Mammographic and Ultrasound Analysis of Breast Masses

Masses

The ACR BI-RADS lexicon defines a breast mass as a three-dimensional (3D) space-occupying lesion seen on at least two mammographic projections. Mass shapes are categorized as oval, round, or irregular ( Figs. 4.2 and 4.3 ; Table 4.1 ). The probability of cancer increases as the mass shape becomes more irregular.

The ACR BI-RADS lexicon defines mass margins as circumscribed (well defined or sharply defined), microlobulated, obscured by surrounding glandular tissue, indistinct, or spiculated ( Fig. 4.4 ; see Table 4.1 ). As the mass margin becomes more spiculated, the probability of cancer increases. Masses with well-circumscribed borders are more likely to be benign ( Fig. 4.5 ; Video 4.2 ). Sharply marginated borders indicate no invasion of the surrounding tissue; few cancers have smooth, well-circumscribed borders. An obscured mass has a border hidden by overlapping adjacent fibroglandular tissue, and that border cannot be assessed ( Fig. 4.6 ). Microlobulated masses have small border undulations, like petals on a flower, and are more worrisome for cancer than are masses with circumscribed margins ( Fig. 4.7 ). An indistinct mass has a margin that can be seen but is fuzzy. Indistinct margins are worrisome for carcinoma because the fuzzy border suggests tumor infiltration of surrounding tissue ( Fig. 4.8 ). Finally, spiculated masses are characterized by thin lines radiating from the central portion of the mass and are especially worrisome for cancer ( Fig. 4.9 ; ). When caused by cancer, mass spiculations are caused by productive tumor fibrosis (desmoplastic reaction) or actual tumor infiltration. Tomosynthesis and the synthesized mammograms reconstructed from the tomosynthesis slices are especially sensitive for detecting spiculated masses ( Fig. 4.10 ; ).

Mass density describes the mass whiteness compared with an equal volume of fibroglandular tissue ( Figs. 4.11 and 4.12 ; see Table 4.1 ).

High-density masses are whiter than fibroglandular tissue, and low-density masses are darker than fibroglandular tissue. High-density masses are especially worrisome for cancer, because they may contain cells with a higher atomic number than normal glandular tissue and fat. Low-density masses and masses with density equal to that of surrounding fibroglandular tissue are less worrisome for cancer. However, low-density cancers, such as mucinous cancers, do exist and mimic breast cysts. These cancers are low density because they contain mucin, which is fluid density.

Fat-containing masses on 2D mammography are almost always benign, except for the rare liposarcoma or tumors surrounding fat on tomosynthesis. Fat-containing masses include lymph nodes, oil cysts (see Fig. 4.12D ), hamartomas, and fat necrosis, all of which are benign. However, tomosynthesis may show fat in both benign and malignant masses, and the fat seen on tomosynthesis may not be evident on 2D mammography ( ). Cancers may appear to contain fat if the cancer shape is very irregular and has trapped fat in between arms of the tumor. This means that the old mammographic rule that masses containing fat are always benign does not apply to tomosynthesis. To avoid misdiagnosis of cancers that contain fat on tomosynthesis slices, radiologists analyze the mass for suspicious margins or shapes and proceed with biopsy based on the worst mass features, even if the mass contains fat (see section: Masses Containing Fat ).

Asymmetry

The ACR BI-RADS mammography lexicon term asymmetry is used for mammography and not for US or MRI, although enhancing findings at MRI may be called symmetric or asymmetric. On mammography, asymmetries are white areas, more in one breast than in the other, and may represent asymmetric fibroglandular tissue or masses obscured by adjacent tissue. The finding must be included in the field of view on two orthogonal projections to qualify as an asymmetry. Asymmetries themselves do not qualify as masses on mammography. The ACR lexicon divides asymmetries into four categories: asymmetry, global asymmetry, focal asymmetry , and developing asymmetry ( Table 4.2 ; Fig. 4.13 ). A focal asymmetry and developing asymmetry have a higher likelihood of representing true masses, including breast cancer.

