Biomechanical study of screw-supported prosthesis in acetabular revision

Abstract

Aims

Acetabular bone defects pose a common challenge in revision total hip arthroplasty (THA). Our team has applied screw-supported prostheses to the revision of severe acetabular defects and achieved good early clinical results. However, in one of our cases, loosening of the screw-cup connection occurred five months after surgery. Therefore, we investigated the mechanical characteristics and weak points of this prosthesis system through finite element analysis (FEA) and mechanical testing, and proposed improvement strategies.

Methods

A screw-supported prosthesis used in clinical practice was modelled, and FEA was performed to obtain its displacement and von Mises stress distribution under gait loading. Emphasis was placed on the stresses at the screw-cup connection to analyze the failure mechanism. Mechanical compression tests were subsequently performed on the prosthesis, and the failure mode of the prosthesis verified the results of the FEA.

Results

FEA identified the posterior iliac and pubic locking connections as weak points prone to fatigue failure. Mechanical tests confirmed that the locking connection failed under 2,500 to 3,000 N loads. The improved sleeve connection enhanced connection strength and reduced failure risk.

Conclusion

The screw-supported prosthesis shows potential for severe acetabular defects, but the screw-cup connection remains a vulnerable site. Enhanced connection designs (e.g. sleeve connection) and bone grafting effectively mitigate failure risk.

Cite this article: Bone Joint Res 2026;15(3):204–214.

Introduction

Approximately 10% to 18% of total hip arthroplasty (THA) procedures worldwide are revisions,¹ with up to 50% involving acetabular bone defects.² Current clinical reconstruction strategies for these defects include hemispherical acetabular component,³ structural bone grafts,⁴ bone cement filling,⁵ metal augments,^6,7 cup-cage,^3,8 and triflange cups.^9,10 While clinically effective, these methods suffer from narrow applicability and high customization demands.^11,12 To overcome these limitations, our team proposed a screw-supported prosthesis inspired by engineering pile foundations and iliac stemmed cups (Figure 1).^13,14 This design implants long screws into deep, healthy periacetabular bone. Unlike standard acetabular screws, these screws engage the cup via a locking mechanism, providing a universal solution for diverse defect types. Based on pelvic anatomy and biomechanical load transfer characteristics (Supplementary Figure a), we positioned four long screws in the posterior ilium, anterior ilium, pubis, and ischium (Figure 1a). The stability of this configuration was validated through finite element analysis (FEA) and biomechanical test (Supplementary Figure b). Preliminary clinical application achieved favourable outcomes, enabling early postoperative weightbearing and functional recovery (unpublished data). Within the total cohort (n = 8), one patient received a prosthesis featuring a five-point star locking mechanism permitting 30° angular variation range in screw locking trajectories for intraoperative adjustment (Figure 1a). However, this five-point star locking connection led to unanticipated screw-cup connection failure at five months postoperatively. This case highlights the connection’s critical role in prosthetic integrity and identifies it as a structural vulnerability.

Fig. 1.

Screw-supported prosthesis designs. a) Five-point star locking prosthesis. b) Sleeve-connected prosthesis.

Understanding force types and stress distributions at screw-cup connections is essential to comprehensively evaluate the prosthesis’s biomechanical behaviour and guide the design refinements. Therefore, this study combines FEA and mechanical experiments to analyze load type, stress distribution, and failure mechanisms of the screw-cup connections. An enhanced sleeve connection design was proposed (Figure 1b) and validated for stability, followed by clinical implementation in revision surgery.

Methods

Prosthesis design

The initial revision employed a Screw-Supported Prosthesis system (WEGO, China) with a five-point star locking connection (Figure 1a), comprising four 8 mm-diameter screws: posterior iliac screw (80 mm), anterior iliac screw (40 mm), pubic screw (60 mm), and ischial screw (40 mm). These screws were paired with a 54 mm metallic acetabular component featuring a five-point star locking structure providing 30° angular adjustability for precise intraoperative placement. The secondary revision used an improved sleeve-connected prosthesis system (Figure 1b), featuring 10 mm thickened screws and a 56 mm titanium mesh cup. Dual fixation via sleeve-thread matching and locking holes enhanced connection strength.

