ABSTRACT
OBJECTIVES: Third-generation cephalomedullary nails currently represent the gold standard in the treatment of unstable trochanteric femur fractures. Recently, an extramedullary rotationally stable screw-anchor system (RoSA) has been developed. It was designed to combine the benefits of screw and blade and to improve stability using a locked trochanteric stabilizing plate (TSP). The purpose of this study was to compare the biomechanical behavior of RoSA/TSP and the proximal femoral nail antirotation (PFNA). METHODS: Standardized AO/OTA 31A2.2 fractures were induced by an oscillating saw in 10 paired human specimens (n = 20; mean age = 85 years; range: 71-96 years). The fractures were stabilized by either the RoSA/TSP (Koenigsee Implants, Allendorf, Germany) or the PFNA (DePuy Synthes, Zuchwil, Switzerland). Femurs were positioned in 25 degrees of adduction and 10 degrees of posterior flexion and were cyclically loaded with axial sinusoidal pattern at 0.5 Hz, starting at 300 N, with stepwise increase by 300 N every 500 cycles until bone-implant failure occurred. After every load step, the samples were measured visually and radiographically. Femoral head migration was assessed. RESULTS: The stiffness at the load up to the clinically relevant load step of 1800 N (639 ± 378 N/mm (RoSA/TSP) vs. 673 ± 227 N/mm (PFNA); P = 0.542) was comparable, as was the failure load (3000 ± 787 N vs. 3780 ± 874 N; P = 0.059). Up to 1800 N, no femoral head rotation, head migration, or femoral neck shortening were observed either for RoSA/TSP or PFNA. Whereas failure of the PFNA subsumed fractures of the greater trochanter and the lateral wall, a posterior femoral neck fracture with a significantly increased femoral neck shortening (1.7 mm vs. 0 mm; P = 0.012) was the cause of failure with RoSA/TSP. This specific kind of failure was induced by a femoral neck weakening caused by the posterior TSP stabilizing screw. CONCLUSIONS: There was no significant difference in biomechanical properties between the RoSA/TSP and the PFNA for the fracture pattern tested. However, failure modes differed between the 2 implants with greater femoral neck shortening observed in the RoSA/TSP group.