Biomimetic Porous Ti6Al4V Implants: A Novel Interbody Fusion Cage via Gel‐Casting Technique to Promote Spine Fusion

Abstract

An interbody fusion cage (Cage) is crucial in spinal decompression and fusion procedures for restoring normal vertebral curvature and rebuilding spinal stability. Currently, these Cages suffer from issues related to mismatched elastic modulus and insufficient bone integration capability. Therefore, a gel‐casting technique is utilized to fabricate a biomimetic porous titanium alloy material from Ti6Al4V powder. The biomimetic porous Ti6Al4V is compared with polyetheretherketone (PEEK) and 3D‐printed Ti6Al4V materials and their respective Cages. Systematic validation is performed through mechanical testing, in vitro cell, in vivo rabbit bone defect implantation, and ovine anterior cervical discectomy and fusion experiments to evaluate the mechanical and biological performance of the materials. Although all three materials demonstrate good biocompatibility and osseointegration properties, the biomimetic porous Ti6Al4V, with its excellent mechanical properties and a structure closely resembling bone trabecular tissue, exhibited superior bone ingrowth and osseointegration performance. Compared to the PEEK and 3D‐printed Ti6Al4V Cages, the biomimetic porous Ti6Al4V Cage outperforms in terms of intervertebral fusion performance, achieving excellent intervertebral fusion without the need for bone grafting, thereby enhancing cervical vertebra stability. This biomimetic porous Ti6Al4V Cage offers cost‐effectiveness, presenting significant potential for clinical applications in spinal surgery.

A biomimetic porous Ti6Al4V is fabricated via a gel‐casting technique. It is compared with polyetheretherketone and 3D‐printed Ti6Al4V materials and their respective Cages. Systematic validation is performed through mechanical testing, in vitro cell, in vivo rabbit bone defect implantation, and ovine anterior cervical discectomy and fusion experiments to evaluate the mechanical and biological performance of the materials.

1. Introduction

Spinal decompression and fusion surgery is the most widely accepted and extensively utilized surgical technique for treating various spinal diseases, including degenerative, traumatic, and neoplastic conditions.^{[ 1} , ² , ^{3 ]} As one of the core components of this technique, the intraoperative implantation of an interbody fusion cage (Cage) is essential for restoring the normal physiological curvature of the vertebrae and rebuilding spinal stability.^{[ 4} , ⁵ , ^{6 ]}

Currently, conventional Cages used in clinical practice are fabricated from polyetheretherketone (PEEK) and titanium alloy materials.^{[ 7} , ⁸ , ^{9 ]} Owing to their favorable elastic moduli (E‐moduli), PEEK Cages have become the most commonly used Cages in clinical settings.^{[ 10 ]} However, both PEEK and titanium alloy materials are biologically inert, lacking intrinsic osteoinductivity and demonstrating limited osteoconductive properties. Consequently, the Cage interior must be filled with autogenous bone to achieve spinal fusion.^{[ 11} , ^{12 ]} Post‐implantation interface fibrous encapsulation often occurs because of the hydrophobic nature of the PEEK material, hindering its bone integration capability and fusion rate.^{[ 13 ]} Conversely, the E‐modulus of a titanium alloy Cage significantly exceeds that of osseous tissue, making it prone to issues such as implant displacement, subsidence, and stress shielding.^{[ 14} , ^{15 ]} In summary, the ideal interbody fusion material should possess three key characteristics: (1) Exceptional biocompatibility and mechanical properties, (2) Outstanding osteoinductivity, bone ingrowth, and osteoconduction to facilitate direct osseointegration at the implant–bone interface without requiring bone grafts, and (3) High fatigue resistance to withstand repetitive mechanical loading of the spine.

Porous metallic structures have been shown to exhibit enhanced osteoconduction, facilitating a direct connection with bone tissue.^{[ 16} , ^{17 ]} Furthermore, they possess a robust ability to withstand cyclic loads or fatigue resistance. In clinical practice, the growing trend is to incorporate porous metals on the surface of implants.^{[ 18} , ¹⁹ , ^{20 ]} With advancements in manufacturing techniques, including powder sintering, combustion synthesis, and plasma spraying, porous titanium materials have continually evolved.^{[ 21} , ²² , ^{23 ]} Among them, one of the most representative is a 3D‐printed porous titanium alloy Cage manufactured using electron beam melting (EBM) technology. By customizing the porosity and structure, its mechanical properties and structure are brought closer to those of bone tissue to enhance osteoconduction.^{[ 24} , ^{25 ]} In addition, personalized customization allows the external structure of the Cage to be tailored to the shape of any vertebral body, thus meeting biomechanical and ergonomic requirements and effectively reducing implant subsidence.^{[ 26 ]} Despite the drawback of uniform pore shape and higher costs associated with the need for customization, successful clinical applications have already been found.^{[ 18 ]}

In this study, a gel‐casting technique was employed to develop a newly coined material,^{[ 27 ]} biomimetic porous Ti6Al4V material, with a non‐uniform porosity and microstructure closely resembling the trabecular bone in terms of pore size, porosity, and structure. This material boasts low production costs, high cleanliness standard, and large‐scale manufacturing suitability. Subsequently, the mechanical performance of this biomimetic porous titanium alloy material was investigated. Furthermore, in vitro cell experiments, rabbit femoral condyle defect implantation, and sheep anterior cervical discectomy and fusion (ACDF) were conducted to observe and compare the biocompatibility, osteogenic capabilities, bone ingrowth, osseointegration, and intervertebral fusion performance of the biomimetic porous Ti6Al4V, PEEK, and 3D‐printed Ti6Al4V materials prepared using EBM technology, as well as their respective Cages (Figure 1 ).

Figure 1.

Schematic illustration of the biomimetic porous Ti6Al4V interbody fusion cage (cage) for spine fusion.

2. Experimental Section

2.1. Preparation of Samples

To fabricate the biomimetic porous Ti6Al4V material, a high‐purity (≥ 99.5%) Ti6Al4V powder was used as the novel biomimetic material in the gel‐casting method rely on ZhongAoHuiCheng Technology Co. The samples were fabricated as previously described,^{[ 27 ]} and the sintering process parameters were optimized. First, a powder suspension was prepared using 300 g of Ti6Al4V, 201 g of H₂O, 6.4 g of agar (3.18% on H₂O), 6 g of Tergitol TMN 10 (2% on Ti), 3 g of Triton (1% on Ti), and 0.36 g of ammonium alginate (0.18% on H₂O) and mixed for 7 min at 70 °C to obtain a fluid foam. The foam was cast into the specified mold and cooled until it formed a gel structure. After demolding, the structure was dried at a pressure of 10³ mbar at 24 °C for 6–7 d. After that, under an argon atmosphere, the material was slowly heated at a rate not exceeding 20 °C h⁻¹ to a temperature between 400 and 600 °C for 2 h. Subsequently, the material was sintered under vacuum via slow heating to 1250 °C for 2 h.

To fabricate the 3D‐printed TiAl6V alloy, the 3D‐printed materials were designed using CAD software (CATIA, USA) with a diamond lattice porous structure. The material designs were saved in Standard Template Library (STL, USA) format and transferred to the Arcam EBM Q10plus device (Arcam, Sweden) for 3D‐printing based on EBM technology. The raw material used for the samples was Ti6Al4V standard powder with particle diameters ranging from 45 to 106 µm. The powder was selectively melted layer‐by‐layer under vacuum using an electron beam, with a layer thickness of 0.05 mm. After shaping, the samples were naturally cooled to 100 °C under a helium atmosphere and then exposed to air to complete the printing process.^{[ 28 ]}

The PEEK materials, PEEK Cages, and anterior cervical internal fixation system were procured from Stryker (USA).

Three types of samples were manufactured in each group for different experiments: disc specimens of 2 mm in thickness and 8 mm in diameter were used for the cell experiments in vitro; rod specimens of 6 mm in height and 5 mm in diameter were used for the in vivo rabbit bone defect implantation models (excluding PEEK material); and Cage specimens of 12 mm in length, 10 mm in width and 5, 6, or 7 mm in height were used for the ovine ACDF models (Figure 2A). For the mechanical performance testing, different biomimetic porous titanium samples with varying porosities were produced.

