Atlas of Mammography

Chapter 3

An Approach to Mammographic Analysis

Once the mammographic images are presented to the radiologist for interpretation, an approach to the analysis and interpretation defines how lesions are identified and how they are described and managed. The assessment and description of the mammographic findings are defined in a standardized lexicon known as the Breast Imaging Reporting and Data System (BI-RADS®). BI-RADS® was developed by the American College of Radiology in the mid-1990s (1) and includes a nomenclature for lesion description and overall assessment. BI-RADS® incorporates three basic components: (a) a description of the parenchymal density, (b) specific lesion descriptions for masses, calcifications, and other findings, and (c) an overall assessment or management recommendation for the patient. The use of this lexicon is ideal: The terms are standardized, and the correct use of the terms to describe an abnormality is helpful to the radiologist to correctly manage the case. The use of the standardized terminology and the requirement for an overall assessment are also very helpful for communication of the findings to the referring physician.

As the Mammography Quality Standard Act was redefined with the publication of the final regulations in 1997 (2), so included was the requirement that every mammography report include the final assessment category for the case. Only one final assessment category can be used for the patient's study. The utilization of the final assessment category is important not only for communication of results to the referring physician and the patient, but also for tracking of outcomes and quality-assurance monitoring.

The BI-RADS® reporting includes a description of tissue composition based on the volume of glandular tissue relative to fat. The four categories of composition include (a) almost entirely fat, (b) scattered fibroglandular densities, (c) heterogeneously dense, and (d) extremely dense. The specific descriptions for masses, calcifications, and other findings will be covered later in this chapter and throughout the book.

Based on the findings for the case, a BI-RADS® final assessment category is applied. The final assessment categories are shown in Table 3.1. For screening mammography, the assessment categories may be BI-RADS® 0 (requiring additional evaluation with diagnostic mammography and/or ultrasound) or BI-RADS® 1 (negative) or 2 (benign).

If lesions are identified that could potentially be malignant, additional evaluation should be performed based on the diagnostic mammogram. BI-RADS® categories for diagnostic mammography range from 2 through 6. For BI-RADS® 2 (benign) lesions, routine mammographic follow-up is obtained. For BI-RADS® 3 (probably benign) lesions, the likelihood for malignancy is less than 2% (3), and the lesion is followed at 6-month intervals for 2 years. For BI-RADS® (suspicious) 4 lesions, the possibility for malignancy ranges from 2% to 95%, but the overall positive predictive value is approximately 20% to 25%. Therefore, biopsy is recommended for the BI-RADS® 4 category. Lesions classified as BI-RADS® 5 have a high probability of being malignant (>95%), and biopsy is necessary. Category 6 was added to the BI-RADS® classification and is used for biopsy-proven cancers that are being followed with mammography before surgery. This category may be used for patients who have a biopsy-proven cancer and are being imaged for other possible lesions, or for a patient being treated with neoadjuvant chemotherapy before surgical management.

If there are multiple lesions, the overall assessment for the case is the highest or worst category applicable to any lesion. For example, if one lesion is a BI-RADS® 3 and another lesion on the same mammogram is a BI-RADS® 4, the assessment category for the case is BI-RADS® 4, suspicious.

Lesions that are correctly analyzed and categorized as BI-RADS® 3, probably benign, have a high likelihood of actually being benign. Sickles (3) prospectively evaluated the value of periodic mammographic follow-up over a 3- to 3.5-year period for nonpalpable, mammographically detected, probably benign lesions. The frequency of probably benign interpretations in this study was 11.2%. In this study (3), 17 of 3,184 (0.5%) cases were subsequently found to be malignant, 15 of which were identified by means of a mammographic interval change. The importance of completely evaluating the lesion with diagnostic


mammography before classifying it as probably benign was emphasized in the study.

TABLE 3.1 BI-RADS® Classification

BI-RADS® Category





Needs further evaluation (Additional views, ultrasound, comparison with old films)



Routine annual screening



Routine annual screening


Probably benign

Short interval follow-up mammography



Biopsy should be considered


Highly suspicious

Appropriate action/biopsy


Proven malignancy

Appropriate action

Vizcaino et al, (4) reported a study of 13,790 women, of whom 795 (5.8%) underwent short-term follow-up for a nonpalpable, probably benign lesion. Only seven lesions (1%) changed and were biopsied. Of these, two cancers were identified (0.3%).

The specific type of lesions considered to be BI-RADS® 3 include clustered round calcifications, a noncalcified circumscribed solid mass, a focal asymmetry, multiple clusters of tiny calcifications, scattered tiny calcifications, and multiple solid circumscribed masses. A suggested algorithm to manage probably benign lesions is shown in Figure 3.1. The importance of performing a diagnostic mammogram and a complete evaluation before classifying a lesion as BI-RADS® 3 cannot be overemphasized. Rosen et al. (5) found that in 52 cancers initially called probably benign, no cases fulfilled the strict criteria for a probably benign lesion. Correct workup and assessment of a lesion to categorize it as BI-RADS® 3 should lead to the follow-up of lesions that are actually benign and—very infrequently—lesions that are malignant.

