Incisor root resorption associated with palatally displaced maxillary canines: Analysis and prediction using discriminant function analysis


Incisor root resorption associated with palatally displaced canine (PDC-IRR) is an important phenomenon that might alter the orthodontic treatment plan. The aims of this study were to investigate the prevalence and characteristics of PDC-IRRs, to compare PDC-IRR and PDC-No-IRR groups, to identify predictors for PDC-IRR, and to produce a valid PDC-IRR prediction model using discriminant function analysis.


Cone-beam computed tomography images for 107 palatally displaced canines (PDCs) and 51 fully erupted nondisplaced canines (NDC) were analyzed for the presence of incisor root resorption (IRR). The PDCs were divided into 2 groups: PDC-IRR (52 canines) and PDC-No-IRR (55 canines). The 3 groups were compared, and the following variables were measured: canine follicle size, contact with adjacent incisors, associated dental abnormalities, sector analysis, canine vertical relation to adjacent root, lateral incisor angulation and inclination, canine angulation and distance to reference planes (dental midline, midpalatal suture, occlusal plane, pterygoid vertical plane, and lateral incisor), and space available for canine in the dental arch. Discriminant functional analysis was used to produce a discriminant function equation to predict PDC-IRR.


Root resorption affected 74% of lateral incisors adjacent to PDCs and 25.5% of lateral incisors adjacent to NDCs. IRR was located lingually in 41% of PDCs and in the mid and apical thirds in 89%. In the PDC-IRR group, 94.2% of canines were in contact with adjacent roots compared with 23.6% in the PDC-No-IRR group ( P <0.001). Canine follicle size was >2 mm in 42% of PDC subjects. Overall mean value of follicle width in PDC canines was 1.79 ± 1.06 mm. Peg-shaped lateral incisors were recorded in 19% of PDC subjects compared with 4% in NDCs ( P = 0.012). The stepwise analysis using cone-beam computed tomography records produced a discriminant function based on 3 variables. The analysis correctly predicted the outcome in 83% of subjects. The variables contributing to the prediction equation were canine contact with adjacent incisors, size of canine dental follicle, and the presence of peg-shaped lateral incisor.


Canines in contact with adjacent incisor roots was the only risk factor detected for PDC-IRR. Very good predictive discrimination (83%) emerged for PDC-IRR subjects. Predictors of PDC-IRR were canine contact with adjacent incisor, size of canine dental follicle, and presence of peg-shaped lateral incisor.


  • Incisor root resorption (IRR) is more common with palatally displaced canines.

  • Canine in contact with adjacent incisor root was the only risk factor detected for IRR.

  • Size of canine dental follicle and presence of peg-shaped lateral incisor were predictors for IRR.

  • Very good predictive discrimination (83%) emerged for PDC-IRR subjects.

  • All PDC-IRR subjects were correctly predicted when the derived equation was tested.

The displaced or ectopic canine is the one diverging from its normal path of eruption, either to erupt in an unusual position or to become impacted buccally or palatally. The reported prevalence of maxillary canine displacement or impaction varies between different populations. It ranges from 1% to 5%. Numerous sources in the literature suggest that the majority of ectopic canines in Caucasians were palatal.

The specific etiology behind palatally displaced maxillary canines (PDC) is unknown. Two broad theories exit: the guidance theory and the genetic theory. Guidance theory of palatal canine displacement suggests that environmental factors give rise to palatal displacement of canines such as congenitally missing lateral incisor, supernumerary teeth, odontome, and other local factors that interfere with its eruption path. In contrast, the genetic theory suggests that it is a genetically predetermined anomaly, with polygenic, multifactorial inheritance. All causative elements and associated phenomena such as long path of eruption, abnormal position of tooth bud, hard tissue obstructions, soft tissue lesions, and anomalies are etiologically construed as genetic, including the aberrant canine itself.

In addition to the orthodontic sequela and its consequences on the occlusion, several complications can result from maxillary canine ectopia. The most common serious complication is external root resorption of incisors, and so often, it is underestimated or missed. Once diagnosed on plain radiography, the resorption would be in an advanced stage.

Incisor root resorption associated with maxillary canine is common in ectopic cases. Both buccally and palatally displaced canines can cause adjacent root resorption. Resorption of lateral incisor can also be caused by normally erupted maxillary canines.

Radiological examination is an essential part of the diagnostic process of impacted canines and associated incisor root resorption. Serrant et al suggested that cone-beam computed tomography (CBCT) is more accurate than either horizontal or vertical parallax for the localization of ectopic maxillary canine teeth. CBCT is capable of providing accurate, submillimeter-resolution images with isotropic voxels, allowing 3-dimensional (3D) visualization of the complex maxillofacial region.

