COLOR AND PULSED DOPPLER US AND TUMOR-MARKER CA-125 IN DIFFERENTIATION BETWEEN BENIGN AND MALIGNANT OVARIAN MASSES

Background. Colour and pulsed Doppler flow imaging have been proposed as methods that may be useful in differentiating benign from malignant ovarian masses. It was hypothesised that the detection of neovascular isation with abnormal, low-resistance blood flow peculiar to malignant tumours is possible, which is characterised with angle-independent Do ppler indices Pl and Rl (pulsatility and resistance index, respectivel y). Tumour marker CA 125 SC (serum concentration) was found to be elev ated in 80-85% of patients with serous epithelial ovarian cancel and i n a lower percentage in other ovarian cancers, with levels over 35U/ml suggestive of malignancy. In ow study we wanted to determine whether colour and pulsed Doppler US and CA 125 SC can be used to differentiat e benign from malignant ovarian masses and whether, by combining the m ethods, the results can be even improved. Methods. ovarian masses iden tified by sonography in 71 patients aged 35 years or more were confirm ed at surgery (n = 61) or endoscopy (n = 4) or followed up to resoluti on with US (n = 6). Colour and pulsed Doppler US were used to identify intratumoral areas of vascularisation and to calculate the lowest Pl and Rl for each ovarian mass. CA 125 SC were measured. Results. In 16 of 18 ovarian malignancies and 28 of 53 benign masses, areas of intrat umoral vascularisation were detected with colour Doppler US (p = 0.002 ). Pl and Rl values displayed considerable overlap between malignant a nd benign lesions and the differences were not significant. Mean CA 12 5 SC was higher, in mali,want than in benign masses (p = < 0.0001). Fo r cut-off at 35 U/ml, sensitivity (SE), specificity (SP), positive pre dictive value (PPV) and negative predictive value (NPV) for ovarian ca ncer were 83%, 74%, 79% and 75% respectively. Either CA 125 SC > 35U/m l or intratumoral colour Doppler signal was detected in all 18 patient s with ovarian cancel; but neither of them was detected in 25 patients all of whom had benign tumours. Thus combining the two methods, SE, S P, PPV and NPV for ovarian cancer were: 100%, 47%, 32% and 100% respec tively. Conclusions. Rl and Pl values thus cannot be used to different iate between benign and malignant ovarian tumours. The determination o f CA 125 serum level is useful in identifying ovarian cancer, CA 125 S C under 35U/ml, together with the lack of detectable colour flow in th e tumour; can reliably exclude ovarian malignancy (NPV = 100%).