Clin. Otolaryngol.

1986, 11, 123-129

Zygomatic fractures: classification and

complications
s. KRISTENSEN A N D K . TVETERAS
Department of Otolaryngology, St Joseph’s Hospital, Esbjerg, Denmark

Accepted for publication 14 October I985

KRISTENSEN
S. & TVETERAS
K. (1986) Clin. Otolaryngol. 11, 123-129
Zygomatic fractures: classification and complications
A retrospective study of zygomatic fractures is presented in order to analyse late
complications and to evaluate the different radiographic classifications. The study
comprises 109 patients with 11 1 zygomatic fractures. The aetiology was violence in
39% and traffic accidents in 28%. Associated fractures of the craniofacial skeleton
occurred in 42% of the patients. Seventy-two patients were available for the follow-
up study. Malar flattening was found in 16% of the patients operated on. Thirty-four
per cent of the patients had sensory disturbances, 6% had enophthalmos, and 1 %
had diplopia. Classifications of zygomatic fractures are reviewed. The fractures in the
current study were grouped in accordance with the classifications of Knight &
North” and Larsen & Thomsen.’ Neither of these classifications was found to be
useful in the peroperative evaluation of the postreductive fracture stability. The most
reliable method of evaluating this stability is the preoperative evaluation, but CT
classification systems may in the future demonstrate their value.
Keywords Zygomatic fractures radiographic classification

The zygomatic bone gives prominence to zygomatic fracture may be comminuted, or

the cheek and takes part in the formation
of the orbit and the maxillary sinus. Due to
its prominent position in the face the
zygoma is frequently subjected to fracture
and dislocation. Thus, zygomatic fractures
are, next to nasal fractures, the most
frequent fractures of the facial skeleton. 1,2
The fracture lines are most frequently
located between the zygoma and the
adjacent facial bones. However, the
separation rarely occurs within the lines of
suture and the designation fracture of the
zygomatico-maxillary or malar complex is
therefore often u ~ e d . ~The, ~ zygomatic
bone will frequently be displaced in the
fracture and this displacement may be
medial, lateral, posterior or inferior, and
the bone may be rotated around a vertical
or a longitudinal axis. Infrequently the
Correspondence: Serren Kristensen MD, Department of Otolaryngology, St Joseph’s Hospital,
DK-6700 Esbjerg, Denmark.
the zygomatic arch may alone be
fra~tured.~.~-’
Fractures of the zygoma are always
caused by direct violence and clinically
often accompanied by a considerable
degree of periorbital and subconjunctival
ecchymosis. Additionally, flattening of the
cheek, diplopia, sensory disturbances of the
infraorbital region, and palpable fracture
displacement of the infraorbital margin are
typical clinical finding^.^,^ The treatment
procedures, which have been the subject of
several investigations during the last
century, are, however, still controversial.
Thus open, closed, direct, and indirect
methods of reduction are d e ~ c r i b e d . The
~.~
fractures may be stable or unstable after
reduction. To predict this postreductive
stability several classifications of zygomatic

