The Probability of And Risk Factors For Femoral Neck Shortening And Internal Fixation Failure After Femoral Neck Fracture Treatment Using Cannulated Screws Combined With A Medial Buttress Plate

Background: Femoral neck shortening and internal xation failure are two most prevalent complications associated with surgical treatment of femoral neck fracture. This study evaluates the probability of and risk factors for these complications after the treatment of femoral neck fracture using cannulated screws combined with a medial buttress plate. Methods: Data of 70 patients with femoral neck fractures treated using cannulated screws combined with a medial buttress plate in our hospital between March 2016 and March 2019 were retrospectively analyzed, and 40 patients who were followed up for no less than 6 months were enrolled in the study. We evaluated the probability of postoperative femoral neck shortening and internal xation failure and conducted multivariate logistic regression analysis to determine risk factors. Results: Among 40 patients, 11 had femoral neck shortening (cid:0) 5 mm and 8 had fracture nonunion or screw cutout. There were signicant differences in screw quality between the shortening and non-shortening groups; There were signicant differences in fracture site and shortening between the xation failure and non-failure groups. According to the multivariate analysis, screw quality was an independent risk factor for femoral neck shortening (odds ratio [OR]: 8.58; 95% condence interval [CI]: 1.35–50.57; P=0.022), and femoral neck shortening was an independent risk factor for internal xation failure (OR: 11.82; 95% CI: 1.66–84.36; P=0.014). Conclusions: Femoral neck fracture treatment using cannulated screws combined with a medial buttress plate led to a femoral neck shortening rate lower than that of other internal xation methods and an internal xation failure rate lower than or equivalent to that of other internal xation methods. The quality of cannulated screws is a risk factor for postoperative shortening of the femoral neck, and femoral neck shortening is a risk factor for postoperative internal xation failure.


Background
Femoral neck fractures account for almost half of all hip fractures and are most common in the elderly; in young adults, these fractures are often caused by high-energy trauma [1]. The treatment of femoral neck fractures is still challenging for clinicians. Although these treatment strategies and methods are constantly improving, a variety of postoperative complications still exist, with the most common being fracture nonunion, malunion, failure of internal xation, and femoral head necrosis [2][3][4][5]. Parker et al. [6] reported that the nonunion rate of displaced femoral neck fractures after internal xation was up to 30.1%, which is a rate that increases with age. At present, surgeons generally believe that obtaining anatomic reduction and rm xation is the key to the treatment of femoral neck fracture [7]. Therefore, a variety of operation strategies and internal xation methods have been developed in recent years, including the "inverted triangle" parallel arrangement of three cannulated screws [8], a commonly used simple and minimally invasive technique, and the proximal femur angle plate and sliding hip screw (SHS) [9]. On the other hand, biplanar non-parallel screws have recently been reported to have better biomechanical and clinical healing rates than parallel screws; however, long-term follow-up results for this approach are still lacking [10,11]. Although most patients can achieve anatomic reduction through closed reduction, this is often di cult in young patients and in cases of comminuted femoral neck fractures [12,13]. For such patients, an incision is required to achieve anatomical reduction, and the most commonly used method is the direct anterior approach (modi ed Smith-Petersen) introduced by Molnar et al. [14]. Since Hassan Mir [15] rst proposed the concept of the medial buttress plate, the clinical application of cannulated screws combined with a medial buttress plate for the treatment of femoral neck fractures, especially displaced femoral neck fractures with high shear force, has been increasing and achieving good results [16]. However, although this treatment approach for femoral neck fractures signi cantly improved postoperative stability [17], postoperative femoral neck shortening, and fracture nonunion/internal xation failure were still prevalent.
There are numerous reports on the probability and risk factors of femoral neck shortening and internal xation failure after the use of multiple screws and SHS xation treatment approach [18][19][20], but not after the use of cannulated screws combined with a medial buttress plate. Therefore, the purpose of this study was to evaluate the probability of and risk factors for femoral neck shortening and internal xation failure after treatment of femoral neck fracture using cannulated screws combined with a medial buttress plate. The ndings of this study could provide a theoretical basis for clinical treatment decision-making.

