Published online before print August 18, 2005, 10.1148/radiol.2371041228
(Radiology 2005;237:281-287.)
© RSNA, 2005
Bone Invasion in Patients with Oral Cavity Cancer: Comparison of Conventional CT with PET/CT and SPECT/CT1
Gerhard W. Goerres, MD,
Daniel T. Schmid, MD,
Bernhard Schuknecht, MD and
Gerold K. Eyrich, MD, DMD
1 From the Department of Radiology (G.W.G.), Division of Nuclear Medicine (G.W.G., D.T.S.), Institute of Neuroradiology (B.S.), and Department of Cranio-Maxillo-Facial Surgery (G.K.E.), University Hospital Zurich, Raemistr 100, CH-8091 Zurich, Switzerland; and Radiology Institute, Kantonsspital Winterthur, Switzerland (D.T.S.). Received July 14, 2004; revision requested September 22; revision received October 20; accepted November 26.
Address correspondence to G.W.G. (e-mail: gerhard.goerres{at}usz.ch).
 |
ABSTRACT
|
|---|
PURPOSE: To prospectively compare the accuracy of helical contrast material–enhanced computed tomography (CT) with that of CT and positron emission tomography (PET) combined and CT and single photon emission CT (SPECT) combined in the detection of bone invasion in patients scheduled to undergo surgery for clinically suspected oral cavity carcinoma with possible bone invasion, with surgical results as the reference standard.
MATERIALS AND METHODS: This study had local ethical committee approval, and all patients gave written informed consent. Thirty-four consecutive patients (17 men, 17 women; mean age, 64.2 years; age range, 46.0–84.6 years) who were clinically suspected of having bone invasion from oral cavity carcinoma prospectively underwent helical contrast-enhanced CT, coregistered PET/CT, and coregistered SPECT/CT. Two radiologists assessed the contrast-enhanced CT images and two nuclear medicine physicians separately assessed the PET/CT and SPECT/CT images in consensus and without knowledge of the results of other imaging tests. The presence of bone involvement as suggested with an imaging modality was compared with histologic findings in the surgical specimen.
RESULTS: With histologic findings as the standard of reference, the accuracy of SPECT/CT (88% [30 of 34 patients]) was lower than that of PET/CT and contrast-enhanced CT (94% [32 of 34 patients] and 97% [33 of 34 patients], respectively). Sensitivity was highest with PET/CT (100% [12 of 12 patients]), and specificity was highest with contrast-enhanced CT (100% [22 of 22 patients]). Fluorine 18 fluorodeoxyglucose (FDG) uptake seen on two sides of the same cortical bone was not a helpful imaging pattern for better identifying bone invasion in patients without evident cortical erosion on CT scans.
CONCLUSION: The assessment of cortical erosion with contrast-enhanced CT and the CT information from PET/CT are the most reliable methods for detecting bone invasion in patients with oral cavity carcinoma. FDG uptake seen on PET/CT images does not improve identification of bone infiltration.
© RSNA, 2005
|
INTRODUCTION
|
|---|
The assessment of invasion of the mandibular or maxillary bone is important in patients with carcinomas of the oral cavity because the surgical procedure is influenced by the presence and extent of bone involvement. Early research revealed that there is often no evidence of tumor invasion in resected bone in patients undergoing surgery for an oral cavity carcinoma (1–3). Therefore, a preoperative evaluation to reliably confirm or exclude tumor invasion into the bone is helpful for planning rim resection of the alveolar margin or partial- or full-thickness segmental mandibulectomy or maxillary bone resection.
Results of previous studies suggested that bone scintigraphy with single photon emission computed tomography (SPECT), orthopantomography, or a combination of different methods had high diagnostic power (3–8). In addition, computed tomography (CT) has been shown to be highly accurate in the identification of bone invasion in patients with carcinomas of the oral cavity (2,5,9). There are false-positive findings, however, and in a relatively recent work (10) it was concluded that orthopantomography and CT fail to enable an accurate prediction of invasion of the mandibula by intraoral squamous cell carcinomas. Positron emission tomography (PET) with use of fluorine 18 fluorodeoxyglucose (FDG) has become a routine diagnostic tool for staging and restaging disease in patients with squamous cell carcinoma of the oral cavity and is also useful for staging lymph nodes and identifying distant metastases (11,12). Because of the lack of anatomic information, however, it is not possible to reliably delineate a primary tumor within the oral cavity and to verify bone invasion on the basis of FDG uptake alone. Therefore, the coregistration of structural (CT) and functional (PET) imaging information could help improve the identification of bone invasion.