Clinical preparation and application

Preoperative 3D reconstruction of the patient’s right pelvic CT data enabled fabrication of customized surgical guides and pelvic models via 3D printing. Simulated surgeries validated guide accuracy (Figure 2a). During the initial procedure, the original incision was reopened to remove the failed prosthesis. Kirschner wire positioning, guided by the surgical template, was confirmed via intraoperative CT. Bone defects were grafted with impaction bone grafting, followed by cup implantation. Screws were inserted into drill holes along the Kirschner wires and fixed to the cup’s pentagram-locking holes. Finally, the liner and femoral components were installed, joint mobility and stability were tested, and the wound was closed after antibiotic irrigation (Figure 2b). The secondary procedure followed similar steps (Figures 2c and 2d). However, due to firm fixation of the posterior iliac sleeve after compression and reaming, the sleeve was retained without screw arthroplasty. Informed consent was obtained from the patients, with approval from the Ethics Committee of Second Hospital of Jilin University for the use of clinical data and publication.

Fig. 2.

Clinical workflow of screw-supported prostheses in two revision surgeries. a) Preoperative planning for the first revision (five-point star locking prosthesis). b) Intraoperative view of the first revision. c) Preoperative planning for the second revision (sleeve-connected prosthesis). d) Intraoperative view of the second revision.

Finite element analysis

Patient-specific CT-based 3D models were used for FEA of both five-point star locking and sleeve-connected prostheses. In HyperMesh 2021 software (Altair Engineering, USA), tetrahedral meshing was performed using C3D10 elements with a global element size of 1 mm for the pelvic bone, refined to 0.3 mm at screw threads. A mesh sensitivity analysis was performed to ensure that results are independent of mesh size (Supplementary Figure c). The acetabular component was positioned at 45° abduction and 15° anteversion, consistent with surgical placement.¹⁵ Material properties for metallic components (Ti6Al4V) included an elastic modulus of 110 GPa and Poisson’s ratio of 0.3.¹⁶ Pelvic bone material properties were derived from CT Hounsfield units using established density-modulus relationships:¹⁷

$${\displaystyle {\displaystyle \rho =0.00069141\ast \mathrm{H}\mathrm{U}+1.026716}}$$

$${\displaystyle {\displaystyle {\displaystyle \mathrm{E}=2017.3{\rho }^{2.46}}}}$$

(where ρ = bone density in mg/cm³, E = elastic modulus in MPa)

Boundary conditions in Abaqus 2021 (Dassault Systèmes, France) (Figure 3a) fixed the pubic symphysis and sacroiliac joints to simulate physiological constraints. Screw-bone interfaces were modelled as bonded contacts to simplify computation, and bone-cup contact was assumed absent to represent worst-case clinical scenarios (despite partial support from intraoperative impaction grafting). Additional analyses included bone-cup contact (friction coefficient μ = 0.3) to evaluate stress reduction effects by graft osseointegration. Dynamic implicit analysis applied loading of hip joint throughout the gait cycle. The five-point star locking connections were modelled as tie constraint. The load profile referenced Bergmann’s gait-phase data (Figure 3b),¹⁸ scaled to patient body weight. Load components (x: medial, y: anterior, z: superior) aligned with the radiological anatomical coordinate system.

Fig. 3.

Finite element analysis (FEA) and biomechanical test configurations. a) Boundary conditions for FEA. b) Applied gait-phase loading profile (timepoints indicate peak stress and joint force magnitudes). c) Surgical simulation on hemipelvic model. d) Biomechanical test setup. BW, body weight.

Fatigue life assessment was conducted using Fe-safe 2021 (Dassault Systèmes). Stress analysis results under gait loading were imported. The load curve was defined as a sinusoidal function (0 to 1 normalized amplitude), simulating cyclic joint motion during walking. Material properties were configured using the built-in Ti6Al4V alloy from the Fe-safe material library, which includes predefined stress-life (S-N) curves accounting for surface finish and notch effects. Fatigue life distribution across the prosthetic structure was calculated via the S-N approach, suitable for high-cycle fatigue regimes (> 10⁴ cycles).

