Fracture reduction with positive medial cortical support: a key element in stability reconstruction for the unstable pertrochanteric hip fractures

Geriatric pertrochanteric hip fractures are still a major orthopedic challenge worldwide [1]. Despite the fact that fracture union rates are high, the functional outcomes tend to be disappointing [2‐4]. A combination of factors, such as medical comorbidities, patient compliance, fracture pattern, quality of the bone, and environmental factors are thought to be responsible for this poor result [5‐9]. Many of these factors cannot be addressed at the time of fracture presentation.

As the operative procedure is a major component in the treatment of patients with hip fractures, understanding the causes of failure is integral to any attempt to achieve an improved functional outcome. In 1980, Kaufer [10] described five major factors related to the treatment outcome, i.e. the bone quality, the fragment geometry, the choice of implant, the quality of reduction, and the placement of the implant in femoral head. However, the stability of the fracture after implant fixation is primarily dependent on the quality of fracture reduction. It is well known that slight valgus position to allow impaction means more stable fracture reduction and implies better outcome. Besides the valgus alignment, it is paramount important to achieve an anatomical contact between the anteromedial cortices of the two major fragments, the head–neck and the shaft [11‐14].

In this paper, we describe the concept of positive medial cortical support (PMCS) in fracture reduction of unstable pertrochanteric fractures treated with cephalomedullary nail. PMCS is defined as the medial cortex of the head–neck fragment is displaced and located a little bit superomedially to the medial cortex of the femur shaft in AP view. PMCS reduction is a key element for stability reconstruction for unstable fractures, as it allows limited sliding of the head–neck fragment after operation (fracture impaction) to contact with the femur shaft and achieve secondary stability, providing a good mechanical environment for fracture healing. PMCS differs from the anatomic reduction of the anteromedial cortex. PMCS is a functional nonanatomic buttress reduction, which is easy to achieve in practice and is used for description of secondary stability after sliding impaction. While exact anatomic reduction is difficult to obtain and is used for primary fracture stability.

According the relationship of medial cortex position in AP radiographs, there were 89 cases (70 %) in positive support, 26 (20.5 %) in neutral support, and 12 (9.5 %) in negative support. There were no statistical differences among the three groups in age, sex ratio, prefracture score of activity of daily living (pre-ADL), walking ability score (WAC), ASA physical risk score, number of medical comorbidities, osteoporosis Singh index, and the position of lag screw or helical blade in femoral head (TAD). The demographics and operative data are given in Table 1.

At 3 months follow-up, there was minimal difference in radiograph measurement between the injured and the normal contralateral extremity, both in PMCS and in NP groups. As for the NMCS group, the mean neck–shaft angle and the femoral neck length lost significantly compared to the normal side.

The mean loss of the femoral neck–shaft angle in PMCS, NP and NMCS groups was 0.7°, 4.8°, and 8.9°, respectively. The differences among these three groups were statistical significant. The same trend was presented in the neck shortening, which was 2.4 mm in PMCS group, 3.5 mm in NP group and 6.7 mm in NMCS group. The PMCS group had least loss both in femoral neck–shaft angle and neck length. Compared with NMCS group, PMCS group also got ground-walking (full-weight bearing walking) much earlier, with better functional outcome at 3 months follow-up and less hip–thigh pain presence (Table 2).

In the operation of unstable pertrochanteric fractures, anatomic reduction is always prior to the recommended positions of variety implants. Although the posteromedial cortex alignment is the key for successful reduction, most implants used today do not have the ability to purchase the less trochanteric fragment. According to the reduction criteria modified by Baumgaetner, most unstable fractures (31A2-3) could only be achieved “acceptable” reduction grade, i.e. good alignment. For these fractures, the Garden alignments and anteromedial contact between the femoral head–neck and shaft fragments are extremely important [1]. However, valgus position in fracture alignment is not synonymous to positive medial cortical support in fragment displacement.

Compression of the bone fragments is beneficial to bone healing. For unstable pertrochanteric fractures, it can be achieved through two approaches: intraoperative fracture compression and postoperative impaction via controlled sliding along the axis of the instrument device (helical blade or lag screw). The former is the maneuver done by the surgeon during surgery to compress the fracture site through which to obtain primary fracture stability, while the latter is the postsurgical compression provided by a fixation device with a sliding capability, in association with muscle contraction and patient weight bearing, attained secondary fracture stability.