The aim of this study was to prospectively compare the accuracy of helical contrast material–enhanced CT alone with that of coregistered PET/CT and coregistered SPECT/CT for detecting bone invasion in patients scheduled to undergo surgery because oral cavity carcinoma with possible bone invasion was suspected on the basis of clinical evaluation, with surgical results as the reference standard.
|
MATERIALS AND METHODS
|
|---|
Patients
Between July 2002 and March 2004, 34 consecutive patients (17 men, 17 women; mean age, 64.2 years; age range, 46.0–84.6 years) who were clinically suspected of having oral cavity carcinoma and possible invasion of the mandibular or maxillary bone were enrolled in our prospective study. All patients were scheduled to undergo surgery for oral cavity carcinoma, contrast-enhanced CT and SPECT/CT of the head and neck, and skeletal scintigraphy and PET/CT performed from the head to the pelvic floor or legs. A summary of patient characteristics is given in Table 1.
In our hospital, bone scanning with SPECT/CT of the head, contrast-enhanced CT, and PET/CT are always performed as part of the routine clinical work-up in such patients to stage the tumor before surgery and as part of the restaging examination when recurrence is suspected. The delay between the PET/CT and SPECT/CT examinations was 0–10 days (mean, 2.7 days ± 2.4 [standard deviation]; median, 2 days), and the delay between contrast-enhanced CT and PET/CT was 0–24 days (mean, 5.0 days ± 5.2; median, 4 days). The study was conducted in accordance with regulations set forth by the institutional review board, and all patients gave informed written consent to participate in this study.
Imaging
Conventional CT was performed in all patients with a helical CT scanner (Somatom VolumeZoom; Siemens, Erlangen, Germany). CT scans were obtained from the base of the skull to the clavicles with the patient in a supine position by using the following parameters: 4 x 1.0-mm collimation, table feed of 2.5 mm per rotation, rotation time of 0.5 second, 300 mAs effective, and 120 kV. Nonionic contrast material (Optiray 300; Guerbet, Zurich, Switzerland) was injected intravenously as an 80-mL bolus at a rate of 2 mL/sec. Image acquisition started at the end of injection to ensure adequate enhancement of vessels and soft tissues. The image data were reconstructed with a section thickness of 1.25 mm, in increments of 0.7 mm, by using a high-spatial-resolution algorithm. Images were displayed as 3-mm-thick sections in transverse and coronal reconstructions. All images were viewed on a workstation and printed on hard copy with soft tissue (window width, 270 HU; window level, 100 HU) and bone (window width, 3200 HU; window level, 600 HU) settings.
Bone scintigraphy with SPECT/CT was performed 3 hours after the injection of a standard activity of 650 MBq technetium 99m (99mTc) 3,3-diphosphono-1,2-propanedicarboxylic-acid (DPD) (Teceos, Schering, Switzerland). First, whole-body images were obtained with the patient supine by using a two-headed whole-body gamma camera (Bodyscan; Siemens). After acquisition of the whole-body images, the patient was placed on a two-headed gamma camera with an integrated CT scanner (Hawkeye Millenium VG8; GE Medical Systems, Milwaukee, Wis) and SPECT images were obtained in a step-and-shoot mode. The angular steps were 3° with a range of 180° per gamma camera head and a 30-second acquisition per step. The image matrix was 256 x 256, and images were reconstructed as 4.42-mm-thick sections by using an iterative algorithm. The CT data from the SPECT/CT examination were reconstructed in the transverse plane as 10-mm-thick sections. The CT data were also used for attenuation correction of the SPECT data. All images were viewed with software (eNtegra 2.5202; GE Medical Systems, Waukesha, Wis) that provided multiplanar reformatted images of PET, CT, and fused data with linked cursors. All images were displayed in contiguous transverse, coronal, and sagittal sections.
The PET/CT examinations were performed with an in-line scanner (Discovery LS; GE Medical Systems, Waukesha, Wis) that enabled the acquisition of coregistered CT and PET images in the same examination. The patients fasted for at least 4 hours before the intravenous injection of a standard activity of 370 MBq of FDG. The blood sugar level in each patient was measured before scanning. In this study, no patients with diabetes were examined and blood sugar levels were within the normal range in all patients. After an uptake time of approximately 60 minutes, imaging began with CT performed during breath holding as previously described (13).
The following parameters were used for imaging: 140 kV, 80 mA, 0.5 second per rotation, a table speed of 38.5 mm/sec, 867-mm coverage, and an acquisition time of 22.5 seconds. PET was initiated immediately after the CT examination, and the CT data were used for the attenuation correction of PET data. Intravenous contrast material was not used for PET/CT because all patients underwent contrast-enhanced helical CT in addition to PET/CT. However, the patients were asked to drink oral contrast material (1000 mL of a 1.5% diluted barium sulfate suspension [Micropaque Scanner; Guerbet]) starting approximately 45 minutes before PET/CT to improve visualization of abnormalities in the abdomen.
The PET data were reconstructed by using an iterative ordered subset expectation maximization algorithm (28 subsets, two iterative steps). The reconstructed section thickness in the transverse plane was 4.25 mm for PET and CT images. All images were viewed with software (Xeleris Functional Imaging Workstation 1.0628; GE Medical Systems, Waukesha, Wis) that provided multiplanar reformatted images of PET, CT, and fused data with linked cursors. All images were displayed in contiguous transverse, coronal, and sagittal sections.
