Material and methods

Seven skeletally mature, periodontally healthy, beagle dogs with complete dentitions, between 1 and 2 years of age and weighing 21-29 pounds, were used because their bone more closely approximates human bone than that of most other experimental models. The protocols were approved by the Institutional Animal Care and Use Committees (No. 2014-0275) at Texas A&M University College of Dentistry.

An electronically generated random number table was used to assign the experimental side of the maxillary arch to receive an archwire expansion appliance. The control side received no appliance. Following 10 days of quarantine, baseline records were obtained, including intraoral photographs, periapical radiographs, and alginate impressions. Periapical radiographs were taken on the experimental sides using a size 4 phosphor plate and a custom holder that stabilized the film and standardized angulations and distances ( Fig 1 ). Alginate impressions were taken using custom trays made of Triad material. Models were poured using die stone for the fabrication of occlusal radiographic guides and appliances.

Measurement of the center of resistance. A, Custom measurement jig for acquiring periapical radiographs of the experimental teeth. Two 10-mm long stainless steel wires were orientated parallel to the mesial and distal roots of the second premolar for image calibration (yellow arrows) . B, The center of resistance was estimated halfway (50%) between the root apex and alveolar crest. The length of the extension arm from the gingival margin was calculated with the formula: (Y/2) +Z, with X referring the crestal bone level, Y is the distance from the root apex to alveolar crest and Z is the distance from the crestal bone to premolar band.

The appliance was designed to produce buccal translation of the second premolar, using the canine and fourth premolar as anchor teeth. Orthodontic band material was custom pinched and welded to fit these 3 teeth. Each tooth’s inner surfaces were microabraded with 60-micron alumina particles, and several small holes were drilled with a 0.25 round bur to enhance band retention. Tubes with a 0.022 in slot size were welded and soldered to the bands on the canine (0° offset, 0° torque, and 0.051-in headgear tube) and fourth premolar (10° distal offset, –14° torque, and 0.045-in headgear tube). To stabilize the canine and fourth premolar, and to protect the second premolar from the cheek, a 0.045 in stainless steel wire was made passive in the headgear tubes. The ends were covered with solder and polished smooth for animal comfort ( Fig 2 ). A 10 mm 0.045 in wire segment was soldered perpendicular to the extension arm at the apical extent to serve as a cheek guard.

Appliance design. A and B, Activated appliance consisting of bands on the canine, second premolar, and fourth premolar, along with a passive stainless steel wire connecting the canine and fourth premolar, a vertical extension arm on the second premolar band, and a Beta-Titanium wire activated to 100 g of force. The active wire was looped at one end, tied to the apical extent of the vertical arm on the second premolar and engaged through the fourth premolar tube. C, Coe-Pak on vertical extension arm for cheek protection. D, Passive wire bonded to first, second, and third premolars during 3-week consolidation period.

A stainless steel vertical extension arm 1.5 mm in diameter was welded onto the second premolar band ( Fig 2 , A ) and positioned 1 mm distal to the crown tip. The distal premolar root is wider and typically tipped distally, while the mesial root extends vertically ( Fig 1 , B ). Since the furcation was located coronal to the alveolar crest, the experimental second premolar was treated as a single-rooted tooth, and the COR was estimated halfway (50%) between the root apex and alveolar crest ( Fig 1 , B ). The apical aspect of the extension arm was positioned at the estimated COR, which were determined by periapical films that were imported into Dolphin Imaging and traced.

The animals were sedated, and vital signs were monitored throughout the procedure. Small notches were made in the cusp tips of the second premolars with a mosquito shaped micro-diamond bur (Brasseler USA, Savannah, GA) to serve as references for the caliper measurements. Amalgam markers, approximately 1.5 mm in diameter, were placed using a 330 carbide bur in the canine, second premolar, and fourth premolar on the experimental side for radiographic measurements of tooth movement. Two 6 mm wide Imtec miniscrew implants were placed in the midline of the palate mesial and distal to the second premolars. They served as references for the radiographic measurements. Retention grooves were placed in each of the banded teeth with a 330 carbide bur. The teeth were etched with 37% phosphoric acid for 30 seconds and rinsed for 10 seconds. The bands were cemented, and excess cement was removed.

A 0.021 in × 0.025 in beta-titanium wire was bent to act as a cantilever delivering a 100-gram buccal force on the second premolars. This force level was based on a previous dog study evaluating the effects of expansion. The force level was regularly checked using a Correx Tension gauge. The wire was first engaged in the second premolar tube and cinched distal to it. The loop at the end of the wire was tied to the tip of the vertical extension with a 0.012 in stainless steel ligature tie. Coe-Pak periodontal dressing was applied to the apical extent of the vertical arm to prevent irritation of the mucosa. The other end of the wire was cinched gingivally distal to the fourth premolar tube. The animals were maintained on a soft diet for the duration of the experiment to prevent appliance loss or damage.

