Focus on stability: biomechanical evaluation of external fixation technique versus locking plate osteosynthesis in 3-part proximal humeral fractures.

Background: The purpose of this study was to investigate the primary stability and displacement tendencies of an external fixation system in 2 different configurations in comparison to locking plate fixation in a 3-part proximal humeral fracture model.

Methods: Twenty-one fresh-frozen human cadaveric proximal humeral specimens were divided into 3 groups of 7 stripped specimens. An unstable 3-part fracture of the proximal humerus was set. Construct A consisted of the Galaxy Fixation Shoulder System, with fixation in accordance with the manufacturer's recommended technique (3 pairs of threaded wires). Construct B was fixed with an additional pair of threaded wires according to the recommendations of Gumina et al (4 pairs of threaded wires). The remaining 7 specimens were fixed with a locking plate. By use of an optical motion capture system, relative motion at the fracture site and failure load were evaluated during a cyclic loading protocol.

Results: Plate osteosynthesis showed the highest cyclic load to failure (895 ± 239 N; range, 597-1135 N), followed by construct B (692 ± 121 N; range, 432-788 N) and construct A (350 ± 190 N; range, 139-615 N). Statistically significant differences were found between plating and construct A (P < .001), between plating and construct B (P = .031), and between construct B and construct A (P = .013). Relative motion at the fracture site, in terms of change in distance during cyclic loading, was lowest with construct B measured at the peak load of 395 N after 11 steps (1.73 ± 1.98 mm; range, 0.53-5.96 mm) and highest with construct A (8.46 ± 10.67 mm; range, 2.26-20.79 mm). Angular change measured at the peak load after 11 steps was lowest with construct B (1.19° ± 0.36°; range, 0.59°-1.7°) and highest with construct A (2.44° ± 1.63°; range, 0.77°-4.04°). Statistically significant differences were found between construct A and construct B and between construct A and plating at various steps of cyclic loading.

Conclusion: In this biomechanical study of unstable 3-part proximal humeral fractures, locking plate osteosynthesis demonstrated the highest possible load application. External fixator application with 8 threaded wires (construct B) revealed the least relative motion at the fracture site, without a statistically significant difference compared with locking plates. It therefore represents another valid technique for treatment regarding primary stability. The use of external fixator application with 6 threaded wires (construct A) resulted in the lowest stability and the highest rates of displacement.