Revision rate of displaced femoral neck fractures in non-elderly adults: a retrospective comparative study of the femoral neck system and cannulated screws

Abstract

Background

Displaced femoral neck fractures are associated with a high revision rate. The new femoral neck system(FNS) offers advantages in fixation stability, potentially reducing the need for revision. The purpose of this study was to compare the revision rate of patients with different reduction quality treated with the FNS and cannulated screws (CS).

Methods

This retrospective study included patients with Garden III or IV femoral neck fractures who underwent osteosynthesis in a level 1 trauma center between July 2019 and June 2023. A total of 141 cases met the inclusion criteria and received treatment with either the FNS (N = 65) or CS (N = 76). The quality of fracture reduction, surgical complications (such as femoral neck shortening, avascular necrosis of femoral head, nonunion of fracture, implant failure and withdrawal), revision surgery and the reasons for revision were analyzed.

Results

The mean age of the 141 cases was 52.0 years (range 18–65); with sixty-five cases being male (46.1%). Eighty-four fractures (59.6%) were classified as Garden type III. Reduction quality was good in 71 cases (50.4%) and fair in 70 cases. The mean follow-up period was 25.9 months (range 12–46). A total of 26 cases(18.4%) underwent revision surgery. The revision rate in cases with good reduction was 11.3% (8/71 cases), with seven cases (four hardware removal and three arthroplasty) in the CS group and one case (arthroplasty for fracture nonunion and implant failure) in the FNS group, a significant difference was found between the two groups(P = 0.041). Among the 18 cases (25.7%, 18/70) with fair reduction who underwent revision surgery, nine cases (six hardware removal and three arthroplasty) in the CS group, and nine cases (arthroplasty for implant failure and cut-out) in the FNS group, and there was no significant difference between the two groups (P = 0.672). The total revision rate between the FNS group (15.4%, 10/65) and the CS group (21.1%, 16/76) was not significantly different (P = 0.387).

Conclusions

The total revision rate between the FNS and CS group showed no difference. However, in cases with good reduction, the revision rate was lower in the FNS group compared to the CS group.

Introduction

Displaced femoral neck fractures are uncommon among non-elderly adults, with an incidence of 4–72 per 100,000 population per year in patients aged 65 years or younger [1, 2]. Considering the population size of 170 million in the Yangtze River Delta region, it is estimated that over 20,000 non-elderly displaced femoral neck fractures occur annually in the region. Femoral neck fractures, especially occurring in young patients, are typically managed with osteosynthesis. A procedure of closed reduction and fixation using percutaneous cannulated screws(CS) remains the benchmark in China and many countries worldwide [3, 4]. Evidence suggests that the quality of reduction is the single most important surgeon-mediated factor affecting clinical outcomes for displaced femoral neck fractures in non-elderly adults [5, 6]. Complications including implant failure, nonunion, malunion and avascular necrosis could occur at rates ranging from 7 to 27%, necessitating revision surgery at a high frequency, between 21.8% and 36.4% [6,7,8].

The femoral neck system (FNS), an innovative minimally invasive internal fixation device for femoral neck fractures, has demonstrated greater stability than CS in biomechanical studies [9]. Regardless of the type of implant employed - whether it’s CS, dynamic hip screws (DHS), or FNS - a particular percentage of femoral neck fracture cases will necessitate revision surgery [3, 7, 10]. Understanding the reasons behind these revisions can assist in minimizing the probability of future surgeries. However, to date, there is a scarcity of articles investigating the rates and causes of these revision surgeries. Some studies have found that FNS accelerates fracture healing, reduces surgical duration and lowers postoperative complications [11,12,13]. Nevertheless, other studies have found that the treatment with FNS did not show significant differences in these aspects compared with CS [14, 15]. Moreover, comparative studies that evaluate the revision rates between FNS and CS fixation methods are few. The aim of this study was to compare revision rates among patients of varying reduction quality treated with either the FNS or CS.

Patients and methods

Institutional and prior ethical committee approval were obtained for the use of data in this study. Medical records of all patients over 18 years with isolated femoral neck fractures treated at this level I trauma center between July 2019 and June 2023 were retrospectively reviewed. Exclusion criteria were arthroplasty treatment, open reduction treatment, non-displaced Garden type I or II fractures, and patients with pathologic fractures. Patients treated without FNS or CS, and patients with less than one-year follow-up or deficient radiographic information were also excluded.

The quality of fracture reduction was measured by the consensus of three observers(ZQ, WF, ZK) using X-rays and/or computed tomography (CT) scan performed on the first postoperative day. Modified from Lindquist and Tornkvist’s Criteria [16], the quality of reduction was considered good if displacement was ≤ 2 mm, if the AP Garden angle was 160–175°, and if the posterior or anterior angulation was ≤ 10°. The quality of reduction was considered fair if displacement was ≤ 5 mm, if the AP Garden angle was 160–175°, and if the posterior or anterior angulation was ≤ 20°. For fractures with either a displacement of > 5 mm, an AP Garden angle outside the interval of 160–175°, or a posterior or anterior angulation of > 20°, the quality of reduction was considered poor.

