Multiple myeloma (MM) is a chronic, malignant B-cell disorder, with a less than 50% 5-year survival rate [1]. This disease is responsible for vertebral compression fractures (VCFs) in 34 to 64% of diagnosed patients [1], and at least 80% of MM patients experience pathological fractures [3]. Even though reduced DXA-derived bone mineral density (BMD) has been observed in MM patients with vertebral fractures [4], the current quantitative standard method is insufficient in MM due to the osteo-destructive bone changes. Finite-element (FE) analysis is a computational and non-destructive modeling and testing approach to determine bone strength using 3D bone models from CT images. Thus, this study aimed to assess the differences in FE-predicted critical fracture load in MM patients with and without VCFs in the thoracic and lumbar segments of the spine.

Multi-detector CT (MDCT) images of two radiologically assessed MM patients (1 with VCFs and 1 without VCFs) were used to generate three-dimensional (3D) models of the whole spine. For each subject, the thoracic segments, 1 to 12 (T1-T12) and lumbar segments, 1 to 5 (L1-L5) were segmented and meshed. Heterogeneous, non-linear anisotropic material properties were applied by discretizing each vertebral segment into 10 distinct sets of materials. A compressive load was simulated by constraining the surface nodes on the inferior endplate in all directions, and a displacement load was applied on the surface nods on the superior endplate [2]. This analysis was performed using ABAQUS version 6.10 (Hibbitt, Karlsson, and Sorensen, Inc., Pawtucket, RI, USA).

The MM subject with VCFs had originally experienced fractures in the T4, T5, T12, L1, and L5 segments whereas the MM subject without VCFs experienced none. The former displayed large and abrupt differences in fracture loads between adjacent vertebrae segments, unlike the latter, which exhibited progressive differences instead (no abrupt changes between adjacent vertebrae segments observed).

Results from this preliminary study suggest that segments at high risk of fracture are collectively involved in an unstable network, which place the vertebral segments with high values of fracture loads (peaks) as well as the adjacent segments at risk of VCF. For instance, the high fracture load at T11 places T10, T11 and T12 at risk of fracture. Accordingly, T12 has already fractured, and T10 and T11 remain at risk. The relative changes between adjacent vertebrae segments that indicate instability (extremely high fracture load values) enables ease of identification of segments at high fracture risk. Clinicians would be able to work with pre-emptive treatment strategies in future as they can focus on more targeted therapy options at the high-risk vertebrae segments [3].