The evaluation of osteoconductive WE43 magnesium as a therapeutic approach for nonunion treatment Authors: Alina Marie Schmitz, Judith Hahnemann, Katharina Schmidt-Bleek, Jakob Fenske, Agnes Ellinghaus, Denise Jahn, Heilwig Fischer, Georg Duda, Max Heiland, Carsten Rendenbach Published: 2026 ABSTRACT AimsNonunion is a challenging condition marked by impaired bone healing and vascularization. While magnesium-based biomaterials show osteogenic and angiogenic potential, their role in nonunion remains controversial. Controlled modification of magnesium-based implants may enable regenerative osteogenesis.MethodsHere, WE43, a magnesium alloy, was investigated in a standardized femoral nonunion model in C57Bl/6 mice. Three groups were compared: WE43 alone, WE43 with cancellous bone (WE43+), and cancellous bone alone (control, CTR). Bone healing was assessed at eight and 12 weeks using micro-CT, histology, and immunohistochemistry. Key outcomes included defect bridging, bone volume, mineralization, trabecular metrics, osteoclast/osteoblast activity, and vascularization.ResultsWE43+ significantly reduced defect size and improved bone formation and microvascular density, nearing control levels. In contrast, WE43 alone showed high osteoclast activity, poor mineralization, and fibrotic healing. Magnesium’s regenerative performance depended strongly on degradation rate and osteoconductive support. Combining WE43 with cancellous bone improved osseointegration and angiogenesis.ConclusionWE43 requires interaction with cancellous bone to support regeneration. Future studies should clarify which properties of cancellous bone enhance magnesium-based healing.Cite this article: Bone Joint Res 2026;15(6):675–685. Article focus To evaluate the regenerative potential of osteoconductive WE43 magnesium alloy in a validated murine nonunion model. To evaluate the regenerative potential of osteoconductive WE43 magnesium alloy in a validated murine nonunion model. To assess whether combining magnesium with cancellous bone (WE43+) enhances bone regeneration and vascularization. To assess whether combining magnesium with cancellous bone (WE43+) enhances bone regeneration and vascularization. To explore the role of magnesium degradation dynamics in impaired or enhanced healing outcomes. To explore the role of magnesium degradation dynamics in impaired or enhanced healing outcomes. Key messages WE43 magnesium alone was insufficient to induce bone healing and led to elevated osteoclastic activity, fibrotic tissue deposition, and poor vascularization. WE43 magnesium alone was insufficient to induce bone healing and led to elevated osteoclastic activity, fibrotic tissue deposition, and poor vascularization. The combination of WE43 with cancellous bone (WE43+) significantly improved defect bridging, bone volume, and vessel formation, approaching control levels. The combination of WE43 with cancellous bone (WE43+) significantly improved defect bridging, bone volume, and vessel formation, approaching control levels. Scaffold composition and degradation control are critical for effective bone regeneration and should guide future therapeutic strategies. Scaffold composition and degradation control are critical for effective bone regeneration and should guide future therapeutic strategies. Strengths and limitations This is the first study to investigate magnesium-based implants in a validated murine nonunion model, addressing a key translational gap. This is the first study to investigate magnesium-based implants in a validated murine nonunion model, addressing a key translational gap. This is a comprehensive multimodal analysis (micro-CT, histology, immunohistochemistry) providing robust insights into bone architecture, remodelling, and angiogenesis. This is a comprehensive multimodal analysis (micro-CT, histology, immunohistochemistry) providing robust insights into bone architecture, remodelling, and angiogenesis. The lack of direct quantification of magnesium degradation and local ion concentrations limits mechanistic interpretation of adverse responses. The lack of direct quantification of magnesium degradation and local ion concentrations limits mechanistic interpretation of adverse responses. Introduction Nonunion is a failure of bone healing from a fracture or osteotomy, commonly classified as hypertrophic and atrophic,1,2 depending on the biological activity at the fracture site. Hypertrophic nonunion often results from insufficient mechanical stabilization, whereas atrophic nonunion is characterized by impaired osteogenesis and vascular activity, fibrotic tissue infiltration, and localized necrosis. Both hypertrophic and atrophic often necessitate surgical intervention and