Figure 2.

Characterization of the biomimetic porous Ti6Al4V. A) Overview of the manufactured samples (disc, rod, and cage types). B) 3D view of the sample. C) Plane porosity analysis of the specimen. D) The absolute, open, and close porosities of samples with different porosities obtained via mercury intrusion porosimetry (n = 3 per group). E) Cross‐section pore diameter distribution of the samples with different porosities. Pore diameter distribution of the samples with different porosities obtained via F) micro‐CT and G) the mercury porosimetry method. H) Microstructures of the samples obtained via scanning electron microscopy (SEM). I) Representative images of the energy‐dispersive spectrometry (EDS) detection data of the samples.

Before use, all implants were cleaned in an ultrasonic cleaner sequentially using ample amounts of acetone, ethanol, and deionized water for 20 min and then dried overnight at 60 °C.

2.2. Sample Characterization

The surface topographies and microstructures of the biomimetic porous Ti6Al4V samples were observed via scanning electron microscopy (SEM) (Bruker, Germany). Micro‐computed tomography (Micro‐CT) (LAJ Testing, China) was employed to measure the pore size and porosity of the materials and conduct a layer‐by‐layer porosity analysis.^{[ 29 ]}

Two methods, Micro‐CT and mercury porosimetry (Quantachrome, USA), were used to assess the permeability and porosity of the biomimetic porous Ti6Al4V materials.^{[ 29} , ^{30 ]}

The distribution of elements on the surface of the biomimetic porous Ti6Al4V samples was determined by energy‐dispersive spectrometry (EDS) (Zeiss, Germany).

2.3. Mechanical Testing

Mechanical tests were performed via a mechanical testing machine (Instron, USA) with a load cell of 20 kN. The static shear and tensile tests were performed by a universal material testing machine (UTM, China). The biomimetic porous Ti6Al4V cuboid samples (length of approximately 15 mm and rectangular cross‐section of 10 mm by 10 mm) were manufactured for the different porosity groups to determine the compressive strength and E‐modulus (n = 3 per porosity). The biomimetic porous Ti6Al4V Cage samples (12 mm in length, 10 mm in width, and 6 in height) were manufactured to examine shear strength (n = 5). The biomimetic porous Ti6Al4V rod samples (8 cm in height and 5 mm in diameter) were fabricated for the tensile strength test (n = 3).

2.4. Cellular Experiments In Vitro

2.4.1. Cell Culture and Inoculation

Osteoblast‐like MC3T3‐E1 cells (MREDA, China) were cultured in Dulbecco's modified Eagle medium/F12 (Gibco, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco, USA) and 1% penicillin–streptomycin (Gibco, USA) in a constant temperature and humidity incubator at 37 °C and 5% CO₂ concentration, respectively. The medium was changed every 2 d and cultured for 96 h. The third passage cells were used in the current study. Then, the MC3T3‐E1 cells were inoculated on the surfaces of disk specimens of the three groups at densities of 1 × 10⁵ cells per specimen in a 12‐well plate. The medium was replaced with an osteogenic medium (α‐minimum essential medium (α‐MEM) (Gibco, USA) supplemented with 10% FBS, 1% penicillin–streptomycin, 1 mM dexamethasone (Solarbio, China), 1 M β‐glycerol phosphate (Solarbio, China), and 10 mM ascorbic acid (Solarbio, China)) on the second day. The medium was changed every 2 d.

2.4.2. Cell Morphology

The disk specimens inoculated with MC3T3‐E1 cells were cultured for 1, 3, and 5 d and then rinsed thrice with phosphate‐buffered saline to remove non‐adherent cells and fixed with 4% paraformaldehyde (PFA) at 4 °C for 4 h. Subsequently, the samples were dehydrated with a series of gradient ethanol solutions (50%, 70%, 90%, 95%, and 100%). Finally, the specimens were coated with a gold layer, and the cell morphologies were observed using SEM (JSM, Japan).

2.4.3. Cell Proliferation and Viability

The cell proliferation and viability assays were performed via the cell counting kit‐8 (Beyotime, China) method and the live/dead cell staining kit (Thermo, USA).

For the CCK‐8 assay, after culturing for 1, 3, 5, and 7 d, the medium in each sample was removed, and the CCK‐8 solution was added (n = 3 per group at each observation time). After incubating for 1.5 h in a dark environment, a 100 µL supernatant was collected to measure the optical density (OD) value at the wavelength of 450 nm using a microplate reader (Thermo, USA).

For the live/dead cell staining assay, after culturing for 1, 3, and 5 d, the cells attached to the samples were stained with a 0.1% fluorescein diacetate (FDA)/propidium iodide (PI) solution and then observed under a confocal laser microscope (CLSM) (Nikon, Japan). In the CLSM observation, living cells fluoresce green, and dead cells fluoresce red.

2.4.4. Cell Osteogenic Differentiation

Osteogenic differentiation of the cells of each specimen was evaluated via alkaline phosphatase (ALP) staining and activity detection (Thermo, USA), mineralized nodule staining (Beyotime, China), real‐time quantitative polymerase chain reaction (RT‐qPCR), and Western bolt of osteogenesis‐related factors.

For ALP staining and activity detection, after osteogenic culturing for 7 and 14 d, the cells of each group were fixed with 4% PFA and then incubated at 37 °C with an ALP enzyme solution for 15 min, Co solution for 5 min, and working vulcanized solution for 3 min. The ALP activity in the stained samples was measured using an inverted fluorescence microscope (Nikon, Japan). The cells were lysed with 1% Triton‐X 100 (Sigma, USA). The protein concentrations were measured using a bicinchoninic acid (BCA) protein assay kit (Beyotime, China). The ALP activity was calculated according to the OD values measured by a microplate reader at 520 nm (n = 3 per group at each observation time).

For mineralized nodule staining, after osteogenic culturing for 28 d, the cells seeded onto the disc specimens were fixed with 4% PFA and then incubated and stained with alizarin red (AR) solution at 25 °C for 10 min. Finally, the stained specimens were exposed, and images were captured using an optical microscope (Olympus, Japan).

For RT‐qPCR, after osteogenic culturing for 4, 7, and 14 d, the cells were harvested, and their total RNA was extracted using Trizol reagent (Sigma, USA) according to the manufacturer's instructions (n = 3 per group at each observation time). The quality of RNAs were detected by Nano‐one (Thermo, USA). Then, the cDNA was synthesized using high‐capacity reverse transcription kit (Thermo, USA) as templates for RT‐qPCR. The RT‐qPCR was performed on CFX96 (Bio‐Rad, USA) using SYBR Green PCR Master Mix (Thermo, USA), and Ct values of target genes were normalized to GAPDH. The osteogenesis‐related genes included the genes of ALP, Runt‐related transcription factor 2 (RUNX2), Type I collagen (COL‐I), osteocalcin (OC), osteopontin (OPN), osteoclast inhibitor (OPG), and osterix (OSX). The 2^−ΔΔCt method was used to calculate the relative RNA expression levels of the target genes.^{[ 31 ]} All primer sequences are listed in Table 1 .

Table 1.

Primer sequences used for RT‐qPCR.

