Mechanical effects of valgus osteotomy angle and re-fixation strategies on re-fixation of the femoral neck fracture: a finite element analysis

Abstract

Aims

This study examined the biomechanical outcomes of osteotomy angles and the internal fixation devices for re-fixation of the femoral neck fracture.

Methods

The finite element model of femoral neck fracture was initially fixed with three cannulated screws, followed by 10° and 20° valgus osteotomy. The models were re-fixed using sliding hip screw (DHS) + anti-rotation screw (AS) or DHS + medial buttress plate (MBP). Physiological loads were applied to all models.

Result

With 10° and 20° valgus osteotomy, the interfragmentary shear stress, sliding distance, and gap decreased, the compressive stress increased. With a further 10° osteotomy angle, the interfragmentary sliding distance in DHS + AS group decreased by 38.4%, and the interfragmentary compressive stress in DHS + MBP group increased by 20.9%. The interfragmentary shear stress and gap showed changes of no more than 10%. For the sliding distance, gap, and shear stress, DHS + AS is slightly superior to DHS + MBP at both osteotomy angle. The Von Mises stress of the internal fixation device for DHS + AS is 22.61% higher than that for DHS + MBP at a 20° osteotomy angle.

Conclusion

Both 10° and 20° osteotomy angles enhance the mechanical strength and stability at the fracture site, which may offer potential options for the clinical selection of valgus osteotomy angles. DHS + AS exhibits better mechanical stability but may carry a higher risk of internal fixation failure. Patients treated with DHS + AS may require more cautious postoperative weight-bearing; however, this should be validated in clinical studies. Meanwhile, with the development of bone tissue engineering, the material strength of internal fixation devices may improve, reducing the risk of internal fixation device failure.

Introduction

High-energy femoral neck fractures (FNF) in young patients present a clinically challenging problem. Operations for these fractures are fraught with difficulties, including substantial fracture displacement, compromised blood supply to the femoral head, and post-reduction malalignment at the fracture site [1]. Consequently, there is a high incidence of postoperative complications with rates as high as 45% [2]. Among these complications, nonunion remains a major cause of revision surgery, with an incidence of approximately 10%–20% [1]. Given the young age and high functional demands of this patient population, hip-preserving revision strategies are generally preferred whenever feasible. According to Pauwels' theory, fractures with a larger Pauwels' angle have a fracture plane closer to being vertical, resulting in greater shear stress and smaller compressive stress on the fracture surface, increasing the risk of nonunion [3]. This unfavorable biomechanical environment is particularly relevant in Pauwels type III fractures and represents a key mechanical target during revision surgery. Valgus osteotomy has been proposed as an effective biomechanical strategy to address this problem by reorienting the fracture line toward a more horizontal direction. By converting shear forces into compressive forces, valgus osteotomy can promote callus formation, improve fracture stability, and facilitate fracture healing [3–6]. Previous relevant studies have verified that osteotomy with re-fixation demonstrates definitive therapeutic efficacy for postoperative complications arising from femoral neck fracture internal fixation procedures [7, 8]. Meanwhile, shortage of bone grafts and compromising blood supply of the femoral head during clinical practice make it more difficult to treat young patients. Therefore, valgus osteotomy with re-fixation has emerged as an effective treatment option for this population [9]. The main principle of valgus osteotomy is to reduce the Pauwels angle to below 20–25 degrees [5]. Tiny osteotomy angles may fail to effectively reduce the shear stress at the fracture site. However, an excessively large osteotomy angle may lead to risks such as postoperative limb length discrepancy, gait deformity, and nonunion at the osteotomy site [10]. Notably, valgisation exceeding the contralateral side by 15° has been associated with kinking of posterior retinacular vessels and increased joint reaction forces, potentially increasing the risk of ONFH [11]. Therefore, the optimal osteotomy angle that balances mechanical benefit and biological safety remains unclear, and whether smaller osteotomy angles, such as 10° or 20°, can provide sufficient biomechanical improvement warrants further investigation. In addition to the osteotomy angle, the choice of internal fixation following valgus osteotomy represents another critical issue in revision surgery. Currently, clinicians typically use DHS alone for the re-fixation of the femoral neck after valgus osteotomy [12, 13]. DHS can cause greater trauma to patients and lacks sufficient anti-rotation capability, despite it being modular and easy to use. It is preferred to use it in combination with an anti-rotation screw (AS) or a medial buttress plate (MBP) for stabilization. Nevertheless, clinicians have not reached a consensus on the appropriate internal fixation during revision surgery. Studies on the angle of valgus osteotomy and the fixation strategies for re-fixation of the femoral neck fracture are scarce, especially with finite element analysis.