pose challenges in both orthopaedic and maxillofacial surgery,3 especially in critical-sized defects or compromised patients. For treatments, autologous bone grafting remains the standard,4 but is limited by availability and donor-site morbidity.5 An alternative, magnesium, can be used as it has biodegradable and bioactive properties – osteoconductive, osteoinductive, and angiogenic – which makes it a promising biomaterial.6-9 However, rapid degradation and gas formation occurs, leading to impaired stability.10 WE43 is a magnesium alloy containing rare earth elements, primarily yttrium (Y) and neodymium (Nd), as well as smaller amounts of gadolinium (Gd) and zirconium (Zr), offering improved degradation behaviour.11-13 This study evaluates WE43’s regenerative potential in a murine nonunion model, hypothesizing enhanced bone repair with scaffold augmentation. Methods Study design In this study, 36 female C57Bl/6 mice (nine to ten weeks old) were treated according to a well-characterized nonunion model, and 24 additional animals were used as cancellous bone donors.14 All animals were obtained from Janvier Labs (France). A two-factorial study included three treatment groups, control (CTR), WE43 magnesium implant, and WE43+ (WE43 plus cancellous bone), and two endpoints (eight and 12 weeks). Animals were randomly allocated to groups. The CTR group received cancellous bone grafts, the WE43 group received only magnesium implants, and the WE43+ group received both treatment (Figure 1). Group allocation was performed by the investigator (HF), who was not involved in outcome assessment. HF was aware of group assignment, but imaging, histological analysis, and data evaluation were conducted blinded. Animals were randomly assigned to treatment groups, and standardized procedures were used throughout. Housing conditions were consistent, though cage positions and assessment order were not formally randomized. Experimental workflow for a mouse femur nonunion model, including defect creation, evaluation, treatment with magnesium (WE43) and/or cancellous bone, and efficacy assessment at eight and 12 weeks after nonunion treatment using radiological and histological methods. Confirmation of the established nonunion was conducted four weeks after defect induction and before the experimental procedures, using in vivo CT imaging and direct visual inspection during the revision surgery. CTR, control; WE43, magnesium alloy containing yttrium and neodymium; WE43+, WE43 plus cancellous bone. Magnesium implant Magnesium implants were fabricated from pre-fabricated WE43 alloy wires (Magnesium alloy rod WE43C T5) supplied by Sino Zheng Industrial Inc (China). The alloy contains 4.04% Y, 2.17% Nd, 0.7% Gd, 0.46% Zr, and less than 1% of other elements (Cu, Fe, Ni, Mn, Zn, Ag), with magnesium as the base metal. Animal model All procedures followed EU Directive 2010/63/EU.15 We have adhered to the ARRIVE guidelines and have included the ARRIVE checklist in the Supplementary Material. Mice were housed under standard conditions at the Research Facility for Experimental Medicine, Charité – Universitätsmedizin Berlin, Germany. Rigid enrichment (plastic shelters and tunnels) was removed postoperatively to prevent injury from the external fixator; the soft nesting material, MouseDis (RISystem AG, Switzerland), was used to the animal’s needs. The external fixator (polyether-ether-ketone (PEEK)), and titanium pins/nails remained in place throughout the experimental period. Surgical procedure All surgeries were performed under isoflurane anaesthesia (1.8 to 2.2% in O₂, 0.9 ml/min) (CP-Pharma, Germany). Body temperature was maintained and depth monitored via respiratory rate. Buprenorphine (0.03 mg/kg) (Bupresol, CP-Pharma) and clindamycin (8 mg/kg) (Ratiopharm, Germany) were administered subcutaneously for analgesia and prophylaxis. To prevent corneal drying during anaesthesia, eye ointment (Bepanthen Eye and Nose Ointment; Bayer, Germany) was applied. Under sterile conditions, a 2.5 mm femoral defect was created in the left leg using two osteotomies and a wire saw, following distractor fixation using MouseDis (RISystem) with four titanium screws. The bone fragment was removed, and the side of the bone fragment was irrigated and closed with non-resorbable Ethicon prolene 5/0 sutures (Johnson & Johnson, USA). Surgery duration was 20 minutes, followed by 30 minutes of postoperative monitoring. Analgesia was maintained with Tramadol (Tramal; Grünenthal, Germany) at a concentration of 0.5 to 1 mg/ml. Four weeks after defect creation, cancellous bone was harvested from donor mice, euthanized with ketamine (75 mg/kg, Inresa Arzneimittel, Germany) and medetomidine (1 mg/kg, Cepetor, CP-Pharma). Both