Intertrochanteric fracture of the femur is a common type of fracture in older individuals (Socci et al., 2017). If the poor quality of fracture reduction, premature weight-bearing exercise after fracture surgery may lead to serious complications such as fracture re-displacement and cut out, but long-term bed-rest immobilization may also increase the incidence of postoperative complications (Frei et al., 2012; Hua et al., 2014; Tan et al., 2016); hence, femoral intertrochanteric fracture is still considered as an “unsolved fracture” type at present (Haidukewych, 2009; Haidukewych, 2010).

The posterior wall and medial wall are important biological structures of the femoral trochanter (Sharma et al., 2014). Previous studies have shown that the integrity of posterior-medial structure is closely related to its stability after fracture surgery and hip joint functional mobility (Ciufo et al., 2017; Sharma et al., 2017). Marmor et al. suggest that in the process of intramedullary nail treatment (Marmor et al., 2013), fractures become more and more unstable as the severity of medial cortical fragmentation increases, but previous studies often regard the whole posterior wall and medial wall as a whole, and lack further understanding of its structure and biological characteristics. However, with the wide application of CT three-dimensional reconstruction technique in treating intertrochanteric fractures and further understanding of the structures of posterior-medial wall (Futamura et al., 2016; Shoda et al., 2017), we believe that the biological characteristics of the posterolateral (the area near the greater trochanter, with the stop point near the medial edge of the quadratus femoris muscle), posteromedial (the area covering the entire small trochanter) and medial intertrochanteric (the pressure-bearing area of the femoral trochanter) parts may be different (Figure 1), and the possible postoperative complications and precautions caused by fractures in different regions may also be different (Ren et al., 2019; Ren et al., 2020). The medial side of the proximal femur is an important supportive structure. Insufficient medial support after surgery may lead to deformities such as the inward rotation of the femoral head and the disappearance of the neck-shaft angle. The posterior side of the proximal femur includes the lesser tubercle and part of the greater tubercle; they serve as attachment points for some tendons. Improper treatment after surgery could result in partial loss of hip joint mobility. With the wide application of CT three-dimensional (3-D) reconstruction techniques in treating intertrochanteric fractures coupled with further understanding of the structures of posterior-medial wall, this issue can now be explored in detail. The purpose of this study was to classify the fracture types in different posterior-medial regions through CT reconstruction, and to understand the fracture patterns between different types through fracture-mapping techniques. We also aimed to further explore the influence of different fracture regions and classification systems on postoperative function and complications of fractures.

A total of 144 males and 204 females (n = 348) with intertrochanteric fractures were included in this study, with a mean age of 73.5 years (Table 1). No patients were allocated to the isolated posteromedial group or posterior + posteromedial group. Instead, we found that almost all patients had medial wall injuries (339/348, 97.41%). Of all of the patients, 213 (61.21%) were allocated to the posterolateral + posteromedial + medial group, and we further divided them into the posterolateral + posteromedial + isolated fragment medial group and the posterolateral + posteromedial + simple medial group. Furthermore, 111 (31.9%) and 102 (29.3%) patients were in the posterolateral + posteromedial + isolated fragment medial group and posterolateral + posteromedial + simple medial group, respectively, and 81 (23.3%), 33 (9.5%), 12 (3.4%), and 9 (2.6%) in the posterolateral + medial group, the isolated medial group, the posteromedial + medial group, and isolated posterolateral group, respectively (Figure 4).

In this study, we retrospectively modeled 348 intertrochanteric fractures imaged by CT scan, and applied 3-D fracture-mapping techniques to macroscopically analyze the fragment morphology of medial, posteromedial, and posterolateral fractures (Xie et al., 2017). We divided posterior-medial fractures into six different classifications for comparison. Our classification system differs from previous classification systems in that this study more accurately presents the fracture line and fragment morphology of posterior-medial fractures through an advanced fracture-mapping technique, which allows us a more accurate understanding of posterior-medial injuries. In posterior-medial injuries, whether medial, posteromedial, or posterolateral, there are different injury patterns, and the resulting postoperative effects are distinct. By analyzing the posterior-medial fracture pattern, we can further improve our understanding of this fracture type.

In recent years, the popularity of 3-D methods for analysis by CT has also grown. In 2017, Shoda et al. (2017) proposed the classification of intertrochanteric fractures by 3D-CT. In 2019, Li et al. (2019) further classified femoral fractures into five classifications using the Hausdorff distance-based K-means approach (Li et al., 2019). However, we believe that a comprehensive fracture classification should not only accurately diagnose fractures, guide treatment, and predict the prognosis of fractures, but also be reproducible and simple for use by clinicians across the scope of practice.

