Development and evaluation of a modified injectable calcium phosphate cement with enhanced injectability and mechanical stability for bone defect repair Authors: Xiaoli Zhou, Dong Wei, Zepei Zhang, Yinghua Liu, Jun Miao, Zhiyong Qian Published: 2026 ABSTRACT This study aimed to optimize an injectable carbonated hydroxyapatite cement for metaphyseal bone defect filling and to evaluate its biomechanical performance in distal radius fracture fixation. Six food- or pharmaceutical-grade additives, including sodium alginate, glycerol, astragalus gum, xanthan gum, citric acid, and hydroxypropyl methylcellulose (HPMC), were screened to improve injectability while maintaining setting behavior and mechanical strength. Among them, 2% HPMC provided the best overall performance, increasing injectability to approximately 95% without markedly affecting setting time or compressive strength. The optimized injectable carbonated hydroxyapatite cement (ICHA) was then tested in a standardized human cadaveric distal radius fracture–defect model using three fixation strategies: K-wire fixation alone, ICHA cement alone, and combined ICHA + K-wire fixation. Within a 10° torsion range, the ICHA and ICHA + K-wire groups showed higher torsional stiffness and maximum torque than the K-wire-only group, although peak torque occurred at smaller torsion angles. Under physiological compression, the groups showed comparable stability, whereas ICHA-containing groups showed greater resistance under higher compression. These findings suggest that 2% HPMC-modified ICHA may serve as an injectable metaphyseal void filler and structural adjunct to conventional K-wire fixation.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12891-026-10234-z. Introduction Bone has a remarkable regenerative capacity; however, this intrinsic repair potential is often insufficient in large segmental or metaphyseal defects caused by severe trauma, tumor resection, osteoporotic fractures, or revision surgery. Autologous bone grafting remains the clinical gold standard for bone reconstruction because of its biocompatibility and osteogenic potential, but its use is limited by donor-site morbidity, restricted graft availability, and difficulty in achieving anatomical conformity [1]. Critical-sized bone defects may lead to nonunion, deformity, and functional impairment, particularly in metaphyseal regions where cancellous bone loss, mechanical instability, and local inflammatory responses may further compromise bone regeneration [2]. Therefore, injectable bone substitutes capable of providing both osteoconductive guidance and immediate structural support are highly desirable. Polymethylmethacrylate (PMMA) acrylic bone cement has been widely used as a defect-filling and load-transfer material in orthopedic procedures. However, PMMA should be regarded primarily as a filler rather than a true adhesive [3]. Its fixation and load-transfer capacity depend mainly on mechanical interdigitation with trabecular or cancellous bone and interlocking with implant surfaces, rather than chemical bonding to bone [4]. Although PMMA provides immediate mechanical support, concerns related to brittleness, non-degradability, thermal effects, and mechanical mismatch have encouraged the development of more biologically compatible alternatives. Injectable calcium phosphate cement (CPC) has emerged as a promising bone substitute because of its biocompatibility, osteoconductivity, degradability, and ability to self-set under mild conditions. Bioactive ion modification, such as magnesium or strontium incorporation, has been reported to improve the mechanical and biological performance of CPCs [5]. However, conventional CPCs still suffer from poor injectability and insufficient mechanical strength. During injection, CPC pastes are prone to phase separation, known as the “filter-pressing effect”, resulting in poor handling properties and inconsistent composition [6]. In addition, the inherent brittleness and low flexural strength of set CPCs restrict their use in load-bearing applications [7]. Polymeric rheology modifiers, including hydroxypropyl methylcellulose, sodium alginate, and hyaluronic acid, have been explored to improve paste cohesion and injectability [8–10], but they may also affect setting kinetics, degradation behavior, and microstructure. Thus, optimizing CPC injectability while preserving mechanical stability remains an important materials challenge. Distal radius fractures (DRFs), especially in osteoporotic bone, represent a typical metaphyseal defect scenario in which cancellous bone voids may contribute to mechanical instability after fracture reduction. In this context, injectable CPC is not intended to serve as a primary fixation device, but rather as a subchondral void filler providing structural support beneath the articular surface. In this study, we employed a problem-oriented additive screening strategy to improve the rheological stability of carbonated hydroxyapatite cement. Six candidate modifiers were evaluated with emphasis on injectability and setting behavior, and hydroxypropyl methylcellulose (HPMC) was identified as the optimal component. The resulting injectable carbonated hydroxyapatite cement was characterized in terms of injectability, setting