Introduction: The majority of midshaft humeral fractures will achieve a satisfactory outcome with non-operative management. However, internal fixation is occasionally required to assist with rehabilitation, particularly in multiply-injured patients. Although the clinical risks and benefits of the locking plate and humeral nail are well known, there is a paucity of data comparing their mechanical properties.

The aim of this study was to determine the torsional and 4-point bending properties of a midshaft humeral osteotomy reconstructed with either an intramedullary nail or locking plate.

Methods: 19 fresh cadaveric humeri were DEXA scanned to ensure similar BMD. Non-destructive 4-point bending was performed on the intact bone to determine stiffness in the sagittal and coronal planes. Load was applied using an MTS MiniBionix 858 (Mechanical Testing Systems, MN) at a rate of 1 mm/min to a maximum of 450 N.

A transverse midshaft osteotomy was created and a spacer ensured a constant 3-mm gap between the bone ends. Reconstruction was performed with either

Trigen humeral nail (Smith & Nephew, TN) – 10 specimens

Non-destructive 4-point bending was repeated, and then each humerus was embedded in a low-melting point alloy proximally and distally for torsional testing. Torque was applied at 5 deg/min until failure. Maximum torque, maximum angle and stiffness were calculated.

All data were analysed with SPSS for Windows (SPSS Inc., Il) using ANOVA.

Results: One specimen in the locking plate group fractured during plate application and was excluded from the study. Non-destructive bending tests showed no significant difference in stiffness of the intact bones between the two groups.

4-point bending: the bones reconstructed with the intramedullary nail were ~50% as stiff as the intact state in both planes. There was no statistically significant difference in stiffness between the intact bones and those reconstructed with the locking plate.

Torsional testing: the locking plate specimens were 3 times as stiff as the intramedullary nail specimens (P< 0.05) and failed at twice the torque (P< 0.05).

Discussion: Humeral intramedullary nails are reported to have an advantage over plates under axial loading (Chen et al, 2002). However, this study demonstrates that locking plates are superior to intramedullary nails in torsion and four-point bending. Although the clinical situation often dictates the most appropriate management, locking plates should be considered in patients when torsional or four-point bending loads are expected to predominate in the post-operative period.