Animals
Six young adult male Wistar rats (9 weeks old, 270-350g) were used in this study. Weekly surveillance was performed in order to notice any changes in weight and guarantee normal feeding. Prior to the experiment, the minimum sample size was calculated using a previous split-mouth study[18, 19]. The primary parameter considered was the mesial displacement of the first molar after 31 days of orthodontic force application. A power analysis in the software G*Power 3.1 (Düsseldorf, Germany) suggested a sample size of 6 animals per group (OF and no OF side) for a repeated measures MANOVA when assuming 95% power with α=0.05.
The animals were housed in three cages under continuous temperature (23º C), a 12-hour schedule alternating light and dark, and standard rat maintenance diet. All animal experiments were performed in the Laboratory Animal Center and the Molecular Small Animal Imaging Center (MoSAIC) of KU Leuven, Belgium, with the approval of KU Leuven Ethical Committee for Animal Experimentation (P197/2019) and in accordance with the EU Directive 2010/63/EU and ARRIVE 2.0. Guidelines.
Interventions
Tooth movement was performed by following a previously published protocol[18]. Briefly, on one randomly-allocated side of the maxilla, a self-drilling mini-screw (2.5 mm x 1.3 mm x 5 mm, DEWIMED, Tuttlingen, Germany) was implanted in the alveolar bone, approximately 2 mm distal to the upper incisors and with an angulation of 45° to the occlusal plane, in order to provide bone anchorage. Healing time of 3 weeks was allowed to ensure stability[20]. Afterwards, a constant orthodontic force (OF) of 25 cN was loaded between the mini-screw and the upper first molar by using a Sentalloy closed coil spring (Ultra-light, Dentsply GAC, Rochecorbon, France) (Figure 1A). At the contralateral side, no OF was applied (no OF side). At this moment, the animals were 12 weeks old.
All interventions were conducted by the same investigator, firstly under sedation of 2.5-5% isoflurane (1000 mg/g, Iso-Vet®, Dechra, Skipton, UK), followed by anesthesia of 100 mg/ml IP ketamine (80 mg/kg, Nimatek, Bladel, Netherlands) and 2% xylazine (10 mg/kg, XYL-M®, V.M.D, Arendonk, Belgium). After the interventions, soft diet and analgesic medication (0.05 mg/kg Buprenorphine) were supplied for 3 days. To prevent unnecessary animal suffering and potential appliance loss, daily surveillance of body weight and intraoral examination under sedation with isoflurane was performed. The rat grimace scale[21] was used to evaluate whether animals were in pain.
Micro-CT
Animals were longitudinally followed up with micro-CT at five time points: right before the application of OF (T0, baseline), and 10 (T1), 17 (T2), 24 (T3), and 31 (T4) days after OF. A low-dose, high-resolution in vivo micro-CT (Skyscan 1278, Bruker, Kontich, Belgium) was used for image acquisition. A high-resolution scan protocol was used at 65 kVp, 500 μA, and 180° with an angular rotation step of 0.5°, resulting in an exposure time of 50 ms. A 1 mm aluminum filter was used to eliminate the beam hardening effect. Flat field correction was performed for calibration based on an empty field of view previous to the actual scans. Animals were placed under inhalation anesthesia of 2.5-5% isoflurane during image acquisition. After image acquisition, the image stacks were reconstructed with NRecon software (version 1.7.1, Bruker). Correction for post-alignment, and ring-artifacts reduction were optimized per scan if needed. Smoothing level and beam hardening were applied with values of 0 and 10%, respectively. The dynamic image range of histogram was set from 0.003 to 0.03.
Longitudinal assessment of OTM
The dynamic changes of OTM were volumetrically evaluated based on the spatial displacement of the teeth from T0 to T4, using T0 as a baseline without depending on other external reference structures. Rigid voxel-based registration steps were performed before the evaluation to ensure that all teeth were in the same coordinate system and were temporospatially comparable at each time point.