The output parameters of FEA include the von Mises stress distribution across the entire pelvis-prosthesis system, the displacement results of the prosthesis, and the axial force results acting on individual screws. The von Mises stress is used to assess the risk of bone fracture or prosthetic failure, with particular emphasis on stress concentrations at the screw-cup connections. Furthermore, the displacement distribution of the prosthesis system and the force analysis of the screws are compared with the prosthesis failure displacement patterns observed in clinical cases via radiograph imaging and CT-derived 3D reconstruction, thereby validating the accuracy of the FEA and elucidating the failure mechanisms of the prosthesis.

Mechanical testing

Mechanical compression tests evaluated the failure modes of the five-point star locking prosthesis (Figures 3c and 3d) and validated FEA results. The pelvic model was stabilized using rapid-curing cement at the pubic symphysis and sacroiliac joints; fixation orientations aligned FEA force application directions. A Sinter 10 KN uniaxial testing machine (WDW-10KN; Sinter, China) applied compressive loads at 2 mm/min (displacement control).¹⁹ Biomechanical testing was repeated three times on the same model, with each test terminated upon structural failure (sudden load drop). Following each test, the prosthesis was disassembled to observe connection failure locations. Additionally, vertical strain at five iliac points near the acetabular component was measured using strain gauges for direct FEA validation. Force measurement accuracy was 0.1 N, and strain data accuracy was 1 µɛ.

Results

Clinical outcomes

Postoperative radiographs and CT scans confirmed precise positioning of the four anchoring screws and restoration of the hip rotation centre following the initial surgery (Figure 4a). At the five-month follow-up, right hip pain prompted imaging that revealed acetabular component loosening with valgus displacement: the posterior iliac and pubic screws remained stable, whereas the anterior iliac and ischial screws exhibited migration, accompanied by complete disengagement of the pubic screw from the cup (Figure 4b). Revision surgery using sleeve-connected prosthesis achieved accurate intraoperative positioning and maintained stable fixation without signs of migration or loosening at six-month follow-up (Figure 4c).

Fig. 4.

Postoperative imaging (anteroposterior radiographs and 3D CT reconstructions). a) Immediate postoperative status with locking prosthesis. b) Mechanical failure of locking prosthesis at five-month follow-up. c) Stable fixation after revision with sleeve-connected prosthesis at six-month follow-up.

FEA of five-point star locking prosthesis

Displacement analysis (Figure 5 and Figure 6b): The acetabular component exhibited superior-lateral displacement, reflecting compressive loading on iliac screws versus tensile forces on pubic/ischial screws. The anterior iliac screw (0.64 mm) and pubic screw (0.46 mm) exhibited greater overall displacement magnitudes than the posterior iliac screw (0.14 mm) and ischial screw (0.11 mm), attributable to deeper osseous embedding of the latter. Displacement vector analysis indicated superior Z-axis (+) movement in the cup’s upper hemisphere and inferior Z-axis (-) displacement in the lower region, consistent with clinical failure patterns. Notably, the ischial screw’s displacement (-0.37 mm Z-axis) matched postoperative CT findings (Figure 5b).

Fig. 5.

Displacement analysis of the five-point star locking prosthesis. a) Contour plot of total displacement magnitude. b) Vector decomposition of displacement components.

Fig. 6.

Key biomechanical parameters from finite element analysis. a) Load-time curves for screws during gait cycle. b) Peak displacement at screw tips. c) Peak stress at screw connection interfaces. d) Peak stress at five-point star locking holes of the acetabular cup.

Screw loading (Figure 6a): The posterior iliac screw sustained peak compressive loads (933.4 N), while the pubic and ischial screws sustained tensile loads (235.4 N/251.1 N, respectively). This loading pattern correlated with failure mechanisms: pubic screw pullout from rigid cortical fixation versus ischial screw migration due to shallow fixation depth and compromised cancellous bone quality.