The first category of asymmetries is asymmetry , which is a small area (less than one quadrant of the breast volume) of fibroglandular-density tissue seen only on one mammographic projection. The asymmetry is either invisible or looks like normal fibroglandular tissue on the orthogonal view. Most one-view asymmetries represent overlapping tissues producing a “fake mass” or summation artifact ( Fig. 4.14 ; ).

Because some asymmetries proved to be cancer, radiologists often recalled women with asymmetries from 2D screening mammography. Breast tomosynthesis decreases asymmetry recalls from screening because tomosynthesis either shows the mass as a true finding by removing glandular tissue in front of and behind the mass, or proves that the mass is fake, comprised of superimposed tissue on contiguous slices (see Fig. 4.14 ; see ). However, not all asymmetries that persist as asymmetries/possible masses on tomosynthesis are true masses on workup. A possible mass may represent a summation artifact even on tomosynthesis, especially if there is suboptimal compression. These summation artifact asymmetries will spread out into their normal glandular components if there is sufficient compression, as with spot compression tomosynthesis ( Fig. 4.15 ; ). This means that tomosynthesis examinations require good compression to correctly demonstrate asymmetries as normal overlapping glandular tissue.

A large study from England ( ) demonstrated a tendency for higher specificity for 2D plus tomosynthesis compared with 2D alone for distortion/asymmetry because tomosynthesis showed either overlapping tissue or an underlying mass. In another study ( ), lesions on tomosynthesis were seen as a more specific and localized pattern (eg, mass or focal asymmetry rather than asymmetry) than those on 2D mammogram; cancers were more often constantly visible on tomosynthesis than on 2D mammography, and asymmetries were more often classified as focal asymmetry on tomosynthesis. Decreased recalls from screening in the author’s own practice were often caused by tomosynthesis showing overlapping tissue that could be dismissed as benign, versus true distortion/mass that might be cancer.

showed that asymmetries that represent normal, benign overlapping tissue are nonpalpable, contain no mass and no architectural distortion, and have no associated suspicious calcifications. Their study showed that asymmetries (formerly called asymmetric breast tissue) are present on mammograms in up to 3% of cases, and if benign, are stable on consecutive studies (see Fig. 4.14 ). In reviewing 8406 2D mammograms, showed that 221/8406 (3%) screens had asymmetric breast tissue and were not cancer if the asymmetry did not form a mass, was nonpalpable, and had no architectural distortion or calcifications. During the 36- to 42-month follow-up study period, 20 patients underwent excisional biopsy for clinical findings showing two breast cancer and one lymphoma cases, all of which were palpable. The remaining 17 biopsies were benign, and there were no breast cancers found in the remaining 201 patients with asymmetries. The study abstract concluded with the statement:

...an asymmetric volume of breast tissue, asymmetrically dense breast tissue, or asymmetrically prominent ducts that do not form a mass, do not contain microcalcifications, or do not produce architectural distortion should be view with concern only when associated with a palpable asymmetry, and are otherwise normal variations.

BI-RADS 2013 states that a global asymmetry is large, containing one quadrant or more of fibroglandular-like breast tissue compared with the same location in the contralateral breast, and is a real finding because it is displayed on two orthogonal projections. Global asymmetries are usually interspersed with fat and have no convex outward borders to suggest a mass. Similar to the smaller benign asymmetry, global asymmetries are benign if they are not new and have no associated architectural distortion, palpable findings, or suspicious calcifications. The nonpalpable global asymmetry is either an intrinsic normal variant ( Fig. 4.16 A ) or is caused by surgical removal of glandular tissue in the contralateral breast ( Fig. 4.16B ). If nonpalpable, the global asymmetry can be assessed as a BI-RADS 2 Benign Finding and returned to screening. However, if the finding is new, palpable, or is actually a mass instead of a global asymmetry, it may represent cancer and needs workup.

A focal asymmetry is defined as a more fibroglandular-like density in one breast compared with the other, both in a corresponding location, is seen on two orthogonal views, and is less than one quadrant in size (smaller than the global asymmetry). The focal asymmetry also lacks outward convex borders seen in masses, and may be interspersed with fat. It may be challenging to identify focal asymmetries, because comparison to the contralateral breast is especially important to identify the asymmetry on two views, and the findings must be included in the field of view on both projections. Some focal asymmetries may be dismissed as benign at screening ( Fig. 4.17 ), whereas others require workup. Scientific publications assess the focal asymmetry as a BI-RADS category 3 Probably Benign finding if called back from screening and is worked up, with a 0.5% to 1% probability of cancer if there are no masses at workup and it is stable over a 2-to 3-year mammographic follow-up period ( ).

