J Oral Maxiliofac Surg

55:811-816, 1997

Occlusal Cant in the Frontal Plane as a

Reflection of Facial Asymmetry
BONNIE L. PADWA, DMD, MD,* MAUREEN 0. KAISER, DMDJ-
AND LEONARD B. KABAN, DMD, MD*

Purpose: The purpose of this study was to compare subjective evaluation

of occlusal canting in frontal photographs with objective radiographic measure-
ments to determine the threshold at which a cant is recognized as abnormal.
Patients and Methods: Standardized frontal photographs (at rest and smiling)
of two groups of orthognathic surgery patients were evaluated. Group 1 con-
sisted of patients with a documented occlusal cant (n = 21), and group 2
consisted of patients with no cant (n = 22). Four untrained and five trained
observers independently, and blind to the diagnoses, assessed patient photo-
graphs to judge the presence or absence of canting. These subjective results
were compared with objective measurements of the angle of the occlusal plane
to the true horizontal on each patient’s posteroanterior (PA) cephalogram.
Results: The mean occlusal cant was 5.0” f 1.6” for group 1 and 1.4” I 0.9”
for group 2. The difference between groups was statistically significant (P <
.Ol). Cants greater than 4” were detected clinically with greater than 90%
frequency by both untrained and trained observers.
Conclusions: The results of this study indicate that 4” is the threshold for
recognition of an occlusal cant by 90% of observers. This information has
significant implications for three-dimensional planning and outcome assess-
ment in orthognathic and craniofacial surgery.

It has been reported that all patients have some de- vestigators found subclinical, but measurable, cranio-
gree of craniofacial asymmetry, including those who facial asymmetry in all subjects. Furthermore, their
are perceived as normal.’ Shah and Joshi used mea- data suggested that the degree of asymmetry in facial
surements on posteroanterior (PA) cephalograms to structures increased with greater distance from the cra-
show statistically significant levels of asymmetry in nium. Structures of the lower face were more asymmet-
total facial structure and in the maxillary area in “nor- ric than those of the upper face. In another study, Ferra-
mal” patients. Peck et aI3 also used PA cephalograms rio et al4 showed variable degrees of soft tissue facial
to analyze craniofacial asymmetry in a group of sub- asymmetry in healthy Caucasian dental students with
jects who they thought had “esthetically pleasing normal dentitions, bilateral angle Class I first perma-
faces.” The group included professional models, nent molar relationships, and no history of craniofacial
beauty contest winners, and performing stars. The in- trauma, surgery, or previous orthodontic treatment.
Thus, it is widely recognized that asymmetry is often
present in the craniofacial complex.
* Assistant Professor, Department of Oral & Maxillofacial Sur-
gery, Harvard School of Dental Medicine, Children’s Hospital, Bos-
Despite this fact, asymmetric features are not always
ton, MA. easily detected because soft tissues may compensate
7 Dental Student, Harvard School of Dental Medicine, Boston, for underlying skeletal imbalances.355*6 In addition, in-
MA.
,’ W. C. Guralnick Professor and Chairman, Department of Oral &
dividuals may mask facial asymmetry by their posture.
Maxillofacial Surgery, Harvard School of Dental Medicine, Massa- Tilting of the head slightly may “correct” a canted
chusetts General Hospital, Boston, MA. occlusal plane.5 It is important to remember that an
Address correspondence and reprint requests to Dr Padwa: Chil- assessment of craniofacial and dental asymmetry
dren’s Hospital, 300 Longwood Ave, Boston, MA 02115.
should be part of the clinical evaluation of patients
0 1997 American Association of Oral and Maxillofacial Surgeons with dentofacial deformity.5
0278-2391/97/5508-0004$3.00/O In this study, an occlusal cant is defined as a diver-