The present study focuses on the revision re-fixation scenario for nonunion following failed primary internal fixation and investigates the biomechanical effects of valgus osteotomy and subsequent re-fixation strategies. This study aimed to test the following hypotheses: Whether smaller osteotomy angles (10° and 20°) can improve stability at the fracture fragments to promote fracture healing and whether good fixation effects can be achieved after valgus osteotomy with DHS + AS or MBP.

Methods

In this study, the need to obtain knowledgeable agreement was not necessary, as CT scans were part of the standard treatment protocol and utilizing these data carried negligible harm to patients. Additionally, our research program was authorized by our Institutional Review Board (IRB). (IRB number: K2023-666).

3D modelling of the femur

This analysis utilized CT scan images from a medical examination of a 36-year-old healthy male patient, which were originally intended to assess the health of his femurs. This patient was selected because the neck-shaft angle of his bilateral femurs, as measured by 3D CT examination, was 125°, which is consistent with the anatomical structure of the femur in the Chinese population [14]. The anthropometric information for this patient comprised a height of 175 cm and a weight of 66.3 kg. In this study, the imagine details were loaded into MIMICS software (Materialize Company, Leuven, Belgium) to reconstruct a 3D model of the unfractured left femur.

Fracture models

The model of femoral neck fracture was established by performing a simulated osteotomy according to the Pauwels fracture classification [15]. The fracture plane was designed to form a 52° angle with the horizontal line within the coronal plane, which is established based on the alignment of the femoral neck and shaft, and this fracture surface exhibited a 16° tilt in relation to the coronal plane.

Implant model

In SolidWorks 2017, we established 3 groups of internal fixation device models with reference to the data of Forwos Medical's DHS, MBP, AS, and cannulated screws (CS). Original cannulated screw fixation group(3CSs): screw diameter of 6.5 mm, thread length of 18 mm, and each with an overall length of 90 mm [16]. DHS + AS Group: The dynamic rod had a diameter of 8 mm. For the 10° and 20° osteotomy groups, the barrel angles were 124° and 130° respectively. Following established AO fixation principles, we selected the DHS barrel angle to match the neck–shaft angle after the osteotomy. The lag screw was planned in a central-to-inferior position close to the calcar, with the side plate seated on the lateral cortex. Accordingly, 124° and 130° barrel angles were used for the 10° and 20° osteotomy models, respectively. This selection is consistent with the 120°–135° fixed-angle DHS range reported in valgus osteotomy fixation [5, 13]. The terminal portion of the locking plate was attached to the femoral shaft with 4 locking screws. The AS had a diameter of 6.5 mm and features partial threading [17]. DHS + MBP Group: The DHS data was the same as that of the DHS + AS group. MBP: A 1/3 tubular steel plate was used, with a width of 10 mm and a thickness of 2.7 mm. Additionally, in the DHS system, the angle between the lag screw and the locking plate was appropriately adjusted to fit anatomical structure after valgus osteotomy [18]. The materials used for the internal fixation devices were all titanium alloy.

Establishment of valgus osteotomy and revision internal fixation models

Original cannulated screw fixation group(3CSs): The three CSs were implanted in a parallel manner, arranged in an "inverted triangle" configuration. Subsequently, Boolean subtraction was used to simulate bone loss resulting from the drilling operation during the insertion of the hollow screws.

Valgus osteotomy models: Based on the 3CSs model, the three CSs were removed. For the osteotomy, an upper cut was made at the lesser trochanter level, and the lower cut was made downwards at the same level. The angle formed by the convergence of these two cuts on the internal side of the femur represented the osteotomy angle. The wedge-shaped bone fragment cut out by these two cuts was removed, and simultaneously, the femoral head is externally rotated to make two cuts meet [6]. Created models with osteotomy angles of 10° and 20°, respectively.