time, and compressive strength, and its biomechanical role as a metaphyseal defect-filling support material was further assessed in a cadaveric distal radius fracture model. Materials and methods Materials Carbonated hydroxyapatite cement was modified by dissolving HPMC in the CHA setting solution to obtain a liquid phase containing 2% HPMC. The modified liquid phase was then mixed with the solid phase at a liquid-to-powder ratio of 0.4 mL/g to prepare injectable carbonated hydroxyapatite (ICHA) cement. The solid phase consisted of 13% calcium carbonate, 84% α-tricalcium phosphate (α-TCP), and 3% monocalcium phosphate (MCP), while the liquid phase contained 0.2 M trisodium phosphate and 2% HPMC. Compared with the unmodified CHA formulation, the modified ICHA cement showed markedly improved injectability. The six additives were selected because they represent commonly used food- or pharmaceutical-grade auxiliary agents with potential effects on the rheological and handling properties of calcium phosphate cement. Sodium alginate, astragalus gum, xanthan gum, and HPMC were selected as polymer- or polysaccharide-based thickening and rheology-modifying agents; glycerol was selected as a small-molecule humectant and plasticizer that may improve paste lubricity; and citric acid was selected as a low-molecular-weight organic acid and chelating agent that may improve fluidity but may also affect the cement setting reaction. The purpose of screening these additives was to improve the poor injectability and phase separation tendency of conventional CHA cement while minimizing adverse effects on setting time and mechanical strength. The reagents used for additive screening included sodium alginate, glycerol, astragalus gum, xanthan gum, citric acid, and HPMC. Sodium alginate was purchased from Beijing Xudong Chemical Factory and was of analytical grade. Glycerol was purchased from Beijing Yanjing Pharmaceutical Factory and was of analytical grade. astragalus gum powder was purchased from Beijing Chemical Reagent Company and was of analytical grade. Xanthan gum was purchased from Beijing Grain Science Institute and was of food grade. Citric acid was purchased from Beijing Chemical Reagent Company and was of analytical grade. HPMC was purchased from Taian Ruitai Cellulose Co., Ltd. and was of pharmaceutical grade. The main materials and equipment used in this study included ICHA cement, 24 human distal radius specimens, and a biomechanical testing system. Kirschner wires with a diameter of 1.5 mm were used for fixation. Bone mineral density was measured using dual-energy X-ray absorptiometry with a LUNAR DPX-L X-ray bone densitometer. Additional instruments included a 2.5 mL disposable syringe with an inner diameter of 8.5 mm and a nozzle size of 10 mm × 2 mm, an electronic balance with an accuracy of 0.01 g, a customized stainless-steel fixture, and an electric saw for specimen preparation. PMMA bone cement was used for comparative analysis. The human distal radius specimens used in this study were obtained from the Department of Anatomy of a medical university. All specimens were cadaveric bone tissues originally donated and preserved for medical education and scientific research. The specimens had been fully anonymized, and no identifiable donor information was available to the investigators. This study involved only human cadaveric specimens and did not involve living participants or patient-related clinical data. Grouping After measurement of bone mineral density (BMD) at the distal radial metaphysis, the 24 distal radius specimens were randomly divided into four groups, with six specimens in each group. Analysis of variance showed no statistically significant difference in BMD among the groups (Table 1). Each specimen was transversely osteotomized at 10 cm proximal to the distal articular surface, and the distal 10-cm segment was retained for mechanical testing. The overall workflow of specimen preparation, group allocation, fracture–defect model establishment, and biomechanical testing is summarized in Supplementary Figure S1.Table 1Bone mineral density measurement(n = 6, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})GroupABCDBMD0.627 ± 0.1080.656 ± 0.1090.689 ± 0.0550.672 ± 0.059ANOVA: F = 0.557, P = 0.650 Table 1Bone mineral density measurement(n = 6, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})GroupABCDBMD0.627 ± 0.1080.656 ± 0.1090.689 ± 0.0550.672 ± 0.059ANOVA: F = 0.557, P = 0.650 Bone mineral density measurement(n = 6, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document}) ANOVA: F = 0.557, P = 0.650 The experiment consisted of two parts (Table 2). The first part was the torsion test, which included three groups: fixation with ICHA cement, fixation with Kirschner wires, and combined fixation with ICHA cement and Kirschner wires. The second part was the compression test, which also included three groups: fixation with Kirschner wires after removal of residual bone cement, fixation with ICHA cement after removal of Kirschner wires, and combined fixation with ICHA cement and Kirschner wires.Table 2Grouping of distal radius specimensGroupnTorsionCompressionA6ICHAK wireB6K wireICHAC6ICHA + K wireD6ICHA + K wire Table 2Grouping of distal radius