Objective: To determine the ability of contrast-enhanced magnetic resonance imaging to predict myometrial invasion, cervical invasion, and pelvic lymph node metastasis in endometrial carcinoma and to analyze factors that lead to errors in this identification.
Design: A retrospective study.
Setting: University general hospital.
Population: A total of 167 women diagnosed with endometrial carcinoma.
Methods: All patients received a preoperative contrast-enhanced magnetic resonance imaging scan. Histopathological findings were used as the definitive diagnosis.
Main outcome measures: The results were compared with histopathological findings, factors that make accurate assessment of myometrial invasion, cervical invasion, and pelvic lymph node metastasis difficult by contrast-enhanced magnetic resonance imaging were analyzed.
Results: The sensitivity, specificity, diagnostic accuracy, positive predictive values, and negative predictive values of contrast-enhanced magnetic resonance imaging were 90.9, 91.8, 91.6, 73.2 and 97.6%, respectively, for identifying deep myometrial invasion; 84.2, 96.0, 94.6, 72.7 and 97.9%, respectively, for identifying cervical invasion; and 45.0, 91.2, 85.6, 40.9 and 92.4%, respectively, for identifying pelvic lymph node metastasis. The main causes of error in contrast-enhanced magnetic resonance imaging were myomas, cornual lesions, deep myometrial invasion, large tumor size, non-endometrioid tumor type, and lower tumor grade.
Conclusion: Contrast-enhanced magnetic resonance imaging has a high accuracy and a low tendency to produce false-negative predictive values. Gynecological oncologists should combine the imaging data and clinical information to make therapeutic decisions and avoid diagnostic errors.
Keywords: Cervical invasion; endometrial carcinoma; error analysis; magnetic resonance imaging; myometrial invasion; pelvic lymph node metastasis.
© 2014 Nordic Federation of Societies of Obstetrics and Gynecology.