Gene	Forward primer	Reverse primer
ALP	CCAACTCTTTTGTGCCAGAGA	GGCTACATTGGTGTTGAGCTTTT
RUNX2	AACGATCTGAGATTTGTGGGC	CCTGCGTGGGATTTCTTGGT
COL‐I	GCTCCTCTTAGGGGCCACT	CCACGTCTCACCATTGGGG
OC	CTGACCTCACAGATCCCAAGC	TGGTCTGATAGCTCGTCACAAG
OPN	AGCAAGAAACTCTTCCAAGCAA	GTGAGATTCGTCAGATTCATCCG
OPG	ACCCAGAAACTGGTCATCAGC	CTGCAATACACACACTCATCACT
OSX	ATGGCGTCCTCTCTGCTTG	TGAAAGGTCAGCGTATGGCTT
GAPDH	AGGTCGGTGTGAACGGATTTG	TGTAGACCATGTAGTTGAGGTCA

For Western blot, after osteogenic culturing for 7 and 14 d, the cells of the different groups were lysed with a radio immunoprecipitation assay lysis buffer (Beyotime, China). The protein concentration in the cell extracts was measured via the BCA standard method (Beyotime, China). The 100 µg protein samples were separated by 10% sodium dodecyl sulphate‐polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride (PVDF) membrane (Millipore, Germany). The PVDF membranes were blocked with 3% bull serum albumin for 2 h at 24 °C after being washed thrice with 0.05% tris buffered saline tween (TBST) (Beyotime, China). After that, the PVDF membranes were incubated overnight with the primary antibodies at 4 °C. Subsequently, the prepared membranes were incubated with corresponding secondary antibodies at 37 °C for 2 h after being washed thrice with TBST. Blots were detected using an enhanced chemiluminescence kit (Beyotime, China), and the signal intensity was quantified with Image J software (n =  3 per group at each observation time). The following primary antibodies were used: rabbit polyclonal anti‐mouse ALP (1:2000, Cat. No. ab229126), rabbit monoclonal anti‐mouse RUNX2 (1;1500, Cat. No. ab192256), rabbit monoclonal anti‐mouse COL‐I (1;1000, Cat. No. ab260043), rabbit polyclonal anti‐mouse OC (1:1500, Cat. No. ab93876), rabbit recombinant multiclonal anti‐mouse OPN (1:1000, Cat. No. ab283656), rabbit polyclonal anti‐mouse OPG (1:2000, Cat. No. ab73400), and rabbit monoclonal anti‐mouse OSX (1:2000, Cat. No. ab209484); the antibodies were obtained from Abcam (UK).

2.5. Animal Experiments for Bone Defect Repair

2.5.1. Animals

Twenty‐four male New Zealand white rabbits (aged 6‐month‐old and weighing 4–5 kg) were used in this study (Peking university third hospital, Beijing, China) and randomly divided into two groups (biomimetic porous Ti6Al4V group and 3D‐printed Ti6Al4V group, n = 12 per group). The animal study protocol was approved by the Ethics Committee of Peking University Third Hospital (Approval No. A2019021).

2.5.2. Construction of the Bone Defect Repair Rabbit Model

All rabbits were anesthetized with 1% pentobarbital sodium (3 mL kg⁻¹) and kept supine. Preoperative preparation was performed according to sterility principles. A 3 cm longitudinal incision was made on the lateral side of the distal femur, and the skin, subcutaneous tissue, and muscle were separated to expose the lateral femoral condyle of the left lower limb. Thereafter, a columned bone defect of Ø5 × 6 mm was drilled perpendicular to the lateral femoral condyle surface using a grinding drill. Different rod specimens were implanted into the defects for the different groups (biomimetic porous Ti6Al4V group: the biomimetic porous Ti6Al4V; 3D‐printed Ti6Al4V group: 3D‐printed Ti6Al4V). Subsequently, the incision was irrigated with a normal saline solution and closed one layer at a time. After surgery, each rabbit was injected intramuscularly with 80000 U of penicillin for 3 d to prevent infection and allow unrestricted cage activity.

2.5.3. Radiologic Evaluation

After 4 weeks of implantation, half of the animals in each group were euthanized via an overdose of pentobarbital sodium (n = 6 per group), and the remainder were euthanized at 12 weeks (n = 6 per group). The femur samples were removed from the lower left limbs and fixed with 4% PFA. The implantation position and condition of bone defect repair were measured using lateral X‐ray images of the left femurs using a C‐arm X‐ray machine (46 kV, 1.42 mA, GE‐OEC Fluorostar, USA). The representative bone defect repair images and quantitative analyses of the crucial indexes of osseointegration, including the ratio of bone volume to tissue volume (BV/TV), ratio of bone surface area to bone volume (BS/BV), trabecular number (Tb.N), trabecular thickness (Tb.Th) and trabecular spacing (Tb.Sp)), were evaluated using a Micro‐CT machine (80 kV, 80 mA, Siemens, Germany) and the multi‐modal 3D visualization software Inveon Research Workplace (Siemens, Germany). The different thresholds (Hounsfield, HU) were divided to distinguish bone tissue, soft tissue, and implants (bone tissue: 1000–2250 HU, soft tissue: 0–1000 HU, implants: > 2250 HU). After 3D reconstruction, the 500 µm area around the implant and internal space was regarded as the region of interest (ROI). The executor was blinded to all groups being assessed.

2.5.4. Push‐Out Test for Biomechanical Examination

To determine the bonding strength between the implant and bone, the push‐out force of the implant was tested using a universal mechanical testing machine (MTS, China). Three samples from each group at 4 and 12 weeks were randomly selected (n = 3 per group at each observation time). First, the internal surface of the implant was exposed using a hard tissue slicer (Exakt, Germany), and the bone tissue surrounding the outer surface of the implant was removed. The samples were placed in the universal MTS, and a push force was applied from the inner side of the implant at a rate of 1.5 mm min⁻¹. The end point of the experiment was a sudden drop in the push‐out force loaded by the MTS. The maximum axial force during this process was recorded and regarded as the push‐out force.^{[ 28 ]}

2.5.5. Histological Analysis

Three samples from each group at 4 and 12 weeks were randomly selected for implant histological analysis via hard section staining (n = 3 per group at each observation time). First, the samples were fixed with 10% formaldehyde solution for 1 d and dehydrated in an alcohol gradient (concentration of 40%, 75%, 95%, and 100% for 3 d each). The samples were then plastic‐dipped in methyl methacrylate (MMA) for 7 d. Subsequently, the samples were embedded using a photocurable embedding instrument. The samples were then prepared via a hard tissue slicer. Toluidine blue staining and hematoxylin‐eosin (H&E) staining were employed to stain the prepared hard tissue sections,^{[ 32} , ^{33 ]} and the stained samples were observed using an optical microscope.

2.6. Animal Experimentation for ACDF

2.6.1. Animals

Eighteen male small‐tailed Han sheep (aged 24‐month‐old and weighing 30–35 kg) were used for this experiment (Peking university third hospital, Beijing, China) and randomly divided into six parallel groups (Groups 1–6, n = 3 per parallel group). In every group, Cages fabricated from three different materials (PEEK, 3D‐printed Ti6Al4V, and biomimetic porous Ti6Al4V Cages) were implanted into the cervical intervertebral spaces of different sheep (C2/3, C3/4, and C4/5). The description of the placement of implants is listed in Table 2 . The animal study protocol was approved by the Ethics Committee of Peking University Third Hospital (Approval No. A2022064).

Table 2.

Placement of different cervical interbody fusion cages.

Segment	No. 1 in each parallel group	No. 2 in each parallel group	No. 3 in each parallel group
C2/3	PEEK Cage	3D‐printed Ti6Al4V Cage	Biomimetic porous Ti6Al4V Cage
C3/4	Biomimetic porous Ti6Al4V Cage	PEEK Cage	3D‐printed Ti6Al4V Cage
C4/5	3D‐printed Ti6Al4V Cage	Biomimetic porous Ti6Al4V Cage	PEEK Cage

2.6.2. Establishment of the ACDF Ovine Model

All sheep were kept in a supine position for induction analgesia using propofol (5 mg k⁻¹g) and then maintained in that state by inhalation with 2–2.5% isoflurane. Meloxicam (0.05 mg k⁻¹g h⁻¹) was continuously pumped to maintain analgesia. Preoperative localization of C2/3–C4/5 was performed using a C‐arm X‐ray machine. Shaving, disinfection, and sterile sheet covering were executed based on sterility principles. First, a longitudinal incision (about 12 cm in length) was made on the right side of the midline of the neck. The skin, subcutaneous tissue, and platysma muscle were cut layer‐by‐layer. Thereafter, the anterior cervical fascia was exposed and longitudinally incised by approaching between the carotid vascular and visceral sheaths. The musculus longus colli was found and opened. Subsequently, four vertebrae and three intervertebral spaces (C2/3–C4/5) were revealed. The discectomy and decompression of these three intervertebral discs (C2/3–C4/5) were performed using the ACDF approach. Thereafter, Cages fabricated from three different materials were implanted in different cervical intervertebral spaces according to the previously described plan in Table 2. Finally, the anterior cervical titanium alloy plate system was installed, and the incisions were sutured layer‐by‐layer. Immediate postoperative X‐rays were taken to ensure the implants were in good position. After surgery, each sheep was injected with 100000 U of penicillin for 3 d to prevent infection and allow unrestricted activity.

