See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/272624579

Extraction of the deciduous canine as an interceptive treatment in children

with palatally displaced canines. Part II: possible predictors of success and
cut-off points for a spont...

Article in The European Journal of Orthodontics · February 2015

DOI: 10.1093/ejo/cju102 · Source: PubMed

CITATIONS READS

53 1,113

3 authors, including:

Julia Naoumova Heidrun Kjellberg

Specialist clinic of Orthodontics, University clinics of Odontology Region Västra Götaland
23 PUBLICATIONS 462 CITATIONS 47 PUBLICATIONS 1,902 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Julia Naoumova on 13 October 2016.

The user has requested enhancement of the downloaded file.

 European Journal of Orthodontics, 2015, 219–229
doi:10.1093/ejo/cju102
Advance Access publication February 20, 2015

Randomized Controlled Trial

Extraction of the deciduous canine as an

interceptive treatment in children with palatally
displaced canines—part II: possible predictors
of success and cut-off points for a
spontaneous eruption
Julia Naoumova, Jüri Kürol and Heidrun Kjellberg
Department of Orthodontics, Institute of Odontology at the Sahlgrenska Academy, University of Gothenburg,
Göteborg, Sweden

Correspondence to: Dr Julia Naoumova, Department of Orthodontics, Institute of Odontology at the Sahlgrenska Academy,
University of Gothenburg, Box 450, Göteborg SE-405 30, Sweden. E-mail: julia.naoumova@vgregion.se

Downloaded from by guest on November 28, 2015

Summary
Aim: To analyse factors affecting the success rate of palatally displaced canines (PDCs) and
eruption time and to find cut-off points to predict when interceptive extraction is beneficial versus
unnecessary.
Materials and methods: Sixty-seven patients, 40 girls, 27 boys (10–13 years) with uni- (45) or
bilateral (22) PDCs, persisting deciduous canine and no previous orthodontic treatment were
randomly allocated for extraction or non-extraction using the block randomization method. There
were no dropped out after the randomization or during the trial. Clinical examination and cone
beam computed tomography was performed at 0, 6, and 12 months. Blinded measurements were
done on baseline images.
Results: Erupted PDCs had a significantly smaller mesioangular angle, shorter distance of canine
cusp tip-dental arch plane, and larger distance of canine cusp tip-midline, and the patients were
younger compared to the non-erupted group. Faster eruption was noted of PDCs in the extraction
group. Spontaneous eruption was achieved without prior deciduous canine extraction with cut-
off points: initial canine cusp tip-midline of 11 mm, canine cusp tip-dental arch plane of 2.5 mm,
or a mesioangular angle of 103 degrees. PDCs with a less favourable position, i.e. an initial cusp
tip-midline of 6 mm, a canine cusp tip-dental arch plane of 5 mm, or a mesioangular angle of 116
degrees, will need surgical exposure despite interceptive extraction of the deciduous canine. The
canine cusp tip-midline had the best predictive measure for assessing the outcome.
Limitations: Decision on where to place the cut-off points may differ from one operator to another,
therefore results from several studies are needed to get average cut-off points.
Conclusions: Deciduous canine extraction is the variable that affects the spontaneous eruption of
the canine most. Canine cusp tip-midline, canine cusp tip-dental arch plane, and mesioangular angle
might be useful for distinguishing when an interceptive extraction of the deciduous canine is beneficial
or when exposure of the PDC should be implemented without previous interceptive treatment.
Registration: This trial was registered in ‘FoU i Sverige’ (http://www.fou.nu/is/sverige), registration
number: 40921.
Protocol: The protocol was not published before trial commencement.

