TL;DR
This study developed and validated a finite element (FE) model using CT scans to predict vertebral bone strength in multiple myeloma (MM) patients. The FE model proved more effective than traditional bone mineral density (BMD) measurements in identifying vertebral fracture risk, particularly by highlighting structural instability between adjacent vertebrae.
What Problem Does This Paper Address?
Multiple myeloma (MM) is a blood cancer that significantly weakens bones, leading to a high incidence of vertebral compression fractures (VCFs). Current methods for assessing fracture risk, primarily bone mineral density (BMD), are often unreliable in MM patients due to the complex nature of bone loss in this disease. This unreliability makes it difficult for clinicians to accurately predict which patients are at high risk of VCFs, hindering timely and targeted interventions. The paper aims to address this gap by developing and validating a more sophisticated tool, finite element (FE) analysis, to provide a patient-specific assessment of vertebral bone strength and fracture risk.
Methods
The researchers employed a two-pronged approach:
- In Vitro Validation: They first developed FE models from CT scans of 12 human vertebrae obtained from donors. These models were then subjected to biomechanical testing to determine their failure load. The FE-predicted failure loads were compared to the experimentally measured failure loads to validate the accuracy of the FE modeling approach.
- In Vivo Application: After validation, the same FE modeling technique was applied to CT scans of vertebrae (T1-L5) from four MM patients (two with existing fractures, two without). This allowed for the analysis of patient-specific vertebral strength and the identification of potential fracture risks within their spines.
Key to the FE analysis was the creation of patient-specific 3D models from DICOM images, followed by meshing and the assignment of heterogeneous, nonlinear, and anisotropic material properties based on Hounsfield Units (HUs) derived from the CT scans. The material properties were linked to bone density using established literature references.
Datasets
- In Vitro Dataset: 12 human vertebrae (T9-T12) from 3 fresh frozen human donors.
- In Vivo Dataset: Whole-body MDCT scans of 4 MM patients (one male, three females) with new diagnoses, identified retrospectively. Two of these patients had vertebral compression fractures (VCFs), and two did not.
Key Insights and Ideas
- FE Analysis Outperforms BMD: The study strongly suggests that FE analysis is a superior tool for assessing vertebral fracture risk in MM patients compared to traditional BMD measurements, which can be misleading in this population.
- Structural Instability as a Predictor: The most significant finding is the identification of "erratic behavior" in fracture load values between adjacent vertebral segments in patients who have already sustained fractures. This suggests that localized regions of structural instability, rather than just overall bone density, are critical indicators of fracture risk.
- Peaks as Warning Signs: The presence of "peaks" in fracture load values between adjacent segments appears to be a key indicator of potential instability, placing those segments and their neighbors at higher risk.
- Geometric Deformations: Patients with existing fractures also exhibited severe geometric deformations in their vertebral models, further highlighting the impact of the disease on bone structure.
- Patient-Specific Risk Assessment: The FE approach allows for a highly individualized assessment of fracture risk, moving beyond generalized population-based risk factors.
Key Results
- In Vitro Validation: FE-predicted failure loads (FFE) showed a significant correlation with experimentally determined failure loads (Fexp) (r = 0.85, P < 0.001), validating the FE model.
- In Vivo Findings:
- Patients with fractures exhibited an "erratic behavior" in FE-predicted fracture loads between adjacent spinal segments, characterized by peaks.
- Patients without fractures showed a more "gradual change" in FE-predicted strength values.
- Segments adjacent to these "peaks" were identified as being at higher risk of fracture.
- Geometrically compromised segments in fractured patients showed higher maximum principal strain values.
- BMD Correlation: While FE-predicted strength correlated moderately with BMD (r=0.79 for lumbar, r=0.58 for thoracic segments), the FE analysis provided more nuanced information about localized instability.
Limitations and Future Work
- Small Sample Size: The study was limited to only four MM patients, which restricts the statistical power and generalizability of the findings. A larger sample size is needed to establish definitive threshold values for relative changes in fracture loads and to solidify the clinical significance of the observed patterns.
- Scan Resolution Differences: The resolution of MDCT scans used for in vitro validation was higher than that used for in vivo MM patient scans, potentially impacting the precision of the in vivo models.
- Simplified Loading Conditions: The FE analysis only considered a single, axial compressive loading condition. Future work should incorporate more complex, in vivo loading scenarios (e.g., flexion, rotation, bending) to provide a more comprehensive assessment of fracture risk.
- Limited Parameters: The study primarily focused on fracture load values. Future research could incorporate geometric parameters like buckling ratio to further indicate structural stability.
- Future Directions: The authors suggest that establishing a threshold value for the relative change in fracture load between adjacent segments could serve as a valuable clinical tool for identifying at-risk segments and guiding preemptive treatment strategies. They also propose that this noninvasive, patient-specific examination can provide crucial information for reducing morbidity and mortality in MM patients.
Summary updated