The reported prevalence of IRR associated with PDCs has increased over the time, because the methods for its radiographic detection has got more advanced. With the application of a 3D imaging, it reached a percentage of 48%. Recent CBCT studies reported 25%-67.6% prevalence for lateral incisor and 5%-18% for central incisor roots resorption adjacent to PDCs in Caucasians.

The true etiology of incisor root resorption associated with palatally displaced canines (PDC-IRR) is not well understood. However, the removal of the canine seems to stop progression of resorption even in severe cases. Therefore, physical proximity of canine and/or its follicle seem to be the most likely risk factors although results contradictory. Ericson and Kurol found that IRR cases showed a more advanced dental development, a more medial canine position in the dental arch and a slightly more mesial horizontal path of eruption. With the use of 3D imaging, the exact contact relationship of the canine crown follicle and the adjacent root can be assessed accurately. Some studies explored the association of IRR with PDC canine follicles larger than 3 mm. They reported no relationship between resorption and enlarged dental follicles. On the other hand, Chaushu et al reported that dental follicles wider than 2 mm increased the risk by 8.3-fold compared with normal dental follicles.

In the present investigation, CBCTs were used to study PDC-IRRs. The aims were to report on the prevalence and characteristics of PDC-IRRs, to compare PDC-IRR and PDC-No-IRR groups, to identify predictors for PDC-IRR, and to produce a valid PDC-IRR prediction model using discriminant function analysis.

Material and methods

Ethical approval was obtained from Institutional Review Board of Jordan University of Science and Technology (JUST). The sample of this study was collected over a period of 2 years by 3 means: database search (existing CBCT that were taken for diagnostic purposes as part of comprehensive orthodontic treatment), referrals by fellow dentists and orthodontists, and screening school-aged children and referring them for further examination and treatment at orthodontic clinics/JUST.

All patients in the study had been referred for CBCT imaging after clinical and radiographic (panoramic or periapical radiographs) examination. The CBCT images taken between January 2010 to October 2016) were screened for the presence of PDCs. Inclusion criteria were nonsyndromic and noncleft patients, with no previous orthodontic treatment involving maxillary incisors, no contributing history of trauma; no root canal treatment and no presence of cysts or other pathologies.

University students and school-aged children were screened (clinical examination, palpation) for the presence of maxillary PDC. Subjects suspected to have PDC were referred to the radiology department, where localization of canine was carried out radiographically (2 posterioanterior radiographs were taken). If posterioanterior radiographs showed a palatal displacement along with some overlapping of the canine over the adjacent incisor, a CBCT image was taken as a part of orthodontic pretreatment records. If those subjects were able to pay the cost of orthodontic treatment, they were referred to postgraduate orthodontic clinics for fixed appliance treatment.

Sample size for this study was calculated using the formula, n = Z 2 (1-α) p (1-p)/d 2 , where Z is the 2-sided Z value required for the 90% confidence interval, which is equal to 1.645, p is the prevalence of maxillary lateral incisor root resorption, which was estimated from previous study ( P = 25%), and d is the allowable error (0.1). Therefore, the sample size for each group calculated using the above formula was 51 canines.

Canines were determined to be displaced or not displaced palatally, as follows: PDCs—canines appearing palatal to a line connecting the roots of adjacent teeth at any level of canine crown (n = 107), and nondisplaced canines (NDCs)—canines that are normal and erupted as should in the dental arch (n = 51).

IRR was registered by sliding along the incisor root in cross-sectional and longitudinal views (both perpendicular to the tooth’s long axis). Any loss of root structure, that is, clear demarcated crater or irregular surface loss (depression) compared with areas above and below was recorded. The degree of IRR severity was determined according to Ericson and Kurol criteria as no resorption, mild, moderate, and severe resorption ( Table I ; Fig ).

