Investigative Modalities of

Mandibular Invasion by Squamous

Cell Carcinoma
Caroline H. C. Acton, BDS, MDSc, FDSRCS,
FRACDS (OMS); Craig Layt, MB, BS, FRACS;
Ray Gwynne, MB, BS, FRACR; Robin Cooke,
MB, BS, MD, DCP, FRACPA, FRCPath; David
Seaton, MB, BS
 Objectives/Hypothesis To compare preoperative
investigations with histological findings in squamous
cell carcinoma (SCC) of the oral mucosa that abuts the
mandible.
Study Design: 2 parts prospective study
Methods: All patients presented to the Queensland
Radium Institute Head and Neck Clinic between 1993
and 1997 with a biopsy-proven SCC that abutted the
mandible
1st part: Investigated clinically, radiologically, and
histologically
2nd part: SPECT bone scans
 Results 67 patients (followed for 55 months)
assessed with orthopantomogram (OPG), computed
tomography (CT) scans, and, in the second part of the
study, SPECT bone scans
36 tumors histological evidence of bony invasion
36 by OPG, 27 confirmed histologically
22 by CT scans, 18 confirmed histologically
24 patients technetium 99m methylene
diphosphonate (MDP) bone scans with planar imaging
and SPECT, 14 histological analysis
3 patients with tumor had this confirmed histologically
 Conclusions:
1st part study: confirms our hypothesis that currently used
investigations, as well as clinical assessment, fail to predict
accurately invasion of the mandible by intraoral SCC
2nd part : suggests that SPECT scanning with high
quantification ratios is promising in the prediction of
tumor involvement
 INTRODUCTION
Cancers of the upper aerodigestive tract constitute 5% of
malignancies squamous cell carcinoma (SCC) being the
most common
The 5-year survival rate remains around 50%
Radical resection to ensure complete ablation of tumor
resulting functional and cosmetic deformity, combined with
poor survival recommendations : less radical bony
resections
Preoperative knowledge of tumor spread expedites
surgical planning and makes appropriate informed consent
possible
 Surgical margin around a SCC is 2 cm
difficult to achieve close to the mandible
lifting the periosteum at the time of surgery considerable value in
ascertaining whether any macroscopic bony involvement has occurred
Slootweg and Muller:
Alveolar invasion by SCC usually occurred from the superior surface of
the mandible; often intact periosteum between tumor and bone
2 modes of invasion: the less aggressive arrosive front (spare
periodontal ligaments as well as the inferior alveolar nerve), and the
diffuse infiltrating pattern
McGregor and MacDonald in 1988:
Specific entry points in the mandible
Presence or absence of teeth in the adjacent bone largely affected
the route of invasion
 Ord et al. : careful case selection would allow
a favorable oncologic outcome with
preservation of mandibular contour.
OPG, CT scans, SPECT may be accurate
when used in combination with meticulous
clinical assessment
The aim of this study was to define clinical
selection criteria to improve the accuracy of
these careful case selections
 MATERIALS AND METHODS
2 part prospective study patients attending the Head
and Neck Clinic at the Queensland Radium Institute
(Brisbane, Australia) between 1993 and 1997 (inclusive)
Patients had a diagnosis of SCC of the floor of mouth,
mandibular alveolus, or retromolar trigone considered
to invade the mandible, and had received no prior
treatment
1st part study :
Preoperative details of clinical factors, tumor factors, and
radiological diagnosis were collected
OPG examination and axial and coronal CT scanning with bone
windows results : evidence of tumor invasion into the
mandible
 2nd part :
Tomographic scintigraphy of the mandible 3
hours after the administration of technetium 99m
methylene diphosphonate (MDP), planar and
tomographic images of the jaw were acquired
Histological examination gold standard of
bone involvement by SCC
Meticulous sectioning of the soft tissue and
decalcified bone gave the clearest histological
assessment of tumor margins
 Radiological and clinical data were analyzed
Sensitivity : the number with a positive test result with
positive histological findings (i.e., true-positive result)
divided by the total number with positive histological
findings
Specificity : the number with a negative test result with
negative histological findings (i.e., true-negative result)
