Sonographic discrimination between benign and malignant adnexal masses in premenopause

The aim of this study was to assess the diagnostic value of sonographic pattern recognition by experts, a standardized morphological scoring system, the risk malignancy index (RMI) and CA 125 assay for the preoperative assessment of ovarian lesions in premenopausal patients.

Diagnostic work-up of 1320 patients who underwent surgical exploration due to an adnexal mass at a tertiary referral center were included. We assessed the discriminative value of pattern recognition, a sonographic morphological scoring system, RMI and CA 125 by calculating sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and Cohen's kappa for each diagnostic approach while using histopathology as the reference standard.

Pattern recognition showed the highest discriminative power with an observed kappa of 0.53. Sensitivity and specificity yielded 0.76 and 0.97 respectively. Combining pattern recognition with CA 125 serum measurement in the context of a triage system diminished the diagnostic value (kappa: 0.24; sensitivity: 0.29 specificity: 0.97). For the RMI we observed a sensitivity of 0.54 and a specificity of 0.96 and estimated kappa value yielded 0.37. Omitting the CA 125 assay and using a morphological sonographic assessment system increased the kappa value to 0.45 with sensitivity and specificity observed at 0.61 and 0.97 respectively.

Expert pattern recognition was found to be the method with the highest discriminative power in assessing an adnexal mass during premenopause. Additional assessment of serum CA 125 diminished the diagnostic accuracy. Standardized morphological sonographic assessment resulted in a moderate diagnostic accuracy. Supplementing the morphological sonographic assessment with CA 125 by using the RMI algorithm did not improve the diagnostic value.