Mandibular cortical bone remodeling characteristics in patients with extraction: A cone-beam computed tomography study

Introduction

This study evaluated the labial and lingual cortical bone remodeling characteristics of mandibular central incisors after retraction, which remain controversial among orthodontists.

Methods

Cortical bone remodeling and central incisor movement of 33 patients (aged 23.64 ± 4.30 years) who underwent mandibular first premolar extraction and incisor retraction at the crestal (S1), midroot (S2), and apical (S3) levels were analyzed using superimposed cone-beam computed tomography images on the basis of voxel-based registration of the mandibular stable region. Multivariate linear regression was used to explore the relationships between labial bone remodeling/tooth movement (BT) ratios and factors such as the ANB angle, mandibular plane angle (Mp-SN), and incisor movement patterns. The patients were divided into 4 groups according to the lingual cortical bone remodeling condition and the relationship between posttreatment incisor roots and the original lingual cortical bone border. At the 3 levels (S1, S2, and S3), the classifications of cortical bone remodeling of the mandibular incisors were calculated; t tests were used to compare the amount of labial and lingual bone remodeling, BT ratios, and lingual bone remodeling/root over the original border (BRo) ratios.

Results

The mean labial BT ratios at all 3 levels were close to 1. Multivariate linear regression indicated that the tooth movement pattern negatively correlated with the BT ratio at the S2 and S3 levels ( P <0.05). Lingual bone apposition occurs when the root penetrates the original lingual cortical bone border in most patients. BRo ratios can more accurately reflect the inherent remodeling ability of the lingual cortical bone than BT ratios. The mean lingual BRo ratios were (1) S1 level: mandibular left central incisor (T31), 0.87 ± 0.25 and mandibular right incisor (T41), 0.86 ± 0.25; (2) S2 level: T31, 0.81 ± 0.12 and T41, 0.80 ± 0.22; and (3) S3 level: T31, 0.76 ± 0.20 and T41, 0.83 ± 0.26. There was no significant difference between labial BT ratios and lingual BRo ratios at the S2 and S3 levels.

Conclusions

The amount of labial cortical bone resorption caused by mandibular incisor retraction showed varied relationships with the amount of tooth movement. Bodily retraction may decrease the labial BT ratios at the S2 and S3 levels. Active lingual cortical bone apposition occurred when the roots penetrated the original lingual border and exhibited strong remodeling ability.

Highlights

•

A new 3D superimposition method was applied to the research of alveolar bone remodeling of retracted mandibular incisors.
•

Both labial and lingual cortical bone remodeled actively during mandibular incisor retraction.
•

A new index had been introduced to describe the inherent lingual bone remodeling ability.

In clinical practice, orthodontists frequently need to retract mandibular incisors over long distances, such as in camouflage orthodontic treatment for bimaxillary protrusion or patients with a Class III malocclusion. However, the mandibular anterior teeth area has the thinnest natural alveolar bone. Therefore, alveolar bone remodeling caused by mandibular incisor movement is a major concern for orthodontists and periodontists. There is a paucity of research on the remodeling of retracted mandibular incisors, with limited results focused on the changes in height or thickness of the labial and lingual alveolar bones, but they do not directly describe the remodeling pattern of the alveolar bone itself. Therefore, orthodontists are concerned about clinical decisions on incisor retraction, which lacks solid theoretical support.

The limited research on the mandible may be attributed to the constraints of the current research methods. Most previous studies estimated alveolar bone remodeling by measuring the morphology of the cortical bone in the longitudinal section view of the mandibular incisors before and after treatment. However, the labial and lingual alveolar bones of the mandibular incisors are extremely thin, making it challenging to eliminate the effect of measurement errors on the thickness of the alveolar bone in the longitudinal section of the tooth. In addition, much information is disregarded as all bone fenestration or dehiscent sites may be recorded as 0 without distinction.