divided by the total number with negative histological
findings
Positive predictive value : the number with a true-positive
result divided by the number of positive results
Negative predictive value : the number with a true-
negative result divided by the number of negative results
 RESULTS
67 patients (47
male and 20
female) with a
mean age of 61.8
years; median
follow-up of 15.5
months
 56 tumors (83.6%)
clinically fixed to
the mandible
34 (60.7%) bone
invasion
2 tumors no
preoperative
evidence of bony
involvement but
histological analysis
showed invasive
carcinoma
 The OPG
revealed
evidence
of tumor
invasion
in 36
cases
(54%)
 27 cases (40.3%)
confirmed
histologically
OPG alone had a
sensitivity of 80%, a
specificity of 72%,
and positive and
negative predictive
values of 75%
 46 CT scan
22 (47.8%) bone invasion
18 (81.8%) histological
confirmation of the CT
findings
 CT scan alone sensitivity of 78%, a specificity of
83%, a positive predictive value of 82%, and a
negative predictive value of 79%
 46 patients (68.7%) both CT
scan and OPG
19 cases (41.3%) bone invasion, 17
(37%) histologically confirmed
19 cases (41.3%) no bone invasion,
and in 15 (32.6%) histologically
confirmed
8 cases (21.2%) two radiological
modalities conflicted
Combining OPG and CT a sensitivity of
81%, a specificity of 88%, a positive
predictive value of 90%, and a negative
predictive value of 79%
 40 (10 male and 4 female)
included in the second part of the
study
6 scans (42.9%) bone involvement
by tumor (Fig. 10), 3 (50%) had
confirmatory histology
8 cases (57.1%) scan did not
suggest bone involvement, 6 (75%)
of these interpretations were proved
correct on histological analysis
The sensitivity of jaw SPECT was
60%, specificity was 67%, positive
predictive value was 50%, and
negative predictive value was 75%
 DISCUSSION
This study define the accuracy of commonly
used methods for assessment of invasion of the
mandible by intraoral SCC
Comparison of the imaging procedures with
clinical examination and histological analysis,
singly and in combination, for sensitivity and
specificity constituted the first part of this study
The second part of the study assessed a newer
investigation that may contribute to our
preoperative diagnosis and planning
 Clinical assessment by an experienced team of
clinicians tends to result in overdiagnosis of invasion
of the mandible leading to a high sensitivity, low
specificity, and predictive values of 57% to 80%
Orthopantomogram (assessed by a single experienced
head and neck radiologist) had a lower sensitivity, was
more specific, and had better positive and similar
negative predictive values when compared with clinical
assessment alone This result must be viewed
objectively because a radiolucency on an OPG may
represent infection, dental disease, or extraction socket
 CT scanning slightly better specificity and positive
predictive value than OPG, but with similar sensitivity and
negative predictive value
Combining CT and OPG as investigative tools similar
sensitivity and a high specificity if the investigations agree
3 key findings in bone scintigraphy
First, in the small group of edentulous patients with available
histological findings the bone scan correctly predicted malignant
involvement
Second, a positive bone scan may have negative histological
result
Third, a normal bone scan in dentate patients does not exclude
disease
 Our study
clinical assessment (in particular, intraoperative
assessment) remains the most sensitive tool for
diagnosis of mandibular invasion
Disadvantage : low specificity
Thus, some mandibles are resected needlessly if
it is the sole modality to be used
Standard radiography and a combination of these
offer sensitivity around 80%
OPG and CT in combination will improve
investigation of these tumors
 The semiquantitative using SPECT to assess
mandibular invasion similar results in our
study, but more data are needed before any
conclusions may be drawn
The discussion of McGregor and MacDonald
about routes of entry into the mandible by SCC
warrants further investigation because we
found little relationship between the presence or
absence of teeth and bony invasion by SCC
 CONCLUSION
Currently there is no ideal method for
identifying which tumors invade the mandible
THANK YOU