The remodeling of cortical bone largely takes place on the bone surfaces of Haversian canals that encase neuronal-vascular bundles that compose a central structure of the osteon. Larger animal models with a well-developed Haversian system, as observed in humans, are ideal to analyze cortical bone remodeling in pharmacological studies of anti-osteoporosis drugs, although they have some limitations in controlling for individual variability in size, weight, age, and number. This study aimed to morphometrically analyze cortical bone remodeling, through artificial intelligence (AI)-driven morphometry, focusing on Haversian canals (H.Ca) in dogs using four regimens of Teriparatide (TPTD) with daily and weekly administrations at lower and higher weekly doses (4.9 μg/kg/week and 19.8 μg/kg/week, respectively) for 9 months. Spatial distribution patterns of cortical remodeling sites were also assessed by applications of geographical information system (GIS).
A micro-computed tomography-based analysis of dog ribs were conducted. Transverse non-decalcified sections were prepared from ribs taken from each group. Differential interferometric and fluorescence images of each whole section were taken, and subsequently analyzed by AI-driven morphometry and GIS. The H.Ca area was significantly enlarged in the daily high dose (DH) group compared to the control group. Osteoid parameters were higher in the daily low dose(DL), DH, and weekly high (WH) groups, with the highest values especially in the DH group. The number and total area of resorption pores were also increased in the DH group compared to the WH group. Furthermore, spatial mapping by GIS suggested that low-density H.Ca areas coincided with expanded H.Ca areas.
Our analytical results revealed that the DH regimen specifically increased the number of eroded pores creating spaces between existing canals. It, therefore suggested that this regimen induced cortical porosity due to the increased resorption phase of osteons and fusion of H.Ca.