Stress distribution (Figure 6c and Figure 7): Maximum von Mises stresses concentrated at the five-point star locking interface, reaching 534.2 MPa (posterior iliac screw threads) and 346.8 MPa (acetabular component locking holes). Fatigue life analysis predicted failure at 65,978 gait cycles under conditions lacking bone graft support. Following osseointegration, bone graft structural support reduced peak stresses by 101.5 MPa (screws) and 78.6 MPa (cup), mitigating fatigue risks. At the screw-bone interface, stresses remained low: posterior iliac screw (56.6 MPa), anterior iliac screw (11.7 MPa), pubic screw (70.1 MPa), ischial screw (28.6 MPa), with no fracture risk identified.

Fig. 7.

von Mises stress distribution on a) the pelvis and b) the prosthesis system.

Mechanical test results of the five-point star locking prosthesis

Three compression trials were conducted (Figure 8a). The first trial induced a single-tooth fracture of the posterior iliac locking hole at 2,808 N. The second trial caused two-tooth fractures of the same hole at 2,516 N and 3,523 N. The third trial resulted in pubic locking hole failure at 2,501 N. Post-test evaluation revealed severe damage in the posterior iliac and pubic connections (screw thread/locking hole fractures), with mild deformation in the anterior iliac and ischial regions (Figure 8b and 8c). Notably, retrieved clinical prostheses exhibited matching failure patterns: damage in posterior iliac/pubic connections versus minimal ischial damage (Figure 8d). Pelvic strain measurements during mechanical testing aligned with FEA-predicted trends (Supplementary Figure d), validating the FEA methodology.

Fig. 8.

Mechanical test results of the five-point star locking prosthesis. a) Force-displacement curves from three compression tests. b) Failure initiation sites (red highlights) at connection interfaces. c) Post test damage morphology at screw-cup connections. d) Retrieved clinically failed prosthesis.

FEA of the sleeve-connected prosthesis

FEA demonstrated excellent stability of the sleeve-connected prosthesis (Figure 9). Maximum stresses were recorded as 246.0 MPa at the titanium mesh cup locking screws, 146.5 MPa at the posterior iliac sleeve-bone interface (specifically at the sleeve tip), and 82.6 MPa at the sleeve locking screws. Stresses at the pubic/ischial screws remained below 28.9 MPa. All values were significantly lower than the titanium alloy yield strength (795 to 827 MPa), confirming negligible mechanical failure risk under physiological loading.

Fig. 9.

von Mises stress distribution in the sleeve-connected prosthesis from finite element analysis.

Discussion

Severe acetabular bone defects remain a challenge in revision arthroplasty. The screw-supported prosthesis anchors screws in deep healthy periacetabular bone to support the acetabular component, potentially overcoming limitations of conventional strategies. Although preliminary mechanical validation and clinical application demonstrated satisfactory outcomes, a five-point star locking mechanism led to catastrophic prosthetic failure within five months. Radiological analysis revealed fractures at the posterior/anterior iliac and pubic screw-cup junctions, identifying this locking interface as the primary failure site. This early failure underscores a critical clinical dilemma: modular connections enable intraoperative adaptability for complex defects but introduce mechanical vulnerability. Specifically, active patients exceeding weightbearing restrictions during graft integration (8 to 12 weeks) face high risks of catastrophic loosening. Targeted investigation into the cup-screw connection biomechanics is thus imperative for design refinement.

Through FEA and biomechanical testing, we elucidated the failure mechanisms of the five-point star locking prosthesis. FEA enabled the assessment of stress states across various regions of the prosthesis, thereby identifying high-risk zones for structural failure. Furthermore, it quantified the magnitude and direction of forces acting on individual screws, providing critical evidence to support iterative design improvements of the prosthesis. More importantly, mechanical testing and radiological evaluations from clinical cases consistently demonstrated failure patterns congruent with the high-risk regions predicted by FEA, thereby validating the accuracy of this computational methodology.

