Staging And Predicting Respectability Of Primary Advanced Ovarian Carcinoma Using Computed Tomography

Objective:To assess the values of staging and predicting respectability of primary advanced ovarian carcinoma using computed tomography(CT). Methods: Sixty-six women with a preoperative CT scan of the abdomen and pelvis and a histologic diagnosis of Stage III or IV ovarian carcinoma were identified from tumor registry databases. All CT scans were analyzed retrospectively without knowledge of the operative finding. CT staging was compared with surgical and pathological findings. Residual disease measuring≤ 2cm in maximal diameter was considered an optimal surgical result. Twenty-eight of 66 patients (42.4%) underwent optimal cytoreduction to≤ 2cm residual disease. Based on statistical probability of each factor predicting cytoreductive outcome , 11 radiographic features were selected for inclusion in the final model. Each predictive index parameter was assigned a numeric value. Sensitivity , specificity, positive predictivevalue(PPV), negative predictive value(NPV), and accuracy were calculated for each Predictive Index score of ≥1 through 7. Receiver operating characteristic(ROC) curve analysis was used to assess the ability of the model to predict surgical outcome. Statistical significance was evalutated using a x2 test. Results: Evaluation of ascite, disease of Douglas pouch, paracolic gutter deposists and hepatic metastases were better with CT, the accuracy was 89. 4%, 93. 9% and 98. 5%;the sensitivity was 84. 8%, 87. 5% and 90. 0%;the specificity was 93.9%, 96.0% and 100% .The overall accuracy of staging of advanced ovarian carcinoma by CT was 87.9%, respectively, with the staging accuracy for stage III and IV patients being 86.8% and 92. 3%. For the prediction of tumor resectablity, CT features of omental extension (spleen, stomach, or lesser sac), paracolic gutter deposists( ≥1cm), liver lesion ≥ 2cm on surface or parenchymal lesion any size, and suprarenal paraaortic lymph nodes(≥1cm) were most strongly associated with surgical outcome. Using the Predictive Index scores, a ROC curve was generated with an area under the curve =0.801+0.053 .In the final predidictive Index model, a Predictive Index score≥2 had the overall accuracy at 71.2% and identified patients undergoing suboptimal surgery with a sensitivity of 68. 4%. Thespecificity , or ability to identify patients undergoing optimal surgery, was 75. 0%. The PPV and the NPV was 78.8% and 63. 6%. The ability of this model to correctly predict surgical outcome was statistically significant(P<0. 01). Conclusion:CT has high accuracy for the staging of advanced ovarian carcinoma .This study shows that prediction of respectability by CT is moderateo The stage and the Predictive Index model may have clinical utility in guiding the management of patients with ovarian carcinoma.