Preoperative PET/CT score can predict complete resection in advanced epithelial ovarian cancer: a prospective study.

Abstract

To assess the ability of preoperative positron emission tomography/computed tomography (PET/CT) scans to predict postoperative residual disease in advanced epithelial ovarian cancer (AEOC). Thirty-one women with suspected AEOC were enrolled in our prospective study before surgery from July 2016 to December 2017. Complete resection was determined as no residual disease (R0) after surgery. A PET/CT scan was obtained within 2 weeks before surgery in our hospital. The PET score was the sum of each score of the radiological criteria from Suidan's model. The correlations between the PET score and tumor burden and surgical complexity were evaluated by Pearson correlation analysis. T-test or Fisher's exact test was used to compare differences in the variables between the complete and incomplete resection groups. Receiver operating characteristic (ROC) curve analysis was performed to assess the accuracy of the PET score for predicting complete postoperative resection. The median [range] of PET score was 2 [0-8], and the PET score in 20 (65%) patients was less than 3. Complete resection was achieved in 11 (35.5%) patients after surgery, including 10 (90.91%) with low PET scores and only 1 (9.09%) with a high score. The PET score had a significant positive correlation with tumor burden [Eisenkop: r=0.603, P<0.001; peritoneal cancer index (PCI): r=0.522, P=0.003] but not with surgery complexity (Aletti: r=0.291, P=0.113). Patients with lower PET scores (P=0.046) and tumor burdens (Eisenkop: P=0.013; PCI: P=0.012) had higher rates of complete resection. The PET score and tumor burden were effective for predicting complete resection. The AUCPET, AUCEisenkop, and AUCPCI were 0.797 (95% CI: 0.633-0.961, P=0.01), 0.847 (95% CI: 0.707-0.988, P=0.003), and 0.811 (95% CI: 0.653-0.969, P=0.007), respectively. However, surgery complexity was not useful for assessing complete resection. The preoperative PET score can noninvasively reflect tumor burden and helps predict complete resection after surgery in AEOC patients.