© The Author 2015. Published by Oxford University Press on behalf of the European Orthodontic Society. All rights reserved.
219
For permissions, please email: journals.permissions@oup.com
220
Introduction
Maxillary canines are important aesthetically and functionally, and
canine impaction, which occurs in about 2 per cent of the popu-
lation (1), is a dilemma for many orthodontists since its aetiology,
localization, response to preventive treatments, and prediction are
complex. According to the reports from two newly published sys-
tematic reviews evaluating the effect of interceptive treatment in
patients with palatally displaced canine (PDC), the available scien-
tific evidence is too weak to support interceptive treatment (2, 3).
Numerous randomized trials have been published since then, all
pointing on that interceptive extraction of the deciduous canine is
beneficial for spontaneous eruption of the PDC (4–11). The need
for surgical exposure and subsequent orthodontic treatment, a treat-
ment that is often difficult (12), time consuming, and more expensive
(13) than an average orthodontic treatment, would thus decrease.
Early diagnosis of canine displacement and prediction of subsequent
impaction is therefore valuable for the clinician in order to be able
to avoid unnecessary measures.
Several authors have tried to find indicators to predict canine
eruption and the risk for impaction. Ericson and Kurol (4) found
in their uncontrolled prospective study, two possible predictors for
a successful outcome after extraction of the deciduous canine: the
mesiodistal location of the crown and the angulation of the tooth.
In the retrospective uncontrolled study by Power and Short (5), the
angulation of the unerupted canine was also analysed as a predictor
for canine eruption. The authors found that, if the tooth is angled
more than 31 degrees in relation to the midline, there is less chance
for spontaneous eruption after extraction of the deciduous canine.
However, the most significant factor contributing to a successful out-
come was the horizontal overlap of the incisor nearest the displaced
canine (4). The initial vertical position of the canine cusp tip to the
occlusal plane is another indicator that has been suggested as an
important predictor for spontaneous eruption of the canine (14, 15).
Lindauer et al. (16) used the location of the canine cusp tip and
its relationship to the adjacent lateral incisor, which is a modifica-
tion of the Ericson and Kurol (4) method, for prediction of canine
impaction. He reported that this method identifies up to 78 per cent
of the canines that will become impacted, all having the canine cusp
tip located in sectors II–IV, i.e. the canine’s cusp tip overlapping or
mesial to the lateral incisor root. Warford et al. (17) verified these
results and added that the angulation of the canine had little sup-
plementary predictive value as an indicator of eventual impaction.
Moreover, several studies have investigated pre-treatment vari-
ables that may affect the duration, the risk for complications, and
the success rate of impacted maxillary canines surgically exposed
and treated with orthodontics. These factors are: age, vertical height,
mesial displacement of the cusp tip, apex position, and transposition
with a lateral incisor or first premolar (12, 15, 18).
Although the results of previous studies suggest that extraction
of the deciduous canine is effective in patients with PDC, not all
permanent canines erupt in spite of this intervention and some PDCs
erupt even if the deciduous canine is not interceptively extracted. It
would thus be desirable to identify which cases would benefit from
early extraction of the deciduous canine.
Thus the aims of this randomized prospective controlled study
using the cone beam computer tomography (CBCT) technique
were to:
Primary outcomes
1. Analyse how factors such as: initial angulations and positions
of the PDC, extraction of the deciduous canine, and age affect
European Journal of Orthodontics, 2015, Vol. 37, No. 2
the success rate and influence the length of time it takes for the
canine to erupt in successful cases
Secondary outcomes
2. To find cut-off points for the cases in which interceptive
extraction of the deciduous canine is not effective, thus where
exposure of the PDC should be considered directly without
preceding interceptive extraction or where spontaneous erup-
tion of the canine will be achieved without prior interceptive
extraction.
It was hypothesized that there are no significant differences in the
initial angulation and position of the PDC regarding the success rate
and the time it takes for the canine to erupt. It was also hypothesized
that there are no age differences between the successful and unsuc-
cessful group.
Materials and methods
Trial design
The study was designed as a randomized controlled trial with equal
patient allocation to following groups: extraction group (EG) or
control group (CG). Patients with unilateral PDC were randomized
for either extraction or non-extraction of the deciduous canine and
patients with bilateral PDC were randomized to have either the right
or the left deciduous canine extracted.
Downloaded from by guest on November 28, 2015
Subjects
Study setting and eligibility criteria.
Patients were consecutively recruited by the consulting orthodon-
tists from 15 Public Dental Clinics in Gothenburg, Västra Götaland
County Council, Sweden from September 2008 to January 2011.
Patients with the following inclusion criteria were included in
the trial:
• Caucasians aged 10–13 years with either maxillary unilateral or
bilateral PDC
• Persisting deciduous canine
• No previous experience of orthodontic treatment
The canine was considered palatally displaced when clinical
palpation of a labial canine bulge was absent and when the canine
crown was diagnosed on intraoral apical radiographs as palatally
positioned (4), using Clark’s rule (19).
Criteria for exclusion were:
• Crowding in the maxilla exceeding 2 mm
• Ongoing orthodontic treatment
• Resorption of the adjacent teeth, grades 3 and 4, according to
Ericson and Kurol (20), caused by the displaced canine either at
the start of or during the trial
• Craniofacial syndromes
• Odontomas, cysts
• Cleft lip and/or palate
A total of 70 patients were invited to enter the study, of whom
three declined to participate before they were randomized. Thus,
in total, 67 patients were randomly allocated to the EG or the
CG. Of these, 45 patients had unilateral PDC (29 girls, mean age
± SD: 11.2 ± 1.1, and 16 boys, mean age ± SD: 11.7 ± 0.8) and 22
patients had bilateral PDC (11 girls, mean age ± SD: 11.5 ± 0.9,
J. Naoumova et al.
and 11 boys, mean age ± SD: 11.6 ± 0.8). A more detailed explana-
tion of the subjects and the method used is given in part I of this
study (11).
Randomization method
Block randomization method was used and the allocations were con-
cealed in sequentially numbered, sealed opaque envelopes opened
by a dental nurse after the written consent was obtained. To ensure
that equal patients were allocated to each group, the generation of
randomization was performed in blocks of 10.
Treatment protocol and process
CBCT images were taken of all PDCs at the Department of Oral
and Maxillofacial Radiology, Institute of Odontology, Sahlgrenska
Academy, Gothenburg, Sweden, at baseline (T0), after 6 months
(T1) and at 12 months (T2). All patients underwent also a clini-
cal examination at the Department of Orthodontics, University
Clinics of Odontology, Sahlgrenska Academy, Gothenburg,
Sweden. If the canine was clinically visible, i.e. emerged though the
gingiva at any occasion from T1 to T2, no further CBCT images
were taken. Imputation measurement values were instead used for
these teeth (11).
An individual treatment plan was made for patients with
unerupted PDCs at T2. In the case that the canine had improved
its position on the radiographic examination, it was followed up
in both groups until it emerged through the gingiva. However, in
the CG, the deciduous canine was extracted if mobility of the tooth
was not present. If the canine showed impaired or no change in its
position at T2, a combined surgical exposure and orthodontic treat-
ment was carried out, regardless of the group to which the patient
belonged.
Blinding
The same oral radiologist measured all the CBCT images and was
unaware of the group to which the patients belonged. Thus, baseline
measurements were blinded, while T1 and T2 images were not, since
the extraction site was visible.
Ethical issues
The study was pre-approved by the research ethics committee of
