Impact of fracture patterns and patient factors on distal interlocking screw backout after retrograde femoral nailing

Abstract

BACKGROUND

Retrograde femoral nailing is a common procedure done to treat native or periprosthetic distal femur fractures. Among intramedullary devices, the DePuy Synthes Retrograde Femoral Nail-Advanced Retrograde Femoral Nailing System is one such newer intervention that claims features that improve implant alignment and distal fixation. However, previous studies evaluating interlock screw backout in this implant have demonstrated varying backout rates.

AIM

To evaluate factors associated with distal interlocking screw backout after using the DePuy Synthes RFN-Advanced Retrograde Femoral Nailing System (DePuy Synthes, Raynham, MA, United States). We hypothesize that higher backout rates might be observed in older patients with medical comorbidities that can affect bone quality, such as smoking and diabetes, and in fracture patterns that may put higher strain on the distal fixation.

METHODS

A retrospective case control study was performed on patients undergoing retrograde femoral nailing from 2021 to 2024 at an academic level 1 trauma center through electronic medical record review. The cohort was stratified into those with distal interlocking screw backout and those without. Categorical outcomes were compared using χ² tests and continuous outcomes were compared with nonparametric tests. Significance was determined as P < 0.1, and variables that differed between groups at this level were included in a subsequent logistic regression model.

RESULTS

A total of 76 patients met inclusion criteria, with 9 patients (11.8%) experiencing distal interlocking screw backout and 67 patients (88.2%) not experiencing backout. The average age of our patients was 41.8 years old. The cohort with distal interlocking screw backout had a statistically significant higher median age compared to the non-backout group, 68 years vs 38 years (P = 0.011). Both anterior-posterior (AP) and lateral (LAT) distance to the fracture were significant at P < 0.1. Distance on the LAT view was 76 mm [interquartile range (IQR): 54-98] for the failure group compared to 120 mm (IQR: 63-180) for the non-failure group (P = 0.052). Distance on the AP view was 63 mm (IQR: 56-80) for the failure group compared to 123 mm (IQR: 61-181) for the non-failure group (P = 0.064).

CONCLUSION

Age and distance to the fracture line from the joint line on AP and LAT radiographs were significantly different between the two cohorts. Further biomechanical testing could be utilized to determine the forces present on the distal interlocking screws with varying fracture patterns.

Key Words: Trauma; Intramedullary nailing; Distal interlocking screw backout; Femoral fracture; Retrograde nailing

Core Tip: The purpose of our study was to evaluate patient and fracture-related factors associated with distal interlocking screw backout after retrograde femoral nailing with the DePuy Synthes Retrograde Femoral Nail-Advanced Retrograde Femoral Nailing System. Older age was significantly associated with higher rates of screw backout, with affected patients averaging 68 years compared to 38 years in the non-backout group. Fracture lines located closer to the joint line on both anterior-posterior and lateral radiographs were associated with increased risk of screw backout. No significant associations were found between screw backout and comorbidities such as smoking, diabetes, or body mass index. These findings suggest that certain fracture patterns and older patients may require careful consideration when using the DePuy Synthes Retrograde Femoral Nail-Advanced Retrograde Femoral Nailing System device.

INTRODUCTION

Femur fractures are an increasingly common traumatic injury with incidence rates reaching 8.7 per 100000 person-years and 6.4 per 100000 person-years within the United States specifically in the past decade[1,2]. Adult femur fractures present in a bimodal distribution, with younger adults suffering from higher energy traumas and geriatric patients sustaining them after lower energy trauma[1-4]. After the age of 60, there is a significant increase in the incidence rate of these injuries[1]. Distal femur fractures in elderly patients are often the result of low energy trauma in osteoporotic bone[2-4]. Distal femur fractures specifically can present unique challenges to treat, as the fracture patterns, combined with frequent comminution and intra-articular involvement, create specific biomechanical problems[3].

