EP529 Transvaginal ultrasound versus magnetic resonance imaging for preoperative assessment of myometrial invasion in patients with endometrial cancer: prospective study

Introduction/Background Endometrial cancer is the most common gynecologic malignancy in developed countries. In 1988, the International Federation of Gynecology and Obstetrics (FIGO) established that endometrial cancer should be surgically staged. However, the therapeutic role of systematic lymph node dissection is still a matter of debate in low risk endometrial cancers. Risk classification is mainly based on tumor histology, tumor grade and myometrial infiltration (MI) depth. Currently, transvaginal ultrasound (TVS) and magnetic resonance imaging (MRI) are the commonest techniques used for assessing preoperatively the depth of MI. Methodology Prospective study in the Obstetrics and Gynecology Department of a tertiary hospital. The objective is to compare the diagnostic accuracy of TVS and MRI in for detecting myometrial infiltration in endometrial carcinoma. Results Between January 2016 and January 2019, 121 patients with diagnosis of endometrial adenocarcinoma type 1 (63,6%) and 2 (36,4%) have been recruited.85.8% of the patients presented a myometrial infiltration of less than 50% while 14.2% of them presented a myometrial infiltracion of more tan 50%. Pooled sensitivity, specificity, LR+ and LR- of TVS for detecting deep myometrial infiltration were 76.5%, 73.8%, 2.92 and 0.32, respectively. Pooled sensitivity, specificity, LR+ and LR- of MRI for detecting deep myometrial infiltration were 76.5%, 87.4%, 6.07 and 0.27, respectively. Pooled sensitivity, specificity, LR+ and LR- of intraoperative anatomopathological study for detecting deep myometrial infiltration were 69.2%, 97.4%, 26.61 and 0.32, respectively. Conclusion Data available so far show that TVS has similar diagnostic performance than MRI for detecting deep myometrial invasion in women with endometrial cancer: both of them have the same sensitivity although MRI has a better specificity (no statistically significant differences: McNemar test p>0.05). Therefore, although patients recruitment has not yet been completed, it seems that TVS could be considered as good enough for being used per preoperative diagnosis in these a prioir low-risk cases. Disclosure Nothing to disclose.