Introduction: Fracture fixation devices are often used to reduce recovery time and regain function. Many tests on new devices are done on composite bones which are available and mechanically consistent. Zdero, et al., studied the screw pull-out, shear stress, and energy to failure in composite (Sawbones®) and cadaveric bones finding no statistical difference for these properties[1]. There has been no study comparing cadaveric and composite bone under cyclic loading and no study comparing composite and canine bones.
Methods: Anterior-posterior holes were drilled 3 cm apart and tapped for 3.5 mm cortical bone screws which were inserted bicortically. Each test construct was loaded into a servohydraulic materials test frame [Bionix 858, MTS Corp, Eden Prairie, MN] with a custom-designed screw grip for axial alignment. Pull-out force was collected under displacement control at 0.1 mm/sec. For the fatigue loading tests, a 20 N pre-load was applied followed by a tensile cyclic load of 200-1000 N for 200,000 cycles at 7 Hz for the composite and cadaveric bone and at 2 Hz for the canine bone. Two sample t-tests were performed comparing the composite to the human and to the canine bone with a significance value of p = 0.05.
Results: The screw pull-out force for the composite, human, and canine were 3894 N, 4746 N, and 1628 N respectively [Table 1]. The screw displacement with fatigue loading for the composite, human, and canine was 0.057 mm, 0.094 mm, and 1.039 mm respectively [Table 2]. The canine testing failed at 84,715 ± 68,312 cycles; the composite and human bones ran to completion. p-value < 0.05 in all comparisons.
Discussion: There is a statistical difference between composite bones and cadaveric bones in regards to screw fatigue. The difference is not thought to be of clinical significance as the average pull-out of about 0.1 mm is not enough to cause failure over the time frame of a device. The canine bones showed significantly greater displacement before pull-out and failed under cyclic load; the composite specimens did neither.
Conclusion: These composite bone models are widely used to test devices and the models had not been validated in terms of screw fatigue. This study suggests that they may not be accurate representations of fatigue behavior, particularly for canine studies.
Table 1: Screw Pull-Out Force
Newtons p-value
Composite n = 8 3895 ± 451
Human n = 4 4746 ± 553 0.044
Canine n = 8 1629 ± 284 4.70E-08
Table 2: Fatigue Results
Displacement p-value Cycles p-value
Composite n=19 0.057 ± 0.0260 200,000
Human n=17 0.094 ± 0.0296 0.0004 200,000
Canine n=6 1.039 ± 0.184 4.75E-05 84,715 ± 68,312 0.002
Synthes Corp; OSU Dept. of Orthopaedics
References:
[1] Zdero R. J Biomech Eng 131, 094503, 2009