BI-RADS Version 2025 Mammography Updates Part 1

Part 1 review of BI-RADS Version 2025 updates for mammography for radiology board exams. Check out the study guide for this episode by clicking here.

BI-RADS version 2025 Mammography Highlights Part 1

True or false? If tissue in one region of the breast is dense, and potentially masking an underlying cancer, then the breast density classification should be based on an assessment of this most dense region, rather than the overall amount of dense parenchymal tissue.

True. If any portion of the breast is dense, the entire breast is considered dense. Additionally, if each breast would vary in density category, the density should be assigned based on the breast that is denser. Breast density should no longer be assigned based on how many quartiles of the breast contain dense glandular tissue—that was how this used to work, but this was changed in BI-RADS 5thedition and that change is continued in version 2025.

True or false? Breast density can be variable year to year.

The answer is true. In version 2025 it is stated that breast density may vary year to year based on various factors including weight gain, or weight loss (something we are currently seeing a lot more of in the era of GLP-1 inhibitors), changes in estrogen replacement therapy, and menopause status. Therefore, breast density should be updated at each examination and is not a lifelong assessment. It is possible to move from not dense to dense to not dense depending on the aforementioned factors.

True or false? While dense breast tissue can mask a cancer from being detected on a mammogram, dense breast tissue, alone, does not increase the risk of developing breast cancer.

False. Having dense breast tissue does increase the likelihood of developing breast cancer. Per the version 2025 BI-RADS, the odds ratio of breast cancer is 0.6 for almost entirely fatty breasts, 1.4 for heterogeneously dense breasts, and 1.6 for extremely dense breasts.

BI-RADS version 2025 recognizes that synthetic 2D mammography is in clinical use. What is synthetic 2D mammography?

Synthetic mammography is when an algorithm is used to synthetically reconstruct a 2D image from the tomosynthesis acquisition rather than obtaining a separate 2D mammogram. This reduces the overall radiation exposure per view versus obtaining the combination of a 2D mammogram plus a tomosynthesis acquisition.

What are some pros and cons of synthetic mammography per BI-RADS version 2025?

Synthetic mammography can sometimes enhance or otherwise alter the appearance of calcifications. Sometimes calcifications appear more conspicuous or coarser on a synthetic 2D image versus an actual 2D mammogram. Synthetic mammography can contain artifacts that mimic calcifications which have been termed “pseudocalcifications” and these are typically not seen on the tomosynthesis slices which can help confirm artifact. Otherwise, an actual 2D image and/or magnification views can be obtained to determine whether calcifications on a synthetic mammogram image are in fact true or artifactual pseudocalcifications. Faint and/or malignant calcifications can sometimes appear less well defined on synthetic mammography.

True or false? The spatial resolution of a synthetic mammogram is lower than that of a 2D mammogram image.

True. The spatial resolution of synthetic mammography is lower than an actual 2D mammogram image, and this is one reason why faint calcifications can sometimes appear less well defined on synthetic mammography.

True or false? Calcifications that are not within the first few or last few tomosynthesis slices cannot be dermal in location.

False. Calcifications that are not within the first few or last few slices of a tomosynthesis data set can still be dermal if they are within skin that is not touching the compression paddle or detector due to the curvature of the breast. If you see calcifications within the first few or last few slices, this can confirm dermal calcifications, but if they look dermal, but are not within the first or last slices, one could get tangential views to confirm whether they are or are not within the skin.

Name three types of calcifications that are no longer included in BI-RADS version 2025.

1.     Popcorn-like calcifications. 2. Dystrophic calcifications. 3. Milk of calcium

If you see what would have previously been characterized as popcorn-like or dystrophic calcifications, how should they be described per BI-RADS 2025?

These are both now simply described as coarse calcifications.

How do coarse calcifications which are considered benign differ from coarse heterogeneous calcifications which are considered suspicious?