123
124 s. KRISTENSEN A N D K. TVETERAS
fractures. using both conventional radio-
graphic and, recently, CT techniques, have
been described in order to select a proper
treatment modality.5- l 1 Thus, Schjelderup
in 1950' classified the zygomatic fractures
into 5 types based on the location of the
hinging attachment to the facial skeleton.
Nysingh' stated that radiographic evidence
of separation at the zygomatico-frontal
suture indicated postreductive fracture
instability. In 1961 Knight & North"
published their classical work, in which
they tried to predict the postreductive
stability and made proposals for the proper
treatment of zygomatic fractures based on
systematic fracture analysis and classi-
fication. This classification has for several
years been the most applied and accepted
grouping of zygomatic f r a c t ~ r e s . ~3.14
.~.'
Rowe & Killey 1968' realized that dis-
placement of the zygoma might be a
consequence of axial rotation or of en bloc
displacement, Their fracture classification
contained 8 types of fracture and 7
subdivisions. Yanagisawa in 19731° modi-
fied the classification of Rowe & Killey'
and based his classification on the analysis
of the displaced fragment in 3 directions,
using 3 different conventional radiographs.
This classification contained 7 types of
fracture with 8 subdivisions, prediction of
postreductive stability. and proposals for
treatment of each group. Although, these
different classifications have led to a better
understanding of the nature of zygomatic
fractures, they have, from a practical point
of view, proved to be of little v a l ~ e . ~ . ' ~modality of zygomatic fractures in this
Based on the work of Schjelderup,'
Nysingh,' and Rowe & Killey,' a
simplified radiographic classification for
practical use with an easy preoperative
grouping into stable and unstable fractures
was therefore proposed by Larsen &
Thomsen in 1978.s In 1983 Fuji &
Yamashiro" proposed a new CT classifi-
cation system based on the displacement of
the zygoma in the antero-posterior
direction. This classification contains 4
groups of fracture with 3 subdivisions. and
confirms the possibility of establishing a
complete three-dimensional pattern of dis-
placement in malar complex fractures.
These c l a s ~ i f i c a t i o n s ~have
~ ' ~ not yet been
the subject of re-examination.
The purpose of this study has been to
evaluate the classifications of Knight &
North" and Larsen & T h ~ m s e n as , ~ well
as to analyse the results of treatment and
late complications in fractures of the
zygoma.
Material and methods
During the years 1978 to 1982, 109 patients
with 111 zygomatic fractures were
admitted to the ENT Department of St
Joseph's Hospital. Registered data of the
109 patients included: age, sex, alcohol
intoxication, and associated fractures of
the cranio-facial skeleton. Additionally, the
aetiology of the 111 fractures was
registered.
At the end of 1984, 78 patients with 80
zygomatic fractures had come for follow-
up. Thus, the observation period was from
2 to 7 years. As this presentation aims inter
alia to evaluate the results of treatment, 6
of these patients (fractures) were excluded
as they initially refused any treatment in
spite of fracture displacement. Con-
sequently the follow-up study numbers 72
patients with 74 zygomatic fractures.
Additionally, registered data of these 72
patients included: initial clinical findings,
treatment procedures, and clinical findings
at follow-up. The routine treatment
retrospective study has been closed
reduction using the Gillies p r o c e d ~ r e ' ~
while simultaneously palpating the
fractured bone in order to evaluate the
postreductive fracture stability. In cases of
instability the fractured zygoma was fixed
by transosseous wiring at the zygomatico-
frontal and/or zygomatico-maxillary suture
lines. For reducing and fixing depressed
orbital floor and comminuted zygomatic
fractures, antral packing with a urografin-
filled Foley-catheter, introduced through
an antrostomy beneath the inferior tur-
Zygomatic fractures 125

