| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Report |
Dept. of Diagnostic Radiology, University Hospital Tuebingen, Hoppe-Seyler-Strasse 3, Tuebingen D-72076, Germany.
Received May 11, 2004;
accepted after revision July 20, 2004.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. MRI is recommended as the method of choice in the preoperative evaluation of SCC of the oral cavity and the oropharynx. PET can provide relevant diagnostic information in case of equivocal findings by MRI or CT. Routine use of PET, however, does not appear to be necessary if optimized MRI is available.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Patients with prior treatment for malignant disease, renal failure, or known diabetes mellitus were excluded. The study protocol was reviewed and approved by the ethics commission of our institution; informed consent was obtained from all patients.
CT and MRI
All patients underwent axial helical CT (Somatom Plus S, Siemens) with 5-mm
slice thickness, 5-mm rotation table feed, and 4-mm increment during IV
contrast infusion. Additional slices were obtained with a gantry tilt parallel
to the teeth to avoid superimposition of metal artifacts in the oropharynx.
The images were documented on laser film using a soft-tissue window setting
(width, 350 H; level, 50 H). MRI was performed on a 1.0- or 1.5-T whole-body
scanner (Magnetom Impact or Magnetom Vision, Siemens), using a Helmholtz head
and neck surface coil. The examination protocol included axial and coronal
fast spin-echo T1-weighted (TR range/TE, 350-600/12) and inversion recovery
fat-saturated T2-weighted imaging (4,500-7,200/60). Additional
contrast-enhanced (gadolinium, 0.1 mmol/kg of body weight) T1-weighted images
with or without fat suppression were obtained in all cases. Standard slice
thickness was 5 mm with a 1-mm inter-slice gap. In selected cases, spin-echo
T2-weighted images (4,500-5,500/120) or thin slices were obtained.
18FDG PET
18FDG PET was performed using a whole-body PET scanner (Advance,
GE Healthcare) with an axial field of view of 148.5 mm. After a fasting period
of at least 12 hr, patients were injected IV with 300-400 MBq of
18FDG. After an uptake time of 45 min, static emission scans (15
min per field of view) of the tumor region and the neck were acquired; 20-min
transmission scans were then obtained. After emptying of the bladder, the
patients underwent whole-body scanning (6 or 7 fields of view; 5-min emission
time per field of view) followed by acquisition of a transmission scan (2 min
per field of view). Attenuation-corrected and iteratively reconstructed images
of the tumor region and neck with a slice thickness of 4.25 mm and whole-body
images with a slice thickness of 8.5 mm were generated using a standardized
scaling with standard uptake value of 0-8.
Image Analysis
PET images were evaluated visually by two experienced nuclear medicine
radiologists. Any focus that was hypermetabolic compared with muscle or
salivary glands was considered suggestive of tumor. CT and MR images were
reviewed by two head and neck radiologists. Malignancy of primary tumors and
lymph nodes was diagnosed using established morphologic criteria including a
lymph node size larger than 10 mm, a conglomeration of a minimum of three
lymph nodes, central necrosis, or indistinct nodal margins for lymph node
metastases [2,
8]. All reviewers were aware of
the inclusion criteria of the study but blinded to all other clinical data,
the surgical findings, and the results of the other imaging techniques. Tumor
and lymph node stages were classified according to the TNM scheme set forth by
the Union Internationale Contre le Cancer
[16].
The level of confidence for diagnosing or excluding malignancy was graded using a 6-point scale (1 = definitely not malignant, 2 = probably not malignant, and so on to 6 = definitively malignant). Sensitivity, specificity, and accuracy values were calculated on the basis of a dichotomous grouping of these data, with a summarizing level of 1-3 as a negative finding and a level of 4-6 as a positive diagnosis of malignancy. Additional relevant findings (tumor localization and extension, bone involvement, other complications) were registered. For lymph node staging, four regions were separately evaluated on each side of the neck: a suprahyoidal region (I) including the submental, submandibular, and upper deep cervical nodes; a supraomohyoidal region (II) including the middle deep cervical nodes; an infraomohyoidal region (III) including the pretracheal, prelaryngeal, and lower deep cervical nodes; and a posterolateral region (IV) including the spinal accessory and transverse cervical chain nodes. All imaging results were compared with the findings of histopathology as the gold standard.
