on all BMD measures (p<0.05). The pearson correlation coefficient for the absolute scores was 0.568 (p=0.000) for BMAD vs BHI, 0.665 (p=0.000) for BMD total body vs BHI, and 0.770 (p=0.000) for BMD lumbar spine vs BHI. The correlation between BMAD and BHI is depicted in Fig. 1. The correlation coefficient for the Z-scores of DXA and BoneXpert (0.127-0.322) was not significant.
Conclusions BHI measured by BoneXpert is correlated to measurements of BMD by DXA. The correlation of Z-scores of BMD measured by the two methods is weaker. Longitudinal studies and assessment of the association of the BMD measurements with outcome (for instance atraumatic fractures) are necessary to further determine the value of BoneXpert in clinical use.
Acknowledgements The authors wish to acknowledge Hans Henrik Thodberg for his support with and the free use of BoneXpert.
Disclosure of Interest C. Nusman: None declared, J. Anink Grant/research support: Travel grant Pfizer, M. van Rossum Grant/research support: Travel grant Pfizer, R. van Rijn: None declared, M. Maas: None declared, L. van Suijlekom-Smit Grant/research support: For the submitted work: Dutch Board of health insurances, Pfizer and Abbvie.