Are facial, maxillary arch and incisor dimensions related in patients with a unilaterally impacted palatal canine? A prospective investigation

Palatal canine impaction has a prevalence of 0.8 to 3%1 with a female predilection.2 Although the aetiology has been proposed to arise from a disturbance in the dental lamina with a polygenic, multifactorial inheritance1 or, due to an anomalous or absent lateral incisor,3 the exact cause has yet to be determined. An early diagnosis and timely intervention prevents resorption of the roots of adjacent lateral incisors.4 Using CT images, mid-facial dimensions, specifically nasal width, have been investigated as a possible aetiological factor related to palatal canine impaction but a control group was not compared.5 In addition, facial soft tissue variables were not assessed but these may be relevant and easier for the clinician to evaluate.

There is conflicting evidence regarding the maxillary transverse dimension in patients with a palatally impacted canine.5–9 While maxillary inter-canine (IC) and inter-molar (IM) widths have been reported to be reduced,5–7 others have detected an increase in maxillary arch width.8,9 Similarly, palatal depth has been found to be smaller7 or of minimal difference relative to control subjects,10 while palatal area has been reported to be reduced on the side of the impacted palatal canine.7 A lesser mesiodistal maxillary lateral incisor width, as well as size discrepancies in female patients has also been implicated.11–13 In the light of this evidence, the relevance of these factors remains unclear.

Dental and maxillary arch parameters in a sample of patients presenting with a palatally impacted canine have been compared to control subjects.7,10,14 No study has collectively compared the possible correlations between facial, maxillary arch and incisor dimensions in patients presenting with a unilaterally impacted palatal canine against a control group or has evaluated 3D facial soft-tissue dimensions within these groups.

The aim of the present prospective study in Caucasian adolescents was to identify possible relationships between facial, maxillary arch and incisor dimensions associated with a unilaterally impacted palatal canine. The null hypothesis stated that, ‘in Caucasian adolescents, there were no differences in the relationships between soft-tissue facial, maxillary arch and maxillary incisor dimensions associated with a unilaterally impacted palatal canine compared to control subjects’.

Ethical approval was granted by the local clinical research ethics committee (ECM 4 (n) 05/11/13). Written informed consent was obtained from each participant. All Caucasian patients diagnosed with a unilaterally impacted palatal canine and referred to a university dental teaching hospital over one calendar year, were prospectively recruited (Table I). Using a parallax technique, one examiner radiographically and subsequently clinically confirmed the diagnosis. Simultaneously, a matched control group of consecutive non-orthodontic subjects without a unilaterally impacted palatal canine was recruited. In the absence of facial soft-tissue parameters related to the impacted canine, the sample size calculation was based on data available for intercanine width. To detect a 1.50 mm difference in inter-canine width between the impacted canine and control groups, assuming a standard deviation of 2.60 mm, at 80% power (P < 0.05), 50 subjects were recruited for each group.7 This was increased to 54 to account for any image quality issues. All assessments were undertaken by one observer.

Facial dimensions

A stereophotogrammetric camera system (www.di4d.com)15 connected to a Dell Dimension 980 personal computer was re-calibrated for each session and images captured using Di3D Capture software (Dimensional Imaging, Glasgow, Scotland, UK). The accuracy of the system was 0.1 mm.16 Following instruction, one operator experienced in 3D image capture, recorded each subject when at rest using the verbal cue recommended by Zachrisson.17 Subjects practiced this twice before each image was taken. The 3D co-ordinates of twelve landmarks were identified from each image (Figure 1A, Table II).18 For each subject, 3D distances and angles were then recorded (Tables II, III).19 To determine the face ratio (face height divided by face width; Figure 1B), 2D facial images were exported from Di3D software and landmarks identified (Paint, Microsoft Corporation, Redmond, Seattle, USA). Facial shape was visually defined using these images as square, tapered or ovoid.20

Figure 1.

Table II.

Facial Landmarks.

Landmark name	Definition
Midpoint between the eyes (mid)	The halfway distance between exocanthion left and right
Soft tissue nasion (n)	Point in the midline of both the nasal root and nasofrontal suture
Exocanthion right (exR)	Outer most point on commissure of right eye fissure
Exocanthion left (exL)	Outer most point on commissure of left eye fissure
Pronasale (prn)	Most protruded point of apex nasi
Subnasale (sn)	Midpoint of angle where lower nasal septum and lips meet
Alare right (alR)	Most lateral point of the right alar contour
Alare left (alL)	Most lateral point of the left alar contour
Alar curvature Right (acR)	Most lateral point in the curved base line of right alar base
Alar curvature Left (acL)	Most lateral point in the curved base line of left alar base
Labiale superioris (ls)	Midpoint on upper vermilion border
Pogonion (pg)	Most anterior midpoint on chin

Table III.

3D distances and angles.

