Are we ready to stratify BI-RADS 4 MRI lesions?

The role of subdivision of category 4 in the American College of Radiology (ACR) Breast Imaging Reporting and Data System (BI-RADS), with a malignancy risk ranging from 2% to 95%, is well-established in mammography and ultrasound. Subdividing BI-RADS category 4 allows more precise risk stratification of breast lesions that are suspicious for malignancy, promoting better understanding and multidisciplinary communication, as well as facilitating the assessment of the radiological-histopathological correlation after biopsy and contributing to quality control audits(¹).

Although not formally defined for use in breast magnetic resonance imaging (MRI) in the ACR BI-RADS, the benefits of subdividing category 4 could be even more significant for MRI. Given the high cost and limited availability of MRI-guided biopsy, enhancing risk stratification in category 4 might be crucial, especially for suspicious imaging changes on MRI that are not identified on mammography or ultrasound. For lesions with low malignancy risk, defined within category 4a, clinical follow-up could be considered as an alternative to MRI-guided biopsy.

In the article “Are we ready to stratify BI-RADS 4 lesions observed on magnetic resonance imaging? A real-world noninferiority/equivalence analysis”, published in the current issue of Radiologia Brasileira, Maltez de Almeida et al.(²), unlike the authors of previous retrospective studies of the topic, presented results of a classification conducted during routine hospital practice, translating to practical applicability. Radiologists participating in the study had access to previous examinations and clinical data, allowing them to incorporate, albeit subjectively, mammographic and ultrasonographic findings into their decisions, which could have facilitated the stratification.

The Maltez de Almeida et al.(²) study encompassed screening and diagnostic breast MRI examinations, with a total of 419 suspicious breast lesions, classified as 4a, 4b, or 4c according to ACR BI-RADS descriptors, which were divided into minor, intermediate, and major findings, achieving positive predictive values (PPVs) of 14.2%, 41.2%, and 77.2%, respectively, with statistical equivalence/noninferiority only for categories 4b and 4c.

Despite the relevance of the topic, there is a scarcity of studies on the stratification of BI-RADS category 4, including retrospective studies and studies that are methodologically heterogeneous. The image characteristics evaluated in the definition of subcategories, the number of descriptors included, the ratio between mass and non-mass lesions, and the ratio between benign and malignant lesions are factors with significant variability across studies(¹,³,⁴,⁵,⁶,⁷,⁸).

In general, the use of one or multiple image characteristics, such as spiculated margins, mass rim enhancement, segmental non-mass enhancement, and a clumped pattern or clustered ring pattern, appears to be effective in dividing breast lesions into categories 4b and 4c. However, the mere absence of these imaging findings that are highly specific for malignancy is not sufficient to define category 4a, for which the PPV shows greater variability (1.8–15.0%) in the literature(⁶). These data illustrate how challenging it can be to establish a set of criteria capable of predicting a malignancy risk < 10% for patients undergoing breast MRI, who commonly present some additional risk factor for breast cancer and therefore have a higher pre-test probability of the disease.

Studies with extensive sampling and evaluation of additional parameters, such as T2 signal intensity and diffusion with apparent diffusion coefficient values, could help refine the criteria for subdividing BI-RADS category 4 for MRI(⁴,⁵,⁹). Ultrafast sequences and their quantitative parameters (e.g., maximum slope, initial enhancement rate, and time between arterial and venous enhancement) have been shown to increase specificity in breast MRI and could eventually play a complementary role(¹⁰,¹¹,¹²). The use of radiomics and deep learning in breast MRI applied to contrast-enhanced sequences and diffusion has also produced promising initial results in differentiating between benign and malignant lesions and in stratifying the risk of suspicious findings(¹³,¹⁴). In addition, correlation with other methods such as mammography and ultrasound could help increase or decrease the PPV of lesions identified on breast MRI, thus facilitating their upgrading or downgrading(¹⁵).

REFERENCES

• ¹
  D’Orsi CJ, Sickles EA, Mendelson EB, et al. ACR BI-RADS Atlas, Breast Imaging Reporting and Data System. Reston, VA, American College of Radiology; 2013.
• ²
  Maltez de Almeida JR, Bitencourt AGV, Gomes AB, et al. Are we ready to stratify BI-RADS 4 lesions observed on magnetic resonance imaging? A real-world noninferiority/equivalence analysis. Radiol Bras. 2023;56:291– 300.
• ³
  Honda M, Kataoka M, Kawaguchi K, et al. Subcategory classifications of Breast Imaging and Data System (BI-RADS) category 4 lesions on MRI. Jpn J Radiol. 2021;39:56–65.
• ⁴
  Strigel RM, Burnside ES, Elezaby M, et al. Utility of BI-RADS assessment category 4 subdivisions for screening breast MRI. AJR Am J Roentgenol. 2017;208:1392–9.
• ⁵
  Maltez de Almeida JR, Gomes AB, Barros TP, et al. Subcategorization of suspicious breast lesions (BI-RADS category 4) according to MRI criteria: role of dynamic contrast-enhanced and diffusion-weighted imaging. AJR Am J Roentgenol. 2015;205:222–31.
• ⁶
  Istomin A, Masarwah A, Okuma H, et al. A multiparametric classification system for lesions detected by breast magnetic resonance imaging. Eur J Radiol. 2020;132:109322.
• ⁷
  Li J, Zheng H, Cai W, et al. Subclassification of BI-RADS 4 magnetic resonance lesions: a systematic review and meta-analysis. J Comput Assist Tomogr. 2020;44:914–20.
• ⁸
  Liu D, Ba Z, Gao Y, et al. Subcategorization of suspicious non-mass-like enhancement lesions (BI-RADS-MRI category4). BMC Med Imaging. 2023; 23:182.
• ⁹
  Asada T, Yamada T, Kanemaki Y, et al. Grading system to categorize breast MRI using BI-RADS 5th edition: a statistical study of non-mass enhancement descriptors in terms of probability of malignancy. Jpn J Radiol. 2018; 36:200–8.
• ¹⁰
  Mann RM, Mus RD, van Zelst J, et al. A novel approach to contrast-enhanced breast magnetic resonance imaging for screening: high-resolution ultrafast dynamic imaging. Invest Radiol. 2014;49:579–85.
• ¹¹
  Abe H, Mori N, Tsuchiya K, et al. Kinetic analysis of benign and malignant breast lesions with ultrafast dynamic contrast-enhanced MRI: comparison with standard kinetic assessment. AJR Am J Roentgenol. 2016;207:1159– 66.
• ¹²
  Onishi N, Kataoka M, Kanao S, et al. Ultrafast dynamic contrast-enhanced MRI of the breast using compressed sensing: breast cancer diagnosis based on separate visualization of breast arteries and veins. J Magn Re-son Imaging. 2018;47:97–104.
• ¹³
  Cui Q, Sun L, Zhang Y, et al. Value of breast MRI omics features and clinical characteristics in Breast Imaging Reporting and Data System (BI-RADS) category 4 breast lesions: an analysis of radiomics-based diagnosis. Ann Transl Med. 2021;9:1677.
• ¹⁴
  Debbi K, Habert P, Grob A, et al. Radiomics model to classify mammary masses using breast DCE-MRI compared to the BI-RADS classification performance. Insights Imaging. 2023;14:64.
• ¹⁵
  Kolta M, Clauser P, Kapetas P, et al. Can second-look ultrasound downgrade MRI-detected lesions? A retrospective study. Eur J Radiol. 2020;127:108976.

Publication Dates

• Publication in this collection
  04 Mar 2024
• Date of issue
  Nov-Dec 2023