A 24-year-old healthy male (height 185 cm, weight 70 kg) was selected as the experimental subject. After he signed the informed consent form, a Siemens SOMATOM Definition AS scanner (128 Slice, 70 kV steps) was used to complete the knee joint scan. The DICOM-format tibial CT images were imported into Mimics 21.0. By setting appropriate grayscale values to distinguish the cortical bone and cancellous bone of the tibia, the three-dimensional model of the original knee joint was constructed. Subsequently, the model was imported into Geomagic Wrap 2021 for surface defect repair, smoothing, and precise optimization of surface functions.

After reconstructing the patient’s knee joint CT data on a computer and simulating the tibia to its true anatomical position, bone filler was placed in the tibial plateau defect to form an integrated structure matching autologous bone and the prosthetic tibia. Precise thickness, width, and height of the filler’s smooth curved transition were determined by biomechanical simulations, defect features, and residual bone mass.

To prevent stress shielding, the non-porous integrated tibial prosthesis (ITP) model was topologically optimized via Magics software (Materialise, Belgium) — adjusting internal mechanical paths and solid distribution to ensure structural stability and reduce stress concentration from redundant material.

Finally, the optimized ITP model was imported into an EOS M 290 metal 3D printer (EOS GmbH, Germany) for prosthesis fabrication using laser powder bed fusion. Equipped with a 400W fiber laser, the printer used a 30 μm layer thickness for precision. The material was ASTM F799-compliant cobalt-chromium-molybdenum alloy (UNS R31537: 63% Co, 28%–29% Cr, 6% Mo, plus trace Si, W), which offers excellent wear/corrosion resistance and mechanical strength, meeting long-term implantation needs.

Based on the anatomical parameters of the subject’s knee joint and with reference to a completed surgical case of tibial defect, a 200 mm tibial plateau defect model with a depth exceeding 20 mm and a defect area exceeding 60% was constructed in SOLIDWORKS (Figure 1a). In accordance with the principles of primary TKA (total knee arthroplasty), the tibial prosthesis implantation parameters were determined as follows: 0° varus (ensuring the proximal tibial resection is perpendicular to the tibial mechanical axis in the coronal plane), 1° posterior slope (the angle between the proximal tibial resection plane and the tibial anatomical axis in the sagittal plane), and 3° external rotation (the posterior edge line of the tibial tray rotates outward by 3° relative to the posterior edge line of the proximal tibial resection in the axial plane) (Dong et al., 2020). To simplify computational processes, modeling of the tibial liner, femoral component, and femoral stem was omitted; additionally, modeling of the ligaments and menisci was also excluded. Three-dimensional modeling and assembly of the knee prosthesis (including a 7 cm short stem), spacer, cone, sleeve, and 3D-printed integrated prosthesis were completed in SolidWorks (all prosthesis parameters were provided by JST Medical). A 2.0 mm-thick bone cement layer was used to fix the tibial tray and augments to the resected upper surface of the tibia, and the bone tissue area separated from the defect region was defined as the bone cement filling area (Figure 1b). The assembly was imported into Hypermesh, where the finite element model was meshed using C3D4 tetrahedral elements (Figure 1c). After exporting as an. hm file, it was imported into Abaqus for stress analysis. This finite element model was validated by comparison with in vitro mechanical test data from Zheng et al. (2020), with a contact stress error within 5%, ensuring its reliability (Zheng et al., 2020). Cortical bone and cancellous bone were fully bonded, and cancellous bone was fully bonded with the bone cement layer. The general contact algorithm was used to model the contact between bone and prosthesis, as well as between prosthesis and augments, with a constant friction coefficient of 0.3 applied to all models. In the finite element analysis, the femoral part of the prosthesis was simplified to a vertically downward load averaging on the plane of the tibial prosthesis. The tibia was truncated, and all degrees of freedom were fixed at a distance of 200 mm (Figure 1d).

