Preoperative differentiation of malignant from benign ovarian tumors: the efficacy of morphology indexing and Doppler flow sonography.

OBJECTIVE The goal of this study was to determine the efficacy of morphology indexing and Doppler flow sonography as methods to predict risk of malignancy in sonographically confirmed ovarian tumors. METHODS Risk of malignancy was assessed preoperatively in 442 ovarian tumors using a new morphology index (MI) based on tumor volume and wall structure. Each tumor was assigned a score of 0 to 10 based on increasing volume and morphologic complexity. Doppler flow studies were performed on 371 of these tumors. Following morphologic evaluation, all ovarian tumors were removed surgically. RESULTS Of 315 tumors with a MI < 5 there was only 1 malignancy (a stage IA granulosa cell tumor <2 cm in diameter) whereas there were 52 malignancies in 127 tumors with a MI > or = 5. Stage of disease was as follows: stage I, 33; stage II, 6; stage III, 14. Risk of malignancy was related directly to MI score, varying from 0.3% in tumors with a MI < 5 to 84% in tumors with a MI > or = 8. A MI value of > or = 5 as indicative of malignancy was associated with the following statistical parameters: sensitivity 0.981, specificity 0.808, PPV 0.409, NPV 0.997. A pulsatility index (PI) < 1.0 as indicative of malignancy was associated with: sensitivity 0.528, specificity 0.776, PPV 0.288, NPV 0.906. A resistive index (RI) < 0.4 as indicative of malignancy was associated with: sensitivity 0.222, specificity 0.867, PPV 0.222, and NPV 0.867. The addition of Doppler flow indices to MI did not improve the accuracy of predicting malignancy. Likewise, the absence or presence of ovarian tumor blood flow was not reliable as a means to differentiate benign from malignant ovarian tumors. CONCLUSIONS Morphology indexing is an accurate and inexpensive method of differentiating benign from malignant ovarian tumors, and can be a valuable adjunct in treatment planning. The addition of Doppler flow studies did not improve diagnostic accuracy of MI.