Controlled fracture impaction by limited sliding, provides secondary axial and torsional stability between the head–neck fragment and the femur shaft. Controlled fracture impaction is particularly important for the maintenance of stable reduction during fracture healing, and is compatible with the subsequent dynamic events of cyclic loading and remodeling across the fracture line. In contrast, fracture collapse, also termed uncontrolled fracture impaction, or excessive sliding, is fracture impaction-displacement, with loss of reduction. Fracture collapse is one of the major reasons for failure of fixation of these fractures.

The concept of nonanatomic positive cortex buttress reduction was firstly introduced by Gotfried [20, 21] for displaced subcapital femoral neck fracture. On the premise of 180° fracture alignment in lateral view, it was defined a displaced subcapital femoral position, AP view, in which the distal femoral neck fragment is positioned medially to the lower-medial edge of the proximal fracture fragment. In this state, the distal fragment can limit the femoral head excessive sliding through cortex-to-cortex buttress [22].

We present a counterpart concept of positive medial cortical support in unstable pertrochanteric fractures. It also demand a 180° fracture alignment in lateral view, while in AP view, contrary to the Gotfried’s standard, the distal femoral shaft fragment is intentionally positioned a little bit laterally to the lower-medial edge of the proximal fracture fragment. Unlike the usual displaced route of the proximal fragment in unstable femoral neck fractures, for pertrochanteric fractures, when sliding begins after surgery, the head–neck fragment is tended to displace laterally, impacted into the comminuted and low-intensity trochanteric region, which finally led to collapse (Fig. 4). As in positive medial cortical support position, the cortex contact between the two main fragments are achieved, meanwhile, the medial cortex of the femoral shaft can resist the femoral head–neck fragment from further sliding laterally (Fig. 5). The anterior cortical contact after head–neck sliding can also provide rigid buttress for secondary stability [23, 24]. However, considering the essence of lateral sliding direction, we think positive medial cortical support maybe more effective than anterior cortical contact [14]. In addition, obtaining both medial and anterior cortical buttress (anteromedial reduction) is the best option for pertrochanteric fragment reduction.

However, exact anatomic reduction of anteromedial cortex is rare in reality. The so-called “anatomic reduction” shown in intraoperative fluoroscopy may actually contain three sub-conditions: some are in exact anatomic cortex-to-cortex position, others in slight positive position, and still others in slight negative position. But as the image resolution was limited, those sub-conditions were hardly to be distinguished clearly. So we used the term “neutral” to instead “anatomic”. After bone resorption of the fracture line, slight negative position might become truly negative position. In our case series, five patients with neutral cortical reduction (5/26, 19 %) became negative reduction later, and their outcomes were lower. In lateral radiographs, these five cases also lost their anterior cortical support. The neck cortex was located posteriorly to the shaft cortex more than one cortical thickness. The other 21 cases had a real anatomic reduction of the medial cortex, and obtained a PMCS.

In our series, using cephalomedullary nail seems easy to get a positive medial cortical support reduction (89/127, 70 %). One possible explanation is that pertrochanteric fracture is a kind of extracapsular fracture, the traction applied to the leg can relatively easy to separate the two main fragments. When insert the nail from the medial edge of the greater trochanter, wedge-open effect [25] may occur between the femoral head–neck fragment and the lateral wall, the nail may push the lateral wall, and move the shaft laterally, which makes the shaft fragment be positioned laterally to the lower-medial edge of the proximal head–neck fragment. Positive medial cortex support reduction and wedge-open effect can increase the femoral off-set theoretically, which is beneficial to the strength of the abductor muscles. However, over distraction and/or open (greater than one cortex) may decrease the impaction area among the fragments, lead to delayed union or nonunion.

Now in practice, for unstable pertrochanteric fractures, we attempt to achieve an ideal anteromedial reduction between the head–neck and shaft fragments, i.e. slight valgus and 160°–180° for alignment, positive or anatomic medial cortical support and smooth anterior cortical contact for displacement, in AP and lateral radiography, respectively (Fig. 6).

Our recommendations for good quality of fracture reduction (Table 3), include slight valgus position in alignment and positive medial cortical support in displacement (AP view), and central axial alignment with smooth anterior cortex contact (sagittal view).

In conclusion, fracture reduction with positive medial cortical support and valgus alignment, allows limited sliding of the head–neck fragment to contact with the femur shaft and achieve secondary stability, providing a good mechanical environment for fracture healing.