The Use of CT Findings to Predict Extent of Tumor at Primary Surgery for Ovarian Cancer

Abstract

The present standard of care for newly diagnosed ovarian carcinoma (OC) is primary surgical cytoreduction followed by platinum-based chemotherapy. Response to chemotherapy and survival is improved in patients with the lowest residual disease after cytoreductive surgery. A recent multicenter study identified certain subgroups of patients at very high risk for major morbidity and mortality and poor short-term outcomes following primary maximal cytoreductive surgery and adjuvant chemotherapy. Patients with high initial tumor dissemination (HTD) or stage IV disease, poor performance, or nutritional status and those 75 years or older are at high risk for serious morbidity and achieve limited survival benefit from maximal cytoreduction. The authors of the present study hypothesized that preoperative computed tomography (CT) scanning can identify patients with HTD or stage IV disease. This information when combined with preoperative risk factors such as age, poor performance, or nutritional status could be used for the initial triage and treatment planning of OC patients. This pilot study was designed to determine whether specific findings available on preoperative CT could predict extent of disease dissemination in patients with advanced OC and surgical complexity required for cytoreduction. A secondary aim was to compare sensitivity and specificity of CT for identifying disease in specific locations as found at primary exploration. Preoperative CT scan data from 1997 to 2003 for patients with advanced OC were examined for disease-related findings and compared with both the findings at primary surgery and the required surgical procedures. The χ2 test was used to assess associations. Forty-six cases were identified that met inclusion criteria. Mean patient age was 66.4 years. At the conclusion of surgery, 76% had residual disease 1 cm or less. Computed tomography findings correlated (sensitivity/specificity) as follows: diaphragm disease (48%/100%), surface liver (100%/93%), omental cake (72%/65%), any sigmoid involvement (54%/100%), ascites (44%/100%), and extrapelvic large bowel involvement (29%/91%). The strongest predictors of HTD were the presence of both diaphragm disease and omental cake; specificity was 100%, and sensitivity was 48%. With a finding of liver, extrapelvic large bowel disease, and spleen involvement on CT, there was a trend for resection at the time of surgery; P = 0.001, P = 0.06, and P = 0.06, respectively. These preliminary findings suggest that the presence of both omental cake and diaphragm disease on CT in patients with advanced OC is highly predictive of HTD and the need for complex extended surgery to achieve low residual disease. Disease at multiple sites also appears to be predictive of a higher likelihood for complex surgery. Accurate prediction of HTD with CT can optimize treatment planning and help triage patients to appropriate surgical centers or alternative primary therapy.