Abstract
Purpose: To determine the utility of MR imaging in the evaluation of lingual carcinomas.
Material and Methods: Eleven patients with lingual carcinoma were evaluated with MR imaging including dynamic study within one week before surgery. Nine patients underwent preoperative chemotherapy or irradiation, and 2 patients had no preoperative treatment. Delineation of the tumor, the contrast between tumor and surrounding tissue, and extent of the tumor were evaluated in a blinded manner. After glossectomy, MR images were correlated with pathological findings. Statistical analysis was performed on the visual assessment ratings.
Results: Blinded evaluation suggested that dynamic and T2-weighted images (T2WI) were significantly superior to postcontrast T1WI in demonstrating the lesion. There were no significant differences between dynamic MR images and T2WIs. However, histopathological correlation showed that in patients with preoperative treatment, dynamic MR imaging demonstrated more accurately the size and extent of the residual tumor than did T2WI and postcontrast T1WI, which tended to demonstrate an area of signal abnormality that was more extensive than the true size of the residual lesion. In patients without preoperative treatment, dynamic MR imaging, T2WI, and postcontrast T1WI gave an equivalent depiction of the extent of the carcinoma.
Conclusion: In patients with no preoperative treatment, dynamic MR imaging does not appear to offer any advantage over T2WI. However, in patients with preoperative treatment, dynamic MR imaging provides a more accurate assessment of the residual tumor than do T2WI and postcontrast T1WI.