Anda di halaman 1dari 13

ORIGINAL ARTICLE

Cone-beam computed tomography


evaluation of bone plate and root
length after maxillary expansion using
tooth-borne and tooth-tissue-borne
banded expanders
Mariana Roennau Lemos Rinaldi,a Fabiane Azeredo,a Eduardo Martinelli de Lima,a Susana Maria Deon Rizzatto,a
Glenn Sameshima,b and Luciane Macedo de Menezesa
Porto Alegre, Rio Grande do Sul, Brazil, and Los Angeles, Calif

Introduction: The objective of this research was to evaluate the buccal bone plate and root length of maxillary
permanent first molars using cone-beam computed tomography after maxillary expansion with different
activation protocols. Methods: Cone-beam computed tomography images of growing patients were obtained
from the orthodontic department of Pontifical Catholic University of Rio Grande do Sul in Brazil. The groups
were Haas-type 2/4 turns, Haas-type 4/4 turns, hyrax-type 2/4 turns, and hyrax-type with alternate rapid
maxillary expansions and constrictions (alt-RAMEC) 4/4 turns a day. Tooth length, periodontal insertion,
alveolar bone thickness, and intermolar distances were evaluated. The data at the start of treatment and
6 months later were compared using generalized linear models. The intergroup differences were determined
by univariate analysis of variance with the Bonferroni adjustment. Results: Tooth length was significantly short-
ened after expansion in all groups (0.28 to 0.51 mm), except for the alt-RAMEC group. Bone level variables
(bone level and bone level at the tooth tip) changed statistically in all groups, except for the Haas 4/4 turns group.
There was significant periodontal attachment loss after rapid maxillary expansion with the hyrax/alt-RAMEC
(5.09 mm). The hyrax/alt-RAMEC and hyrax groups had more dehiscences, fenestrations, and exposures of
the root. Conclusions: The consequence of rapid maxillary expansion using the hyrax was alveolar bone
resorption, especially in the hyrax/alt-RAMEC group, whereas the Haas expander caused mild root
resorption. (Am J Orthod Dentofacial Orthop 2018;154:504-16)

T
he maxillary expansion procedure is widely used to by the separation of the midpalatal, circumzygomatic,
correct posterior maxillary transverse discrepancies and circumaxillary sutures. The aims of the transverse
that are usually associated with different kinds of correction during the mixed dentition are to eliminate
malocclusion, such as a Class II or Class III molar relation- arch length discrepancies and basal bone deficiencies
ship, open bite, or crowding.1-7 Early treatment of this and to facilitate facemask protraction.8-10 However,
condition offers the possibility of orthopedic correction Liou and Tsai8 proposed an alternative method for disar-
ticulation of the circumaxillary sutures using alternate
a
expansion and constriction (alt-RAMEC) of the maxillary
Department of Orthodontics, School of Dentistry, Pontifical Catholic University
of Rio Grande do Sul, Porto Alegre, Brazil. arch. They reported that the maxillary sutures were
b
Advanced Dental Education Program in Orthodontics, Herman Ostrow School of less disarticulated using conventional rapid maxillary
Dentistry, University of Southern California, Los Angeles. expansion (RME) compared with the alt-RAMEC
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
tential Conflicts of Interest, and none were reported. method. In addition, the facemask therapy associated
Supported by the doctoral fellowships program of the Coordenaç~ao de Aperfei- to the alt-RAMEC procedure was 3 times more effective
çoamento de Pessoal de Nıvel Superior, Brazil. to displace A-point anteriorly than with RME in a
Address correspondence to: Luciane Macedo de Menezes, Av. Ipiranga, 6681,
Predio 6 (Faculdade de Odontologia), Porto Alegre, RS CEP 91530-000, Brazil; sample of growing Class III patients with cleft lip and
e-mail, luciane.menezes@pucrs.br. palate.8
Submitted, August 2017; revised and accepted, December 2017. The orthopedic expansion is obtained when a high-
0889-5406/$36.00
Ó 2018 by the American Association of Orthodontists. All rights reserved. force system is applied on the midsagittal maxillary
https://doi.org/10.1016/j.ajodo.2017.12.018 suture using tooth-tissue-borne (Haas type) or tooth-
504
Lemos Rinaldi et al 505