For all imaging examinations, the patients were asked to remove all metal parts such as artificial dentures and bridgework from the mouth whenever possible. All three imaging tests were scheduled to occur within a maximum of 4 weeks in all patients, and none of the patients underwent treatment between the different examinations.
Image Interpretation
The contrast-enhanced CT scans were displayed in gray scale and printed on film. Two radiologists (B.S., with 7 years of experience, and a second radiologist at our Institute of Neuroradiology with 5 years of experience) viewed the CT scans obtained in all patients with knowledge of the clinical data but without knowledge of the results of the other imaging examinations. The bone and soft-tissue windows were assessed in consensus. Bone invasion was suggested when contrast-enhanced tumor tissue was visible outside the cortical bone and the cortical bone was seen to be partially eroded or totally destroyed.
The SPECT images were displayed in gray scale, in which dark areas corresponded to regions with increased tracer uptake and bright areas to regions with low or absent tracer uptake. The CT images were assessed by using bone and soft-tissue windows. The SPECT/CT images were displayed with tracer uptake overlaid in color on the CT images. The quality of image coregistration was visually controlled. The window settings were those used to evaluate bone and soft tissues. Two nuclear medicine physicians (G.W.G. and D.T.S., both with 2 years of experience in reading SPECT/CT images) visually assessed all images on a computer screen. Images were read in consensus and with knowledge of clinical information but without knowledge of the results of other imaging examinations. Bone invasion was suggested when areas with locally increased tracer uptake corresponding to increased bone turnover were detected at the mandible or at the maxillary bone. The images obtained with whole-body skeletal scintigraphy, which is used as a routine staging test to detect areas with increased tracer uptake suggestive of distant metastases, were assessed separately.
The PET images were displayed in gray scale, in which dark areas corresponded to regions with increased FDG uptake and bright areas to regions with low or absent FDG uptake. The CT scans were displayed by using the different window settings needed to evaluate lungs, bone, and soft tissues. PET/CT images were displayed with FDG uptake overlaid in color on the CT scans. Two nuclear medicine physicians (G.W.G. and D.T.S., both with 4 years of experience in reading PET/CT images) visually assessed the PET, CT, and coregistered PET/CT images on a computer screen in consensus. Although they were blinded to the results of SPECT/CT and contrast-enhanced CT, clinical information was available. The quality of image coregistration was visually controlled on screen. Invasion of tumor into the bone was suggested when FDG uptake was present adjacent to a cortical bone that had a visible defect on CT images or when FDG uptake was visible on the outside of the cortical bone and within the bone marrow in the same region but there was no detectable cortical erosion.
In addition, meaningful findings at whole-body skeletal scintigraphy and whole-body PET—that is, areas with increased tracer uptake suggestive of distant metastases, were evaluated and documented by the same readers in consensus (G.W.G., with 7 years of experience in PET and 12 of years experience in conventional nuclear medicine, and D.T.S., with 4 years of experience in PET and conventional nuclear medicine). All suspicious findings identified by the two readers on the different image sets were documented to enable comparison with those identified by using other methods.
Data and Statistical Analysis
Images were classified as positive or negative for bone invasion. Equivocal readings were not permitted. All patients underwent surgical resection of the tumor. Bone resection was always performed when results of at least one of the above-mentioned imaging tests were suspicious for bone invasion. In patients in whom bone invasion was not suspected on the basis of results of any of the imaging modalities, bone resection was performed only when intraoperative findings were suspicious for invasion. Fast-frozen histologic slices of the soft tissues adjacent to the bone were always obtained intraoperatively for evaluation of whether there was tumor-free soft tissue adjacent to the bone. This is the common way to rule out bone involvement in such patients and can be considered the reference standard. If this method revealed that the tumor was removed with a rim of normal tissue (R0 resection), the surgeon did not perform bone resection.
With use of this standard of reference, it was possible to construct 2 x 2 contingency tables. These data were used to calculate sensitivity, specificity, accuracy, and positive and negative predictive values for each imaging method. The performance parameters for the imaging techniques were calculated for all patients together. Furthermore, the performance parameters were separately calculated after patients in whom metal artifacts rendered image interpretation difficult were excluded.
In addition to evaluating the performance of each technique in the identification of bone invasion in patients suspected of having oral cavity carcinoma, we performed a separate analysis of the information obtained with whole-body skeletal scintigraphy and PET/CT.
|
RESULTS
|
|---|
Thirty-one patients had squamous cell carcinoma. In one patient (patient 20), an aggressive tumor with possible invasion into the bone was clinically suspected, and contrast-enhanced CT and PET/CT revealed the presence of bone invasion, but histologic examination revealed a giant cell epulis. Therefore, no TNM stage was recorded for this patient (Table 1). Two patients (patients 21 and 34) had adenoid cystic carcinomas. In all patients, high FDG uptake was visible at the site of the primary tumor. The primary tumor was at the upper alveolar ridge in four patients (patients 1, 21, 25, and 34) and in the retromolar trigone reaching the upper alveolar ridge in two (patients 29 and 31). In all other patients, the primary tumor was at the lower alveolar ridge and mandibular bone invasion was suspected at clinical examination.
