Differentiation of benign and malignant adnexal masses: relative value of gray-scale, color Doppler, and spectral Doppler sonography.

Abstract
The purpose of this study was to evaluate prospectively the relative usefulness of color Doppler, spectral Doppler, and gray-scale sonography in differentiating benign from malignant adnexal masses. A total of 170 adnexal masses in 161 patients were classified prospectively as suggestive of or not suggestive of malignant tumor on the basis of gray-scale morphology, internal flow versus peripheral or no flow, and spectral Doppler pulsatility, as measured by a pulsatility index (PI) threshold of 1.0 and a resistive index (RI) threshold of 0.4. Surgical pathology revealed 123 benign masses and 46 malignant masses. One malignant mass was confirmed by cytologic evaluation of ascitic fluid. On gray-scale analysis, 46 of the 47 malignant masses were classified as suggestive of tumor, and 76 of the 123 benign masses were classified as not suggestive of tumor (sensitivity, 98%; specificity, 62%; negative predictive value [NPV], 99%; and positive predictive value [PPV], 50%). The use of internal color flow as a predictor of malignant tumor yielded a sensitivity of 77%, a specificity of 69%, an NPV of 89%, and a PPV of 49%. The PI and RI values were significantly lower (p < .0001) in malignant masses than in benign masses, although the values overlapped considerably. For a PI of less than 1.0, sensitivity was 67%, specificity was 66%, NPV was 83%, and PPV was 46%. For an RI of less than 0.4, sensitivity was 24%, specificity was 90%, NPV was 73%, and PPV was 50%. In our series, a gray-scale prediction of benignity was reliable (NPV = 99%), and a prediction of malignancy was unreliable (PPV = 50%). Internal color flow was not useful as a predictor of malignancy (PPV = 49%). Although the absence of internal or peripheral color flow suggested benignity (NPV = 94%), only 17 (16 benign) of the masses (about 10%) had no flow. Spectral Doppler analysis with RI and PI was not useful, as no reliable discriminatory value with both high sensitivity and high specificity could be found for either parameter because of the overlap in values obtained for benign and malignant masses.