Abstract
PURPOSE: The purpose of this study was to evaluate the factors that are associated with the accuracy of magnetic resonance (MR) imaging for predicting myometrial invasion and lymph node metastasis in women with endometrial carcinoma. MATERIALS AND METHODS: We retrospectively reviewed the medical records and preoperative MR imaging reports of 128 women who had pathologically proven endometrial carcinoma. We compared the MR imaging and the histopathology findings. RESULTS: The sensitivity, specificity and accuracy for identifing any myometrial invasion (superficial or deep) were 0.81, 0.61 and 0.74, respectively; these values for deep myometrial invasion were 0.60, 0.94 and 0.86, respectively. The sensitivity, specificity and accuracy of MR imaging for detecting lymph node metastasis were 50.0%, 96.6% and 93.0%, respectively. The patients who were older, had more deliveries and a larger tumor size more frequently had incorrect prediction of deep myometrial invasion (p=0.034, p=0.044, p=0.061, respectively). A higher tumor grade, a histology other than the endometrioid type, myometrial invasion on MR findings and a larger tumor size were associated with a more frequent false-negative prediction of lymph node metastasis (p=0.018, p=0.017, p=0.002, p=0.047, respectively). A larger tumor size was also associated with more frequent false-positive results (p=0.009). CONCLUSIONS: There are several factors that make accurate assessment of myometrial invasion or lymph node metastasis difficult with using MRI; therefore, the patients with these factors should have their MR findings cautiously interpreted.