811
812
gence of the occlusal plane from the transverse (hori-
zontal) axis. On physical examination it becomes most
apparent when the patient smiles. At rest, however, the
presence of an elevated labial commissure or alar base
on one side is often an indication of vertical skeletal
asymmetry. This should be documented during routine
evaluation of patients for orthodontic or orthognathic
surgical treatment.
To measure occlusal canting, a wooden tongue de-
pressor can be placed across the right and left posterior
teeth, and the parallelism or the angle of the tongue
depressor to the interpupillary plane can be docu-
mented. Alternatively, the vertical distance between
the maxillary canines and the medial canthi of the eyes
can be measured.’ These methods require assessment
of the patient’s eyes to ensure that discrepancies be-
tween right and left sides are not related to asymmetry
of the orbits or globes.
Analysis of the PA cephalogram also can be used
to determine occlusal cant. A line is drawn connecting
the occlusal surfaces of the left and right maxillary
first molars. The angle of this plane relative to the
transverse axis of the skull, that is, the angle of occlusal
cant, is measured. This documents the skeletal asym-
metry without the influence of the overlying soft tissue.
Standardized PA cephalometric analyses do not in-
clude evaluation of the relationship of the occlusal
plane to the horizontal. This is an important deficiency,
because leveling the occlusal plane, when necessary,
should be a goal of surgical and orthodontic therapy.
It has been suggested that a level occlusal plane is
a prerequisite for success in all orthognathic surgical
procedures and that failure to level it during surgery to
correct dentofacial deformities may have a detrimental
effect on masticatory function.7 In addition, the pres-
ence of an occlusal cant may be unmasked in some
patients by correction of more severe deformities in
the mandible. For this reason, bimaxillary surgery may
be required to achieve a satisfactory result even when
other maxillary parameters are within normal limits.
Patients with hemifacial microsomia (HFM) typi-
cally exhibit significant occlusal canting. The hypo-
plastic mandible impedes normal vertical growth of
the maxilla and midface, and progressive asymmetric
growth leads to increased distortion of the mandible
and the midface.’ The maxilla is shortened, with the
occlusal plane canting upward on the affected side as
much as 10” or more in many patients.
It has been shown that normal subjects with healthy
dentitions have varying degrees of dentoalveolar asym-
metry. Ferrario et al5 used digitized photographic anal-
ysis to determine angulation of the occlusal plane in
healthy Italian dental students of both sexes. They
found mean angulations between 2.15” and 2.90” in
these normal patients, with no statistically significant
differences between males and females. However, they
OCCLUSAL CANT IN THE FRONTAL PLANE
did not assess whether these levels of cant were clini-
cally noticeable. In 1968, Heimansohn’ proposed a the-
ory that in all healthy patients there is a normal func-
tional tilt of the transverse occlusal plane so that it is
not parallel to the horizontal skull axis.’ He measured
his own cant at 4”, despite having a normal appearance.
He recommended that occlusal canting should be re-
corded before major restorative treatment to preserve
the patient’s natural occlusal tilt and to prevent postop-
erative dysfunction. Thus, the evaluation of occlusal
canting may be important not only in patients with
obvious craniofacial asymmetry, but also in normal
patients who require extensive dental treatment.
To our knowledge, there have been no studies to
quantify the minimal degree of occlusal canting that
is clinically detectable. We hypothesized that there is
a threshold angle of the occlusal plane to the horizontal
above which a cant will be predictably noticed on phys-
ical examination. The purpose of this study is to deter-
mine this threshold for trained and untrained observers
by comparing subjective evaluation of standardized
photographs with measurements on a PA cephalogram.
Patients and Methods
The subjects for this study were obtained from a
group of 188 patients evaluated for orthognathic sur-
gery at Massachusetts General, Children’s, and Brig-
ham & Women’s Hospitals between January 1985 and
February 1996. To be included in the study, a patient’s
chart had to contain a resting and smiling frontal facial
photograph in natural head position (in slide format),
and a corresponding PA cephalogram in occlusion. The
PA cephalogram was taken in natural head position
with a standard head holder on the same day as the
photographs or within 6 months provided that no sur-
gery or orthodontic treatment was performed on the
patient during that period. Patients who met the inclu-
sion criteria, but whose smiling photographs did not
show their teeth adequately, were excluded.
As part of the routine orthognathic workup, each
patient was examined by an attending oral and maxillo-
facial surgeon who documented occlusal canting with
a wooden tongue depressor placed across both the ante-
rior (canine region) and posterior (premolar/molar)
dentition. In this way, any discrepancy between these