specimensGroupnTorsionCompressionA6ICHAK wireB6K wireICHAC6ICHA + K wireD6ICHA + K wire Grouping of distal radius specimens Preparation of distal radius fracture models Distal radius fractures are defined as trabecular bone fractures occurring within 3 cm of the radiocarpal joint surface. In this study, a distal radius fracture with bone defect model was created 2.5 cm proximal to the wrist joint surface [11–17]. Using the dorsal edge of the distal articular surface as reference, a point 2.5 cm proximally was marked. Two transverse lines were drawn at 0.5 cm distal and proximal to this point, perpendicular to the long axis of the radius. Using an electric saw, the dorsal cortical bone between these two lines was cut to form a 1 cm-wide wedge-shaped osteotomy perpendicular to the long axis, while maintaining the integrity of the palmar cortex. After removing the dorsal wedge bone segment, a bone defect was created. Then, the palmar cortex was transversely cut using a thin bone knife but kept in cortical contact, simulating a dorsal compressive fracture of the distal radius (Fig. 1A).Fig. 1Distal radius fracture model. A: Dorsal wedge osteotomy of the radius; B: Cross fixation with K-wires; C: Fixation with ICHA cement; D: Combined fixation with ICHA cement and K-wires Distal radius fracture model. A: Dorsal wedge osteotomy of the radius; B: Cross fixation with K-wires; C: Fixation with ICHA cement; D: Combined fixation with ICHA cement and K-wires Kirschner wire fixation group After fracture reduction and maintaining palmar cortical contact, two 1.5 mm Kirschner wires were used for cross fixation. One wire was inserted from the radial styloid process obliquely toward the ulnar side of the proximal shaft at a 60° angle to the long axis; the other was inserted from the ulnar corner of the distal radius obliquely toward the radial side of the proximal shaft, also at 60°, thereby stabilizing the distal fragment to the proximal shaft (Fig. 1B). ICHA cement fixation group After fracture reduction with palmar cortical contact, the liquid and powder phases of the bone cement were mixed at the specified ratio and quickly loaded into a 5 mL syringe. The cement was injected into the defect, ensuring stability of the fracture fragments without displacement. After 10 min of initial setting, the specimen was allowed to be moved. Approximately 3 mL of bone cement was used per defect (Fig. 1C). ICHA cement + Kirschner wire combined fixation group After cross-fixation of the fracture fragments with Kirschner wires, the mixed bone cement was injected into the bone defect according to the same procedure (Fig. 1D). After establishment of the distal radius fracture model and completion of fixation, X-ray imaging was performed using an F78-IIIA 500-mA conventional diagnostic X-ray radiography unit (Beijing Wandong Medical, Beijing, China). Standard anteroposterior and lateral radiographs were obtained for each distal radius specimen before biomechanical testing. The radiographs were used to confirm fracture reduction, alignment, defect filling, and the position of the K-wires and/or ICHA cement, and to verify the consistency of the fracture model and fixation configuration. All specimens were then stored at 37 °C under 100% humidity for 24 h before being subjected to biomechanical testing. Methods Evaluation of injectability The six additives described above were separately dissolved in the CHA cement liquid phase at different concentrations for injectability testing. Injectability was evaluated using the injectability coefficient (Inj%), which was defined as the percentage of cement paste extruded from the syringe relative to the initial total mass loaded. As shown in Fig. 2, a 2.5 mL syringe was fixed vertically in a stainless-steel holder mounted on the platform of a mechanical testing system. For each test, 2.0 g of cement powder was mixed with the liquid phase at a liquid-to-powder ratio of 0.4 mL/g and then loaded into the syringe. The injection test was initiated 120 s after mixing. The plunger was advanced at a constant speed of 15 mm/min until the maximum extrusion force reached 100 N. The Inj% was calculated according to the mass of cement paste extruded from the syringe. Each experiment was performed in triplicate, and the mean value was recorded.Fig. 2Evaluation method of injectability. The syringe was placed vertically on the platen of a biomechanical testing machine and pushed downward at a speed of 15 mm/min. The test was stopped when the maximum pushing force reached 100 N, and the percentage of bone cement extruded relative to the total amount was recorded Evaluation method of injectability. The syringe was placed vertically on the platen of a biomechanical testing machine and pushed downward at a speed of 15 mm/min. The test was stopped when the maximum pushing force reached 100 N, and the percentage of bone cement extruded relative to the total amount was recorded Setting time measurement According to ISO 3107 testing standards for zinc phosphate cement, a 400 g conical Vicat needle with a tip diameter of 1 mm was used. The initial setting time was defined as the time when the needle could no longer penetrate a 2 mm-thick specimen, and the final setting time was defined as the time when the needle failed to