The DHS + AS revision model(DHS + AS): Based on the valgus osteotomy model, the DHS was placed at the center of the femoral neck, ensuring that the Tip-Apex Distance (TAD) was less than 10 mm, while also guaranteeing that the lag screw was positioned more than 10 mm above the point of the osteotomy cut. Subsequently, the locking plate was attached to the femoral shaft [19]. Simultaneously, an anti-rotation screw was placed 15 mm above the DHS dynamic hip screw, ensuring that the TAD was less than 10 mm away from the cortical bone of the femoral head [20]. To obtain two models: one with a 10° osteotomy angle using DHS + AS (10°DHS + AS), and a 20° osteotomy angle using DHS + AS (20°DHS + AS).

The DHS + MBP revision model(DHS + MBP): Based on the valgus osteotomy model, the DHS was inserted using the same method. Simultaneously, a MBP was positioned along the inner aspect of the femoral neck, spanning across both the original fracture line and the valgus osteotomy fracture line. Subsequently, three cortical screws were implanted, respectively, on the medial aspect of the femoral shaft, beneath the femoral head, and on the medial side of the femoral neck, to secure the MBP [21] (Fig. 1). To obtain two models: one with a 10° osteotomy angle using DHS + MBP (10°DHS + MBP), and a 20° osteotomy angle using DHS + MBP (20°DHS + MBP).

Fig. 1.

Following the performance of a valgus osteotomy for a Pauwels III FNF, stabilization was achieved with internal implants. A Fixation using DHS + AS after a 10° valgus osteotomy. B 10° valgus osteotomy. C Fixation using DHS + MBP after a 10° valgus osteotomy. D Fixation using DHS + AS after a 20° valgus osteotomy. E 20° valgus osteotomy. F Fixation using DHS + MBP after a 20° valgus osteotomy

Material property assignment and mesh generation

In the software ANSYS Workbench 19.0, we created a new static analysis project and imported the complete model into the Geometry module. The assignment of material properties was conducted with consideration to prior pertinent research studies [22]. The material properties of the femoral model were considered to be linearly elastic and isotropic, while the implant material was titanium alloy. The material parameters for each component were listed (Table 1). The femoral geometry was spatially differentiated into cortical and cancellous bone using offset and Boolean operations in Mimics, and the corresponding material properties were assigned to each region accordingly. The mesh of the model was automatically generated, with the element size set to 2.0 mm for bone tissue and 0.5 mm for the internal fixation devices. All components were meshed using three-dimensional tetrahedral solid elements. The quantity of nodes and elements within the model were listed (Table 2).

Table 1.

Material properties of Titanium–6% Aluminium–7% Niobium (Ti-6Al-7Nb)

Structure	Modulus of elasticity (GPA)	Poisson’s ratio
Ti-6Al-7Nb	105	0.34
Cortical bone	16.80	0.30
Cancellous bone	0.84	0.20

Table 2.

The information pertaining to elements that are constructed from finite element models

Finite elements model		10°DHS + AS	20°DHS + AS	10°DHS + MBP	20°DHS + MBP	3CSs
Number of nodes	Implants	2,136,189	2,151,649	2,082,414	2,060,508	771,844
	Femur	928,095	918,368	935,396	960,702	959,355
	Summary	3,064,284	3,070,017	3,017,810	3,021,210	1,731,199
Number of elements	Implants	1,487,307	1,496,735	1,439,003	1,432,387	500,530
	Femur	616,294	612,634	624,367	635,044	643,096
	Summary	2,103,601	2,109,369	2,063,370	2,067,431	1,143,626

To ensure that the simulation results were not dependent on mesh density, a mesh convergence analysis was additionally performed. A representative configuration (10° valgus osteotomy with DHS + AS fixation) was selected for this analysis because all four configurations were generated using the same meshing strategy and exhibited highly comparable mesh scales (total node and element counts differed by < 2%). During mesh refinement, key response quantities were monitored, including the 95th percentile (P95) von Mises stress of the femur together with the interfragmentary gap at the fracture surface (Fig. 2) [23]. In the final refinement step, reducing the femoral mesh size from 2.05 mm to 2.00 mm changed the interfragmentary gap from 0.14422 mm to 0.14367 mm (|Δ|= 0.38%), while the femoral P95 von Mises stress changed from 117.67 MPa to 115.65 MPa (|Δ|= 1.72%). Therefore, the selected mesh density (femur: 2.0 mm; implants: 0.5 mm) was considered sufficiently converged and was applied consistently to all models.