Table I
Methods used to measure root resorption, canine follicle width, proximity and contact relationship presence of dental anomalies and sector analysis
Variables Method
Root resorption
Severity of IRR No resorption. Intact root surfaces, cementum layer may be lost.
Mild resorption. Up to half of the dentine thickness to the pulp
Moderate resorption. Resorption midway to the pulp or more, pulp lining being unbroken
Severe resorption. Pulp is exposed by the resorption
Location of IRR Cross-sectional. Recorded as distal, mesial, buccal, lingual or distolingual
Vertically. Recorded at the deepest point corresponding to cervical, middle, or apical thirds of incisor root.
Extent of IRR Registered whether resorption extended for less than a third of root length, one third to half, or more than a half the root length.
Canine follicle width. Graded in 1-mm intervals, in accordance with Chaushu et al Grade 1—follicle width is 0-1 mm
Grade 2—follicle width is 1-2 mm
Grade 3—follicle width is 2-3 mm
Grade 4—follicle width is 3 mm or more.
Proximity and contact relationship Location of contact on the cross-section of the adjacent root was recorded to be distal, mesial, buccal, lingual or distolingual.
Presence of dental abnormalities Agenesis of teeth, supernumeraries, impacted teeth other than third molars, peg-shaped lateral incisors were recorded.
Sector analysis. In accordance with Lindauer et al Sector I. Cusp tip distal to a line tangent to the distal aspect of lateral incisor. It is the reference for normality as it corresponds to the primary canine.
Sector II. Mesial to sector I, with the cusp tip distal to midline of lateral incisor.
Sector III. Mesial to sector II, with the cusp tip distal to central incisor distal aspect.
Sector IV. Mesial to sector II, with the cusp tip distal to midline of central incisor.
Sector V. Any position mesial to sector IV.

Severity of incisor root resorption: A, mild resorption; B, moderate resorption; and C, severe resorption.

To determine a clearly defined cutoff point in choosing the PDC-IRR group, subjects had to show moderate or severe root resorption, as this could be identified more clearly and with a higher sensitivity. Mild resorption was grouped with the PDC-No-IRR group to overcome any error that may affect the diagnostic capacity of the CBCT images in cases of mild resorption. Accordingly, PDCs were divided into 2 groups: group 1 (PDC- IRR)—PDCs associated with moderate to severe IRR (52 canines), and group 2 (PDC-No-IRR)—PDCs associated with no or mild IRR (55 canines).

A total of 82 Caucasian patients (19 male and 63 female) were included in the study. Age ranged from 14 to 31 years with an average age of 20.84 ± 8.54 years. PDC was present bilaterally in 25 patients and unilaterally in the remaining patients (24 patients on the right side and 33 on the left side).

A total of 51 patients with unilateral PDC had normally erupted canines on the contralateral side (11 male and 40 female; average age 21.79 ± 5.54 years). These canines were included in the study, analyzed, and served as a control (NDC). The remaining 6 unilateral PDC patients had buccally displaced canines on the contralateral side, and thus these canines were excluded.

Two similar CBCT apparatuses were used. They were CS 9500 Cone Beam 3D System (Carestream Health, Rochester, NY) with a flat panel detector. The CBCT images were 0.2 mm slices of a medium field of view, where the maxillofacial area was examined at a tube voltage of 90 kV, a tube current of 10 mA, and exposure time of 8.01 seconds. The imaging area was a cylinder with a height of 15 cm and a diameter of 9 cm providing 0.2 mm cubic voxels.

Examinations were performed by 360° rotations with patient in an upright position and Frankfort horizontal plane parallel to ground. CBCT orthogonal views were used for linear measurements. 3D reconstruction view was used to aid in assessing peg-shaped laterals. Oblique views passing through the long axis of canines and incisors were used to evaluate tooth length, IRR, and buccolingual and mesiodistal widths of crowns. Dental pantomographic reconstruction view was used to record some angular measurement, peg-shaped laterals, sector and overlap. Custom focal trough (image layer) selection for panoramic reconstruction was symmetrical and coinciding with maxillary teeth for all patients. Dental pantomographic reconstructed images were examined at a thickness of 15-30 mm.

Tables I and II present the variables included in the study and methods used to measure them.

Table II
Methods used to measure canine vertical relationship, lateral incisor- and canine-related variables, canine angulation, and canine distance to reference planes
Variables Method
Canine vertical relationship to adjacent incisor root Recorded as its tip corresponds to cervical, mid, or apical third of adjacent root.
Lateral incisor-related variables Mesiodistal and buccolingual width of the lateral incisor
Lateral incisor length. Measured on oblique view in mm
Mesiodistal tipping of lateral incisor. Angulation to midline measured on coronal and DPT view.
Buccolingual inclination of lateral incisor. Angulation to occlusal plane measured on sagittal view
Canine-related variables Mesiodistal and buccolingual width of the canine.
Canine length. Measured on oblique view in mm
Canine angulation to reference planes Angle to midline ( Alpha angle): Angulation of canine to midline measured on the coronal view.
Angle to occlusal plane ( Delta angle): Angulation of canine to OP, measured on the sagittal view.
Angle to midpalatal suture ( Gamma angle): Angulation of canine to midpalatal suture measured on the axial view.
Canine angulation to lateral incisor ( Beta angle): Angle between canine long axis and lateral incisor, measured on coronal view.
Canine distance to reference planes Canine distance to occlusal plane: Vertical distance from canine tip to OP in mm, measured on the sagittal view. Canines below the plane were given a negative value.
Canine distance to dental midline: Horizontal distance from canine tip to midline in mm, measured on the coronal view. Canines crossing the midline were given a negative value.
Canine distance to midpalatal suture: Horizontal distance from canine tip to midpalatal suture in mm, measured on the axial view. Canines crossing the suture were given a negative value.
Canine distance to pterygoid vertical (d4): perpendicular distance from the distal aspect of canine crown to pterygoid vertical plane. Pterygoid vertical plane is a perpendicular tangent to the posterior border of the pterygomaxillary fissure.