Surgical intervention for femur fractures has changed significantly in the past few decades. Intramedullary devices have been the standard for appropriate diaphyseal fractures, while the angled blade plate was the early option for distal femur fractures. However, treatment strategies have evolved with efforts to provide more minimally invasive care, anatomical fixation, and encourage secondary healing[5,6]. The invention of locking screws in the 1990s allowed for a more minimally invasive approach but was marred by nonunion due to high construct rigidity. Treatment options include plates that accommodate both locking and compression screws, intramedullary rods, or a combination of the two[7]. However, treatment of distal femur fractures specifically remains challenging with high rates of nonunion, malunion, and implant failure despite various treatment modalities. Among intramedullary devices, the DePuy Synthes Retrograde Femoral Nail-Advanced Retrograde Femoral Nailing System (RFNA) is one such newer intervention that claims features that improve implant alignment and distal fixation. However, previous studies evaluating interlock screw backout in this implant have demonstrated varying backout rates. Bhale et al[8] and Minhas et al[9] found high interlock backout rates of 24.5% (n = 110) and 30% (n = 27). A similar study done in a younger patient cohort that included femoral shaft and distal femur fractures found a backout rate at 8.9% (n = 101)[10].

While previous studies have looked at RFNA failure rate, they have found varying failure rates without an associated cause. The purpose of our study thus is to investigate patient, radiographic, and injury related factors associated with distal interlocking screw backout after using the RFNA System. We hypothesize that higher backout rates might be observed in older patients with medical comorbidities that can affect bone quality, such as smoking and diabetes, and in fracture patterns that may theoretically put higher strain on the distal fixation.

MATERIALS AND METHODS

Study approval was obtained through protocol review by our Institutional Review Board. A retrospective review of CPT codes was used to identify patients undergoing retrograde intramedullary nail fixation for femoral shaft fractures using the RFNA device. This was verified by reading the operative report. Electronic medical record review was used to collect patient demographic data, perioperative data, and post-operative follow up data, including complications. Radiographic review was undertaken by two independent reviewers to determine fracture characteristics and classifications. Two patients were excluded from the analysis as the RFNA was not the primary mode of fixation for their fracture. One patient was determined to be an outlier and was removed from the sample. Categorical variables recorded included patient sex, medical comorbidities, length of surgical stay, length of hospital stay, Orthopaedic Trauma Association/AO Foundation fracture classification, implant complications, and need for reoperation. Continuous variables recorded included patient age, patient body mass index (BMI), and distance from the joint line to the fracture line on the anterior-posterior (AP) and lateral (LAT) radiographic views. Screw backout was defined as any of the distal interlocking screw heads migrating any distance away from the implant on subsequent post-operative radiographs.

Statistical analysis

Univariate analysis was performed for variables kept for analysis, comparing the groups that had screw failure and that did not. Wilcoxon-rank-sum-tests for continuous variables were used to test differences in continuous variables (age, distance to joint line, etc.), and Fisher’s exact test was used to test the difference in binary variables (diabetes, smoking status, etc.). We treated this analysis as a first stage variable selection for creating a model. Significance of the variables was determined as P < 0.1, and variables that were different between groups at this level were included in a subsequent logistic regression model. Outcomes considered were both distances from the joint line, diabetes and smoking status. Covariates for the regression models were chosen using a variable selection technique to maintain model stability, and the outcomes were analyzed individually. All analysis was done using R version 4.4.0 – Puppy Cup.

RESULTS

A total of 76 patients met inclusion criteria, with 9 patients (11.8%) experiencing distal interlocking screw backout and 67 patients (88.2%) not experiencing backout. Gender within our study population was nearly even, with 37 male patients (48.7%) and 39 female patients (51.3%) (Table 1). The average age of our patients was 41.8 years old. Within both the non-failure and failure group, a majority of injuries affected the left leg (57% and 67%, respectively) (Table 2). Of our 9 patients with screw backout, 7 required revision surgery for hardware removal. Of these 7 hardware removal revision surgeries, one of these patients additionally had a fracture nonunion complication, and another additionally had a wound healing complication that was exacerbated by the backed-out screw (Figure 1).