borne (hyrax type) appliances. Heavy forces such as 10 International, Hatfield, Pa) at 120 kV, 8 mA, scanning
kilogram-force generated during a turn of the expander time of 40 seconds, and 0.3-mm voxel dimension. The
screw11 are responsible for transverse skeletal changes data for each patient were saved in DICOM format,
with suture opening and minimum orthodontic move- and the images were stored in compact disks.
ment.1,12,13 The CBCT data of 77 patients of the original sample
The application of orthodontic forces induces a local were analyzed, and 16 examinations were excluded
process of inflammation, which is essential for tooth because the roots of the permanent first molars were
movement. This biomechanical reaction includes the 4 incomplete at the apical third when the initial CBCT
cardinal signs and symptoms of inflammation: redness, were taken. Therefore, the total sample size of this study
heat, swelling, and pain.14 Although orthodontically consisted of 61 children, distributed as described in
induced root resorption is an undesirable risk of treat- Table I.
ment, it is considered an unavoidable consequence of The patients were distributed in 4 groups, according
the forces applied for tooth movement.14,15 No to the daily screw activation protocol and expander ap-
regeneration is possible when the tooth root loses pliances used for RME: Haas-type 2/4 turns (n 5 11),
apical tissue beneath the cementum layers14; however, Haas-type 4/4 turns (n 5 16), hyrax-type 2/4 turns
root resorption severe enough to create a clinical prob- (n 5 18), and hyrax-type with alternate rapid maxillary
lem is unusual in orthodontics.16 expansions and constrictions (alt-RAMEC) activation
Previous studies have reported RME with root resorp- protocol with 4/4 turns a day (n 5 16) (Table I).
tion and evaluated this phenomenon with different im- A complete turn of the screw was done at the instal-
aging techniques such as radiographs,17 histologic lation of the appliance in all groups (0.8 mm). The pa-
analysis,18,19 scanning electron microscope,20 and tients of the alt-RAMEC group were instructed to
cone-beam computed tomography (CBCT).21 Conven- perform screw activation for a week and deactivation to-
tional computed tomography and CBCT have also wards closing at the same daily rate over the next week.
been used for skeletal, dentoalveolar, and periodontal Maxillary expansions and constrictions were repeated
change analyses resulting from maxillary expan- for 7 weeks in this hyrax-type group.
sion.22-27 Some studies have indicated that RME can Activations were performed until up to 8 mm of
change the buccal cortical bone level23,28,29 and cause expansion in the Haas-type 2/4, Haas-type 4/4, and
root resorption,21 but no studies have compared how hyrax-type 2/4 groups. The alt-RAMEC group had a total
different RME screw activation protocols can affect of 6.4 mm of screw opening. Overcorrection of the trans-
both bone level and root length. verse dentoskeletal discrepancy in all groups was
The aim of this study was to evaluate the buccal bone achieved. At the end of the RME active phase, the screws
plate and root length of maxillary permanent first molars were stabilized, and all patients used the same expanders
using CBCT in 4 groups of patients, divided according to for 6 months during the retention period.
their treatment protocols for RME. The CBCT analysis was performed using the InVivo5
software program (Anatomage Dental, San Jose, Calif) as
MATERIAL AND METHODS similarly proposed by Bernd.30 The region of the perma-
The ethical committee of the Pontifical Catholic Uni- nent first molar was evaluated. The long axis of the me-
versity of Rio Grande do Sul in Brazil approved this siobuccal root of the maxillary permanent first molar was
study. A sample of growing patients was selected from used as the reference for the standardization of the CBCT
the database of previous randomized clinical trials in images taken at T1 and T2 (Fig 1; Table II).
the Department of Orthodontics at the School of In the axial view, using the Section mode tool in the
Dentistry, with random allocation of each subject to InVivo5, the reference line (horizontal) available in the
the groups. program was positioned at the center of the mesiobuccal
The inclusion criteria for the study were transverse root of the maxillary first molar (Fig 1, A). In the sagittal
maxillary deficiency, mixed or early permanent denti- view, the reference line (vertical) was positioned on the
tion, and no surgical or other treatment that could inter- long axis of the mesiobuccal root of the tooth (Fig 1,
fere with the RME effects during the expansion period. B), resulting in a coronal image with adequate visualiza-
Patients with congenital malformations, periodontal tion of the alveolar buccal cortical bone and molar root
diseases, or metallic restorations in the permanent first axis to be measured (Fig 1, C and D). Then the reference
molars were excluded. line (vertical) was also placed on the long axis of the me-
CBCT images were taken before treatment (T1) and siobuccal root of the maxillary first molar (Fig 1, C). The
6 months after jackscrew stabilization (T2). The images reference protocol for the CBCT analysis was based on
were taken using an i-CAT scanner (Imaging Sciences the individual positioning of each tooth, and this step

American Journal of Orthodontics and Dentofacial Orthopedics October 2018  Vol 154  Issue 4
506 Lemos Rinaldi et al

Table I. Distribution of the 4 groups of patients


Group n Boys Girls Mean age (y 6 SD) RME activation protocol (mm/day) Total activation (mm)
Haas 2/4 11 2 9 11.6 6 1.5 0.4 8.0
Haas 4/4 16 6 10 11.3 6 1.7 0.8 8.0
Hyrax 18 7 11 11.1 6 1.25 0.4 8.0
Hyrax/alt-RAMEC 16 10 6 10.3 6 1.98 0.8 6.4
Total 61 25 36 11.0 6 1.6 - -

Fig 1. CBCT slices were oriented using the software program tools, and dentoalveolar variables were
measured: A, in the axial view, the reference line was positioned on the center of the mesiobuccal root
of the maxillry permanent first molar at the right side (horizontal line); B, in the sagittal view, the refer-
ence line (vertical line) was positioned on the long axis of the mesiobuccal root of the tooth; C, as the
result of axial and sagittal repositioning, a coronal image was obtained; D, dentoalveolar variables eval-
uated (TL, BL, BLC, B5, and B10).

was repeated according to the orientation of the right regions of the maxillary permanent first molars (Fig 2).
and left molars. In the axial view, the trifurcation region was used as
Transverse measurements were also made. CBCT im- an anatomic reference at the level where the mesial
ages were reoriented with the Frankfort horizontal plane and distal roots are separated. Thus, intermolar root
parallel to the floor in the sagittal view. In addition, the width (IRW) was the distance between the buccal sur-
infraorbital plane, formed by a line tangent to the lowest faces of the mesiobuccal root canals of the maxillary
point on the inferior margin of the bony orbits and par- right and left permanent first molars, measured in the
allel to the floor, was used as a reference plane in the cor- axial view at the trifurcation level (Fig 2, A). At the
onal aspect. Then the intermolar widths were evaluated same sagittal position, however, in the coronal view,
at 2 vertical levels: the trifurcation and the cuspal the intermolar cuspal width (ICW) was the distance

October 2018  Vol 154  Issue 4 American Journal of Orthodontics and Dentofacial Orthopedics
Lemos Rinaldi et al 507

Table II. Linear measurements at T1 and T2 used to analyze RME dentoalveolar effects
Measurement Definition
TL Tooth length Distance between the tip of mesiobuccal cusp to the apex of the
mesiobuccal root of the maxillary first molar
BL Buccal alveolar bone level Distance from the buccal cementoenamel junction and the most
cervical portion of the buccal alveolar bone crest
BLC Buccal bone level to the cusp tip Distance from the buccal cusp tip to the buccal alveolar crestal bone.
B5 Alveolar bone thickness 5 mm from cementoenamel junction Distance between the outer surface of the mesiobuccal root of the
maxillary first molar and the outer surface of the buccal cortical
bone, perpendicular to the tooth length line.
B10 Alveolar bone thickness 10 mm from cementoenamel junction Distance between the outer surface of the mesiobuccal root of the
maxillary first molar and the outer surface of the buccal cortical
bone, perpendicular to the tooth length line.
IRW Intermolar root width Transverse distance between the maxillary right and left first molars
at the furcation level, in the axial view.
ICW Intermolar cuspal width Transverse distance between the tips of the right and left mesiobuccal
cusps of the maxillary first molars, in the coronal view.

