COMPARISON OF US, CT, MRI IN DETECTION AND STAGING OF OV

点击翻译按钮获取中文摘要

This application is in response to RFA:CA-92-02 of the Radiologic Diagnostic Oncology Group IV entitled Ovarian Cancer and Pediatric Solid Tumors. It is responsive to ovarian cancer only. Our long term objective is to evaluate the relative roles of current and emerging imaging modalities in the diagnosis, staging and serial monitoring of ovarian cancer. Currently neither ultrasonography (US), computed tomography (CT) or magnetic resonance imaging (MRI) are widely used as screening methods for ovarian cancer. CT has been shown to carry a relatively high false negative rate for residual or recurrent disease. Magnetic resonance imaging has not been routinely used to stage ovarian cancer. The potential roles of ultrasonography, computed tomography, and magnetic resonance imaging will be determined in terms of their ability to detect and differentiate benign from malignant ovarian masses as well as their ability to stage extent of tumor by prospective trials. Using receiver operator characteristic (ROC) curves, we will compare the performance of each modality separately and then jointly. Appropriate diagnostic criteria for detection and staging of ovarian cancer will be determined as well as the optimal role of each imaging modality. We will test the hypothesis that CA-125 in combination with Color Flow Doppler US, CT or MRI can differentiate benign from malignant ovarian masses. In addition we will assess whether or not more effective cytoreductive surgery can be achieved by more accurate pre-operative staging using either optimized CT or optimized MRI imaging. The role of CT and MRI in monitoring patients with ovarian cancer after primary surgery will also be tested through sequential examinations and correlation with serial CA-125 tumor antigen and second look operations. We hypothesize that CA-125 in combination with CT and MRI may detect residual disease and thus avoid the need for a second-look surgery in most patients. Detailed surgical and pathologic correlations are proposed as measures of truth for both loco-regional extent and distant peritoneal spread of tumor. The data acquired through this prospective study will form the basis for developing a decision tree analysis for optimal imaging evaluation of ovarian cancer.