2.6.3. Imaging Examination

Anterior and lateral X‐rays of the cervical vertebra were taken immediately after surgery to evaluate the position of the implant and fixation via a C‐arm X‐ray machine (58 kV, 2.78 mA).

To evaluate the condition of bone ingrowth, cervical vertebra CTs were taken at 2 months after surgery using a CT machine (120 kV, 105 mA, Sinovision, China).

After 3 and 6 months of implantation, three parallel groups were euthanized via an overdose of pentobarbital sodium. The functional spinal unit (FSU) samples were removed from the cervical vertebrae (C2/3−C4/5) via the en‐block excision method. The bone ingrowth and fusion conditions near the Cages were assessed by Micro‐CT (80 kV, 80 mA) and the multi‐modal 3D visualization software Inveon Research Workplace (n = 9 per group at each observation time). The different thresholds (Hounsfield, HU) were divided to distinguish bone tissue, soft tissue, and metal implants (bone tissue: 1000–2250 HU, soft tissue: 0–1000 HU, metal implants: >2250 HU). After 3D reconstruction, the 1000 µm area around the implant and internal space was regarded as the ROI. The executor was blinded to all groups being assessed.

2.6.4. Biomechanical Test

Six cervical FSU samples with different implants at 3 and 6 months after surgery were randomly selected for the biomedical test (n = 6 per group). The muscles of each FSU were removed, leaving the intervertebral discs, ligaments, and articular capsules. Screws were inserted that penetrated through the upper and lower vertebral body and embedded in the MMA for fixation to the biomechanical testing machine (MTS, China). The 3D displacement of each segment was determined via an optical measurement system (Optotrak, Canada). Flexibility tests were executed at pure moments of 2.5 Nm in the following three motion planes: flexion−extension, lateral bending to the right/left, and axial rotation to the right/left. The moment was applied at a rate of approximately 0.5° s⁻¹. The range of motion (ROM, defined as the angular displacement for the minimum and maximum bending moment) of each FSU sample was tested.^{[ 34} , ^{35 ]}

2.6.5. Hematological Examinations

To evaluate the degree of inflammation and osteogenic activity, three venous blood samples from each parallel group were collected at 1 month for hematological examination (n = 3 per parallel group). Examinations of the blood routine as well as C‐reactive protein and ALP levels were performed.

2.6.6. Histological Analysis

FSU samples with different implants were randomly selected at 3 and 6 months after surgery for implant histological analysis via hard section staining and analysis (n = 3 per group at each observation time). The steps were the same as those described in Section 2.5.5. Goldner's trichrome staining, toluidine blue staining, and methylene blue‐acid fuchsin staining were employed to stain the prepared hard tissue sections.^{[ 32} , ³⁶ , ^{37 ]} To evaluate the biosecurity via paraffin section H&E staining, three sheep from the remaining three parallel groups were randomly selected at 6 months for histological analysis of the major organs (heart, liver, spleen, lung, and kidney) (n = 3, randomly selected).^{[ 38 ]}

2.7. Statistical Analysis

Data were presented as mean ± standard deviation. Comparison of continuous data between the two groups was performed via independent Student's t‐test, which was analyzed using SPSS software version 20 (IBM, USA). Differences at P‐value < 0.05 were considered statistically significant.

3. Results

3.1. Sample Characterization

3.1.1. Morphological Characteristics and EDS Results

Different types of biomimetic porous Ti6Al4V samples were successfully fabricated using the gel‐casting technique (Figure 2A), and then a series of characterization of the material's properties were executed (Figure 1). The analyses of Micro‐CT and mercury porosimetry were conducted to determine the potential of the fabrication technique to create high‐porosity and interconnected structure (Figure 2B–G). In the samples with different porosity, the open porosity of the samples increased with the increase of the absolute porosity, indicating good sample interconnectedness (Figure 2D). The two plots of cumulative volume versus pore diameter distribution in different porosity samples demonstrated that the pores had good dispersity, and the average diameters of the largest pores exceeded 200 µm (200–600 µm) (Figure 2F,G). The microstructure images of the samples were captured using SEM, revealing loose and porous structures and uneven pore sizes (Figure 2H). The EDS data showed that the sample was made up of Ti6Al4V (Figure 2I). Thus, these data confirmed that the porous Ti6Al4V sample had a structure similar to that of bone trabecular tissue and a porous interconnected network with proper pore size and distribution, which was conducive to osteocyte migration and proliferation.^{[ 39 ]}

3.1.2. Mechanical Properties

To explore the mechanical performance of the biomimetic porous Ti6Al4V samples, compression, static shear, and tensile tests were performed. Typical compressive stress–strain curves for the biomimetic porous Ti6Al4V samples with different porosities are shown in Figure 3A. The minimum compressive strength exceeded 40 MPa. Representative stiffness–relative density curves and E‐moduli of the samples are presented in Figure 3B. The data demonstrated that the compressive E‐modulus increased with the decrease of porosity, and the average E‐moduli of the samples ranged from 1 to 3.5 GPa. Furthermore, they were in the range of the compressive E‐modulus of cancellous and cortical bone (0.5–20 GPa) and close to that of cancellous bone.^{[ 40 ]} Moreover, the static shear properties of the samples are summarized in Table 3 . The data showed that the average shear strength was 5343.9 ± 325.0 N mm⁻¹. In addition, as shown in Table 4 , the mean tensile strengths of the samples were 27.1 ± 3.1 MPa. These results indicated that the biomimetic porous Ti6Al4V was similar to cancellous bone in compressive E‐modulus with excellent compressive, shear, and tensile strengths. The biomimetic porous Ti6Al4V was thus suitable for bone implantation in bone defects and intervertebral fusion.

Figure 3.

Mechanical properties of the biomimetic porous Ti6Al4V. A) Typical compressive stress–strain curves for samples with different porosities (n = 3 per group). B) Representative stiffness–relative density curves and elastic moduli (E‐moduli) of the samples (n = 3 per group).

Table 3.

Static shear properties of the biomimetic porous Ti6Al4V (n = 5).

Sample No.	Yield load [N]	Yield displacement [mm]	Ultimate load [N]	Ultimate displacement [mm]	Stiffness [N mm−1]
1	3341	1.4	7351	2.8	5503.7
2	2955	1.4	7291	2.7	5518.9
3	2911	1.4	6228	2.5	5353.8
4	3335	1.5	6498	2.6	4779.8
5	3366	1.5	7313	3.0	5563.2
Mean value	3182	1.4	6936	2.7	5343.9
Standard deviation	228	0.1	532	0.2	325.0

Table 4.

Tensile properties of the biomimetic porous Ti6Al4V (n = 3).

Sample No.	Diameter [mm]	Tensile strength [MPa]	Ultimate load [N]
1	3.01	28.2	201
2	3.03	23.6	170
3	3.02	29.5	211
Mean value	3.02	27.1	194
Standard deviation	0.01	3.1	21

3.2. Cellular Experiments In Vitro

3.2.1. Cell Morphology

The morphology of the MC3T3‐E1 cells in the three groups of disc specimens was observed via SEM. As shown in Figure 4A, a few cells had already proliferated and attached to the surface of the discs in all groups after 1 d. These cells were mainly spherical and elliptical in shape. After 3 d, more cells adhered and diffused into the sample pores in good condition and exhibited extending pseudopods. The cells mainly presented fusiform. The number of cells in the 3D‐printed Ti6Al4V and biomimetic porous Ti6Al4V groups was large in comparison with that in the PEEK group. Five days later, an increased quantity and significant morphological changes in the cells were observed; the cells were widely and confluently interconnected and covered the surface and pores of the discs. The above representative cell morphology changes are indicated by red arrows in the SEM images.

Figure 4.