Fig 2. Three-dimensional CBCT slices used for intermolar measurements: A, intermolar root width
(IRW) measured at the trifurcation region (axial view); B, in the sagittal view, the reference line (vertical
line) was positioned on the long axis of the mesiobuccal root of the maxillry first molars; C, intermolar
cuspal width (ICW) measured at the mesiobuccal cusp tips of the maxillary permanent first molars (cor-
onal view).

between the mesiobuccal cusp tips of the reference teeth Intergroup statistical analyses of the mean differences
(Fig 2, C), similar to the method proposed by Rungchar- between T1 and T2 were determined by univariate anal-
assaeng et al.31 ysis of variance, controlling the initial variations among
the groups, with Bonferroni adjustments at a signifi-
Statistical analysis cance level of 5%.
Intraexaminer reliability of the measurements was
determined by intraclass correlation coefficients. CBCT
images of 10 patients were randomly selected, and dou- RESULTS
ble assessments of each parameter at T1 and T2 were The mean ages of the patient groups are described in
performed by the same operator (M.R.L.R.) at a 10-day Table I, and no statistically significant difference was
interval. found among them.
Means and standard errors for all parameters and the The measurements of all variables were considered
ICW/IRW ratio were calculated, and data at T1 and T2 reliable, and the intraclass correlation coefficient values
were compared using generalized linear models. ranged from 0.87 to 0.99.

American Journal of Orthodontics and Dentofacial Orthopedics October 2018  Vol 154  Issue 4
508 Lemos Rinaldi et al

Table III. Tooth mean length, and minimum, maximum, and mean differences among groups between T1 and T2
Mean difference T2-T1 (95% CI) Mean difference
T1 mean (SD) T2 mean (SD) P* (minimum; maximum) Group (groups) Py
TL
Haas 2/4 19.20 (1.0) 18.80 (0.93) 0.000 0.40 (0.60; 0.18) Haas 4/4 0.11 1.00
Hyrax 0.12 1.00
Hyrax/alt–RAMEC 0.22 1.00
Haas 4/4 20.17 (1.61) 19.66 (1.66) 0.000 0.51 (0.69; 0.33) Haas 2/4 0.11 1.00
Hyrax 0.23 0.55
Hyrax/alt–RAMEC 0.33 0.13
Hyrax 20.08 (1.41) 19.80 (1.48) 0.000 0.28 (0.42; 0.12) Haas 2/4 0.12 1.00
Haas 4/4 0.23 0.55
Hyrax/alt–RAMEC 0.1 1.00
Hyrax/alt-RAMEC 19.51 (1.63) 19.33 (1.38) 0.159 0.18 (0.43; 0.07) Haas 2/4 0.21 1.00
Haas 4/4 0.33 0.13
Hyrax 0.1 1.00

Statistically significant (P #0.05).


*Generalized estimating equations. Bonferroni adjustment for multiple comparisons; yAnalysis of covariance adjusted for baseline values. Each
Wald chi-square test evaluates the simple effects of time on each level of the combination of other factors shown. These tests are based on the
linearly independent pairwise method comparisons between estimated marginal means.

Tooth length was significantly shortened after treat- In addition, anatomic defects related to the maxillary
ment in all groups, except in the hyrax/alt-RAMEC permanent first molars and their adjacent buccal cortical
group. Nonetheless, the mean difference among the 4 plates were observed at T2 in the CBCT scans, probably
groups showed no significant difference (Table III; because of the RME procedure (Fig 5). These undesired
Fig 3). Buccal alveolar bone level (BL) increased signif- findings were classified and quantified with their per-
icantly in the 4 groups, especially in the hyrax/alt- centages in Table VII.
RAMEC group (5.09 mm) (Fig 3). Buccal cortical bone
level at the cusp tip (Fig 3) also increased significantly, DISCUSSION
except in the Haas 4/4 group, and alveolar bone thick- Apical root resorption is frequently associated with
ness (B5 and B10) decreased significantly in the 4 orthodontic movement. The literature supports the vari-
groups (Table IV; Fig 3). ability of its incidence due to individual predispositions
Both the IRW and ICW variables showed statistically (genetic factors), deleterious habits (eg, tongue thrust),
significant differences between T1 and T2 in all 4 increased overjet, and susceptibility to traumatic dental
groups. The Haas 2/4 group had the greatest mean dif- injuries especially to the maxillary incisors, as well as
ference between T1 and T2 for IRW and ICW (7.12 and forces applied during orthodontic procedures.32-36
9.16 mm, respectively; Fig 3). The IRW mean for the Root resorption can be influenced by root shape when
Haas 2/4 group was statistically greater than the means the apical thirds are triangular, pipette shaped,
obtained for the Haas 4/4 and hyrax/alt-RAMEC groups dilacerated, or short.14 External root resorption is
(5.46 and 5.13 mm, respectively; Table V). considered severe when it exceeds 4 mm or one third
The lowest mean difference between pretreatment of the original root length,33,37,38 or up to 30% of root
and posttreatment ICW was found in the Haas 4/4 group loss. Another important factor is the number of teeth
(6.46 mm), and it was significantly different from the involved; short roots of all 4 anterior teeth have higher
Haas 2/4 and hyrax groups (Table V; Fig 4). risks for future problems with additional insertion
The ICW/IRW ratios were significantly different for all loss.31 The major concern related to root resorption is
4 groups when the difference between the treatment pe- the dental insertion loss, compromising the maintenance
riods was compared (T2-T1). Buccal inclination of the of the tooth.33,39-41
molars at the cuspal level was higher than at the root In this study, the mesiobuccal root was analyzed
level. No statistical difference was observed among the because this root is directly related to the buccal bone
groups (Table V; Fig 3). Significant differences were plate. This root is bulkier than the distobuccal root and
observed after the RME procedure with the Haas and more susceptible to dentoskeletal changes related to
hyrax appliances. However, the buccal cortical bone RME.
levels were significantly increased in the hyrax groups Root resorption associated with RME was previously
(Table VI). studied by invasive methods, extracting premolars after

October 2018  Vol 154  Issue 4 American Journal of Orthodontics and Dentofacial Orthopedics
Lemos Rinaldi et al 509

Fig 3. Comparison between groups after RME. Generalized estimating equations with the Bonferroni
adjustment for multiple comparisons were used, and analysis of covariance was adjusted for baseline
values. Each Wald chi-square test evaluates the simple effects of time on each level of the combination
of other factors shown. These tests were based on the linearly independent pairwise method compar-
isons between estimated marginal means. Groups not connected by the same letter are significantly
different (P #0.05). *Statistically significant between T1 and T2.

expansion and observing the roots by optical microscopy used as anchorage after RME. Other authors reported
or scanning electron microscopy and, more recently, by that root damages were more associated with teeth
CBCT images.18,19,21,42 Barber and Sims42 reported with longer periods of fixed retention; in the repaired
resorption of the buccal surface of the roots of premolars resorptive defects, cellular cementum with Sharpey's