Recent studies have used the 3-dimensional superimposition method of the stable region in cone-beam computed tomography (CBCT) on the basis of voxel-based registration to study alveolar bone remodeling in retracted maxillary incisors. This study applied this technique, in conjunction with cross-sectional measurements of cortical bone remodeling, to address the limitations of the single longitudinal section method and view the bone remodeling surrounding the roots.

In this study, CBCT scans of patients undergoing orthodontic mandibular incisor retraction were evaluated using this new method to investigate the labial and lingual cortical bone remodeling characteristics of mandibular central incisors after retraction and explore the overall remodeling patterns of the alveolar bone.

Material and methods

This retrospective study was approved by the Institutional Review Board of Peking University Hospital of Stomatology (approval no. PKUSSIRB-201631110). According to a 10-sample pretest and the 10 patients per variable rule, the minimum sample size of this study was 30. This study included 33 adult patients from the Orthodontic Department of Peking University Hospital of Stomatology (29 women and 4 men, aged 23.64 ± 4.30 years; treatment time: 31.46 ± 6.69 months). The inclusion criteria were (1) the removal of bilateral mandibular first premolars during orthodontic treatment with the edge of mandibular central incisors retracted >4 mm without obvious intrusion; (2) crowding in the mandibular arch <4 mm and crowding in the anterior teeth area <2 mm; (3) no evidence of periodontal or gingival problems at the beginning of orthodontic treatment; (4) no history of trauma to the mandibular anterior teeth; (5) treatment with 0.022-in MBT brackets, leveled and aligned with nickel-titanium, 0.019 × 0.025-in stainless steel archwires for space closing and en-masse retraction of all anterior teeth (canine-to-canine) using sliding mechanics; and (6) availability of pretreatment (T0) and posttreatment (T1) CBCT scans obtained using a NewTom VG scanner (Aperio Services, Verona, Italy) with the following scanning parameters: 15 × 15-cm field of view, 110 kVp, 1-3 mA, 10-second scan time, and 0.3 mm voxel size. The exclusion criteria were (1) root length (distance from the cementoenamel junction to apex) of mandibular central incisors <9 mm on T1 CBCT images and (2) low CBCT image quality, in which the cortical bone and root could not be defined accurately. The institution’s adoption of standard electronic medical records ensured that the recording format and methods for each patient were consistent.

In the Dolphin Imaging 3D program (version 11.9; Dolphin Imaging and Management Solutions, Chatsworth, Calif), the head position at the time of the T1 CBCT scan was adjusted as follows: the plane determined by sella, nasion, and basion was the central sagittal plane, and the plane determined by bilateral gonion and gnathion was the axial plane. The T0 and T1 CBCT scans were superimposed using voxel-based CBCT registration, consistent with the mandibular body’s basal bone extending from the external part of the symphysis to the first molar ^, ( Fig 1 ).

The image was rotated in the central sagittal plane, such that the long axis of the mandibular symphysis was perpendicular to the ground. The T1 and T0 axial slices and a scale ruler were exported individually from Digital Imaging and Communications. These slices were taken at intervals of 3, 6, and 9 mm below the cementoenamel junction of the mandibular right central incisor. These planes were categorized as the crestal (S1), midroot (S2), and apical (S3) levels of the mandibular central incisor.

In Procreate (version 4.3; Savage Interactive, Hobart, Tasmania, Australia), we used the slice (T0) to delineate the labial and lingual borders of the cortical bone and incisor edge. The resulting sketch and slice (T1) were combined into a single superimposed image and saved in the JPEG format. This process was repeated for slices at the S1, S2, and S3 levels, and each was performed independently ( Fig 2 ).

The measurements were performed using the ImageJ software (version 1.48; National Institutes of Health and the Laboratory for Optical and Computational Instrumentation, University of Wisconsin, Madison, Wis). The scale was set before measurement. All measured items were along the reference line that passed through the centers of the roots of the incisors on T0 and T1 scans. Measured parameters and their definitions are detailed in Table I and Figure 3 .

