Staging of early endometrial carcinoma: assessment with T2-weighted and gadolinium-enhanced T1-weighted MR imaging.

This study evaluated the usefulness of T2-weighted and gadolinium-enhanced T1-weighted magnetic resonance (MR) images correlated with patients' menopausal status in assessing the depth of myometrial invasion in stage I endometrial carcinoma. MR images of 46 patients with stage I endometrial carcinoma were retrospectively reviewed. Twenty-five patients were premenopausal, and 21 were postmenopausal. The staging accuracy without regard to menopausal status was 59% for T2-weighted images and 61% for gadolinium-enhanced T1-weighted images. However, when staging accuracy was evaluated separately in the premenopausal and postmenopausal patient groups, T2-weighted imaging had an accuracy of 80% in the premenopausal group and gadolinium-enhanced T1-weighted imaging had an accuracy of 81% in the postmenopausal group. Therefore, T2-weighted imaging was more accurate in premenopausal patients and gadolinium-enhanced T1-weighted imaging was more accurate in postmenopausal patients. The overall accuracy of staging with MR imaging improved to 80% when patients' menopausal status was considered. Therefore, menopausal status should be considered when T2-weighted and gadolinium-enhanced T1-weighted MR images are used to stage early endometrial carcinoma.

MR imaging analysis of heterogeneous leiomyomas of the uterus.

Thirty-six leiomyomas from the same number of patients that were heterogeneous on MR imaging were evaluated for analyzing their MR patterns and for differentiating each type of secondary changes by means of MR imaging-pathologic correlation. The tumors with a mean diameter of 9 cm could be classified into 4 patterns depending on the morphological appearance of signal intensity: speckled (n = 14); nodular (n = 11); cystic (n = 9); or indeterminate (n = 2). Speckled pattern was associated with a mild degree of hyaline or myxoid degeneration or focal necrosis. Nodular pattern was caused by necrosis or cellular leiomyoma, and cystic pattern was related to severe hyaline or myxoid degeneration or necrosis. Each type of secondary changes within leiomyomas showed distinctive MR findings, if they were severely involved. However, use of an additional contrast-enhanced study was necessary in some instances for further clarification. MR imaging has a potential in distinguishing each type of secondary changes that occur in leiomyomas. Various degenerative changes occur in approximately 65% of uterine leiomyomas, and are caused mainly by alteration in the blood supply originating from rapid growth, pregnancy, mechanical accident, and postmenopausal atrophy. These changes include hyaline, mucoid, or myxoid degeneration, calcification, cystic changes, necrosis (red degeneration), and fatty metamorphosis. It is well known that the presence of degenerative changes within leiomyomas can be predicted on MR imaging by a heterogeneous signal intensity on T2-weighted images, although clear distinction of each type of degeneration can not be made by this modality. Recently, cellular leiomyoma, one of the variants of leiomyomas, was also reported to cause heterogeneous signal intensity. However, because various other uterine tumors can also have similar signal intensity on MR imaging, further evaluation for the heterogeneous leiomyomas appears to be necessary. The purpose of our study was to analyze the patterns of heterogeneous leiomyomas and to differentiate each type of secondary changes by means of MR imaging-pathologic correlation.

Gestational trophoblastic tumors of the uterus: MR imaging--pathologic correlation.

The purpose of this study is to evaluate the MR findings of gestational trophoblastic tumors (GTTs). Ten patients who confirmed the diagnosis (4 choriocarcinomas and 6 invasive moles) constituted the basis of our study. The MR findings from these patients were prospectively correlated with the histopathology of resected tumor specimens. MR findings of 4 patients with choriocarcinomas were well-defined, hemorrhagic masses with central necrosis. In contrast, 6 patients with invasive moles showed ill-defined, permeative masses with densely enhanced solid components and tiny cystic lesions. The trophoblastic proliferation, coagulation necrosis, and molar villi had varying signal intensities on T1- and T2-weighted images. Our results suggest that MR imaging is a promising tool for noninvasive morphologic analysis of GTTs.