Four patients had previously undergone treatment for an oral cavity carcinoma with or without surgery (patients 3, 19, 23, and 32). Patient 3 had undergone radiation therapy 9 years earlier. Patient 19 had undergone combined chemotherapy and radiation treatment and received a reconstructive autologous fibular bone graft 2 years earlier. Patient 23 had undergone combined radiation and chemotherapy treatment and surgery (but did not receive a bone graft) 2 years earlier. The T stage was assessed at autopsy in this patient because she died of fulminant mediastinal infection before surgery. Results of autopsy helped confirm the presence of mandibular bone infiltration. Patient 32 had undergone three surgical interventions during the previous 8 years for an oral cavity cancer with local recurrence. In this patient, autologous bone grafting had been performed 7 years earlier. An actual TNM stage was not available for this patient because the tumor was considered to be a local recurrence of the previous primary tumor (which, at that time, was classified as stage T4N0). Conversely, the actual tumors in patients 3, 19, and 23 were considered to be metachronous secondary cancers.
Histologic Results
Histopathologic confirmation of the tumor was available for all patients. Bone resection was performed in 15 patients. In the other patients, in whom fast-frozen histologic slices of the soft tissues adjacent to the bone revealed tumor-free soft tissue, no bone specimen was available. Bone invasion was confirmed in 12 of the 34 patients, and clinically suspected bone infiltration was ruled out in 22 patients. Thus, the prevalence of bone invasion was 35% in our patients. In patients 5 and 8, radiation treatment with chemotherapy was performed as an induction treatment before surgery. In these two patients, bone involvement was also ruled out during surgery.
Modality Performance
A summary of the 2 x 2 contingency data and modality performance parameters is given in Table 2. PET/CT had the highest sensitivity and a negative predictive value of 100% (Table 2). Contrast-enhanced CT, however, had a higher specificity, a positive predictive value of 100%, and the highest accuracy of all imaging tests (Table 2).
There were two false-positive findings with PET/CT (in patients 12 and 13). Neither patient had previously undergone surgery, and both were undergoing staging examinations for squamous cell carcinoma in the lower alveolar ridge (Fig 1). With SPECT/CT, three false-positive findings and one false-negative finding were reported. False-positive findings were reported in patients 13, 19, and 28. Patient 19 had previously undergone reconstructive bone surgery 2 years earlier, and the homogeneously increased tracer uptake in the bone was due to normal postoperative changes in vital bone. In patient 20, a false-negative finding was reported with SPECT/CT. This patient had giant cell epulis and there was no increased tracer uptake visible in the mandible; however, bone erosion was depicted with PET/CT and contrast-enhanced CT. Only one false-positive finding was reported with contrast-enhanced CT (in patient 34). In this patient, the PET/CT study did not depict bone erosion, but FDG uptake, a finding that is suggestive of tumor invasion, was clearly visible at both sides of the cortical bone.
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. Patient 12. False-positive PET/CT finding in 55-year-old man who underwent imaging to stage a squamous cell carcinoma in the lower alveolar ridge (pT2). (a) Transverse unenhanced CT image from PET/CT shows metal artifacts. (b) Transverse PET image from PET/CT shows avid FDG uptake adjacent to the mandibular bone. (c) Coregistered image of a and b shows that FDG uptake (arrowheads) seems to be present not only adjacent to the bone but also within the bone marrow cavity. (d) Transverse coregistered PET/CT scan obtained adjacent to c (a section without metal artifacts is shown) shows FDG uptake in the soft tissue adjacent to the bone. Corresponding transverse (e) CT scan, (f) SPECT scan, and (g) coregistered SPECT/CT scan show no increased uptake of 99mTc DPD. No bone erosion is visible.
|
|
View larger version (60K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. Patient 12. False-positive PET/CT finding in 55-year-old man who underwent imaging to stage a squamous cell carcinoma in the lower alveolar ridge (pT2). (a) Transverse unenhanced CT image from PET/CT shows metal artifacts. (b) Transverse PET image from PET/CT shows avid FDG uptake adjacent to the mandibular bone. (c) Coregistered image of a and b shows that FDG uptake (arrowheads) seems to be present not only adjacent to the bone but also within the bone marrow cavity. (d) Transverse coregistered PET/CT scan obtained adjacent to c (a section without metal artifacts is shown) shows FDG uptake in the soft tissue adjacent to the bone. Corresponding transverse (e) CT scan, (f) SPECT scan, and (g) coregistered SPECT/CT scan show no increased uptake of 99mTc DPD. No bone erosion is visible.