binate, was routinely applied, succeeded by Results

postoperative radiographic examination.
This procedure was carried out while The annual incidence of zygomatic
simultaneously observing the position of fractures in the 5-year period was 10.4 per
100000 inhabitants. Table 1 demonstrates
the orbital floor through a subperiorbital
the age and sex distribution of the 109
approach, and the free passive movement
patients. Ninety-two (84%) patients were
of the eye was finally secured by per-
male, 17 (16%) female, and the age varied
forming the ‘forced duction test’.16-18 If
between 14 and 83 years. Table 2 shows
herniation of the orbital contents into the
the causes of the fracture in this series.
maxillary sinus was suspected radiographic
Ninety-three per cent of the zygomatic
tomography of the orbital floor and antral
fractures caused by violence occurred in
sinoscopy under local anaesthesia were
men, among which 73% were under the
carried out preoperatively. In all cases the
influence of alcohol. The fractures occurred
surgical treatment was performed within 10
with equal frequency on each side. The
days of the fracture.
associated fractures of the craniofacial
The radiographs, which included the
skeleton are illustrated in Table 3, and
posterior-anterior, submento-vertical,
these fractures occurred in 46 (42%)
lateral, and posterior-anterior oblique
patients.
(Water’s) projections3 were in all cases
Table 4 demonstrates the initial clinical
‘blindly’ reviewed and the zygomatic
manifestations in the 72 patients (74
fractures grouped in accordance with the
classification of Knight & North12 and
Larsen & T h o m ~ e n .Knight
~ & North12 Table 1. Age and sex distribution of 109 patients with
111 zygomatic fractures
classified the fractures into 6 groups: (1)
undisplaced fractures (6%), (2) displaced Age groups Male Female Total
arch fractures (lo%), (3) unrotated body 10-19 10 2 12
20-29 21 1 28
fractures (33%), (4) medially rotated body 30-39 28 3 31
fractures (1 1%), (5) laterally rotated body 4049 10 3 13
fractures (22%), and (6) comminuted 50-59 I 5 12
60-69 8 1 9
fractures (18%). According to Knight & 70-79 2 1 3
North12 fractures in group 1 required no 80-89 0 1 1
reduction, fractures in groups 2 and 5 Total 92 (84%) 17 (16%) 109
never required postreductive fixation,
fractures in group 4 always required Table 2. Zygomatic fractures related to aetiology
fixation, and fractures in groups 3 and 6
required fixation in 40 and 70% of the n YO
cases respectively. Violence 43 39
Larsen & Thomsen5 divided the Traffic accident 30 28
Occupational 14 13
fractures into 3 groups: (1) fractures Fall 12 11
showing minimal or no displacement Sport 10 9
(l6%), (2) fractures showing great dis-
placement and disruption at the fronto-
Table 3. Associated craniofacial fractures in 46 of 109
zygomatic suture and, furthermore, patients with zygomatic fractures
comminuted fractures (12%), and (3) all
other kinds of zygomatic fractures (72%). n
Cranial fracture 9
According to Larsen & Thomsen,5 Nasal fracture 25
fractures in group A required no reduction, Maxillary fracture 23
fractures in group B reduction and Orbital floor fracture I
Mandibular fracture 6
fixation, and fractures in group C Total 70
reduction without fixation.
I26 s. KRISTENSEN AND K . T V E T E R ~ S

Table 4. Acute clinical manifestations of the 72 patients (74 zygomatic

fractures) who came for follow-up related to treatment procedures

NO Reduction
treatment Redurtion andjxation Total
Number of fractures 37 26 11 74
Malar flattening 0 26 11 37
Palpable displacement 13 23 10 46
Sensory disturbance 18 24 6 48
Trismus 0 0 1 1
Ocular displacement 0 0 3 3
Diplopia 3 8 5 16

Table 5. Clinical manifestations at follow-up of the 72 patients (74

zygomatic fractures) related to treatment procedures. ( ) = subjectively
evaluated

N O Reduction
treatment Reduction and jxation Total
Number of fractures 37 26 11 74
Malar flattening 1 (1) 3 (1) 3 (2) 7 (4)
Palpable displacement 13 I1 4 28
Sensory disturbance I 15 3 25
Trismus 0 0 0 0
Enophthalmos 2 2 0 4
Diplopia 0 0 1 1

fractures), who came for follow-up, related
to the treatment modality. Sensory
disturbances of the infraorbital nerve,
diplopia (in one or more of the 9 ocular
cardinal positions), palpable fracture dis-
placement of the infraorbital margin, and
trismus (less than 30 mm between the
incisor teeth) occurred most frequently in
fractures with dislocation.
Table 5 shows the clinical manifestations
at follow-up related to the treatment
procedure. Enophthalmos (2 mm or more)
was found in 4 patients without diplopia.
Seven patients appeared with flattening of
the cheek (evaluated objectively), and in 4
of these the flattening was also noticed by
the patient himself (evaluated subjectively).
Among these 4 patients, 1 had received no
treatment, and 1 had bilateral fractures.
The patient with diplopia on upward view
initially presented with fracture and
depression of the orbital floor.
Table 6 illustrates the 74 zygomatic
fractures, classified according to Knight &
North,12 related to the treatment
modalities and the occurrence of cheek
flattening at follow-up. As illustrated, 6
(60%) of the fractures classified in Group

Table 6 . The 74 zygomatic fractures classified in accordance

with Knight & North'' related to treatment procedures.
( ) = fractures associated with flattening of the cheek at
follow-up evaluated objectively