In our study and classification, we did not describe the lateral wall, because we believe that intertrochanteric fractures involving the lateral wall are a special type of femoral fracture, and their fracture patterns and prognosis are considerably different from other intertrochanteric fracture models. In addition, we believe that the posteromedial structure plays an important biomechanical role in ensuring the stability of the proximal femur (Sharma et al., 2014; Ciufo et al., 2017; Sharma et al., 2017). The loss of support in the posteromedial side is an important cause of femoral head collapse, femoral neck shortening, and internal fixation failure, and studies of the posterior-medial fracture line and fracture fragments have already been reported. In 2017, Sharma et al. (2017) described the size, shape, and fracture mode of the smaller trochanteric fragments. Xiong et al. (2019) further summarized the extended fragment of the lesser trochanter and posterior cortex. Their study provided further insight into the characteristics of posterior-medial fracture structures, but further analyses of the clinical morphology of different fragment morphologies are lacking. Our study classification targets only the posterior-medial fragment of intertrochanteric fractures, and we believe that the inferential pattern of fracture-mapping technique gives us a new spatial perspective on the fracture pattern and morphology of the posterior-medial fragments after intertrochanteric fractures, and will ultimately allow us to link fracture morphology to clinical features.

In our study, posterior-medial intertrochanteric fractures were typed into six classifications. Both type I and type II fracture lines involve the entire posterior-medial intertrochanteric region, leaving the posterior-medial cortex unsupported and the fracture extremely unstable. Moreover, the sliding distance of cephalic nail and FNSA change in the radiographic findings of type I and type II were significantly different from those of the other groups (Table 2; Figure 7). However, even if the extent of type I and type II fracture involvement is consistent, the clinical results are different due to the uniqueness of intertrochanteric fractures. The concept of the “isolated fragment” that we introduced divides intertrochanteric fractures involving the entire posterior-medial fragment into two classifications. We can find in the fracture-mapping technique that the fracture of type I is more complex and is a posterior-medial comminuted fracture. Type II fractures, on the other hand, are relatively simple, and most of the posteromedial aspect remains a complete fragment. There is also a significant difference between type I and type II in the radiography parameters of FNSA change (Figure 7). We believe that type I fractures have a worse outcome than type II fractures and more surgical complications, but further studies are needed to confirm this hypothesis. Although the type IV fracture shown in Figure 7 involves the medial wall fracture, there is no significant difference in the imaging indicators. We trust that the posterior wall will play a partial supporting role in the case of medial wall fracture. At this time, the fracture is still stable when the calcar femorale is intact. As expected, we found that the radiographic appearance of type VI fractures was significantly different from that of the other classifications, further confirming the important role of the posterior-medial fragment in maintaining the stability of intertrochanteric fractures. However, the stability of intertrochanteric fractures involving medial wall injury was not universally affected; medial wall injury accounted for (97.4%) of the patients with posterior-medial fractures, but the radiographic results in patients with type III, IV, and V fractures suggested that the fracture type was relatively stable. HHS function scores indicated a significant difference between type I fracture and other fracture classifications, which also confirmed that neck shortening significantly reduced Harris hip scores (Weil et al., 2012; Slobogean et al., 2017; Felton et al., 2019; Shin et al., 2020). Moreover, the postoperative HHS function score of type II fracture with the same posterior-medial involvement was satisfactory.

We (and many researchers) believe that the entire pathophysiological process of fracture is related to muscle and ligament attachment at the fracture site (Bair and Zafar Gondal, 2020; Slagstad et al., 2020). This belief guided our classification of posterior-medial fragment into posterolateral, posteromedial, and medial groups due to muscle attachment in the intertrochanteric region and its different roles between the trochanters (Marmor et al., 2013; Ehrnthaller et al., 2017; Ren et al., 2019), where both posteromedial and posterolateral are muscle attachments and the medial region is an important supporting structure. Through fracture-mapping, we found that the direction of the fracture line was consistent with this mechanistic view, confirming the reliability and accuracy of our classification. In addition, we also considered the piriformis, obturator internus, obturator externus, superior gemellus, inferior gemellus, and the quadratus to be the external rotator muscle groups with the end points at the piriformis fossa and intertrochanteric ridge of the femur. The gluteus medius and gluteus minimus were the abductor muscle groups with end point at the greater trochanter, while the iliac muscle and psoas were the flexor muscle groups with end points at the lesser trochanter. Therefore, we retrospectively analyzed the postoperative hip range of motion of patients and found that different fracture classifications had unique postoperative range of motion effects, and even the posterolateral fracture alone had a significant effect on hip abduction function. This finding can further guide individualized postoperative functional rehabilitation of patients using hip joint exercise.

This study has several limitations. First, our study is a retrospective study in which we reviewed patients managed operatively for intertrochanteric fractures, while most patients with type VI fractures were managed conservatively and not included in our analysis. As a result, we had a lower proportion of type VI fractures in our study. Second, some patients with type I fractures could not be systematically classified because the fracture was severely displaced and separated into very small fragments. In such patients, fracture morphology of type I is less complex than that of type I fractures, which is not captured by fracture-mapping technique. Finally, the isolated posteromedial + posteromedial group was not represented in our study population. This may be related to our insufficient number of patients, or other yet undefined mechanisms that need to be further elucidated.

In conclusion, the results of this morphologic study helps to further identify and recognize the characteristics of posterior-medial intertrochanteric fracture patterns, which may be strongly associated with different clinical outcomes. More prospective randomized controlled trials are needed to verify the effectiveness of this new classification system.