leave a visible indentation on the specimen surface. The test was conducted at 37 °C under 100% humidity. These conditions were selected to approximate the physiological environment after cement implantation. A temperature of 37 °C represents normal human body temperature, while 100% humidity simulates the moist in vivo environment. Because the setting and hardening of calcium phosphate cement depend on hydration, dissolution, and precipitation reactions, a high-humidity environment was used to prevent premature drying of the cement paste, which could otherwise affect the setting behavior of the cement. Measurement of compressive strength The cement powder and liquid were mixed at a liquid-to-powder ratio of 0.4 mL/g until homogeneous and poured into stainless-steel molds (8 mm × 12 mm). The surfaces were leveled, sealed with plates, and clamped. Specimens were cured at 37 °C and 100% humidity for 20 min, demolded, and stored under the same conditions for 1 week. Compressive strength was measured using a biomechanical testing machine at a loading rate of 0.5 mm/min. Each group contained five specimens, and the mean value was recorded. Biomechanical testing The torsion test was performed as shown in Fig. 3. Briefly, the proximal shaft of the radius was vertically secured to the upper platen of the biomechanical testing machine using a custom-made clamp. The platen was then lowered until the distal radius was inserted approximately 1.5 cm into the lower clamp, without initial contact between the clamp and the bone. At this position, axial displacement, axial load, torque, and torsion angle were all set to zero.Fig. 3Torsion test setup for biomechanical evaluation Torsion test setup for biomechanical evaluation The lower clamp was then carefully adjusted to contact the distal radius while ensuring that the torque remained zero. The screws were subsequently tightened to secure the clamp to the base. The torsion test was conducted at a loading rate of 5°/min and stopped when the torsion angle reached 10°. These conditions were selected to provide a standardized quasi-static loading protocol for evaluating the initial rotational stability of distal radius fracture fixation constructs. The distal radius is subjected to rotational loading during forearm pronation and supination; therefore, torsion testing was used to assess whether ICHA cement could provide adequate rotational stability. The relatively slow loading rate was chosen to minimize dynamic or inertial effects, and the 10° limit was selected to compare early fixation stability within a controlled angular range while avoiding excessive damage or catastrophic failure of the cadaveric specimens. A torque–angle curve was generated, and the slope of the linear region of the curve was defined as torsional rigidity. The maximum torque within 10° and its corresponding torsion angle were also recorded. A representative torque–torsion angle curve is shown in Supplementary Figure S2 to illustrate the data extraction process. For the compression test, PMMA bone cement powder and liquid monomer were mixed according to the manufacturer’s instructions. For the Zimmer PMMA bone cement used in this study, the powder and liquid monomer were mixed at the manufacturer-recommended ratio of approximately 40 g powder to 20 mL liquid monomer. Before complete setting, the PMMA cement was shaped to match the distal radial articular surface, forming customized joint-loading spheres. The concave indentations were created while the PMMA cement was still in the moldable dough phase. A smooth curved stainless-steel rod was gently pressed onto the PMMA surface to create a shallow concave indentation. This indentation was designed to match the loading platen during axial compression testing, thereby providing stable contact and reducing specimen slippage or eccentric loading. Figure 4A shows the PMMA joint-surface mold prepared to match the distal radial articular surface, and Fig. 4B shows the PMMA joint-surface mold placed on the distal radius specimen during the compression test. The proximal radius was fixed downward in the biomechanical testing machine, with the distal articular surface facing upward. The customized PMMA loading sphere was placed on the distal articular surface, and the platen was positioned in contact with the concave surface of the sphere. Four markers were attached to the palmar, dorsal, radial, and ulnar sides of the proximal and distal bone fragments. Two digital cameras were used to record the rotation of the distal fragment in the coronal and sagittal planes. The axial compression test setup is illustrated in Supplementary Figure S3. The setup included the upper loading platen, PMMA joint sphere, distal fragment, proximal fragment, lower clamp, and two cameras positioned to monitor fragment displacement and rotation during loading.Fig. 4Wrist articular surface molded with PMMA cement. A: PMMA joint-surface mold. B: PMMA joint-surface mold placed on the distal radius specimen Wrist articular surface molded with PMMA cement. A: PMMA joint-surface mold. B: PMMA joint-surface mold placed on the distal radius specimen Compression testing was performed in two stages. In the first stage, the platen was lowered at a rate of 2 mm/min until an axial