|
|
View larger version (125K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1c. Patient 12. False-positive PET/CT finding in 55-year-old man who underwent imaging to stage a squamous cell carcinoma in the lower alveolar ridge (pT2). (a) Transverse unenhanced CT image from PET/CT shows metal artifacts. (b) Transverse PET image from PET/CT shows avid FDG uptake adjacent to the mandibular bone. (c) Coregistered image of a and b shows that FDG uptake (arrowheads) seems to be present not only adjacent to the bone but also within the bone marrow cavity. (d) Transverse coregistered PET/CT scan obtained adjacent to c (a section without metal artifacts is shown) shows FDG uptake in the soft tissue adjacent to the bone. Corresponding transverse (e) CT scan, (f) SPECT scan, and (g) coregistered SPECT/CT scan show no increased uptake of 99mTc DPD. No bone erosion is visible.
|
|
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1d. Patient 12. False-positive PET/CT finding in 55-year-old man who underwent imaging to stage a squamous cell carcinoma in the lower alveolar ridge (pT2). (a) Transverse unenhanced CT image from PET/CT shows metal artifacts. (b) Transverse PET image from PET/CT shows avid FDG uptake adjacent to the mandibular bone. (c) Coregistered image of a and b shows that FDG uptake (arrowheads) seems to be present not only adjacent to the bone but also within the bone marrow cavity. (d) Transverse coregistered PET/CT scan obtained adjacent to c (a section without metal artifacts is shown) shows FDG uptake in the soft tissue adjacent to the bone. Corresponding transverse (e) CT scan, (f) SPECT scan, and (g) coregistered SPECT/CT scan show no increased uptake of 99mTc DPD. No bone erosion is visible.
|
|
View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1e. Patient 12. False-positive PET/CT finding in 55-year-old man who underwent imaging to stage a squamous cell carcinoma in the lower alveolar ridge (pT2). (a) Transverse unenhanced CT image from PET/CT shows metal artifacts. (b) Transverse PET image from PET/CT shows avid FDG uptake adjacent to the mandibular bone. (c) Coregistered image of a and b shows that FDG uptake (arrowheads) seems to be present not only adjacent to the bone but also within the bone marrow cavity. (d) Transverse coregistered PET/CT scan obtained adjacent to c (a section without metal artifacts is shown) shows FDG uptake in the soft tissue adjacent to the bone. Corresponding transverse (e) CT scan, (f) SPECT scan, and (g) coregistered SPECT/CT scan show no increased uptake of 99mTc DPD. No bone erosion is visible.
|
|
View larger version (74K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1f. Patient 12. False-positive PET/CT finding in 55-year-old man who underwent imaging to stage a squamous cell carcinoma in the lower alveolar ridge (pT2). (a) Transverse unenhanced CT image from PET/CT shows metal artifacts. (b) Transverse PET image from PET/CT shows avid FDG uptake adjacent to the mandibular bone. (c) Coregistered image of a and b shows that FDG uptake (arrowheads) seems to be present not only adjacent to the bone but also within the bone marrow cavity. (d) Transverse coregistered PET/CT scan obtained adjacent to c (a section without metal artifacts is shown) shows FDG uptake in the soft tissue adjacent to the bone. Corresponding transverse (e) CT scan, (f) SPECT scan, and (g) coregistered SPECT/CT scan show no increased uptake of 99mTc DPD. No bone erosion is visible.
|
|
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1g. Patient 12. False-positive PET/CT finding in 55-year-old man who underwent imaging to stage a squamous cell carcinoma in the lower alveolar ridge (pT2). (a) Transverse unenhanced CT image from PET/CT shows metal artifacts. (b) Transverse PET image from PET/CT shows avid FDG uptake adjacent to the mandibular bone. (c) Coregistered image of a and b shows that FDG uptake (arrowheads) seems to be present not only adjacent to the bone but also within the bone marrow cavity. (d) Transverse coregistered PET/CT scan obtained adjacent to c (a section without metal artifacts is shown) shows FDG uptake in the soft tissue adjacent to the bone. Corresponding transverse (e) CT scan, (f) SPECT scan, and (g) coregistered SPECT/CT scan show no increased uptake of 99mTc DPD. No bone erosion is visible.
|
|
In patients 5, 12, 16, 25, 28, 30, and 34, it was difficult to interpret the contrast-enhanced CT scans and the CT data from PET/CT and SPECT/CT scans owing to metal artifacts. When these seven patients were excluded from the analysis, the sensitivity, specificity, accuracy, and positive and negative predictive values were 100% (10 of 10 patients), 94% (16 of 17 patients), 96% (26 of 27 patients), 91% (10 of 11 patients), and 100% (16 of 16 patients), respectively, for PET/CT. For SPECT/CT, the corresponding parameters were 90% (nine of 10 patients), 88% (15 of 17 patients), 89% (24 of 27 patients), 82% (nine of 11 patients), and 94% (15 of 16 patients). For contrast-enhanced CT, a sensitivity and specificity of 100% (10 of 10 and 17 of 17 patients, respectively) resulted in an accuracy of 100% (27 of 27 patients). Therefore, the negative and positive predictive values also reached 100% (16 of 16 and 11 of 11 patients, respectively).