Sahlgrenska Academy at the University of Gothenburg, Sweden (Dnr
578-08) and by the radiation protection committee, Sahlgrenska
Academy at the University of Gothenburg, Sweden. Children and
parents received verbal and written information, and informed
consent was provided by the child and the parent or by an adult
with parental responsibilities and rights in accordance with the
Declaration of Helsinki.
Measures of treatment effect
Linear and angular measurements on CBCT images, in an axial, a
sagittal, and a frontal view (Figure 1) (11) were made. The method-
ology for the radiographic measurements is described in a previous
study, where also the reproducibility and validity of the 3D measure-
ments used in this study were assessed (21).
The following outcome measures were assessed:
Primary outcomes
• Differences in radiographic measures of erupted (defined as:
emerged through the gingiva) and non-erupted canines, i.e. suc-
cessful and unsuccessful cases.
221
• Radiographic variables that might influence the duration of the
canine eruption between T1 and T2 and after T2.
• The effect the age of the patient has on the success rate and the
duration of the canine eruption.
Secondary outcomes
• Cut-off points for angles and/or linear measurements that could
be used for predicting the cases in which an interceptive extrac-
tion is unnecessary and when surgical exposure should be done
directly without a preceding interceptive extraction.
An intention-to-treat analysis was made, i.e. all randomized
patients were included in the groups to which they were randomly
assigned regardless of the treatment they actually received, and
regardless of subsequent withdrawal from treatment or deviation
from the protocol, i.e. patients that had extraction of the deciduous
canine in the CG at T2 were still considered to belong to the non-
extracted group.
Statistics
Sample size calculation
The sample size calculation was based on the alpha significance level
of 0.05 and a beta of 0.10 to achieve 90 per cent power to detect
a difference of 5 degrees (SD = 6.38) over time in the sagittal and
mesioangular angles measured in the CBCT images, between the
EG and the CG. According to the calculation a total of 60 patients
Downloaded from by guest on November 28, 2015
with unilateral PDC were needed, i.e. 30 patients in each group (i.e.
EG and CG). Since the inclusion criteria were both unilateral and
bilateral PDC, the more bilateral PDC that were included the fewer
patients were needed. To compensate for possible dropouts during
the study and to increase the power further, a total of 70 patients
were enrolled.
The data were statistically analysed by SAS, version 9.3 for
Windows (SAS Institute Inc., Cary, North Carolina, USA). Arithmetic
means and standard deviations were calculated for the numerical
variables. Dependent and independent t-tests were used to compare
changes in time between and within the groups, respectively. The
independent t-test was used to test the significance of baseline means
(i.e. angular and linear measurements and age) between canines
that erupted between T1 and T2 and after T2 and successful versus
non-successful outcome. Multiple stepwise regression analysis and
logistic regression analysis were used to detect possible predictors of
what influences the eruption of the PDC. The time of eruption of the
PDC was entered as the dependent variable. Logistic regression is a
multivariate statistical method that is used when the dependent vari-
able is dichotomous, as in this case, successful versus unsuccessful
outcome, and the explanatory variables are of any kind. The logistic
regression is based on calculating the odds of the outcome as the
ratio of the probability of having the outcome divided by the prob-
ability of not having it. To determine cut-off points for successful
outcome, i.e. spontaneous eruption without interceptive extraction
of the deciduous canine, or unsuccessful outcome in spite of extrac-
tion of the deciduous canine, the receiver-operating characteristic
(ROC) curve analysis was calculated on the CG and the EG. An
ROC curve is a graphical tool for displaying the accuracy of a medi-
cal diagnostic test, and the goal of the analysis is to determine a cut-
off value in order to minimize the number of false positives and false
negatives, i.e. maximizing the sensitivity (Sn) and specificity (Sp).
The ROC curve plots the true-positive rate (sensitivity) on the Y-axis
and the false-positive rate (1 − specificity) on the X-axis. When the
ROC curve climbs towards the upper left hand corner of the graph, a
222
Figure 1.The different radiographic measurements done on the axial, sagittal, and coronal views of the CBCT images (11). CBCT, cone beam computer tomography.
greater discriminatory ability of the test will be attained, i.e. both the
sensitivity and the specificity will be maximized. Sensitivity in this
study is referred to as the percentage of PDCs in the CG, not erupt-
ing spontaneously in that the radiographical predictors identified
them as not erupting spontaneously, or as the percentage of PDCs in
the EG needing surgical exposure in that the radiographical predic-
tors identified them as needing exposure. Specificity is the percentage
of PDCs in the CG erupting spontaneously that the radiographic
predictors identified as erupting spontaneously, or as the percentage
of PDCs in the EG not needing surgical exposure in that the radio-
graphic predictors identified them as not needing surgical exposure.
The accuracy of the clinical test can be revealed by measuring the
area under the curve (AUC). The larger the area, the better the diag-
nostic tests. An area of 1.0 gives an ideal test, because it achieves
both 100 per cent sensitivity and 100 per cent specificity. A rough
guideline for interpretation of the AUC is: 0.50–0.75 = fair, 0.75–
0.92 = good, 0.92–0.97 = very good, 0.97–1.00 = excellent (22).
Harms
No detrimental effects were detected during the study.
Results
Following randomization, there were no dropped out throughout
the rest of the study.
The success rate between the groups, the amount of canines that
had imputation values, surgical exposure followed by orthodontic
treatment or followed until submerge through the gingiva are pre-
sented in part I of this trial (11). As written in part I, the whole mate-
rial was analysed as two independent observations since a difference
in the correlation between extraction and successful outcome was
not found between independent observations and dependent paired
observations in the bilateral group.
European Journal of Orthodontics, 2015, Vol. 37, No. 2
Downloaded from by guest on November 28, 2015
Primary outcomes
Comparison of baseline variables between erupted versus non-
erupted PDCs
Table 1 shows a comparison of baseline variables of those PDCs
that had erupted or had not at the end of the total observation
period (24 months). The mesioangular angle and the distance of
the canine cusp tip-dental arch plane were smaller and the distance
of the canine cusp tip-midline was larger in the group in which the
PDCs had erupted, indicating that a more favourable position of
the PDC at baseline resulted in eruption of the PDC, regardless of
whether the deciduous canine was extracted or not. However, in the
CG, where the PDC had erupted, and where the deciduous canine
was not extracted, the angles and distances mentioned above devi-
ated less from the ‘ideal’ position in comparison to the EG, indi-
cating that if the deciduous canine was extracted the PDCs may
erupt in spite of a more deviating baseline position. For the non-
erupted PDCs, no significant differences between the EG and the
CG were seen for the angular and linear measurements. Patients
in the ‘erupted’ group were significantly younger. Online supple-
mentary figure 1 exemplifies the above-mentioned results with one
clinical case.
Duration of the canine eruption
PDCs closer to the occlusal plane at baseline, i.e. there were higher
values of the variable vertical position, erupted earlier in the EG
(P = 0.030). A significant change in vertical position (P < 0.004) was
seen between the groups in favour of the EG. Age at baseline did not
have an effect on the time when the PDCs erupted (Table 2). Online
supplementary figure 2, illustrates how the eruption time can differ
between the sites in a bilateral case with a similar degree of canine
displacement.
Using multiple stepwise regression analysis on the erupted
teeth only, the vertical position of the canine showed a significant
Table 1. Differences in baseline means (T0) between erupted and non-erupted canines.

