Diagnostic Value of Pelvic MRI for Assessment of the Depth of Myometrial Invasion and Cervical Involvement in Endometrial Cancer: Comparison of New Versus Old FIGO Staging

BACKGROUND Endometrial carcinoma is a highly prevalent gynecologic malignancy. The International Federation of Gynecology and Obstetrics (FIGO) staging system underwent significant revision on 2009. Key changes in the FIGO staging system include simplification of stage I endometrial cancer and removal of cervical mucosal invasion as a separate stage. MRI is a noninvasive diagnostic method for preoperative staging of endometrial cancer. OBJECTIVES The main purpose of this study was to investigate the diagnostic efficacy of pelvic MRI in determining the depth of myometrial invasion and cervical involvement in endometrial carcinoma. The other aim was to compare the accuracy of pelvic MRI using the old and new FIGO staging systems in endometrial carcinoma. PATIENTS AND METHODS Between November 2010 and January 2012, 54 patients underwent primary surgical staging in our department due to endometrial adenocarcinoma. Pre-operative pelvic MRI was performed and MRI staging was done according to old and new FIGO staging, separately. The sensitivity, specificity, positive and negative predictive values as well as the accuracy of MRI for deep myometrial invasion and cervical infiltration were calculated. MRI accuracy was also compared for old and new FIGO staging. Pathological staging was the standard of reference. RESULTS The mean age was 53.31 (SD = 11.52) and the most common histological subtype was the endometrioid type of endometrial adenocarcinoma (90.8%). In the evaluation of deep tumoral invasion of the myometrium (> 50%), sensitivity, specificity, diagnostic accuracy and positive and negative predictive values of MRI were 82.35%, 94.59%, 90.74%, 87.5% and 92.1%, respectively. For cervical stromal involvement, these values were 54.54%, 100%, 90.74%, 100% and 89.58%, respectively. In case of cervical mucosal involvement (in old FIGO staging), the positive predictive value was only 50% and the accuracy decreased to 74.07%. Agreement between MRI and the final histology using the old and new FIGO classification was appropriate with Kappa = 0.62 and 0.72, respectively (P < 0.001). CONCLUSION Using 2009 FIGO classification increases the accuracy of pelvic MR imaging for preoperative staging of patients with early stages of endometrial cancer.