Fixation of three-part proximal humeral fractures: a biomechanical evaluation

To examine the biomechanical stability of three constructs currently used for the management of three-part proximal humerus fractures. Tension band wires (TBW) with supplemental Enders nails, modified cloverleaf plate and screws, and intramedullary (IM) nailing with proximal and distal interlocks were tested to determine relative stability.

A reproducible three-part fracture was made in fresh-frozen stripped proximal humeri. The fracture was stabilized using TBW/Enders nail (n = 6), plate/screws (n = 5), or IM nailing (n = 5).

Mechanical testing was performed with a small preload followed by deflection of five millimeters at a rate of one millimeter per second in flexion, extension, and varus and valgus relative to the humeral shaft. A load-displacement curve was obtained. Torsional testing was performed in internal and external rotation, and torque-rotation curves were recorded.

In cantilever bending, the plate/screws construct and the IM nail construct were superior to the TBW/Enders nail construct for all parameters except extension. There was no statistically significant difference between the IM nail and the plate/screws groups. Torsional stiffness testing revealed that the plate/screws and the IM nail were superior to the TBW/Enders nail construct. There was no statistical difference between the IM nail and the plate/screws groups.

In a cadaveric model of three-part proximal humerus fractures stripped of soft tissue, plate/screws fixation and IM nailing provide greater torsional and bending stiffness than does fixation with TBW/Enders nail. There was no statistically significant difference in torsional or bending stiffness between IM nailing with interlocks and plate/screws fixation in this model.