As stated in BI-RADS 2013, focal asymmetries that are less than 1 cm are of concern because they may represent nonpalpable cancers ( Fig. 4.18 ; ). It is of particular concern if the focal asymmetry recalled from screening has associated architectural distortion or calcifications. On occasion a focal asymmetry is found to be a true mass at diagnostic mammography or targeted US and might be cancer ( Fig. 4.19 ).

A developing asymmetry is a focal asymmetry that, when compared with older mammograms, is new, larger, or more conspicuous than on prior studies (BI-RADS 2013; Fig. 4.20 ).

The 2007 study by showed that developing asymmetries were present in 0.16% (292 cases) of 180,801 screening mammograms and 0.11% (32 cases) of diagnostic mammograms. At screening mammography, 12.8% of the developing asymmetries were cancer. The Leung and Sickles study and a follow-up study by showed that the developing asymmetry in the diagnostic setting has a likelihood of 26.7% of malignancy if found in follow-up of BI-RADS category 3 lesions, if shown after a benign concordant biopsy or if developing in the first 5 years after breast conservation. Because the percentage of cancer is above the 2% Probably Benign category 3 threshold, developing asymmetries should undergo biopsy if they are not normal overlapping tissue at workup.

Occasionally, nonpuerperal (nonlactating) mastitis may show a developing asymmetry with rare associated architectural distortion ( ), which can be difficult to distinguish from breast carcinoma. Puerperal or postpartum mastitis caused by lactation also produces the same radiologic features as developing asymmetry but clinically, women present with a clinical history of fever and pain, and their mammograms commonly show breast edema with the developing asymmetry.

Architectural Distortion

Architectural distortion is defined as linear alterations of breast parenchyma pulled into a central focus, without a definite visible mass, resulting in radiating spiculations or thin lines pointing toward the center, like a star ( Fig. 4.21 ). Distortion also describes the pulling in or straightening of any edge of the glandular tissue boundary with fat (see Fig. 4.10 ). This was called the “tent sign” and attributed to Dr. Lazslo Tabar. When associated with asymmetry or calcifications, architectural distortion is even more suspicious for cancer. In contrast, architectural distortion associated with a history of surgical biopsy represents a postbiopsy scar and is benign. Because postbiopsy scars can be indicated by placement of a linear metallic marker on the skin over the postbiopsy scar, many facilities place radiopaque skin markers on skin scars to show that underlying distortion represents a benign scar. However, architectural distortion is suspicious for malignancy or radial scar if there is no history of trauma or surgery, and it requires biopsy. Architectural distortion is seen on 2D mammography and is even more exquisitely depicted on tomosynthesis. Architectural distortion seen on tomosynthesis may be invisible or subtle on 2D mammography, but still represents cancer or radial scar ( Fig. 4.22 ; Video 4.8 ).

For findings seen only on tomosynthesis and undetected by US, biopsy may be done under tomosynthesis guidance ( ). Architectural distortion may be a finding by itself or may be an associated finding (see section: Associated Features of Masses ).

Associated Features of Masses

Associated findings of masses worrisome for cancer ( Box 4.1 ) include skin or nipple retraction ( Figs. 4.23 and 4.24 ; ), skin thickening ( Fig. 4.25 ), or trabecular thickening ( Fig. 4.26 ); axillary adenopathy ( Fig. 4.27 ); architectural distortion; and calcifications ( Fig. 4.28 ).

Associated calcifications within or adjacent to a suspect mass are important for two reasons. First, suspicious, pleomorphic calcifications inside a benign-appearing mass may be the only clue that the mass is a cancer. Second, if the mass is cancer, calcifications around it may represent ductal carcinoma in situ (DCIS). Patient management includes sampling both the mass and surrounding suspicious calcifications ( Box 4.2 ). If both prove to be cancer, the surgeon will excise the extent of the suspicious calcifications to remove the cancer in its entirety (see Fig. 4.28B–C ).

