Fibula allograft with cannulated screw fixation versus ordinary cannulated screw fixation for femoral neck fractures: a 10-year retrospective comparative study

Inclusion criteria are as follows: (i) patients aged 18–60 years; (ii) unilateral displaced transcervical femoral neck fracture; (iii) patients were treated with allogeneic fibula combined with cannulated screw fixation or ordinary cannulated screw fixation; (iv) postoperative follow-up time of more than 5 years.

Exclusion criteria are as follows: (i) surgical area with poor skin conditions or skin diseases; (ii) severe osteoporosis; (iii) pathological fractures or old fractures (more than 14 days); (iv) comorbid rheumatoid arthritis, osteomyelitis, severe cardiovascular and cerebrovascular diseases; (v) incomplete clinical data.

A total of 186 candidates who were diagnosed as femoral neck fractures at the First Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine from January 2006 to January 2012 were initially involved in this study. Sixteen of them did not fulfill the inclusion criteria, 9 were lost to follow-up, and 5 withdrew from this study for other reasons (Fig. 1). The demographic characteristics and clinical profiles including age, gender, the mechanism of injury, the side of fracture, time to surgery, body mass index (BMI), and follow-up duration were recorded. Finally, 156 eligible subjects were selected and assigned to the combination group (fibula allograft combined with cannulated screw fixation, n = 76) and the control group (ordinary internal fixation with parallel cannulated screw in an inverted triangle configuration, n = 80) according to surgical protocols. All participants provided written informed consent, and the current study was approved by the Ethics Committee of the First Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine(No: JK《2019》061).

Operations were performed by the same surgeon group. All subjects underwent lumbar anesthesia. The patient was placed supine with legs apart on an orthopedic table. A manual reduction for the fracture was performed, and the C arm X-ray guidance system was employed to confirm that the hip joint was in a good reduction position. The affected leg was slightly abducted and internally rotated. A longitudinal incision about 5 cm was made below the tip of the greater trochanter. One or two threaded Kirschner wires were inserted into the femoral head and neck to temporarily fix the fracture. The above procedures including drilling and insertion were to keep an anatomically reduced position and avoid inversion and recline. For subjects in the control group, three parallel guide needles were inserted into the femoral head along the longitudinal axis of femoral neck in an inverted triangle configuration. After confirming a proper screw placement, three 7.3-mm cannulated screws were screwed and finally pressed evenly, the distal threads completely passed the fracture line and the top of the screw reached 5–10 mm below the femoral head cartilage. Moreover, the screw should be as close to the cortex as possible [12]. Notably, to reduce stress concentration, the screw entry point should not be lower than the horizontal plane of the lesser trochanter. For subjects in the combination group, the first cannulated screw was placed within 3 mm from the anterior and inferior cortex of the femur. Subsequently, allogeneic fibula graft was inserted into the weight-bearing area of the anterior and upper femur. The top of the fibula graft should reach 5–8 mm below the femoral head cartilage. The diameter of the fibula graft was 11–12 mm. Because the bone tissue has a certain elasticity, the inner diameter of the implanted fibula column should be about 0.5–1 mm smaller than the allogeneic fibula column so that the fibula column could be implanted into the weight-bearing position by compression. The second cannulated screw were placed just below the allofibular column and tightly fixated. Notably, the top of the cannulated screw inside and below the fibula should not exceed the distal end of the fibula, and the two screws alongside the grafted fibular column formed a powerful mechanical construct (Fig. 2).

To avoid postoperative infections, antibiotics were carefully used according to an individual’s condition during the perioperative period. All subjects were administered by the subcutaneous injection of low-molecular-weight heparin to prevent venous thrombosis after surgery. Postoperative rehabilitation started from nonweight-bearing bilaterally in the first 6 weeks to partial weight-bearing in the following 6 weeks with crutches, and to full weight-bearing with or without crutches according to different phases of fracture healing. Each subject was followed up clinically and radiologically at 3, 6, and 12 months and every year postoperatively. At each follow-up visit, AP plain radiographs, computed tomography (CT) and magnetic resonance imaging (MRI) of the affected hip joint were reexamined (Figs. 3 and 4).

Medical records consisted of operative time, intra-operative blood loss, incidence of postoperative and bone healing complications, fracture healing time, femoral neck shortening and the time course of recovery of full-weight-bearing. The HHS was used to evaluate postoperative recovery of hip joint function [13]. The EQ-5D was used to evaluate living quality [14]. Nonunion was defined as persistence of the fracture line 6 months after the surgical procedure [15]. Femoral head necrosis was judged according to the criteria described by Slobogean et al. [16]. Femoral neck shortening was identified by measuring the difference in the length of on the affected side compared with the normal side on pelvic orthotopic X-ray photographs [17]. Fixation failure was defined as fracture displacement (> 5 mm) and femoral neck shortening (> 10 mm), nail withdrawal, or varus deformity (> 10°) compared with initial reduction [18].