Figure 1 Radiographic images of screw backout and nonunion. A: Lateral view radiograph of injury; B: Postoperative anterior-posterior (AP) view radiograph; C: Screw backout captured on AP view radiograph at 11 months postoperative visit prior to revision surgery; D: Postoperative AP view radiograph following revision surgery for hardware removal; E: Postoperative AP view 6 months after revision surgery; F: Postoperative lateral view 6 months after revision surgery.

Table 1 Patient demographics, n (%)/mean (interquartile range).

Variable	No backout (n = 67)	Backout (n = 9)	P value
Diabetes			0.12
No	59 (88)	6 (67)
Yes	8 (12)	3 (33)
Smoker			0.6
No	56 (84)	7 (78)
Yes	11 (16)	2 (22)
Age (years)	38 (26-61)	68 (62-70)	0.011
Body mass index	27 (24-36)	28 (26-34)	0.4
Gender			> 0.9
Female	34 (51)	5 (56)
Male	33 (49)	4 (44)

Table 2 Distal interlocking screw backout rate, n (%)/mean (interquartile range).

Variable	No backout (n = 67)	Backout (n = 9)	P value
Fracture distance from joint line (mm)
Lateral measurement	120 (63-180)	76 (54-98)	0.052
Anterior-posterior measurement	123 (61-181)	63 (56-80)	0.064
Laterality			0.7
Left	38 (57)	6 (67)
Right	29 (43)	3 (33)

Screw backout was affected by both fracture characteristics and patient comorbidities. There was a significant difference in the distance from fracture site between the failure and non-failure groups in the univariate analysis. This was observed on both the LAT and AP views. The screw failure group had a median LAT distance of 76 mm [interquartile range (IQR): 54-98] while the non-failure group had a median LAT distance of 120 mm (IQR: 63-180) (P = 0.052). For the AP view, median distance within the failure group was 63 mm (IQR: 56-80) and the non-failure group had a median AP distance of 123 mm (IQR: 61-181) (P = 0.064; Table 2). Patients in the screw failure group also had a significantly higher median age compared to the non-failure group. The average patient with screw backout was 68 years old compared to an average age of 38 in those with no backout (P = 0.011; Table 1). The association between AO Classification and screw backout was assessed using Fisher’s exact test. Observed counts and expected counts were calculated. No significant association was found in this analysis between classification and screw backout (P = 0.273).

Comorbidities including smoking status, diabetes diagnosis, and BMI were also investigated for association with screw failure. These were chosen due to their potential impact on bone quality. However, no statistically significant differences were noted between these groups. Out of 76 patients, 13 were smokers. The proportion of smokers within the non-failure and failure groups was nearly equivalent at 16% and 22% respectively (P = 0.6). Patients with diabetes were found in higher proportion within the failure group (33%) compared to the non-failure group (12%), however the difference was not significant (P = 0.12). The average BMI was similarly equal, with the non-failure group having an average of 27 and the failure group having an average of 28 (P = 0.4; Table 1).

DISCUSSION

In this retrospective case control study, we aimed to investigate patient, radiographic, and injury-related factors associated with distal interlocking screw backout after retrograde femoral nailing with the RFNA System. The study particularly focused on evaluating the role of fracture line in relation to the joint line, and patient characteristics such as age, BMI, and comorbidities to identify any factors which can lead to increased risk of screw backout. We found that patients experiencing screw backout were significantly older than those without backout, with a median age of 68 years vs 38 years (P = 0.011). Additionally, the distance from the fracture to the joint line on both AP and LAT radiographs was shorter in the screw failure group, suggesting that patients with more distal fracture lines may be more prone to backout (P < 0.1). No significant differences were observed between groups regarding gender, smoking status, diabetes, BMI, or surgery time. Logistic regression did not identify any independent predictors of screw backout, likely due to the limited sample size.