regions, indicating a step or irregularity in the dental
arch, was demonstrated. Only patients who had a corre-
sponding cant in the canine and molar regions were
used in this study. The sample of 43 suitable patients
was divided into two groups. Group 1 (Fig 1) were
patients who were found to have an occlusal cant dur-
ing their evaluation (n = 21) and group 2 (Fig 2) were
patients with no occlusal cant (n = 22). Each patient’s
date of birth, sex, and date of frontal photographs and
PA cephalogram were recorded.
PADWA, KAISER, AND KABAN
FIGURE 1. Frontal photographs of a group
Frontal resting and smiling photographs of the 43
study patients were projected on a dual-projection
screen in random order. The slides were evaluated in-
dependently by four medical residents (untrained ob-
servers) and five oral and maxillofacial surgeons (4
attending surgeons and one resident). The observers
were instructed to examine each pair of slides while
answering a questionnaire. The questions included: 1)
Do you notice anything abnormal about this patient’s
face (Yes/No)? 2) Do you notice any of the following
(Yes/No): facial asymmetry, crooked smile, tilting of
the base of the nose (one side higher than the other),
tilting of the corners of the mouth, tilting of the plane
of the teeth, deviation of the chin point?
The occlusal cant was measured on each patient’s
PA cephalogram (natural head position) using the fol-
lowing method: The orbital rims, crista galli, nasal
septum, piriform apertures, mandibular angles, and
maxillary molars in occlusion were traced. A vertical
reference line was constructed from the top of the crista
galli through a point on the superior nasal septum at the
level of the equator of the globe and ethmoid sinuses.A
horizontal line was drawn perpendicular to the vertical
line and tangent to the supraorbital rims. In patients
813
1 patient in natural head position: A, resting and B, smiling.
with orbital dystopia, the horizontal line was drawn
tangent to the normal supraorbital rim. The occlusal
plane was drawn asa tangent to the cuspsof the maxil-
lary first molars. The angle of the maxillary occlusal
plane was measuredin relation to the horizontal line
(Fig 3). When maxillary first molars were not present
or not in occlusion, the best available pair of posterior
maxillary teeth in occlusion was used. All PA cephalo-
grams were traced and measuredby two authors (B.P.
M.K.). The occlusal cant measurementswere then cor-
related with the subjective photographic evaluations.
Statistical analysis of the data was performed using
an unpaired Student’s t-test for values in which a mean
could be determined. Subjective data and other values
where no mean could be obtained were analyzed using
P values basedon the chi-squared test of independence.
P values less than .05 were considered statistically
significant.
Results
There were 13 female (62%) and 8 male (38%) pa-
tients who had an occlusal cant recognized on physical
examination (group 1). The mean age of the group was
814
FIGURE 2. Frontal photographs of a group
18.1 +- 10.3 years, with a range of 6.1 to 36.8 years.
There were 11 female (50%) and 11 male patients
(50%) without a cant (group 2). The mean age in group
2 was 29.1 k 7.3 years (range, 17.2 to 44.6 years).
The mean cant measured on the PA cephalogram for
the total patient sample was 3.1” t 2.2” (range, 0” to
8.0”). The average occlusal cant measured on the PA
cephalogram for group 1 was 5.0” 1. 1.6” (range, 3.0
to 8.0”) and for group 2 was 1.4” 5 0.9” (range, 0” to
2.5”). The difference between the groups was statisti-
cally significant (P < .Ol).
The frequency of detection of an occlusal cant (un-
trained and trained observers) is presented in Figure
4. Detection rates were calculated by dividing the num-
ber of positive responses (positive response = ob-
served cant) for untrained or trained observers by the
total number of possible answers for the respective
observer group. Untrained observers detected canting
in 55% of observations in the total patient sample,
whereas trained observers detected canting in 58% of
OCCLUSAL CANT IN THE FRONTAL PLANE
2 patient in natural head position: A, resting and B, smiling.
observations (P = NS). Untrained observers noted
canting in 82% of observations in group 1 patients,
and 30% of observations in group 2 patients. Trained
observers noted a cant in 90% of observations in group
1 and 27% in group 2. The detection of occlusal cant-
ing in group 1 patients was significantly greater than
in group 2 patients for both trained and untrained ob-
servers (P < .05).
The average cant detected by all untrained observers
in the total patient sample was 4.1”, whereas the aver-
age cant detected by all trained observers was 4.2”.
The difference between untrained and trained observ-
ers was not statistically significant.
The threshold for clinical appreciation of an occlusal
cant was estimated by plotting the detection rates for
each degree of occlusal cant (Fig 5). The detection
rates were calculated for two cant categories: 0” to <4”
(n = 29) and 4” and above (n = 14) (Fig 6). Untrained
observers noticed occlusal cants greater than 4” with
a 90% frequency, and trained observers detected these
PADWA, KAISER, AND KABAN 815