American Journal of Orthodontics and Dentofacial Orthopedics October 2018  Vol 154  Issue 4
510 Lemos Rinaldi et al

Table IV. Means and differences among groups between T1 and T2 for BL, BLC, B5, and B10
Mean difference T2-T1 (95% CI) Mean difference
T1 mean (SD) T2 mean (SD) P* (minimum; maximum) Group (groups) Py
BL
Haas 2/4 1.25 (0.40) 2.53 (2.26) 0.003 1.28 (0.44; 2.12) Haas 4/4 1.05 1.00
Hyrax 0.52 1.00
Hyrax/alt-RAMEC 3.80* 0.00
Haas 4/4 1.24 (0.31) 1.47 (0.61) 0.020 0.23 (0.03; 0.41) Haas 2/4 1.05 1.00
Hyrax 1.57 1.00
Hyrax/alt-RAMEC 4.86* 0.00
Hyrax 1.08 (0.32) 2.88 (3.46) 0.002 1.80 (0.68; 2.89) Haas 2/4 0.52 1.00
Haas 4/4 1.57 1.00
Hyrax/alt-RAMEC 3.29* 0.00
Hyrax/alt-RAMEC 1.31 (0.29) 6.40 (5.33) 0.000 5.09 (3.31; 6.87) Haas 2/4 3.90* 0.00
Haas 4/4 4.86* 0.00
Hyrax 3.29* 0.00
BLC
Haas 2/4 8.26 (0.78) 9.51 (2.29) 0.003 1.24 (0.43; 2.06) Haas 4/4 1.14 1.00
Hyrax 0.74 1.00
Hyrax/alt-RAMEC 4.10* 0.00
Haas 4/4 8.06 (0.71) 8.16 (0.90) 0.377 0.10 (0.12; 0.33) Haas 2/4 1.14 1.00
Hyrax 1.88 0.14
Hyrax/alt-RAMEC 5.24* 0.00
Hyrax 8.14 (0.69) 10.12 (3.79) 0.001 1.98 (0.86; 3.09) Haas 2/4 0.74 1.00
Haas 4/4 1.88 0.14
Hyrax/alt-RAMEC 3.36* 0.00
Hyrax/alt-RAMEC 7.75 (0.81) 13.09 (5.32) 0.000 5.34 (3.59; 7.09) Haas 2/4 4.10* 0.00
Haas 4/4 5.24* 0.00
Hyrax 3.36* 0.00
B5
Haas 2/4 1.95 (0.55) 0.60 (0.46) 0.000 1.35 (1.54; 1.15) Haas 4/4 0.57* 0.00
Hyrax 0.45* 0.03
Hyrax/alt-RAMEC 0.27 0.56
Haas 4/4 1.71 (0.54) 0.93 (0.42) 0.000 0.78 (0.97; 0.59) Haas 2/4 0.57* 0.00
Hyrax 0.12 1.00
Hyrax/alt-RAMEC 0.30 0.25
Hyrax 1.80 (0.98) 0.90 (0.82) 0.000 0.90 (1.12; 0.69) Haas 2/4 0.45* 0.03
Haas 4/4 0.12 1.00
Hyrax/alt-RAMEC 0.18 1.00
Hyrax/alt-RAMEC 1.42 (0.68) 0.34 (0.50) 0.000 1.08 (1.26; 0.89) Haas 2/4 0.27 0.56
Haas 4/4 0.30 0.25
Hyrax 0.18 1.00
B10
Haas 2/4 3.16 (1.59) 1.88 (1.48) 0.000 1.28 (1.92; 0.64) Haas 4/4 0.17 1.00
Hyrax 0.30 1.00
Hyrax/alt-RAMEC 0.37 1.00
Haas 4/4 3.03 (1.11) 1.92 (1.25) 0.000 1.11 (1.42; 0.78) Haas 2/4 0.17 1.00
Hyrax 0.47 1.00
Hyrax/alt-RAMEC 0.54 0.98
Hyrax 3.37 (2.15) 1.79 (1.38) 0.000 1.58 (2.11; 1.03) Haas 2/4 0.30 1.00
Haas 4/4 0.07 1.00
Hyrax/alt-RAMEC 0.17 1.00
Hyrax/alt-RAMEC 2.52 (2.14) 0.87 (1.23) 0.000 1.65 (2.25; 1.05) Haas 2/4 0.37 1.00
Haas 4/4 0.54 0.98
Hyrax 0.07 1.00
Statistically significant (P #0.05).
*Generalized estimating equations. Bonferroni adjustment for multiple comparisons; yAnalysis of covariance adjusted for baseline values. Each
Wald chi-square test evaluates the simple effects of time on each level of the combination of other factors shown. These tests are based on the
linearly independent pairwise method comparisons between estimated marginal means.

October 2018  Vol 154  Issue 4 American Journal of Orthodontics and Dentofacial Orthopedics
Lemos Rinaldi et al 511