femora were exposed and bisected, and trabecular bone collected from the metaphysis, epiphysis, and the bone marrow. Reoperation followed the same anaesthesia protocol. Nonunion was confirmed by in vivo micro-CT and intraoperative inspection using modified radiological union score of tibia (mRUST) scoring system.16 The defect gap (0.78 ± 0.18 mm) was refined with an oscillating saw and re-enlarged to 2.5 mm. Group-specific procedures:WE43 group: Magnesium implant inserted and fixed.CTR group: Defect filled with autologous cancellous bone.WE43+ group: Combination of WE43 implant and cancellous bone.The soft-tissue closure was anatomical. Reoperation time averaged 30 minutes. Animals were euthanized at eight or 12 weeks post-treatment. WE43 group: Magnesium implant inserted and fixed. WE43 group: Magnesium implant inserted and fixed. CTR group: Defect filled with autologous cancellous bone. CTR group: Defect filled with autologous cancellous bone. WE43+ group: Combination of WE43 implant and cancellous bone.The soft-tissue closure was anatomical. Reoperation time averaged 30 minutes. Animals were euthanized at eight or 12 weeks post-treatment. WE43+ group: Combination of WE43 implant and cancellous bone. The soft-tissue closure was anatomical. Reoperation time averaged 30 minutes. Animals were euthanized at eight or 12 weeks post-treatment. WE43 group: Magnesium implant inserted and fixed. WE43 group: Magnesium implant inserted and fixed. CTR group: Defect filled with autologous cancellous bone. CTR group: Defect filled with autologous cancellous bone. WE43+ group: Combination of WE43 implant and cancellous bone.The soft-tissue closure was anatomical. Reoperation time averaged 30 minutes. Animals were euthanized at eight or 12 weeks post-treatment. WE43+ group: Combination of WE43 implant and cancellous bone. The soft-tissue closure was anatomical. Reoperation time averaged 30 minutes. Animals were euthanized at eight or 12 weeks post-treatment. Radiological analysis Micro-CT imaging was performed in vivo (week 4) and post-mortem (weeks 8 and 12). In vivo scans were acquired using a MiLabs U-XTruhr scanner (MILabs, Netherlands) at 11 µm isotropic voxel size (reconstructed with MILabs Rec v11.00). Following euthanasia, specimens were fixed in 4% paraformaldehyde (diluted from 16% PFA; Electron Microscopy Science, USA) prepared in phosphate‑buffered saline (PBS) (Waldeck, Germany) for 24 hours. Ex vivo scans were performed using a SkyScan 1172 system (Bruker, USA) at 7.46 µm resolution and reconstructed in NRecon (v1.7.0.3, Bruker) with standardized artifact correction. A 3 mm volume of interest (VOI) was defined between the distractor screw holes, bounded by bone cortex and callus. Quantitative and qualitative analyses were conducted in ImageJ (v1.53, National Institutes of Health, USA). The healing progress was further assessed using mRUST,17,18 evaluating all four cortices (anterior, posterior, medial, lateral) in anteroposterior and lateral views. Cortices were scored 1 to 4 (total range: 4 to 16); scores > 10 indicated possible healing, while < 7 reflected persistent nonunion. The use of mRUST has been validated for murine femoral models, including C57BL/6 mice.19 Histological preparation After micro-CT imaging, samples were rinsed in PBS and incubated at 4°C for 24 hours to remove the fixative. Tissues harvested at eight weeks were processed as cryosections; those at 12 weeks underwent plastic embedding. Cryosectioning involved stepwise sucrose infiltration (10%, 20%, and 30% for 24 hours each), embedding in super cyroembedding medium (SCEM, Section-Lab, Japan), flash-freezing, and sectioning at -23°C using the Kawamoto20 method with Cryofilm type 3C(16UF) (Section-Lab). Plastic embedding included an ethanol dehydration sequence; 70% for seven days, 80% for seven days, 96% for 14 days, and 100% for 14 days, followed by xylene and embedding in Technovit 9100 New (Kulzer, Germany) then methyl methacrylate resin (MMA) infiltration under vacuum, and polymerization at -4°C. Sections (~500 µm) were ground and polished to 100 µm and measured with a digital micrometer.17 This method preserved the magnesium wire, which cannot be retained in cryosections. Plastic-embedded sections were stained with Giemsa (Merck, Germany)13 after 1% formic acid (Merck) pretreatment and mounted with Vitro-Clud mounting medium (R. Langenbrinck, Germany) for long-term storage. Cryosections were stained using four protocols: movat pentachrome, Giemsa, tartrate-resistant acid phosphatase (TRAP), and immunofluorescence staining. Movat pentachrome staining was conducted according to standard protocols.17 Plastic-embedded sections were stained with Giemsa following