[Preoperative determination of the structure of pelvic tumors with color-coded Doppler ultrasound and conventional transvaginal ultrasound diagnosis].

In 265 adnexal masses results of conventional transvaginal sonography and coloured doppler sonography were compared with histopathological results. Statistical evaluation was based on lowest PI and RI as well as highest peak systolic velocity from signals derived of all detectable tumour vessels. Cut-off values from actual literature and cut-off values established in Aachen (PI < 0.69, RI < 0.45) were used. Histopathologic evaluation showed 210 (79.2%) benign and 55 (20.8%) malignant ovarian tumours. A cut-off at < 0.69 for the PI led to a sensitivity of 79.6% and a specificity of 58.2%. For RI sensitivity was 66.7% and specificity was 68.7% at a cut-off at < 0.45. The sonomorphologic evaluation following a scale published by Sassone et al. [20] had a sensitivity of 85.2% and a specificity of 67.1%. There was no statistical significant difference between mean values for peak systolic velocity in benign and malignant tumours. The exclusive evaluation with Doppler sonography leads to a high percentage of misdiagnosis. A combination with sonomorphologic evaluation does not lead to an important improvement of preoperative diagnosis as there is a bright overlap between benign and malignant tumours. In contrast Doppler sonography might add important information in the preoperative evaluation of early ovarian cancer.