Dual energy X-ray absorptiometry derived bone quantity in an ovine model of osteoporosis.

Abstract

Dual energy X-ray absorptiometry (DXA) is the most widely used bone quantity investigation in human osteoporosis; it is so ubiquitous that many clinicians now mistakenly equate DXA derived areal BMD with bone strength. Bone strength partly depends on bone mass and the distribution of this mass in terms of size and shape but also depends on bone quality factors of microarchitecture, material properties, accumulated microdamage and remodelling rate. In osteoporosis research, sheep have emerged as a useful lar ge animal model due to size and physiological similarities to humans; and, as a flock animal, large numbers of sheep can be reared at reasonable costs. The aim of this research, part of the ‘Bone for Life’ group of studies, was to evaluate; DXA scanner val idity in the ovine animal model, DXA derived bone quantity (aBMD) change in an ovine animal model of osteoporosis, and following on from these preliminary studies; to examine the influence of DXA derived aBMD on cortical and trabecular bone mechanical prop erties. In initial validation studies of DXA scanner reliability and assessment of natural variation in aBMD in the Irish sheep population ex - vivo bone samples were sourced from heal thy Irish mixed breed ewes (age >5yr) at routine commercial sacrifice. Fur ther samples were also harvested from control (CON) and ovariectomised (OVX) sheep in a specially reared flock at two time points, 12 and 31 month post - ovariectomy, to determine the short - and long - term effect of ovariectomy on bone quantity (aBMD); and in further collaborative studies, bone quantity and quality relationships to mechanical properties. The DXA scan protocol consisted of standardised alignment of the whole sample (WB) on a holding jig within a tank filled with normal saline, followed by basel ine repeat data acquisition scans using a standard algorithm. The same algorithm allowed further analysis (SRA) of six sub - regions down the length of selected long bones femur (FMR), tibia (TBR), humerus (HMR) and radius (RDR); and, the six vertebrae spina l samples (LS); as well as smaller region of interest (ROI) analyses in areas of high trabecular bone content. Statistical analysis was made of initial repeat test data to determine DXA scanner reliability, establish measurement limits; and assess biologic al variability; and for the specially reared flock, inter - group differences in aBMD were analysed for effect of ovariectomy, and then re - analysed in conjunction with bone quality variables to assess the strength of aBMD and bone quality relationships to me chanical test data. xxiii The results of preliminary studies confirmed an acceptable level of; repeatability (ICC >0.9), reliability (%TEM 3 - 5%) and measurement limits (95%LOA ± 0.035 g.cm - 2 ) of the adapted DXA scanning protocol for large ovine bone samples. In a representative sample of the Irish ewes evaluation of biological variability revealed; an aBMD range of approximately 0.60 to 1.20 g.cm - 2 , minimal inter - breed variation and a small seasonal effect toward lower density in Spring months at some sites. Ana lysis of inter - group differences between CON and OVX detected no significant difference in aBMD post - ovariectomy at any spinal or long bone site at 12 month; but did detect a significant reduction in aBMD in the proximal humeral region and an unexpected fi nding of increased aBMD in the mid - region of the tibia, both with large effect size, at the 31 month time point. In follow on studies; aBMD (r 2 ~0.83) and then structural geometric properties of CSMI (r 2 ~0.75) and section modulus (r 2 ~0.65) had the stronges t association with breaking strength (Fmax) of ovine radii in three point bending. In microdamage analysis of radii exposed to cyclic loading, and stratified as high, medium or low aBMD, results also suggested significantly higher microcrack numerical and surface density in bones with lower aBMD. Analysis of compressive strength of isolated trabecular cores was less clear, but significant inter - group differences in trabecular thickness of large effect size were detected in the CON vs. OVX comparison at the 31 month time point, and in terms of relationships to strength, trabecular number and spacing (r 2 ~0.4 to 0.5) were better related to compressive strength and modulus than any DXA or microCT derived bone quantity variable. In conclusion this work has further validated DXA bone quantity measurement and defined measurement limits for scanning and sub - regional analysis of whole bone samples in this animal model. Bone quantity data for Irish s heep were of a similar range to those seen in human female populations, and exhibited some seasonal variation again similar to human populations living in northern latitudes. In this animal model ovariectomy alone however , did not prove a sufficient enough stimulus in the short - term to produce a significant reduction in bone quantity detectable by DXA in time frames suitable for research; however longer term changes detected in the proximal humerus and tibia require further evaluation. In terms of relations hips to strength DXA derived aBMD data was the better predictor of bending strength in a typical long bone mid - shaft site ; but in compression of isolated trabecular cores, bone quality factors were more closely associated to mechanical properties.