EP23.19: Diagnostic accuracy of ultrasound scoring systems in differentiating benign versus malignant adnexal masses

To assess the diagnostic accuracy of O-RADS, RMI 4, IOTA SR and ADNEX Rules in differentiating benign and malignant Adnexal Masses (AM) with histopathology as the reference standard. This prospective study recruited 100 women with AM visiting the gynecology department at a tertiary care centre in Northern India. Patients with functional ovarian cysts or corpus luteal cysts < 3cm were excluded. Clinical presentation and tumour markers were noted and transabdominal ultrasonography was performed by an expert radiologist. Morphological features and colour Doppler parameters were recorded to score the AM using O-RADS, RMI 4, IOTA SR and ADNEX Rules. Scoring systems were statistically analysed for sensitivity, specificity, positive and negative predictive values, likelihood ratios and diagnostic accuracy to differentiate benign and malignant histopathologies. Scores for borderline tumours were assessed separately. 100 women (mean age 42 ± 16.8 years, 25 postmenopausal) were recruited. Histopathology was reported as benign, borderline and malignant in 45,14 and 41 respectively. O-RADS had the highest sensitivity (98%, 95% CI: 87.4-99.5) followed by IOTA SR (95%, 95% CI: 83.8-98.6), RMI 4 (88%, 95% CI: 74.4-94.6) and ADNEX (85%, 95% CI: 71.5-93.1). However, ADNEX had maximum specificity (95%, 95 CI: 85.1-98.7), followed by RMI 4 (89%, 95% CI:76.5-95.1), IOTA SR (87%, 95% CI: 73.82-93.74) and O-RADS (53%, 95% CI: 39.1-67.1). Overall, ADNEX and IOTA SR had the highest and similar diagnostic accuracy (91%, 95% CI: 82.7-95.2), followed by RMI 4 (88%, 95% CI: 79.9-93.5) and O-RADS (74%, 95% CI: 64.2-82.4). Likelihood ratios [LR(+), LR (-)] of ADNEX [19.2, 0.15] and IOTA SR [7.1, 0.05] were the highest. Our findings demonstrated the ADNEX and IOTA SR models to have the highest diagnostic accuracy, followed by RMI4 and O-RADS. The established radiological scoring systems were developed using a European database. Our study comprehensively compares four scoring systems in Indian ethnicity, on various histopathologies.