An original risk of ovarian malignancy index and its predictive value in evaluating the nature of ovarian tumour

Abstract

Background: We carried out this study to evaluate the predictive value of an original risk of ovarian malignancy index (ROMI) devised by us. Method: Our ROMI was the simple sum of points from the three-stage gradation of serum cancer antigen 125 (CA-125), data from the patient’s familial and personal history and the ultrasound characteristics of the tumour (i.e. tumour size ? 6 cm, multilocularity, tumour with ? ¼ solid areas, dense and opalescent liquid, septum or papillary vegetation ? 3 mm), ascites, bilaterality, an unclear margin with respect to the surrounding tissue and thickness of the capsule ? 3 mm. Its originality lay in the three-stage CA-125 gradation, namely < 35 U/ml (1 point), 35-129 U/ml (3 points) and ? 130 U/ml (5 points). The study group comprised 274 patients divided into a group with benign tumours (BOT) (n = 205), and a group with malignant tumours (MOT) (n = 69). Both groups were subdivided into three subgroups of ROMI ? 11 (low risk), ROMI 12-14 (unclear risk) and ROMI ? 15 (high risk). Results: The cut-off ROMI of ? 11 showed high sensitivity, specificity and area under the curve (AUC) of 0.74, 0.93 and 0.83, respectively. The cut-off ROMI of ? 14 demonstrated extremely high specificity of 0.985, but lower sensitivity and AUC, of 0.57 and 0.78, respectively. Conclusion: Our newly devised ROMI and its cut-off of ? 11 is very effective in excluding, as well as confirming, ovarian cancer.