Locking compression plate fixation of femoral intertrochanteric fractures in patients with preexisting proximal femoral deformity: a retrospective study

Inclusion criteria: (1) patients aged over 18 years old; (2) patients with fresh fractures; (3) X-ray film and CT scan demonstrating femoral intertrochanteric fractures and preexisting proximal femoral deformity; (4) no knee dislocation, no vascular or nerve injury; (5) complete clinical demographic and follow-up data. Exclusion criteria: (1) patients with open fractures; (2) patients who refused surgery; (3) incomplete clinical demographic and follow-up data.

A single-center, case series, retrospective study was conducted at the Lower Limb Surgery Ward of the Traumatic Orthopedic Department, Xi’an Honghui Hospital from January 2013 to July 2019. All eligible data including age, gender, mechanism of injury, AO classification, days before operation, complications, time of bone healing, HOOS, and SF-36 scores were collected and reviewed. Thirty-seven patients, including 24 males and 13 females aged from 23 to 69 years old with a mean age of 47.5 years, were enrolled in this study. According to the distribution of the mechanism of injury, 31 cases resulted from accidental falling from a standing height and 6 cases from a road traffic accident. The causes of the preexisting proximal femoral deformities, included 8 cases of proximal femoral fibrous dysplasia, 9 cases of poliomyelitis sequela, 18 cases of malunion, and 2 cases of internal fixation implant failure combined with coxa vara after intramedullary nailing fixation. There were 6 cases of 31-A2.1, 6 cases of 31-A2.2, 20 cases of 31-A3.1, 5 cases of 31-A3.2 based on AO classification of intertrochanteric fractures. All patients or their families provided written informed consent for participation before surgery and the study was approved by the Ethics Committee of Xi’an Honghui Hospital.

The patients were placed under epidural anesthesia or general anesthesia in the supine position on an orthopedic surgical table. The proximal femoral lateral approach was performed after routine disinfection and toweling with an average length of 15 cm ranged from 12 to 18 cm depending on the pattern of the fracture. The skin, subcutaneous tissue, and fascia were then bluntly separated from the lateral femoral muscle layer-by-layer to expose the fracture site. The lateral side of the proximal femur was measured using a testing template after the fracture was successfully restored, which was then used as a reference to reshape the locking plate. The plate was placed on the lateral side of the proximal femur and locking screws were placed into each appropriate hole according to the position of the fracture line. For patients with poliomyelitis sequela, the main challenge was the presence of an abnormal femoral neck angle and anteversion angle and an extremely narrow medullary cavity, which caused difficulty in achieving effective fixation using a 4.5-mm bone plate. Under these circumstances, we resorted to epiphyseal plate fixation so that the proximal end could be fixed with multiple screws and at different angles. Patients with proximal femoral fibrous dysplasia were treated by completely removing lesions in the femur, followed by reshaping of the proximal femoral locking compression plate or reverse distal femoral locking compression plate [9], which was then used for fixation after implantation of the autologous iliac bone combined with allogeneic bone. For patients with malunion, the fracture was reduced under direct vision based on the location of the fracture and fixed with reverse distal femoral locking compression plate. Patients with implant breakage and coxa vara after proximal femoral nail fixation were fixed with reverse femoral LISS bone plate after removal of the original failure internal fixation plate. A 3D model was printed prior to surgery and was used for bone plate reshaping, decision-making of plate length, and the number of screws. The reduction and internal fixation of the fractures were evaluated using intraoperative C-arm fluoroscopy. Once satisfactory reduction was achieved, irrigation and suction drainage were performed and the incision was subsequently sutured.

After cold compress for 24 h and routine second-generation cephalosporin antibiotics were administered postoperatively for 48 h to prevent infection, and the drainage tube was removed. Comorbidity diseases, such as osteoporosis, were treated using alendronate and vitamin D3. Physical prevention (quadriceps muscle contraction and relaxation exercise, and CPM training) and thrombolytics or anticoagulant therapy were administered after surgery to prevent deep vein thrombosis of the lower extremities. If the patient presented with deep vein thrombosis, different methods of treatment were provided according to the location of the thrombosis. The inferior vena cava filter implantation was performed if thrombosis located proximal to the level of the popliteal vein to decrease the risk of pulmonary embolism, otherwise only low molecular weight heparin was administered. X-rays of the anteroposterior and lateral femurs were taken after the postoperative removal of the drainage tube to investigate reduction and implant location, and the wound sutures were removed after 2 weeks. Follow-up took place once a month for the first 3 months after surgery, then every 3 to 6 months after the first 3 follow-ups visits of the surgery, and finally every 6 to 12 months after 1 year of surgery. The presence of a callus formation observed through X-ray examination implied that the patient was ready for partial weight-bearing activity, while full weight-bearing depended on the condition of the fracture healing.