Table V. Means, and minimum, maximum, and mean differences among groups between T1 and T2 for IRW, ICW
and ICW/IRW ratio
Mean difference T2-T1 (95% CI) Mean difference
T1 mean (SD) T2 mean (SD) P* (minimum; maximum) Groups (groups) Py
IRW
Haas 2/4 44.98 (2.04) 52.10 (2.55) 0.000 7.12 (6.44; 7.79) Haas 4/4 1.66* 0.01
Hyrax 0.88 0.47
Hyrax/alt-RAMEC 1.99* 0.00
Haas 4/4 47.62 (2.59) 53.08 (2.42) 0.000 5.46 (4.8; 6.11) Haas 2/4 1.66* 0.01
Hyrax 0.78 0.54
Hyrax/alt-RAMEC 0.33 1.00
Hyrax 46.22 (2.61) 52.46 (3.08) 0.000 6.24 (5.62; 6.83) Haas 2/4 0.89 0.47
Haas 4/4 0.78 0.54
Hyrax/alt-RAMEC 1.11 0.10
Hyrax/alt-RAMEC 47.01 (4.20) 52.14 (4.41) 0.000 5.13 (4.54; 5.71) Haas 2/4 1.99* 0.00
Haas 4/4 0.33 1.00
Hyrax 1.11 0.10
ICW
Haas 2/4 46.24 (2.28) 55.40 (2.84) 0.000 9.16 (8.14; 10.18) Haas 4/4 2.70* 0.00
Hyrax 0.29 1.00
Hyrax/alt-RAMEC 1.77 0.10
Haas 4/4 48.99 (2.93) 55.45 (3.08) 0.000 6.46 (5.94; 6.97) Haas 2/4 2.70* 0.00
Hyrax 2.41* 0.00
Hyrax/alt-RAMEC 0.93 0.99
Hyrax 48.99 (2.34) 57.86 (3.12) 0.000 8.87 (8.08; 9.65) Haas 2/4 0.29 1.00
Haas 4/4 2.41* 0.00
Hyrax/alt-RAMEC 1.48 0.14
Hyrax/alt-RAMEC 48.92 (3.93) 56.31 (4.23) 0.000 7.39 (6.18; 8.58) Haas 2/4 1.77 0.10
Haas 4/4 0.93 0.99
Hyrax 1.48 0.14
ICW/IRW ratio
Haas 2/4 1.028 (0.04) 1.064 (0.05) 0.006 0.036 (0.01; 0.10) Haas 4/4 0.020 1.00
Hyrax 0.007 1.00
Hyrax/alt-RAMEC 0.004 1.00
Haas 4/4 1.029 (0.04) 1.044 (0.04) 0.024 0.015 (0.00; 0.29) Haas 2/4 0.020 1.00
Hyrax 0.028 0.31
Hyrax/alt-RAMEC 0.024 0.58
Hyrax 1.061 (0.05) 1.104 (0.05) 0.000 0.043 (0.03; 0.06) Haas 2/4 0.007 1.00
Haas 4/4 0.028 0.31
Hyrax/alt-RAMEC 0.003 1.00
Hyrax/alt-RAMEC 1.042 (0.04) 1.082 (0.06) 0.002 0.040 (0.01; 0.06) Haas 2/4 0.004 1.00
Haas 4/4 0.024 0.58
Hyrax 0.003 1.00

Statistically significant (P #0.05).


*Generalized estimating equations. Bonferroni adjustment for multiple comparisons; yAnalysis of Covariance adjusted for baseline values. Each
Wald chi-square test evaluates the simple effects of time on each level of the combination of other factors shown. These tests are based on the
linearly independent pairwise method comparisons between estimated marginal means.

fibers was observed.18,19 Studies have assessed maxillary first molars (18.60 mm3). The less affected
volumetric dimensions of dental roots after RME with root was the distobuccal root of maxillary first molars
CBCT images.21,43 Baysal et al21 evaluated root resorp- (9.47 mm3). In another study, the teeth were segmented
tion in 25 patients, and the 3-dimensional images of to measure their surface areas.43 The authors observed
the permanent first molars and first and second premo- loss of structure in the roots of the maxillary permanent
lars were cautiously segmented. The difference between first molars and premolars after RME. The analyses
the before and after RME root volumes was statistically showed that the main locations of the changes were at
significant for all roots investigated. The maximum vol- the apex of the maxillary teeth and the buccal aspect
ume loss was observed for the mesiobuccal roots of of the roots.

American Journal of Orthodontics and Dentofacial Orthopedics October 2018  Vol 154  Issue 4
512 Lemos Rinaldi et al

relation to the 8 mm activated at the expander screw.


The same comparison showed that buccal inclination
was 15.4% (0.99 mm) greater in relation to the
6.4 mm of total screw activation using the hyrax/alt-
RAMEC protocol. Maybe the clinical response using a
short-term activation protocol, as used in the Haas 4/4
group (with approximately 10 days of expansion), could
contribute to preventing excessive buccal tipping.
Buccal or palatal bone plate measurements were
inaccessible before the advent of CBCT. The accuracy
and reliability of this method to detect bone alterations
have been tested.45-47 Most studies agree that bone
measurements using CBCT have appropriate accuracy
Fig 4. After RME, the ICW/IRW ratio showed that the and reliability for clinical studies. Nevertheless, the
crowns were more buccally inclined than the furcation
accuracy of the measurements of alveolar bone height
level of the roots of the maxillary permanent first molars.
and bone thickness are related to the voxel size.
Alveolar bone height loss might be overestimated with
In our study, the Haas 4/4 group showed greater a 0.4-mm voxel size of the CBCT. This situation can be
amounts of resorption (0.51 6 0.09 mm), with no signif- observed if the bone thickness is smaller than the
icant difference compared with the other groups CBCT voxel size.48,49 By using a voxel size of 0.38 mm,
(Table III). The range of shortening observed was compa- the alveolar bone height can be measured to an
rable with the values found by Akyalcin et al43 in a sam- accuracy of about 0.6 mm, and root fenestrations can
ple of patients with tooth-borne RME appliances be identified with greater accuracy than dehiscences.
(0.36-0.52 mm). Sameshima and Sinclair44 also found Sun et al49 found that dehiscences were systematically
resorption of the mesial buccal roots of first molars overestimated on the CBCT measurements when the ver-
after treatment with full fixed orthodontic appliances tical diameter of the dehiscence was over 3 mm. When a
(0.12 6 1.34 mm). The smallest root shortening fenestration was found on CBCT, they observed that the
observed in our study was 0.18 6 0.12 mm for the alveolar defect was confirmed in approximately 20% of
hyrax/alt-RAMEC group, with no statistical significance the cases. Decreasing CBCT voxel size from 0.4 to
(Table III). 0.25 mm could improve the image accuracy for evalua-
The IRW and ICW measurements were related to the tions of the alveolar bone and its adjacent structures.50
amount of expansion, whereas the ICW/IRW ratio However, high-resolution CBCT images increase the pa-
referred to the tooth's buccal tipping. The maxillary tient's radiation dose, and it is important to consider
expansion had a triangular pattern with a wider base whether this is necessary for the diagnosis, treatment
at the cuspal level of the maxillary first molars. Signifi- plan, and clinical outcome.
cant buccal inclination of the crowns was observed in Garib et al23 reported significant decreases in the
all groups, but it was greater in the hyrax and hyrax/ buccal bone plate thickness of the banded supporting
alt-RAMEC groups. The mean ICW values in both groups teeth (first premolars and permanent first molars) for
were 44% (2.26 mm) and 42.1% (2.63 mm) higher than both tooth-tissue-borne and tooth-borne expanders,
IRW, respectively. The cuspal tipping amounts were ranging from 0.6 to 0.9 mm, after 3 months of expan-
28.6% (2.04 mm) and 18.3% (1 mm) higher than root sion. Brunetto et al28 compared RME with slow maxillary
buccal tipping for the Haas 2/4 and Haas 4/4 groups, expansion and reported loss and reduction of height and
respectively (Fig 4). However, no significant difference thickness of bone in both groups.
was found in the ICW/IRW ratio among the groups In this study, bone thickness measurements (B5 and
(Table V; Fig 4). B10) decreased significantly in all groups between T1
The total amount of screw activation was standard- and T2, varying from 1.35 mm in B5 and 1.65 mm in
ized at 8 mm for the Haas 2/4, Haas 4/4, and hyrax B10. The Haas 4/4 group had the lowest variation in
groups (Table I). The results showed that buccal cusp the measurements of bone plate height (BL and BLC)
tipping in the Haas 2/4 group was 14.5% (1.16 mm) and thickness (Table IV). The distance from the buccal
greater than the expansion at the screw. Similar behavior cementoenamel junction of the maxillary first molars
was observed in the hyrax group, where the ICW was and the most cervical portion of the buccal alveolar
10.8% (0.87 mm) greater than the screw opening. How- bone crest (BL) increased by a mean of 18.5%
ever, the Haas 4/4 group had a smaller ICW (6.46 mm) in (0.23 mm; P 5 0.020) in this group. Therefore, the