pre-etching with 1% formic acid (diluted in distilled water from 98% to 100% formic acid) (Merck) to enhance the dye.17 Stained sections of the manipulated femur were mounted using Vitro-Clud and sealed with a coverslip for preservation. TRAP staining involved incubation in a preheated (37°C) 40 mM sodium acetate buffer (pH 5) containing 10 mM disodium tartrate dihydrate and Fast Red Violet LB Salt, followed by Mayer’s haematoxylin counterstain (Carl Roth, Germany).17 Immunofluorescence staining was employed to identify osteoblasts using a primary polyclonal anti-mouse osteocalcin antibody (ALX-210-333 C100; Enzo, USA, 1:4,000) followed by a secondary Alexa Fluor 647-conjugated anti-rabbit antibody (A-31573; Thermo Fisher Scientific, USA, 1:1,000). Endothelial cells were detected via endomucin labelling (Rat anti-mouse Endomucin clone V.7C7, 14-5851-82, 0.5 mg/ml, Thermo Fisher Scientific), visualized with an Alexa Fluor 555-conjugated secondary antibody (Goat-anti-rat AF 555, ab150166, 2 mg/ml, Abcam, UK). Prior to staining, samples were permeabilized with Triton X-100, blocked with bovine serum albumin (BSA, Fraction V, IgG-free, NZ-origin, Carl Roth), and counterstained with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI, Thermo Fisher Scientific). Appropriate controls were included to validate the specificity and reliability of the staining procedure. Brightfield imaging was performed using Leica DM6 B (Leica, Germany) and Zeiss Axioskop 40 (Zeiss, Germany) microscopes. Fluorescent images were captured using a Keyence BZ-X810 (Keyence, Japan) with DAPI, TRITC, and Cy5 filters. Histological analysis Histological quantification was performed using Fiji/ImageJ with custom macros. A 3 mm central region between the distractor screws was defined as the region of interest (ROI). Tissue fractions (bone, cartilage, marrow, connective tissue) were quantified via thresholding. Osteoclasts were manually counted per mm of bone surface. Osteoblasts were quantified from skeletonized osteocalcin signals. Vessel density and area were measured in endomucin-stained sections within the same ROI. Statistical analysis Due to small sample sizes (n = 6/group), non-parametric comparisons were made using the Mann-Whitney U test. A significance threshold of α = 0.05 (two-sided) was used. Data were processed in Microsoft Excel (Microsoft, USA) and analyzed using GraphPad Prism (v. 10.1.0, GraphPad Software, USA). Results Verification of the nonunion model A murine nonunion model was established by creating a 2.5 mm critical size in the femur. Nonunion was confirmed four weeks post-surgery using in vivo micro-CT and intraoperative inspection. Among 36 defects, only two showed complete osseous bridging (mRUST ≥ 10), which was confirmed radiologically (Figure 2). The remaining 34 animals exhibited impaired healing: 15 (41.6%) had mRUST < 7 (nonunion), and 19 (52.7%) scored between 7 to 9 (probable nonunion). The mean defect gap was 0.78 mm (SD 0.18), with a mean mRUST score of 7.06 (SD 0.6). These results confirm the reproducibility and reliability of the nonunion model for subsequent therapeutic evaluation. a) Boxplot of defect gap sizes (mm) four weeks post-induction (micro-CT). Median and interquartile range are highlighted. b) Modified radiographic union score for tibial fractures (mRUST) scores at four weeks. The red dashed line (mRUST = 6) indicates nonunion, while the black dashed line (mRUST = 10) marks complete bridging. Most defects remain nonunion, confirming the model. Radiological quantification of bone in the defect gap At eight weeks, the WE43 group showed markedly larger defect gaps compared to WE43+, which demonstrated near-complete closure comparable to the CTR (Figure 3). This pattern persisted at 12 weeks. Quantitative micro-CT analysis revealed that WE43 consistently exhibited lower mRUST scores, bone volume, mineral density, bone surface area, and trabecular parameters (Figure 4). The elevated trabecular number and reduced trabecular separation indicated fragmented, immature bone formation rather than effective consolidation. Although trabecular thickness was initially reduced in WE43, it reached values comparable to the other groups by 12 weeks, suggesting delayed but partial reduction. In contrast, the WE43+ group showed quantitative metrics closely approaching those of the CTR, reflecting enhanced regeneration facilitated by the addition of cancellous bone (Figure 5). This improvement likely extends beyond osteoconduction, as autologous graft material provides osteogenic progenitor cells and osteoinductive cues (e.g. mesenchymal stem cells, bone morphogenetic proteins) that synergize with magnesium’s bioactivity. a) Micro-CT-based