Fig. 2.

Convergence study results in FEA

Boundary conditions

The interfaces within the internal fixation assembly were defined as bonded contact. The contacts between the fracture surfaces and between the implant and bone were defined as frictional contact. The fracture interface was assumed to be initially closed with direct bone–bone contact, and no interposed fibrous tissue was modeled; any opening/closure under loading was governed by the contact algorithm and quantified by the interfragmentary gap outcome. The friction coefficients were set to 0.46 for the bone–bone interface and 0.42 for the bone–implant interface, in accordance with previously published femoral neck fracture finite element models [24]. The distal end of the femur is designated as an immovable, fixed interface.

Loading condition

During normal gait, while in single-leg stance, a load of 1989 N was applied to the femoral head, following the methodology employed in prior research [25]. This load represents the resultant hip joint contact force vector and was equivalent to 300% of body weight [25, 26]. Concurrently, the greater trochanter was subjected to an abductor force, other muscle forces such as the iliopsoas and vastus lateralis were not modeled. Considering the anatomical configuration of the acetabulum and hip joint, the body weight load was distributed onto the spherical surface of the femoral head. Specifically, the 1989 N resultant force was applied as a distributed load over the hemispherical articular surface of the femoral head rather than as a point load, to avoid artificial stress concentration. The load is positioned at an angle of 20° with respect to the femoral shaft axis within the coronal plane, and at an angle of 17° within the sagittal plane (Fig. 3) [27].

Fig. 3.

Within the finite element model, the femur is subjected to loading in a stance position with one leg. A A hip joint contact force vector with a magnitude of 1989 N, equivalent to 300% of the body weight for an individual weighing 66.3 kg, was applied as a distributed load over the hemispherical articular surface of the femoral head, forming a 20° angle with the femoral shaft axis in the coronal plane. B The resultant hip joint contact force formed a 17° angle with the femoral neck axis in the horizontal plane of the femur. C An abductor force of 800 N was applied at the greater trochanter

Comparative parameters

The following measurement indicators were used in the finite element analysis: interfragmentary shear stress, interfragmentary compressive stress, interfragmentary sliding distance, interfragmentary gap, the peak value and distribution pattern of Von Mises stress within the internal fixation device. Gap and sliding were calculated as the normal and tangential components of the relative displacement between the opposing fracture surfaces, and the maximum value across the interface was reported.

Statistical analysis

Finite element simulations are deterministic and yield a single solution for each configuration; therefore, inferential statistics (e.g., p values) were not applied. Instead, outcomes were summarized as absolute values and percentage changes relative to predefined reference models (defined in Table 3). Numerical stability was ensured by mesh convergence analysis.

Table 3.

The interfragmentary motion and stress on the fracture surface with the angle of valgus osteotomy and the choice of internal fixation device. Δ% = (X − R)/R × 100%, where R is the reference value. R = 3CS for the 10° rows; for the 20° rows, R includes 3CS and the corresponding 10° construct, as indicated in parentheses

