| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Division of Otorhinolaryngology, Head and Neck Surgery, Department of Surgery,
The University of Hong Kong Medical Centre, Queen Mary Hospital, 102 Pokfulam
Rd., Hong Kong.
2 Department of Radiology, The University of Hong Kong Medical Centre, Queen
Mary Hospital, Hong Kong.
3 Department of Pathology, The University of Hong Kong Medical Centre, Queen
Mary Hospital, Hong Kong.
Received May 27, 2003;
accepted after revision September 22, 2003.
Address correspondence to A. P.-W. Yuen.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Eighteen patients with oral tongue cancer
underwent preoperative MRI of the tongue. After surgery, the glossectomy
specimens were serially sectioned. The radiologic tumor thickness of
contrast-enhanced T1-weighted and T2-weighted images was compared with the
histologic tumor thickness using our proposed tumor thickness staging
classifications. These included stage I (tumor 
3 mm), stage II (> 3 mm
but 9 mm) and stage III (> 9 mm).
RESULTS. The overall accuracy in assessment of proposed tumor thickness staging using contrast-enhanced T1-weighted and T2-weighted images was 83% and 56%, respectively. The radiologic tumor thickness as measured on contrast-enhanced T1-weighted and T2-weighted images had significant correlation with histologic tumor thickness (R = 0.938 and 0.941, respectively).
CONCLUSION. MR images provide satisfactory accuracy for the measurement of tumor thickness and staging of oral tongue cancer. Preoperative MRI is recommended to assist in treatment planning for patients with this disease.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We have previously shown that tumor thickness is a significant independent prognostic factor in predicting subclinical nodal metastasis, local recurrence, and patient survival in oral tongue cancer, among other parameters such as the diameter, width, length, area, and volume of the tumor [3, 4]. Using the 3- and 9-mm divisions as cutoff points, we found that a tumor with a thickness of 3 mm or smaller has an 8% probability for subclinical nodal metastasis and a 0% probability of local recurrence and indicates a 100% probability of 5-year actuarial disease-free survival for the patient; a tumor with a thickness of more than 3 mm but less than or equal to 9 mm has a 44% probability of subclinical nodal metastasis and a 7% probability of local recurrence and indicates a 76% probability of 5-year actuarial disease-free survival for the patient; a tumor more than 9 mm thick has a 53% probability of subclinical nodal metastasis and a 24% probability of local recurrence and indicates a 66% probability of a 5-year actuarial disease-free survival for the patient. Accurate preoperative assessment of tumor thickness is thus invaluable in planning treatment of an oral tongue carcinoma.
High-resolution MRI has been used for thickness assessment and staging of colorectal, stomach, and esophageal cancers [5–7]. The involvement of tongue cancer of the mucosal epithelium, lamina propria, and muscles is clearly depicted on MRI [8], thereby making measurement of the thickness of the tongue carcinoma possible. However, the accuracy of such measurement is not well defined. We sought to evaluate the correlation between MRI-measured and histologically measured tumor thickness and to assess the accuracy of MRI in helping to stage the tumor in accordance with the 3- and 9-mm divisions.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
All patients were examined with a 1.5-T MRI system at a median of 5 days (range, 0–23 days) before glossectomy. Coronal gadolinium-enhanced T1-weighted imaging (short TR/TE, 3-mm slice thickness, and no spacing) and coronal T2-weighted imaging with fat suppression (long TR/TE, 3-mm slice thickness, and no spacing) were performed. A horizontal line joining the two tumor–mucosa junctions was drawn as a reference line. We measured the tumor thickness by drawing perpendicular lines from the reference line to the point of maximal tumor projection and invasion and then calculated the greatest radiologically determined tumor thickness by adding these two parameters.