Patients
Data from 70 femoral neck fracture patients who were treated with cannulated screws combined with a medial buttress plate in our hospital from March 2016 to March 2019 were retrospectively analyzed. All patients were Garden III/IV or Pauwels II/III. The inclusion criteria were: (1) having follow-up data of ≥6 months and (2)

Surgical technique
All internal xations were performed using the standard basic technique and were completed by the same senior physician. Each patient was placed in the supine position on a radiolucent table. The healthy limb was placed on the abductor frame, the C-arm was set up in advance, and it was veri ed that good anteroposterior and lateral or axial positions of the hip joint could be obtained. All patients rst underwent closed reduction not exceeding three times, upon failure of which, open reduction using direct anterior approach was adapted. A 10-15 cm skin incision was made approximately 2 cm lateral and 2 cm distal to the anterior superior iliac spine. The incision extended distally toward the lateral aspect of the patella. The space between the tensor fascia latae and sartorius was identi ed after subcutaneous incision, and the lateral femoral cutaneous nerve was carefully protected. The rectus femoris was then exposed, and its lateral origin was cut. The sartorius and rectus femoris were pulled to the medial side, and the tensor fascia latae were pulled to the lateral side. The ascending branch of the lateral femoral circum ex artery was ligated, the hip capsule was exposed, and a "T"-shaped capsulotomy was sharply made through the anterior capsule, following which the hematoma was evacuated. A 2.5 mm Kirschner wire was placed into the femoral head and used as a joystick to manipulate the femoral head, aiding in achieving anatomic reduction. Traction, internal rotation, and adduction of the affected limb were performed when necessary. Once the femoral neck fracture had been anatomically reduced, a 2.5 mm Kirschner wire was drilled from the lateral thigh parallel to the femoral neck to maintain the reduction. Then, three cannulated screw guide needles were drilled under the supervision of uoroscopy, and these guide needles were arranged in parallel and in an inverted triangle; the distance to the femoral neck cortex was not more than 5 mm. The appropriate length of the guide needles was determined after which three 6.5 mm cannulated screws, corresponding to that length, were screwed in. The fracture end was directly examined, con rming fracture reduction without loss. Hip abduction, rotation, and exion were also con rmed, such that the medial side of the femoral neck can be exposed. A suitable length of one third tubular or reconstructed medial buttress plate had been pre-bent and was placed on the medial side of the femoral neck. First, a cortical bone screw was drilled into the distal end of the fracture line to press the plate to the medial side of the femoral neck. Then, a cortical bone screw was placed into the femoral head through the plate, and the nal screw passed through the hole at the distal end. Fluoroscopy con rmed proper positioning of the plate and screw. The incision was rinsed and sutured after drainage.
After surgery, a tolerable range of hip motion was immediately permitted. Partial weight bearing of the affected limb was allowed 6 weeks after surgery, but only under the guidance of a physician. Patients were postoperatively followed-up at 6 weeks, 3 months, 6 months, and 12 months, and then every 6 months thereafter.