quantification of defect gap size revealed significantly larger defects in the magnesium alloy (WE43) group compared to WE43 plus cancellous bone (WE43+) and control (CTR) groups at both eight and 12 weeks (*p < 0.05). WE43+ showed near-complete defect closure, comparable to the CTR. Notably, the apparent increase in gap size in the 12-week CTR group was due to two non-responders, which skewed the mean value. b) Modified radiographic union score for tibial fractures (mRUST) scoring indicated significantly delayed bone healing in the WE43 group relative to WE43+ and CTR at both timepoints (*p < 0.05). WE43+ achieved scores approaching the CTR, reflecting enhanced structural consolidation (*p < 0.05). Quantitative micro-CT analysis of bone healing parameters at eight and 12 weeks in magnesium alloy (WE43), WE43 plus cancellous bone (WE43+), and control (CTR) groups. Trabecular parameters, including trabecular thickness (Tb.Th), trabecular number (Tb.Nb), and trabecular separation (Tb.Sp), indicated inferior bone microarchitecture in WE43, with WE43+ showing comparable values to the CTR at 12 weeks (*p < 0.05). BMD, bone mineral density; BS, bone surface; BV, bone volume; BV/TV, bone volume per tissue volume; TV, tissue volume. a) Histological sections of magnesium alloy (WE43), WE43 plus cancellous bone (WE43+), and control (CTR) groups at eight and 12 weeks. Movat’s staining (left): bone appears yellow, cartilage blue-green, connective tissue green-yellow, and bone marrow brown. Giemsa staining (right): bone is stained pink; connective tissue ranges from blue to violet. The CTR group showed superior bone bridging. Scale bars: 1 mm. b) Box plots showing fractions of bone, bone marrow, connective tissue, and cartilage in WE43, WE43+, and CTR groups at eight and 12 weeks. WE43 showed the lowest bone formation and highest connective tissue content; WE43+ showed intermediate values. Bone marrow was significantly lower in WE43. Cartilage was minimal in all groups. c) Region of interest definition and tissue segmentation example. Magnesium (grey) was manually segmented as void; all other tissues were thresholded. Tissue fractions were calculated from the total tissue area (bone + marrow + connective tissue + cartilage), excluding voids and magnesium. Bone marrow was only quantified if enclosed by new bone. Temporal analysis showed a significant increase in bone surface area in WE43 (p = 0.041), suggesting limited remodelling. In WE43+, trabecular thickness increased over time (p = 0.002), indicating ongoing maturation. No significant changes were observed in the CTR group, suggesting that bone formation there had already plateaued. Tissue composition in the defect region At eight weeks, Movat-stained cryosections revealed the highest bone formation in the CTR group, followed by WE43+ and WE43. Although WE43+ contained more bone than WE43, this difference did not reach statistical significance (Figure 6). Bone marrow development was lowest in WE43 but comparable between WE43+ and the CTR. Connective tissue content was most abundant in WE43, exceeding CTR levels, while WE43+ displayed intermediate values. Cartilage was scarcely present in WE43 and WE43+ and absent in controls. At 12 weeks, these trends persisted. The CTR group showed the highest bone retention, with WE43+ maintaining more bone than WE43. Bone marrow formation stayed limited in WE43, but reached near-CTR levels in WE43+. Connective tissue content again peaked in WE43, suggesting continued fibrotic remodelling, whereas WE43+ retained a more balanced composition. Cartilage remained minimal across all groups. These findings indicate distinct regenerative profiles, with WE43 favouring fibrotic healing and WE43+ reflecting a more balanced trajectory. Representative tartrate resistant acid phosphatase (TRAP)-stained sections (left) show osteoclast distribution; osteocalcin (OCN) fluorescence labelling (right, red = OCN, blue = 4′,6-diamidino-2-phenylindole (DAPI)) indicates osteoblast presence across magnesium alloy (WE43), WE43 plus cancellous bone (WE43+), and control (CTR) groups. Box plots summarize osteoclast and osteoblast densities relative to bone surface (BS), illustrating group-specific difference. ObN/BS, osteoblast number per bone surface; OcN/BS, osteoclast number per bone surface. Bone remodelling activity Bone remodelling was assessed at eight weeks by quantifying osteoclasts and osteoblasts within a standardized 3 mm ROI (Figure 7). Osteoclast density was highest in WE43, significantly exceeding CTR, while WE43+ showed intermediate values. Osteoblast density was highest in WE43+, followed by CTR and WE43, though differences were not statistically significant. These findings suggest