Erupted (n = 40 patients, 48 PDC) Non-erupted (n = 27 patients, 41 PDC)

J. Naoumova et al.

EG (n = 31), CG (n = 17), Differences EG (n = 14), CG (n = 27), Differences P value, erupted versus

Variables at T0 mean ± SD mean ± SD Mean (95% CI) P value, EG-CG* mean ± SD mean ± SD Mean (95% CI) P value, EG-CG* non-erupted*

Mesioangular angle (°) 105.5 ± 2.1 100.7 ± 5.9 4.8 (4.2, 1.4) 0.027* 115.4 ± 6.8 112.8 ± 9.1 2.6 (15.9, 6.9) 0.409 0.000***
Sagittal angle (°) 99.4 ± 8.1 104.9 ± 7.9 −1.1 (3.3, 6.8) 0.625 93.5 ± 10.1 98.2 ± 12.7 −4.8 (11.1, 9.8) 0.538 0.364
Vertical position (mm) 15.2 ± 2.4 14.3 ± 2.2 0.9 (2.1, 2.0) 0.178 16.6 ± 4.0 15.6 ± 2.3 1.0 (4.2, 1.9) 0.502 0.168
Canine cusp tip-dental arch plane (mm) 2.9 ± 1.2 2.0 ± 1.7 0.9 (0.1, 1.3) 0.051* 4.1 ± 1.7 3.7 ± 0.7 0.3 (1.7, 1.3) 0.707 0.000***
Canine root apex-dental arch plane (mm) 2.4 ± 1.2 2.5 ± 1.4 −0.1 (1.6, 1.1) 0.130 2.9 ± 1.8 2.3 ± 1.9 0.6 (3.1, 1.5) 0.599 0.324
Canine cusp tip-midline (mm) 10.5 ± 2.2 12.9 ± 1.7 −2.4 (0.7, 1.7) 0.048* 6.5 ± 2.6 7.7 ± 1.8 −1.2 (1.2, 1.5) 0.311 0.000***
Age, years 11.0 ± 0.9 10.7 ± 0.99 0.1 (0.6, 0.7) 0.676 12.0 ± 1.0 11.9 ± 1.0 1.0 (2.3, 0.8) 0.112 0.000***

CG, control group; CI, confidence interval; EG, extraction group; PDC, palatally displaced canine; SD, standard deviation.
*P value <0.05 is considered as statistically significant. ***P value <0.001.

Table 2. Differences in baseline means of canines erupted between T1 and T2 and those erupted after T2 in the extracted group (EG) and control group (CG), respectively.

EG (N = 45 PDCs) CG (N = 44 PDCs)

Erupted between T1 Erupted between P value*:erupted

and T2 Erupted after T2 Differences T1 and T2 Erupted after T2 Differences between T1 and
Variables at T0 Mean ± SD Mean ± SD Mean (95% CI) P value* Mean ± SD Mean ± SD Mean (95% CI) P value* T2 versus after T2

Mesioangular angle (°) 101.5 ± 6.5 103.7 ± 5.2 −2.2 (1.4, 4.8) 0.210 105.0 ± 6.8 104.2 ± 5.9 0.8 (6.9, 4.5) 0.785 0.282
Sagittal angle (°) 100.9 ± 8.1 98.0 ± 8.2 2.9 (7.9, 6.7) 0.248 98.4 ± 7.8 101.4 ± 8.1 −3.1 (5.2, 6.1) 0.447 0.618
Vertical position (mm) 16.1 ± 2.4 14.5 ± 2.3 1.6 (3.1, 1.9) 0.030* 15.5 ± 0.8 13.8 ± 2.5 1.7 (3.9, 1.6) 0.114 0.004**
Canine cusp tip-dental arch plane (mm) 2.1 ± 1.2 2.7 ± 1.2 −0.5 (0.2, 0.9) 0.129 1.1 ± 1.1 1.7 ± 1.6 −0.5 (0.9, 1.1) 0.463 0.236
Canine root apex-dental arch plane (mm) 2.9 ± 1.8 2.1 ± 1.5 0.8 (1.8, 1.4) 0.743 3.4 ± 1.6 3.1 ± 1.0 0.3 (1.5, 0.9) 0.602 0.905
Canine cusp tip-midline (mm) 10.6 ± 2.7 10.4 ± 1.8 0.2 (1.6, 1.8) 0.120 10.6 ± 1.8 11.1 ± 1.7 −0.5(1.2, 1.3) 0.519 0.181
Age, years 11.3 ± 0.9 11.3 ± 0.9 0.0 (0.5, 0.7) 0.964 11.2 ± 0.7 10.9 ± 0.8 0.3 (0.7, 0.5) 0.510 0.272

CI, confidence interval; PDC, palatally displaced canine; SD, standard deviation.
*P value <0.05 is considered as statistically significant. **P value <0.01.
223