Measurements
Femoral neck shortening was de ned as a shortening distance >5 mm and was measured using the method described by Zlowodzki et al. [21]; using the screw with the greatest amount of lateral protuberance in the followup radiographs compared with initial xation (Figure 1). Internal xation failure was de ned as fracture nonunion or screw cutout requiring reoperation. All radiographic evaluations and measurements were performed by all the authors of this study. In case of disagreements, the corresponding author made the nal decision. For continuous data, the average of all measurements was calculated after eliminating the two most extreme values, and this average was used in subsequent analyses. Digimizer 4.5.1 (MedCalc Software, Belgium) was used to measure the shortening distance. All images were calibrated using the known screw diameter (6.5 mm). To determine the associated risk factors for shortening, we assessed patients' age, sex, reduction quality, cannulated screw quality, fracture site, and Pauwels classi cation. In addition, shortening was evaluated as a risk factor for internal xation failure. Reduction quality was based on the Garden classi cation [22] and was rated as anatomic, acceptable (within the range of 155°-180° in both anteroposterior and lateral radiographs), borderline acceptable (<155° or >180° in either view), or unacceptable (<150° or >185° in the anteroposterior radiograph alone). Screw position was scored based on subjective evaluation of screw location adequacy and spread in an inverted triangle con guration based on the intraoperative uoroscopic images [23]. The screw position was given a score of 3 if it was excellent and represented optimal surgical technique (three screws obtained 3-point xation), 2 if it was sub-optimal but acceptable (two screws obtained 3-point xation), and 1 if it was poor (only one or no screw obtained 3-point xation) [18]. Since the number of cases scored with 1 point was small, the cases were divided by score into two groups only: "=3" and "≤2." Similarly, because there were fewer cases of basicervical femoral neck fractures, the cases were divided by fracture site into subcephalic and non-subcephalic (transcervical and basicervical). According to the Pauwels classi cation [24], all patients in the study were diagnosed as either Pauwels II or III, and no surgery for type I was required. Similarly, all patients were either Garden types III/IV, and no surgery for Garden type I/II was required.

Statistical Analyses
Independent two-sample t-test was performed for continuous variables, and the chi-square or Fisher exact test was performed for categorical variables. For all tests, a 2-sided p-value of 0.05 was considered signi cant.
Multivariate regression analysis was performed for variables with a p-value less than 0.1. Statistical analyses were performed using SPSS statistics software (version 22.0; SPSS, Chicago, IL, USA).

Results
Among 40 patients enrolled in the study, 11 had femoral neck shortening of > 5 mm and 8 had femoral neck shortening of > 10 mm. Moreover, 8 patients had fracture nonunion or screw cutout, requiring reoperation (Fig. 2).
Among them, 6 patients underwent joint replacement and 2 underwent valgus-producing osteotomy. The average time from surgery to identi cation of nonunion was 4 months (range: 3-5 months). Due to open reduction, postoperative X-rays showed that all fracture reductions were either anatomic or acceptable, and none were borderline acceptable and unacceptable.
According to univariate analyses, there were no differences in age, sex, fracture site, Pauwels type, and reduction quality, but there were signi cant differences in screw quality between the shortening and non-shortening groups. Similarly, there were no differences in age, sex, Pauwels type, reduction quality, and screw quality, but there were signi cant differences in fracture site and shortening between the xation failure and non-failure groups ( Table 1).