More recently, Leung and Sickles (6) studied the need for recall or further evaluation in 1,440 consecutive cases of multiple circumscribed masses identified on screening mammography. These cases were prospectively managed as benign and were followed with mammography at 1 year; any that were palpable were aspirated. Two interval cancers (0.14%) were identified, which is less than the incident rate of breast cancer in the United States. Based on this, the authors did not recommend recall for women with multiple, similar nonpalpable, circumscribed masses.

Search Patterns

Cancers are missed on mammography for a variety of reasons, including technical, perceptive, and interpretative ones. In a study by Bird et al. (7) of the reasons for missed breast cancers, the authors found that 43% of the errors were caused by the lesions being overlooked by the radiologists. Missed breast cancers were more likely to be manifested as a developing density on mammography. In a study of breast cancers that were missed by radiologists, Goergen et al. (8) found that missed cancers were significantly lower in density than detected cancers, and missed cancers were more often seen on only one of the two views.

Bilateral synchronous breast cancers do occur and are often diagnosed through mammography. However, magnetic resonance imaging (MRI) has made an important impact in this area by detecting mammographically and clinically occult cancers in the contralateral breast. In a study of 77 patients with bilateral breast cancers, 43% were found to be synchronous (9), and the remainder were metachronous. When one lesion of suspicion is identified, it is imperative that the radiologist modify the search pattern to specifically look for other foci of tumor in the same breast and for contralateral disease.

It is important for the radiologist who is interpreting mammography to develop careful and consistent search patterns that will assist one in identifying potential abnormalities. Because the mammographic appearance of the breasts is quite variable from patient to patient, it is important that films be placed as mirror images during interpretation. In this way, one will be able to identify most easily a subtle mass or asymmetry that may be the only sign of carcinoma. With the films placed as mirror images (Fig. 3.2), a systematic comparison of the various areas of each breast should be carried out.

A search pattern that emphasizes the importance of symmetry as a parameter of benignity is as follows. The mediolateral oblique (MLO) views are imaged side by side, and an overall assessment of positioning and image quality is made. Then the images are viewed overall for any obvious abnormalities. Even if an obvious finding is


identified, a meticulous review of the remainder of the images is performed. By segmenting the image into identical sections, small portions of the breasts are compared with each other. For an experienced reader, this process becomes nearly instantaneous.


Figure 3.1 Algorithm for management of a BI-RADS® 3 lesion.

Initially, image segmentation can be performed by placing an opaque overlay on the images to emphasize the segment; as one develops the skills, this function is performed subconsciously. In addition to emphasizing the review of the images for symmetry, the amount of information in a segment versus the whole breast is more easily managed by the eye-brain axis (Fig. 3.3). This technique can be performed rapidly and incorporated into the global overview that an experienced breast imager uses




when interpreting a mammogram. The same pattern is used on the craniocaudal (CC) views that are arranged side by side as mirror images of each other (Fig. 3.4).


Figure 3.2 Bilateral MLO (A) and CC (B) views of a screening mammogram. The approach to interpretation begins with viewing the images as mirror images with both MLOs together and both CCs together. Comparison of like areas of each breast is performed using a segmentation approach as shown. By focusing on segments of the breasts together, one may be better able to identify subtle areas of asymmetry or subtle masses.


Figure 3.3 Segmentation of MLO images on a patient for screening mammography shows a focal asymmetry located in the axillary tail of the left breast (arrow). Although the area is less dense than some of the surrounding parenchyma, it is clearly asymmetric in comparison with the same region of the opposite breast. Pathology showed invasive lobular carcinoma.


Figure 3.4 The initial step in evaluation of the mammogram, after determining that the quality is adequate, should be to assess the symmetry of the breasts. This can be accomplished best by orienting the images together, as mirror images of each other. The viewer should make a systematic comparison of the breasts from side to side, determining if there are any asymmetries. In A, both MLO views are shown, and there is a masslike area of asymmetry in the left lower-inner quadrant (arrow). On the CC views (B), the mass is more clearly seen as being asymmetric (arrow) in comparison with the opposite breast. This lesion was an infiltrating carcinoma.

Literature on search patterns by radiologists has shown that experienced versus nonexperienced readers are different. Mammographic search patterns for different readers vary in terms of duration of gaze, scan pattern, and detection times for abnormalities. Krupinski (10) found that true positives and false positives on mammography were associated with prolonged eye gaze duration, and that false negatives were associated with longer eye gaze durations than true negatives. Therefore, gaze duration was a useful predictor of missed lesions in mammography (10). Barrett et al. (11) showed a rapid identification of microcalcifications by experienced readers as measured by pupil direction and dilatation. Less experienced readers spend more time reviewing images and covering more image area than more experienced radiologists (10).