increased bone resorption in WE43, and enhanced osteoblastic activity in WE43+, despite comparable osteoblast numbers. Micro-CT reconstructions of fracture healing at eight and 12 weeks (W) in magnesium alloy (WE43), WE43 plus cancellous bone (WE43+), and control (CTR) groups. The WE43 group exhibited limited bone bridging and the WE43+ group showed improved bone continuity. The CTR group demonstrated the most advanced healing with greater bone formation and structural integrity. Microvascularization At eight weeks, immunohistochemical analysis revealed the lowest vessel density and smallest vessel area in the WE43 group (Figure 8). In contrast, WE43+ exhibited significantly higher vessel density, with values comparable to the CTR. Quantification was normalized to the total defect area and included all tissue types. The data indicated impaired angiogenesis in WE43, whereas WE43+ supported vascularization similar to native bone healing. Immunohistochemical analysis of vascularization in the defect region after eight weeks (red = endomucin, blue = 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI)). The magnesium alloy (WE43) group exhibited significantly lower vessel density and reduced vessel cross-sectional area compared to both WE43 plus cancellous bone (WE43+) and the control (CTR) group. While WE43+ and the CTR group showed no significant differences. Nb, number. Collectively, these results demonstrate that WE43 alone does not support effective regeneration, likely due to rapid degradation and the local accumulation of byproducts. In contrast, combining WE43 with cancellous bone enhanced bone formation, remodelling, and vascularization, validating the importance of biological context in magnesium-based therapy. Discussion This study provides novel insight into the regenerative potential of WE43 magnesium alloy in a validated murine nonunion model. Despite extensive research on magnesium alloys in the context of standard fracture healing,17 their application in established nonunions remains poorly understood. To the best of our knowledge, this is the first study to evaluate magnesium-based implants in a validated preclinical model of atrophic nonunion, thus addressing a critical translational gap. The findings demonstrate that the administration of magnesium alone, specifically WE43, failed to induce effective bone regeneration. WE43 persistently showed wide defect gaps, low mRUST scores, and increased fibrotic tissue indicating impaired healing, suggesting that rapid degradation of the WE43 alloy and the associated local tissue response may have contributed to impaired bone regeneration.21-23 Conversely, combining WE43 with cancellous bone (WE43+) markedly enhanced early bone formation and vascularization at eight weeks, approaching outcomes observed in the CTR. The results suggest that magnesium does not compromise the graft’s osteogenic capacity and may promote early-stage healing, thereby corroborating earlier reports of magnesium’s pro-osteogenic and pro-angiogenic properties.18,19,24 By 12 weeks, regenerative progression in the WE43+ group had plateaued, while CTR achieved the most advanced consolidation. Although both WE43 and WE43+ did not outperform autologous bone grafting, the early benefit observed in WE43+ reflects a meaningful interaction between the implant and the osteogenic environment. Autologous cancellous bone remains the gold standard for critical-sized defects due to its osteogenic, osteoinductive, and osteoconductive characteristics,21,22 however its use is limited by graft availability and donor-site morbidity. In this context, magnesium may serve as a bioactive adjunct, potentially reducing graft volume or improving healing in large defects or compromised hosts. These findings hold translational relevance for reconstructive and trauma surgery, where nonunion and critical-sized defects remain major challenges. Magnesium-based materials such as WE43 may act as bioactive adjuncts to enhance graft performance and early vascularization, particularly in maxillofacial and orthopaedic applications. Successful clinical translation will depend on achieving controlled degradation and validating long-term safety in large animal and human studies. The stagnation observed in WE43+ between eight and 12 weeks may result from a mismatch between the kinetics of implant degradation and the evolving biological requirements of bone repair. The rapid corrosion of solid WE43 likely caused the local accumulation of degradation products, such as hydrogen gas and magnesium ions, which exceeded the tissue’s regenerative capacity. This could have created a hostile microenvironment during remodelling. While there was no structural progress