This retrospective study investigated the risk factors leading to anteromedial cortical support failure in follow-ups for pertrochanteric fractures treated with cephalomedullary nails. The reduction quality and residual calcar fracture gapping between the head-neck and femoral shaft fragments in the AP and lateral views were reliable predictors of anteromedial cortical support alterations. These findings further developed the concept of cortical support theory with quantification.
Anteromedial cortical support reduction
Anteromedial cortical support reduction was introduced by Chang et al. [11] in 2015 for pertrochanteric fractures. Anteromedial cortical support reduction is a non-anatomic buttress reduction that allows limited sliding along the lag screw/helical blade axis and achieves secondary stability, providing mechanical stability to share loads from the implant and biological environment for fracture healing [1, 13, 17]. In contrast, the mean loss of the femoral neck-shaft angle, neck shortening, varus deformity, healing complications, and implant failure in the negative medial cortical support group was significantly higher than positive/neutral medial cortical support groups [10, 12, 27]. Similarly, our research revealed that mechanical complications (including varus displacement and excessive lateral migration) were frequently observed in the negative anteromedial cortex support group.
Postoperative alteration of anteromedial cortical support reduction
In 2018, Chang et al. [13] revealed the discrepancy of anteromedial cortical support reduction in pertrochanteric fractures between immediate postoperative fluoroscopy and postoperative 3D CT reconstructions. Acceptable anteromedial cortical contact in both AP and lateral views (positive or neutral) endured an approximately 20% loss rate of cortical support postoperatively. Furthermore, Chen et al. [17] added a 30-degree oblique view (displaying the anteromedial inferior corner cortices) to estimate anteromedial cortex contact status. The change rate still reached up to 10%. However, the possible risk factors predicting the change in the final cortical contact are still out of reach. Therefore, this research preliminarily revealed that fracture reduction and calcar fracture gapping played an essential role in inducing the alteration of anteromedial cortical support.
Importance of fracture reduction quality
In 1980, Prof. Kaufer [28] proposed five factors influencing the treatment outcomes of trochanteric fractures: bone quality, fracture morphology, implant choice, implant placement, and fracture reduction quality. The first two factors belong to fracture characteristics and are non-modifiable. The last three factors are modifiable and could be controlled by the orthopaedist. Among the five factors, fracture reduction is the first manipulation during operation and is regarded as paramount compared with other factors [29].
In 1995, Baumgartner et al. [24] proposed the criteria for fracture reduction quality, including both the alignment and displacement degree of the main fragments. The fracture reduction quality was rated as good, acceptable, and poor. Based on the alignment and medial/anterior cortex contact status, Chang and colleagues further proposed [11] a new fracture reduction criterion for trochanteric fractures, which is more reliable in indicating postoperative mechanical complications than the Baumgaertner reduction system [12]. Studies have revealed that poor reduction quality is an independent risk factor for implant failure and inferior outcomes [19, 30]. Similarly, this study found that good reduction quality was an independent protective predictor of anteromedial cortical support alteration in pertrochanteric fractures treated with intramedullary nails (OR, 0.097; 95% CI, 0.022–0.430; p = 0.002). Our result was consistent with previous studies concerning the importance of fracture reduction quality. Therefore, we should emphasize the importance of reduction quality before the intramedullary nailing process.
Calcar fracture gapping
During the surgery, anteromedial calcar fracture gapping was frequent. Zhang et al. [31, 32] reported that an inferior medial gap (mean and standard deviation, 9.2 ± 4.6 mm) appears during intramedullary nailing in basicervical trochanteric fractures, called the reverse wedge effect. In addition, inferior placement of the guidewire is beneficial for acquiring an ideal Cal-TAD. However, when the spiral blade is hammered into the head and neck bone, it easily impinges on the bone block due to the narrow local space and promotes separation of the bone block, which is called the impingement effect [33]. It is worth noting that calcar fracture gapping is also a subset criterion for fracture reduction quality. Baumgaertner reduction quality criteria considered the displacement of fragments in the AP or lateral view of more than 4 mm as unacceptable criteria [24, 34].
Existing larger calcar fracture gapping is associated with a poor prognosis. Parry et al. [20] revealed that calcar gapping and fracture classification was related to over sliding of lag screws leading to revision surgeries, which is consistent with our results. Ciufo et al. [19] also revealed that basicervical gapping (defined as cortical diastasis > 3 mm) and mal-reduction were risk factors for cutting out in trochanteric fractures treated by cephalomedullary nails. Lobo-Escolar et al. [35] demonstrated that diastasis of fragments > 3 mm postoperatively was significantly more common in the cut-out group treated with femoral intramedullary nailing. In this research, we discovered that calcar fracture gapping in the AP (OR, 2.022; 95% CI, 1.456–2.808; p < 0.001) and lateral (OR, 1.437; 95% CI, 1.060–1.947; p = 0.019) views were risk factors leading to the negative transformation of anteromedial cortical support.
Mechanisms lead to the change
Limited fracture gapping is beneficial for subsequent fragment sliding and secondary stability to achieve medial cortex-to-cortex contact. However, the cortex-to-cortex buttress will not be realised if the space is larger than one cortical thickness. Due to the biomechanical characteristics of the hip joint, the head-neck fragment obliquely slides in the inferior lateral direction until it contacts the intramedullary nail, which gains implant support. Therefore, the existing larger calcar gap might lead to the loss of anteromedial cortical support during the sliding process [1]. In addition to calcar fracture gapping, other factors might also interfere with sliding and change anteromedial cortical support. Possible factors include the capacity to activate head-neck sliding, the orientation of sliding, external rotation of the femoral shaft, and rotation during sliding.
Study limitations
The limits of this research were the retrospective analysis and the number of cases enrolled. The intraoperative fluoroscopic images were low in resolution, which possibly led to an error during the measurement. However, intraoperative CT was not always available in the majority of operating theatres. The pixel ratio calculation method was beneficial for decreasing the error. Osteoporosis severity was not analysed in this research, as we considered that patients over 60 years old who suffered from low-energy fractures could be diagnosed with severe osteoporosis.