Each glossectomy specimen was pinned on a foam board during formalin fixation to prevent shrinkage of the tissue. The preserved specimen was cut into 3-mm sections in the coronal plane and stained with H and E. The tumor border was outlined, and the image on each slide was analyzed using a computerized image analyzer (Metamorph Imaging System, Universal Imaging, Downingtown, PA). The tumor thickness was assessed in the same way as it was measured on coronal MR images. We drew a horizontal line to connect the two tumor–mucosa junctions (Fig. 1) and then drew vertical lines perpendicular to the horizontal line to measure the maximal thickness of both exophytic and endophytic parts of the tumor. We then calculated the greatest histologically determined tumor thickness by adding these two parameters.
|
The tumors were divided into three groups in accordance with the previously
proposed staging system for tumor thickness: stage I (tumor 3 mm), stage
II (> 3 mm but 9 mm), and stage III (> 9 mm). We evaluated both
MRI-determined and histologically determined tumor thicknesses. Pearson's
product-moment correlation analysis was performed on paired data between
measurements on MRI (contrast-enhanced T1-weighted spin-echo and T2-weighted
spin-echo sequences) and histologic sections. Scatterplots of tumor
thicknesses from both sources were obtained and analyzed using the linear
regression method with 95% confidence interval.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
|
We obtained scatterplots of the radiologically determined tumor thickness (contrast-enhanced T1-weighted spin-echo and T2-weighted spin-echo sequences) against the histologically determined tumor thickness (Figs. 5 and 6). Pearson's correlation analysis showed a statistically significant correlation (correlation coefficient [R] = 0.938, p < 0.0005) between tumor thickness determined with contrast-enhanced T1-weighted spin-echo images and tumor thickness determined with histologic slides. The least squares regression equation was Y = –0.336 + 0.957 X. A high correlation was also observed between thickness determined with T2-weighted spin-echo images and thickness determined with histologic slides (R = 0.941, p < 0.0005). The least squares regression equation was Y = 0.203 + 0.803 X. The Y intercept points on both contrast-enhanced T1-weighted spin-echo and T2-weighted spin-echo scatterplots were so small that the regression lines almost passed through the zero points.
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Different methods have been proposed for preoperative evaluation of the thickness of tongue tumors. Simple clinical palpation to gauge the thickness of a tongue tumor is unreliable, inaccurate, and inconsistent. Satisfactory results for reliable evaluation of staging and tumor thickness for oral tongue cancer were reported with use of sonography [12, 13]. However, the results are not easily replicable because of nonvisualization of superficial lesions and reduced accuracy for evaluation of the extension of large tumors. Excessive pain experienced by patients during sonographic examination also renders accurate placement of the sonographic probe difficult unless the examination is performed with the patients under general anesthesia. Interpretation of sonographic results is also more operator-dependent.
MRI has been shown to satisfactorily depict oral tongue carcinoma because the tumor has higher signal intensity than the normal tongue tissue [8, 14–17]. The use of gadopentetate dimeglumine and high-resolution MRI further improves tumor visualization. The structures of normal tongue tissue and the oral tongue carcinoma can be clearly shown, with good differentiation between the tumor and the mucosal epithelium, lamina propria, and tongue muscles.
We found that MRI can also be used to satisfactorily measure thickness of oral tongue carcinomas. The radiologically determined thickness measured on both T1- and T2-weighted images had a strong positive correlation with the histologically determined thickness. Although both T1- and T2-weighted–determined tumor thicknesses strongly correlated with the histologically determined thickness, the mean thickness of the tumor measured on T1-weighted images was 0.8 mm larger than the histologically determined thickness, whereas the tumor thickness measured on the T2-weighted image was 2 mm larger than the histologically determined thickness.
The possible cause of the discrepancy between the histologic and radiologic thicknesses may be due to the shrinkage and distortion of glossectomy specimen during formalin fixation despite the immediate stretching and pinning of specimen on a foam board [18]. The larger discrepancy between the thickness measured on T2-weighted spin-echo MR images and that measured on histologic sections is probably due to fact that on the T2-weighted images, the signals of the peritoneal inflammation and edema show higher intensity than on the T1-weighted images.