Statistical Analysis
A receiver operating characteristic (ROC) analysis for all tumor lesions
and lymph node regions was performed for all techniques and observers involved
using a commercial software program (MedCalc version 5.0, MedCalc). All
p values were calculated on the basis of a sign test. A p
value of less than 0.05 indicated a significant difference.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
18FDG PET.—18FDG correctly detected 51 of the 59 tumors (Table 2). Eight tumors were missed, including six lesions equal to or smaller than 10 mm in diameter. 18FDG PET findings were false-positive in three locations. Similar results were achieved with MRI, whereas CT showed a considerably higher number of false-negative findings (Figs. 1A, 1B, and 1C).
|
|
|
|
Three of 20 pT1 tumors and one of 18 pT2 tumors were not detected on any of the three techniques. Another pT2 tumor of the right tonsillary region was missed by CT and MRI and at the initial clinical examination and was exclusively depicted by PET.
MRI.—MRI showed false results in eight suspected tumor sites. PET allowed correct classification in three of these sites, including one of the five false-negative and two of the three false-positive MRI findings. MRI allowed correct diagnosis in six cases misinterpreted by PET (false-negative, four; false-positive, two). False-positive results with PET included a dental focus and an inflammatory reaction after a biopsy performed 7 weeks before PET.
CT.—CT failed to show a primary tumor in 23 cases. PET correctly detected 17 of these tumors, and MRI revealed 18 tumors. Six false-negative CT diagnoses were not corrected by PET, including two cases with true-positive MRI findings.
Confidence level.—With respect to the confidence level for tumor diagnosis, CT (mean ± SD, 3.84 ± 1.69) was significantly inferior compared with MRI (5.10 ± 1.26) or PET (5.09 ± 1.70).
A borderline rating (i.e., ratings 3 or 4) indicating a low level of confidence in the scan interpretation was obtained for PET in only five cases, compared with 17 for CT and 12 for MRI. PET provided an improved level of confidence and corrected a false diagnostic decision using CT data in eight patients. In the cases of equivocal MRI findings, PET provided a superior level of confidence in five patients and corrected a false diagnosis in one patient.
Tumor stage.—Tumor stage was correctly classified by CT in 31 and by MRI in 42 of the 59 malignant lesions. PET did not allow evaluation of the morphologically defined tumor stage.
In five patients, malignant disease was simultaneously present at two different sites, including lung carcinoma in one patient and esophageal carcinoma in another two patients (Fig. 2). Four of these tumors were correctly identified by PET and confirmed at endoscopy or surgery. PET revealed further suspicious findings in the lung hila in two patients, on the maxillary alveolar rim in one patient with a pT1 mouth floor carcinoma, and within a cervical vertebral arch in one patient. All of them were finally classified as benign considering all of the clinical information that was available.
|
Lymph Node Staging
Histologic findings.—The extent of surgical neck dissections
was limited to one or two regions in 69 of the 109 neck sides that underwent
surgery. In 50 of the 65 patients, a bilateral neck dissection was performed.
As a result, surgical and histologic findings were available for 293 of the
512 lymph node regions. A summary is given in
Table 1. Lymph node metastases
were found in 40 regions of 32 neck sides, or in 31 of 59 patients.
Five lymph node regions were misclassified with all three imaging techniques (two false-negative and three false-positive diagnoses). PET findings were false-positive in only six regions, whereas the number of false-positive findings was two to three times higher for CT and MRI (Figs. 3A, 3B, and 3C). Table 2 summarizes the results of PET, MRI, and CT. Although the positive predictive values were moderate (PET = 0.85, CT = 0.65, and MRI = 0.73), high negative predictive values (PET = 0.98, CT = 0.97, and MRI = 0.99) were achieved by all techniques.
|
|
|
18FDG PET.—PET allowed correctly classification of 281 of the 293 lymph node regions. In 10 regions, a correct diagnosis was achieved exclusively with PET, whereas in five regions, PET was the only technique that provided a false result.
MRI.—False-positive MRI findings were corrected by PET in 11 lymph node regions. Metastatic disease was missed in another region by MRI but detected by PET. In contrast, PET provided false information in eight regions that were correctly diagnosed with MRI, including five regions with metastatic disease that were not detected by PET.
CT.—An incorrect diagnosis was obtained on CT for 25 lymph node regions. PET allowed the correct diagnosis in 19 of them, including five false-negative and 14 false-positive CT findings. CT allowed correct classification in five regions misclassified with PET, including two regions with a false-negative PET diagnosis.
