Session III - Femur/Knee

Optimal Proximal Interlocking Location for Retrograde Femoral Intramedullary Nailing

Toni M. McLaurin, MD; Peter J. Symbas, MD; William C. Hutton DSc, Emory University School of Medicine, Atlanta, GA

Purpose: Despite the continued gains in popularity of retrograde femoral intramedullary nailing, questions about potential problems with proximal interlocking in the subtrochanteric region, such as risks of peritrochanteric fracture because of the nail end or locking screws acting as stress-risers, do remain. The purpose of this study was to evaluate the optimum location, from a biomechanical standpoint, for proximal interlocking screws in a retrograde femoral nail.

Methods: Twenty-three synthetic femurs and fourteen cadaveric femurs (age, 69 to 92 years) were obtained and divided into three groups: group I, nine synthetic and five cadaveric femurs with 42 cm ¥ 12 mm titanium retrograde intramedullary nails implanted; group II, nine synthetic, and five cadaveric femurs with 38 cm ¥12 mm nails implanted; group III, five synthetic and four cadaveric femurs with no intramedullary nails implanted. The 42-cm nail was inserted with the proximal end 1 cm proximal to the lesser trochanter and both proximal locking screws at or above the lesser trochanter. The 38-cm nail was inserted with the proximal end 1 cm distal to the lesser trochanter and both proximal locking screws distal to the lesser trochanter. The femurs were mounted on a materials testing system machine at an angle reproducing the actual force vector experienced by the proximal femur during axial loading. Each femur was axially loaded to failure and then examined to determine where failure occurred with respect to the femoral neck, the lesser trochanter, the proximal end of the nail, and the proximal interlocking screws.

Results: The nine synthetic femurs with the 42-cm nails inserted (group I) showed a mean load to failure of 2335 N. The mean load to failure in the nine synthetic femurs with the 38-cm nails inserted (group II) was 2006 N. The difference between these two means was significant (P = 0.0080). All synthetic femurs of group I and of group II failed at the proximal end of the nail or at the proximal interlocking screw interface. All synthetic femurs of group III failed in the subtrochanteric region with a mean load to failure of 2904 N. All cadaveric femurs of groups I, II, and III failed at the femoral neck with a mean load to failure of 3478 N across all three groups.

Discussion: There is a significant difference in the amount of force needed to break synthetic femurs with two different length retrograde nails inserted, requiring more force to break the femurs with the 42-cm nails inserted. Our results may provide biomechanical evidence that hardware in the subtrochanteric region does act as a stress-riser because all the synthetic femurs in groups I and II failed at the proximal extent of the nail or at the level of the proximal locking screw and all in group III had a higher mean load to failure. The results may also support the use of a longer nail, as the mean load to failure was greater in the 42-cm nail group. However, all of the cadaveric femurs failed at the femoral neck with a higher mean load to failure than seen in any of the synthetic femurs, and the site of failure was not influenced by the presence, absence, or proximal location of an implant. Perhaps the femoral neck, not the subtrochanteric region, is the weakest area when the femur is loaded in axial compression (independent of the actual location of any implants), although age-related loss of inherent mechanical stiffness in our older cadavers may have caused our specimens to behave differently than younger cadaveric bone might. Because the synthetic femurs did not accurately simulate the mechanical properties of the cadaveric specimens, these results reinforce the limitations in using synthetic bone as a model, but may also illustrate that these particular bone models are designed to be more representative of young human femurs which have not lost their mechanical stiffness in the femoral neck region.

Conclusions: Despite limitations, we believe that biomechanical data obtained from the use of synthetic femurs is useful and may support placement of the proximal end of the nail and proximal locking screws at or above the level of the lesser trochanter in the younger patient.