In vitro biocompatibility of the samples for the three groups (PEEK, 3D‐printed Ti6Al4V, and biomimetic porous Ti6Al4V). A) Typical SEM images of MC3T3‐E1 cells after seeding on the samples for 1, 3, and 5 d (the representative cell morphology changes are indicated by red arrows). B) Live/dead cell staining (fluorescein diacetate/propidium iodide (FDA/PI) dual‐fluorescence) microscopy images of cells after seeding on the samples for 1, 3, and 5 d (green: live cells; red: dead cells). C) Proliferation of cells cultured on the samples for 1, 3, 5, and 7 d as measured by cell counting kit‐8 (CCK‐8) assay (n = 3 per group at each observation time). Data are expressed as mean ± standard deviation. Two of the three groups are compared via an independent t‐test. ^* p < 0.05.

3.2.2. Cell Proliferation and Viability

To investigate the cell proliferation in different groups during the culture period, CCK‐8 assay was performed. The results demonstrated that cell proliferation presented an increasing trend with time. Except that the cell proliferation of the 3D‐printed Ti6Al4V group was remarkably decreased relative to that of the PEEK group after culturing for 1 d; the assay failed to show a marked difference among the three groups after culturing for 1, 3, 5, and 7 d (Figure 4B). Consistent with the CCK‐8 results, the fluorescence microscopy images of the FDA‐PI merge showed a homogenous distribution of the living cells (dyed green) and relatively few dead cells (dyed red) on the surfaces of the three groups. The green fluorescence intensity increased with time (from 1 to 5 d) (Figure 4C). Together with the SEM results, these data confirmed that the materials all showed good biocompatibility for cell growth, and the biomimetic porous Ti6Al4V could meet the biological conditions as a scaffold for cell growth.

3.2.3. Osteogenic Differentiation In Vitro

To explore the osteogenic differentiation capability of the cells after osteogenic culturing, ALP staining and activity detection, mineralized nodule staining, RT‐qPCR, and Western blot of osteogenesis‐related factors (including ALP, RUNX2, COL‐I, OC, OPN, OPG, and OSX) were conducted. The ALP staining and activity examination results indicated that ALP activity was enhanced with the increase in time (from 7 to 14 d) (Figure 5A,B). Except that there was no remarkable difference in ALP activity between the biomimetic porous Ti6Al4V and 3D‐printed Ti6Al4V groups, there were significant differences among the three groups at each time point. The biomimetic porous Ti6Al4V group showed the highest activity (Figure 5B). Mineralized nodule staining using AR was performed to assess the mineralization capacity after 28 d of incubation. Several prominent mineralized nodules were observed in the biomimetic porous and 3D‐printed groups (Figure 5C).

Figure 5.

In vitro osteogenic differentiation of MC3T3‐E1 cells for the three groups. A) Representative images of alkaline phosphatase (ALP) staining of the cells seeded on the samples after osteogenic culturing for 7 and 14 d. B) Quantitative analyses of the relative ALP activities of the samples (n = 3 per group at each observation time). C) Typical alizarin red (AR) staining for the cells seeded on the samples after osteogenic culturing for 28 d. D–F) Relative mRNA levels of osteogenesis‐related genes (ALP, RUNX2, COL‐I, OC, OPN, OPG, and OSX) in the cells after osteogenic culturing for (D) 4, (E) 7, and (F) 14 d measured via RT‐qPCR analyses (n = 3 per group at each observation time). G–J) Relative protein levels of osteogenesis‐related factors (ALP, RUNX2, COL‐I, OC, OPN, OPG, and OSX) in the cells after osteogenic culturing for (G,H) 7 and (I,J) 14 d measured via Western blot analyses (n =  3 per group at each observation time). H,J) Quantitative analyses showing the differences in cell protein levels among the three groups after osteogenic culturing for (H) 7 and (J) 14 d. Data are expressed as mean ± standard deviation. Two of the three groups are compared via an independent t‐test. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, and ^**** p < 0.0001.

The osteogenesis‐related factors were examined in the MC3T3‐E1 cells of different groups using RT‐qPCR at the gene level and Western blot at the protein level. The RT‐qPCR analysis showed that compared with those in the PEEK group, the mRNA levels of osteogenesis‐related genes (including ALP and OPN) were all remarkably upregulated in the cells of the biomimetic porous Ti6Al4V and 3D‐printed Ti6Al4V groups after culturing for 4 d. Furthermore, the mRNA levels of ALP, RUNX2, OC, OPG, and OSX were all significantly increased in the biomimetic porous Ti6Al4V group after culturing for 7 d. In addition, the mRNA levels of ALP, RUNX2, COL‐I, OC, OPN, OPG, and OSX all had a significant increase in the biomimetic porous Ti6Al4V group after osteogenic culturing for 14 d (Figure 5D–F). Consistent with the RT‐qPCR results, Western blot further demonstrated that the protein levels of these osteogenesis‐related factors in the cells of the biomimetic porous Ti6Al4V and 3D‐printed Ti6Al4V groups also markedly increased in comparison with those in the PEEK group after osteogenic culturing for 7 and 14 d, including ALP, RUNX2, COL‐I, OC, OPN, OPG, and OSX (Figure 5G,I). Quantitative analyses specifically showed the differences in the protein levels of the cells among the three groups. The biomimetic porous Ti6Al4V group showed the highest protein levels of osteogenesis‐related factors among the three groups (Figure 5H,J).

Thus, these data revealed that the biomimetic porous Ti6Al4V was more beneficial in promoting osteogenic differentiation than PEEK and 3D‐printed Ti6Al4V. The results of RT‐qPCR and Western blot preliminary indicated the molecular mechanism of cell osteogenesis.

3.3. Animal Experiment for Bone Defect Repair

3.3.1. General Observation

The femoral condyle defect rabbit models were constructed, and rod specimens of different materials (biomimetic porous Ti6Al4V and 3D‐printed Ti6Al4V) were implanted (Figure 6A). All rabbits recovered well after surgery, with no infection or postoperative complications observed during the experimental period. The rabbits were euthanized at the scheduled time points (4 and 12 weeks after surgery) to obtain samples.

Figure 6.

In vivo osseointegration analyses of the samples for two different groups (3D‐printed Ti6Al4V and biomimetic porous Ti6Al4V) via the bone defect repair rabbit model. A) Surgical implantation of rod specimens into the lateral femoral condyle defects of rabbits. B) Lateral X‐ray images of the femur specimens of the two groups 4 and 12 weeks after implantation. C) Representative Micro‐CT and 3D reconstruction images of the two groups 4 and 12 weeks after surgery (new bone tissue is indicated in blue and implants are indicated in white in the 3D reconstruction images). D–H) Quantitative analyses of the critical indexes of osseointegration via Micro‐CT, including the (D) ratio of bone volume to tissue volume (BV/TV), (E) ratio of bone surface area to bone volume (BS/BV), (F) trabecular number (Tb.N), (G) trabecular thickness (Tb.Th), and (H) trabecular spacing (Tb.Sp) (n = 6 per group at each observation time). I) Representative axial push‐out force of the implants of the two groups via biomechanical push‐out tests 4 and 12 weeks after implantation (n = 3 per group at each observation time). J,K) Histological analyses of the bone defect repair rabbit models 4 and 12 weeks after surgery (n = 3 per group at each observation time). (J) Quantitative analysis of BV/TV of the two groups 4 and 12 weeks after surgery. (K) Typical histological staining (toluidine blue staining and hematoxylin‐eosin (H&E) staining) images of the femoral condyle defects 4 and 12 weeks after implantation. Data are expressed as mean ± standard deviation. The two groups are compared via an independent t‐test. ^* p < 0.05, ^*** p < 0.001, and ^**** p < 0.0001.