load of 100 N was reached. The displacement–load curve and rotation data were recorded. In the second stage, after the load was reset, the platen was lowered at a rate of 5 mm/min until defect closure. The compressive strength at 2 mm displacement and the maximum compressive strength were recorded (Fig. 5).Fig. 5Compression test setup for biomechanical evaluation Compression test setup for biomechanical evaluation Methodological details and assumptions All procedures were standardized to reduce methodological variability. The modified calcium phosphate cement was prepared using the same liquid-to-powder ratio, mixing protocol, and testing sequence for all specimens. Injectability and setting time were evaluated using syringe extrusion and Vicat needle-based methods, respectively. For biomechanical testing, a standardized human cadaveric distal radius fracture–defect model was established with consistent osteotomy site, defect size, cement filling, fixation method, specimen orientation, and loading protocol. Radiographs were used to confirm cement placement, defect filling, and fixation morphology. Torsion testing was performed under quasi-static loading to assess initial rotational stability. The main assumption was that the cadaveric model and quasi-static protocol could reasonably represent the local anatomy and immediate mechanical environment of distal radius fracture fixation. Statistical analysis All data were analyzed using SPSS software version 11.0. One-way analysis of variance (ANOVA) was used to compare parameters among groups, including torsional rigidity, maximum torque, maximum torsion angle, compressive stiffness, fragment rotation angle, compressive strength at 2 mm displacement, and maximum compressive strength. For post hoc pairwise comparisons, the Least Significant Difference (LSD) test was used when the assumption of homogeneity of variance was satisfied. The LSD test compares differences between group means based on the pooled error variance from the ANOVA model. When variances were unequal, Tamhane’s T2 test was applied. Tamhane’s T2 is a conservative multiple-comparison method that does not require equal variances and is suitable for unequal variance conditions or small sample sizes. A value of P < 0.05 was considered statistically significant. Results Injectability test results In the absence of additives, CHA cement exhibited an injectability of 70%. The addition of sodium alginate, glycerol, or astragalus gum led to a gradual decrease in injectability with increasing concentration, whereas appropriate amounts of xanthan gum, citric acid, or hydroxypropyl methylcellulose significantly enhanced injectability. For xanthan gum, 0.2% increased injectability to 80%, but concentrations above 1% made the cement too viscous to inject. Citric acid at 0.5% rendered the mixture too dry, while concentrations above 2% improved injectability up to 95%. Hydroxypropyl methylcellulose began to improve injectability above 1%, reaching 95% at 2%, though further increases slightly reduced injectability due to increased viscosity, which remained above 90% (Fig. 6A-F).Fig. 6Injectability coefficients of CHA cements containing different concentrations of additives. (A) Sodium alginate; (B) Glycerol; (C) Astragalus gum; (D) Xanthan gum; (E) Citric acid; and (F) Hydroxypropyl methylcellulose (HPMC) Injectability coefficients of CHA cements containing different concentrations of additives. (A) Sodium alginate; (B) Glycerol; (C) Astragalus gum; (D) Xanthan gum; (E) Citric acid; and (F) Hydroxypropyl methylcellulose (HPMC) Setting time test results With increasing concentrations of sodium alginate, glycerol, astragalus gum, and xanthan gum, the setting time of CHA cement was slightly shortened. In Fig. 7, TI represents the initial setting time, and TF represents the final setting time. In contrast, citric acid progressively prolonged the setting time as its concentration increased, and at concentrations above 3%, the final setting time exceeded 60 min. HPMC had no significant effect on the setting time; across the tested concentrations, the initial setting time remained around 9 min and the final setting time around 15 min (Fig. 7A-F).Fig. 7Setting times of CHA cements containing different concentrations of additives. TI represents the initial setting time, and TF represents the final setting time. (A) Sodium alginate; (B) Glycerol; (C) Astragalus gum; (D) Xanthan gum; (E) Citric acid; and (F) Hydroxypropyl methylcellulose (HPMC). TI indicates initial setting time, and TF indicates final setting time Setting times of CHA cements containing different concentrations of additives. TI represents the initial setting time, and TF represents the final setting time. (A) Sodium alginate; (B) Glycerol; (C) Astragalus gum; (D) Xanthan gum; (E) Citric acid; and (F) Hydroxypropyl methylcellulose (HPMC). TI indicates initial setting time, and TF indicates final setting time Compressive strength test results Based on the results of injectability and setting time tests, HPMC was preliminarily selected as an additive to improve the injectability of CHA cement. The effect of different HPMC concentrations on the compressive strength of CHA cement was then evaluated. The addition of HPMC at varying