The biomechanics of the T2 femoral nailing system: a comparison of synthetic femurs with finite element analysis

Intramedullary nails are commonly used to repair femoral fractures. Fractures in normal healthy bone often occur in the young during motor vehicle accidents. Although clinically beneficial, bone refracture and implant failure persist. Large variations in human femur quality and geometry have motivated recent experimental use of synthetic femurs that mimic human tissue and the development of increasingly sophisticated theoretical models. Four synthetic femurs were fitted with a T2 femoral nailing system (Stryker, Mahwah, New Jersey, USA). The femurs were not fractured in order to simulate post-operative perfect union. Six configurations were created: retrograde nail with standard locking (RS), retrograde nail with advanced locking 'off' (RA-off), retrograde nail with advanced locking 'on' (RA-on), antegrade nail with standard locking (AS), antegrade nail with advanced locking 'off' (AA-off), and antegrade nail with advanced locking 'on' (AA-on). Strain gauges were placed on the medial side of femurs. A 580 N axial load was applied, and the stiffness was measured. Strains were recorded and compared with results from a three-dimensional finite element (FE) model. Experimental axial stiffnesses for RA-off (771.3 N/mm) and RA-on (681.7 N/mm) were similar to intact human cadaveric femurs from previous literature (757 +/- 264 N/mm). Conversely, experimental axial stiffnesses for AS (1168.8N/mm), AA-off (1135.3N/mm), AA-on (1152.1 N/mm), and RS (1294.0 N/mm) were similar to intact synthetic femurs from previous literature (1290 +/- 30 N/mm). There was better agreement between experimental and FE analysis strains for RS (average percentage difference, 11.6 per cent), RA-on (average percentage difference, 11.1 per cent), AA-off (average percentage difference, 13.4 per cent), and AA-on (average percentage difference, 16.0 per cent), than for RA-off (average percentage difference, 33.5 per cent) and AS (average percentage difference, 32.6 per cent). FE analysis was more predictive of strains in the proximal and middle sections of the femur-nail construct than the distal. The results mimicked post-operative clinical stability at low static axial loads once fracture healing begins to occur.