Downloaded from by guest on November 28, 2015

224
correlation with the duration of the eruption, explaining 11.4 per
cent (P = 0.006, R2 = 0.114) of the overall variance. Adding the vari-
able canine root apex-dental arch plane increased the explanation
rate to 17.4 per cent (P = 0.041, R2 = 0.174), and adding the extrac-
tion of the deciduous canine to the other two factors explained in
all 24.1 per cent (P = 0.026, R2 = 0.240) of the variance in the
duration of the eruption. No further predictors, i.e. location of the
canine cusp tip (canine cusp tip-dental arch plane, P = 0.094 and
canine cusp tip-midline, P = 0.529), age (P = 0.120), and angula-
tion (mesioangular angle, P = 0.226), were associated with time of
eruption.
Secondary outcomes
Predictors and cut-off points for successful outcome.
A logistic regression (Table 3) was done to disclose how possible
confounders might affect the correlation between the main outcome
(i.e. eruption versus non-eruption) and the radiographic variables,
age and the intervention (i.e. extraction versus non-extraction). Only
significant baseline variables from Table 1 that affected the main
outcome were included in the analysis. The probability of sponta-
neous eruption of the PDC in the EG was 18 times that in the CG,
indicating that the extraction of the deciduous canine was the vari-
able that most affected the main outcome, followed by the age of
the patient. Every unit of change in millimetre of the canine cusp
tip-midline and every unit of change in degrees of the mesioangular
angle also increased the odds ratio for successful outcome by 2.995
and 1.187, respectively.
ROC curve analyses were made in patients with erupted PDCs in
the CG and in patients in the EG who had surgical exposure of the
PDCs since the canine did not spontaneously erupt. Radiographic
variables that showed a correlation between the radiographic find-
ings and age and the main outcome achieved from the regression
analysis were analysed (Table 3).
ROC curve analysis of ‘easy cases of PDCs’ in the CG.
The area under the ROC curve for the variable canine cusp-midline
was 0.920 [95 per cent confidence interval (CI): 0.812–0.990], dem-
onstrating a very good discriminatory power and therefore a poten-
tial use as a diagnostic test in determining whether a PDC will erupt
spontaneously or not (Figure 2). An initial distance of the canine
cusp tip to the midline of 11 mm was crucial for spontaneous erup-
tion (sensitivity, 0.850; specificity, 0.739; P = 0.000) without prior
extraction of the deciduous canine.
The initial distance of the canine cusp tip-dental arch plane
(Figure 2) of 2.5 mm was essential for spontaneous eruption
(sensitivity, 0.950; specificity, 0.783; P = 0.000) with an AUC
of 0.803 (95 per cent CI: 0.789–0.998), indicating good validity
of the test.
Table 3. Logistic regression of the significant baseline means from Table 1 that affected the main outcome, i.e. eruption of the PDC.
Variables at T0 β df
Extraction versus non-extraction 2.833
Mesioangular angle (°) −0.872
Canine cusp tip-dental arch plane (mm) −0.504
Canine cusp tip-midline (mm) 1.097
Age, years −1.732
CI, confidence interval; df, degree of freedom; OR, odds ratio.
*Sig.: statistical significance; P < 0.05. **P < 0.01. ***P < 0.001.
European Journal of Orthodontics, 2015, Vol. 37, No. 2
The best compromise between sensitivity and specificity of the
mesioangular angle may be found at a cut-off point of 103 degrees.
Based on this cut-off point, the sensitivity was 0.800 and the speci-
ficity 0.739 (P = 0.003). The AUC of 0.770 (95 per cent CI: 0.789–
0.998) suggested good accuracy of the test.
The AUC for the age of the patient was 0.676 (95 per cent CI:
0.515–0.837), indicating only a fair validity of the test. For a success-
ful outcome, the cut-off point of the initial age of the patient was set at
11.5 years (sensitivity, 0.700; specificity, 0.608; P = 0.049) (Figure 2).
ROC curve analysis of ‘severe cases of PDCs’ in the EG.
An initial distance of the variable canine cusp to the midline of 6 mm
was essential for surgical exposure, despite extraction of the decidu-
ous canine (sensitivity, 0.917; specificity, 0.853; P = 0.000). The AUC
of 0.930 (95 per cent CI: 0.856–1.000) suggested very good validity
of the test (Figure 3). The best compromise for the variable canine
cusp tip-dental arch plane was found at the cut-off point of 5 mm
(sensitivity, 0.875; specificity, 0.818; P = 0.000) with an AUC of
0.801 (95 per cent CI: 0.979–0.987) (Figure 3).
The optimum cut-off point of the initial mesioangular angle was
determined to be 116 degrees in terms of being necessary for surgi-
cal exposure (sensitivity, 0.864; specificity, 0.750; P = 0.052) with
an AUC of 0.783 (95 per cent CI: 0.780–0.985), indicating good
validity of the test (Figure 3). The AUC for the age of the patient
was 0.619 (95 per cent CI: 0.556–0.882), i.e. suggesting fair valid-
ity of the test. With a cut-off point of the initial age of the patient
Downloaded from by guest on November 28, 2015
of 12.5 years, the sensitivity was 0.719 and the specificity 0.641
(P = 0.038) (Figure 3). Figure 4, exemplifies the above-mentioned
results with one clinical case.
Discussion
Part I of this prospective randomized clinical longitudinal trial con-
cluded that extraction of the deciduous canine in patients with PDC
will result in spontaneous eruption of the permanent canine in 69 per
cent on the extraction side compared to 39 per cent on the control
side (11). Since extraction of the deciduous canine does not always
lead to spontaneous correction of the eruption path, a reliable pre-
diction of the intervention could decrease the amount of unnecessary
interceptive extractions as some PDC could erupt spontaneously
without preceding extraction of the deciduous canine. In severely
malpositioned PDCs, a decision can be made to extract the decidu-
ous canine simultaneously with the surgical exposure of the PDC,
thereby avoiding double treatment and a delay in treatment, and also
minimizing the costs for bringing the impacted canine into occlusion.
Possible factors and cut-off points were therefore investigated in this
continuation of the study. The results showed that the permanent
canines erupted spontaneously significantly more often in younger
OR 95% CI Sig.*
1 18.877 4.846–29.999 0.0014***
1 1.187 1.017–1.386 0.0295*
1 2.237 0.923–5.181 0.0610
1 2.995 1.108–8.096 0.0306*
1 5.649 1.464–21.739 0.0120**
J. Naoumova et al. 225

Downloaded from by guest on November 28, 2015

Figure 2. ROC curve analysis for spontaneous eruption of the canine without prior interceptive extraction of the deciduous canine. The optimum cut-off points
(marked with a star) of the variables: canine cusp tip to midline (11 mm; sensitivity, 0.850; specificity, 0.739), canine cusp tip-dental arch plane (2.5 mm; sensitivity,
0.950; specificity, 0.783), mesioangular angle (103 degree; sensitivity, 0.800; specificity, 0.739), age of the patient (11.5 years of age; sensitivity, 0.700; specificity,
0.608). Sensitivity (true-positive rate) indicates the percentage of non-erupted PDCs that are correctly identified as non-erupted, and the specificity (false-positive
rate) indicates the percentage of spontaneously erupted PDCs that were correctly identified to spontaneously erupt. The diagonal line shows what could be
expected from simple guessing, i.e. pure chance. PDC, palatally displaced canine; ROC, receiver-operating characteristic.