October 2018  Vol 154  Issue 4 American Journal of Orthodontics and Dentofacial Orthopedics
Lemos Rinaldi et al 513

Table VI. Means and mean differences among groups between T1 and T2
Mean difference T2-T1 (95% CI) Mean difference
T1 mean (SD) T2 mean (SD) P* (minimum; maximum) Expander types (expanders) Py
TL
Haas 17.33 (1.31) 17.86 (1.06) 0.000 0.53 (0.27; 0.77) Hyrax 0.07 1.00
Hyrax 17.69 (1.94) 18.15 (1.61) 0.005 0.46 (0.14; 0.78) Haas 0.07 1.00
BL
Haas 1.00 (0.47) 1.60 (1.43) 0.069 0.60 (0.04; 1.23) Hyrax 1.21 0.80
Hyrax 0.11 (0.32) 2.92 (2.93) 0.039 1.81 (0.08; 3.52) Haas 1.21 0.80
BLC
Haas 7.70 (0.16) 8.46 (1.67) 0.000 0.76 (0.07; 1.44) Hyrax 1.68 0.29
Hyrax 7.97 (0.23) 10.41 (2.99) 0.000 2.44 (0.68; 4.19) Haas 1.68 0.29
B5
Haas 2.15 (0.71) 1.07 (0.68) 0.000 1.08 (1.27; 0.87) Hyrax 0.29 1.00
Hyrax 1.84 (0.91) 0.54 (0.36) 0.000 1.30 (1.70; 0.92) Haas 0.29 1.00
B10
Haas 4.70 (0.49) 2.68 (1.30) 0.000 2.02 (2.78; 1.28) Hyrax 0.69 0.98
Hyrax 2.96 (0.67) 1.62 (1.02) 0.019 1.34 (2.45; 0.22) Haas 0.69 0.98
IRW
Haas 46.86 (2.18) 52.56 (1.98) 0.000 5.70 (4.97; 6.43) Hyrax 0.90 1.00
Hyrax 46.45 (2.19) 53.05 (1.86) 0.000 6.60 (5.97; 7.24) Haas 0.90 1.00
ICW
Haas 48.21 (1.91) 57.63 (2.99) 0.000 9.42 (8.26; 10.58) Hyrax 1.31 1.00
Hyrax 49.20 (1.28) 59.93 (0.45) 0.000 10.73 (9.48; 11.98) Haas 1.31 1.00

Statistically significant (P #0.05).


*Generalized estimating equations. Bonferroni's corrected multiple comparisons; yAnalysis of covariance adjusted for baseline values. Each Wald
chi-square test evaluates the simple effects of time on each level of the combination of other factors shown. These tests are based on the linearly
independent pairwise method comparisons between estimated marginal means.

preservation of crestal bone seems to be related to the RAMEC protocol for RME was more prone to cause de-
lower buccal tipping. The BL increased by 388.5% fects at the buccal bone plate. Despite that, the hyrax/
(5.09 mm; P 5 0.000) in the hyrax/alt-RAMEC group, alt-RAMEC group was the exception among the groups
and it was significant compared with the other groups evaluated. With no significant difference in tooth length
(Table IV). A deleterious alveolar effect can be due to measurements between the treatment periods, it is plau-
the highest amount of buccal tipping of the crowns, re- sible that the expansion and contraction protocol might
sulting in crestal bone resorption and loss of periodontal have caused adverse effects to the bone plate in this
attachment of anchored teeth in patients who used the sample, but not to the roots, maybe because the amount
repetitive weekly protocol of alt-RAMEC and a rigid of expansion was smaller compared with the other
expander appliance such as the hyrax. groups.
In addition, occasional findings were observed in the The Haas type of expander differs from the hyrax type
CBCT evaluations, and the hyrax/alt-RAMEC group by the presence of an acrylic pad close to the palate to
showed more dehiscences, fenestrations, and exposures reinforce the anchorage for a greater orthopedic
of the buccal aspect of the roots. The hyrax group also response during RME.12 In this study, both groups had
had many subjects with these kinds of defects. It seems no statistically significant differences for any measure-
to be related to the increased trend of buccal inclination ment, except for B5, which was greater in the Haas-
of the crowns after RME and the structural rigidity of the type groups. This could be explained by the acrylic pad
appliance. Maybe the maxillary permanent first molars against the palate, related to a more bodily movement
and their adjacent alveolar bone could be preserved if of the teeth, instead of inclination.23,28
the deciduous first molars were banded and used as Dental, skeletal, and periodontal stability in patients
anchorage instead of the permanent teeth. after RME is a major concern for clinicians. According
The voxel size of the CBCT images used in this study to Phillips,39 2 mm of resorption at the apical root
was 0.3 mm, and the measurements were taken on the length reduces the total insertion area of the root by
long axis of the coronal images, with clear visualization only 6% to 9%, and the apical loss of more than
of the alveolar buccal cortical bone and maxillary molar 2 mm can be equivalent to 1 mm of crestal bone
roots, in both axial and coronal views. The hyrax/alt- loss. Akyalcin et al26 found that the changes in the

American Journal of Orthodontics and Dentofacial Orthopedics October 2018  Vol 154  Issue 4
514 Lemos Rinaldi et al

Fig 5. Examples of anatomic defects related to the maxillary permanent first molars and adjacent alve-
olar bone tissue observed at T2, but not at T1: A, fenestration; B, dehiscence; C and D, complete
disruption of the buccal bone plate.

