Abstract
Background
Structural bone allografts are an established treatment method for long-bone structural defects arising from such conditions as trauma, sarcoma, and osteolysis following total joint replacement. However, the quality of structural bone allografts is difficult to non-destructively assess prior to use. The functional lifetime of structural allografts depend on their ability to resist cyclic loading, which can lead to fracture even at stress levels well below the yield strength. Because allograft bone has limited capacity for remodeling, optimizing allograft selection for bone quality could decrease long-term fracture risk. Raman spectroscopy biomarkers can non-destructively assess the three primary components of bone (collagen, mineral, and water), and may predict the resistance of donor bone allografts to fracture from cyclic loads.
The purpose of this study was to prospectively assess the ability of Raman biomarkers to predict number of cycles to fracture (“cyclic fatigue life”) of human allograft cortical bone.
Methods
Twenty-one cortical bone specimens were from the mid-diaphysis of human donor bone tissue (bilateral femurs from 4 donors: 63M, 61M, 51F, 48F) obtained from the Musculoskeletal Transplant Foundation. Six Raman biomarkers were analyzed: collagen disorganization, type B carbonate substitution (a surrogate for mineral maturation), matrix mineralization, and 3 water compartments. Specimens underwent cyclic fatigue testing under fully reversed conditions at 35 and 45MPa (physiologically relevant stress levels for structural allografts). Specimens were tested to fracture or to 30 million cycles (“run-out”), simulating 15 years of moderate activity (i.e., 6000 steps per day). Multivariate regression analysis was performed using a tobit model (censored linear regression) for prediction of cyclic fatigue life. Specimens were right-censored at 30 million cycles.
Results
All of the 6 biomarkers that were evaluated were independently associated with cyclic fatigue life (p < 0.05). The multivariate model explained 70% of the variance in cyclic fatigue life (R2=0.695, p<0.001,). Increasing disordered collagen (p<0.001) and loosely collagen-bound water compartments (p<0.001) were associated with decreased cyclic fatigue life. Increasing type B carbonate substitution (p<0.001), matrix mineralization (p<0.001), tightly collagen-bound water (p<0.001), and mineral-bound water (p=0.002) were associated with increased cyclic fatigue life. In the predictive model, 42% of variance in cyclic fatigue life was attributable to degree of collagen disorder, all bound water compartments accounted for 6%, and age and sex accounted for 17%.
Conclusions
Raman biomarkers of three bone components (collagen, mineral, and water) predict cyclic fatigue life of human cortical bone. Increased baseline collagen disorder was associated with decreased cyclic fatigue life, and was the strongest determinant of cyclic fatigue life. Increased matrix mineralization and mineral maturation were associated with increased cyclic fatigue life. Bound-water compartments of bone contributed minimally to cyclic fatigue life. These results are complementary with prior Raman studies of monotonic testing of bone that reported decreased toughness and strength with increased collagen disorder and increased stiffness with increased bone mineralization and mineral maturation. This model should be prospectively validated. Raman analysis is a promising tool for the non-destructive evaluation of structural bone allograft quality and may be useful as a screening tool for selection of allograft bone.
Acknowledgements
Supported by a grant from the Musculoskeletal Transplant Foundation. The Dudley P. Allen Fellowship (JYD), Wilbert J. Austin Professor of Engineering Chair (CMR) and the Leonard Case Jr. Professor of Engineering Chair (OA) are gratefully acknowledged.