At histology, DCIS constituting more than 25% of an invasive ductal cancer is called an extensive intraductal component (EIC); such a cancer is called EIC-positive (EIC ⁺ ). Because EIC tumors have an increased risk of local recurrence, breast-conserving surgery is less successful. This is one of the reasons to always look for calcifications when a suspicious mass is present.

Other important associated mammographic findings suggestive of cancer are skin thickening, which may indicate breast edema or focal tumor invasion; skin retraction or nipple retraction as a result of focal tumor tethering; axillary adenopathy indicating axillary lymph node metastases; and architectural distortion.

Ultrasound Technique and Analysis of Masses

The ACR BI-RADS US lexicon describes terms and features of breast masses that are key for the diagnosis of cancer ( Table 4.3 ). As shown by , ultrasonographic features are basically different between malignant and benign solid masses ( Box 4.3 ). Illustrations of these features are shown in Chapter 5 , with the most important US features of cancer detailed under the categories of mass shape (irregular), margin (not circumscribed), and orientation (not parallel to the chest wall).

US evaluation of breast masses begins with determining whether the mass is cystic or solid. Simple cysts are anechoic (all black inside), round or oval shaped with circumscribed margins, have an abrupt interface with surrounding tissue, have a thin posterior wall, and are enhanced through sound transmission (have a white tail on the side of the mass opposite of the transducer, like a comet tail). Simple cysts are dismissed as benign.

In contrast, solid breast masses have internal echoes and could be either malignant or benign. To determine whether a mass is cancer on US, the radiologist evaluates the mass shape, margins, and orientation. The most suspicious mass shape is irregular. The most suspicious mass margin is not circumscribed, including margins that are indistinct (including echogenic halo), angular, microlobulated, and spiculated. The most suspicious orientation is not-parallel orientation in which the long axis of the mass is perpendicular to the chest wall. Not parallel is also known as “taller than wide” by and is of concern that the mass is growing through normal tissue planes. Other suspicious descriptors are a heterogeneous or complex cystic and solid internal echo pattern, posterior acoustic shadowing , or architectural distortion of the surrounding breast tissue. The presence and location of associated calcifications are helpful in describing masses on US.

After scanning, the technologist takes representative pictures of the mass and labels the images to clarify the mass’ location in the breast. The ACR Practice Parameter for the Performance of a Breast Ultrasound Examination, Labeling, includes facility name and location, examination date, patient’s first and last name, which breast was scanned (left or right), position of the mass in terms of breast clock face or diagram annotation, the finding’s location in centimeters from the nipple, and the technologist’s initials. Other information helpful in finding the mass on subsequent studies includes the scan angle (radial or antiradial and transverse or longitudinal). The technologist captures the image without and with measuring calipers on the mass ( Box 4.4 ). It is also helpful to indicate whether the mass is palpable or nonpalpable. The ACR Practice Parameter encourages real-time scanning by the interpreter. Many facilities capture cine or moving images in addition to static images.

Benign US findings include no malignant features, oval or round shape, an abrupt circumscribed margin, intense homogeneous internal hyperechogenicity, fewer than four gentle lobulations, parallel (wider than tall) configuration (parallel to the chest wall), no posterior acoustic shadowing, and a thin echogenic capsule ( ). Because benign and malignant features in solid masses overlap, common sense plays a major role in patient management for solid masses, especially if the mass looks benign but the clinical scenario is suspicious (for example, new benign-appearing mass in a patient with a strong family history of breast cancer).

Correlating Palpable and Nonpalpable Masses on Mammography and Ultrasound

A common clinical problem is the palpable breast mass. The ACR Appropriateness Criteria for Palpable Breast Masses ( ) specifies guidelines to tailor imaging examinations for palpable masses. For example, the ACR Criteria recommends initial mammography for women aged ≥40 years old, usually with US as the second test. For pregnant or lactating women, or women aged ≤30 years old, US is the first study. Either US or mammography can be done first in women between these ages.

