Imaga analysis of the distribution of intraperitoneally administered fluids in patients with ovarian cancer

Patients are selected for intraperitoneal chemotherapy based on the condition of the peritoneal cavity and the volume of residual disease, as determined by second-look surgery. Several techniques have been described to assess the distribution of intraperitoneal fluids. They rely on point, as opposed to area, measurements. We performed serial radionuclide scans of 15 patients undergoing intraperitoneal chemo for ovarian cancer. Scans were performed following intraperitoneal administration of 500-1000 ml of ⁹⁹Tc-labeled human serum albumin in normal saline. Each patient underwent three scans at least 4 weeks apart. Radionuclide scans were digitized for computerized analysis. Area in picture elements (pixels) and optical density (mass) of the ⁹⁹Tc-labeled regions were determined and analyzed for similarities in the pattern of distribution. Area analysis was used to divide the patients into two groups. Group 2 patients (n=7) had more patients with <0.5 cm of residual disease (P=.051) and no evidence of disease (P=0.1) compared with group 1 patients (n= 8). No differences in overall survival were found (P=0.21). No objective measurement of distribution of intraperitoneally administered fluids exists. Computer image analysis may be an improvement over static scoring techniques to assess non-invasive imaging studies.