Fracture No Reduction Reduction Total

ii'pe treatment andjxation
1 36 (0) 0 (0) 0 (0) 36 (0)
2 0 (0) 1 (0) 0 (0) 1 (0)
3 0 (0) 6 (2) 0 (0) 6 (2)
4 1(1) 6 (0) 4 (1) 11 (2)
5 0 (0) IO(1) 3 (0) 13 (1)
6 0 (0) 3 (0) 4 (2) 7 (2)
Total 37 (1) 26 (3) 11 (3) 74 (7)
Zygomatic fractures
Table 7. The 74 zygomatic fractures classified in accordance
with Larsen & Thomsen5 related to treatment procedures.
( ) = fractures associated with flattening of the cheek
evaluated objectively at follow-up
4, postreductively were evaluated as stable
and consequently not fixed. Additionally, 3
(23%) of the fractures, classified in Group
5 , postreductively were evaluated as un-
stable and therefore fixed following
reduction.
Table 7 illustrates the 74 zygomatic
fractures grouped in accordance with the
classification of Larsen & Thornsen,'
related to the treatment procedures and the
occurrence of malar flattening at follow-up.
As demonstrated, 7 of the fractures,
classified in Group B, postreductively were
evaluated as stable and consequently not
fixed. Among these, only 1 appeared with
malar flattening at follow-up. Further, 5 of
the fractures, classified in Group C,
postreductively were evaluated as unstable
and therefore fixed following the reduction.
Discussion
It is remarkable that 40% of the zygomatic
fractures in this study were caused by
violence, and that the patients in 73% of
these cases were under the influence of
alcohol. Most authors state that traffic
accidents are the most frequent cause of
zygomatic fractures.'.' Violent behaviour
is often associated with drunkenness and
the medical literature demonstrates that
40-53% of the victims of violence are
under the influence of alcoho1.20,21
The frequency of associated fractures of
the craniofacial skeleton (42%) is consider-
ably higher in this study than in other
corresponding ~ t u d i e s . ~ , This
' illustrates
the importance of a careful clinical and
radiological examination of the whole
127
craniofacial skeleton in patients with zygo-
matic fractures. The associated fractures
occurred most frequently in zygomatic
fractures caused by traffic accidents (57%).
The material in the present study
contains a remarkably high number of
fractures with minimal or no displacement.
As no association between aetiology and
degree of fracture displacement could be
demonstrated, this result may be explained
by the fact that the patient material was
entirely unselected.
At follow-up, 6 (16%) of the patients,
who had undergone surgery, presented
with flattening of the cheek (Table 5), and
this is in accordance with similar
s t ~ d i e s . ~ ~ ' ~Three
~ ' ~ ~of' 'the patients with
malar flattening had a comminuted
fracture treated with reduction and
fixation. As also stated by other authors,
this fracture type is difficult to fix in the
correct anatomical position. 3,699
Reduction by the Gillies p r o c e d ~ r e ' ~
alone in this study proved to be successful,
evaluated objectively, in 88% and
subjectively in 96% of the cases. However,
1 of the untreated patients at follow-up
presented with malar flattening evaluated
subjectively. The displacement of this
fracture was initially not discovered,
probably due to haematoma and oedema
in the acute stage. This illustrates the
importance of re-examining patients with
apparently undisplaced zygomatic fractures
1 week after f r a ~ t u r e . ~ , * . ' ~
Fractures of the zygoma usually affect
the infraorbital foramen, which is the
weakest point of the malar complex.
Sensory disturbances of the infraorbital
128 s. KRISTENSEK AND K . TVETERAS
region are therefore frequently present in
the acute stage.23 In this study the
frequency of sensory disturbances in this
region was reduced from 65% initially to
34% at follow-up. This reduction was
independent of treatment and illustrates
the ability of spontaneous regeneration of
the infraorbital nerve, as also stated by
Nordgaard.'