Variable		Types of internal fixation devices
	osteotomy angle	DHS + AS	DHS + MBP	3CS
Maximum interfragmentary shear stress (amount of change (reference))	10°	6.4794 MPa(Δ% = −36.7%(R = 10.2280))	6.6886 MPa(Δ% = −34.6%(R = 10.2280))	10.2280 MPa
	20°	6.9396 MPa(Δ% = −32.2%(R = 10.2280), Δ% = + 7.1%(R = 6.4794))	6.4245 MPa(Δ% = −37.2%(R = 10.2280), Δ% = −4.0%(R = 6.6886))
Maximum interfragmentary compressive stress(amount of change (reference))	10°	14.962 MPa(Δ% = + 78.4%(R = 8.3853))	12.373 MPa(Δ% = + 47.6%(R = 8.3853))	8.3853 MPa
	20°	16.445 MPa(Δ% = + 9.9%(R = 14.962), Δ% = + 96.1%(R = 8.3853))	14.952 MPa(Δ% = + 20.9%(R = 12.373), Δ% = + 78.3%(R = 8.3853))
Maximum interfragmentary sliding distance (amount of change (reference))	10°	0.036475 mm(Δ% = −15.8%(R = 0.043328))	0.030293 mm(Δ% = −30.1%(R = 0.043328))	0.043328 mm
	20°	0.022456 mm(Δ% = −48.1%(R = 0.043328), Δ% = −38.4%(R = 0.036475))	0.028127 mm(Δ% = −35.1%(R = 0.043328), Δ% = −7.1%(R = 0.030293))
Maximum interfragmentary gap (amount of change (reference))	10°	0.14367 mm(Δ% = −11.5%(R = 0.16228))	0.12998 mm(Δ% = −19.9%(R = 0.16228))	0.16228 mm
	20°	0.15118 mm(Δ% = −6.8%(R = 0.16228), Δ% = + 5.2%(R = 0.14367))	0.14972 mm(Δ% = −7.7%(R = 0.16228),Δ% = + 15.2%(R = 0.12998))
The equivalent von-Mises stress of fixation (amount of change (reference))	10°	635.67 MPa(Δ% = −6.1%(R = 676.88))	676.88 MPa
	20°	575.95 MPa(Δ% = + 22.6%(R = 469.71)	469.71 MPa

R reference

Results

With the single-leg stance, the maximum interfragmentary shear stress in the 3CSs group was 10.228 MPa. The 10°DHS + AS group and 10°DHS + MBP group exhibited a 36.7% and 34.6% reduction in maximum interfragmentary shear stress compared to the 3CSs group. The 20°DHS + AS group and 20°DHS + MBP group demonstrated a 32.2% and 37.2% reduction in maximum interfragmentary shear stress compared to the 3CSs. The interfragmentary shear stresses were similar between the 10° DHS + MBP and 20° DHS + MBP configurations. (Figs. 4 and 7; Table 3).

Fig. 4.

Interfragmentary stress distributions at the fracture site. Top row (A1–A5): shear stress; bottom row (B1–B5): compressive stress. Configurations: 10° DHS + AS (A1/B1), 20° DHS + AS (A2/B2), 10° DHS + MBP (A3/B3), 20° DHS + MBP (A4/B4), and 3CSs (A5/B5)

Fig. 7.

Bar-chart comparison of key interfragmentary outcomes across fixation configurations. Shown are the maximum interfragmentary shear stress, maximum interfragmentary compressive stress, maximum interfragmentary sliding distance, maximum interfragmentary gap, and the maximum von Mises stress of the fixation. Results are presented for 10° and 20° valgus osteotomy models and the original 3CSs fixation. Group labels are provided in the legend (upper right)

The maximum interfragmentary compressive stress in the 3CSs group was 8.3853 MPa. The 10°DHS + AS group and 10°DHS + MBP group showed a 78.4% and 47.6% increase in maximum interfragmentary compressive stress compared to the 3CSs group. The 20°DHS + AS group and 20°DHS + MBP group exhibited a 96.1% and 78.3% increase in maximum interfragmentary compressive stress compared to the 3CSs group (Figs. 4 and 7; Table 3).

The maximum interfragmentary sliding distance in the 3CSs group was 0.043328 mm. The 10°DHS + AS and 10°DHS + MBP groups demonstrated a 15.8% and 30.1% reduction in maximum interfragmentary sliding distance compared to the 3CSs group. The 20°DHS + AS group and 20°DHS + MBP group showed a 48.1% and 35.1% reduction in maximum interfragmentary sliding distance compared to the 3CSs group (Figs. 5 and 7; Table 3).

Fig. 5.

Interfragmentary motion at the fracture site. Top row (A1–A5): sliding distance; bottom row (B1–B5): gap. Configurations: 10° DHS + AS (A1/B1), 20° DHS + AS (A2/B2), 10° DHS + MBP (A3/B3), 20° DHS + MBP (A4/B4), and 3CSs (A5/B5)

The maximum interfragmentary gap in the 3CSs group was 0.16228 mm. The 10°DHS + AS group and 10°DHS + MBP group exhibited an 11.5% and 19.9% reduction in maximum interfragmentary gap compared to the 3CSs group. The 20°DHS + AS group and 20°DHS + MBP group demonstrated a 6.8% and 7.7% reduction in maximum interfragmentary gap compared to the 3CSs group (Figs. 5 and 7; Table 3).

