Abstract
The endplate region of the vertebra, which includes the bony endplate (BEP) and underlying subchondral trabecular bone (STB), is critically involved in vertebral fracture (VF). While evidence abounds that failure initiates in the endplate region, the relative risk of failure of the BEP vs. STB has not been established. In this study, micro-finite element models were constructed of L1 vertebrae (n = 21) that were mechanically tested in a prior study and given experimentally matched boundary conditions corresponding to the vertebra's yield point. Volumes of interest (VOIs) were defined corresponding to the BEP and STB; the remainder was defined as the mid-vertebral body (MVB). The proportion of elements within each VOI that yielded was defined as the VOI yield fraction, and this value divided by the yield fraction of the entire model was defined as the normalized yield fraction. While yield fraction did not differ across VOIs (p = 0.179), normalized yield fraction was greater in the BEP than STB and MVB (p < 0.001), indicating a higher risk of yield in the BEP compared to the other two VOIs. None of the yield fractions was correlated with BEP or STB microstructure, and tension (rather than compression) was the dominant mode of tissue level yield. These findings indicate that the BEP, more so than the STB, is likely the site of VF initiation and that current methods of screening for VF risk, because they omit specific analysis of the BEP, are missing the region that matters the most. The endplate region of the vertebra, which includes the bony endplate (BEP) and underlying subchondral bone (SB), is critically involved in vertebral fracture (VF). While evidence abounds that failure initiates in the endplate region, the relative risk of failure of the BEP vs. SB has not been established. In this study, micro-finite element models were constructed of L1 vertebrae (n = 21) that had been mechanically tested in a prior study, and they were given experimentally matched boundary conditions corresponding to the vertebra's yield point. Volumes of interest (VOIs) were defined corresponding to the BEP and SB; the remainder was defined as the mid-vertebral body (MVB). The proportion of yielded elements within each VOI was defined as the VOI yield fraction, and this value divided by the yield fraction of the entire model was defined as the normalized yield fraction. While yield fraction did not differ across VOIs (p = 0.179), normalized yield fraction was greater in the BEP than SB and MVB (p < 0.001), indicating a higher risk of yield in the BEP compared to the other two VOIs. None of the yield fractions was correlated with BEP or SB microstructure, and tension (rather than compression) was the dominant mode of tissue level yield. These findings indicate that the BEP, more so than the SB, is likely the site of VF initiation and that current methods of screening for VF risk, because they omit specific analysis of the BEP, are missing the region that matters the most.