Coarse calcifications are typically larger, typically greater than 2-3 mm in greatest diameter, and have smoother margins than coarse heterogeneous calcifications. Coarse heterogeneous calcifications are smaller, generally 0.5 to 1 mm in greatest diameter, and tend to coalesce. Note that fine pleomorphic calcifications are typically even smaller than coarse heterogeneous calcifications and are usually less than 0.5 mm in diameter. BI-RADS 2025 does state that numerous bilateral groups of coarse heterogeneous calcifications can often be dismissed as benign due to bilaterality and multiplicity.

If you see what would have previously been characterized as milk of calcium calcifications, how should they be described per BI-RADS 2025?

These are now simply described as layering calcifications and remain typically benign. Remember that a key for layering calcifications is to see a change in morphology of the calcifications on orthogonal views, typically appearing smudgy on a CC view, and layering on a lateral view.

What is required for calcifications to be considered grouped per BI-RADS version 2025?

Grouped calcifications typically mean at least 5 calcifications within a 1 cm area or a greater number within a 2 cm span that appear discrete and non-random.

True or false? Identifying a grouping of calcifications is more important when considering need for biopsy than is determining the morphology of calcifications.

False. Morphology of individual calcifications is more important than absolute number. If morphology is suspicious, biopsy should still be performed (and by extension should be called back for additional imaging) even if the number of calcifications is fewer than 5.

What is the BI-RADS version 2025 definition of architectural distortion?

Architectural distortion is defined as thin, straight, dense and /or lucent lines radiating from a point, and/or focal retraction or straightening along the edge of the breast parenchyma with no definite mass visible.

True or false? The positive predictive value of architectural distortion seen on tomosynthesis is lower than architectural distortion seen on 2D digital mammography.

True. Tomosynthesis shows more architectural distortion than 2D digital mammography, but those seen on 2D digital mammography are more likely to be malignant. However, even though the positive predictive value of tomosynthesis-only architectural distortion is lower, tissue sampling is still generally required unless there is a history of trauma or surgery to account for the architectural distortion.

What type of asymmetry is no longer included in BI-RADS version 2025?

Developing asymmetry is no longer included. By means of justification, the new BI-RADS version 2025 states that delineation of change over time will no longer be embedded in descriptor terminology. Instead, a developing asymmetry would now be described as an asymmetry (or global or focal asymmetry) that is enlarging, becoming denser, or more conspicuous.

True or false? Normal axillary lymph nodes on mammography demonstrate a cortical thickness measuring up to 3 mm in size.

False (or perhaps indeterminant). Per BI-RADS version 2025 the 3 mm cortical thickness threshold that may be used with ultrasound is not validated for use with mammography.

True or false? Multiple dilated ducts are always considered benign.

False. Although this is a typically benign finding, especially if in patients who are pregnant or lactating, if new, asymmetric, or associated with clinical symptoms or other mammographic abnormalities, diagnostic imaging may be necessary to evaluate for malignancy.

Is a solitary dilated duct still considered suspicious in BI-RADS version 2025?

It depends. A solitary dilated duct is defined as a unilateral single or branching tubular structure that typically measures greater than 5 mm in diameter, oriented towards the nipple. In the 5th edition of BI-RADS a solitary dilated duct was considered a BI-RADS Category 4A finding, necessitating tissue sampling. In BI-RADS version 2025, it is stated that this may be considered benign if there are no associated clinical or additional imaging findings or change over time. If presenting at baseline, or if new, or associated with clinical or other mammographic abnormalities, evaluation with diagnostic imaging may be warranted to exclude malignancy. I personally have some questions with that last statement, and BI-RADS version 2025 does not state whether biopsy needs to be performed, or if this is a Category 4A lesion. All we get is “further evaluation with diagnostic imaging may be warranted to exclude malignancy”. If diagnostic. Imaging shows a fluid-filled dilated duct is this no longer a 4A lesion, but a 3, or a 2? And presumably, if the patient has symptoms, you would not be encountering a solitary dilated duct on a screening exam, but on a diagnostic imaging study already, so what then? If I can get an answer to this question, I will make sure to share it with you on a future podcast episode.