ABSTRACT
BACKGROUND: Current implants for clavicle fractures are known to cause poor cosmesis and irritation, which may require implant removal. Low-profile shape-memory staples provide an attractive alternative, but their biomechanical utility in clavicle reconstruction is unknown. We hypothesized that shape-memory reconstructions would be more compliant compared to traditional constructs but would also outperform conventional plates during cyclic loading to failure. METHODS: This study was performed with 36 synthetic clavicles and 12 matched pairs of cadaveric specimens. The synthetic study tested four reconstructions: a single superiorly placed staple (n = 6), a single anteroinferiorly-placed staple (n = 6), a 3.5 mm reconstruction plate (n = 12), and two orthogonally placed staples (n = 12). The cadaveric study tested three constructs: reconstruction plate (n = 8), two orthogonal staples (n = 8), and a 2.7 mm reconstruction plate combined with a superior staple (n = 8). Non-destructive 4-point bending, compression, and torsion assays were performed prior to destructive cantilever bending and cyclic torsion tests. FINDINGS: The single staple and double staple groups demonstrated significantly decreased resistance to bending (p < 0.001) and torsion (p ≤ 0.027) when compared to reconstruction plate groups. The double staple group sustained significantly fewer cycles to failure than the reconstruction plate group in cyclic torsional tests (p = 0.012). The synthetic models produced higher stiffness and failure mechanisms that were completely different from cadaveric specimens. INTERPRETATION: Shape memory alloy implants provided inadequate stiffness for clavicle fixation but may have utility in other orthopaedic applications when used as a supplementary compression device in conjunction with traditional plated constructs. Synthetic bones have limited capacity for modeling fragility fractures.