The variation in crown-root morphology of anterior teeth assessed with cone-beam computed tomography

ABSTRACT Objective: To determine the discrepancy of crown-root morphology of anterior teeth, using cone-beam computed tomography (CBCT), and to provide a guidance for proper torque expression. Methods: A total of eligible 200 CBCT were imported into Invivo v. 5.4 software, to obtain the middle labio-lingual sections of anterior teeth. AutoCAD 2007 software was applied to measure the crown-root angulation (Collum angle) and the angle formed by a tangent to the center of the labial surface and the long axis of the crown (labial surface angle). SPSS 18.0 was used for statistical comparisons of the two measurements, at the level of p< 0.05, and the Pearson correlation analysis was applied to investigate the association between the two measurements. Results: The value of Collum angle in maxillary central incisor was close to 0°. Significantly negative Collum angle in lateral incisors and maxillary canine, and positive value in mandibular canine were detected (p < 0.001). The labial surface angle in canine was significantly greater than the intra-arch incisors (p< 0.001), and no significant difference was detected between the central and lateral incisors (p > 0.05). Notably, there was also a significant positive correlation between the two measurements. Conclusions: The crown-root angulations were greatly different among anterior teeth. Accompanying the obvious crown-root angulations, the canines both in maxillary and mandibular arches presented considerable labial surface curvatures. Hence, equivalent deviation during bracket bonding might cause greater torque expression error and increase the risk of alveolar fenestration and dehiscence.


INTRODUCTION
Appropriate anterior torque expression is essential for establishing stable occlusion and satisfying esthetics. In the past two decades, researchers have focused at the influence of the morphology of anterior alveolar height and thickness on anterior torque, 1 while the tooth morphological variation has been frequently ignored. In 1984, Bryant et al 2 firstly noticed the variability in the maxillary permanent central incisor morphology by establishing three anatomic features, and investigated the discrepancies among different Angle malocclusions. Thereafter, two features were widely adopted by the following studies. [3][4][5][6] One feature was the crown-root angulation (Collum angle, CA) formed by the long axis of root and crown, in the labiolingual direction. Previous studies found the connection between root apex and incisal edge did not pass through the middle point of the connection between labial and lingual CEJ 2 (Fig 1), which further restricted the degree to which the roots of these teeth can be lingually torqued when close to the maxillary palatine cortical bone plate. 2,7,8 Our recent study concluded that the maxillary central incisor in patients with sagittal skeletal Class II malocclusion and mandibular incisor with Class III malocclusion presented remarkable crown-root angulation. 6 Hence, the diversity in CA might result in unmanageable root position, increasing the incidence of dehiscence and fenestration, and The other feature was the labial surface angle (LSA, representing the facial contour of anterior teeth), which was formed by a tangent to the bracket site on the labial surface and the long axis of the crown, from a proximal view. 2 The significant amount   6 and the analysis were performed with AutoCAD. Furthermore, the regularity of crown-root morphology and effect on torque expression among different anterior teeth were verified.

STUDY DESIGN
The design was a cross-sectional and retrospective study using dental records.

MEASURING THE CAPTURED IMAGE
The CBCT data were three-dimensionally adjusted using Invivo It was essential that the bunch of cutting lines (green) was set vertical to the connection between mesial and distal edges of crown, and located at the center in horizontal view ( Fig 2D).
Thus, the median of the nine images would be selected as the measuring one ( Fig 2E).

MARKER AND MEASUREMENT
AutoCAD (Autodesk, San Rafael, CA) software was used as  was assessed with the Dahlberg's formula: 6,16 in which d i was the difference between the first and second measurement on the i th sample, and n was the whole sample number. As a result, the intra-observer errors were 0.27° for CA and 0.39° for LSA, and inter-observers error were 0.47° for CA and 0.52° for LSA. The values indicated that the analysis was reliable, since all the measurements presented no significant difference, according to the t-test (p > 0.05).

