1. Lesions observed in mammography with spiculated margins, linear calcification morphology, irregular shape, segmental calcification distribution, or linear calcification distributions have high positive predictive values (81%, 81%, 73%, 74%, 68%, respectively).
2. Using the BI-RADS lexicon (Table I) , radiologists had very comparable positive predictive values for detecting carcinoma in lesions categorized as a 4 or 5.
Original Date of Publication: July 1998
This study summary is an excerpt from the book 2 Minute Medicine’s The Classics in Radiology, 1e (The Classics Series).
Study Rundown: Seeking to better standardize characterization and reporting of mammography readings, the ACR developed BI-RADS nomenclature. This study sought to evaluate the positive predictive value of the various descriptive terms BI-RADS offers for mammogram interpretation. Results showed that among lesions that would typically be followed up by physicians , several lesional features were highly associated with cancerous biopsies, with positive predictive values for carcinoma ranging from 68-81%. Utilization of BI-RADS terminology also showed different radiologists characterized lesions into various final assessment categories describing the suspicion of cancerous lesions at statistically similar rates.
Breast Imaging-Reporting and Data System (BI-RADS) terminology was developed by the American College of Radiology to standardize the description of mammographic findings and final assessment categories. The first edition was released in 1993. Subsequent editions expanded the system to breast ultrasound and MRI. Tables I–III below reproduce the assessment framework and fifth-edition (2013) lexicons summarized in the book.
Please click to read study in AJR
In-Depth [prospective cohort]: Patients at Memorial Sloan Kettering with impalpable breast lesions seen upon initial mammographic interpretation were referred for surgical tissue biopsy. Prior to surgery, 1 of 5 academic radiologists evaluated the mammographic lesion images. Lesions that had already undergone a prior biopsy were excluded from the study and image interpretation. In total, 492 lesions from 466 women (median age 52, range 27- 92) were included in the study, with 73 (16%) of the women having had breast cancer previously. Lesions were characterized using BI-RADS terminology describing mass margins and shape, and calcification morphology and distribution. The most worrisome descriptor in the radiologist’s interpretation was used to give the lesion a final classification on a scale of 1 (normal mammography) to 5 (highly suggestive of malignancy).
Carcinoma was confirmed in 225 (46%) of the lesions. Within subcategories of lesions, 50% (79/158) of noncalcified masses, 86% (12/14) of calcified masses, and 42% (134/320) of calcifications without mass were carcinomas. BI-RADS final assessments by reading radiologists placed 26% (129/492) of lesions in category 5, 72% (355/492) in category 4, and 2% (8/492) in category 3. Carcinoma was confirmed in 81% (105/129) of category 5, 34% (120/355) of category 4, and 0% (0/8) of category 3 lesions. Amongst the BI-RADS features used to assign final assessment grades to lesions, high positive predictive values were found in 5 of them: spiculated margins (81%), linear calcification morphology (81%), irregular shape (73%), segmental calcification distribution (74%), or linear calcification distribution (68%).
Table I. BI-RADS classification (fifth-edition framework).
| Category | Descriptor | Management | Cancer probability |
|---|---|---|---|
| 0 | Incomplete: need additional imaging and/or prior mammograms for comparison | Further imaging needed | N/A |
| 1 | Negative | Routine screening | 0% |
| 2 | Benign | Routine screening | 0% |
| 3 | Probably benign | Short-interval follow-up | ≤2% |
| 4 | Suspicious | Consider biopsy | 4A: >2% to ≤10% 4B: >10% to ≤50% 4C: >50% to <95% |
| 5 | Highly suggestive of malignancy | Biopsy or surgery | ≥95% |
| 6 | Known biopsy-proven malignancy | Appropriate action (e.g., surgery) | N/A |
Table II. BI-RADS fifth-edition (2013) mammography lexicon.
| Feature | Descriptors |
|---|---|
| Breast composition | Entirely fatty; scattered areas of fibroglandular density; heterogeneously dense, which may obscure masses; extremely dense, which lowers sensitivity |
| Mass shape | Oval; round; irregular |
| Mass margin | Circumscribed; obscured; microlobulated; indistinct; spiculated |
| Mass density | Fat; low; equal; high |
| Asymmetry | Asymmetry; global; focal; developing |
| Architectural distortion | Distorted parenchyma with no visible mass |
| Calcification morphology: typically benign | Skin; vascular; coarse “popcorn-like”; large rod-like; round; rim; dystrophic; milk of calcium; suture |
| Calcification morphology: suspicious | Amorphous; coarse heterogeneous; fine pleomorphic; fine linear or fine linear branching |
| Calcification distribution | Diffuse; regional; grouped; linear; segmental |
| Associated features | Skin retraction; nipple retraction; skin thickening; trabecular thickening; axillary adenopathy; architectural distortion; calcifications |
| Location | Laterality; quadrant and clock face; depth; distance from the nipple |
Table III. BI-RADS fifth-edition (2013) ultrasound lexicon.
| Feature | Descriptors |
|---|---|
| Tissue composition (screening only) | Homogeneous fat; homogeneous fibroglandular; heterogeneous |
| Mass shape | Oval; round; irregular |
| Mass margin | Circumscribed; not circumscribed (indistinct, angular, microlobulated, spiculated) |
| Mass orientation | Parallel; not parallel |
| Mass echo pattern | Anechoic; hyperechoic; complex cystic and solid; hypoechoic; isoechoic; heterogeneous |
| Mass posterior features | No features; enhancement; shadowing; combined pattern |
| Calcifications | In mass; outside mass; intraductal |
| Associated features | Architectural distortion; duct changes; skin changes (thickening, retraction); edema; vascularity (absent, internal, rim); elasticity (soft, intermediate, hard) |
| Special cases | Simple cyst; clustered microcysts; complicated cyst; mass in or on skin; foreign body including implants; intramammary or axillary lymph nodes; AVMs; Mondor disease; postsurgical fluid collection; fat necrosis |
Liberman L, Abramson A, Squires F, Glassman J, Morris E, Dershaw D. The breast imaging reporting and data system: positive predictive value of mammographic features and final assessment categories. American Journal of Roentgenology. 1998;171(1):35-40.
Additional Review:
ACR. ACR BI-RADS® Atlas – American College of Radiology [Internet]. 2016. Available from: http://www.acr.org/Quality-Safety/Resources/BIRADS.
© 2 Minute Medicine, Inc. All rights reserved. No works may be reproduced without expressed written consent from 2 Minute Medicine, Inc. Inquire about licensing here. No article should be construed as medical advice and is not intended as such by the authors or by 2 Minute Medicine, Inc.
![ABCD2 Score: Predicting Early Stroke Risk After Transient Ischemic Attack (TIA) [Classics Series]](https://www.2minutemedicine.com/wp-content/uploads/2013/05/web-cover-classics-with-logo-medicine-BW-small-jpg-750x536.jpg)




![PFO closure equivalent to medical management in prevention of recurrent stroke [PC and RESPECT trials]](https://www.2minutemedicine.com/wp-content/uploads/2013/03/Echokardiogram_von_Atriumseptumdefekt_Ostium_secundum1-75x75.jpg)