The finite element models were categorized into four groups based on component types: Group A (Spacer group), Group B (Cone group), Group C (Sleeve group), and Group D (Integrated prosthesis group). All models were meshed using C3D4 tetrahedral elements. To ensure geometric similarity and computational accuracy of the models, a global mesh size of 1 mm was set, and all materials were assumed to be continuous, isotropic, and homogeneous linear elastic bodies (Zheng et al., 2020). Table 1 details the mechanical parameters of each component in the model.

The distal surface of the tibia was fully fixed in all directions. Based on the results of peak stress on the knee joint reported in relevant literature, the load when standing on both legs is 1 time the body weight (700N) (Northon et al., 2018),The load when the knee joint is flexed is 2.5 times the body weight (1750N) (Mesfar and Shirazi-Adl, 2005),The load when climbing stairs is 3 times the body weight (2100N) (Makani et al., 2022),The load during jogging is 3.5 times the body weight (2450N) (Miller et al., 2014). It is applied averagely on the tibial prosthesis platform.

In the systematic assessment of AORI type Ⅲ tibial plateau defect reconstruction, the initial stability of the ITP and the other three bone implants was compared and analyzed by quantifying the contact stress at the interface between the tibial prosthesis and the host bone, the settlement of the tibial plateau, and the relative displacement. Meanwhile, the KSS/KOOS knee joint scoring system was used to clinically evaluate the postoperative knee joint function. Additionally, imaging methods were employed to closely observe key indicators such as radiolucent lines, displacement, aseptic loosening, osteolysis, and bone growth between the prosthesis and the bone surface, establishing a dual evaluation system integrating biomechanics and clinical function. Gait analysis was conducted using the Walkway MW - 1,000 type 2.4 m × 0.6 m sheet - shaped foot pressure - ground gait analyzer from ANIMA Company on a 6.4 m × 0.6 m dedicated walkway (including 2 - meter front and rear auxiliary paths). The subjects completed two walking tests at a comfortable and natural speed. The system collected basic gait parameters, such as walking speed, step length, step width, and cadence, as well as unilateral gait parameters, including stride time, stance time, swing - phase time, double - limb support - phase time, stride, and step length. The research design included a longitudinal comparison of the basic gait information of the preoperative group and the 1 - year postoperative group, as well as a transverse and longitudinal cross - comparison of the basic and unilateral gait parameters of the affected and contralateral sides of the OA group before and 1 year after surgery. This comprehensive assessment of the impact of surgical intervention on the patient’s gait pattern provided an objective and quantitative basis for clinical efficacy evaluation.

Initial stability was evaluated by quantifying the contact stress at the prosthesis-bone interface, tibial plateau settlement, and relative micromotion. The contact interface between the prosthesis and bone is divided into cortical bone and cancellous bone. For the peak stress at the cortical bone interface: the maximum stress peaks of ITP at the contacting cortical bone interface are 5.425 MPa (700 N), 13.64 MPa (1750 N), 16.32 MPa (2100 N), and 19.06 MPa (2450 N); those of the spacer are 6.66 MPa (700 N), 16.74 MPa (1750 N), 20.03 MPa (2100 N), and 23.4 MPa (2450 N); those of the cone are 5.716 MPa (700 N), 14.37 MPa (1750 N), 17.19 MPa (2100 N), and 20.08 MPa (2450 N); those of the sleeve are 5.734 MPa (700 N), 14.44 MPa (1750 N), 17.27 MPa (2100 N), and 20.18 MPa (2450 N); the maximum stress concentration point is at the distal end of the cortical bone. For the maximum stress peaks at the contacting cancellous bone interface: those of ITP are 0.868 MPa (700 N), 2.182 MPa (1750 N), 2.612 MPa (2100 N), and 3.050 MPa (2450 N); those of the spacer are 1.211 MPa (700 N), 3.044 MPa (1750 N), 3.645 MPa (2100 N), and 4.255 MPa (2450 N); those of the cone are 1.080 MPa (700 N), 2.714 MPa (1750 N), 3.249 MPa (2100 N), and 3.794 MPa (2450 N); those of the sleeve are 1.118 MPa (700 N), 2.810 MPa (1750 N), 3.364 MPa (2100 N), and 3.928 MPa (2450 N); the maximum stress concentration point is at the contact area between the cancellous bone and the distal end of the extension stem. For the peak stresses at the prosthesis interface: those of ITP are 25.59 MPa (700 N), 64.31 MPa (1750 N), 77.01 MPa (2100 N), and 89.9 MPa (2450 N), with the stress concentration point at the corner of the extension stem; those of the spacer are 18.82 MPa (700 N), 47.3 MPa (1750 N), 56.63 MPa (2100 N), and 66.12 MPa (2450 N); those of the cone are 19.17 MPa (700 N), 48.19 MPa (1750 N), 57.69 MPa (2100 N), and 67.37 MPa (2450 N); those of the sleeve are 30.62 MPa (700 N), 76.67 MPa (1750 N), 92.18 MPa (2100 N), and 107.60 MPa (2450 N); the maximum stress concentration is at the corner of the extension stem (Figures 3–8).

