Relating micromechanical properties and mineral densities in severely suppressed bone turnover patients, osteoporotic patients, and normal subjects

Abstract

Mineralization of bone, from the tissue level to whole bones, is associated with mechanical properties. The relationship between bone tissue mineralization and micromechanical properties may be affected by age, disease, and drug treatment. Patients with severely suppressed bone turnover (SSBT) suffered atypical fractures while on bisphosphonate treatment. The role of tissue level mineralization in predicting material level properties of SSBT bone may be different from that of other osteoporotic patients and of normal subjects. The aim of this study was to compare the relationships between mineralization and micromechanical properties of bone biopsies from patients with SSBT, bisphosphonate-naive osteoporotic patients with typical vertebral fracture, and normal young and age-matched subjects. We used nanoindentation and quantitative backscattered electron microscopy to characterize the elastic modulus, contact hardness, plastic deformation resistance, and tissue mineralization of the biopsies at site-matched locations within each biopsy. The linear mineralization-mechanical property relationships were different among the groups with respect to the intercepts for only cortical bone tissue but not the slopes for cortical and trabecular bone tissues. For a given mineral density, there was a trend of greater plastic deformation resistance in SSBT cortical bone compared to young normal bone. Similarly, there was a trend of greater plastic deformation resistance in osteoporotic trabecular bone compared to young normal bone for a given mineral density. The age-matched normal group had higher elastic modulus and a trend of higher contact hardness compared to the young normal group for a given mineral density. However, the mechanical property-mineralization relationships within an individual were weak, and only 21 of 53 biopsies that were analyzed had at least one significant association between mineralization and a mechanical property measurement for either cortical or trabecular bone tissues. The average properties of microstructural regions (deep and superficial remodeling packets in trabecular bone; osteonal and interstitial regions in cortical bone) were consistent with mineral accumulation with tissue age, with the exception of the SSBT group. SSBT trabecular bone deep packets had higher hardness and plastic deformation resistance than superficial packets, but mineralization levels and tissue modulus were not different between packet types. We conclude that relationships between mineral and mechanical properties were different between fracture and normal groups and between young and old normal groups, and that atypical fracture may be associated with changed microstructural material properties and tissue level mineralization compared to osteoporotic patients with vertebral fracture and normal subjects. We hypothesize that tissue level bone quality may be an important determinant in fracture risk, such that tissue mineral density may predict different material properties in different patient groups.

Fig. 1

Representative nanoindentation load-displacement curve for 1 μm indent. Berkovich diamond tip loaded into the surface at a target strain rate of 0.05 s⁻¹ to 1 μm depth. Tip was held at maximum load for 10 seconds, unloaded and held at 10% of maximum load for 100 seconds to calculate thermal drift. Elastic modulus and contact hardness were calculated from the unloading portion of the curve.

Fig. 2

A 4 by 5 array of indents made in a cortical bone region of a biopsy from a young normal subject. Secondary electron images (600×) were captured to identify indent locations (a). Corresponding images were captured in backscattered electron mode for mineralization quantification (b). Circles mark the 2 indents that were discarded from the analysis due to cracks at those locations. Indent marked by the square corresponds with the image analysis example in Fig 3. Scale bar represents 50 μm.

Fig. 3

A magnified image of an indent impression from the young normal biopsy array in Fig 2b (a). Each square represents one image pixel of the 50 by 50 pixel image. The triangular indent impression is visible in the backscattered image and influences the measurement of mineral density. To estimate the mineralization at the indent location, a 24 by 24 pixel region encompassing the indent impression is excised, and this region is reconstructed using a thin-plate cubic spline fit to the surrounding 50 by 50 pixel region. The 50 by 50 pixel region with the fitted pixel intensities is shown in (b). The mineralization was calculated from the average gray-level of the fitted 24 by 24 region. Gray-level contrast of these images was enhanced to demonstrate the effect of the fitting algorithm. Pixel resolution is 0.138 μm.

Fig. 4

Simple linear regressions between mineralization and elastic modulus (E), contact hardness (H_c), and plastic deformation resistance (H), corresponding with an indentation array in the cortical tissue of a young normal biopsy (Fig 2). All three relationships were not significant (p>0.05).

Fig. 5

Indentations made in cortical and trabecular bone tissues were classified by region in order to analyze whether regional differences in mean properties were the same among groups. Trabecular bone indents were located in superficial (S) or deep (D) packets, the former were packets that had a portion touching the surface of the trabecula. Cortical bone indents were located in osteonal (O) or interstitial (I) regions. The lines drawn on the backscattered images delineate between regions (300×, scale bar = 100 μm).

Fig 6

Average modulus (E), contact hardness (H_c), plastic deformation resistance (H), and mineralization (Z) of each region (osteonal, interstitial; superficial, deep) for each biopsy plotted by group. E, H_c, and H were greater in interstitial regions compared to osteonal regions, and greater in deep packets compared to superficial packets (ANOVA, Region effect p<0.05; Table 4). Cortical mineral density was not different by region. Trabecular mineral density was greater in deep packets compared to superficial packets, except in SSBT (ANOVA, Region*Group interaction p=0.07; Table 4). Error bars indicate plus or minus one standard deviation. YN=young normal, AMN=age-matched normal, OP=osteoporotic, SSBT=severely suppressed bone turnover.