Session IV - Femur

*A Biomechanical Analysis of Fixation Constructs in High Angle Femoral Neck Fractures

Michael Sirkin, MD (b-Synthes USA); Mark G. Grossman, MD (b-Synthes USA); Regis L. Renard, MD (b-Synthes USA); Christopher T. Sabatino, MS (b-Synthes USA); Christopher Doumas, BA (b-Synthes USA); Mark C. Reilly, MD (b-Synthes USA); Fred F. Behrens, MD (b-Synthes USA), Orthopaedic Trauma Service, New Jersey Medical School, Newark, NJ

Purpose: In Pauwels' classification of femoral neck fractures, type III fractures (OTA type 31-B2.3) are those which are the most vertically oriented (>70°). These fractures are typically caused from high-energy trauma in young adults. Aseptic necrosis rates range from 20-90% and non-union rates from 17-62%. Anatomic reduction and rigid internal fixation are mandatory to minimize complications while maximizing functional outcomes. To date, no study has investigated the in vitro strengths of different fixation techniques for this fracture type. The purpose of this study is to compare four different constructs in high angled femoral neck fractures to determine the optimal fixation technique.

Materials and Methods: A preliminary study and subsequent powers analysis was performed to determine the number of specimens needed. Six pairs of femurs from fresh cadavers were used (average age 41, range 18-58). All soft tissues were removed and the matched pairs were potted in a cylinder of polymethylmethacrylate. Scoring the femoral neck and applying a catastrophic impaction force reproducibly created high angled femoral neck fractures. The fractures were anatomically reduced and internally fixed with one of four constructs. Techniques included three parallel 7.0mm cannulated lag screws (CS), a 135° sliding hip screw with a 2-hole side plate (DHS), a 95° sliding condylar screw with a 5-hole side plate (DCS), and two 7.0mm cannulated lag screws along the axis of the femoral neck with a transverse 4.5 mm cortical lag screw directed into the calcar (XCS). A balanced incomplete blocking design was utilized to distribute the four constructs within the six pairs. Mechanical testing was performed using an MTS machine on the fixed fractures in simulated single-legged stance (0° flexion, 15° adduction, and 5° internal rotation). A displacement gage was attached directly to the femur to measure inferior head translation. Each specimen was cyclically loaded 0 N to 1000 N of compression at 1mm/sec for 50 cycles to reduce creep. After cycling, the specimens were tested to failure at a rate of 1 mm/sec, and a load vs. displacement curve was generated. The force required to create 2 mm, 4 mm, and 6 mm of femoral head displacement was examined. The stiffness within 50 N through 2000 N for each specimen was calculated. The data was statistically analyzed using an analysis of variance (ANOVA) with Fisher's Protected LSD post-hocs.

Results: The force required to displace the XCS model 2 mm (190%, p=0.011), 4 mm (177%, p=0.037), and 6 mm (199%, p=0.05) was significantly greater than the CS model. Significant differences were also found between the DCS and CS models at 2 mm (175%, p=0.018) and 4 mm (192%, p=0.023) of displacement. At 2mm of displacement the DHS model was significant weaker than the DCS (139%, p=0.05)and XCS (151%, p=0.027) models. The DCS and XCS were 140% stiffer than the CS and 170% stiffer than the DHS, but these differences were not statistically significant. The results are summarized in Table 1. The mode of failure of the constructs was different between groups. The XCS and DCS models failed by bending of the 4.5 mm cortical screw directed into the calcar. The DHS models failed by bending of the hip screw at the screw-barrel interface. The CS models failed by movement of the implants within bone or by bending of the most inferior lag screw.

Table 1: Results Summary -Values with an (!) indicate a significant difference with CS (p<0.05), Values with an (*) indicate a significant difference with DHS (p<0.05).

	Force at 2 mm	Force at 4 mm	Force at 6 mm	Stiffness
	Displacement (N)			(N/mm)
CS	1789.93 ± 369.57	1883.79 ± 93.78	1949.09 ± 184.94	2127.41 ± 1749.10
DHS	2243.18 ± 491.29	2692.30 ± 848.97	3323.29 ± 811.15	1740.23 ± 475.75
DCS	3123.94 ± 980.92*!	3621.10 ± 1442.22!	3632.70 ± 303.10	3080.93 ± 574.60
XCS	3396.02 ± 474.68*!	3337.54 ± 161.65!	3862.14 ± 466.58!	2964.57 ± 1741.33

Discussion: Our data show the most commonly used devices to fix femoral neck fractures, the sliding hip screws and cannulated lag screws, are less stiff and significantly less strong than either the DCS or XCS constructs. When placing the DCS, a lag screw is placed through the side plate into the calcar. The XCS model also utilizes a lag screw for calcar fixation. This screw accomplishes 2 functions. The first is to create an inferior buttress that resists displacement. The second, by being 90° to the fracture, is to achieve optimum compression. An anatomically reduced fracture with interdigitation and compression will significantly oppose displacement. Stronger, stiffer fixation provides a more favorable environment for bone healing. Optimal fixation can be accomplished with either 95° sliding condylar screw or with 3 crossed screws, 2 lag screws placed parallel to the neck and a cortical lag screw placed into the inferior femoral neck.

Conclusion: Our study demonstrates the ideal fixation of vertically oriented shear fractures of the femoral neck is either a 95° condylar screw or 3 crossed screws. With increased strength and stiffness, patients can be mobilized with less fear of fracture displacement. In addition, the added stability will place fractures in their optimal environment for healing and hopefully decrease the reported high rates of avascular necrosis and non-union.