Étude biomécanique de la réduction trochantérienne

Abstract

The greater trochanter (GT) serves as an anchor for several muscles, the most important one being the gluteus medius. A common operative or postoperative complication of a total hip arthroplasty is the fracture of the GT. The reduction of this fracture is complicated by the pull of the abductor muscles attached to the GT which exert forces in competition with its alignment. Some authors suspect that the abduction would produce posterior-anterior (PA)displacement and rotation (laterally) of the GT fragment during extension movements such as sit-to-stand or stair climbing. Current implants present a lateral and superior approach and would not sufficiently block the fractured GT fragment towards the PA direction. A new trochanteric implant (Y3) was developed at the research center of the Sacré-Coeur Hospital in Montréal. In addition to the classic lateral plate, it splits into a Y-shape with an anterior branch and offers the possibility to fix the bone with locking head screws or with cable cerclages. Modern techniques of GT fixation advocate the use of plates fastened with cable cerclages around the femur and GT. Actually, the use of screws is virtually nonexistent due to the small size of the GT fragment and the common presence of osteoporosis. The problematic of this biomechanical study of trochanteric reduction revolves around three research questions. First, the physiological movements of leg extension exert efforts towards the PA direction on the GT fragment fractured and instrumented, which could explain the high rate of postoperative complications associated with current implants. GT movements resulting from these efforts have not been characterized. Secondly, the Y shape of the Y3 implant has an anterior branch that would reduce the PA movements of the GT and possibly those of other directions. Thirdly, the fixation of the implant by locking head screws in the shaft bone and the GT has not been studied. The usefulness of cable cerclages in addition to or in replacement of these screws also remains to be explored. Two specific experimental test beds simulating a hip extension were developed in order to address the three research questions. Both protocols used a similar experimental setup that applied a load to activate the abductors towards the PA direction (simulating stair climbing) on a fractured and instrumented GT. Two main responses were measured. Migration (MIG) was defined as the residual GT movement after cyclic stair climbing movement compared to the initial state. Displacement (DISP) was defined as the amplitude of movement between unloaded and loaded states within a single cycle. The first protocol was a paired study with cadaveric specimens, while the second involved an experimental plan with synthetic specimens. The first protocol associated with the first two research questions (paired study) was performed on 20 cadaveric femurs (10 pairs) to compare two types of fixation: 1) an antero-lateral (AL) implant which includes an anterior branch and 2) a lateral (L) implant without any anterior branch. Biological variability (size, shape and bone density) was included in this section. No recurrence has been made for a total of 20 trials. The second protocol was a complete experimental design with two factors at three levels (3 ² = 9 trials) allowing to estimate all main effects and quadratic interactions. It aimed to verify the first and third research questions. A small variation along synthetic specimens permitted to control the anatomical parameters. With two repetitions, a total of 27 trials were conducted. Finally, a comprehensive validation study of the two experimental protocols was performed. The paired study on cadaveric specimens showed that migration in translation decreased significantly in the PA direction (P = 0,02) when the antero-lateral (AL) implant with the anterior branch was used. The AL implant also significantly decreased rotational MIG around the superior axis (P = 0,01) and the lateral axis (P = 0,02). Translational DISP also declined significantly in the PA (P = 0,04) and lateral directions (P = 0,03) when the GT fragment was fixed with the AL implant. Rotational DISP was significantly reduced around the superior axis (P = 0,04) when using the AL implant. The experimental design using synthetic specimens showed that the GT fragment MIG was the most affected by the fixation method. Varying the fixation from cable cerclages only to locking head screws only showed a very significant reduction in translational and rotational MIG (all P << 0,01 except P = 0,66 for the rotation around the superior axis). The bone shaft fixation method significantly reduced the superior translation when screws were used (P << 0,01). DISP was also significantly reduced by varying the fixation from cable cerclages only to locking head screws only, but to a lesser extent (P << 0,01 for superior translation and rotation around the PA axis and P = 0,01 for rotation around the lateral axis; all other P > 0,07). The bone shaft fixation method showed no significant reduction of all DISP. No significant interaction was found between the GT and bone shaft fixation methods. In conclusion, the GT fragment underwent multidirectional movements during hip extension. The lateral (L) implant failed in rotation around the superior axis and in PA translation. The anterior branch significantly limited this failure. Finally, adding cable cerclages is recommended only on the bone shaft when screw fixations are unachievable. For the reduction of the fractured GT, a Y-shaped plate with lateral and anterior branches attached with locking head screws is an effective alternative to current surgical procedures with lateral plates.

Date	17 Dec 2013
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Yvan Petit (Supervisor) & Yves Laflamme (Co-supervisor)