The performance of SPECT/CT was always worse than that of PET/CT or contrast-enhanced CT, and the exclusion of images with interfering metal artifacts resulted in only a marginal improvement in accuracy (from 86% to 89%). In addition, the exclusion of PET/CT studies with metal artifacts did not improve test accuracy in a relevant way (accuracy improved from 94% to 96%). Conversely, no false-positive or false-negative findings occurred with contrast-enhanced CT when patients whose images had metal artifacts were excluded from the analysis.
In patients who had previously undergone treatment for oral cavity carcinomas, the three imaging modalities performed as follows: In patients 3 and 23, bone infiltration was correctly identified with all three imaging examinations (Fig 2). In patient 32, all three imaging examinations enabled us to correctly rule out the presence of bone invasion. In patient 19, PET/CT and contrast-enhanced CT enabled us to correctly rule out the presence of bone involvement, but SPECT/CT findings were false-positive.
View larger version (167K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Patient 3. Example of correct assessment of bone invasion with all three imaging methods in 55-year-old man with pT4 squamous cell carcinoma of the mandible. This patient had undergone radiation therapy 9 years previously for a squamous cell carcinoma in the oral cavity. (a, b) Transverse contrast-enhanced CT scans obtained with (a) soft tissue and (b) bone levels clearly depict the presence of bone invasion (arrow). (c) Transverse coregistered PET/CT scan shows avid FDG uptake at the mandibular bone, and bone erosion (arrow) is evident. (d) Transverse coregistered SPECT/CT scan shows avid uptake of 99mTc DPD (arrow), a finding that is suggestive of bone invasion.
|
|
View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Patient 3. Example of correct assessment of bone invasion with all three imaging methods in 55-year-old man with pT4 squamous cell carcinoma of the mandible. This patient had undergone radiation therapy 9 years previously for a squamous cell carcinoma in the oral cavity. (a, b) Transverse contrast-enhanced CT scans obtained with (a) soft tissue and (b) bone levels clearly depict the presence of bone invasion (arrow). (c) Transverse coregistered PET/CT scan shows avid FDG uptake at the mandibular bone, and bone erosion (arrow) is evident. (d) Transverse coregistered SPECT/CT scan shows avid uptake of 99mTc DPD (arrow), a finding that is suggestive of bone invasion.
|
|
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2c. Patient 3. Example of correct assessment of bone invasion with all three imaging methods in 55-year-old man with pT4 squamous cell carcinoma of the mandible. This patient had undergone radiation therapy 9 years previously for a squamous cell carcinoma in the oral cavity. (a, b) Transverse contrast-enhanced CT scans obtained with (a) soft tissue and (b) bone levels clearly depict the presence of bone invasion (arrow). (c) Transverse coregistered PET/CT scan shows avid FDG uptake at the mandibular bone, and bone erosion (arrow) is evident. (d) Transverse coregistered SPECT/CT scan shows avid uptake of 99mTc DPD (arrow), a finding that is suggestive of bone invasion.
|
|
View larger version (112K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2d. Patient 3. Example of correct assessment of bone invasion with all three imaging methods in 55-year-old man with pT4 squamous cell carcinoma of the mandible. This patient had undergone radiation therapy 9 years previously for a squamous cell carcinoma in the oral cavity. (a, b) Transverse contrast-enhanced CT scans obtained with (a) soft tissue and (b) bone levels clearly depict the presence of bone invasion (arrow). (c) Transverse coregistered PET/CT scan shows avid FDG uptake at the mandibular bone, and bone erosion (arrow) is evident. (d) Transverse coregistered SPECT/CT scan shows avid uptake of 99mTc DPD (arrow), a finding that is suggestive of bone invasion.
|
|
Whole-Body Examinations
Whole-body skeletal scintigraphy did not depict distant bone metastases in the 34 patients. Conversely, PET/CT of the thorax and abdomen depicted distant metastases in one patient (patient 23). In this patient, who had an advanced-stage squamous cell carcinoma at the inferior alveolar ridge, soft-tissue metastases were identified in the lung, thoracic wall, and mediastinum. The metastases were verified by means of ultrasonographically guided biopsy of a thoracic wall metastasis and at autopsy.