3.3.2. Imaging Evaluation

X‐ray and Micro‐CT were performed to validate the osseointegration and bone defect repair. The X‐ray images at 4 and 12 weeks after surgery confirmed that the cylindrical implants of the biomimetic porous Ti6Al4V and 3D‐printed Ti6Al4V were both well placed and well fixed. Bone growth was good at the surgical site, without implant dislocation and fracture (Figure 6B). Compared with the X‐ray results, the Micro‐CT data clearly showed excellent bone growth and osseointegration. Not only did the new bone grow well on the implant−bone interface, but it also grew inside many pores in the implant (Figure 6C). The 3D reconstruction images showed that the new bone (indicated in blue) grew extensively around the implant (indicated in white) and in the internal interconnecting pores. The implant achieved good osseointegration, which was more pronounced at 12 weeks than at 4 weeks (Figure 6C). The quantitative analysis of the key indexes of bone growth illustrated that compared with the 3D‐printed Ti6Al4V group, the BV/TV ratio and Tb.N increased remarkably, whereas Tb.Sp declined significantly for the biomimetic porous Ti6Al4V group at each time point (4 and 12 weeks). However, there were no significant differences in the BS/BV ratio and Tb.Th between the two groups (Figure 6D–H). These data demonstrated that the biomimetic porous Ti6Al4V had a better in vivo osseointegration performance than the 3D‐printed Ti6Al4V.

3.3.3. Biomechanical Push‐out Test

A push‐out test was conducted to reflect the bonding strength and stability between the implant and the host bone. As shown in Figure 6I, the push‐out axial force of each group increased as time increased. At each time node (4 and 12 weeks), the axial force of the biomimetic porous Ti6Al4V group was markedly greater than that of the 3D‐printed Ti6Al4V group. Together with the radiologic examination results, these data confirmed that both materials had good bonding with the host bone via osseointegration, and the biomimetic porous Ti6Al4V had an advantage over the 3D‐printed Ti6Al4V with regard to osseointegration in vivo.

3.3.4. Histological Evaluation

A histological analysis was performed via hard section staining to further explore bone ingrowth and osseointegration. Two staining methods using either toluidine blue or H&E were used. The toluidine blue staining images showed that the bone tissue (dyed blue) was well developed on the implant–bone interface with good morphology, and some of the bone had grown into the implants at 4 weeks. After 12 weeks, more bone tissue had grown into the pores of the implants in both groups (Figure 6K). Consistent with the toluidine blue staining results, the H&E staining showed new bone trabeculae wrapped around and growing into the implants, and the osteocytes had clear cytoplasm and regular nuclei. In addition, few inflammatory cells were observed (Figure 6K). The statistical analysis showed that the BV/TV ratio of the biomimetic porous Ti6Al4V group was markedly higher than that of the 3D‐printed Ti6Al4V group at 12 weeks (Figure 6J). These data suggested that the two materials both had good bone ingrowth and osseointegration properties, but the biomimetic porous Ti6Al4V was stronger.

3.4. Animal Experiment for ACDF

3.4.1. General Observation

All the modeling operations and implantation were carried out successfully (Figure 7A), and the physical activity and temperament of all sheep were in good condition after surgery. Most of the sheep were free of wound seepage and infection, except for one sheep that suffered from early postoperative infection (after dressing the wound several times, it gradually recovered). The sheep all showed no signs of nerve damage during the experimental period. Their feeding and behavior were good, and their body weight increased normally. The sheep were euthanized at the scheduled time points (3 and 6 months after surgery) to obtain samples.

Figure 7.

Imageology analyses of the interbody fusion cage (cage) specimens of the three different groups (PEEK cage, 3D‐printed Ti6Al4V cage, and biomimetic porous Ti6Al4V cage) via the in vivo anterior cervical discectomy and fusion (ACDF) ovine model. A) Surgical implantation of the different cages into three intervertebral spaces (C2/3–C4/5) using the ACDF surgical method. B) Preoperative localization and immediate postoperative X‐ray images. C) Transverse‐section CT images of the three groups 2 months after implantation. D) Representative 3D reconstruction CT images of the cervical spine 2 months after surgery. E) Representative Micro‐CT three‐view (axial, sagittal, and coronal views) images of the three groups 3 and 6 months after implantation. F) 3D reconstruction Micro‐CT images of the three groups 3 and 6 months after surgery (new bone tissue is indicated in blue and Cages are indicated in white). G–K) Quantitative analyses of the important indexes of intervertebral fusion via Micro‐CT, including the (G) BV/TV ratio, (H) BS/BV ratio, (I) Tb.N, (J) Tb.Th, and (K) Tb.Sp (n = 9 per group at each observation time). Data are expressed as mean ± standard deviation. Two of the three groups are compared via an independent t‐test. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, and ^**** p < 0.0001.

3.4.2. Radiographic Analyses

To evaluate the Cage position and cervical intervertebral fusion in the three groups, X‐ray, CT, and Micro‐CT were conducted. The X‐ray images immediately after surgery illustrated that the position of the internal fixation was good (Figure 7B). Furthermore, the cervical vertebra CT showed no looseness, shift, or subsidence of internal fixation. Moreover, bone tissue grew into the Cages, and the intervertebral space partially fused at the operative levels 2 months after surgery. Furthermore, obvious bone hyperplasia occurred in front of the internal fixation devices (Figure 7C). This result was also verified by the 3D reconstruction CT images (Figure 7D). To directly and clearly observe the degree of bone ingrowth and intervertebral fusion, Micro‐CT was used to evaluate the bone tissue in the Cage and on the surface between Cage and vertebral interbody 3 and 6 months after surgery. The three views (axial, coronal, and sagittal views) of the Micro‐CT images show that osteogenesis was more obvious in both the Cage and on the surface of the biomimetic porous Ti6Al4V and 3D‐printed Ti6Al4V Cages than that in the PEEK Cage (Figure 7E). The 3D reconstruction Micro‐CT images show that the new bone tissue (indicated in blue) grew and interconnected within the Cage (indicated in white). There was sufficient bone growth on the surface of the Cage. These changes were more significant at 6 months than at 3 months after surgery (Figure 7F). The statistical analysis of the crucial indexes of bone growth and intervertebral fusion based on Micro‐CT demonstrated that compared with the PEEK group, the BV/TV ratio and Tb.N were significantly increased, whereas Tb.Sp was markedly decreased in the biomimetic porous Ti6Al4V and 3D‐printed Ti6Al4V groups at each time point (3 and 6 months). Furthermore, the biomimetic porous Ti6Al4V Cage dramatically outperformed the 3D‐printed Ti6Al4V Cage on all of the above indexes at each time point, except for Tb.N at 3 months, where no significant difference was found (Figure 7G–K). Therefore, these results demonstrated that the biomimetic porous Ti6Al4V Cage had better bone ingrowth and intervertebral fusion properties when compared with those in PEEK and 3D‐printed Ti6Al4V Cages.

3.4.3. Biomechanical Stability

Biomechanical examinations for cervical FSUs were performed 6 months after surgery to evaluate the ROM and degree of stability of the fused cervical vertebra for the different types of Cages. As shown in Figure 8C, there were remarkable differences in the ROM of flexion–extension, lateral bending, and axial rotation among the three groups. Compared with the PEEK and 3D‐printed Ti6Al4V Cages, the biomimetic porous Ti6Al4V Cage could reduce the ROM of the fused cervical vertebrae; it could enhance the biomechanical stability of the cervical vertebra. Combined with the radiographic analysis results, these data suggest that the biomimetic porous Ti6Al4V and 3D‐printed Ti6Al4V Cages exhibited stronger bone ingrowth and intervertebral fusion performance than the PEEK Cages. Moreover, the biomimetic porous Ti6Al4V Cage was superior to the 3D‐printed Ti6Al4V Cage with regard to promoting cervical intervertebral fusion and stability.

Figure 8.

Histological analyses and biomechanical stability of cervical vertebra samples and biosecurity of the cages for the three groups via the ACDF ovine model. A,B) Histological analyses of the ACDF ovine models 3 and 6 months after implantation (n = 3 per group at each observation time). (A) Representative histological staining (Goldner's trichrome staining, toluidine blue staining, and methylene blue‐acid fuchsin staining) images of the cervical vertebra samples of the three groups 3 and 6 months after implantation. (B) Quantitative analysis of the BV/TV ratio of the two groups 3 and 6 months after surgery (because autogenous bone grafts existed in the PEEK group, a true calculation of the amount of new bone tissue was not possible; therefore, only the 3D‐printed Ti6Al4V cage and biomimetic porous Ti6Al4V cage groups were compared). C) Comparison of range of motion of functional spinal units (FSUs) of cervical vertebra in the three groups in the flexion–extension (FE), lateral bending (LB), and axial rotation (AR) motion planes 3 months after surgery (n = 6 per group). D–H) Hematological examination results for the ACDF ovine models 1 months after surgery, including the amounts of (D) serum ALP, (E) red blood cells (RBCs), (F) white blood cells (WBCs), (G) hemoglobin (HGB), and (H) immunocytes (n = 3 per parallel group). I) Histological staining (hematoxylin‐eosin staining) images of the major organs 6 months after surgery (n = 3, randomly selected). Data are expressed as mean ± standard deviation. Two of the three or six parallel groups are compared via an independent t‐test. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, and ^**** p < 0.0001.