patients with a smaller mesioangular angle, shorter distance of the
canine cusp tip to the dental arch plane and greater distance of the
canine cusp tip to the midline. Cut-off points of four radiographical
variables—canine cusp-midline, canine cusp tip-dental arch plane,
mesioangular angle, and age of the patient—were found for when
the canine will spontaneously emerge or not emerge. Furthermore,
canines closer to the occlusal plane, i.e. higher values of the vari-
able of vertical position, erupted significantly earlier. We can thereby
reject all the null hypotheses.
Previous studies (4–8, 10, 17) have assessed the effect of extract-
ing the deciduous canine mainly on the grounds of panoramic
radiographs using the original or a modified measurement method
developed by Ericson and Kurol (4). Although 3D radiographs with
a different measurement method were used in the current study,
some of our results are comparable to findings from earlier studies.
The alpha angle at baseline has been reported to be associated with
success in facilitating canine eruption (5, 17) which could be com-
pared to the mesioangular angle showing the same pattern in this
study. Smailienė et al. (15) reported an alpha angle of more than 20
degrees and Power and Short (5) an angle of more than 31 degrees as
the critical angle of inclination for a reduced chance of spontaneous
eruption. In the present study, a mesioangular angle of 116 degrees
was crucial for surgical exposure and 103 degrees for spontane-
ous eruption without previous extraction of the deciduous canine.
226 European Journal of Orthodontics, 2015, Vol. 37, No. 2

Downloaded from by guest on November 28, 2015

Figure 3. ROC curve analysis for PDCs needing surgical exposure despite of interceptive extraction of the deciduous canine. The optimum cut-off points (marked
with a star) of the variables: canine cusp tip to midline (6 mm; sensitivity, 0.917; specificity, 0.853), canine cusp tip-dental arch plane (5 mm; sensitivity, 0.875;
specificity, 0.818), mesioangular angle (116 degree; sensitivity, 0.864; specificity, 0.750), age of the patient (12.5 years of age; sensitivity, 0.719; specificity, 0.641).
Sensitivity (true-positive rate) indicates the percentage of PDCs that needs surgical exposure who are correctly identified as needing surgical exposure and the
specificity (false-positive rate) indicates the percentage of spontaneously erupted PDCs who were correctly identified as spontaneously erupted. The diagonal
line shows what could be expected from simple guessing i.e. pure chance. PDC, palatally displaced canine; ROC, receiver-operating characteristic.

The differences in angles between these two studies can be explained
by the different way the angles are measured. Nevertheless, there are
studies reporting that the pre-treatment alpha angle is not associated
with a successful outcome, which is contradictory to the results of
the present study (18, 23).
Sector measurement is another valuable prognostic indicator (4,
10, 15) and, according to Power and Short (5), the most important
single factor determining success. This was also seen on the ROC
curve analysis that was used in the present study, but expressed as
the canine cusp tip-midline. An initial distance of the canine cusp tip-
midline of 6 mm was determined as the cut-off for surgical exposure.
Therefore, in such cases, one might consider surgical exposure and
extraction of the deciduous canine at the same occasion. However,
if the distance is 11 mm, one could follow the canine eruption with
radiographs at 6-month intervals, as previously recommended, but
without extracting the deciduous canine.
Smailienė et al. (15) found that a vertical distance of less than
12 mm from the crown tip to the occlusal plane had the strongest
impact on spontaneous eruption. This conclusion is not in accord-
ance with the present study or with other earlier studies, which found
that the vertical distance of the PDC is not a prognostic indicator for
the outcome after deciduous canine removal (4–6).
In this study we had the possibility to assess the ‘depth’ of impac-
tion, i.e. the canine cusp tip-dental arch plane, since CBCT images
J. Naoumova et al. 227

Downloaded from by guest on November 28, 2015

Figure 4. Bilateral PDCs in a girl aged 12 years 7 months with more severe displacement on the right then the left side at baseline (T0). Extraction of the right
deciduous canine was done (indicated as X). On the control side, a worsening of the canine position was seen on the radiographic measurements at the
12 months control (T2), i.e. the mesioangular and sagittal angle had increased and the distance: canine cusp tip-midline had decreased while canine cusp tip-
dental arch plane had improved slightly. The canine on the extracted side had worsening of the sagittal angle, more or less no changes of the mesioangular
angle and slight improvement of the vertical position, canine cusp tip-dental arch plane and canine cusp tip to the midline. None of the permanent canines
erupted at the T2 and were surgically exposed. Retrospectively extraction of the deciduous canine on the left side would be more beneficial according to the
baseline cut-off point.