Table VII. Anatomic defects related to the maxillary permanent first molars and their adjacent cortical buccal plates
after RME
Anatomic defects

Number of teeth evaluated Dehiscence Fenestration Complete disruption Total % of teeth


Haas 2/4 22 2 1 1 4 18
Haas 4/4 32 0 0 0 0 0
Hyrax 36 5 6 1 12 33
Hyrax/alt-RAMEC 32 6 4 9 19 59
Total teeth 122 13 11 11 35 28

alveolar bone after RME were reversible in the long RME in patients with incomplete root development
term, with no evident deleterious effects to the alveolar of the maxillary permanent first molars and its relation-
buccal cortical bone. ship to the buccal bone plate shall be further

October 2018  Vol 154  Issue 4 American Journal of Orthodontics and Dentofacial Orthopedics
Lemos Rinaldi et al 515

investigated; since the number of patients with incom- 10. Proffit WR, Fields HW, Sarver DM. Contemporary orthodontics. St
plete roots was small, those CBCT images were excluded Louis: Elsevier Health Sciences; 2014.
11. Isaacson RJ, Ingram AH. Forces produced by rapid maxillary
from this study.
expansion. Angle Orthod 1964;34:261-70.
An individual risk assessment for each patient is rec- 12. Haas AJ. Palatal expansion: just the beginning of dentofacial or-
ommended before RME, and clinical examinations of the thopedics. Am J Orthod 1970;57:219-55.
periodontal conditions after RME should be considered 13. Haas AJ. Long-term posttreatment evaluation of rapid palatal
routine procedures. expansion. Angle Orthod 1980;50:189-217.
14. Brezniak N, Wasserstein A. Orthodontically induced inflammatory
root resorption. Part I: the basic science aspects. Angle Orthod
CONCLUSIONS 2002;72:175-9.
15. Roscoe MG, Meira JB, Cattaneo PM. Association of orthodontic
1. Although the differences were not significant, force system and root resorption: a systematic review. Am J Orthod
greater amounts of root resorption were found in Dentofacial Orthop 2015;147:610-26.
the Haas 4/4 group. 16. Killiany DM. Root resorption caused by orthodontic treatment: an
evidence-based review of literature. Semin Orthod 1999;5:128-33.
2. There were no statistically significant differences
17. Handelman CS. Nonsurgical rapid maxillary alveolar expansion in
among the Haas 2/4 and hyrax groups. adults: a clinical evaluation. Angle Orthod 1997;67:291-308.
3. There was significant alveolar bone resorption and 18. Langford SR, Sims MR. Root surface resorption, repair, and peri-
consequent loss of attachment after RME with the odontal attachment following rapid maxillary expansion in man.
hyrax/alt-RAMEC compared with the other groups, Am J Orthod 1982;81:108-15.
19. Erverdi N, Okar I, K€ uc€
ukkeles N, Arbak S. A comparison of two
although no significant root resorption was
different rapid palatal expansion techniques from the point of
observed. root resorption. Am J Orthod Dentofacial Orthop 1994;106:47-51.
4. In general, the consequence of RME using the hyrax 20. Langford SR. Root resorption extremes resulting from clinical
appliance was alveolar bone resorption, whereas the RME. Am J Orthod 1982;81:371-7.
Haas appliance caused root resorption. 21. Baysal A, Karadede I, Hekimoglu S, Ucar F, Ozer T, Veli I, et al. Eval-
uation of root resorption following rapid maxillary expansion using
5. The hyrax/alt-RAMEC and hyrax groups had dehis-
cone-beam computed tomography. Angle Orthod 2012;82:488-94.
cences, fenestrations, and exposure of the root at 22. Garib DG, Henriques JF, Janson G, Freitas MR, Coelho RA. Rapid
the buccal aspect. maxillary expansion—tooth tissue-borne versus tooth-borne ex-
panders: a computed tomography evaluation of dentoskeletal ef-
fects. Angle Orthod 2005;75:548-57.
REFERENCES 23. Garib DG, Henriques JF, Janson G, de Freitas MR, Fernandes AY.
1. Haas AJ. The treatment of maxillary deficiency by opening the Periodontal effects of rapid maxillary expansion with tooth-
midpalatal suture. Angle Orthod 1965;35:200-17. tissue-borne and tooth-borne expanders: a computed tomography
2. Haas AJ. Rapid palatal expansion: a recommmended prerequisite evaluation. Am J Orthod Dentofacial Orthop 2006;129:749-58.
to Class III treatment. Trans Eur Orthod Soc 1973;311-8. 24. Garib DG, Navarro R, Francischone CE, Oltramari PV. Rapid maxil-
3. Tollaro I, Baccetti T, Franchi L, Tanasescu CD. Role of posterior lary expansion using palatal implants. J Clin Orthod 2008;42:
transverse interarch discrepancy in Class II, Division 1 malocclusion 665-71.
during the mixed dentition phase. Am J Orthod Dentofacial Orthop 25. Weissheimer A, de Menezes LM, Mezomo M, Dias DM, de Lima EM,
1996;110:417-22. Rizzatto SM. Immediate effects of rapid maxillary expansion with
4. Baccetti T, Franchi L, McNamara JA, Tollaro I. Early dentofacial Haas-type and hyrax-type expanders: a randomized clinical trial.
features of Class II malocclusion: a longitudinal study from the de- Am J Orthod Dentofacial Orthop 2011;140:366-76.
ciduous through the mixed dentition. Am J Orthod Dentofacial Or- 26. Akyalcin S, Schaefer JS, English JD, Stephens CR, Winkelmann S. A
thop 1997;111:502-9. cone-beam computed tomography evaluation of buccal bone
5. Lineberger MW, McNamara JA, Baccetti T, Herberger T, Franchi L. thickness following maxillary expansion. Imaging Sci Dent 2013;
Effects of rapid maxillary expansion in hyperdivergent patients. 43:85-90.
Am J Orthod Dentofacial Orthop 2012;142:60-9. 27. Garib DG, Menezes MH, Silva OG, Santos PB. Immediate peri-
6. Kanomi R, Deguchi T, Kakuno E, Takano-Yamamoto T, odontal bone plate changes induced by rapid maxillary expansion
Roberts WE. CBCT of skeletal changes following rapid maxillary in the early mixed dentition: CT findings. Dental Press J Orthod
expansion to increase arch-length with a development- 2014;19:36-43.
dependent bonded or banded appliance. Angle Orthod 2013;83: 28. Brunetto M, Andriani JS, Ribeiro GL, Locks A, Correa M, Correa LR.
851-7. Three-dimensional assessment of buccal alveolar bone after rapid
7. Diouf JS, Ngom PI, Sonko O, Diop-B^a K, Badiane A, Diagne F. In- and slow maxillary expansion: a clinical trial study. Am J Orthod
fluence of tonsillar grade on the dental arch measurements. Am J Dentofacial Orthop 2013;143:633-44.
Orthod Dentofacial Orthop 2015;147:214-20. 29. Pangrazio-Kulbersh V, Jezdimir B, de Deus Haughey M,
8. Liou EJ, Tsai WC. A new protocol for maxillary protraction in cleft Kulbersh R, Wine P, Kaczynski R. CBCT assessment of alveolar
patients: repetitive weekly protocol of alternate rapid maxillary ex- buccal bone level after RME. Angle Orthod 2013;83:110-6.
pansions and constrictions. Cleft Palate Craniofac J 2005;42: 30. Bernd GP. Immediate effects of rapid maxillary expansion on alve-
121-7. olar height and width with Haas-type and hyrax-type palatal ex-
9. Graber LW, Vanarsdall RL, Vig KW. Orthodontics: current princi- panders in tomographies [dissertation]. Porto Alegre, RS, Brazil:
ples and techniques. St Louis: Elsevier Health Sciences; 2011. Pontifical Catholic University of Rio Grande do Sul; 2011.

