Osteocyte Lacunar Biomarkers in Human Rare Bone Diseases

Abstract

Osteoporosis is a metabolic disease that is characterized by an increased risk of skeletal fracture. While typically occurring in postmenopausal women, rare forms of the disease also manifest in younger premenopausal women such as idiopathic osteoporosis (IOP) and idiopathic low bone mineral density (ILBMD). The weakening of the skeletal system and subsequent increased fracture risk can be explained by a reduction of overall bone density as well as changes in the bone tissue microarchitecture. However, open questions remain regarding if and how the disease impacts bone on a cellular level – specially with respect to the morphology of the lacunar pores and their resident osteocytes. With recent developments in ultra-high-resolution micro-computed tomography (microCT), it is now possible to image embedded three-dimensional (3D) structures with a resolution on the order of one micrometer. This technology is uniquely positioned to image the cellular structures of ex-vivo bone biopsies, such as lacunae, with the resolution and scale of synchrotron computed tomography (SR-CT) without requiring an entire beamline facility. However, a rigorously validated methodology to image lacunar pores using ultra-high-resolution microCT remains to be developed. Therefore, this thesis has been divided into three aims: (i) To develop a method to perform large-scale quantification of osteocyte lacunar morphological biomarkers (ii) To develop a method for the in silico quantification of osteocyte lacunar mechanical biomarkers (iii) To create osteocyte lacunar biomarkers for the characterization of human rare bone diseases. With respect to the first aim, a method was developed to image, extract, and measure lacunar morphometries. First, a novel sample preparation methodology was developed to machine embedded human bone biopsies into a cylindrical format that fits within the physical constraints of the microCT device. Image acquisition was next optimized to create high quality images with clearly visible lacunae, which involved parameters such as x-ray beam energy and filter selection. Following image acquisition, a sample-specific imaging threshold was then applied to binarize the image and segment lacunar structures. Finally, 3D morphometric parameters were measured in both cortical and trabecular regions of a group of 31 transiliac biopsies originating from healthy premenopausal women. Lacunar morphometric parameters were categorized as global, local, and population-based measurements. Rigorous validation was performed for the entire methodology and included measures of lacunar detection accuracy (true positives (TP), false positives (FP), false negatives (FN)), reproducibility (precision error (PE) and intraclass correlation coefficients (ICC)), and biological sensitivity (detectible lacunar differences between regions of cortical and trabecular bone). The second aim investigated the biomechanical significance of the embedded lacunar structures. Two computational models were created that simulate microcrack initiation and propagation through mineralized bone tissue, which ultimately intersected several lacunae. These models were constructed by combining the microCT image data with micro-finite element analysis (microFE). The first model, called the microcrack gradient model (MGM), was based purely on 3D mechanical gradients derived from the original microCT image density values. The second model, named the scissor model (SM), explored the hypothetical release of interstitial fluid from the dense network of canaliculi as a microcrack propagates through the tissue matrix. Both models had unique advantages and disadvantages, however, the microcracking simulated with the SM resembled previous experimental results more closely than with the MGM. The third and final aim of this thesis was to apply the validated imaging methodology developed in the first aim to a large set (103 in total) of three cohorts of transiliac bone biopsies: healthy cohort (n=39), IOP cohort (n=45), and ILBMD cohort (n=19). This was a true large-scale application of the imaging methodology, and the analysis included a total of 26.2 million lacunae. While lacunar morphometries did not significantly differ between cohorts, lacunae in cortical bone were significantly larger, flatter, and more densely packed together than in trabecular regions. This data corroborated the results of the lacunae examined in the first aim, which displayed the same trend of lacunar morphometries between regions in a healthy cohort. Across all three cohorts, lacunar sphericity was strongly negatively correlated with bone volume (BV/TV). While only a cross-sectional study, this strong correlation indicates that lacunar morphology is indeed connected with BV/TV. Finally, the same analysis was applied to the healthy cohort (n=39), but grouped according to adiposity (nlow = 13, nmid = 14, nhigh = 12). Similar to the first biopsy grouping, lacunar morphology differed between cortical and trabecular regions, but not between adiposity tertiles. Furthermore, a strong negative correlation persisted regarding lacunar sphericity and BV/TV. To conclude, the application of the large-scale lacunar imaging methodology using microCT was successful and improved our understanding of lacunar morphology in premenopausal women suffering from rare forms of osteoporosis. Firstly, the imaging methodology was demonstrated to be able to measure lacunar morphometries accurately, reproducibly, and sensitively in human transiliac bone biopsies using microCT technology. Secondly, the biomechanical importance of lacunar structures in the mineralized bone matrix with respect to simulated microcracking was explored and the importance of the canalicular network was highlighted. Finally, the validated lacunar imaging methodology was applied to both healthy vs. affected cohorts (healthy, IOP, ILBMD) as well as to the healthy cohort alone but divided into tertiles based on adiposity. Ultimately, the development, validation, and application of this large-scale lacunar imaging methodology using widely accessible microCT technology has two main accomplishments: it has revealed new insights into the lacunar morphology of premenopausal women with rare forms of osteoporosis and it has also empowered researchers with access to microCT systems to explore additional hypotheses related to lacunar morphology.