concentrations had minimal impact on cement strength. At 1%, compressive strength showed a slight increase, followed by a minor decrease at higher concentrations, but remained above 30 MPa (Fig. 8).Fig. 8Compressive strength of cements containing different concentrations of HPMC Compressive strength of cements containing different concentrations of HPMC Torsion test results Both the ICHA cement fixation group and the ICHA + Kirschner wire combined fixation group demonstrated markedly greater torsional rigidity and maximum torque compared with the Kirschner wire–only group. In contrast, the maximum torsion angle corresponding to the peak torque was smaller in the ICHA-containing groups, indicating that these fixations provided higher stiffness but lower angular tolerance. No significant mechanical differences were observed between the ICHA cement group and the combined fixation group (Tables 3 and 4).Table 3Torsional stiffness of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})GroupnRigidity (N⋅m/Deg)K wire60.03814 ± 0.01566aICHA60.2657 ± 0.06352bICHA + K wire60.2421 ± 0.05289cb vs. a: P = 0.000; c vs. a: P = 0.000; c vs. b: P = 0.413Table 4Maximum torque and torsion angle of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})GroupnMaximum torque (N⋅m)Angle (Deg)K wire60.4086 ± 0.2033a9.5733 ± 0.8781dICHA60.9516 ± 0.2760b4.3452 ± 1.7940eICHA + K wire60.9285 ± 0.2445c5.0302 ± 1.6318fb vs. a: P = 0.002; c vs. a: P = 0.002; c vs. b: P = 0.872; e vs. d: P = 0.000; f vs. d: P = 0.000; f vs. e: P = 0.438 Table 3Torsional stiffness of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})GroupnRigidity (N⋅m/Deg)K wire60.03814 ± 0.01566aICHA60.2657 ± 0.06352bICHA + K wire60.2421 ± 0.05289cb vs. a: P = 0.000; c vs. a: P = 0.000; c vs. b: P = 0.413 Torsional stiffness of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document}) b vs. a: P = 0.000; c vs. a: P = 0.000; c vs. b: P = 0.413 Table 4Maximum torque and torsion angle of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})GroupnMaximum torque (N⋅m)Angle (Deg)K wire60.4086 ± 0.2033a9.5733 ± 0.8781dICHA60.9516 ± 0.2760b4.3452 ± 1.7940eICHA + K wire60.9285 ± 0.2445c5.0302 ± 1.6318fb vs. a: P = 0.002; c vs. a: P = 0.002; c vs. b: P = 0.872; e vs. d: P = 0.000; f vs. d: P = 0.000; f vs. e: P = 0.438 Maximum torque and torsion angle of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document}) b vs. a: P = 0.002; c vs. a: P = 0.002; c vs. b: P = 0.872; e vs. d: P = 0.000; f vs. d: P = 0.000; f vs. e: P = 0.438 Compression test results A load of 100 N represents the physiological pressure exerted on the distal radius articular surface under normal wrist conditions. At this load, the compressive stiffness and rotation angles of the distal bone fragment in both the coronal (radial-ulnar) and sagittal (palmar-dorsal) planes were comparable among the groups, indicating that ICHA cement provided fixation strength similar to that of Kirschner wires (Tables 5 and 6). When the applied load exceeded 100 N, specimens fixed with ICHA cement-either alone or in combination with Kirschner wires-demonstrated greater resistance to compression compared with those fixed solely with Kirschner wires (Table 7).Table 5Compressive stiffness of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})GroupnRigidity(N⋅m/mm)K wire6416.5443 ± 304.5180aICHA6418.3789 ± 90.2952bICHA + K wire6534.7798 ± 328.4943cANOVA: P = 0.680Table 6Rotation angle of the distal radial fragment(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})FixationnSagittal (deg)Coronal (deg)K wire62.20 ± 1.131.26 ± 0.83ICHA62.14 ± 2.061.25 ± 0.60ICHA + K wire62.46 ± 0.461.93 ± 1.28ANOVA: Coronal: P = 0.384; Sagittal: P = 0.916Table 7Compressive strength of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})GroupnCS of 2 mm displacement (N)Maximum CS (N)K wire6115.5514 ± 86.5997a394.8335 ± 95.1180dICHA6375.3478 ± 138.5467b871.1942 ± 142.3753eICHA + K wire6386.2538 ± 168.5871c943.6239 ± 356.6995fb vs. a: P = 0.005; c vs. a: P = 0.004; c vs. b: P = 0.891; e vs. d: P = 0.003; f vs. d: P = 0.001; f vs. e: P = 0.591 Table 5Compressive stiffness of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})GroupnRigidity(N⋅m/mm)K wire6416.5443 ± 304.5180aICHA6418.3789 ± 90.2952bICHA + K wire6534.7798 ± 328.4943cANOVA: P = 0.680 Compressive stiffness of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document}) ANOVA: P = 0.680 Table 6Rotation angle of the distal radial fragment(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})FixationnSagittal (deg)Coronal (deg)K wire62.20 ± 1.131.26 ± 0.83ICHA62.14 ± 2.061.25 ± 0.60ICHA + K wire62.46 ± 0.461.93 ± 1.28ANOVA: Coronal: P = 0.384; Sagittal: P = 0.916 Rotation angle of the distal radial fragment(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document}) ANOVA: Coronal: P = 0.384; Sagittal: P = 0.916 Table 7Compressive strength of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document})GroupnCS of 2 mm displacement (N)Maximum CS (N)K wire6115.5514 ± 86.5997a394.8335 ± 95.1180dICHA6375.3478 ± 138.5467b871.1942 ± 142.3753eICHA + K wire6386.2538 ± 168.5871c943.6239 ± 356.6995fb vs. a: P = 0.005; c vs. a: P = 0.004; c vs. b: P = 0.891; e vs. d: P = 0.003; f vs. d: P = 0.001; f vs. e: P = 0.591 Compressive strength of the