When US is the only study for a palpable mass, one correlates US findings with the palpable mass at real-time imaging. One method to correlate US findings to the mass is to place an examining finger or a cotton-tipped swab directly on the palpable mass and scan over the finger or cotton-tipped swab on the mass to generate a ring-down shadow. Subsequent removal of the finger or cotton-tipped swab from under the probe produces a scan of the palpable finding. Then the radiologist, technologist, and patient have no doubt that the palpable finding has been scanned, because this technique ensures that the transducer is placed directly on the palpable finding.

To correlate palpable US and mammographic findings when US is the first study, the radiologist or technologist scans the palpable mass and if there is a US-detected mass, the sonographer places a finger or cotton-tipped swab on the skin over the mass. The sonographer then marks the skin with indelible ink and places a metallic skin marker, such as a BB, over the palpable finding. Subsequently, the mammography technologist performs the mammogram. If the marker is at or near the mammographic finding, the palpable, mammographic, and US findings all correlate with each other.

To correlate nonpalpable US findings with mammographic findings when mammography is the first study, the sonographer looks at the mammogram to determine where to scan and what to expect on US. If there is a US finding that might correlate with the mammographic finding, the sonographer places a finger, cotton-tipped swab, or large unwound paper clip under the transducer so that a ring-down shadow is superimposed over the finding. The sonographer removes the transducer, marks the skin over the mass with an indelible ink marker, and places a metallic skin marker on the ink spot. The mammography technologist takes orthogonal mammographic views. The skin marker over the US finding should be in the same location as the mammographic finding on the films. It should be expected that even if the mammogram and US findings are the same, the mammographic finding might be 1 cm or more away from the skin marker on the films because the skin marker will be compressed away from the mass on the mammogram by the compression paddle.

Sometimes it is still uncertain whether a US and mammographic finding are one and the same even after US/mammography skin marking. If the patient agrees to a biopsy of the US finding, the radiologist places a metallic marker into the mass using a US-guided, percutaneously placed needle after the biopsy ( Fig. 4.29 ). Postbiopsy mammograms will show the marker in the mass if the US and mammography findings are the same. If the mass was only seen by tomosynthesis, marker placement in the mass should be confirmed by postbiopsy tomosynthesis.

Alternatively, a retractable hookwire may be placed in the mass under US. A mammogram with the wire in place will show that the US finding and the mammographic finding represent the same mass. The radiologist can subsequently remove the retractable hookwire.

The mammography and US report for a breast mass should describe if the mass is palpable; the size, shape, margin, and density of the mass; its location and associated findings; and any change from previous examinations, if known. The report should also include US finding descriptors and whether it correlates with the mammographic finding. Finally, each report that includes a mammogram should be assigned an ACR BI-RADS final assessment code indicating the level of suspicion for cancer and follow-up management recommendations ( Box 4.5 ).

Masses with Spiculated Borders and Sclerosing Features

Masses with spiculated borders and sclerosing features suggest malignancy ( Box 4.6 ). The benign (but high-risk) radial scar also appears as a spiculated mass on both 2D mammography and tomosynthesis and is a common cause of false-positive breast biopsies. On occasion, inflammatory or fibrotic lesions also may appear as spiculated masses

Cancer

Invasive Ductal Carcinoma

Invasive ductal carcinoma is the most common breast cancer and accounts for approximately 90% of all cancers. Also known as invasive ductal carcinoma not otherwise specified (NOS), invasive ductal carcinoma usually grows as a hard irregular mass in the breast ( Fig. 4.30 ; ). The classic appearance of invasive ductal carcinoma is a high-density irregular or spiculated mass, occasionally containing or associated with adjacent pleomorphic calcifications representing DCIS. On the mammogram, the mass should be about the same size and density on two orthogonal mammographic views. Spot compression magnification views may show unsuspected calcifications in or around the mass or unsuspected irregular borders.

Mammographic spiculated masses are often round, irregular, or spiculated on US and commonly produce acoustic shadowing as a result of either productive fibrosis or tumor extension. When present, acoustic spiculation looks like thin radiating lines extending from the tumor into surrounding breast structures, occasionally causing tissue distortion. In a dense white breast, the US spicules are dark against the white glandular tissue. In a fatty breast, the spicules are white against the dark fatty background.