Equivocal findings.—Equivocal findings (i.e., ratings of 3 or 4) were obtained with CT in 27 lymph node regions. PET allowed the correct diagnosis with a high level of confidence (ratings of 1, 2, 5, or 6) in 21 of these regions. MRI allowed the correct diagnosis with a high level of confidence in seven of these regions and revealed borderline ratings in 30 lymph node regions (Table 3). A correct diagnosis with a high confidence level resulted from PET in 22 of these regions and from CT in 11.
|
ROC analysis.—Results of ROC analysis of the performance of lymph node classification with CT, MRI, and PET are shown in Figure 4. The areas under the curves are not significantly different.
|
Summarized results of a per-neck-side analysis.—PET allowed correction of nearly all false-positive or overestimated CT (Table 4) or MRI (Table 5) diagnoses and was more efficient for this purpose than MRI or CT, respectively. False-negative findings with CT or MRI were less frequent and harder to correct with PET or with corresponding MRI or CT, respectively. Contralateral lymph node disease occurred in four patients and was correctly identified by all three techniques. Again, PET was the most specific technique, causing false-positive diagnosis in one case compared with two cases for MRI or four cases for CT.
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Tumor Staging
Large primary tumors.—Large primary tumors of the oral
cavity or the oropharynx can be detected easily by clinical examination.
Additional information about tumor extension into the deep spaces, the
relationship to adjacent structures, and bone infiltration is needed for
treatment planning. Both MRI and CT met these requirements in all four tumors
of our series with a diameter larger than 4 cm. PET had no additional value in
this situation because of the lack of morphologic information.
Small tumors.—For small tumors, the reported sensitivity and specificity of CT and MRI have been low [24]. In early reports, PET was more accurate [7, 21, 22]; however, the number of patients was low and the technical quality of MRI was limited. Recent studies showed similar results for PET and CT [14, 20]. In our series, the sensitivity and level of confidence of PET diagnoses were significantly higher than those of CT. However, the sensitivity of PET did not reach the figures reported by other groups, which range from 88% to 100% [7, 15, 20, 22]. This may be explained by the relatively small size of the primary tumors involved in our study. The low sensitivity of CT may be explained by the high number of small tumors in the oral cavity where diagnosis is frequently impaired by superimposing dental metal artifacts. In contrast, MRI was not affected by such artifacts. The high sensitivity of MRI in our series contrasts with results from previous reports [7, 22] and may be a result of technical improvement. MRI and PET provided similar sensitivity, levels of confidence, and potential to correct uncertain or false CT findings. PET was superior to MRI at one synchronous tumor site in the head and neck region that was exclusively detected by PET and in five of the 12 tumor lesions for which PET provided a superior level of confidence.
CUP syndrome.—PET may be superior to morphologic imaging in patients with unknown sites of the primary carcinoma (CUP syndrome) [11, 25] but may also provide false-positive findings in up to 11% of these cases [11, 26]. In our study, as in the series by AAssar et al. [11], tumors were predominantly localized in the oral cavity. All were detected by clinical examination. The potential advantage of PET may become relevant in head and neck regions where anatomic imaging methods and clinical examinations have lower sensitivity.
Secondary tumors.—Synchronous secondary tumors are found in about 8% of all head and neck malignant carcinomas [11, 27]. In our series, a simultaneous malignancy was histologically confirmed in five (8.5%) of the 59 patients, including three lesions outside the head and neck region. All of the latter were clearly detected by a PET whole-body scan. They were missed by the initial CT and MRI examinations of the head and neck region. However, all synchronous malignancies were also detected by routinely performed preoperative panendoscopy or chest radiography. Thus, PET had no relevant impact on the planning treatment of the primary tumors.
Lymph Node Staging
The prognosis for patients with head and neck cancer is strongly influenced
by the presence of lymph node metastases
[7]. Similar to the findings of
previous reports [10],
metastatic lymph node disease was confirmed in approximately half of the
patients in our series. Because of limited morphologic information, PET did
not allow reliable differentiation of N1, N2a, or N2b stages. More important
for treatment planning, however, is regional analysis of lymph node disease.
In our study, we defined four lymph node regions on each neck side as
described by previous reports
[2,
10,
28].