were taken. The cut-off point for spontaneous eruption without a
previous deciduous canine extraction was set to 2.5 mm while a dis-
tance of 5 mm was essential for surgical exposure. To our knowl-
edge, there are no studies in the literature that have evaluated the
changes in ‘depth’; hence we cannot compare our results.
Younger patients, i.e. 10–11 years, compared to 11- to 12-year-
olds had significantly more successful eruption of the canine than the
older ones, which is in accordance with earlier findings (4, 10, 24).
However, age at baseline did not have an effect on the time when the
PDCs erupted and the ROC curve showed that a cut-off point for
age has only fair validity as a factor for determining the outcome.
The only significant difference in baseline means for canines that
erupted between T1 and T2 and canines that erupted after T2 in the
EG was the vertical position.
PDCs closer to the occlusal plane erupted earlier in the EG than the
ones further away, indicating a faster eruption in the EG than in the CG
(Table 2). Furthermore, the multiple stepwise regression analysis revealed
that the vertical position, root apex-dental arch plane, and extraction of
the deciduous canine were the most important factors for the duration
of time it took for the canine to erupt. The only available research on the
duration of the canine eruption and possible predictors influencing it are
retrospective studies investigating the time it took to successfully treat
ectopic palatal canines with surgical exposure and orthodontic traction
(12, 18, 25–27). It is therefore not possible to compare our results with
previous studies. The finding that the vertical position of the canine is
associated with the duration of the eruption after surgical exposure, fol-
lowed by orthodontic traction, has been highlighted by Stewart et al.
(12) and Zuccati et al. (18) The alignment of canines positioned 14 mm
or more above the occlusal plane increased the treatment duration from
24 to 31 months (12). Furthermore, the more mesial the cusps of the
PDCs are before treatment and the older the patient is, the longer is the
orthodontic treatment of the impacted canines (18, 25–27).
228
Clinical implications
The average cost of treatment of PDC with surgical exposure fol-
lowed by orthodontic treatment in Sweden has been reported to
be € 3200 per case, with an estimation of a total annual cost of
€ 600 0000 (13). This very high cost for the patient or for the soci-
ety in countries where orthodontic treatment is free of charge for
children and adolescents can be avoided if interceptive extraction is
done in the right cases.
In severely displaced canines in older patients, as can be deter-
mined using the cut-off points, surgical exposure without previous
extraction of the deciduous canine can be considered and thereby
reduce the number of interventions. In addition, patients do not have
to wait for an unsure spontaneous eruption that might lead to a
more severely displaced palatal canine during the observation period
and that might cause resorption of the adjacent teeth. The sensitivity
and specificity will change if the cut-off points are changed, but it
would be reasonable to suspect that a spontaneous eruption will not
occur in cases positioned worse than the cut-off points. In contrast,
one might consider not extracting the deciduous canine if the PDC
is very slightly displaced, i.e. according to the cut-off points, but in
doing this one must be aware that it is necessary to keep control
over the eruption path and intervene with extraction if X-rays indi-
cate that the position of the canine becomes worse during eruption.
Cases ‘in between’ could clinically be interpreted such that intercep-
tive extraction is advantageous, in that faster spontaneous eruption
can be expected.
When some radiographic variables are above or below the cut-
off points for successful and unsuccessful outcomes, respectively,
looking at the variable with the highest validity is recommended, i.e.
the canine cusp tip-midline followed by the canine cusp tip-dental
arch plane and the mesioangular angle.
A comparison of the CBCT images with the available intraoral
radiographs and panoramic radiographs in these patients in order
to form a thumb of rule for when and in which cases interceptive
treatment will be successful are to be presented in an upcoming
study.
Limitations
Although the ROC curves have many advantages in terms of charac-
terizing the accuracy of a diagnostic test, they also have some limita-
tions. By changing the cut-off point, one can change the number of
true-positive and true-negative values, i.e. the sensitivity and speci-
ficity. Choosing a higher cut-off point would reduce the sensitivity
but increase the specificity. Decisions as to where to place the cut-off
points are subject to discussion and disagreement. Decisions should
be based on the costs, risks, and benefits associated with the propor-
tion of those diagnosed as true positive and false positive (28). In
the present study, we used cut-off points with strict values in order
to maximize the sensitivity and specificity. Getting a more average
specificity, sensitivity and ROC for the above-mentioned cut-off
points require pooling results of several studies examining the same
test. As shown in the clinical cases (Figures 2–4), some variables
can be above and others below the cut-off point for successful and
unsuccessful outcome. In these cases it is important to decide which
variables are the most important. According to the results, the canine
cusp tip-midline has the best discriminatory power for whether the
PDC will spontaneously erupt or not and would therefore be a more
important variable to look at when deciding to extract or not to
extract the deciduous canine. The canine cusp tip-dental arch plane
had the next best discriminatory power, followed by the mesioan-
gular angle.
European Journal of Orthodontics, 2015, Vol. 37, No. 2
Generalization
The primary results obtained from the present sample can be gener-
alized on a Caucasian population aged 10–13 years old. However,
the cut-off points may differ in other samples, depending on where
the operator chooses to place the cut-off, which limits the generaliza-
tion of the determined cut-off points in the present study.
Conclusions
• Extraction of the deciduous canine is the variable that affects the
spontaneous eruption of the canine most.
• Erupted PDCs showed an initial smaller mesioangular angle,
shorter distance of canine cusp tip-dental arch plane, and larger
distance of canine cusp tip-midline in younger patients than the
non-erupted group.
• PDCs positioned closer to the occlusal plane, with a shorter
distance of the canine root apex-dental arch plane and with a
combination of extraction of the deciduous canine, show faster
eruption.
• A spontaneous eruption in cases positioned worse than the cut-
off points will most probably not occur. On the other hand,
