807P - Predicting surgical outcomes in patients with advanced ovarian cancer through MRI based preoperative evaluation

Background

Ovarian cancer (OC) is the most lethal gynecologic tumor and is characterized by a widespread metastasis in abdominal cavity. Accurate preoperative evaluation is important for the specification of surgical procedures and prediction of surgical outcomes.Currently, a variety of imaging techniques can be used for preoperative evaluation of ovarian cancer such as the Computed Tomography (CT) Suidan score. However, the Magnetic resonance imaging (MRI) based preoperative evaluation was not fully explored. Here, our study explores the feasibility of MRI as a preoperative assessment in OC patients.

Methods

This is a prospective, non-randomized, single-center trial of 134 epithelial ovarian cancer (EOC) patients treated with surgical cytoreduction between 2018 and 2022. All patients underwent preoperative MRI scan. The imaging score consists of lesion size scores in 33 areas of the abdominal cavity.

Results

Complete cytoreduction (CC) was achieved in 81 (60.4%) of 134 patients. CC patients with a median MRI score of 10(4 - 44) and the median MRI score of patients who were incompletely cytoreduced (IC) was 15(5 - 48). By comparing imaging scores with those seen in surgery, we found that MRI evaluation is in good agreement with intraoperative observation. However, for areas such as the small bowel mesentery, the MRI score is lower than the intraoperative score, while for lymph node metastases the MRI score is higher than the intraoperative score. Through multi-factor analysis, we have screened clinical and imaging criteria related to IC including CA125 level, prognostic nutritional index (PNI), diaphragmatic surface of spleen lesion, gallbladder fossa lesion and small bowel mesentery lesion. Based on these criteria, we have constructed preoperative imaging scores and the ‘predictive score’ was assigned to each criterion based on its multivariate OR.

Conclusions

Our research explores the feasibility of incorporating tumor load into MRI based preoperative assessment for OC patients. Also, we identified 5 criteria associated with IC, and developed a predictive model which may be helpful in treatment planning.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.