In total, 24 Ti implants that were 3.3 mm in diameter and 10.0 mm long were used in this study. Twelve were from Deep Implant System, Inc. (Seongnam, Korea) and had been subjected to coarse-grit sandblasting and acid etching. These sandblasted and acid-etched (SLA) implants served as control implants. The remaining 12 implants were SLActive implants from Institut Straumann AG (Basel, Switzerland). These modified SLA (modSLA) implants not only underwent grit blasting and acid etching, but were also rinsed under nitrogen protection and then sealed in glass tubes containing isotonic sodium chloride (NaCl). This modification increases the hydrophilicity of the implants, as it yields a hydroxylated/hydrated surface that retains high surface energy. Of the 12 implants in each group, 2 were used for surface analysis and 10 were implanted in rabbit tibiae.

To separate contact and distance osteogenesis physically, half of the SLA and modSLA implants were implanted within grade 4 Ti tubes that were 4.0 mm in diameter and 6 mm long and had a wall thickness of 0.2 mm (Figure 1A). Thus, these Ti tubes had an inner diameter of 3.6 mm. Implantation of a SLA or modSLA implant 3.3 mm in diameter within such a Ti tube therefore left a 0.15 mm gap between the implant surface and the inner Ti tube surface (Figure 1B-D).

Two SLA and modSLA implants and 2 Ti tubes were subjected to surface analysis using field emission scanning electron microscopy (FE-SEM) and energy-dispersive spectroscopy (EDS). The former was used to generate overall surface images (S-4700, Hitachi, Tokyo, Japan), while the latter was used to analyze the components and contents of the modified surfaces (EX220, Horiba Ltd., Kyoto, Japan).

This study was approved by the Animal Research Committee of Seoul National University, Seoul, Korea. All animal experiments were conducted in accordance with the guidelines of the Institute of Laboratory Animal Resources at Seoul National University (approval number: SNU-130619-4) and the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines for reporting in vivo animal experiments [7]. Five male New Zealand white rabbits (age, 1–2 years; body weight, 2.6–3.0 kg) with no signs of disease were used. All rabbits were anesthetized via an intramuscular injection of tiletamine/zolazepam (15 mg/kg, Zoletil^® 50, Virbac Korea Co., Ltd., Seoul, Korea) and xylazine (5 mg/kg, Rompun™, Bayer Korea Ltd., Seoul, Korea). Before surgery, the skin over the proximal tibia was shaved and washed with betadine, after which an antibiotic (Cefazolin, Yuhan Co., Seoul, Korea) was administered intramuscularly. Lidocaine was injected locally into each surgical site. The skin was then incised, and the tibiae were exposed by muscle dissection and periosteal elevation. Drills and profuse sterile saline irrigation were used to prepare implant sites on the flat tibial surface.

In 4 rabbits (the handling of the fifth rabbit is described below), drilling was performed bicortically, with final diameters of 4 mm at the upper cortical bone (to prevent physical involvement, including friction) and 3 mm at the lower cortical bone (to promote stability). If used, the Ti tubes were inserted prior to implant placement; otherwise, implants were installed directly into the tibial bone as shown in Figure 1B. Each of the 4 rabbits received 2 sets of implants: 2 were SLA implants and 2 were modSLA implants. One SLA implant and 1 modSLA implant were inserted within Ti tubes. Thus, in total, 8 SLA and 8 modSLA implants were inserted in the 4 rabbits, and 4 implants of each group were encased in Ti tubes. The implants were inserted into the right and left tibiae of each rabbit (Figure 1C) according to the split-plot design shown in Figure 1D. The muscle and fascia were then sutured with resorbable 4-0 Vicryl sutures, and the outer dermis was closed with a nylon suture.

The rabbits were housed separately after surgery. After a 4-week period of bone healing, they were anesthetized and sacrificed via an intravenous overdose of potassium chloride. The tibiae were exposed, and the implants were surgically removed en bloc with an adjacent bone collar. The implants/bone were immediately fixed in 10% neutral formaldehyde. Histomorphometry specimens were prepared by embedding the implants and bone in light-curing resin (Technovit 7200 VLC, Kultzer, Wehrheim, Germany) [8] and then treating them according to the method described by Donath and Breuner [9]. Thus, undecalcified specimens were ground to a thickness of 50 μm, stained with hematoxylin and eosin (H&E), and subjected to a general histological evaluation under a light microscope (Olympus BX, Olympus, Tokyo, Japan). The image analysis software (Kappa PS30C Imagebase, Kappa Opto-electronics GmbH, Gleichen, Germany) that was associated with the light microscope was used to calculate the bone-to-implant contact (BIC) ratio for each implant. Thus, a stretch of the implant surface measuring 2 mm in length was examined to determine how much of that length was in direct contact with bone. It should be noted that the top of the 2-mm stretch was level with the upper surface of the cortical bone (Figure 1E). The BIC ratios of the left and right sides of each implant were determined. The mean BIC ratio for each implant was then calculated.

In the fifth rabbit, 4 SLA implants were placed in both tibiae without Ti tubes as described above, except that the final drill was used to make holes 3 mm in diameter and the implants were only engaged at the upper cortex. The rabbit was sacrificed after 10 days, and the implants were removed by counterclockwise rotation after demineralization. Paraffin wax-embedded sections were generated from the bone blocks from which the implants had been removed, and the sections were then subjected to immunohistochemistry according to a standard protocol. Thus, sections were deparaffinized and rehydrated in a graded alcohol series, followed by antigen retrieval in 10 mM Tris/1 mM ethylene diamine tetra-acetic acid (pH 9.0) at 98°C for 20 minutes (microwave oven). The sections were then washed under running tap water for 10 minutes and treated with 0.3% hydrogen peroxide in methanol for 10 minutes to block intrinsic peroxidases; the method followed the protocol that was supplied with a mouse- and rabbit-specific horseradish peroxidase/3,3′-diaminobenzidine (DAB) avidin-biotin complex immunohistochemistry detection kit (ab64264, Abcam^®, Cambridge, UK). Subsequently, the sections were blocked using a protein blocking reagent from the same kit and then incubated overnight at 4°C with an anti-rabbit antibody specific for bone morphogenetic protein (BMP)-2 (1:800 dilution, ab6285, Abcam^®). The next day, the sections were washed 3 times for 5 minutes in 0.01 M phosphate-buffered saline (PBS) (pH 7.2–7.4) and incubated with a biotinylated goat anti-polyvalent secondary antibody (1:200 dilution, ab64264, Abcam^®) at room temperature for 20 minutes. After 3 other 5-minute washes in 0.01 M PBS, the sections were incubated at 37°C for 20 minutes with streptavidin peroxidase from the detection kit. After 3 additional 5-minute washes in 0.01 M PBS, the sections were incubated for 10 minutes at room temperature with a mixture consisting of 20 mL of 2.5% DAB chromogen and 1 mL of DAB substrate from the detection kit. They were then counterstained with hematoxylin, dehydrated through ethanol and xylene, and mounted with Permount under a cover glass. The remaining sections from the bone blocks were subjected to Masson trichrome staining.

Statistical analyses were performed using R 3.3.2 statistical software (R Foundation for Statistical Computing, Vienna, Austria). The effect of different implant surfaces and the presence or absence of Ti tubes on the BIC ratio was assessed using 2-way analysis of variance (ANOVA). When the 2-way ANOVA indicated a significant difference between the groups, pairwise multiple comparisons were conducted using the Tukey test. The level of significance was set at P=0.050.