The FEA results showed that under gait loading, the cup demonstrated tendencies for superolateral displacement and valgus tilt (Figure 5b), consistent with radiological findings in follow-up examinations. This suggests that screws in the iliac region may serve as the primary load-bearing components. Force analysis identified the posterior iliac screw as the primary load-bearing component, explaining its predisposition to failure. The pubic and ischial screws experienced pull-out forces, accounting for the observed screw dislodgement in follow-up imaging (Figure 4b). Stress distribution analysis revealed stress concentration at the five-point star locking connections, with peak stresses reaching 534 MPa at the posterior iliac connection – below the yield strength of Ti6Al4V,²⁰ but sufficient to cause fatigue fracture after approximately 66,000 gait cycles. For patients experiencing prosthesis failure within five months postoperatively, this equates to only 400 daily loading cycles before failure. Subsequent compression mechanical testing confirmed our FEA results. The posterior iliac column screw, as the primary load-bearing component, exhibited initial failure at its connection site, followed by the pubic screw connection. These failures occurred under loads of approximately 2,500 to 3,000 N, a critical threshold indicating that even moderate running or jumping could lead to connection failure when relying solely on screws for cup support. These findings provide clinicians with two pivotal strategies for prosthesis design refinement. First, incorporating reinforcement structures such as buttress plates in primary load-bearing zones of the iliac region can distribute mechanical loads away from iliac screws, thereby mitigating fracture risks at posterior column screw connections. Second, implementing optimized connection geometries can minimize stress concentration at screw connections.

In clinical practice, the second connection design optimization scheme was validated and implemented to enhance structural integrity and mitigate stress concentration. A sleeve connection design was introduced, which reinforced the connection through multi-threaded locking between screws and sleeves, combined with multiple locking screws securing the sleeve to the titanium mesh cup. FEA demonstrated more uniform stress distribution and reduced stress concentration in the sleeve-connected prosthesis, markedly improving overall system stability. The clinical application of the sleeve-connected prosthesis in revision surgery also demonstrated exceptional stability, with six-month radiological follow-up confirming the absence of migration or loosening.

Notably, FEA confirms that viable bone grafts supporting the acetabular component reduce stress concentrations at the screw-cup junction, effectively mitigating the risk of fatigue fracture. Osseointegrated grafts typically achieve load-bearing capacity at 8 to 12 weeks postoperatively.²¹ Clinicians should note that premature weightbearing (< 3 months) induces micromotion at the prosthesis-bone interface, thereby inhibiting bone ingrowth onto the graft surface and promoting fibrous tissue formation at the interface.²² Consequently, during the early postoperative phase following revision surgery, full weightbearing should be avoided, with serial follow-up assessments recommended to evaluate graft viability and monitor changes in bone mineral density.

This study has several limitations. First, the patient-specific pelvis used in mechanical testing was fabricated via resin-based 3D printing, resulting in material properties distinct from those of natural bone. However, given the study’s focus on evaluating failure patterns at the cup-screw interface, we deliberately used patient-specific models rather than standardized biomechanical testing models (e.g. Sawbones) to accurately replicate the anatomical fixation geometry of the prosthesis. Second, in the FEA, cup-screw connections were modelled using tie constraints without accounting for pre-load forces, as intraoperative screw tightness varies clinically and is quantitatively challenging to standardize. Consequently, the interfacial stresses derived from FEA represent stress concentrations induced solely by patient weightbearing; nonetheless, they provide valuable mechanistic insights into biomechanical behaviour and guide design refinements for connection mechanisms.

In conclusion, this study demonstrates that in screw-supported prosthesis reconstruction for severe acetabular defects, the connection sites between the posterior iliac/pubic screws and the acetabular component are critical weak points prone to fatigue failure. To mitigate implant failure risks, two strategies are essential: 1) ensuring bone graft incorporation to provide biological support for the acetabular component; and 2) adopting reinforced connection designs such as the sleeve mechanism to enhance mechanical stability.

Data Availability

The data that support the findings for this study are available to other researchers from the corresponding author upon reasonable request.