observed in WE43 from eight to 12 weeks, WE43+ exhibited increased trabecular thickness, indicating continued, albeit limited, remodelling activity. Histological analysis revealed that WE43 did not induce the typical signs of a foreign body reaction, such as dense inflammatory infiltrates or fibrous encapsulation. However, the increased connective tissue observed in the WE43 group suggests impaired osteogenesis, which may have been triggered by accelerated local degradation. Interestingly, bone formation occurred in the WE43+ group despite presumed exposure to the same degradation products, suggesting that graft-derived cells and signals may buffer or compensate for adverse effects. This finding lends weight to the idea that the regenerative performance of magnesium is highly context-dependent. Biomaterial-based scaffolds, particularly those made from hydroxyapatite or synthetic substitutes, are recognized for their ability to promote bone regeneration in nonunion models by providing structural support for cellular processes.23,25,26 Magnesium builds on this concept by providing both mechanical support and biological stimulation. However, using solid WE43 wires may be suboptimal. High ion concentrations and gas formation can hinder tissue integration.23,27,28 Micro-CT and histological analyses suggest that excessive degradation limits long-term regenerative potential. Quantitative vessel analysis confirmed reduced vascularization in the WE43 group, whereas WE43+ restored angiogenesis to near CTR levels. However, since quantification was normalized to total defect area, conclusions regarding bone-specific vascularization are limited. To address these challenges, alternative delivery strategies are necessary. Incorporating finely milled magnesium particles into cancellous bone grafts could facilitate gradual ion release, preserving the biological benefits while reducing byproduct accumulation. Micro-CT data from the present study showed stagnation in WE43 from weeks eight to 12, while indicating ongoing bone maturation in WE43+. This finding reinforces the importance of degradation control. Surface modifications, including plasma electrolytic oxidation, have been clinically validated as a means of reducing magnesium degradation. This process has the potential to improve performance in nonunion applications.13,20,24 Exploring hybrid approaches, such as combining magnesium with hydrogels, platelet-rich fibrin, or bone chips, is also recommended. As magnesium and its alloy WE43 have not previously been evaluated in a nonunion model, the implant configuration and defect-filling strategy employed here may not yet reflect the most effective therapeutic approach. Completely filling the defect cavity may have restricted the physiological migration of osteogenic and inflammatory cells, which is essential for coordinated healing. Improving the architecture of the defect and the spatial organization of the implant-graft interface could enhance regenerative outcomes. Despite the robust nonunion phenotype confirmed at four weeks, there was occasional spontaneous bridging. This was probably due to the regenerative capacity of young, immunocompetent mice, combined with the use of autologous graft material. Notably, the WE43+ group exhibited early healing comparable to that observed in the CTR group. This suggests that magnesium did not impede regeneration and may even have facilitated early biological processes. This phenomenon may be more evident in less favourable biological conditions, such as in aged or metabolically impaired animals, where spontaneous healing is less probable. Such scenarios offer a more accurate reflection of clinical reality and signify an important direction for future studies. This study has several limitations. The lack of direct quantification of magnesium degradation limits mechanistic interpretation. The exclusive use of young, female, immunocompetent mice may overestimate regenerative potential and reduce clinical generalizability. Finally, the small sample size and absence of mechanical testing restrict conclusions on implant stability. Magnesium-based biomaterials show potential for nonunion treatment, particularly when combined with osteogenic and angiogenic components. While the magnesium alloy (WE43) alone impaired bone regeneration, its combination with cancellous bone (WE43+) improved bone formation and vascularization. These findings highlight the importance of controlling degradation dynamics and optimizing implant–tissue interactions. To support clinical translation, future studies should refine magnesium composition, develop strategies for controlled degradation, and assess long-term outcomes in large animal models.