The overestimation of the T2-weighted tumor thickness caused the lower rate of concordance between the T2-weighted spin-echo findings and the histologic findings, although the correlation remained strong. The T1-weighted tumor thickness, however, had a higher rate of concordance (83%) with the histologic thickness than did the T2-weighted tumor thickness (56%) in the assessment of tumor thickness and staging using the 3- and 9-mm cutoff values. Therefore, we recommend T1-weighted imaging for assessing preoperative tumor thickness and staging and for planning of case management.
Accurate and consistent assessment of tumor thickness is essential for
optimal preoperative planning. In our previous study, we found that the risk
of subclinical nodal metastasis for patients with a tumor up to 3-mm thick was
8% [4]. Partial glossectomy
alone achieved nearly 100% local control of the disease and patient survival.
Stage I ( 3 mm) tumors were upstaged on both contrast-enhanced T1-weighted
spin-echo and T2-weighted spin-echo images. Because these tumors are so small,
even an artifact of 1 mm may contribute significantly to tumor upstaging.
For tumors thicker than 3 mm but less than or equal to 9 mm, the risk of subclinical nodal metastasis was 44%; elective neck dissection may be considered to improve nodal control. The high risk of local or regional failure in the tumor thicker than 9 mm may necessitate postoperative adjuvant radiotherapy to both the oral cavity and the neck. Elective neck dissection is unnecessary for patients in this stage because of the effectiveness of external radiotherapy in nodal control. Thus, accurate differentiation between stage II and stage III tumors is essential for optimal treatment planning.
On contrast-enhanced T1-weighted spin-echo images, the stage 2 and stage 3 tumors correlated reasonably well with the stages on the histologic sections. T2-weighted spin-echo images showed the high signal intensity of peritumoral inflammation and edema of the adjacent tongue tissue, which led to overestimation of tumor thickness. Six tumors judged to be stage II on the basis of the histologic sections would have been upstaged on the basis of the T2-weighted spin-echo images. Contrast-enhanced T1-weighted spin-echo imaging study is therefore more accurate for staging oral tongue cancer.
Despite the lower accuracy of T2-weighted spin-echo images compared with the accuracy of the contrast-enhanced T1-weighted spin-echo images for staging, the images from the two sequences have complementary qualities. T2-weighted spin-echo imaging is sensitive in the detection of superficial oral tongue cancer, but caution is needed in interpretation of findings if overestimation of tumor thickness is to be avoided. High-resolution MRI with fat-suppression techniques can further improve tumor assessment [19–21]. High correlations were observed between the tumor thickness determined radiologically (using either the contrast-enhanced T1-weighted spin-echo or the T2-weighted spin-echo images) and the tumor thickness determined from the histologic slides.
MRI allows good visualization of oral cancer and is the imaging technique of choice for studying head and neck tumors [22]. It provides better resolution of detailed architecture than either CT or sonography, especially in 3D visualization of soft-tissue lesions. MRI also assists in visualization of nodal metastasis. Our study showed the application of preoperative MRI assessment of tumor thickness in oral tongue cancer and that such MRI assessment could, in turn, translate into better prediction of prognosis and optimal preoperative treatment planning. The main drawbacks include the relatively small number of cases in our study, the inevitable shrinkage of tumor tissue during slide preparation, and the enhancement of the adjacent inflamed and edematous tongue tissues, especially on T2-weighted spin-echo images. Despite these drawbacks, preoperative MRI assessment for tongue cancer is useful. Both contrast-enhanced T1-weighted spin-echo and T2-weighted spin-echo images provide useful complementary information with which to evaluate the tumor thickness.
In conclusion, the assessment of tumor thickness on MRI before surgery can assist one in deciding whether the patient who has oral tongue cancer but no nodal involvement in the neck should be placed under observation or treated with prophylactic neck dissection. In addition, MRI also can be used to evaluate whether elective postoperative radiotherapy to improve local and nodal control is necessary.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Okura, S. Iida, T. Aikawa, T. Adachi, N. Yoshimura, T. Yamada, and M. Kogo Tumor Thickness and Paralingual Distance of Coronal MR Imaging Predicts Cervical Node Metastases in Oral Tongue Carcinoma AJNR Am. J. Neuroradiol., January 1, 2008; 29(1): 45 - 50. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