DPT , dental pantomographic; OP , occlusal plane.

CBCT images were evaluated throughout a period of a month, by one examiner (AA). All images were evaluated in dimmed light using a screen with 1920 × 1200 pixels. Measurements were made using installed CBCT software, CS 3D imaging viewer (CS 3D imaging viewer, 3.2.9, Carestream Health). When necessary, the window settings were adjusted to optimize the images for evaluation and zoomed in as much as needed for a careful evaluation.

A random sample of 10 CBCT images (10% of total sample) where re-evaluated after 1-month interval. All measurements were repeated by the same examiner in the same conditions to test intraexaminer reliability. Kappa test was used for the categorical data and Dahlberg error for the double measurement was used to calculate the standard error of the method. Kappa values were above 92% for all the measured categorical variables. Dahlberg error ranged from 0.16° for canine angulation to occlusal plane to 0.018 mm for canine distance to occlusal plane.

Statistical analysis

Statistical analysis was performed with the use of the Statistical Package for Social Science computer software (version 23, SPSS Inc, Chicago, Ill). The Shapiro-Wilk test was used to assess normality of numerical data. The values were ≥0.05, indicating the data was normal.

Descriptive statistics were calculated for all the measured variables for each group. Differences between groups were assessed using ANOVA and chi-square test for linear and categorical data, respectively. Bonferroni post-hoc multiple comparisons test was used for numerical data to identify differences between the groups. The level of significance was set at P ≤0.05. Discriminant functional analysis (stepwise analysis) was used to produce a discriminant function score. The function was statistically significant at P <0.001. The discriminant function equation derived from original sample was applied to a fresh sample of 10 patients (8 female and 2 male; average age 18.52 ± 3.40 years) who fulfilled the same selection criteria as the main sample. PDCs were present bilaterally in 2 patients and unilaterally in the remaining subjects.

The data for each patient were entered in the equation. A positive discriminant score indicated prediction of IRR. Percentage of correct prediction was calculated.


Sex differences in PDC-IRRs were not detected (χ 2 = 0.190, P value = 0.663). A total of 111 maxillary incisors (92 laterals and 19 centrals) had some degree of root resorption associated with canine eruption ( Table III ). All 19 resorbed central incisors belonged to the PDC group. Root resorption affected 25.5% of lateral incisors adjacent to NDCs ( P <0.001). Root resorption affected 74% of lateral incisors adjacent to PDCs, together with 18% of central incisors. IRR in the PDC group was significantly higher than the NDC group ( P <0.001).

Table III
Frequency of IRR in the studied groups (PDC and NDC)
Group (number of canines) Lateral incisors IRR severity Total Central incisors IRR severity Total IRR
None Mild Moderate Severe None Mild Moderate Severe
Bilateral PDC (50) 12 (24) 16 (32) 8 (16) 14 (28) 38 (76) 46 (92) 2 (4) 0 (0) 2 (4) 4 (8)
Unilateral PDC (57) 16 (28) 14 (24.5) 10 (17.5) 17 (30) 41 (72) 42 (74) 5 (9) 2 (3) 8 (14) 15 (26)
Total PDC (107) 28 (26) 30 (28) 18 (17) 31 (29) 79 (74) 88 (82) 7 (7) 2 (2) 10 (9) 19 (18)
NDC (51) 38 (74.5) 6 (12) 7 (14) 0 (0) 13 (25.5) 51 (100) 0 (0) 0 (0) 0 (0) 0 (0)
Only gold members can continue reading. Log In or Register to continue

Jan 7, 2020 | Posted by in Orthodontics | Comments Off on Incisor root resorption associated with palatally displaced maxillary canines: Analysis and prediction using discriminant function analysis
Premium Wordpress Themes by UFO Themes