On MRI, the usual appearance of invasive ductal cancer is a brightly enhancing irregular mass with or without spiculation; enhancement is initially rapid, with a late-phase plateau or washout curve. Rim enhancement and heterogeneous enhancement are other worrisome signs for invasive ductal cancer on MRI.

Invasive Lobular Carcinoma

Invasive lobular carcinoma (ILC) is most commonly seen as an equal-density or high-density noncalcified mass, occasionally showing spiculations or ill-defined borders ( Fig. 4.31 ; see also Fig. 4.1 ). and showed that 50% of ILC may have a density equal or lower than surrounding glandular tissue and showed that some ILC may contain lucent areas. ILC has a higher rate of bilaterality and multifocality than does invasive ductal cancer. It accounts for less than 10% of all invasive cancers, but historically it is the most difficult breast cancer to see on mammograms ( Box 4.7 ). ILC gives radiologists a bad name because it can be missed by mammography, at a rate reported by to be as high as 21%. This failure can be partly explained by the growth pattern of the carcinoma. Classically, ILC grows in single lines of tumor cells infiltrating the surrounding glandular tissue and may not produce a mass, making it difficult to see by mammography and difficult to feel by physical examination. It usually does not contain microcalcifications. It often infiltrates the breast, is often seen on only one view, and may cause only subtle distortion of the surrounding glandular tissue. When actually seen on the mammogram, ILC masses are often of equal or higher density than fibroglandular tissue and are seen because of the mass itself or its effect on surrounding tissue, such as architectural distortion and straightening of Cooper’s ligaments. As with any mass, distortion, flattening, and tenting of glandular tissue caused by ILC are most easily seen in locations in which Cooper’s ligaments extend out into surrounding fat, such as in the retroglandular fat or along the edge of the normal, scalloped fibroglandular tissue (see Figs. 4.10 and 4.25 ).

BOX 4.7

Features of Invasive Lobular Carcinoma

• 10% of all breast cancers

• Grows in single-cell files

• Hardest tumor to see on mammography

• Often seen on one view

• Causes mass or architectural distortion

• Calcifications not a feature

On US, ILC is a hypoechoic, irregular, spiculated, or ill-defined mass that may or may not have acoustic shadowing. When ILC becomes very large, only the acoustic shadowing may be apparent; the mass itself can be difficult to see because of its large size. On MRI, ILC looks like a spiculated mass but may have variable enhancing patterns; it can look like a mass, like a distortion of tissue, or like nodular regions connected by strands of tissue. Its enhancement kinetics can be similar to those of normal breast tissue and can thus be a cause of false-negative MRI examinations. However, ILC is detected more readily by MRI compared with mammography and US, rendering ILC a common indication for breast MRI when assessing extent of disease. In a 2015 study, showed that women with ILC were significantly associated with having other foci of cancer in the involved breast when evaluated with MRI ( p = 0.02).

Tubular Carcinoma

Tubular carcinoma is a generally slow-growing tumor with a bilateral incidence of 12% to 40%. On mammography, tubular cancer is a dense or equal-density spiculated mass with occasional microcalcifications ( Fig. 4.32 ). On occasion it may be apparent on the previous mammogram because of its slow growth. Although controversial, some pathologists believe that radial scars may be a precursor to tubular carcinoma. Generally, tubular carcinoma has a good prognosis and a lower incidence of metastases than does invasive ductal cancer. On US, tubular cancers are hypoechoic, irregular masses that occasionally produce acoustic shadowing.