Comparison of CA among different types of intra and inter-dental arch teeth (Tables 3 and 4)
In   indicating less identical crown-root angulation. The CAs in lateral incisor and canine were considerably negative, indicating that the long axis of root deviated labially from the direction of the long axis of crown. However, there was no significant difference between the lateral incisor and canine (p = 1.000) (Fig 4A). In mandible, the mean values of CA in central incisor (-3.97 ± 4.49°) and lateral incisor(-6.50 ± 4.03°) were negative and presented significant discrepancy between each other (p = 0), while CA in canine was significantly positive (3.70 ± 4.91°, p = 0). Thus, the long axis of crown deviated labially from the direction of root in mandibular central and lateral incisors, while deviated lingually in canine (Fig 4A). Comparing the values  between the same types of interdental arch teeth, significant discrepancies were observed between upper and lower central incisors (p = 0), and similar result was also detected between canines (p = 0). However, no significant discrepancy in the value of CAs between upper and lower lateral incisors were found (p = 0.72) (Fig 4B).

Comparison of LSA among different types of intra and inter-dental arch teeth (Tables 3 and 4)
In maxilla, the mean values of LSA for canine presented significantly greater (20.07 ± 3.66°) than for central (15.51 ± 2.91°) and lateral incisors (15.92 ± 3.50°), suggesting a greater facial curvature of crown. No significant statistical difference was detected between the central and lateral upper incisors (p = 0.20) (Fig 4C).
In mandible, the results distribution was similar to maxillary distribution. The mean value of LSA in canine (18.27 ± 3.07°, p = 0) was also significantly greater than central (14.40 ± 3.20°, p = 0) and lateral incisors (14.76 ± 3.25°, p = 0), and no significant difference between the central and lateral lower incisors was found (p = 0.25) (Fig 4C). Further comparison of the values between same types of inter-dental arch teeth demonstrated that the all of the three LSAs in upper teeth appeared greater than for the same types of opposite teeth (p = 0.04, p = 0.03, p = 0), indicating a greater labial surface curvature in maxillary anterior teeth (Fig 4D).

Pearson correlation test between the CA and LSA within the same tooth (Table 5)
The consistency of the difference distribution suggested that there might be some extent of correlation between the two measurements within the same tooth. Thus, the association between CA and LSA within the same tooth was verified using the data from all the samples. As a result, except for the maxillary central incisor, where there was no significant correlation between the two measurements (r = -0.101, p = 0.078), the Pearson correlation test indicated that the two measurements were significantly and positively correlated in other anterior teeth, even though the correlation coefficient was relatively weak (Fig 5). Together, the meaningful positive correlation suggested a regular effect on torque limitation, which was analyzed in the following discussion.

DISCUSSION
The preadjusted, or straight-wire, appliance is widely accepted by orthodontics, based on the assumption that the morphologies of teeth crowns among the individuals present discrete variation. Thus, without considering the variation in facial surface contours, orthodontists believe that the bracket preadjusted torque promotes a coincident extent of labio-lingual inclination with the same setting. 6   In the same tooth, at different heights from incisal edge, each 0.5-mm increase leads to a torque reduction of about 2°. 5 It was demonstrated that the significant variation of LSA would cause a wide range of torque (from 12.3° to 24.9°) when measuring at 4.5 mm from the incisal edge. 9 In the present study, we integrated all the anterior teeth and determined the LSAs formed at the uniform bonding sites of straight-wire brackets.
We found that both in maxilla and mandible, the LSA of canine was significantly greater than the intra-dental arch central and with remarkable crownroot angulation (Fig 6).
The cause of variable Thus, based on a previous study, we suppose that the position of tongue and lip relative to crown and the extent of anterior overbite may be regularly associate with the variable crown-root morphology. Therefore, further research is necessary to verify the potential relationship.
However, it is worth highlighting that the present study mainly focused on the sagittal variation of tooth morphology, but the variation and stress analysis of tooth in three-dimension during torqueing will be recommend to more reasonably explain the effect on the periodontal ligament and alveolar bone. Moreover, the movement restriction of root can be caused by the poor alveolar bone, which should also be integrated into torqueing. The authors report no commercial, proprietary or financial interest in the products or companies described in this article.