Patient charts were reviewed to collect demographics data, time interval from injury to internal fixation, implant device, and to detect surgical complications such as implant-related infection and implant failure. Avascular necrosis was considered an outcome event if femoral head segmental collapse or subchondral radiolucency was present. Radiographic nonunion was declared in the presence of a persistent fracture line at six months with no interval progression toward radiographic healing or with evidence of instrumentation cutout indicating failure of the fracture to unite. Shortening of femoral neck were judged according to the criteria described by Zlowodzki et al. [17]. Full weight-bearing time was categorized based on whether it was achieved before or after twelve weeks post-operation. Revision surgery and the reason were recorded in the follow-up, including clinic and telephone interviews.

SPSS 25.0 was used for statistical analysis. Continuous data are expressed as mean ± SD. Categorical data are summarized by ratios and percentages. The quality of reduction and the rate of revision surgery were compared using the χ² test. A p-value of less than 0.05 was considered statistically significant.

Table 1 Patient demographic and revision data by methods of fixation

Results

Five hundred and fifteen medical records of femoral neck fracture patients were reviewed. Two hundred and fifty-one cases were treated with arthroplasty, sixteen cases were not isolated femoral neck fracture, fifty-five cases were nondisplaced fractures, and fifteen cases were treated with other osteosynthesis methods or open reduction. A total of 178 patients were enrolled in the study, and one-year follow-up was available for 141 participants (79%).

The mean age of these 141 cases was 52.0 years (range 18–65); with sixty-five cases being male (46.1%). The mean body mass index (BMI) of 23.8 ± 2.9 kg/m². Eighty-four fractures (59.6%) were classified as Garden type III. The median time interval from injury to surgery was 2.2 days (range 1 to 4 days). Reduction quality was good in 71 cases (50.4%), and fair in 70 cases. The CS group included 76 cases (54.0%), and FNS was used in 65 cases. There were no significant differences between the groups in terms of age, sex, BMI, Garden classification, time interval, or reduction quality (Table 1).

The mean follow-up was 25.9 months (range 12–46). At the last follow-up, 26 cases (18.4%) had undergone revision surgery. This included ten cases of hardware removal for femoral neck shortening and implant irritation, ten cases of arthroplasty for implant failure or cut-out, and six cases of arthroplasty for avascular necrosis and intolerable pain. In the case of revision surgery, smoking was found in six cases, while twenty smoking cases were identified in the non-revision group, with no significant difference in rates (P = 0.363). The incidence of alcohol consumption was eight and thirty-six in revision and non-revision surgery cases, respectively, also showing no significant difference (P = 0.958). Full weight-bearing before 12 weeks post initial operation was seen in four revision surgery cases and twelve non-revision cases, with no significant difference in the rate (4/26 vs. 12/115) (P = 0.472). Another eleven patients(four cases in the CS group and seven in the FNS group) had radiographic evidence of avascular necrosis with tolerable pain, and no revision surgery has been performed yet.

The revision rate in good reduction cases was 11.3% (8/71 cases), with seven cases (four hardware removal and three arthroplasty) in the CS group and one case (arthroplasty for fracture nonunion and implant failure) in the FNS group (Fig. 1).

Fig. 1

Preoperative X-ray (Fig. 1-A), as well as coronal and axial CT cuts (Fig. 1-B), demonstrated a displaced femoral neck fracture in a 64-year-old male. One day after the operation, anteroposterior and lateral X-ray (Fig. 1-C) and CT cuts (Fig. 1-D) showed a good closed reduction. CT scans (Fig. 1-E, F) taken at two mouths and X-ray images (Fig. 1-G) at six mouths after the operation revealed an impaction of the femoral neck. X-ray (Fig. 1-H) and CT scans (Fig. 1-I) at twelve mouths after the operation indicateed nonunion of the fracture and implant cut-out. X-ray image (Fig. 1-J) showed a revision surgery of arthoplasty

A significant difference was found between the CS (18.4%, 7/38) and the FNS (3.0%, 1/33) group (P = 0.041). Among the seventy cases of fair reduction, eighteen cases(25.7%) had revision surgery, with the revision rate being significantly higher than that of the good reduction cases (P = 0.027). Nine cases in the CS group (six hardware removal and three arthroplasty) and nine cases (arthroplasty for implant failure and cut-out) in the FNS group. The rate between the the CS (23.7%, 9/38) and the FNS (28.1%, 9/32) group was not different (P = 0.672). The total revision rate between the FNS group(15.4%, 10/65) and the CS group(21.1%, 16/76) was not significantly different(P = 0.387).

Discussion

Although displaced femoral neck fractures are uncommon in non-elderly adults, the size of the population living in the Yangtze River Delta means that surgeons still face a large number of cases at our trauma center. According to current guidelines and surveys of members of the Orthopaedic Trauma Association and European-AO International affiliated trauma centers, osteosynthesis is the appropriate treatment for patients under 65 years [4, 18]. The simple procedure of closed reduction and percutaneous screw fixation remains the first choice, and a novel FNS has also been recommended by many authors [3]. Regardless of the implant device chosen, the importance of adequate reduction of femoral neck fracture has been emphasized in previous reports [6, 16, 19]. The quality of reduction in displaced femoral neck fractures is known to influence patient outcomes and postoperative complications, and studies have demonstrated that worse fracture reduction is associated with poorer outcomes [20]. The result of this study also supported this, as the revision rate of good reduction cases was significantly lower than that of the fair reduction group.