3.4.4. Hematological Examination

Blood tests were performed to evaluate the degree of inflammation and osteogenic activity of all sheep 1 month after surgery. The results demonstrate that the blood routine of all parallel groups was normal, and there were no significant differences between the important indexes (Figure 8E–H). The levels of C‐reactive protein of all sheep were in the normal range (< 5.0 mg L⁻¹), indicating that there were no infections. In addition, serum ALP levels in all groups were higher than normal, suggesting that there was active osteogenesis in the ovine models (Figure 8D). These data confirmed the successful establishment of ACDF models and promising in vivo biosecurity of the implants.

3.4.5. Histological Analysis

Three staining methods were conducted to further investigate the degree of bone ingrowth and intervertebral fusion 3 and 6 months after surgery, including Goldner's trichrome staining, toluidine blue staining, and methylene blue‐acid fuchsin staining. The data of Goldner's trichrome staining revealed the following: First, for the PEEK Cage group, the new bone tissue (dyed deep green) of the Cage was integrated with the upper and lower endplates at 6 months owing to the inserted autogenous bone (dyed green) of the Cage, but there were some small gaps in the Cage–vertebra interface, suggesting undesirable bone fusion. Second, there was obvious bone ingrowth for the 3D‐printed Ti6Al4V Cage group, and the quantity of new bone tissue was barely satisfactory. Finally, for the biomimetic porous Ti6Al4V Cage group, the bone growth inside the Cage was denser and more extensive. Although there were also gaps in the Cage–vertebra interface, the condition of the intervertebral fusion was slightly better than the PEEK Cage and 3D‐printed Ti6Al4V Cage groups. These results were confirmed by the toluidine blue and methylene blue‐acid magenta staining images (Figure 8A). The statistical analysis showed that the BV/TV ratio of the biomimetic porous Ti6Al4V Cage group was significantly higher than that of the 3D‐printed Ti6Al4V Cage group at 3 and 6 months (because there were autogenous bone grafts during the operation in PEEK group, it was impossible to truly calculate the amount of new bone tissue. Therefore, only the 3D‐printed Ti6Al4V Cage group and biomimetic porous Ti6Al4V Cage group were compared) (Figure 8B).

The H&E staining of the major organs of the sheep was conducted to assess the biosecurity of the Cages. The images produced by H&E staining illustrated that all tissues were normal in structure and morphology, without any apparent inflammatory response or other injury (Figure 8I). These data suggested that compared with the traditional PEEK and 3D‐printed Ti6Al4V Cages, the biomimetic porous Ti6Al4V Cage had good biocompatibility and better bone ingrowth and intervertebral fusion performance.

4. Discussion

With the continuous progression of the world's aging population, the incidence of fractures, bone defects, and degenerative spinal diseases is rising, leading to a rapid increase in the demand for bone tissue regeneration engineering.^{[ 41} , ^{42 ]} Consequently, implant materials have become a focal point for researchers and clinical practitioners.^{[ 43} , ⁴⁴ , ^{45 ]} The ideal implant material should possess comprehensive properties such as high mechanical strength, low E‐modulus, excellent abrasion and corrosion resistance, and good biocompatibility. As essential grafts in bone tissue regeneration engineering, metal materials have a profound research foundation. Among metals, titanium alloys are considered one of the most suitable materials for orthopedic implants and meet the requirements for implant materials better than other competing materials, such as stainless steels, Cr‐Co alloys, CP niobium, and tantalum.^{[ 46 ]} They display good biocompatibility and sufficient mechanical strength to withstand the mechanical loads applied to them.^{[ 47 ]} However, their robustness and high stiffness can result in a mismatch of E‐moduli between implants and the surrounding bone, leading to issues such as stress shielding‐induced fractures and Cage subsidence.^{[ 48} , ^{49 ]} Hence, maintaining an appropriate equilibrium between strength and hardness to best match the properties of the bone is crucial.

The E‐modulus is an inherent property that is not easily altered. Porous materials have been developed and applied to further reduce the E‐moduli of titanium alloys because the amount of material supporting the same cross‐section area is significantly less than that of block materials. Hence, deformation will be greater if stress increases, and stiffness will be reduced. Porous titanium alloys reduce stiffness mismatch to alleviate inherent issues of large metal biomaterials. Moreover, they can achieve better biological fixation by enhancing osseous tissue ingrowth into the pores of the implant, ensuring even stress transfer between bone and the implant to achieve complete bone ingrowth and long‐term biological fixation.^{[ 50 ]} Currently, there are various methods for producing porous titanium alloys, such as powder metallurgy,^{[ 51 ]} gas foaming,^{[ 52 ]} slurry sintering,^{[ 53 ]} spark plasma sintering,^{[ 54 ]} freeze‐drying,^{[ 55 ]} and rapid prototyping.^{[ 56 ]} One of the more well‐known methods is the 3D printing of porous titanium alloys fabricated via EBM, which has found extensive clinical applications.^{[ 57 ]} It is additive manufacturing and can customize the shape of materials. Customizing the porosity makes its mechanical properties and structure better suited to bone tissue for treating related diseases (such as bone defect and bone tumor resection, etc.).^{[ 28} , ^{58 ]} In addition, this method has the advantages of environmental friendliness, one‐stop manufacturing, and low logistics cost. However, the drawbacks include regular porosity and relatively high customization costs.

The current study employed a gel‐casting method to prepare porous titanium material with pore sizes ranging from 200 to 600 µm and porosities of 60–80% in a high‐temperature sintering furnace.^{[ 27 ]} The irregular porosity and microstructure of this material closely resemble the structure of trabecular bone. The mechanical testing results demonstrate that the mechanical properties and E‐modulus of this material are conducive to bone ingrowth. The irregular porous scaffolds have the ability to simulate the complex and anisotropic microstructures of bone tissues, which can help cell adhesion and regeneration, and promote osseointegration. In vitro and in vivo experiments demonstrate that this material exhibits good biocompatibility and bone integration performance, suggesting significant potential for clinical applications in orthopedics. In addition, this material has the advantage of low production cost and high cleanliness, making it suitable for large‐scale production.

For orthopedic implant materials, mechanical performance is a primary consideration, which entails properties closely matching those of human bone tissue and providing sufficient mechanical strength.^{[ 59 ]} In this research, the biomimetic porous titanium alloy material exhibited a compressive strength exceeding 40 MPa, which is higher than that of normal cancellous bone (2–5 MPa).^{[ 60 ]} The tensile strength of the material exceeded 23 MPa, and the shear strength exceeded 30 MPa, indicating excellent mechanical performance that is sufficient for mechanical support. Furthermore, the E‐modulus of the material ranged from 1 to 3.5 GPa, which closely matches the E‐modulus of normal cancellous bone (2 GPa),^{[ 60 ]} thus preventing the occurrence of “stress shielding” and ensuring good biological fixation between the material and surrounding bone tissue.