The relative micromotion between the tibial prosthesis and the tibia—defined as the displacement difference between two adjacent nodes (one on the bone and one on the tibial plateau)—were recorded. For the ITP, the peak micromotions relative to the host bone were 15.2 μm (700 N), 38.3 μm (1750 N), 45.7 μm (2100 N), and 53.6 μm (2450 N). For the spacer, the peak micromotions relative to the host bone were 16.6 μm (700 N), 41.7 μm (1750 N), 49.9 μm (2100 N), and 58.3 μm (2450 N). For the cone, the peak micromotions relative to the host bone were 19.17 μm (700 N), 48.19 μm (1750 N), 57.69 μm (2100 N), and 67.37 μm (2450 N). For the sleeve, the peak micromotions relative to the host bone were 17.8 μm (700 N), 44.6 μm (1750 N), 53.3 μm (2100 N), and 62.3 μm (2450 N), respectively (Figures 9–11).

Clinical functional recovery was evaluated using the KSS scoring system, X-ray imaging to assess prosthesis position and bone integration, and gait analysis.The patient was followed up for 24 months. The preoperative KSS knee score was 46.5 ± 4.8. It was 80.1 ± 6.6 at 3 months postoperatively and 85.4 ± 5.5 at the last follow-up (Figure 12). At the last follow-up, there was no swelling, infection or postoperative complications at the surgical site. X-rays of the knee joint in anteroposterior position of all patients showed no adverse conditions such as radiolucency, loosening, or osteolysis around the 3D-printed ITP and bone surface (Figures 13A–C).

In terms of basic gait information, the preoperative mean walking speed was 0.49 m/s, with step lengths of 45.4 cm, cadences of 85.8 steps/min, and step widths of 10.3 cm. At 1 year postoperatively, the mean walking speed increased to 1.08 m/s, step lengths increased to 52.6 cm, cadences rose to 122.8 steps/min, while there was no significant improvement in step widths. In the unilateral gait information, the preoperative stance times of the contralateral and affected sides were 0.71 s and 0.68 s, respectively; the swing phase times were 0.35 s and 0.40 s, respectively; the double-limb support phase times were 0.18 s and 0.17 s, respectively. Both sides had a stride time of 1.07 s, a stride of 88.0 cm, and step lengths of 45.4 cm and 45.9 cm, respectively. At 1 year postoperatively, the stance time, swing phase time, and double-limb support phase time of the contralateral and affected sides tended to be consistent, being 0.61 s, 0.36 s, and 0.12 s, respectively. Both sides had a stride time of 1.00 s, a stride increased to 105.6 cm, and step lengths of 52.7 cm and 52.6 cm, respectively (Table 2).