Production of the additive-manufactured Ti-6Al-4V/PEEK composite porous cage
The innovative composite porous interbody cage was produced using the polymer core material PEEK, which is a favorable biomedical material with high strength and toughness. The surface of the PEEK polymer substrate was modified through laser grooving to enhance the bonding strength of the subsequent coating layer (Fig. 1-A and B). The surface was then coated with a metallic (Ti alloy) interfacial layer through low-temperature arc ion plating and plasma spraying to ensure sufficient thickness of the metallic interface layer (300 μm).
Subsequently, a 3-mm-thick 3D Ti-6Al-4V porous scaffold was constructed above the metallic interfacial layer with intended porosity parameters by using the SLM EOSINT M 270 model (EOS GambH-Electro Optical Systems, Krailling, Germany).
The microstructure of the metallic interfacial layer was observed through multifunctional field-emission scanning electron microscopy (SEM), and its chemical composition was analyzed through SEM with energy-dispersive spectroscopy (SEM-EDS). The shear strength of the bonding interface between the metallic layer and PEEK substrate exceeded 30 MPa (Fig. 1-C).
The results of the compressive mechanical test on the composite porous interbody cage are shown in Fig. 1-F and Table 1, and those of the torsional mechanical test are shown in Fig. 1-G and Table 2.
The porosity rate was defined using the following equation:
where P% is the porosity rate, Di_porous is the measured density of the metallic layer (measured mass divided by measured volume), and Dsolid is the density of the metallic layer in the solid condition. The parameters of different porosity rates are listed in Table 3. Schematics of the porous structure and interporous distance are illustrated in Fig. 1-D and E. Based on these parameters, composite cages with metallic layers of different porosities were produced. The geometric parameters of metallic and composite cages in this study were derived from the commercialized PEEK cage (Anterior Cervical Interbody Fusion Cage®, BAUI biotech, New Taipei City, Taiwan).
This animal study was approved by the Ethics Committee of the Biomedical Technology and Device Research Laboratories of Industrial Technology Research Institute in accordance with National Animal Welfare Legislation (Approval No. PIG-1040106), and the study protocol conformed to the National Institute of Health guidelines for the use of laboratory animals. Twenty 5-month-old female pigs (Lanyu 50, Taiwan) from different litters, weighing 35–45 kg, were used in this study following Zou et al.’s protocol . All the pigs were obtained commercially from PigModel Animal Technology Co., Ltd. (Miaoli, Taiwan). Each pig underwent anterior intervertebral lumbar fusion at three levels: L2–L3, L4–L5, and L6–L7. Each level was randomly implanted with one of the five test cages. Each of the five groups comprised 12 specimens. In the first group, we tested a commercialized pure PEEK interbody device with an autologous iliac crest bone graft (PEEK_NonP, group 1). In the second group, a Ti alloy/PEEK composite cage with nonporous Ti alloy endplates embedded with an autologous iliac crest bone graft was tested (Comp_NonP, group 2). The third, fourth, and fifth groups used composite cages embedded with an autologous iliac crest bone graft with porosities of 40%, 60%, or 80% on both Ti alloy endplates (Comp_40%P [group 3], Comp_60%P [group 4], and Comp_80%P [group 5]). Each fusion segment was additionally secured with pedicle screws (Lumbar Trans-Pedicle Screw Fixation System®; Wiltrom Biotech Co., Ltd., New Taipei City, Taiwan) in implanted levels. All pigs were kept in single pens throughout the 6-month observation period and were subsequently euthanized. Tetracycline (20 mg/kg; SIGMA-Aldrich, Merck Group, Germany) was injected intravenously at 4 and 2 weeks before euthanasia to label bone growth.
The pigs were euthanized under deep anesthesia with intravenously injected KCl (1–2 mEq/kg). Plain radiographs of anteroposterior and lateral views and computed tomography (CT) of the lumbosacral spine were taken at euthanasia. The whole lumbar spinal column from L1 to L7 was removed en bloc, stripped of the soft tissue, transported to the laboratory, and stored at −20°C for further examination. The pigs were bred for scientiﬁc purposes and handled according to the regulation of the Institutional Animal Care and Use Committee (IACUC: PIG-106022) on animal experimentation.
Before anesthesia, the pigs were premedicated intramuscularly with 5 mg/kg Zoletil 50 (Zolazepam + Tiletamine) + 2.2 mg/kg Xylazine for induction. After orotracheal intubation, anesthesia was maintained through the inhalation of isoﬂurane (1.5%). Cephalosporin (1 g, intravenously) was administered 30 minutes before surgery as a prophylactic antibiotic.
