INITIAL FAILURE IN HIP FRACTURES OCCUR IN THE SUPEROLATERAL CORTEX: EVIDENCE FROM AN EX-VIVO MODEL

Abstract

Evidence suggests that femoral neck fractures initiate in the superolateral cortex, where it is significantly thinner in older than younger individuals (Mayhew, et al. Lancet 2005). Thus, we sought to determine the relative time-course of crack initiation and propagation during a simulated hip fracture.

Four unembalmed frozen, human cadaveric specimens (mean age = 78 yrs) were loaded to failure in sideways fall configuration at a rate of 100 mm/sec using a materials testing system. Images of the fracture were captured with two high-speed video cameras at a resolution of 384x384 pixels, and sample rate of 9,111 Hz (frames/second).

Test A: The load-displacement (L-D) curve had three distinct peaks: at the first peak (4390 N), the head and neck rotated slightly. At the second peak (4607 N), a visible local compressive fracture appeared in the superior cortex of the proximal neck. At the third peak (3582 N), a neck-spanning tensile failure occurred in the inferior neck. Test B: At the first and second peak loads (1714 N and 3040 N) fluid was released from the posterior then superior and inferior surfaces. The third peak load (3361 N) corresponded to a local compressive failure in the lateral superior neck, followed by a neck-spanning tensile failure medially. Test C: The L-D curve was linear until ultimate load (3038 N). A compressive crack first appeared on the anterior-superior surface of the neck cortex, then fractured in the inferior neck. Test D: The L-D curve was linear until ultimate load. A small local crack appeared in the superior cortex of the proximal neck at ultimate load (3841 N).

We found that during ex vivo simulations of hip fracture, the femur failed initially in the superior cortex of the neck, and then failed in the inferior cortex. This is the first study to demonstrate, with high speed video data, the location of crack initiation and its propagation. These preliminary data support the hypothesis of Mayhew et al. (Mayhew, et al. Lancet 2005) in terms of fracture development and could relate to clinically relevant fracture types.