|
DISCUSSION
|
|---|
Our data suggest that contrast-enhanced CT and PET/CT are better than SPECT/CT for the identification of invasion of the mandibular and maxillary bones in patients with oral cavity cancers. In this study, the performances of PET/CT and contrast-enhanced CT were comparable; however, PET/CT had the highest sensitivity and contrast-enhanced CT the highest specificity of all imaging tests. The difference between the CT scans obtained during contrast-enhanced CT and those obtained at PET/CT was the section thickness (3 mm for contrast-enhanced CT vs 4.25 mm for PET/CT) and the fact that intravenous contrast material was not used with whole-body PET/CT. The thicker sections of in-line PET/CT can cause problems in image interpretation, especially when areas with marginal bone invasion must be assessed in detail. Therefore, small areas of bone erosion may be missed with thicker sections. In this study, we tried to overcome this potential drawback of whole-body PET/CT by using an additional criterion of FDG uptake in postulating the presence of bone invasion: FDG uptake identified on both sides of a cortical bone without evident cortical bone erosion was thought to be indicative of infiltration. This pattern led to the correct identification of bone infiltration in only one patient (patient 34) and to two false-positive findings. This highly sensitive method of image interpretation explains the 100% negative predictive value obtained with PET/CT in this study. We believe, however, that this additional interpretation criterion is of limited value because the interpretation of the extent of an FDG-avid area is dependent on several factors, as follows: First, the quality of image coregistration is of utmost importance and a mismatch of a few millimeters will lead to an incorrect assumption of FDG uptake at two sides of a cortical bone. Second, "up-leveling" the PET image in gray scale (thus rendering the dark areas in the image that correspond to sites of FDG uptake darker) will lead to an overestimation of the extent of the FDG-avid area. Third, partial volume effects can cause visualization of FDG activity in an image section adjacent to the tumor and, therefore, the extent of tumor tissue can be overestimated.
The most important and most reliable image criterion in our study was the identification of cortical erosion and bone destruction at contrast-enhanced CT. If only bone invasion is to be assessed, however, the use of intravenous contrast material seems unnecessary.
The performance of SPECT/CT was worse than that of contrast-enhanced CT and PET/CT because interpretation of SPECT/CT images mainly relied on assessment of increased tracer uptake in the bone. The quality of the CT scans from SPECT/CT was far too poor to enable the reliable identification of small areas of bone erosion. Therefore, we believe that CT scans from combined bone SPECT/CT are of limited value in the evaluation of patients suspected of having bone invasion from oral cavity carcinoma. Furthermore, the low-dose protocol with thick sections that we used was susceptible to metal artifacts. In addition, in this study we included patients who had previously undergone therapies, and two of these patients (patients 19 and 32) had received bone grafts to replace resected mandibular bone. In such patients, the uptake of bone-avid tracers can be increased due to remodeling processes in the healing bone graft. False-positive findings are possible in patients who have recently undergone tooth extraction. In our study, we had one false-negative finding at bone SPECT in a patient with evident bone erosion due to giant cell epulis, a benign proliferation of the gingiva containing numerous multinucleated giant cells. If only patients without previous surgical and dental interventions and with a verified malignant lesion had undergone scanning, however, the sensitivity and specificity of SPECT and SPECT/CT would be much higher. All patients in this study with a correct SPECT/CT diagnosis also had a correct diagnosis on the basis of the other imaging tests. Therefore, we suggest that the acquisition of additional SPECT/CT scans is helpful only in patients with uncertain findings at contrast-enhanced CT or PET/CT.
The effect of metal artifacts was limited in this study because, with all three imaging tests, only one false-negative or false-positive finding was avoided when data in the seven patients whose images showed metal artifacts were excluded from the analysis. This is in line with results of a recent study (14) that suggested that metal artifacts do not worsen the visibility of FDG-avid tumors in the oral cavity. Metal artifacts, however, were the limiting factor in the performance of contrast-enhanced CT because a correct diagnosis was obtained in all patients whose images were free of metal artifacts.
The use of a whole-body imaging protocol can be an advantage of PET/CT in the detection of distant metastases in the same imaging session. It has previously been shown that the identification of distant metastases in patients who undergo whole-body PET for the staging of squamous cell carcinoma in the oral cavity or in the head and neck area can add relevant clinical information (11,12,15). In this study, however, the detection of distant metastases had no clinical effect on further treatment because the single patient in our study with such metastases died of fulminant mediastinitis.
A limitation of this study was that bone resection was performed only in patients with positive imaging tests and when the findings at intraoperative assessment were suspicious for bone involvement. The use of intraoperative fast-frozen histologic slices for evaluating tumor-free soft tissue adjacent to the bone, however, is a common way to rule out bone involvement in these patients. Therefore, we believe that the methodologic approach used in this study did not introduce a bias in patient selection. In addition, the prevalence of bone invasion in this study was 35%, and, therefore, our data relied on a relatively small number of positive findings. Because patient characteristics such as a history of maxillofacial surgery with bone transplants can have an influence on the performance of imaging tests, diagnostic accuracy may be subject to variability when other patient populations are scanned.
