A three-dimensional comparison of intramedullary nail constructs for osteopenic supracondylar femur fractures.

OBJECTIVES: This study developed a new 6 degree-of-freedom, unconstrained biomechanical model that replicated the in vivo loading environment of femoral fractures. The objective of this study was to determine whether various distal fixation strategies alter failure mechanisms and/or offer mechanical advantages when performing retrograde intramedullary nail (IMN) stabilization of supracondylar femur fractures in osteoporotic bone.

METHODS: Forty fresh-frozen human femora were allocated into 2 groups of matched pairs: "locked" (fixed angle locking construct with both distal locking screws rigidly attached to the IMN) versus "unlocked" (conventional locking technique with 2 distal locking screws targeted through the distal locking screw holes of the IMN) and "locked" versus "washer" (fixed angle locking with the most distal screw exchanged for a bolt with condyle washers) distal fixation of a retrograde IM nails. A comminuted fracture (OTA 33-A3) was simulated with a wedge osteotomy. Bone density measurements were completed on all specimens before instrumentation. Instrumented femurs were loaded axially to failure, whereas 6 degree-of-freedom translations and angulations were measured using Roentgen stereophotogrammetric analysis.

RESULTS: Mean (± SD) load born by "locked" specimens (1609 ± 667 N) at clinical failure was 38.1% greater (P = 0.09) than the corresponding mean load born by "unlocked" specimens (1165 ± 772 N). Clinical failure for the "washer" group (1738 ± 772 N) was 29.9% greater (P = 0.07) than the corresponding mean of the "locked" counterparts (1338 ± 822 N). Failure load was most clearly related to bone density in the "unlocked" fixation group.

CONCLUSIONS: Predicting failure load based on bone density using a least squares estimate suggests that the washer construct provides superior fixation to other treatment techniques. The failure mechanism for a comminuted, supracondylar fracture cannot be analyzed accurately with a 1-dimensional measurement. The most common failure mechanism in this model was medial translation and varus angulation.