Six male beagle dogs with a weight between 10 and 13 kg and a mean age of 14 months were used in the present study. The animals were housed in the Animal Facility for Oral and Maxillofacial Reconstruction and Regeneration Experiments, School of Stomatology, Southwest Medical University at Luzhou, China. The appearance, behavior, reactivity and social interactions of the dogs were monitored during the entire treatment period. Each animal was kept in a separate cage with a light:dark cycle of 12:12 h and without excessive or potentially startling noises. This study was approved by the Animal Experiment Ethics Committee of Southwest Medical University, Luzhou, China (No. 2021-0528-1), and it is reported according to The ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines. All methods were performed in accordance with the relevant guidelines and regulations.
Titanium Ring and Retaining Screw
The titanium ring of the present study was first designed by 3-matic Research (Release 11 for Windows, Materialise, Leuven, Belgium) following the creation of the original STL format file. A unique design feature of the titanium ring was that its inner ring was designed as a frustum of a cone with a central hole, which matched with the nonhex implant with a morse taper (Dentium Superline, Ltd, Gyeonggi-do, Korea) and retaining screw (Fig. 6). A pore size of 500 µm was created at the top of the titanium ring, with no pores on the sides.
The 3D printer (M290, EOS, Munich, Germany) used a high-energy laser to melt the titanium metal powder into slices of the required three-dimensional shape, and then the sintering machine accumulated these slices layer by layer to obtain the required titanium ring. The retaining screw stemmed from the abutment screw (Dentium Superline, Ltd., Gyeonggi-do, Korea) and could easily secure the titanium ring to the implant after machining (Fig. 7). Subsequently, the final titanium ring and retaining screw were checked and polished. Before grafting in beagle dogs, all products were sterilized by high temperature and high pressure, reaching the requirements of a surveillance standard for sterilization (ISO 15883:2006).
Surgical Procedures
The study was designed as a pilot experimental study. All surgeries were performed under general anesthesia, and surgical interventions were performed under systemic anesthesia using pentobarbital sodium (i.m. 30 mg/kg). In the first surgery, three premolars (P2-P4 per hemimandible) were atraumatically extracted by root separation from both sides of the mandible. All animals received 125 mg Amoclav (clavulanic acid and amoxicillin, Hanvet Pet Health, Shanghai, China) as antibiotics intraoperatively and twice a day for 7 days after the surgery. Appropriate therapy was administered when the behavior of the animals indicated adverse reactions or pain. The sutures were removed 7 days after the surgery.
After 3 months of healing, the second surgery involved a midcrestal incision and full-thickness muco-periosteal flap elevation. Two standard three-dimensional ridge defects of 7.2 mm (mesio-distal and buccal-lingual) to 3.5 mm (apico-coronal) were created on P2 and P4 of each side of the mandible. Following the drill protocol of the manufacturer, the first guide drill was navigated and fixed 5 mm below the bone defect, and then the final drill was used to enlarge the diameter of the hole to 2.8 mm at 50 rpm while the autogenous bone particles were harvested. Bone level implants (Ø 3.6 mm, length 8 mm) were inserted in the hole to 5 mm to achieve adequate primary stability, with 3 mm exposed. A titanium ring was secured to the implant with a retaining screw, and a combination of deproteinized natural bovine bone mineral (DBBM) (Bio-oss, geistlish Pharma, Wolhusen, Switzerland) and autogenous bone particles was used to fill the void between the implant and the titanium ring. The flap was relaxed by releasing the periosteum and then sutured to allow for submerged healing of all implants.
Normal saline was used to irrigate the wound once a day for 3 weeks after surgery without applying anesthesia. Unfortunately, all titanium rings were exposed 3 weeks after the operation due to the dehiscence of the flap (Fig. 8). Subsequently, the experimental sites were randomly divided into two groups, and each group included p2 of one mandible and p4 on the other in the same beagle dog. In the titanium ring removal group, the small flap of soft tissue was elevated to remove the titanium ring, and then the soft tissue was sutured again after normal saline irrigation at 4 weeks. No surgical treatments were carried out in the titanium ring nonremoval group (holding group). However, daily flushing was still performed with normal saline.
Sample Preparation
After an osseointegration period of 3 months, the animals were euthanized by an intramuscular overdose of pentobarbital sodium (100 mg/kg). Block specimens of the entire implant and surrounding bone (mesiodistal 3 mm) were harvested except for loose implants after elevating the full-thickness mucoperiosteal flap. Furthermore, the trimmed mandible bone blocks containing the implants were fixed in 4% paraformaldehyde and analyzed by means of micro-CT and histology.
Micro-CT Analysis
The biopsies were nondestructively examined first with a radiographic computed microtomography (micro-CT) system. The samples were placed in a vertical position to the direction of the radiation and scanned using a micro-CT system (Venus, VNC-102; PINGSENG Health care, Kunshan, China) as follows: X-ray source, 80 kV, 80 µA; distance from the source to the object, 90 mm; distance from the source to the detector, 410 mm: slice thickness, 0.08 mm; voxel size, 0.049×0.049×0.08 mm. The micro-CT images (270 slices) were then reconstructed using 3D structural analysis software (Avatar; PINGSENG Health care, Kunshan, China). Threshold values were set for the segmentation between 5400 and 22800 for the implant and 2500 and 5400 for the bone tissue. The region of interest (ROI) was selected corresponding to the dimensions of the titanium ring, with a diameter of 7.2 mm and the base at 3 mm from the implant shoulder (V1). Bone directly under V1 with a diameter of 7.2 mm and a height of 5 mm was measured separately (V2). Then, the bone volume (BV, mm3), BV/total volume (BV/TV), bone surface/TV (BS/TV) and bone mineral density (BMD, g/mm3) were determined in V1. Only BMD was calculated in V2.
Histological Preparation
First, the bone samples were fixed for one week, and the fixation liquid was replaced overnight. Following rinsing in running tap water for 24 hours, trimming, and dehydrating in ascending concentrations of ethanol from 70–100%, the bone samples were made transparent by using ETC-3 (Combine, Changsha, China) for less than 4 hours, and then the bone samples were processed for the production of undecalcified ground sections and embedded in a mixture of methylmethacrylate, dibutyl violet phthalate and benzoylperoxide. The embedded tissue blocks were cut buccolingually along the implant axis using a slow-speed rotating diamond serrated blade (LEICA1600 VC-50, Frankfurt, Germany) into approximately 0.08 mm thick ground sections. The sections were stained with toluidine blue and photographed under a light microscope equipped with a digital imaging system (CellSens; OLYMPUS, Tokyo, Japan).
Statistical Analysis
To quantify the effects of the titanium rings on the outcomes, absolute values were used. Total bone was defined as newly formed bone (including osteoid and woven bone) and residual graft bone. All of the quantitative variables were characterized using descriptive statistics (median, minimum and maximum) instead of means and standard deviations, and a normal distribution was not assumed considering the low sample size. Kruskal–Wallis nonparametric analysis of variance was applied to estimate the differences in BAH over the buccalingual aspects of the implants in the two groups. In addition, the difference in BAH over the two groups was evaluated by the Wilcoxon rank-sum test. The Wilcoxon signed-rank test was used to estimate the differences in BMD over the V1 and V2 and in BAH over the buccolingual aspects of the implant between the removal group and the holding group, respectively. P values < 0.05 were considered statistically significant. The statistical analysis was conducted using SPSS for Windows Release 26.0, standard version (IBM SPSS, Chicago, IL, USA).