First, the follow-up micro-CT scans at T1, T2, T3 and T4 were manually superimposed with the corresponding baseline scan at T0 based on the maxillary structures as a reference, followed by an optimized automatic superimposition by MTM Scaffold Strain (KU Leuven, Leuven, Belgium)[22, 23]. The structural compatibility of maxillary anatomical reference points was inspected to verify the validity of the manual and automatic superimposition. Second, the maxilla, mini-screw, first and second molars at each time point were delineated by the same investigator and saved as volumes of interest (VOI) in CTAnalyser software (version 1.17.5, Bruker, Kontich, Belgium). Third, the selected VOIs were automatically segmented using an adaptive threshold algorithm[22, 23] and imported as individual 3D standard tessellation language (stl) models, which were loaded in 3-Matic (Materialise, Leuven, Belgium) for longitudinal assessment of OTM and bone morphometry. The complete workflow is shown in Figure 1B.
To assess first and second molar displacement, six reference points on the cusps and five on the root apices were created. Their displacement over the different time points was defined as occlusal and apical movement, respectively. The occlusal plane was defined by the mesial cusp, distobuccal and distolingual points. The angle between the occlusal planes was defined as angular movement. The occlusal, apical and angular movements of the first and second molars at T0-T1, T0-T2, T0-T3, T0-T4, T1-T2, T2-T3, T3-T4 were measured on both OF and no OF side. The OTM rate was calculated by the occlusal movement divided by the time of OTM. The longitudinal measurements of the above-mentioned parameters and their dynamic changes are shown in Figure 2.
For assessment of mini-screw displacement, two reference points were created on the center of the screw head and the screw tip in the T0 model. The line connecting these two points was defined as the central axis. Then, the distances of the head and tip points from T0 to T1 were measured as the head and tip movement in the T0-T1 period, respectively. The angle between the central axes in T0 and T1 was measured as the angular movement. In this way, the displacement of mini-screw at T0-T2, T0-T3, T0-T4, T1-T2, T2-T3, T3-T4 was also measured.
Longitudinal assessment of bone morphometry
Alveolar bone morphometry was quantified using the method proposed by Chatterjee et al, as shown in Figure 3[24]. The registered micro-CT images were transferred to CTAnalyser and a volume of interest (VOI) containing only the molar region of each hemi maxilla was selected by defining a top and a bottom slice. The top slice was defined as the slice 2 mm mesial from that where the cusp of the first molar appeared. The bottom slice was defined as that where the mesial cusp of the third molar appeared. The alveolar bone within this region was semi-automatically selected and segmented using an automatic global threshold algorithm.
To evaluate the BMD, phantoms (standard hydroxyapatite blocks) of 0.25 g/cm³ and 0.75 g/cm³ were scanned to perform BMD calibration with respect to the attenuation values. The BMD at the different time points was calculated by linear extrapolation using: where x1 and x2 are the greyscale indices of standard hydroxyapatite 0.25 and 0.75 g/cm3, y1 and y2 are known as 0.25 and 0.75 g/cm3, and X and Y are the grey indices.
The dynamic changes of the following parameters were also evaluated in VOIs for the assessment of bone morphometry in CTAnalyser:
- Bone volume fraction (BV/TV, %): the proportion of the VOI occupied by binarised solid objects in 3D within the VOI, which directly reflects the bone volume.
- Bone surface density (BS/TV, mm-1): the ratio of surface area to total volume in 3D within the VOI.
- Trabecular number (Tb.N, mm-1): the number of traversals across a trabecular or solid structure made per unit length on a random linear path through the VOI
- Trabecular thickness (Tb.Th, mm): essentially the thickness of the solid voxels as defined by binarisation within the VOI.
- Trabecular separation (Tb.Sp, mm): essentially the thickness of the spaces among trabecular bone.
Statistical analysis
Two-way repeated measures MANOVA and post-hoc Tukey’s test were used to compare the OTM and bone morphometry between different time points (T0-T4) and between the OF and no OF side. Non-parametric analysis (Friedman test) was performed to compare the displacement of the mini-screw among different time points. The correlation of the dynamic changes between bone morphometric parameters and OTM rate was evaluated by Pearson correlation coefficients. Non-parametric statistical methods were used when normality was not confirmed using the Kolmogorov-Smirnov test. Statistical analysis was performed in GraphPad (version 8.4.3, San Diego, USA).