Prevalent role of porosity and osteonal area over mineralization heterogeneity in the fracture toughness of human cortical bone

Abstract

Changes in the distribution of bone mineralization occurring with aging, disease, or treatment have prompted concerns that alterations in mineralization heterogeneity may affect the fracture resistance of bone. Yet, so far, studies assessing bone from hip fracture cases and fracture-free women have not reached a consensus on how heterogeneity in tissue mineralization relates to skeletal fragility. Owing to the multifactorial nature of toughening mechanisms occurring in bone, we assessed the relative contribution of heterogeneity in mineralization to fracture resistance with respect to age, porosity, and area fraction of osteonal tissue. The latter parameters were extracted from quantitative backscattered electron imaging of human cortical bone sections following R-curve tests of single-edge notched beam specimens to determine fracture toughness properties. Microstructural heterogeneity was determined as the width of the mineral distribution (bulk) and as the sill of the variogram (local). In univariate analyses of measures from 62 human donors (21 to 101 years), local but not bulk heterogeneity as well as pore clustering negatively correlated with fracture toughness properties. With age as covariate, heterogeneity was a significant predictor of crack initiation, though local had a stronger negative contribution than bulk. When considering all potential covariates, age, cortical porosity and area fraction of osteons explained up to 50% of the variance in bone׳s crack initiation toughness. However, including heterogeneity in mineralization did not improve upon this prediction. The findings of the present work stress the necessity to account for porosity and microstructure when evaluating the potential of matrix-related features to affect skeletal fragility.

Introduction

The fracture resistance of bone does not solely depend on the quantity of bone, or bone mineral density, but also on the integrity of the bone tissue (Donnelly et al., 2014). In particular, owing to its hierarchical organization, bone is able to resist fracture through the combination of multiple toughening mechanisms that interact at several length scales. For example, interfibrillar sliding at the nanoscale allows the mineralized collagen fibrils to deform without failing, while osteons on the order of hundreds of microns create natural barriers to the propagation of cracks through bone tissue (O’Brien et al., 2003, Ural and Vashishth, 2014). With aging and disease progression, changes can occur at any of the hierarchical levels of organization, thereby lowering the fracture resistance of bone (e.g., accumulation of non-enzymatic collagen crosslinks impeding interfibrillar sliding). Complicating the clinical assessment of fracture resistance or fracture risk, heterogeneity exists at the multiple length scales of organization, but not all heterogeneity may significantly contribute to the age-related loss of fracture resistance.

Investigations of fracture processes in material science indicate that microstructural heterogeneity (i.e., the spatial variation of compositional properties in a material) improves resistance to fracture: crack propagation is straight in a homogenous material with less energy dissipation while it is tortuous deflecting at interfaces in a heterogeneous material, thereby requiring more energy to grow (Dimas et al., 2014, Hossain et al., 2014). Changes in bone metabolism associated with diseases or drug treatment are known to significantly affect the distribution of bone mineralization at the micron length scale (Roschger et al., 2008). In this context, a loss in heterogeneity in mineralization has been posited as a possible factor contributing to an increase in fracture risk (Ettinger et al., 2013). However, in the current literature, there is a lack of consensus on how heterogeneity in tissue mineralization relates to skeletal fragility. Indeed, heterogeneity in mineralization at the micro-structural level was found either to be greater (Bousson et al., 2011) or lower (Gourion-Arsiquaud et al., 2013, Milovanovic et al., 2014) in treatment naïve, hip fracture cases compared to fracture-free women. The inconsistency of these results highlights the need to account for other contributing factors to conclusively state that bone fragility is directly related to the distribution of tissue mineralization.

Among the factors that contribute to skeletal fragility, vascular porosity is greater in patients with a history of fracture compared to non-fracture controls (Ahmed et al., 2015, Bala et al., 2014, Bell et al., 1999, Milovanovic et al., 2014, Shigdel et al., 2015) or in bone specimens with lower fracture toughness properties, as determined from mechanical tests of cadaveric tissue (Granke et al., 2015, Yeni et al., 1997). Moreover, image analysis of bone cross-sections of the femoral neck suggests that greater cortical porosity in fragile bone may result from the merging of spatially clustered remodeling osteons (Bell et al., 2000, Jordan et al., 2000). Therefore, not only the extent but also the spatial distribution of porosity may decrease the fracture resistance of bone, a phenomenon described for porous media (Bilger et al., 2005).

To date, the question of whether heterogeneity mineralization affects bone׳s resistance to fracture remains undetermined for two reasons. First, published data is confined to comparing tissue composition between fracture and control cases, but to date, no studies have attempted to relate heterogeneity in mineralization and the mechanical properties of cortical bone. Second, most studies investigating associations between tissue composition and fracture status or risk do not adjust for other microstructural features (i.e., porosity or bone volume fraction). Given the possible contribution of age, porosity, and tissue microstructure to skeletal fragility, we aimed to determine whether microstructural heterogeneity in mineralization significantly explains bone׳s ability to resist fracture after adjusting for these factors. To address this, we analyzed cross-sections of human cortical bone specimens for which fracture toughness properties had been previously determined (Granke et al., 2015). Upon imaging these cross-sections by quantitative backscattered electron imaging (qBEI), image processing techniques quantified lacunar and vascular porosities, pore clustering, population and local heterogeneity in mineralization, as well as area fraction of osteonal tissue.