Table I

Definition of measurement items

Parameter	Definition	Measurement method
Labial side
BR_lab	Cortical bone remodeling on the labial side	The distance between the labial borders of the cortical bone on T0 and T1 scans
TM_lab	Tooth movement on the labial side	The distance between the labial borders of the incisor root on T0 and T1 scans
BT ratio_lab	Bone remodeling ratio on the labial side	BR_lab/TM_lab
IMP	Incisor movement pattern	The IMP represents the type of incisor retraction, defined as TM_lab at the S3 level/TM_lab at the cervical level. The higher the IMP value, the more likely the incisors were to exhibit bodily retraction. In contrast, lower IMP values indicated tipping
Lingual side
TM_ling	Tooth movement on the lingual side	The distance between the lingual borders of the incisor root on T0 and T1 scans
BR_ling	Cortical bone remodeling on the lingual side	The distance between the lingual borders of the cortical bone on T0 and T1 scans If there was a bone defect on the lingual side of the root, the lingual cortical bone height was defined as the average height of the mesial and distal bone around this root; otherwise, the bone height at the lingual cortical border was considered
BT ratio_ling	Bone remodeling ratio on the lingual side	BR_ling/TM_ling
ROB_ling	Distance of the root border from the original cortical bone border after root retraction over the original border on the lingual side	The distance between the lingual borders of the root on the T1 scan and cortical bone on the T0 scan
BRo ratio_ling	Bone remodeling ratio over the original border	BR_ling/ROB_ling
REP_ling	Root exposure perimeter	The circumference of the root without bone coverage on the T1 scan

Four different remodeling patterns were observed on the lingual side of the mandibular incisors. All patients were classified according to whether the root penetrated the original cortical bone border and whether cortical bone remodeling and bone defects were detected ( Fig 4 ).

1.

Type I: No lingual bone remodeling without root penetration of the original lingual cortical bone border.
2.

Type II: Lingual bone remodeling without bone defects when the root penetrates the original lingual cortical bone border.
3.

Type III: Lingual bone remodeling with bone defects when the root penetrates the original lingual cortical bone border.
4.

Type IV: No lingual bone remodeling when the root penetrates the original lingual cortical bone border.

Statistical analysis

One investigator (S.W.) performed repeated measurements for 25 randomly selected CBCT scans at an interval of 2 weeks. Another investigator (L.L.) performed interexaminer calibration and measured the 25 samples twice at an interval of 2 weeks.

Statistical analysis was performed using the SPSS (version 21.0; IBM, Armonk, NY) software package. Intraclass correlation coefficients (ICCs) were used to evaluate method reliability, computed using a 2-way random model and absolute agreement. The measurement error was evaluated using the method of moments estimator (MME) formula.

The relationships between labial cortical bone remodeling and the possible related factors were explored using multiple linear regression. Bone remodeling ratio on the labial side (BT ratio_lab) of mandibular left central incisor (T31) and mandibular right central incisor (T41) at all 3 levels were selected as the dependent variables. The ANB angle, mandibular plane angle (Mp-SN), and T31 or T41 incisor movement patterns (IMPs) were the independent variables. A histogram of residuals and a scatter diagram of dependent variables and standardized residuals were used to investigate the normality and homogeneity of variance in the 6 regression models.

The data were normally distributed and showed homogeneity of variance. Therefore, the paired t test was used to test the difference between the amount of labial and lingual bone remodeling, BT ratio_lab, and bone remodeling ratio over the original border (Bro ratio_ling) at the S1, S2, and S3 levels. For all tests, the significance level was set at P <0.05.

Results

The cephalometric measurements of the enrolled patients before and after orthodontic treatment are shown in Table II . Before orthodontic treatment, the labiolingual thicknesses of alveolar bone at the midpoint of T31 and T41 at the S1, S2, and S3 levels were 5.91 ± 1.01 mm, 5.97 ± 1.14 mm, and 6.68 ± 1.42 mm, respectively.