Discussion
Femoral neck shortening is a common phenomenon after internal xation of femoral neck fractures. Weil et al. [19] reported a shortening rate of 56% (> 5 mm) and a severe shortening rate of 22% (> 10 mm) after using multiple cannulated screws to x femoral neck fractures. Moreover, Stockton et al. [20] found that femoral neck shortening and severe shortening rates could go up to 54% and 32%, respectively, after using multiple cannulated screws or an SHS + anti-rotating screw, with the former being lower than the latter. Zlowodzki et al. [25] also found, in a multi-center joint study, that the postoperative shortening and severe shortening rates of femoral neck fractures xed by cannulated screws were up to 66% and 30%, respectively. Furthermore, another study showed that the shortening of the femoral neck permanently affects hip function after surgery, primarily impacting gait velocity and gait symmetry [26], with older patients tending not to have the ability to compensate for the impact.
Therefore, determining the appropriate internal xation method to minimize the shortening of the femoral neck has become a research focus.
Since Hassan Mir [15] rst proposed the concept of the medial buttress plate of the femoral neck, relevant anatomical and biomechanical studies have been continuously conducted, verifying the effectiveness of this method [16,17]. Currently, the focus of debate in this type of surgeries is whether open reduction can impair the blood supply of the femoral head, thus increasing the possibility of femoral head necrosis. In this regard, the direct anterior approach has been proven safe and reliable. In 2019, Utnam et al. [27] further con rmed through human autopsy that the placement of the buttress plate on the medial side of the femoral neck had been safe and did not destroy the blood supply to the femoral head, with the optimal placement position being at 6 o'clock of the femoral neck.
In our study, 40 patients with complete follow-up data who received cannulated screws combined with a medial buttress plate were selected for analysis. For this cohort, we found that the postoperative femoral neck shortening rate (> 5 mm) of this internal xation method was 27.5% and that the severe shortening rate was 20%.
It can be seen that the shortening rate of this internal xation method is lower than that of other previously reported internal xation methods; however, the severe shortening rate is less than or equal to that in previous reports [19,20,25]. These ndings suggest that cannulated screws combined with a medial buttress plate can provide higher stability but cannot prevent the shortening of the femoral neck after internal xation.
Stockton et al. [20] conducted a study on femoral neck shortening after xation using the SHS + anti-rotating screw and multiple screws, suggesting that the degree of fracture displacement and internal xation method are independent risk factors for femoral neck shortening. Similarly, our study involving Garden III/IV fractures showed that the quality of cannulated screws was an independent risk factor for femoral neck shortening and that there was no signi cant correlation with age, sex, fracture site, Pauwels type, and reduction quality. These ndings suggest that despite the use of a medial buttress plate, the quality of the cannulated screws still plays the most important role. Risk factors for femoral neck shortening after internal xation have not been reported.
Postoperative internal xation failure of femoral neck fracture is another issue, with a reported failure rate of 10-30% [28]. The nonunion rate of xation using three cannulated screws alone was 21.8% [29]. In our study, all patients had Pauwels type II/III or Garden III/IV fractures, although all had displaced fractures and high shear force, with an internal xation failure rate that was 20% lower than that of xation using simple cannulated screws. The nonunion rate of fractures after xation with cannulated screws, SHS + screw, and locking compression plate is up to 11.9% according to a previous report [30]. However, many patients with Garden II femoral neck fractures had been included in this previous report, and thus, the results are not directly comparable with those of our study.
Numerous studies [28, 29,31] have reported that the internal xation method, degree of fracture displacement, and poor reduction are risk factors for postoperative internal xation failure of femoral neck fractures. However, the risk factors for postoperative internal xation failure of femoral neck fractures, speci cally treated using cannulated screws combined with a medial buttress plate, have not yet been reported. Our study found that femoral neck shortening is an independent risk factor for internal xation failure; however, it was not related to age, sex, reduction quality, Pauwels type, fracture site, or screw quality. Therefore, the prevention of femoral neck shortening should be the main approach for preventing internal xation failure. Nevertheless, more stable internal xation methods still need to be studied.
This study has some limitations. First, due to the lack of multicenter data, as well as our low number of cases and short follow-up times, femoral head necrosis could not be analyzed. Second, since all patients treated with this internal xation method were Garden type III/IV or Pauwels type II/III, Garden type I/II and Pauwels type I have been excluded from the comparative analysis. Due to the adaption of open reduction, all reduction qualities have been categorized as anatomic and acceptable, whereas borderline acceptable and unacceptable were excluded from the comparative analysis. Finally, because image examination in some patients was not standard, there was inevitable error in this data subset.

Conclusions
Despite its limitations, our study demonstrated that the treatment of femoral neck fractures using cannulated screws combined with a medial buttress plate led to a femoral neck shortening rate that is lower than that observed in other internal xation methods and an internal xation failure rate that is lower than or equivalent to that of other internal xation methods. The quality of cannulated screws is a risk factor for postoperative shortening of the femoral neck, and femoral neck shortening is a risk factor for postoperative internal xation failure. Availability of data and materials The datasets supporting the conclusions of this article are included within the article. Raw data can be requested from the corresponding author.

Competing interests
The authors declare that they have no competing interests.