At a 10° valgus osteotomy angle, the Von Mises stress in the internal fixators of the DHS + MBP group and DHS + AS group was similar, whereas at a 20° valgus osteotomy angle, the Von Mises stress in the DHS + MBP group was 22.61% lower than that in the DHS + AS group (Figs. 6 and 7; Table 3).

Fig. 6.

Von Mises stress distributions of the internal fixation constructs. Magnified views highlight the regions of peak stress, and arrows indicate the corresponding locations on the overall construct. A 10° DHS + AS; B 20° DHS + AS; C 10° DHS + MBP; D 20° DHS + MBP

Discussion

Despite the expansion of the age criteria for hip replacement surgery, osteotomy with re-fixation remains the preferred option for young patients with initial internal fixation failure of FNF. The valgus osteotomy we know today is based on Pauwels' theory, which states that excessive shear stress at the fracture site is often accompanied by increased transverse motion and gap widening at the fracture interface. To clarify the biomechanical meaning of the interfragmentary motion metrics, we interpret the gap and sliding distance as the normal and tangential components of the relative displacement between the two opposing fracture surfaces. For two corresponding points on the proximal and distal fracture surfaces with displacement vectors and , the interfragmentary relative displacement is defined as:

1

Let n denote the unit normal vector of the local fracture plane. The interfragmentary gap is quantified as the maximum opening along the normal direction:

2

whereas the sliding distance is defined as the magnitude of the tangential relative motion:

3

Consistent with prior femoral neck fracture FE studies, vertical relative motion between fracture surfaces is interpreted as gap distance and parallel motion as sliding distance [28]. Accordingly, by performing a valgus osteotomy to flip the proximal fracture fragment, the unfavorable shear force at the fracture site can be converted into a favorable compressive force, thereby promoting fracture healing [29, 30]. In this finite element analysis, an adduction femoral neck fracture model with a Pauwels angle of 52° was established, which is characterized by high interfragmentary shear stress and a high risk of postoperative nonunion. In both the DHS + AS group and the DHS + MBP group, after performing 10° and 20° osteotomies, the interfragmentary shear stress decreased by more than 30% compared to the 3CSs group, which may promote fracture healing. Based on previous relevant research, regular axial motion at the fracture surface promoted callus formation in some ways, thereby accelerating fracture healing, while asymmetric axial motion and shear motion were detrimental to fracture healing [31]. Another animal experiment evaluating the motion between fracture fragments also demonstrated that pure axial compression motion promoted bone healing more effectively than pure shear motion, while torsional shear motion significantly impaired the process [32]. In both the DHS + AS group and the DHS + MBP group, the change in interfragmentary shear stress did not exceed 10% and the changes of interfragmentary gap didn’t exceed 0.02 mm after increasing the osteotomy angle by 10°. This indicates a plateau effect in interfragmentary shear stress and gap with increasing correction angle. Our findings suggest that a modest closing-wedge valgus correction of 10° already achieves the primary biomechanical objective of valgus osteotomy for femoral neck nonunion, namely shifting the fracture-site loading toward a more compression-favorable pattern by reducing the functional Pauwels angle. This interpretation is in line with established concepts of valgus osteotomy, in which shear forces are converted into compressive forces across the nonunion interface to promote consolidation [6]. When the wedge angle increased from 10° to 20°, the additional reductions in interfragmentary shear stress and gap in our models were limited, suggesting diminishing returns for these endpoints. Importantly, this “smaller-correction-first” strategy is also supported by clinical evidence [33]. Surgical planning should also consider the alignment and limb-length changes associated with valgus correction, and meticulous soft-tissue handling remains important to avoid compromising femoral head perfusion.