Table II

Cephalometric measurements before and after treatment

Variables	T0	T1	ΔT0 − T1	t	P values
SNA (°)	79.63 ± 3.49	80.07 ± 3.90	−0.44	−0.663	0.514
SNB (°)	75.79 ± 4.12	75.92 ± 4.36	−0.13	−0.199	0.844
ANB (°)	3.85 ± 3.41	4.14 ± 2.93	−0.30	−1.037	0.310
FH-NPo (°)	88.91 ± 4.33	88.38 ± 4.62	0.52	0.894	0.380
NA-APo (°)	7.62 ± 8.00	7.60 ± 7.22	0.02	0.041	0.967
U1-NA (mm)	7.11 ± 3.43	0.98 ± 3.21	6.13	11.202	<0.001 ^∗∗∗
U1-NA (°)	28.60 ± 8.87	14.86 ± 9.08	13.74	8.547	<0.001 ^∗∗∗
L1-NB (mm)	8.89 ± 2.96	3.87 ± 1.85	5.02	11.929	<0.001 ^∗∗∗
L1-NB (°)	32.03 ± 6.83	23.20 ± 6.07	8.83	7.733	<0.001 ^∗∗∗
U1-L1 (°)	115.50 ± 9.01	137.80 ± 6.59	−22.30	−14.932	<0.001 ^∗∗∗
SN-U1 (°)	108.25 ± 8.75	94.93 ± 9.30	13.32	7.395	<0.001 ^∗∗∗
Mp-SN (°)	40.54 ± 5.38	39.45 ± 5.52	1.09	1.865	0.074
Mp-FH (°)	27.62 ± 5.76	27.53 ± 5.93	0.09	0.123	0.903
L1-Mp (°)	95.70 ± 7.10	87.82 ± 6.49	7.88	6.319	<0.001 ^∗∗∗
SGn-FH (°)	62.50 ± 4.30	62.88 ± 4.65	−0.38	−0.608	0.549
Pog-NB (mm)	0.38 ± 1.65	1.08 ± 1.51	−0.70	−3.741	0.001 ^∗∗

∗∗ P <0.01.

∗∗∗ P <0.001.

For examiner 1 (S.W.), the intraexaminer ICCs ranged from 0.934-0.986, and the MME error ranged from 0.11-0.25 mm. For examiner 2 (L.L.), the intraexaminer ICCs ranged from 0.935-0.978, and the MME error ranged from 0.14-0.27 mm. Both the examiners demonstrated good self-stability.

The interexaminer ICCs ranged from 0.924-0.986, and the MME error ranged from 0.11-0.28 mm ( Table III ).

Table III

ICCs and MME error of repeated measurement

Variables	Lingual						Labial
	TM_ling		BR_ling		ROB_ling		REP_ling		TM_lab		BR_lab
	ICC	MME error	ICC	MME error	ICC	MME error	ICC	MME error	ICC	MME error	ICC	MME error
Examiner 1
S1	0.964	0.20	0.966	0.18	0.977	0.17	0.910	0.31	0.969	0.18	0.956	0.21
S2	0.974	0.19	0.958	0.20	0.972	0.21	0.966	0.32	0.950	0.24	0.958	0.21
S3	0.986	0.17	0.943	0.21	0.986	0.11	0.969	0.13	0.934	0.25	0.981	0.12
Examiner 2
S1	0.962	0.20	0.943	0.23	0.972	0.19	0.961	0.23	0.943	0.27	0.935	0.27
S2	0.964	0.24	0.959	0.23	0.973	0.20	0.968	0.20	0.958	0.23	0.955	0.19
S3	0.973	0.22	0.953	0.19	0.978	0.14	0.966	0.15	0.963	0.22	0.960	0.16
Examiner 1 − examiner 2
S1	0.952	0.28	0.929	0.27	0.977	0.17	0.910	0.31	0.932	0.21	0.932	0.25
S2	0.932	0.27	0.938	0.24	0.972	0.21	0.966	0.32	0.969	0.22	0.959	0.20
S3	0.960	0.21	0.943	0.20	0.986	0.11	0.969	0.13	0.972	0.19	0.924	0.11