Compressive stress, tensile stress and shear force are generated at the proximal segment of the femur during force conduction, because of the femoral neck shaft angle. Stress trabecula bone, tension trabecula bone and calcar femorale are derived based on the loading and offered the ability of mechanotransduction. When femoral neck fracture occurred, the normal stress conduction is lost because of these destroyed truss structures. Therefore, it is generally accepted that anatomic reduction with stable fixation offers patients the best biomechanical microenvironment needed for fracture healing without complication to maintain the native hip14, 15.
The mechanical stability of fracture fixation structure is closely related to the implant, reduction quality as well as the structure and material characteristics of bone9. Kunapuli et al.16 found that the medial femoral neck buttress plate enhanced the stability of cannulated screw and sliding hip screw constructs, significantly increased stiffness and load to failure. Giorfano et al.17 conducted a mechanical study to analyze the role of the medial buttress plate in Pauwels type III femoral neck fractures by comparing the resistance of two fixation methods using three cannulated screws, and the results showed that the medial buttress plate combined with multiple cannulated screw fixation provided a mechanically superior structure for Pauwels type III fractures. Additionally, Nwankwo et al.15conducted a mechanical test with ten matched pairs of young cadaveric femurs and concluded that the application of medial plates to Pauwels Type III femoral neck fractures significantly decreased angular displacement and shear.
In the present study, two patients (4.3%) suffered femoral neck shortening caused by implant failure in MCS group, while none was observed in CMBP group. Three cases (6.4%) developed postoperative nonunion with cannulated screw back out in MCS group and one (4.5%) in CMBP group. Nonunion is the main complication following internal fixation of the femoral neck fractures18. The 91% union rate in CMBP group is favorable comparable to the study for Pauwels type III fractures, in which union rates ranging from 73.7–89%5, 12, 19, 20.
We speculate that the satisfactory outcomes probably derived from anatomic reduction under the direct vision and the stable fixation under the combination of intramedullary and extramedullary in the CMBP group. Thus, there is a potential benefit that combine a medial buttress plate with cannulated screws in the anteromedial region of the femur neck improves the mechanical resistance of fixation in unstable fractures, especially in Pauwels type Ⅱ and type III.
Preservation of the blood supply to the femoral head is most important in the treatment of femoral neck fractures. The medial circumflex artery (MFCA) is the main vascular supply to the femoral head, and the superior retinacular artery (SRA) is the most important branch21–23. The SRA and its branches are closely located to the femoral neck and are easily destroyed in displaced femoral neck fractures22, 24. It has also been shown that the inferior retinacular artery (IRA) also plays an important role in femoral head perfusion21, 25. IRA was present in all hips and the point of entry into the hip capsule as well as intra-articular course were consistent. The study by Sara et al.23 determined the anatomical (especially in vascular) feasibility of this approach to the treatment of femoral neck fractures via an anterior approach to the hip. They described the location of the vessels around the femoral neck by using a clock-face system, with the IRA coursing posteromedially between 7:00 and 8:00, placing a straight plate medially or slightly anteriorly will not disrupt the IRA. In addition, lesser trochanter can be used as a positioning reference and the safe zone for placing the straight plate is just anterior to the lesser trochanter (5:00–6:00) when IRA is less well-visualized. The results of these studies not only provided an anatomical basis for the application of the medial femoral neck plate, but will also change the misconception that the plate would disrupts the blood supply to the femoral head.
In the present study, for patients in the CMBP group, the reconstruction plate was placed anteromedially at the femoral neck via a modified Smith-Petersen approach, which is consistent with the recommendations of Sara et.al23. Compared with the MCS group, open reduction allowed emptying of the intra-articular hematoma, reducing the risk of avascular necrosis of the femoral head due to high intra-articular pressure26. In addition, the femur is usually slightly externally rotated during the procedure to aid in visualization of the femoral neck. At this time, the Weibrecht's ligament and the IRA that travels within are naturally fall away from the femoral neck, reducing the risk of damaging the IRA during the placement of the plate. During the follow-up, the absence of avascular necrosis in the CMBP group supports that this surgical approach does not increase the osteonecrosis rate of Garden type III-IV fractures. On the other hand, postoperative nonunion is probably related to blood supply, 2 (4.3%) patients in the MCS group and 1 (4.5%) in the CMBP group experiencing nonunion excluding implant failure. This supports that the medial buttress plate does not increase the nonunion rate.
As shown in Table 1, there is no doubt that the CMBP group has longer operation time and more intraoperative blood loss. This is acceptable given the benefits that the medial buttress plate brings to the patient in terms of maintenance of fracture stability. Which should be noticed is that, similar to the results reported by Lucas14, 2 patients in the CBMP group had impingement with hip flexion, which is probably due to the plate being too proximal.
There are some limitations to our study. First, given that this is retrospective data from the same institution with a relatively small sample size, prudent attitude should be adopted regarding the interpretations of our outcomes. Second, in order to get more persuasive and generalizable results, more comprehensive studies should be conducted, such as multicenter trials with large sample sizes.