Intermodality concordance of O-RADS US and MRI in patients with adnexal-ovarian lesions at St. Paul's Hospital Millennium Medical College, Addis Ababa, Ethiopia

Abstract

Abstract Background: Ovarian cancer is a common and complex disease affecting predominantly postmenopausal women worldwide. Early disease is characterized by asymptomatic or non specific symptoms; therefore, imaging plays a crucial role in accurate discrimination of benign vs. malignant adnexal masses, thereby early detection of ovarian cancer. For this US and MRI evaluation, the major imaging modalities with ORADS-US and MRI, which are widely accepted lexicons for accurate characterization of ovarian-adnexal lesions. Objective: To examine the intermodality concordance of O-RADS US and MRI among patients presenting with adnexal-ovarian lesions at St. Paul's Hospital Millennium Medical College, Addis Ababa, Ethiopia. Methods: An institutional based retrospective cross sectional study of patients record review was conducted at selected hospital in Addis Ababa, Ethiopia. Data were entered and analyzed using Statistical Package for Social Science (SPSS) version 26. Descriptive analysis was used to generate frequencies, mean and standard deviations for background characteristics. Percent agreement and Cohen’s Kappa statistic were used to quantify the level of inter-modality agreement. Additionally, Pearson's chi square test and two-tailed point-biserial correlation coefficient (ρ) were used to evaluate the association of potential factors with inter-modality agreement. Results: A total of 225 women with adnexal masses were included in this study, with a subset of 60 women having both US and MRI. Overall, the median age was 37, with an interquartile range of 29 to 47 years. The agreement between the two approaches for assigning a specific risk group was substantial (к = 0.67 with a percentage of agreement of 75% and weighted к = 0.79 with a percentage agreement of 91.7%). No statistically significant association of menopausal status (χ2 = 0.043; p = 0.836) and lesion size (ρ =1.30; p=0.320) with inter-modality agreement was observed. Conclusion: US and MRI had substantial concordance for assigning a specific risk group, with minimal disagreement. While this finding deserves further research, ultrasonographic studies may offer a reasonable diagnostic alternative when the administration of MRI is not possible. Keywords: adnexal mass; ultrasound; MRI; concordance; O-RADS