3D distances	Definition	3D angles	Definition
Eye distance	Distance between exR-exL	Mid-face angle	Angle between exR-pg-exL
Nose width	Distance between alR-alL	Face convexity	Angle between s-sn-pg
Basal width	Distance between acR-acL	Nose prominence	Angle between n-pr-sn
Total face height	Distance between n-pg	Philtrum depth	Angle between prn-sn-ls
Lower face height	Distance between ls-pg
Mid- face height	Distance between ls- mid

Maxillary incisor dimensions (Figure 2)

Figure 2.

Maxillary central and lateral incisor shape

These were classified as square, tapered, round, square/tapered or square/round.

Maxillary central and lateral incisor ratio

This was calculated by dividing crown height by crown width.22

The sample characteristics are provided in Table IV in relation to the 54 subjects (37 females, 17 males) in each group. There was no significant difference between the groups with respect to age, gender and malocclusion (P < 0.05). There was no disagreement in radiographic localisation of the impacted palatal canine between the first and second assessments. The landmark error for the 3D linear and angular measurements was 0.58 mm. Similar small errors were identified for the dental parameters of palatal depth (0.09 mm), palatal area (0.42 mm²) and anterior Bolton ratio (0.14%). The landmark error for shape assessment was 0.05 mm with complete agreement for classification of the face, arch and tooth shape. None of these errors were considered clinically significant.

The present prospective case-control study assessed soft-tissue facial, maxillary arch and incisor dimensions in relation to patients presenting with a unilaterally impacted palatal canine. The null hypothesis was rejected on the basis that mean nasal basal width and the anterior Bolton ratio differed significantly between patients presenting with a unilaterally impacted palatal canine compared with a control group. No other differences were found between the groups and no correlation was found between the variables.

Bilateral palatal10,23 or buccal canine impactions were not included due to their different aetiologies.10,23,24 Their exclusion is consistent with previous studies and avoided heterogeneity of the sample, which would have adversely affected the results.5,25 The reliability of radiographic canine localisation using vertical parallax was excellent. Mean 3D landmark identification error was low (0.58 mm), similar to findings of a previous 3D facial investigation.26 Landmark identification errors on digital models along with the errors of measurement and shape assessments were similarly small and lower than formerly recorded.27

In the present study, the severity of unilaterally impacted palatal canines was not determined by cone-beam CT. There were no facial, maxillary arch nor incisor morphological variables excluded. All malocclusion groups were included making the findings able to be generalised for Caucasians. No previous prospective investigation has assessed the individual and collective relationships between facial, maxillary arch and incisor dimensions associated with a unilaterally impacted palatal canine. Bias was minimised by the recruitment of consecutive Caucasian patients presenting with a unilaterally impacted canine with whom comparison was made using prospectively recruited control subjects matched for age, gender and malocclusion and without a unilaterally impacted palatal canine.28

The greater female to male ratio in the unilaterally impacted palatal canine group supported previous findings1 and, as there were no significant differences in baseline characteristics, the males and females were combined and analysed following the protocol of an earlier investigation.7 There was a greater prevalence of Class I malocclusion (44%) cases which is broadly consistent with existing studies.7,25,29 A greater prevalence of palatal canine ectopia, however, has been identified in patients presenting with a Class II division 2 malocclusion.30,31

There was no significant difference in either face shape nor ratio between the groups, although an ovoid or square shape was more common in the impacted canine group. Kim et al.5 found no difference in either mean nasal cavity width nor mean nostril width between palatally and buccally impacted maxillary canines but did not incorporate an untreated control group. In the present study, mean nasal basal width was smaller by almost 2 mm in the unilaterally impacted palatal canine group compared to the control group. Facial soft tissue scanning was conducted in the present study, whereas Kim et al.5 used CT scanning. The means by which a narrower nasal basal width influences the developmental path of the maxillary canine remains to be explored.

No significant dental arch differences were identified between the unilaterally impacted palatal canine and control groups in support of previous results.10,32,33 However, in contrast, the mean intercanine width was not significantly reduced in the impacted canine group.7,32,34 A significantly larger mean anterior Bolton ratio was noted in the impacted canine group compared to the control group.35,36 The mean anterior Bolton ratio in the control group (76.6) was similar to that originally recorded by Bolton.37 Although there were no significant differences between the groups in relation to the tooth ratio analyses, the larger anterior Bolton ratio incorporates crown evaluation of the six maxillary anterior teeth and is not specific to the lateral incisor. Leonardi et al. found a shorter maxillary lateral incisor root associated with a palatally impacted canine, which may be more relevant to the guidance theory.38

No correlation was found between any of the variables assessed in the impacted canine group and the differences identified between groups were small. Consequently, the aetiology of a unilaterally impacted palatal canine would appear more likely to be genetically controlled1 rather than due to facial, maxillary arch or dental morphological anomalies. Future studies should consider the investigation of potential biomarkers for risk factors related to palatal canine ectopia.

Patients presenting with a unilaterally impacted palatal canine had a narrower mean nasal basal width and a larger mean anterior Bolton ratio compared to a control group. The clinical significance of the differences was considered minor. Facial, maxillary arch and incisor dimensions were neither individually nor collectively correlated with a unilateral impacted palatal canine, which may lend support to an underlying genetic aetiology.