Many authors have reported the clinical outcomes and rate of revision surgery of different implant device. In Griffin’s study, the revision rate of CS was 36.4% for patients aged 65 years and over with either displaced or undisplaced fractures [21]. Siavashi et al. focused on non-elderly adults and displaced fractures with a minimum of one year follow-up, and found the revision rate of CS was 28.6%(8/28) [8]. Ponkilainen et al. reported an overall screw fixation removal rate of 18.5%, without considering arthroplasty [22]. In our study, the revision rate CS was 21.1%, slightly lower than their results. The revision rate of dynamic hip screw, which was not included in this study, was reported from 6.7 to 13.2% [8, 22, 23]. Stegelmann et al. reported the cumulative incidence of FNS failure at one year was 13% [24], while Davidson et al. reported a revision rate of 8.8% at a mean follow-up duration of 6.9 months, and in this same study, they performed a systematic review and found a revision rate of 9.2% when combining their population with 278 patients from 10 outside studies [25]. The revision rate of FNS in our study was 15.4%, higher than their results, but with a longer follow-up of 25.9 months than them.

According to the literature, reasons for revision surgery included implant fixation failure, screw loosening or migration, implant irritation, avascular necrosis of the femoral head, nonunion, and infection [8, 21]. In this study, the top three causes for revision surgery were implant irritation, implant cut-out, and avascular necrosis.

Implant irritation was only presented in the CS group in this study. During the femoral neck healing process, the shortening of the femoral neck and withdrawal of CS happened simultaneously. In contrast, FNS cases were also associated with some degree of radiologic femoral neck shortening during follow-up, however, as the dynamic design of the integrated bolt and anti-rotation screw allows for up to 20 mm of guided collapse, FNS has an advantage in resisting irritation, which screws do not have. Shortening of the femoral neck due to fracture impaction may result in the loss of the abductors’ moment of force on the greater trochanter, resulting in weakness, pain, and patient dissatisfaction [26]. Given that the patients with shortening and irritation in this study underwent internal fixation removal, a relatively simple surgery, we believe that the number of reoperations was increased in the CS group, but the hip function was not necessarily worse than that in the FNS group (Fig. 2).

Fig. 2

Preoperative X-ray images (Fig. 2-A) demonstrated a displaced femoral neck fracture in a 61-year-old male. One day after the operation, X-ray images (Fig. 2-B) showed a fair closed reduction. X-ray images at seven months (Fig. 2-C) and twenty months (Fig. 2-D) after the operation revealed a femoral neck shortening and implant withdrawal. X-ray image (Fig. 2-E) showed a revision surgery of implant removal

In this study, cut-out was the second cause of revision surgery, found in the FNS group and fair reduction patients, which is consistent with prior studies [10, 25]. As the biomechanical advantages of FNS have been previously reported, FNS presents better overall structural stability than three cannulated screws with angular stability and rotational stability [9], it is inexplicable why cut-out happened in the FNS group instead of the CS group in fair reduction patients. In follow-up imaging data, the femoral shaft end of the fracture was firmly fixed with the FNS, but the femoral head end of the fracture was separated from the FNS system. The reasons for this may include complete weight bearing before the fracture union or insufficient holding force on the femoral head end of the fracture.

In our study, patients in both groups had femoral head avascular necrosis. Six patients received arthoplasty for intolerable pain, while another eleven cases have avascular necrosis in radiogrghy but had not received surgery yet. As Siavashi reported, AVN and its rate are separate phenomena compared with reduction and fixation, it is mainly related to the force of impact, primary displacement, and the time between fracture and reduction/fixation [8].

There are limitations to this study. Firstly, it is a single-center retrospective study with a relatively small sample size and insufficiently long follow-up period. Moreover, many effective internal fixation methods for femoral neck fractures, such as dynamic hip screw and dynamic helical hip system, were not included in this study [27]. Secondly, our findings are vulnerable to selection bias due to loss of 21% of eligible subjects to clinical follow-up within one year, although this rate of attrition represents reasonably achievable follow-up for a retrospective study [28]. Thirdly, the weight-bearing protocol was the same for both groups, however, we did not collect precise weight loading times and used twelve weeks as a benchmark for premature loading. Although this is significant, it is not entirely accurate, and premature loading could be a considerable factor contributing to implant failure. Furthermore, although one year of prospective follow-up should be long enough to identify most fracture-related complications, avascular necrosis has been reported to develop years after injury, and the subsequent femoral head collapse may also occur in a further delayed time frame.

In conclusion, with a mean follow-up period of 25.9 months, the total revision rate of the FNS and CS group showed no difference. However, in the good reduction cases, the revision rate in the FNS group was lower than that in the CS group.