Introduction/Background Endometrial cancer (EC) is currently the most frequent cancer affecting the female genital system. The treatment of choice is represented by surgery, which consists of bilateral hysterectomy and adnexectomy with lymphadenectomy (L) according to risk factor for recurrence. The systematic lymphadenectomy procedure is often associated with postoperative complications, such as lymphedema, lymphocysts, lymphorrhoea. Recently, sentinel lymph node mapping have overcome the complications associated with L, though there is limited access to this tecnique. The aim of the study was to evaluate the role of radiomic analysis of pelvic adipose tissue at CT in predicting the incidence of post-oeprative complications of L.

Methodology Consecutive patients who underwent surgical treatment of endometrial cancer at Careggi University Hospital between January 2016 and Decemeber 2019 were enrolled. Only patients underwent to pelvic lymphadenectomy were enrolled. Exclusion criteria were bulky nodes at the preoperative imagings. Staging CT images were used fot the radiomic analysis; pelvic adipose tissue was identified and segmented, so the images were imported to the 3D Slicer software. Subsequently, the extractions of the three radiomics features (busyness, flatness, elongation) of the rea of interest were carried out.

Results Twenty seven patients were enrolled. Five patients developed post-operative complications. The value of Busyness, Flatness and elongation correlated with postoperative complications (p= 0.04, p= 0.021, p=0.03, respectively).

Conclusion Our preliminary study shows that radiomic might be useful to predict whether a patient will develop any complications associated with the lymphadenectomy. Consequently pre-operative imaging might be used also to select which patient benefit the most from sentinel node study instead of L.

Disclosures The authors have no conflicts of interest to declare.