Under aseptic conditions, the autologous bone graft was harvested from the right iliac crest with the pig placed in a prone position and prepared as morselized cancellous bone chips. Under ﬂuoroscopic control, the intervertebral space in implanted levels was identiﬁed before surgical intervention. The facet joints of the neighboring vertebrae at this level were exposed through a posterior midline incision and paraspinal bilateral intramuscular approach. Pedicle screws (5 mm in diameter and 30 mm in length) were inserted into the neighboring vertebrae transpedicularly. The incision in the back was carefully sutured and the pigs were closely cared for 1 month to allow complete recovery. After the pigs' condition become stable, we performed the 2nd stage operation. With the pigs placed in the left decubitus position, a retroperitoneal anterior approach was used. The rectus abdominis muscle and its sheath were incised and retracted. The innermost layer, the fascia of transverse abdominis, was carefully dissected to prevent damage to the peritoneum lying immediately underneath. After the peritoneum and its contents were separated and retracted, the quadratus lumborum and psoas major muscles could be viewed. The anterior lumbar spine was easily identiﬁed by its thick and shiny anterior longitudinal ligament. After the ligation and cutting of segmental vessels, the L2–L3, L4–L5, and L6–L7 intervertebral discs were excised together with the cranial and caudal endplates, ring apophysis, and part of the anterior longitudinal ligament. Thereafter, the bone graft was morselized and packed into the central holes of the respective interbody cage devices. The fusion device–bone graft complexes were then implanted at each intervertebral disc. After insertion of the three implants, the abdominal muscles and the rectus abdominis sheath were carefully sutured, and the skin was closed using running sutures. Prophylactic cephalosporin (1.0 g, intravenously) and analgesic ketorolac (30 mg, intramuscularly) were administered before and immediately after surgery. All pigs were kept in individual pens and fed a normal diet containing 1.4% calcium and 0.7% phosphorus (percent of food weight). Pain control medication was administered for 7 days postoperatively (400 mg ibuprofen, two tablets/day) and as required afterward.
After euthanasia, five specimens were retrieved from each group and scanned using micro-CT (Skyscan 1272® at 8-μm/pixel, Bruker Micro-CT, Kontich, Belgium). A 360° scan with a high voltage of 90 kVp, current of 111 μA, and output of 10 W was conducted. Image reconstruction was performed using a graphics processing unit-based reconstruction software, GPU-NRecon. Ring-artifact and beam-hardening correction were also performed using GPU-NRecon. The reconstructed cross sections were reoriented, and regions of interest were further selected. Automatic thresholding and 3D/2D structure and pore analyses were performed using CTAn software. We performed the analysis with 1.4-mm (113 slices) images. Tissue volume (TV, mm3), bone volume (BV, mm3), percent bone volume (BV/TV, %), bone surface (BS, mm2) area, and bone surface area per total volume (BS/TV, 1/mm) were measured 0–500 mm above the metallic implant bone. With the nonporous implant as a template, the outer bone was defined by those exterior to the nonporous implant surface (including the bone outside the doughnut and bone of the doughnut hole), as illustrated in Fig. 2. The inner bone was defined by the bone formed interior to the surface of the implant (the doughnut body). 3D visualization was performed using Avizo software (Version 9.4, Thermo Fisher Scientific, MA, USA).
Three specimens were retrieved from each group and scanned using backscattered-electron SEM (BSE-SEM) at 6 months postoperatively. The specimens were decalcified before the procedure, embedded using Technovit 9100 (Kulzer, Wehrheim, Germany), and then cut into thin 1-mm slices. The slices were carefully polished and coated with carbon for BSE-SEM (DSM940; Carl-Zeiss AG, Oberkochen, Germany) analysis. Multiple images were merged using Photoshop CC (Adobe, San Jose, CA, USA). We then converted images into grayscale and analyzed them using the 2D analysis function of CTAn software (Bruker Skyscan, Konitch, Belgium). We defined the implant area as a region of interest (bone ingrowth area). Next, we expanded the bone ingrowth area by approximately 500 mm and then excluded the bone ingrowth area. The surrounding area was defined as the bone ongrowth area. Morphometric indices of ingrowth area, ongrowth area, and total area (ingrowth + ongrowth) were analyzed.
Four specimens were retrieved from each group for histological analysis at 6 months postoperatively. All these harvested samples were fixed in 10% formalin for 14 days and sequentially dehydrated with increasing concentrations of ethanol (70%, 95%, and 100%) for at least 1 day and infiltrated for 5 days with polymethylmethacrylate. After embedding, the samples were cut horizontally, perpendicular to the axis of bony endplates, at the level of the respective bone–implant interfaces. The sections were cut to approximately 150 μm in thickness by using an IsoMetTM Low Speed saw (Buehler, Lake Bluff, IL, USA) and ground to 60 μm with a grinding and polishing machine. The ground sections were then stained with Sanderson’s Rapid Bone Stain (Dorn & Hart Microedge Inc., Loxley, AL, USA) and then counterstained with acid fuchsin. All bone–implant interfaces were examined carefully under a light microscope. In addition, the sections were examined through fluorescence microscopy to identify new bone formation, which was labeled with tetracycline.
All experimental data are presented as the mean ± standard deviation, with values from more than three experiments. The Wilcoxon rank sum test and Fisher’s exact test were used for nonparametric analysis. Data with more than two groups were compared through one-way analysis of variance and Tukey’s post hoc test for repeated measures. The correlation was examined as Pearson correlation and Spearman correlation; p < 0.05 was considered statistically significant. The power value was set to 0.8. Sample size calculation showed that the animal study required 11 in each group based on data by Zou et al. . Statistical analysis was performed using PASW software (version 18.0; SPSS, Chicago, IL, USA).