Data were collected every 2 weeks, starting on the day of appliance placement. Interpremolar width was measured in triplicate at the cusp notches using a digital caliper and averaged (replicate analyses produced an intraclass correlation [ICC] of 0.99). Duplicate measures of pocket depths were taken at the mesiobuccal, buccal, distobuccal, mesiolingual, lingual, and distolingual aspects of the second premolars and averaged. Force levels were recorded, a new segment of wire was activated to 100 grams, and occlusal radiographs were obtained.

After the radiographic and caliper measurements showed approximately 1.5 mm of buccal tooth movement, the appliance was made passive by removing the active beta-titanium wire and second premolar band, and bonding a passive 0.030 in stainless steel wire to the first, second, and third premolars. Occlusal radiographs, caliper measurements, and photographs were obtained, this was followed by a 3-week consolidation period, during which no tooth movement was performed. Three weeks was chosen because new alveolar woven bone becomes evident after 2 weeks, and immature lamellar bone can be seen after 3-4 weeks.

Regions of new bone formation were identified using fluorescent bone labels (calcein and alizarin) that were administered 3 times to each dog at 2-week intervals; thus, making it possible to evaluate new bone formation during tooth movement and the consolidation phase. Following the consolidation period, final records were obtained, the animals were sacrificed, and the maxilla was stored in 4% paraformaldehyde at 4°C.

Occlusal radiographs were imported into ViewBox 4 and calibrated using a 20-mm ruler ( Fig 3 , A ). The perpendicular distances from the amalgam markers on the canine, second premolar, and fourth premolar were measured to a sagittal reference line drawn passing midway through the 2 palatal miniscrews. Repeated measurements, performed on separate days, produced ICCs of 0.97, 0.99, and 0.99 for the canines, second premolars, and fourth premolar measures, respectively. The initial and final maxillary models were laser scanned using Ortho Insight 3D to quantify tipping. Three standardized landmarks were digitized ( Fig 3 , B ) and the angle connecting them was measured (ICC = 0.982).

Radiographic measurements of tooth movement. A, Radiographic measurements of tooth movement: reference line ( red line ) was constructed through the midpoints of the 2 palatal implants. The perpendicular distances from amalgam markers on the canine ( yellow line ), second premolar ( black line ), and fourth premolar ( blue line ) to the reference line were measured. B, Dental casts were laser scanned and digitized to evaluate the tipping of the second premolars, calculated as the angular difference between preexperimental and postexperimental measurements. Second premolar angulation was defined by connecting the following points: r (the most occlusal midline point on the palatal rugae adjacent to the second premolar), c (the most cervical point on the palatal aspect of the second premolar, taken at its mesiodistal center), and t (the cusp tip).

With the overlying soft tissues intact, the maxilla was sectioned to produce blocks (26 mm wide × 19 mm high) containing the first, second, and most of the third premolars. The blocks, which included 3-4 mm of bone apical to the root tips, were placed into microscopic computed tomographic (microCT) tubes with an internal diameter of 27 mm and kept immersed in 0.5% paraformaldehyde. The tubes were loaded into the microCT 35 desktop scanner and scanned at 30 μm resolution, using 55 kVp, 145 μA, and 600 ms integration time. Three-dimensional reconstructions, as well as 2-dimensional slices (30 μm thick), were generated using the Scanco microCT v6.0 software. For three-dimensional reconstruction, the grayscale images were smoothed by a Gaussian filter with a sigma value of 0.9 and a support value of 1. The threshold boundaries for the scans were set between 260 and 1,000 Hounsfield units.

The following linear measurements were taken from 2-dimensional axial slices of both the experimental and control premolars: (1) buccal bone height measured from the most lingual aspect of buccal bone at the level of the apex to the mesiodistal center of the buccal bone crest on both mesial and distal roots; (2) total tooth height measured from the cusp tip to the root apex; (3) total root height measured from the center of the pulpal canal at the furcation to the root apex; and (4) buccal bone thickness measured from the most lingual to the most buccal aspect of buccal bone at the cervical (3 mm apical to the level of the furcation), mid (exactly halfway between the cervical and apical measurements), and apical (measured at the tip of the apex) levels. Duplicate measurements taken on different days were averaged (ICCs of the duplicate measures ranged from 0.97-0.99).

The experimental roots were randomly assigned for fluorescence microscopy or decalcified staining. Two control specimens were randomly allocated for staining, and 5 were allocated for fluorescence microscopy. The specimens analyzed with fluorescence microscopy were fixed in 4% paraformaldehyde, dehydrated in graded alcohol, embedded in methyl methacrylate, and allowed to polymerize. Blocks were sectioned with an IsoMet diamond saw to a thickness of approximately 150-200 μm. Sections were made in a coronal plane, parallel to the long axis of the root, polished to a thickness of about 75-100 μm, and examined using the Leica TCS SP5 confocal microscope.

Specimens prepared for staining were decalcified in 0.5 M calcium disodium ethylenediaminetetraacetate dehydrated in graded alcohol and lastly butyl-alcohol, and infiltrated and embedded in paraffin blocks. The blocks were sectioned with a microtome parallel to the long axis of the root at a thickness of 5-10 μm and mounted on glass slides. Hematoxylin and eosin(H&E) and tartrate-resistant acid phosphatase (TRAP) staining was performed, as well as bone sialoprotein (BSP) immunostaining.