interceptive extraction might be unnecessary in cases positioned
better than the cut-off points. Cases ‘in between’ could be inter-
preted that interceptive extraction is beneficial.
• The canine cusp tip-midline had the best discriminatory power as
to whether the PDC will spontaneously erupt or not, followed by
Downloaded from by guest on November 28, 2015
the canine cusp tip-dental arch plane and the mesioangular angle.
Age is a fair predictor of the outcome.
Supplementary material
Supplementary material is available at European Journal of
Orthodontics online.
Funding
Local Research and Development Board for Gothenburg and Södra
Bohuslän; Health & Medical Care Committee of the Regional
Executive Board, Västra Götaland Region.
Acknowledgement
This work is dedicated to Prof. JK who suddenly and sadly passed away during
the work of this study.
References
1. Thilander, B. and Myrberg, N. (1973) The prevalence of malocclusion in
Swedish schoolchildren. Scandinavian Journal of Dental Research, 81,
12–21.
2. Parkin, N., Benson, P.E., Shah, A., Thind, B., Marshman, Z., Glenroy, G.
and Dyer, F. (2012) Extraction of primary (baby) teeth for unerupted pala-
tally displaced permanent canine teeth in children. The Cochrane Data-
base of Systematic Reviews, 12, CD004621.
3. Naoumova, J., Kurol, J. and Kjellberg, H. (2011) A systematic review of
the interceptive treatment of palatally displaced maxillary canines. Euro-
pean Journal of Orthodontics, 33, 143–149.
4. Ericson, S. and Kurol, J. (1988) Early treatment of palatally erupting max-
illary canines by extraction of the primary canines. European Journal of
Orthodontics, 10, 283–295.
5. Power, S.M. and Short, M.B. (1993) An investigation into the response of
palatally displaced canines to the removal of deciduous canines and an
 J. Naoumova et al.
assessment of factors contributing to favourable eruption. British Journal
of Orthodontics, 20, 215–223.
6. Baccetti, T., Leonardi, M. and Armi, P. (2008) A randomized clinical study
of two interceptive approaches to palatally displaced canines. European
Journal of Orthodontics, 30, 381–385.
7. Sigler, M.L., Baccetti, T. and McNamara, J., Jr. (2011) Effect of rapid max-
illary expansion and transpalatal arch treatment associated with decidu-
ous canine extraction on the eruption of palatally displaced canines: a
2-center prospective study. American Journal of Orthodontics and Dentof-
acial Orthopedics, 139, e235–e244.
8. Bonetti, G.A., Zanarini, M., Incerti, P.S., Marini, I. and Gatto, M.R. (2011)
Preventive treatment of ectopically erupting maxillary permanent canines
by extraction of deciduous canines and first molars: a randomized clini-
cal trial. American Journal of Orthodontics and Dentofacial Orthopedics,
139, 316–323.
9. Baccetti, T., Sigler, L.M. and McNamara, J.A., Jr. (2011) An RCT on treat-
ment of palatally displaced canines with RME and/or a transpalatal arch.
European Journal of Orthodontics, 33, 601–607.
10. Bazargani, F., Magnuson, A. and Lennartsson, B. (2014) Effect of intercep-
tive extraction of deciduous canine on palatally displaced maxillary canine: a
prospective randomized controlled study. The Angle Orthodontist, 84, 3–10.
11. Naoumova, J., Kurol, J., Kjellberg, H. (2014) Extraction of the decidu-
ous canine as an interceptive treatment in children with palatal displaced
canines - part I: shall we extract the deciduous canine or not? European
Journal of Orthodontics, 37, 209–218.
12. Stewart, J.A., Heo, G., Glover, K.E., Williamson, P.C., Lam, E.W. and
Major, P.W. (2001) Factors that relate to treatment duration for patients
with palatally impacted maxillary canines. American Journal of Ortho-
dontics and Dentofacial Orthopedics: Official Publication of the Ameri-
can Association of Orthodontists, Its Constituent Societies, and the Ameri-
can Board of Orthodontics, 119, 216–225.
13. Bazargani, F., Magnuson, A., Dolati, A. and Lennartsson, B. (2013) Pala-
tally displaced maxillary canines: factors influencing duration and cost of
treatment. European Journal of Orthodontics, 35, 310–316.
14. Sajnani, A.K. and King, N.M. (2012) Early prediction of maxillary canine
impaction from panoramic radiographs. American Journal of Orthodon-
tics and Dentofacial Orthopedics, 142, 45–51.
15. Smailienė, D., Sidlauskas, A., Lopatienė, K., Guzevičienė, V. and
Juodžbalys, G. (2011) Factors affecting self-eruption of displaced perma-
nent maxillary canines. Medicina (Kaunas), 47, 163–169.
View publication stats
229
16. Lindauer, S.J., Rubenstein, L.K., Hang, W.M., Andersen, W.C. and Isaac-
son, R.J. (1992) Canine impaction identified early with panoramic radio-
graphs. Journal of American Dental Association, 123, 91–97.
17. Warford, J.H., Jr, Grandhi, K.R. and Tira, E.D. (2003) Prediction of
maxillary canine impaction using sectors and angular measurement.
American Journal of Orthodontics and Dentofacial Orthopedics, 124,
651–655.
18. Zuccati, G., Ghobadlu, J., Nieri, M. and Clauser, C. (2006) Factors associ-
ated with the duration of forced eruption of impacted maxillary canines:
a retrospective study. American Journal of Orthodontics and Dentofacial
Orthopedics, 130, 349–356.
19. Clark, C.A. (1910) A method of ascertaining the relative position of
unerupted teeth by means of film radiographs. Proceedings of the Royal
Society of Medicine, 3, 87–90.
20. Ericson, S. and Kurol, J. (2000) Incisor root resorptions due to ectopic
maxillary canines imaged by computerized tomography: a comparative
study in extracted teeth. The Angle Orthodontist, 70, 276–283.
21. Naoumova, J., Kjellberg, H. and Palm, R. (2014) Cone-beam computed
tomography for assessment of palatal displaced canine position: a meth-
odological study. The Angle Orthodontist, 84, 459–466.
22. Hanley, J.A. and McNeil, B.J. (1982) The meaning and use of the area
under a receiver operating characteristic (ROC) curve. Radiology, 143,
29–36.
23. Baccetti, T., Crescini, A., Nieri, M., Rotundo, R. and Pini Prato, G.P.
(2007) Orthodontic treatment of impacted maxillary canines: an appraisal
of prognostic factors. Progress in Orthodontics, 8, 6–15.
24. Bruks, A. and Lennartsson, B. (1999) The palatally displaced maxillary
canine. A retrospective comparison between an interceptive and a correc-
tive treatment group. Swedish Dental Journal, 23, 149–161.
Downloaded from by guest on November 28, 2015
25. Olive, R.J. (2005) Factors influencing the non-surgical eruption of pala-
tally impacted canines. Australian Orthodontic Journal, 21, 95–101.
26. Fleming, P.S., Scott, P., Heidari, N. and Dibiase, A.T. (2009) Influence of
radiographic position of ectopic canines on the duration of orthodontic
treatment. The Angle Orthodontist, 79, 442–446.
27. Becker, A. and Chaushu, S. (2003) Success rate and duration of orthodon-
tic treatment for adult patients with palatally impacted maxillary canines.
American Journal of Orthodontics and Dentofacial Orthopedics, 124,
509–514.
28. Metz, C.E. (1978) Basic principles of ROC analysis. Seminars in Nuclear
Medicine, 8, 283–298.