Group 1 Group 2 All

patients

FIGURE 4. Frequency of detection of an occlusal cant for patients

in groups 1 and 2 for untrained and trained observers. *:k = signifi-
cant difference between group 1 and 2 patients (P < .05).

PA cephalograms have been used extensively in

studies of craniofacial asymmetry to show mild de-
grees of asymmetry in normal, healthy subjects.2-4~‘0-12
Asymmetric features are easily noticed in patients with
certain craniofacial deformities, such as hemifacial mi-
crosomia but may be undetectable by gross observation
FIGURE 3. Tracing of a PA cephalogram in natural head position in other subjects.’
for a patient in group 1. The following landmarks were identified The maxilla, mandible, and dentoalveolar processes
and traced: supraorbital rims (ST), crista galli (cg), nasal septum
have been shown to exhibit a greater degree of asym-
(IX), piriform apertures (pa), mandibular angles (ma), and maxillary
molars (mm). A vertical reference line was drawn from the top of metry than other facial regions.3.5 Asymmetric growth
cg through a point on the superior ns at the level of the equator of may result in occlusal canting in the frontal or coronal
the globe and ethmoid sinuses (dotted line). A horizontal line was plane. Occlusal canting has been observed in healthy
drawn perpendicular to the vertical and tangent to sr. The occlusal subjects with normal masticatory function.‘,” Some
plane was drawn as a tangent to mm cusps. Occlusal cant was
determined by measuring the angle of the maxillary plane in relation
to the horizontal line. This patient had an occlusal cant of 0”.

cants with a frequency of 98%. The difference in detec-

tion rates between untrained and trained observers for
each category were not statistically significant. How-
ever, the differences in detection rates between the two
categories of cant were statistically significant for both
observer groups (P < .Ol).

Discussion

Cephalometric radiography and facial photographs

are well-established clinical and research tools for as- ol 1 I I 1

sessment of craniofacial morphology and deformities.’ 0” I” 2” 3” 4” 5’ 6’ 7’ 8”

They provide points and landmarks for measurement
Degteeofocclusalcant
and complete analytical evaluation.’ They are both
convenient and inexpensive when compared with other FIGURE 5. Frequency of detection by degree of cant for untrained
state-of-the-art methods for three-dimensional evalua- and trained observers. Each point in the plot represents the detection
tion such as computed tomography (CT) scanning. rate for each degree of cant.
816 OCCLUSAL CANT IN THE FRONTAL PLANE