Continuous data were expressed as mean ± standard deviation (SD) and analyzed using the independent-samples t test. Categorical data were expressed as absolute numbers or percentages and analyzed using χ² test. Statistical analyses were performed using SPSS 19.0 software (SPSS Inc, USA). A P value < 0.05 was considered statistically different.

Finally, 156 subjects (102 males and 54 females) were selected in our study, and they all completed visits during the follow-up, as shown in Fig. 1. Of all subjects, 76 received allograft fibular cannulated screw fixation and 80 accepted ordinary cannulated screw fixation, with a mean age of 40.3 ± 5.30 and 41.0 ± 6.30 years for the two groups (Table 1). There were no statistically significant differences in age, gender, the mechanism of injury, the side of fracture, time to surgery (3.77 ± 0.82 vs. 3.64 ± 1.0 days, P = 0.383), BMI (23.5 ± 2.58 vs. 24.1 ± 2.50, P = 0.17), Garden classification, Pauwels classification and follow-up duration (11.3 ± 1.42 vs. 10.9 ± 1.72 years, P = 0.207) between the combination and control groups (P > 0.05). The baseline clinical and demographic characteristics of these subjects are summarized in Table 1.

The combination group showed significantly longer mean operative time than the control group (61.0 ± 5.46 vs. 55.9 ± 5.67 min, P < 0.001). This suggested that fibula allograft combined with cannulated screw fixation slightly prolonged the operation time. There was more mean intra-operative blood loss in the combination group than that in the control group, with a significant difference between the two groups (66.4 ± 5.61 vs. 63.6 ± 8.03 ml, P = 0.013). No remarkable difference in the length of hospital stay was found between the two groups (6.80 ± 1.83 vs. 6.51 ± 1.49, P = 0.209) (Table 2).

The mean healing time of femoral neck fractures in the combination group was significantly shorter than that in the control group (3.5 ± 0.84 vs. 3.86 ± 0.91 months, P = 0.016), indicating a superiority of the combination strategy in functional recovery. The time course of recovery of full weight-bearing was significantly shorter in the combination group than that in the control group (3.6 ± 0.39 vs. 4.1 ± 0.5 months, P < 0.001). As for postoperative hip function of the two groups, there was a nonsignificant difference in the HHS score between the two groups at 3-month (60.4 ± 7.10 vs. 59.0 ± 6.36, P = 0.613) and 1-year (80.8 ± 7.15 vs. 79.3 ± 6.99, P = 0.195) follow-up. Strikingly, the HHS scores at 3-year and 8-year follow-up in the combination group (90.2 ± 2.86 vs. 91.8 ± 4.85) were significantly higher than those in the control group (85.4 ± 5.49 vs. 86.8 ± 6.62), respectively (P < 0.001). With respect to postoperative quality of life, the two groups showed similar EQ-5D scores at 3-month (0.41 ± 0.079 vs. 0.38 ± 0.087, P = 0.061) and 1-year follow-up (0.72 ± 0.053 vs. 0.70 ± 0.062, P > 0.153). The combination group exhibited significantly higher EQ-5D scores at 3-year (0.83 ± 0.47 vs. 0.81 ± 0.59, P = 0.005) and 8-year (0.85 ± 0.50 vs. 0.82 ± 0.07, P = 0.003) follow-up than the control group (Table 3).

The rate of femoral head necrosis was significantly lower in the combination group than that in the control group (30.3% vs. 37.5%, P = 0.049). That table shows that a high percentage of patients in the control group progressed to femoral osteonecrosis in 37.5%, although 63.5% of patients eventually healed (Table 4). In addition, the rate of fracture nonunion significantly decreased in the combination group compared with the control group (2.6% vs. 11.3%, P = 0.036). However, there was no statistical difference in the rate of fixation failure between the two groups (P = 0.336). The rates of femoral neck shortening > 5 mm in the horizontal (x) and vertical (y) were higher in the control group than that in the combination group (x: 20.0% vs. 6.6%, P = 0.014; y: 17.5% vs. 5.3%, P = 0.017). About 6.6% subjects receiving fibula allograft with cannulated screw fixation for femoral neck fracture underwent total hip replacement surgery, which was significantly lower than the proportion of 17.5% in the control group (P = 0.031) (Table 4). The Kaplan–Meier survival analysis was performed to assess the survival rate of the femoral head, with the total hip replacement as the end point. The survival rate was compared between the two groups using the Mantel-Cox log-rank test. As shown in Fig. 5, the 10-year survival rate of the femoral head in the combination group was significantly higher than that in the control group (93.4% vs 80%, P = 0.039).