Displacement of the eye in zygomatic
fractures in the acute stage is caused by
dislocation of the zygomatic bone with
displacement of the palpebral ligaments
and depression of the orbital floor. How-
ever. this displacement of the eye may be
obscured by intraorbital haematoma and
edema.".'^ Thus. in these cases a
depressed orbital floor is only diagnosed by
sinoscopy of the antrum and/or
conventional or computerised radiographic
tomography. Further, enophthalmos may
occur as a late complication of zygomatic
fractures due to post-traumatic fibrosis and
atrophy of the intraorbital fat.""* In this
investigation slight enophthalmos of 1 mm
is not included, as such small differences in
the eye postion may occur normally.16 The
high frequency of enophthalmos in other
studies may be explained by the inclusion
of enophthalmos of this degree.13*22
Diplopia in the initial stage of zygomatic
fractures is, in most cases, due to contusion
of 1 or more extraocular muscles or their
nerves (transitory neuroplegia) and intra-
orbital haematoma or oedema. In these
cases the double vision most often needs no
treatment, as the symptom spontaneously
disappears within 1 week.".23 However,
diplopia may also be a symptom of
incarceration of orbital tissue within a
fractured and displaced orbital floor, which
always needs surgery. The diagnosis of this
incarceration is, therefore, of essential
importance in fractures of the
zygoma. ',18*'3 Persistent diplopia after
zygomatic fractures is mainly caused by
intraorbital fibrosis and adhesion
formation. However, this symptom is
seldom encountered and occurs mainly for
lateral vision
In this series, antral packing with a
Foley-catheter in cases with orbital floor
fractures was employed in 3 patients
without complications. However, damage
to the ocular contents and blindness have
been reported from the manipulation of
sharp, bony fragments."
As illustrated in Table 6, the post-
reductive fracture stability in group 2, 4
and 5 in the classification of Knight &
North" in this study was not in
accordance with the fracture stability
evaluated peroperatively. This discrepancy
has also been demonstrated by other
authors.' 3.14 Consequently the classi-
fication is regarded as being of little clinical
value.5 9 1 3914
The postreductive fracture stability,
evaluated radiologically as described by
Larsen & Thomsen', in this study was not
in accordance with the postreductive
fracture stability evaluated peroperatively
(Table 7). Thus, among the patients
operated on, 12 (7 reduced and 5 reduced
and fixed) cases of discrepancy between the
fracture stability evaluated radiologically
and peroperatively were apparent. At
follow-up, only 1 of the 7 (simple) reduced,
but radiologically evaluated unstable
fractures was associated with malar
flattening. The remaining 6 were in this
way evaluated correctly peroperatively.
The 5 fractures, radiologically classified as
stable, were evaluated unstable per-
operatively and fixed, Thus, in the current
study we would not have been able to
predict the postreductive fracture stability
preoperatively by grouping the fractures
according to the Larsen & Thomsen5
classification in 11 (30%) cases out of the
37 operated on. Larsen and Thomsen5v6
justified the proposed classification by the
good correlation between the preoperative
grouping, peroperative findings, and
follow-up results, but did not
systematically review all radiographs.
Dingman & Natvig3 advised that most
fractures of the zygoma should be treated
with open reduction and direct wire
fixation. Dealing with zygomatic fractures
Zygomatic fractures
in this way will obviously result in
overtreatment of postreductive stable
fractures. In order to facilitate the choice
of treatment and to prevent overtreatment
the various detailed7-'0,'2 and the
simplified' preoperative classifications were
proposed. On the basis of the current
follow-up study and review of the literature