Complete removal of all metastatic lymph nodes is a prerequisite to achieve curative treatment. Morphologic imaging methods, including CT and MRI, are reported to provide a high rate of false-negative diagnoses, which can be explained by micrometastases within otherwise normal lymph nodes [2, 8]. PET, as a functional imaging method, might solve this problem. In previous studies, PET was able to detect metastatic disease in lymph nodes measuring only 4-6 mm [8, 10, 12, 29]. On the other hand, false-negative PET results were reported in large lymph nodes up to 20 mm in diameter [12] or in necrotic lymph nodes [8]. The reported sensitivities of PET for nodal disease range from 67% to 91% [8, 10, 12-14, 21, 22, 29, 30]; similar values were found for CT (67-90%) [2, 8, 10, 12, 13] and MRI (71-91%) [2, 7, 8, 10, 30]. The results of our series are within this range.
High specificity and high negative predictive value for the diagnosis of lymph node disease are required to restrict the extent of a neck dissection so that subsequent morbidity can be minimized [22]. The reported specificity of PET ranges from 88% to 100% [10, 12, 21, 29, 30] compared with a wide range of reported specificity values for CT (38-97%) and MRI (48-94%) [2, 10, 12, 31]. In contrast, all three imaging techniques yielded high specificity (93-98%) and high negative predictive value in our study. This discrepancy might be explained by the inhomogeneity of the examination protocols [2] or the limited number of patients [30] included in these studies.
ROC analysis did not reveal significant differences among the performances of PET, CT, and MRI. However, with respect to surgical planning, PET provided additional information in 20 (31.3%) of 64 patients, whereas CT or MRI findings were equivocal. Furthermore, the potential of PET to improve or correct CT results was superior to that of MRI due to superior specificity. As a consequence, an ipsilateral neck dissection would have been avoided in nine patients and a contralateral neck dissection would have been avoided in four patients in whom metastatic disease was falsely diagnosed with CT but correctly excluded with PET.
In conclusion, morphologic imaging techniques are crucial for therapy planning in primary head and neck carcinomas. The highest sensitivity and optimal anatomic information of the tumor site are provided by MRI using recent technology. Diagnostic performance in lymph nodes is similar for MRI, CT, and PET. However, in case of equivocal findings by MRI or CT, PET provides relevant information for determining the extent of surgical neck dissection. As a consequence, PET can be recommended as an additional diagnostic procedure in these cases. Furthermore, our findings suggest that combining structural and metabolic information with coregistered CT-FDG PET using a dedicated scanner might be the method of choice in the future.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Marur and A. A. Forastiere Head and Neck Cancer: Changing Epidemiology, Diagnosis, and Treatment Mayo Clin. Proc., April 1, 2008; 83(4): 489 - 501. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. G.C. Troost, W. V. Vogel, M. A.W. Merkx, P. J. Slootweg, H. A.M. Marres, W. J.M. Peeters, J. Bussink, A. J. van der Kogel, W. J.G. Oyen, and J. H.A.M. Kaanders 18F-FLT PET Does Not Discriminate Between Reactive and Metastatic Lymph Nodes in Primary Head and Neck Cancer Patients J. Nucl. Med., May 1, 2007; 48(5): 726 - 735. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. M. Blodgett, C. C. Meltzer, and D. W. Townsend PET/CT: Form and Function Radiology, February 1, 2007; 242(2): 360 - 385. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Quon, N. J. Fischbein, I. R. McDougall, Q.-T. Le, B. W. Loo Jr., H. Pinto, and M. J. Kaplan Clinical Role of 18F-FDG PET/CT in the Management of Squamous Cell Carcinoma of the Head and Neck and Thyroid Carcinoma J. Nucl. Med., January 1, 2007; 48(1_suppl): 58S - 67S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Gregoire, K. Haustermans, X. Geets, S. Roels, and M. Lonneux PET-Based Treatment Planning in Radiotherapy: A New Standard? J. Nucl. Med., January 1, 2007; 48(1_suppl): 68S - 77S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Mukherji and C. R. Bradford Controversies: Is There a Role for Positron-Emission Tomographic CT in the Initial Staging of Head and Neck Squamous Cell Carcinoma? AJNR Am. J. Neuroradiol., February 1, 2006; 27(2): 243 - 245. [Full Text] [PDF]
|
|
|
|
|
|
D. Pauleit, A. Zimmermann, G. Stoffels, D. Bauer, J. Risse, M. O. Fluss, K. Hamacher, H. H. Coenen, and K.-J. Langen 18F-FET PET Compared with 18F-FDG PET and CT in Patients with Head and Neck Cancer J. Nucl. Med., February 1, 2006; 47(2): 256 - 261. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