radius(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\bar X \pm S$$\end{document}) b vs. a: P = 0.005; c vs. a: P = 0.004; c vs. b: P = 0.891; e vs. d: P = 0.003; f vs. d: P = 0.001; f vs. e: P = 0.591 Discussion Commonly used methods for evaluating the injectability of calcium phosphate cements (CPCs) include the injectability coefficient, pushing force, and injection pressure. The injectability coefficient reflects the proportion of cement paste extruded from the syringe relative to the initial amount loaded [18–20], whereas pushing force and injection pressure represent the mechanical resistance during extrusion [21, 22]. Because of its simplicity and direct relevance to clinical handling, the injectability coefficient was used in the present study. Injectability is influenced by several factors, particularly the liquid-to-powder ratio. Although increasing the liquid content can improve paste flowability, excessive liquid may compromise setting behavior and mechanical strength [23]. Previous studies have shown that CPC handling properties can be improved by adjusting phase composition or incorporating inorganic or organic modifiers. Khairoun et al. [24] reported that modifying the proportions of CaCO₃ or Na₂HPO₄ improved handling without adversely affecting strength. In the present study, the basic CHA cement composition was maintained, and six food- or pharmaceutical-grade additives were screened. Among them, HPMC, xanthan gum, and citric acid markedly improved injectability, whereas astragalus gum, glycerol, and sodium alginate showed limited effects. The effects of these additives appeared to be closely related to their rheological properties and interactions with the cement system. Xanthan gum improved injectability at a low concentration of 0.2%, which is consistent with the findings of Van Landuyt [25]. However, higher concentrations increased viscosity excessively and reduced handling performance. Citric acid improved injectability at concentrations above 2%, but also prolonged the setting time, possibly by lowering the pH and adsorbing onto reactive particles, thereby delaying hydroxyapatite formation. By contrast, HPMC improved paste cohesion and flowability without markedly extending the setting time, consistent with the findings of Cherng [26] This effect may be attributed to the alignment of polymer chains during extrusion, which facilitates inter-chain sliding and improves paste injectability. Therefore, HPMC was selected as the optimal modifier for subsequent preparation of ICHA cement. Management of distal radius fractures depends on fracture morphology, bone quality, and functional demands. Treatment options include closed reduction and casting, limited open reduction with external fixation, and internal fixation. Stable fixation and accurate reduction are essential for maintaining fracture alignment. In comminuted or osteoporotic fractures, immobilization alone may be insufficient, and poor bone quality can increase the risk of secondary displacement [27]. Bone grafts and PMMA cement have been used to fill metaphyseal defects, transmit load, and improve structural stability [28, 29]. However, PMMA fixation depends largely on cement penetration and mechanical interlocking at the cement–bone interface [30]. Insufficient penetration may weaken mechanical interdigitation and reduce load-transfer efficiency, whereas uncontrolled flow may increase the risk of cement leakage. In addition, PMMA is relatively brittle, and its fracture and fatigue-related properties may affect crack initiation, cement mantle integrity, and long-term interface stability under repeated loading [31]. Cement distribution and mantle thickness are also important factors affecting mechanical stability. In cemented fixation, an overly thin PMMA mantle may increase stress concentration and cracking risk [32], whereas an excessively thick mantle may alter load transfer and increase thermal effects during polymerization [33]. Although PMMA mantle thickness was not optimized in the present study, these considerations highlight the importance of controlling cement distribution and defect-filling geometry in injectable bone cement applications. Compared with autologous bone grafts and PMMA cement, CPCs offer several advantages, including good biocompatibility, low exothermic reaction, in situ setting, and osteoconductivity. These characteristics make CPCs attractive materials for metaphyseal defect filling. Previous studies have shown that 2% HPMC-modified injectable carbonated hydroxyapatite cement can set under physiologically relevant moist conditions without marked heat generation, exhibits good resistance to washout in an aqueous environment, and achieves compressive strength comparable to that of cancellous bone. In vitro and in vivo biocompatibility evaluations, including bone marrow stromal cell culture, rabbit muscle implantation, hemolysis testing, and pyrogen testing, further support its potential applicability under physiologically relevant conditions [34] Clinical studies have also supported the use of CPCs in distal radius fractures. Sanchez-Sotelo [35] reported better clinical outcomes in patients treated with CPC than in those treated with plaster fixation alone. Zimmermann [36] showed that CPC combined