these preoperative classifications seem
redundant and unreliable. Conclusively,
the most reliable and safe method of
evaluating the postreductive stability in
zygomatic fractures is the peroperative
evaluation, as also stated by Nysingh.'
The application of computerized tomo-
graphy in the diagnosis and classification
of malar complex fractures has recently
been introduced.' 1924 Thus, CT techniques
offer several advantages over conventional
radiographic met hods, the most important
being the ability to demonstrate soft tissue
relations as well as those of the skeleton.
The value of CT classification systems in
the evaluation of postreductive stability in
zygomatic fractures may be demonstrated
in the future.
References
s., JBRGENSEN K. & BRAHE
1 ILLUM P., KRISTENSEN
PEDERSENC. (1983) Role of fixation in the
treatment of nasal fractures. Clin. Otolaryngol. 8,
191-195
2 STARKHAMMER H. & OLOFSSONJ. (1982) Facial
fractures: a review of 922 cases with special
reference to incidence and aetiology. Clin. Oto-
laryngol. 7 , 405409
3 DINGHAM R.O. & NATVIGP. (1964) Surgery of
Facial Fractures. W.B. Saunders, Philadelphia
4 POZATEK Z.W., KABANL.B. & GURALNICK W.C.
(1973) Fractures of the zygomatic complex: an
evaluation of surgical management with special
emphasis on the eyebrow approach. J . Oral. Surg.
31, 141-148
5 LARSENO.D. & THOMSEN M. (1978) Zygomatic
fractures I. A simplified classification for practical
use. Scand. J. Plast. Reconstr. Surg. 12, 55-58
129
6 LARSENO.D. & THOMSEN M. (1978) Zygomatic
fractures 11. A follow-up study of 137 patients.
Scand. J . Plast. Reconstr. Surg. 12, 59-63
7 SCHJELDERUP H. (1950) Fractures of the upper and
middle thirds of the facial skeleton. Acta. Chir.
Scand. 99,447452
8 NYSINGH J.G. (1960) Zygomatico-maxillary
fractures with report of 200 consecutive cases.
Arch. Chir. Neerl. 12, 157-168
9 ROWEN.L. & KILLEYH.C. (1968) Fractures of the
Facial Skeleton. E. & S. Livingstone, Edinburgh
10 YANAGISAWA E. (1973) Pitfalls in the management
of zygomatic fractures. Laryngoscope 83, 527-529
11 FUJIIN. & YAMASHIRO M. (1983) Classification of
malar complex fractures using computed tomo-
graphy. J . Oral. Surg. 41, 562-567
12 KNIGHTJ. & NORTHJ. (1961) The classification of
malar fractures: an analysis of displacement as a
guide to treatment. Br. J . Plast. Surg. 13, 325-339
13 BALLEV., CHRISTENSEN P.H., GREISEN0. &
JBRGENSEN P.S. (1982) Treatment of zygomatic
fractures: a follow-up study of 105 patients. Clin.
Otolaryngol. 7,411416
14 LUND K. (1971) Fractures of the zygoma: a
follow-up study on 62 patients. J . Oral. Surg. 29,
557-560
15 GILLIESH.D., KILNERT.P. & STONE D. (1927)
Fractures of the malar-zygomatic compound: with
a description of a new x-ray position. Br. J. Surg.
14, 65 1-659
16 NORN M.S. (1974) External Eye-Methods of
Examination. Scripter, Copenhagen
17 EMERYJ.M., NOORDENG.K. & SCHLERNITZAUER
D.A. (1971) Orbital floor fractures: long-term
follow-up of cases with and without surgical
repair. Trans. Am. Acad. Ophthalmol. Otolaryngol.
75, 802-812
18 RANKOW R.M. & MIGNOCNA F.V. (1975) Surgical
treatment of orbital floor fractures. Arch.
Otolaryngol. 101, 19-22
19 ADEKEYE E.O. (1980) Fractures of the zygomatic
complex in Nigerian patients. J . Oral. Surg. 38,
596599
20 REICHH. & JENSENJ. (1981) Accidents involving
violence. Acta. Orthop. Scand. 52, 435438
21 ANDERSEN E. & OLANOF.J.A. (1983) Injuries due
to violence treated in the casualty department,
Esbjerg Hospital. Ugeskr. LcPger. 145,48-50
22 ALTONEN M., KOHONEN A. & DICKHOFF K. (1976)
Treatment of zygomatic fractures: internal wiring
- antral packing - reposition without fixation. J .
Maxillofac. Surg. 4, 197-21 5
23 NORDCAARDJ.O. (1976) Persistent sensory
disturbances and diplopia following fractures of
the zygoma. Arch. Otolaryngol. 102, 80-82
24 FUJIIN. & YAMASHIRO M. (1981) Computed
tomography for the diagnosis of facial fractures. J .
Oral. Surg. 39, 735-738