Assessment of the performance of the O-RADS MRI score for the evaluation of adnexal masses, with technical notes

Pereira, Patrick Nunes; Yoshida, Adriana; Sarian, Luís Otavio; Barros, Ricardo Hoelz de Oliveira; Jales, Rodrigo Menezes; Derchain, Sophie

doi:10.1590/0100-3984.2021.0050

Acessibilidade / Reportar erro

Brasil

Español English

sumário « anterior atual seguinte »

Sumário

Original Article • Radiol Bras 55 (3) • May-Jun 2022 • https://doi.org/10.1590/0100-3984.2021.0050 copy

Assessment of the performance of the O-RADS MRI score for the evaluation of adnexal masses, with technical notes

Desempenho do escore O-RADS RM para avaliação de massas anexiais, com notas técnicas

Authorship SCIMAGO INSTITUTIONS RANKINGS

Abstract

Objective:

To assess the performance of the Ovarian-Adnexal Reporting and Data System Magnetic Resonance Imaging (O-RADS MRI) score in the evaluation of adnexal masses and to provide technical notes about its current MRI parameters and concepts.

Materials and Methods:

This was a prospective study of 226 patients with 287 adnexal masses (190 submitted to surgery or biopsy and 97 followed for at least one year). We calculated the sensitivity, specificity, positive predictive value, and negative predictive value for the O-RADS MRI score, using ≥ 4 as the cutoff for malignancy. We performed a technical analysis of the main updates to the score, announced in September 2020 by the American College of Radiology, in comparison with the original (2013) version.

Results:

We found that an O-RADS MRI score of 4 or 5 was associated with malignancy of an adnexal mass, with a sensitivity of 91.11% (95% CI: 83.23-96.08), specificity of 94.92% (95% CI: 90.86-97.54), positive predictive value of 89.13% (95% CI: 81.71-93.77), negative predictive value of 95.90% (95% CI: 92.34-97.84), and overall accuracy of 93.73% (95% CI: 90.27-96.24).

Conclusion:

Our findings support the use of the O-RADS MRI score for evaluating adnexal masses, especially those considered indeterminate on ultrasound. The updates made recently to the O-RADS MRI score facilitate its interpretation and will allow its more widespread use, with no loss of diagnostic accuracy.

Keywords:
Magnetic resonance imaging; Adnexal diseases/diagnostic imaging; Ovarian neoplasms/diagnostic imaging

Score-risk category	PPV	MRI findings
1 - Normal ovaries	—	Non-ovarian lesion. Follicle (simple cyst ≤ 3 cm), corpus luteum, or hemorrhagic cyst in a premenopausal woman. Unilocular cyst with any type of fluid content (no enhancing wall or solid tissue^* * Solid tissue is defined as a lesion component that enhances and conforms to one of these morphologies: papillary projection; mural nodule; irregular septation/wall; or other larger solid portions. ).
2 - Almost certainly benign	< 0.5%	Unilocular cyst with simple or endometrial fluid content (smooth enhancing wall and no enhancing solid tissue). Lesion with lipid content^† † Minimal enhancement of Rokitansky nodules in a lipid-containing lesion does not change the classification to O-RADS MRI 4. and no enhancing solid tissue. Lesion with “dark T2/dark DWI” solid tissue (homogeneously hypointense on T2 and DWI).
3 - Low risk	≈ 5%	Unilocular cyst with proteinaceous, hemorrhagic or mucinous fluid content (smooth enhancing wall and no enhancing solid tissue). Multilocular cyst with any type of fluid content and no lipid content (smooth septa, wall enhancement, and no enhancing solid tissue). Lesion with solid tissue (excluding T2 dark/DWI dark): low-risk (type 1) time-intensity curve on DCE MRI.
4 - Intermediate risk	≈ 50%	Lesion with solid tissue (excluding T2 dark/DWI dark): intermediate-risk (type 2) time-intensity curve on DCE MRI; if DCE MRI is not feasible, score 4 is any lesion with solid tissue (excluding T2 dark/DWI dark) that is enhancing ≤ myometrium at 30-40s on non-DCE MRI. Lesion with lipid content with large volume enhancing solid tissue.
5 - High risk	≈ 90%	Lesion with solid tissue (excluding T2 dark/DWI dark): high risk time-intensity curve on DCE MRI if DCE MRI is not feasible, score 5 is any lesion with solid tissue (excluding T2 dark/DWI dark) that is enhancing > myometrium at 30-40s on non-DCE MRI. Peritoneal, mesenteric, or omental nodularity or irregular thickening, with or without ascites.

Characteristic	n (%)
Means of establishing the diagnosis
Imaging follow-up findings (≥ 1 year)	97 (33.80)
Histopathological results	190 (66.20)
Benign disease	100 (34.84)
Ovarian
Endometrioma	1 (0.35)
Benign germ cell tumor	20 (6.97)
Cystadenoma	39 (13.59)
Stromal tumor	8 (2.79)
Endometrioma/endometriosis	11 (3.83)
Ovarian torsion or necrosis	3 (1.04)
Functional non-neoplastic cysts	7 (2.44)
Non-ovarian	11 (3.83)
Malignant disease	90 (31.36)
Ovarian borderline	17 (5.92)
Serous tumor	11 (3.83)
Mucinous tumor	4 (1.39)
Seromucinous tumor	2 (0.70)
Invasive malignant	73 (25.44)
Ovarian cystadenocarcinoma	44 (15.33)
Ovarian stromal tumors	5 (1.74)
Ovarian germ cell tumors	3 (1.04)
Anaplastic tumor	1 (0.35)
Metastasis	7 (2.44)
Non-ovarian tumor	13 (4.53)

Statistic^† † The sensitivity, specificity, positive predictive value, and negative predictive value were computed for dichotomized scores: scores of 1, 2, and 3 (benign) vs. scores of 4 and 5 (malignant).	Value	95% CI
Sensitivity	91.11%	83.23-96.08
Specificity	94.92%	90.86-97.54
Positive likelihood ratio	17.95	9.78-32.94
Negative likelihood ratio	0.09	0.05-0.18
Disease prevalence	31.36%	26.03-37.07
Positive predictive value	89.13%	81.71-93.77
Negative predictive value	95.90%	92.34-97.84
Accuracy	93.73%	90.27-96.24

Result	Key imaging finding
False negative (n = 8)
Five borderline tumors	No clearly solid component in four and a type 1 time-intensity curve in one
One mucinous ovarian carcinoma	No clearly solid component
One malignant Brenner tumor	Type 1 time-intensity curve
One endometrioid ovarian carcinoma	Type 1 time-intensity curve
False positive (n = 10)
Four serous cystadenomas	Type 2 time-intensity curves
Three fibromas	Type 2 time-intensity curves
One thecoma	Type 2 time-intensity curve
One round ligament myoma	Type 2 time-intensity curve
One peritoneal inclusion cyst	Type 2 time-intensity curve

Publicação do Colégio Brasileiro de Radiologia e Diagnóstico por Imagem Av. Paulista, 37 - 7º andar - conjunto 71, 01311-902 - São Paulo - SP, Tel.: +55 11 3372-4541, Fax: 3285-1690, Fax: +55 11 3285-1690 - São Paulo - SP - Brazil
E-mail: radiologiabrasileira@cbr.org.br

Acompanhe os números deste periódico no seu leitor de RSS

[1] Correspondence: Dr. Patrick Nunes Pereira. Departamento de Tocoginecologia - FCM-Unicamp. Rua Tessália Vieira de Camargo, 126, Cidade Universitária. Campinas, SP, Brazil, 13083-887. patricknunes@gmail.com.