American Journal of Orthodontics and Dentofacial Orthopedics October 2018  Vol 154  Issue 4
516 Lemos Rinaldi et al

31. Rungcharassaeng K, Caruso JM, Kan JY, Kim J, Taylor G. Factors 42. Barber AF, Sims MR. Rapid maxillary expansion and external root
affecting buccal bone changes of maxillary posterior teeth after resorption in man: a scanning electron microscope study. Am J Or-
rapid maxillary expansion. Am J Orthod Dentofacial Orthop thod 1981;79:630-52.
2007;132:428.e1-8. 43. Akyalcin S, Alexander SP, Silva RM, English JD. Evaluation of
32. Sameshima GT, Sinclair PM. Characteristics of patients with severe three-dimensional root surface changes and resorption
root resorption. Orthod Craniofac Res 2004;7:108-14. following rapid maxillary expansion: a cone beam computed to-
33. Weltman B, Vig KW, Fields HW, Shanker S, Kaizar EE. Root resorp- mography investigation. Orthod Craniofac Res 2015;18(Suppl
tion associated with orthodontic tooth movement: a systematic re- 1):117-26.
view. Am J Orthod Dentofacial Orthop 2010;137:462-76: 44. Sameshima GT, Sinclair PM. Predicting and preventing root
discussion: 11A. resorption: part I. Diagnostic factors. Am J Orthod Dentofacial Or-
34. Nakano Y, Yamaguchi M, Fujita S, Asano M, Saito K, Kasai K. Ex- thop 2001;119:505-10.
pressions of RANKL/RANK and M-CSF/c-fms in root resorption 45. Cook VC, Timock AM, Crowe JJ, Wang M, Covell DA. Accuracy of
lacunae in rat molar by heavy orthodontic force. Eur J Orthod alveolar bone measurements from cone beam computed tomogra-
2011;33:335-43. phy acquired using varying settings. Orthod Craniofac Res 2015;
35. Bao X, Liu Y, Han G, Zuo Z, Hu M. The effect on proliferation and 18(Suppl 1):127-36.
differentiation of cementoblast by using sclerostin as inhibitor. Int 46. Leung CC, Palomo L, Griffith R, Hans MG. Accuracy and reliability
J Mol Sci 2013;14:21140-52. of cone-beam computed tomography for measuring alveolar bone
36. Lim WH, Liu B, Hunter DJ, Cheng D, Mah SJ, Helms JA. Downre- height and detecting bony dehiscences and fenestrations. Am J Or-
gulation of Wnt causes root resorption. Am J Orthod Dentofacial thod Dentofacial Orthop 2010;137(Suppl):S109-19.
Orthop 2014;146:337-45. 47. Timock AM, Cook V, McDonald T, Loo MC, Crowe J, Benninger BL,
37. Remington DN, Joondeph DR,  Artun J, Riedel RA, Chapko MK. et al. Accuracy and reliability of buccal bone height and thickness
Long-term evaluation of root resorption occurring during or- measurements from cone-beam computed tomography imaging.
thodontic treatment. Am J Orthod Dentofacial Orthop 1989; Am J Orthod Dentofacial Orthop 2011;140:734-44.
96:43-6. 48. Sun Z, Smith T, Kortam S, Kim DG, Tee BC, Fields H. Effect of bone
38. Lupi JE, Handelman CS, Sadowsky C. Prevalence and severity of thickness on alveolar bone-height measurements from cone-beam
apical root resorption and alveolar bone loss in orthodontically computed tomography images. Am J Orthod Dentofacial Orthop
treated adults. Am J Orthod Dentofacial Orthop 1996;109:28-37. 2011;139:e117-27.
39. Phillips JR. Apical root resorption under orthodontic therapy. 49. Sun L, Zhang L, Shen G, Wang B, Fang B. Accuracy of cone-beam
Angle Orthod 1955;25:1-22. computed tomography in detecting alveolar bone dehiscences
40. Kalkwarf KL, Krejci RF, Pao YC. Effect of apical root resorption on and fenestrations. Am J Orthod Dentofacial Orthop 2015;147:
periodontal support. J Prosthet Dent 1986;56:317-9. 313-23.
41. Lee KS, Straja SR, Tuncay OC. Perceived long-term prognosis of 50. J€ager F, Mah JK, Bumann A. Peridental bone changes after ortho-
teeth with orthodontically resorbed roots. Orthod Craniofac Res dontic tooth movement with fixed appliances: a cone-beam
2003;6:177-91. computed tomographic study. Angle Orthod 2017;87:672-80.

October 2018  Vol 154  Issue 4 American Journal of Orthodontics and Dentofacial Orthopedics