In addition to summary sensitivity and specificity of 92 percent and 91 percent respectively for characterization of indeterminate adnexal lesions, the meta-analysis on pelvic magnetic resonance imaging (MRI) revealed higher summary malignancy rates for O-RADS MRI 4 and 5 lesions than predicted.
Recognizing that the diagnosis of eight to 30 percent of incidental adnexal lesions remains elusive after transvaginal ultrasound, researchers recently performed a meta-analysis to evaluate subsequent use of the Ovarian-Adnexal Reporting and Data System (O-RADS) magnetic resonance imaging (MRI) classification to characterize pelvic MRI findings for indeterminate adnexal lesions.1
For the meta-analysis, recently published in Radiology, researchers reviewed data from 12 studies with a total of 4,520 adnexal lesions from 3,731 women. The study authors found that pelvic MRI and use of the O-RADS classification had a 92 percent sensitivity and a 91 percent specificity for diagnosing adnexal lesions.2
“ … MRI interpretation through the O-RADS MRI system allows us to achieve an excellent level of performance both for the exclusion and confirmation of malignancy,” wrote lead study author Stefania Rizzo, M.D., Ph.D., who is affiliated with the Imaging Institute of Southern Switzerland and the Universita della Swizzera Italiana in Lugano, Switzerland, and colleagues.
For the different O-RADS MRI categories, Rizzo and colleagues noted malignancy rates of 0.1 percent for O-RADS MRI 2 lesions and 6 percent for O-RADS MRI 3 lesions. They also found that malignancy rates for O-RADS MRI 4 (60 percent) and O-RADS MRI 5 lesions (96 percent) were higher than respective predictions of 49 percent and 89 percent.2
(Editor’s note: For related articles, see “Transvaginal Ultrasound or MRI: Which is More Effective in Evaluating Endometrial Cancer?” or “Key Challenges with the O-RADS Ultrasound Classification System.”)
In regard to the O-RADS MRI 4 category, the study authors said different protocols for dynamic contrast-enhanced (DCE) MRI sequences can affect specificity. They noted that studies that emphasized post-contrast series at 30-second intervals had a 93 percent specificity whereas studies in which sequences were performed with 15 seconds or less temporal resolution had a 90 percent specificity.2 Rizzo and colleagues said diffusion-weighted imaging may help remedy discrepancies with use of the O-RADS MRI 4 category.
“Recent studies showed how the O-RADS MRI 4 category carries the highest rate of misinterpretation errors and how the integration of quantitative parameters from diffusion-weighted imaging could achieve a useful discrimination between subgroups of lesions with different malignancy rates, potentially establishing O-RADS MRI 4a and 4b categories,” pointed out Rizzo and colleagues.3,4
Noting that limited availability of category-specific data thwarted multi-threshold modeling, the researchers acknowledged dichotomization of MRI readings for the O-RADS MRI 4 threshold in the reviewed studies. They also suggested that the higher incidence of malignant lesions in the study may have been attributed to 10 of the 12 reviewed studies coming from specialist centers, which could limit broader extrapolation of the findings.
1. Meys EMJ, Kaijser J, Kruitwagen RFPM, et al. Subjective assessment versus ultrasound models to diagnose ovarian cancer. A systematic review and meta-analysis. Eur J Cancer. 2016;58:17-29.
2. Rizzo S, Cozzi A, Dolciami M, et al. O-RADS MRI: a systematic review and meta-analysis of diagnostic performance and category-wise malignancy rates. Radiology. 2022 Nov 22;220795. doi: 10:1148/radiol.220795. Online ahead of print.
3. Thomassin-Naggara I, Belghitti M, Milon A, Abdel Wahab C, Sadowski R, Rockall AG, EURAD study group. O-RADS MRI score: analysis of misclassified cases in a prospective multicentric European cohort. Eur Radiol. 2021;31(12):9588-9599.
4. Hottat NA, Badr DA, Van Pachterbeke C, et al. Added value of quantitative analysis of diffusion-weighted imaging in Ovarian-Adnexal Reporting and Data System Magnetic Resonance Imaging. J Magn Reson Imaging. 2022;56(1):158-170.