Definition of a dynamic laparoscopic model for the prediction of incomplete cytoreduction in advanced epithelial ovarian cancer: Proof of a concept

Abstract

Objective To develop an updated laparoscopy-based model to predict incomplete cytoreduction (RT > 0) in advanced epithelial ovarian cancer (AEOC), after the introduction of upper abdominal surgery (UAS). Patients and methods The presence of omental cake, peritoneal extensive carcinomatosis, diaphragmatic confluent carcinomatosis, bowel infiltration, stomach and/or spleen and/or lesser omentum infiltration, and superficial liver metastases was evaluated by staging laparoscopy (S-LPS) in a consecutive series of 234 women with newly diagnosed AEOC, receiving laparotomic PDS after S-LPS. Parameters showing a specificity ≥ 75%, PPV ≥ 50%, and NPV ≥ 50% received 1 point score, with an additional one point in the presence of an accuracy of ≥ 60% in predicting incomplete cytoreduction. The overall discriminating performance of the LPS-PI was finally estimated by ROC curve analysis. Results No-gross residual disease at PDS was achieved in 135 cases (57.5%). Among them, UAS was required in 72 cases (53.3%) for a total of 112 procedures, and around 25% of these patients received bowel resection, excluding recto-sigmoid resection. We observed a very high overall agreement between S-LPS and laparotomic findings, which ranged from 74.7% for omental cake to 94.8% for stomach infiltration. At a LPS-PIV ≥ 10 the chance of achieving complete PDS was 0, and the risk of unnecessary laparotomy was 33.2%. Discriminating performance of LPS-PI was very high (AUC = 0.885). Conclusions S-LPS is confirmed as an accurate tool in the prediction of complete PDS in women with AEOC. The updated LPS-PI showed improved discriminating performance, with a lower rate of inappropriate laparotomic explorations at the established cut-off value of 10.