occlusal cants are detected with greater than 90% fre-

100 quency and 3” is the threshold for greater than 50%
90 frequency.
The preceding findings have significant implications
80 for three-dimensional planning and outcome assess-
70 ment in orthognathic and craniofacial surgery. For ex-
60 ample, in a patient with significant craniofacial defor-
mity, an occlusal cant that is not recognized
50 q Untrained
preoperatively may become apparent after ortho-
40 II Trained
gnathic surgery, when the more obvious abnormalities
30 have been corrected. Preoperative measurement of the
cant in all such patients will ensure that corrective
20 surgery is properly planned.
10 Occlusal cants within the 0” to 3” range have been
0 observed in normal, healthy patients5 Thus, normal
< 4 Degrees > 4 Degrees
masticatory function is possible within this range. The
Degree of Occlusal Cant cant will not be noticeable, and occlusal canting of this
magnitude probably does not have detrimental effects
FIGURE 6. Frequency of detection for occlusal cants less than 4”
and for those greater than 4”. ** = significant difference between on postoperative outcome. Heimansohn’ suggested that
group 1 and 2 patients (P < .Ol). normal individuals have a natural tilt to the occlusal
plane, and that alteration of this tilt through restoration
of the dentition and placement of prostheses may con-
tribute to the development of temporomandibular joint
individuals have clinically discernible levels of occlu-
dysfunction.” Therefore, awareness of the presence
sal cant, but most normal patients have cants that are of even mild levels of occlusal cant is important in
too small to detect by simple clinical observation. In restorative dentistry as well as in orthodontics and sur-
this study, the mean occlusal cant measured radio-
F-Y.
graphically for those patients judged to be symmetric
by clinical examination (group 2) was 1.4” -t 0.9”. This
is consistent with data in the literature suggesting that References
mild degrees of occlusal canting exist in clinically
symmetrical patients.5’10
The ability to detect an occlusal cant was lower for 1. Erickson KL, Bell WH, Goldsmith DH: Analytical model sur-
untrained observers than for trained observers in group gery, in Bell WH (ed): Modern Practice of Orthognathic and
Reconstructive Surgery. Philadelphia, PA, Saunders, 1992, p
1 patients (82% vs 90%, respectively), although not 156
significantly. Untrained and trained observers also had 2. Shah SM, Joshi MR: An assessment of asymmetry in the normal
similar detection rates for cants in group 2 subjects craniofacial complex. Angle Orthod 48:141, 1978
3. Peck S, Peck L, Kataja M: Skeletal asymmetry in esthetically
(30% for untrained, 27% for trained; P = NS). These pleasing faces. Angle Orthod 61:43, 1991
findings suggest that differences in detection of occlu- 4. Ferrario VF, Sforza C, Poggio CE, et al: Distance from symme-
sal canting by clinical examination are dependent on try: A three-dimensional evaluation of facial asymmetry. J
Oral Maxillofac Surg 52:1126, 1994
the degree of canting and not necessarily the level of 5. Ferrario VF, Sforza C, Miani A, et al: Craniofacial morphometry
expertise of the observers. The untrained observers that by photographic evaluations. Am J Orthod Dentofac Or-
thoped 103:327, 1993
participated in the study were members of the medical
6. Yogosawa F: Predicting soft tissue profile changes concurrent
profession. Although they were not involved in any with orthodontic treatment. Angle Orthod 60: 199, 1990
area of dentistry, oral and maxillofacial surgery, or 7. Polley JW, Figueroa A, Cohen M: New ideas and innovations:
Leveling of the occlusal plane during model surgery. J Cra-
plastic surgery, their answers might have been biased
niofac Surg 4:266, 1993
by an increased capacity, relative to the general public, 8. Kaban LB, Moses MH, Mulliken JB: Surgical correction of
to notice abnormalities in patients. hemifacial microsomia in the growing child. Plast Reconstr
Surg 82:9, 1988
There are no published reports on the minimal level 9. Heimansohn HC: A new occlusal lateral plane. Dental Digest
of occlusal cant that is predictably detected by clinical 74:345, 1968
examination. In this study, the detection rate for both 10. El Mangomy NH, Shaheen S, Mostafa YA: Landmark identifi-
cation in computerized posteroanterior cephalometrics. Am J
untrained and trained observers surpassed 50% at 3” Orthod Dentofac Orthoped 91:57, 1987
of occlusal canting (Fig 5). Cants greater than 4” were 11. Ferrario VF, Sforza C, Poggio CE, et al: Facial three-dimen-
detected 90% of the time by untrained observers and sional morphometry. Am J Orthod Dentofac Orthop 109:86,
1996
98% of the time by trained observers. Based on these 12. Vig PS, Hewitt AB: Asymmetry of the human facial skeleton.
data, we conclude that 4” is the threshold at which Angle Orthod 2:45, 1975