The phenomenon of patients requiring distal interlocking screw removal has been previously documented prior to the use of the Synthes RFNA device, however this was usually due to symptomatic hardware, not hardware migration or failure. Shah et al[11] retrospectively reviewed 31 patients who received retrograde intramedullary nailing for distal femur fractures, demonstrating that while all patients achieved successful union, 6 of the 31 required symptomatic distal interlock screw removal. Bhale et al[8], Minhas et al[9], and Alcaide et al[10], reported relatively high interlocking screw backout rates unique to the Synthes RFNA system, with backout rates reported between 13% to 30%. Notably, the younger cohort study cited, which included femoral shaft and distal fractures, showed a lower backout rate (8.9%), aligning with this study’s finding that age may play a significant role in construct stability[1-4]. The observed association between fracture line location and screw backout has not been heavily emphasized in prior literature, suggesting a potential technical consideration for implant usage that warrants further study.

Distal femur fractures, classically defined as the distal 8-15 cm of the femur, have historically been at higher risk for implant failure, nonunion, or mal-union compared to diaphyseal femur fractures, often due to mechanical failure[12]. We hypothesized that fractures in this area of the femur may see higher implant failure with the use of the RFNA device, and this was borne out within our analysis. In our dataset, patients with distal interlocking screw backout were more likely to have fractures closer to the supracondylar area (63 mm from the joint line vs 123 mm, P < 0.1). We suggest that this is likely related to differing stresses on the implant when the fracture is closer to the diaphyseal-metaphyseal junction, as opposed to within the diaphysis. Ideal construct stiffness for supracondylar femur fractures is not yet known and is likely affected by a complex interplay of bone quality, fracture characteristics, and native biology. Future study is needed to further elucidate this relationship and how it affects implant stability.

Medical comorbidities can compromise bone quality and increase complication and failure rate of orthopedic implants. Bone healing after trauma is inhibited on a cellular level in patients with diabetes[13]. This impact on healing potential therefore places implants at a higher risk in these patients. Diabetes, hypertension, and obesity have been identified as significant predictors of oligotrophic nonunion after intramedullary nailing following femoral shaft fractures[14]. Diabetes has also specifically been identified as a risk factor for aseptic loosening following total knee arthroplasty[15]. Cigarette smoking similarly is a known risk factor for delayed skeletal healing and complications[16]. It is associated with poor bone density, wound healing, collagen synthesis, and osteoblast production due to its toxic by-products. It also is associated with peri-implant bone loss, implant failure, and delayed union[16,17]. Multiple studies have described delayed union following fracture repair in smokers, with smoking found associated with up to double the rate of nonunion in certain fractures[16]. While we did not find an association between modifiable risk factors or comorbid conditions and distal interlock screw backout in our cohort, this may have been due to our cohort size. Larger cohort sizes may reveal statistically significant risk for comorbid factors that are trending toward significance in our study, such as a diabetes diagnosis (P = 0.12). Comorbidity burden and patient demographics should be considered during surgical planning. Modifiable risk factors such as BMI and smoking status should be optimized pre-operatively and implant usage considered carefully in patients with significant unmodifiable comorbidities.

This study was not without limitations. First, the sample size of 76 was relatively small, limiting the power of logistic regression analysis to detect independent predictors of screw failure. Second, the retrospective nature of the study introduces inherent biases, including variability in surgical technique and post-operative protocols. Follow-up imaging and clinical assessment also varied among patients, possibly underestimating the true complication rate. Third, while radiographic measurements were performed by two independent physicians, some degree of measurement error is possible. Finally, patient activity level and bone quality – both of which likely influence screw stability – were not fully captured or controlled for in this analysis, potentially skewing the data for older patients[18,19]. Given that bone frailty and osteoporosis are more prevalent in this population, it will be important to account for these factors in future studies.

CONCLUSION

This study suggests that older age and more distal fracture sites may increase the risk of distal interlocking screw backout following retrograde femoral nailing with the RFN-Advanced system. Surgeons should consider usage of RFNA device carefully, particularly in very distal fractures and in older patients. Future studies with larger cohorts and biomechanical analyses are needed to confirm these findings and better define technical modifications, as well as to identify forces present on distal interlocking screws within varying fracture patterns.