Prior research has shown that the presence of excessive shear motion at the site of the fracture has the potential to hinder the process of bone healing [34]. After the valgus osteotomy with re-fixation, the interfragmentary sliding distance decreased compared to the 3CSs model. In the DHS + AS group, the interfragmentary sliding distance was reduced by 0.014 mm after increasing the osteotomy angle by 10°. This phenomenon can be ascribed to the minimal influence of AS on interfragmentary sliding motion, due to the healing and bone resorption processes at the fracture site [35]. Therefore, for the DHS + AS, a larger osteotomy angle may be necessary to more effectively inhibit interfragmentary sliding, enhancing mechanical stability at the fracture surface. In terms of interfragmentary compressive stress, all models exhibited an increase of over 45% compared to the 3CSs group. In the DHS + AS and DHS + MBP groups, a 10° difference in osteotomy angle resulted in an increase of near 2 MPa in compressive stress. Valgus osteotomy promotes fracture healing by reducing unfavorable shear stress at the fracture site and increasing compressive stress [36]. In previous practices, two distinct osteotomies were utilized: the first being the McMurray osteotomy, executed precisely superior to the lesser trochanter, and the second, the Shanz-type osteotomy, carried out specifically inferior to the lesser trochanter [37]. For the osteotomy angle, some authors have suggested that the Pauwels angle should be ensured to be less than 30° after the osteotomy [30]. However, an excessively large osteotomy angle may lead to risks such as implant failure and nonunion of the osteotomy surface after surgery. In a prior clinical study conducted by Hartford et al., eight patients with nonunion of FNF underwent valgus osteotomies. After the surgery, the mean angle of the fracture plane diminished to 41 degrees, and there was a notable enhancement in the average Harris Hip Score, which rose from 24 to 73 points. No complications in the form of ONFH were observed in the patients [38].Similarly, a study by Magu et al. performed 20-degree intertrochanteric wedge osteotomies on 48 patients with old FNF and fixed them with 120-degree double-bladed plates. Ultimately, 46 patients achieved fracture healing, 40 patients regained normal walking gait, and the average Harris Hip Score was 86.7 points [39]. This is in line with our study results, which show that smaller osteotomy angles (10° and 20°) can improve stability at the fracture fragments to promote fracture healing. Therefore, before performing a valgus osteotomy, there are more options for calculating the osteotomy angle, to achieve an optimal mechanical effect at the fracture site while maximizing the preservation of the patient's bone mass. Beyond the correction angle, the choice of fixation construct is equally critical because it determines how interfragmentary stability is achieved and at what implant-stress cost.

In both the 10° and 20° osteotomy groups, when analyzing the interfragmentary compressive stress, the DHS + AS group exhibited higher compressive stress than the DHS + MBP group. When analyzing the interfragmentary sliding distance, the DHS + AS group also showed lower sliding distances compared to the DHS + MBP group. According to the findings of a previous study on cadaveric samples and finite element analysis, the DHS + AS internal fixation group exhibited the minimal amount of femoral head displacement and fragmentary movement under static loading, providing a stable fixation structure [40]. The results of a similar biomechanical test revealed that, for unstable FNF, the DHS + AS group demonstrated a higher failure load and superior biomechanical stability in comparison to cannulated screw fixation [41]. For the maximum Von Mises stress, the two internal fixation devices show comparable Von Mises stress in the 10° osteotomy group, whereas in the 20° osteotomy group, the DHS + MBP group showed lower Von Mises stress compared to the DHS + AS group. In both internal fixation groups, stress primarily concentrated on the DHS's locking plate and cortical screws. In Fig. 6, the peak von Mises stress in the DHS + AS constructs was located at the barrel–side plate junction, whereas in the DHS + MBP constructs the stress hotspot shifted to the medial buttress plate adjacent to the osteotomy site along the medial cortex. This spatial pattern is clinically relevant because stress concentration at the barrel–side plate junction has been reported as a critical region that may predispose sliding hip screw constructs to fatigue-related failure under repetitive loading [42]. A medial buttress plate can share load with the DHS and redistribute stress within the fixation system, which helps explain the lower peak implant stress observed in the combined constructs [43]. And the MBP exhibited greater efficacy in stress dispersion than the AS. This result aligned with the finite element analysis findings reported by Huang et al., indicating that, under a load of 2100N, the DHS + AS group experienced peak stress levels surpassing other internal fixation devices [44]. Additionally, Zeng et al.'s study showed that incorporating MBP significantly reduced stress concentration on the implants by providing an alternative pathway for load distribution across fracture fragments, lowering the risk of internal fixation failure [45].However, the incorporation of MBP carries the risk of augmenting local tissue injury and compromising the blood flow to the femoral head. This concern is biologically plausible because the inferior retinacular artery, a terminal branch of the medial femoral circumflex artery, courses along the posteroinferior femoral neck within the Weitbrecht ligament and may be vulnerable during medial exposure. Cadaveric anatomical mapping has shown that the inferior retinacular artery typically lies between the 7:00 and 8:00 clockface positions, suggesting a relatively safe corridor for a buttress plate at 6:00 provided meticulous capsular elevation and avoidance of posterior retractor placement [46]. Consistent with this, a contrast-enhanced MRI cadaveric study reported only minimal perfusion reduction following calcar plating at 6:00 compared with a non-plated control, but emphasized the importance of preserving the inferior retinacular artery and using gentle retraction [47]. Clinically, a recent systematic review of MBP-augmented fixation reported low pooled AVN rates; however, the evidence is largely derived from small and heterogeneous series, and potential vascular and soft-tissue morbidity should still be considered when selecting MBP and planning the surgical approach [48]. In terms of interfragmentary shear stress, no significant advantage was observed between the two fixation modes.