In conclusion, our data suggest that the identification of bone involvement in patients with oral cavity carcinomas is reliably performed with helical CT and thin sections. In patients who undergo PET/CT for whole-body staging or repeat staging, the CT information from PET/CT is reliable, whereas FDG uptake does not help better identify bone invasion.
|
FOOTNOTES
|
|---|
Abbreviations: DPD = 3,3-diphosphono-1,2-propanedicarboxylic acid • FDG = fluorine 18 fluorodeoxyglucose
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, G.W.G., D.T.S.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, G.W.G., D.T.S., G.K.E.; clinical studies, D.T.S., G.K.E.; statistical analysis, G.W.G.; and manuscript editing, G.W.G., D.T.S., B.S.
|
References
|
|---|
- O'Brien CJ, Carter RL, Soo KC, Barr LC, Hamlyn PJ, Shaw HJ. Invasion of the mandible by squamous carcinomas of the oral cavity and oropharynx. Head Neck Surg 1986;8:247–256.[Medline]
- Close LG, Merkel M, Burns DK, Schaefer SD. Computed tomography in the assessment of mandibular invasion by intraoral carcinoma. Ann Otol Rhinol Laryngol 1986;95:383–388.[Medline]
- Lewis-Jones HG, Rogers SN, Beirne JC, Brown JS, Woolgar JA. Radionuclide bone imaging for detection of mandibular invasion by squamous cell carcinoma. Br J Radiol 2000;73:488–493.[Abstract]
- Zieron JO, Lauer I, Remmert S, Sieg P. Single photon emission tomography: scintigraphy in the assessment of mandibular invasion by head and neck cancer. Head Neck 2001;23:979–984.[CrossRef][Medline]
- Imola MJ, Gapany M, Grund F, Djalilian H, Fehling S, Adams G. Technetium 99m single positron emission computed tomography scanning for assessing mandible invasion in oral cavity cancer. Laryngoscope 2001;111:373–381.[CrossRef][Medline]
- Brown JS, Griffith JF, Phelps PD, Browne RM. A comparison of different imaging modalities and direct inspection after periosteal stripping in predicting the invasion of the mandible by oral squamous cell carcinoma. Br J Oral Maxillofac Surg 1994;32:347–359.[CrossRef][Medline]
- Leitha T, Glaser C, Pruckmayer M, et al. Technetium-99m-MIBI in primary and recurrent head and neck tumors: contribution of bone SPECT image fusion. J Nucl Med 1998;39:1166–1171.[Abstract/Free Full Text]
- Schimming R, Juengling FD, Altehofer C, Schmelzeisen R. Diagnosis of questionable mandibular infiltration by squamous epithelial carcinomas: 3-D 99mTc-DPD SPECT reconstruction and (18F) fluoride PET study—diagnostic advantages or unnecessary expense? HNO 2001;49:355–360.[CrossRef][Medline]
- Brockenbrough JM, Petruzzelli GJ, Lomasney L. DentaScan as an accurate method of predicting mandibular invasion in patients with squamous cell carcinoma of the oral cavity. Arch Otolaryngol Head Neck Surg 2003;129:113–117.[Abstract/Free Full Text]
- Acton CH, Layt C, Gwynne R, Cooke R, Seaton D. Investigative modalities of mandibular invasion by squamous cell carcinoma. Laryngoscope 2000;110:2050–2055.[CrossRef][Medline]
- Goerres GW, Schmid DT, Gratz KW, von Schulthess GK, Eyrich GK. Impact of whole body positron emission tomography on initial staging and therapy in patients with squamous cell carcinoma of the oral cavity. Oral Oncol 2003;39:547–551.[CrossRef][Medline]
- Hlawitschka M, Neise E, Bredow J, et al. FDG-PET in the pretherapeutic evaluation of primary squamous cell carcinoma of the oral cavity and the involvement of cervical lymph nodes. Mol Imaging Biol 2002;4:91–98.[CrossRef][Medline]
- Goerres GW, Kamel E, Heidelberg TN, Schwitter MR, Burger C, von Schulthess GK. PET-CT image co-registration in the thorax: influence of respiration. Eur J Nucl Med Mol Imaging 2002;29:351–360.[CrossRef][Medline]
- Goerres GW, Schmid DT, Eyrich GK. Do hardware artefacts influence the performance of head and neck PET scans in patients with oral cavity squamous cell cancer? Dentomaxillofac Radiol 2003;32:365–371.[Abstract/Free Full Text]
- Schmid DT, Stoeckli SJ, Bandhauer F, et al. Impact of positron emission tomography on the initial staging and therapy in locoregional advanced squamous cell carcinoma of the head and neck. Laryngoscope 2003;113:888–891.[CrossRef][Medline]
This article has been cited by other articles:
|
|
|
|
 
A Auluck, S Shetty, R Desai, and M Mupparapu
Recurrent ameloblastoma of the infratemporal fossa: diagnostic implications and a review of the literature
Dentomaxillofac. Radiol.,
October 1, 2007;
36(7):
416 - 419.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
