The effectiveness and limitations of the Risk of Malignancy Index in evaluation of adnexal masses

Cho, Geum Joon; Lee, Jong Hyun; Kim, Ok Kyong; Lee, Sang Hoon; Lee, Kyu Wan

Abstract

OBJECTIVE
The purpose of this study was to evaluate the effectiveness and limitations of the Risk of Malignancy Index (RMI) for discriminating malignant from benign adnexal masses.
METHODS
This study was based on 382 women who have visited Korea University Hospital for surgical exploration of adnexal masses between July 2000 and December 2003. RMI was based on menopausal status, serum CA 125 levels, and ultrasound feature. A cutoff level of 200 was chosen as the threshold for discriminating malignant from benign adnexal masses. This setup was evaluated by sensitivity, specificity, positive predictive (PPV) and negative predictive values (NPV) with respect to the ability to distinguish malignant from benign adnexal masses.
RESULTS
The sensitivity was 57.1%, specificity 93.3%, PPV 74.7% and NPV 86.3%. If stage I disease is considered "benign", the sensitivity was 95.1%, specificity 89.3%, PPV 52.7%, and NPV 99.3%. The statistical significance of RMI over other factors was proven. However, it had limitations of high false negative rate (68.4%) in stage 1 ovarian malignancy including borderline malignancy. Moreover, its relationship with RMI1, RMI2, and RMI3 had no statistical significance.
CONCLUSION
RMI is a simple, easily applicable method for the primary evaluation of patients with adnexal masses and an effective method to selectively refer to an oncological unit for further management due to its relatively high specificity observed from this study. However this method has significant limitations in stage I ovarian malignancy including borderline malignancy. Other methods should be evaluated to increase the diagnostic accuracy.