Notably, the higher peak implant stress observed in the DHS + AS construct—particularly at stress-concentration regions such as the plate–barrel junction and adjacent cortical screws—may translate into a reduced fatigue margin under cyclic loading. Clinical reports and biomechanical analyses indicate that fatigue failure of sliding hip screw systems may occur at stress-riser regions, and cyclic loading can exceed the endurance limit of the implant and lead to fatigue-related failure [49]. From a mechanics perspective, higher implant stress reflects greater load transfer through the screw–plate–bone interfaces. Accordingly, constructs relying on intimate plate–bone contact and higher local compressive stress may compromise periosteal vascularity [50] and may therefore contribute to local tissue irritation.

Our study has several limitations: ① When constructing the femoral model, we assumed it to be a material with mean continuity and isotropic elasticity, even though human bone is actually anisotropic. However, this assumption was considered reasonable given that our study aimed to explore overall trends. ② Our results were not validated through in vitro mechanical experiments or clinical trials, but the finite element model was deemed acceptable based on convergence studies. ③ The surgical procedures for osteotomy and internal fixation simulated in this finite element analysis were idealized, with both the osteotomy and fracture reduction assumed to be in perfect condition. In addition, the fracture interface was modeled as a zero-gap bone–bone contact without an interposed fibrous layer, which may underestimate interface compliance and therefore affect the absolute magnitudes of interfragmentary motion and stress. Accordingly, our results should be interpreted primarily as comparative differences among constructs rather than absolute clinical thresholds. ④ The loading condition was simplified to a quasi-static single-leg stance and only an abductor force was applied; time-varying activities and other muscle forces were not simulated, which may limit the extrapolation of absolute values.

Conclusion

In summary, both the 10° and 20° osteotomy angle models showed improvements over the original internal fixation model in terms of shear stress, compressive stress, sliding distance, and gap at the fracture surface. Moreover, there was no significant improvement in the mechanical effect on the fracture site when the osteotomy angle was increased by 10°. As for the internal fixation devices, DHS + AS had some advantages in increasing interfragmentary compressive stress and reducing interfragmentary sliding distance. Consequently, patients treated with DHS + AS may warrant a more cautious early postoperative progression of weight-bearing; however, any specific weight-bearing regimen should be individualized and validated in future clinical studies. Simultaneously, with the progress in bone tissue engineering, the strength of implant materials may be enhanced, which holds promise for reducing the risk of implant failure.

Abbreviations

DHS

Sliding hip screw

AS

Anti-rotation screw

MBP

Medial buttress plate

FNF

Femoral neck fractures

ONFH

Osteonecrosis of the femoral head

Authors’ contributions

H.L. &; F.Z.: Conceptualization, Design and implement experiments, Methodology, Data curation, Writing; T.W. &; H.T.: Methodology, Data analysis, Writing; B.Q. & S.G.: Conceptualization, Writing, Supervision, Project administration, Funding acquisition. All authors reviewed the manuscript.

Funding

This research was funded by the Chongqing Municipal Science and Technology Bureau (Grant No.cstc2020jscx-sbqwX0010).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study was conducted in compliance with the Declaration of Helsinki and was approved by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University. The need for informed consent to participate was waived for all participants by the ethics committee and the Institutional Review Board (IRB) (Research Ethics Committees of the First Affiliated Hospital of Chongqing Medical University; Approval No. K2023-666). All methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.