with Kirschner wire fixation maintained fracture reduction more effectively than pinning alone and improved wrist motion and grip strength after two years. These findings suggest that CPCs may serve as useful adjuncts for maintaining metaphyseal support after fracture reduction. Under physiological conditions, the distal radius is exposed to axial compressive forces transmitted from the carpus and torsional stresses generated during forearm pronation and supination. Therefore, both compression and torsion tests were used in this study to evaluate the mechanical performance of ICHA cement. In torsion testing, ICHA cement provided stronger fixation than Kirschner wires within a limited angular range, but failure occurred beyond this range, indicating that its effective fixation range may be narrower than that of wire fixation. In compression testing, ICHA cement and Kirschner wire fixation showed comparable stiffness under a physiological load of 100 N, while ICHA cement exhibited higher resistance under greater compressive loading. These results suggest that ICHA cement can provide effective early mechanical support when used as a metaphyseal defect-filling material. The novelty of this study is that ICHA cement was not only evaluated as an injectable CHA-based bone substitute, but was also directly compared with K-wire fixation and combined ICHA + K-wire fixation in a human cadaveric distal radius fracture–defect model. The results showed that ICHA cement provided higher torsional stiffness and maximum torque than K-wire fixation alone within the tested angular range, while showing comparable stability under physiological compression and greater resistance under higher compression. The combined ICHA + K-wire construct also showed favorable biomechanical performance. These findings suggest that ICHA cement may function as a metaphyseal defect filler and structural support material that complements conventional K-wire fixation. Limitations and future perspectives The present study mainly evaluated the initial injectability, setting behavior, compressive strength, and ex vivo biomechanical performance of ICHA cement under controlled conditions. Long-term fatigue behavior, in vivo degradation, bone remodeling, cement porosity, viscosity–time profiles, and patient-specific loading conditions were not directly assessed. Future studies should combine long-term mechanical testing, in vivo validation, and coupled thermo-chemo-mechanical-fatigue modeling to simulate cement setting, heat generation, shrinkage, interfacial stress transfer, fatigue crack initiation, and mechanical degradation. In addition, AI-assisted optimization and digital twin frameworks incorporating patient-specific imaging, bone mineral density, defect morphology, cement distribution, and loading conditions may help optimize cement formulation and predict long-term fixation stability and failure risk. Conclusion In conclusion, the incorporation of suitable additives markedly improved the injectability of CHA cement. Among the formulations tested, 2% HPMC achieved the best balance between injectability, setting behavior, and mechanical stability. In the cadaveric distal radius fracture model, ICHA cement provided higher torsional and compressive fixation strength than K-wire fixation under the tested conditions. Moreover, the combined use of ICHA cement and K-wires produced the most favorable biomechanical performance. These findings suggest that ICHA cement may serve as a bioactive injectable metaphyseal void filler and structural support material to complement conventional fixation in distal radius fractures. Supplementary Information Supplementary Material 1: Figure S1. Schematic workflow of specimen acquisition, preparation, group allocation, model creation, and biomechanical testing. Supplementary Material 2: Figure S2. Representative torque–torsion angle curve of the distal radius fixation construct during quasi-static torsional loading. Supplementary Material 3: Figure S3. Schematic illustration of the axial compression test setup. The figure shows the upper loading platen, PMMA joint sphere, distal fragment, proximal fragment, lower clamp, and the positions of the two cameras used to monitor displacement and rotation. The loading protocol included an initial loading phase at 2 mm/min up to 100 N to record load–displacement and rotation, followed by reloading at 5 mm/min until defect closure to measure mechanical strength. Supplementary Material 1: Figure S1. Schematic workflow of specimen acquisition, preparation, group allocation, model creation, and biomechanical testing. Supplementary Material 2: Figure S2. Representative torque–torsion angle curve of the distal radius fixation construct during quasi-static torsional loading. Supplementary Material 3: Figure S3. Schematic illustration of the axial compression test setup. The figure shows the upper loading platen, PMMA joint sphere, distal fragment, proximal fragment, lower clamp, and the positions of the two cameras used to monitor displacement and rotation. The loading protocol included an initial loading phase at 2 mm/min up to 100 N to record load–displacement and rotation, followed by reloading at 5 mm/min until defect closure to measure mechanical strength.