Gray scale and color flow Doppler characterization of uterine tumors.

The aim of this study was to investigate gray scale and color flow characteristics of a group of patients with a suspected uterine pathologic condition. One hundred and twenty-two consecutive patients at the Women's Cancer Center, University of Minnesota, undergoing transvaginal sonography and color flow Doppler imaging for suspected uterine corpus abnormality made up the study group. After gray scale morphologic assessment, color flow Doppler imaging of the tumor and uterus was performed, including the ipsilateral uterine artery. Malignant tumors were confirmed pathologically in all 35 patients who had them. In comparing patients with benign versus malignant tumors, gray scale morphologic assessment confirmed that malignant uterine tumors (31 endometrial cancers and four sarcomas) were more likely to have a thickened echoic endometrium (P=< 0.0001), be enlarged (P=0.004), to be retroverted (P=0.02), and to lack a subendometrial halo (P<0.0001). Patients with four benign and 13 malignant tumors demonstrated increased flow when assessed by CFD. The calculated sensitivity of increased color flow in predicting malignancy was 39%, with a specificity of 92%, a positive predictive value of 77%, and a negative predictive value of 71%. No difference existed between the benign and malignant groups for the systolic, diastolic, and mean velocities and for the calculated pulsatility index and resistive index in both sampled uterine and intramyometrial or tumor vessels. Using a cutoff level of positivity of 1.0 for the pulsatility index and 0.6 for the lowest obtained resistive index, the sensitivity, specificity, positive predictive value and negative predictive value were 34%, 73%, 50%, and 58% for the pulsatility index and 23%, 74%, 39%, and 56% for the resistive index. Useful clinical information was obtained from CFD in only 13 (15%) patients. Multivariate analysis failed to confirm an association between color flow findings and menopausal status or use of estrogen replacement therapy. This study confirms the clinical usefulness of transvaginal sonography in the assessment of uterine pathologic conditions and that CFD generally is unhelpful in distinguishing benign from malignant uterine tumors