Various indexes, including days before surgery, bone formation condition, complications (internal fixation loosening, fracture, nonunion, and deep infection), and fracture healing time, were assessed through the hip disability and osteoarthritis score (HOOS) after fixation with the locking compression plate, which is generally adopted for total hip arthroplasty for hip osteoarthritis. However, it shows good discriminability in evaluating the outcomes of hip and femoral fractures [10, 11]. Hip function was evaluated based on 5 aspects: pain, symptoms, activity limitations-daily living, sport and recreation function, and quality of life [12]. When the patient was first admitted to hospital, they were asked to self-evaluate the quality of life less than 6 months before injury using the SF-36 Health Questionnaire [13], and at the last follow-up, the postoperative quality of life. The 8 SF-36 subscales included physiological function, role-physical, bodily pain, general health, vitality, social function, role-emotional, and mental health. A higher score indicated lower functional damage and a better quality of life.

Statistical analysis was performed using SPSS 24.0 statistical software (IBM, USA). A Shapiro-Wilk test for normality was conducted on all continuous data, and continuous data with a normal distribution were described as mean ± standard deviation, whereas categorical data was described in several cases (percentages). The hip scores of the HOOS and the SF-36 Health Questionnaire were expressed as mean ± standard deviation. The HOOS and the SF-36 Health Questionnaire scores before injury and at the last follow-up were compared using Student’s t test and the Mann-Whitney U test for those without normal distribution. The χ2 test was performed for categorical variables. A p value of < 0.05 was selected as the threshold of statistical significance.

The AO/OTA classification of the 37 patients showed that 68% of the patients had an extra articular intertrochanteric fracture, including 20 cases of type 31-A3.1 and 5 cases of type 31-A3.2, whereas the other 32% of the patients suffered comminuted extra articular pertrochanteric fractures, including 6 cases of type 31-A2.1 and 6 case of type 31-A2.2. None of the 37 fractures were extremely comminuted, and the type 3 of the AO/OTA subgroup was not presented. Radiographic data analysis indicated that all fractures showed various degrees of deformity of the coxa vara. Previous studies [19, 20] have shown that the hip joint loaded approximately 3 to 5 times of the normal body weight due to the presence of a coxa vara deformity, leading to excessive loading on the tension side of the femur, which may cause local overload. Therefore, fractures of type A3 were the main pattern in our study due to the onset of preexisting coxa vara deformities before the injury and local overloading on the tension-side of the proximal femur due to additional weight-bearing. Meanwhile, the physical load increased on the tension side of the proximal femur when the tension-side suffered external loading until it exceeded the ultimate limit of bone toughness. Then the tension side would break first and the pressure side broke as soon as the force was conducted to the internal side. In addition, the affected limb was already disabled before the injury, therefore it would have been subconsciously protected during daily life. During the course of intertrochanteric fracture due to external force, the patient was expected to take protective actions to decrease the severity of injury.

In this study, we applied a locking compression plate to treat intertrochanteric fractures in 37 patients with proximal femoral deformity, and drew similar conclusions on the advantages of this option: 1) the shape of the plate can maximally match proximal femoral deformity without damaging the locking hole. 2) the locking compression plate did not affect the stability and grip strength of fracture fixation even if it could not completely and perfectly match proximal femoral deformity after reshaping. 3) due to the original disability in the affected limb, less daily stress stimulation and the onset of disused osteoporosis, screw loosening and cutting, which are likely to occur if regular plates are used. However, these problems may be solved by using a locking plate.

For these patients, the opportune moment of postoperative weight-bearing should be taken seriously into account: (1) although the locking compression plate was very strong, it had to be reshaped to fit the particular fracture. This reshaping process might partially damage the plate and compromise its stiffness and toughness. If the internal fixation is subjected to weight-bearing overload before the fracture heals, the bone plate may break easily; (2) due to the disability of the affected limb before injury, the contralateral healthy limb became the main weight-bearing extremity when walking. Due to the relatively poor bone quality and quantity in the affected side, premature weight-bearing could result in failed internal fixation and screw loosening. Some authors [12, 21] have reported that coxa vara deformity often occurs after locking compression plate internal fixation due to severe osteoporosis and weight-bearing activity before the internal femoral cortex establishes adequate buttress function. For this reason, the standing timing and weight-bearing for such patients needs to be determined according to the type of fracture, osteoporosis, and fracture healing conditions.