Good biocompatibility is a prerequisite for biomaterials used in medical implants.^{[ 61 ]} In the in vitro experiments in the current study, MC3T3‐E1 cells were co‐cultured with three different materials. An evaluation of cell morphology, proliferation, and vitality was performed. The results show that the cells in all three groups continued to proliferate, adhering to the material surfaces and pores, and exhibited extended pseudopodia, forming extensive and smooth connections. This indicates that all three materials possess good biocompatibility. Moreover, an essential parameter for orthopedic implants, namely bone integration, was also assessed in the experiments. ALP staining and activity detection, mineralized nodule staining, RT‐qPCR, and Western blot of osteogenesis‐related factors (including ALP, RUNX2, COL‐I, OC, OPN, OPG, and OSX) were conducted to show the osteogenic differentiation. ALP production is a recognized standard for early bone formation.^{[ 62 ]} The results of the ALP staining and activity examination indicate that ALP activity is enhanced by the increase in time (from 7 to 14 d). There were significant differences among the three groups at each observation time, and the biomimetic porous Ti6Al4V group showed the highest activity, suggesting that the biomimetic porous Ti6Al4V material is conducive to early bone formation. Mineralized nodule staining at 28 d revealed the later‐stage bone development,^{[ 62 ]} with a considerable number of mineralized nodules observed in the biomimetic porous Ti6Al4V and 3D‐printed Ti6Al4V groups, indicating successful later‐stage bone development. In addition, the examination of osteogenesis‐related factors for both gene (RT‐qPCR) and protein (Western blot) levels showed that the biomimetic porous Ti6Al4V group has significantly higher levels of ALP, RUNX2, OC, OPN, OPG, and OSX compared to the other two groups. Combined with the previously mentioned results, these findings demonstrated the excellent osseointegration property of the biomimetic porous Ti6Al4V material.

In the biological environment, various factors influence the bone integration process, including host bone density, mechanical stress, and hormone secretion. Therefore, in vivo experiments are essential for evaluating material bone integration performance.^{[ 63 ]} In this experiment, a rabbit bone defect model was established. Imaging, histological analysis, and mechanical testing were conducted to assess osseointegration. The results of the X‐ray and Micro‐CT indicated that the biomimetic porous Ti6Al4V achieved excellent bone integration, with a notable amount of new bone tissue observed. Bone ingrowth was good at the surgical site, without implant dislocation and fracture. Not only did the new bone grow well on the implant–bone interface, but it also grew inside many pores of the implant. The quantitative analysis of the key indexes of bone growth illustrated that compared with the 3D‐printed Ti6Al4V group, the BV/TV ratio and Tb.N were remarkably increased, whereas Tb.Sp was markedly decreased in the biomimetic porous Ti6Al4V group at each time point (4 and 12 weeks). Consistent with the Micro‐CT results, histological staining of the hard tissue sections showed extensive new bone ingrowth into the material, indicating successful osseointegration of the biomimetic porous Ti6Al4V. Push‐out testing results indicated that the axial force of the biomimetic porous Ti6Al4V group was remarkably greater than that of the 3D‐printed Ti6Al4V group at each time node (4 and 12 weeks). These data confirmed both materials had good bonding with host bone via osseointegration, and the biomimetic porous Ti6Al4V had an advantage over the 3D‐printed Ti6Al4V regarding osseointegration in vivo. In the microstructure study, it can be seen that pores of irregular shape are formed between the granules in the biomimetic porous Ti6Al4V scaffold, which have tortuous walls with numerous lateral channels going through the granules. The presence of irregular‐shaped pores with tortuous walls may favor osseointegration.

In spinal fusion surgery, Cages play a crucial role as essential substitutes for intervertebral discs, ensuring the postoperative fusion of spinal segments.^{[ 64} , ^{65 ]} An ideal Cage should possess excellent mechanical properties, biocompatibility, superior osteoconduction performance, fatigue resistance, and the ability to achieve direct bone‐implant interface osseointegration without bone grafting. Although PEEK Cage is widely used in intervertebral fusion owing to its adequate mechanical and 3D‐printing capabilities, its biological activity performance remains suboptimal.^{[ 66 ]} Building upon the confirmation of the material's excellent biocompatibility and in vitro and in vivo osseointegration performance, this study designed a biomimetic titanium alloy Cage. An ACDF ovine model was established to compare with PEEK and 3D‐printed Ti6Al4V Cages, demonstrating its bone ingrowth and intervertebral fusion capabilities. In contrast to easily establishing osseointegration in a bone defect rabbit model under static conditions, the in vivo experiments on bone ingrowth, intervertebral fusion, and biomechanics in the ACDF ovine model aimed to anatomically recreate the high‐functioning cervical spine segment as a more appropriate model for simulating human cervical motion segments.^{[ 67 ]} In the ACDF experiment, at 3 and 6 months postoperatively, Micro‐CT results revealed that osteogenesis was more evident in the biomimetic porous Ti6Al4V and 3D‐printed Cages than in the PEEK Cage. The 3D reconstruction Micro‐CT images showed that the new bone tissue grew and interconnected within the Cage with sufficient bone growth on its surface. The changes were more significant at 6 months than 3 months after surgery. Histological staining of hard tissue sections also demonstrated consistent results, specifically for the biomimetic porous Ti6Al4V Cage group, where bone growth within the Cage was denser and more extensive. Although gaps in the Cage−vertebra interface were observed, the condition of the intervertebral fusion was slightly better in the biomimetic porous Ti6Al4V Cage than in the PEEK and 3D‐printed Ti6Al4V Cage groups. Moreover, in biomechanical testing, there were remarkable differences in the ROM during flexion–extension, lateral bending, and axial rotation among the three groups. The results showed that a biomimetic porous Ti6Al4V Cage could enhance the biomechanical stability of the cervical vertebra. In addition, the results of hematological examination and histological analysis of major organs demonstrated that the biomimetic porous Ti6Al4V Cage had good biosecurity in vivo. For the statistical methods about the all experiments, comparison of continuous data between the two groups was performed via independent Student's t‐test. P‐value < 0.05 were considered statistically significant. It means any difference between the two groups is due to systematic factors rather than to chance factors. For variability, the data in our study are all of low variability, which is a desirable result because it means that the population information can be better predicted from the sample data.

There were several limitations to this study. First, the number of animals and the observation time were limited. Thus, some postoperative complications of the implant had perhaps not yet appeared. A larger number of animals with a more extended observation period after implantation might be helpful to better assess the biosecurity of the materials and degree of cervical fusion. Second, owing to the implantation of different materials into different cervical vertebral segments in each sheep, mechanical performance testing was only feasible for individual FSUs. A comprehensive evaluation of the mechanical properties of the entire cervical spine was lacking. Third, histological sections revealed less new bone formation than radiographic images. This discrepancy may be attributed to issues related to sectioning and the production process, which will be addressed in future work to ensure more accurate preparation of histological sections.

5. Conclusion

In this study, the mechanical performance of a biomimetic porous Ti6Al4V material was investigated and compared to those of PEEK and 3D‐printed Ti6Al4V materials. The biomimetic porous Ti6Al4V material fabricated through the gel‐casting method demonstrated excellent mechanical properties, including sufficient compressive, shear, and tensile strength and an E‐modulus closely resembling that of normal bone trabecular tissue. Compared to PEEK and 3D‐printed Ti6Al4V materials, this material exhibited superior biocompatibility, bone ingrowth, and osseointegration performance. The spinal fusion Cage produced from this material showcased enhanced bone ingrowth, intervertebral fusion, and reduced cost, presenting significant potential for clinical applications in spinal surgery.

Conflict of Interest

The authors declare no conflict of interest.

Author Contributions

X.D. and X.L. contributed equally to this work and are considered as co‐first authors. XG.L., B.Z., and Z.L. contributed equally to this work and are considered as co‐corresponding authors. X.D.: Investigation, conceptualization, methodology, validation, formal analysis, and writing—original draft. X.L.: Methodology, validation, formal analysis, and data curation. Y.L.: Methodology, validation, and data curation. L.W.: Investigation and formal analysis. F.J.: Methodology and validation. F.S.: Visualization and validation. Y.M.: Methodology and data curation. C.L.: Validation and data curation. G.J.: Visualization and resources. M.W.: Resources. Z.L.: Conceptualization, project administration, and writing—review & editing. B.Z.: Conceptualization, supervision, funding acquisition, and writing—review & editing. XG.L.: Conceptualization, supervision, project administration, funding acquisition, and writing—review & editing.

Dou X., Liu X., Liu Y., Wang L., Jia F., Shen F., Ma Y., Liang C., Jin G., Wang M., Liu Z., Zhu B., Liu X., Biomimetic Porous Ti6Al4V Implants: A Novel Interbody Fusion Cage via Gel‐Casting Technique to Promote Spine Fusion. Adv. Healthcare Mater. 2024, 13, 2400550. 10.1002/adhm.202400550

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.