Postbiopsy Scar

On mammograms, a new postbiopsy scar is round, ill-defined, and contains air and fluid; after healing, an old postbiopsy scar looks like a spiculated mass that is impossible to distinguish from cancer. In the immediate postoperative period, postbiopsy scars show air and fluid in a round or oval mass at the biopsy site, with adjacent skin thickening where the surgeon has closed the skin. Because the surgeon does not close the excised cavity with sutures, the cavity is left to fill with fluid (seroma) that may contain blood. The fluid is resorbed to varying degrees over time, with fluid occasionally persisting for several years as a round or oval mass in the biopsy site. In other cases the fluid resorbs completely, and the surrounding glandular tissue is drawn into a central dense nidus of scar tissue. As a result, the mammogram shows a centrally dense spiculated mass (the scar) with straightening of the surrounding Cooper’s ligaments and indrawing of normal glandular tissue, simulating a spiculated breast cancer ( Fig. 4.33 ; ). In some patients, no dense central nidus occurs, and the scar appears as a focal architectural distortion with radiating thin white lines from a central point. On US, a postbiopsy scar is a hypoechoic mass with acoustic spiculation and shadowing, similar to cancer. There should be thickening of the patient’s skin where the skin scar lies and distortion of subcutaneous tissue extending from the patient’s skin scar, down the plane of the incision, and finally leading to the postbiopsy scar.

The postbiopsy scar is not of concern for cancer if it occupies a surgical site ( Box 4.8 ). To distinguish postbiopsy scars from cancer, the radiologist reviews surgical biopsy locations on the breast history form and reviews older films to see if the scar is at the same location as the resected finding. Some facilities place a radiopaque linear metallic scar marker on the patient’s skin scar to show the skin scar location on the mammogram (see Fig. 4.33 ; see ). The metallic linear scar marker should be on top of the architectural distortion thought to represent postbiopsy scar. If the mammographic scar does not correspond to a postbiopsy site, it is a mass, should be considered suspicious, and should undergo workup and biopsy ( Fig. 4.34 ). However, increased density, growth, or changing of the normally convex postbiopsy scar to a concave margin suggests breast cancer recurrence and needs workup (see Chapter 8 ).

BOX 4.8

How to Confirm Postbiopsy Scars

To determine whether a spiculated mass is a postbiopsy scar, correlate the “scar” to a linear scar marker on the skin showing the location of the previous biopsy, and correlate the scar location with the targeted finding on old prebiopsy mammograms. If the postbiopsy scar is nowhere near the linear scar marker, or if the scar has developed in a location other than the targeted, excised finding, it is not a scar and should undergo biopsy.

Radial Scar

A radial scar has nothing to do with a postbiopsy scar but is a common name for a benign proliferative breast lesion that looks like a spiculated cancerous mass or postbiopsy scar on mammograms. Both radial scars and their larger variants, called complex sclerosing lesions, may include adenosis and hyperplasia. In autopsy series, small radial scars are common but often may not be apparent on 2D mammography. The central part of a radial scar undergoes atrophy, resulting in a scarlike formation, with pulling in of the surrounding glandular tissue that produces a spiculated mass. On occasion, because of entrapment of breast ductules, the radial scar may be difficult for pathologists to distinguish from infiltrating ductal carcinoma. However, both epithelial and myoepithelial cells in benign radial scars distinguish them from breast cancer. Radial scars may contain or be associated with atypical ductal hyperplasia, atypical lobular hyperplasia, lobular carcinoma in situ, or cancer. This is one of the rationales for surgical excision. Some pathologists believe that a radial scar may be a precursor to tubular carcinoma and should be excised, although this position is controversial.

On mammography, a radial scar appears as a spiculated mass with either a dark or white central area that may or may not have associated microcalcifications ( Figs. 4.35 and 4.36 ; ). It is a myth that radial scars have dark centers in the mass on mammography and can be distinguished from breast cancers, which have white-centered masses. Scientific studies have shown that radial scars and breast cancer can both have either white or dark centers on mammograms. This means that all spiculated masses not representing a postbiopsy scar should be sampled histologically ( Box 4.9 ).

BOX 4.9

Radial Scar versus Cancer

Radial scars cannot be distinguished from cancer on mammography. Spiculated masses not representing postbiopsy scar tissue require a histologic diagnosis.

Because tomosynthesis is extremely sensitive in detecting spiculation or distortion, many radial scars are now found by tomosynthesis that were invisible on 2D mammography. Unfortunately, a radial scar cannot be distinguished from cancer even on tomosynthesis, thus it requires a biopsy and is a cause for false-positive studies.

On US, a radial scar is a hypoechoic mass, with or without acoustic shadowing (see Figs. 4.35 and 4.36 ; ).