At present, most authors have focused on osteotomy orthopedic strategies for proximal femoral deformity without fracture and have emphasized on postoperative rehabilitation of the hip joint. To the best of our knowledge, there are only a few reports on management options for this particular type of fracture with preexisting deformities. In non-fracture cases, abnormal axial alignment and rotation are the main predicaments for orthopedic correction of proximal femoral deformity [22]. Some patients, such as those with congenital developmental dysplasia of the hip, are also affected by structural abnormalities, such as dysplasia, small bones, increased anteversion angle, and a backward great trochanter. The choice of osteotomy and internal fixation mainly depends on the degree of anatomical abnormalities [3]. Deng et al. [6] described that in patients with femoral deformities without associated fractures, THA, precise osteotomy, and correct choice of prostheses were key factors for femoral reconstruction and fixation. Kim et al. [23] reported that total hip arthroplasty with trochanteric ostectomy rendered favorable results for patients with deformity of the proximal femur. For these patients with more complicated pathological factors and osteotomy options, in addition to the deformities on the multi-planar surface, should be taken into account in order to maximally improve the function of the affected hip joint. 3D osteotomy of the proximal femur [24] can be used to treat most deformities because the original intertrochanteric or subtrochanteric bone is continuous. For patients with proximal femoral deformity who have suffered fractures, osteotomy cannot be achieved because the proximal femoral fracture line is likely to cross the osteotomy line. As a result, spreading, closing and rotation between the bone fragments after the osteotomy procedure cannot be performed following the original plan, which likely makes the case more complicated and increases the difficulty and risk of the surgery. Therefore, for fracture patients with preexisting proximal femoral deformity, it is not recommended for osteotomy and orthopedic treatment to be performed directly over the course of open reduction and internal fixation after fracture. Patients can be subject to osteotomy and orthopedic surgery if needed after the fracture has healed and the affected limb is able to bear weight-bearing activities.

In patients with femoral trochanter deformity, soft tissues around the hip joint are adapted to this deformed condition and a new level of homeostasis must be established before fracture. If the original condition of bone deformity is not maintained, the balance between soft tissues and the hip joint may be disequilibrated and unpredictable issues, such as pain, abnormal gait, and lifestyle changes, may emerge. Our aim was to enable the patients to recover with a painless hip joint instead of a movable but painful one. Therefore, it is believed that we have reached a consensus on the primary intention of the Harris Score Scale [25], in which the distribution score is more focused on postoperative pain and joint function changes instead of joint activity. This was further confirmed by the finding that no significant differences were found among the 37 patients based on the HOOS score of the hip joint between pre-injury and last follow-up after fracture fixation. These patients were disabled before surgery, and their contribution to the family and society had certainly been compromised. By maintaining the original skeletal malformation in our fixation surgery, we could address family and social issues raised by fractures to provide the lowest economic burden. At the end of the follow-up period, no significant differences were found in the 8 dimensions of quality of life between pre-injury and after surgery, which indicated that after the fractures had healed, the quality of life of the patient was restored to pre-injury levels. In terms of the quality of life of patients, we achieved our primary objectives of treatment.

There are some limitations of our study that should be clarified. The sample size of our retrospective study was relatively small, which may enhance the probability of bias in the statistical results. Thus, further prospective studies with larger sample sizes are required to validate our findings. When applying the locking compression plates for the treatment of these patients, flexibility for reshaping the locking compression plate was limited since the procedure must be performed without damaging the thread of the locking nail. For this reason, some locking screws could not be screwed in at the designated direction and sometimes the screws had to be screwed in another direction or position. Instead of being drilled and screwed in the designated locking sleeve direction, the screws must be drilled and screwed into the fixed direction as required and sometimes cold-welding must be simultaneously performed to achieve stability. The rigidity and toughness of the compression plates changed after reshaping. Therefore, since patients could not fulfill early weight-bearing after surgery and take care of themselves during rehabilitation, the resulting family and social burden was increased to some degree.