When a potential abnormality is perceived, an analysis ensues that includes its features, possible etiologies, and a plan for evaluation. Many lesions are benign at this point and do not require further workup. Other lesions are equivocal or suspicious and necessitate further mammographic imaging and possibly imaging with other modalities. If the lesion has various features on additional views, some of which are worrisome and others of which are often associated with benign lesions, biopsy is usually warranted. When imaging a lesion in the breast, it is best to err on the side of caution by managing the case based on the worst mammographic features (Fig. 3.5).

Selection of Appropriate Views

It is important first to correlate any clinical findings with the mammographic findings. The location of skin lesions, scars, or palpable masses should be indicated by the technologist, and the radiologist must correlate the position of such entities with the imaging findings (Fig. 3.6). A knowledge of the clinical history is also important, particularly when working up clinically suspicious lesions or palpable masses and in the defining the differential diagnosis for certain findings. For example, a history of surgery at the site of an area of spiculation could be compatible with a scar. Without the history of surgery, the same finding is of concern for malignancy (Fig. 3.7).

The mammographic examination should be tailored to the individual patient (12). Particular attention should be paid to the deep aspects of the breasts, and if breast parenchyma extends posteriorly to the edge of the film on both views, exaggerated CC views should be obtained also to evaluate the deep areas completely. In thin patients with small dense breasts, the posterior lower aspect of the breasts may not be adequately visualized on the MLO




view; an additional mediolateral (ML) view often demonstrates this posterior tissue.


Figure 3.5 HISTORY: A 65-year-old woman with a palpable left breast mass.

MAMMOGRAPHY: Left MLO (A) and CC (B) views show a round dense mass corresponding to the palpable lesion. The mass is associated with dense coarse calcifications, but its margins are highly spiculated on the magnification CC view (C).

IMPRESSION: Highly suspicious for malignancy.

HISTOPATHOLOGY: Invasive ductal carcinoma.

NOTE: When approaching a lesion, it is important to judge it by its worst features. In this case, the pattern of calcification is benign, but the margins of the mass are very malignant.


Figure 3.6 A skin lesion is evident on mammography and marked with a BB. It is important for the technologist to document observed skin lesions with a radiopaque marker. Skin lesions are often evident on mammography and may occasionally be confused with parenchymal abnormalities.


Figure 3.7 HISTORY: A 53-year-old woman who is status post–right mastectomy with reconstruction and left reduction mammoplasty. She presents for routine mammography.

MAMMOGRAPHY: Left MLO (A), ML (B), and MLO spot-magnification (C) views. The patient is status post-reduction mammoplasty, and typical postoperative changes are seen. Scars are present inferiorly and in the periareolar area, and there are areas of distortion of the architecture. Clustered microcalcifications are present superiorly and are associated with some distortion (arrows) of the parenchyma. On the magnification view (C), these calcifications are coarse and pleomorphic, and they are surrounding a fat-containing round mass. The constellation of findings is consistent with fat necrosis.

IMPRESSION: Fat necrosis, post-reduction mammoplasty.

NOTE: Clinical correlation is key in suggesting the correct diagnosis in this case. The pattern of calcifications could be mistaken for malignant pleomorphic calcifications. The history of surgery and the fat-containing region in the area of calcifications are the factors that are indicative of a benign lesion.

In the evaluation of each breast, attention should be paid to the skin thickness, the symmetry of the subcutaneous fat, and the presence of asymmetric tissue, nodules, or calcifications (13). The architecture of the breasts should be symmetrical, with the fibroglandular tissue oriented to the nipple and with Cooper's ligaments appearing as thin arcs traversing the fat. Correlation with clinical examination and the location of any scars is important in assessing the presence of an area of asymmetry or architectural distortion. Focal distortion—including linear densities oriented in a different direction from the other structures, or focal puckering in or out of the glandular


tissue—may indicate an underlying carcinoma. Carcinomas may infiltrate into the fat and parenchyma, producing thickening of Cooper's ligaments. Such involvement can produce skin thickening or retraction that may first be evident on mammography. Central carcinomas also may fix the nipple-areolar complex, which is evident to the technologist on compression of the breast as nipple retraction but which is not evident in the noncompressed state.

Additional positioning, including ML, lateromedial oblique (LMO), or rolled CC views, may be necessary to determine the presence and location of a lesion. An ML view may rotate a lesion in tangent, demonstrating distortion or nodularity obscured by glandular tissue on the MLO view. The ML view is also advantageous to demonstrate (a) the relative medial or lateral location of a lesion based on how it moves in relationship to the MLO position, (b) the posteroinferior aspect of the breast, and (c) the medioposterior aspect of the breast (Fig. 3.8). The LMO view is useful for palpable lesions deep in the inner quadrant not seen on mammography or for lesions identified on mammography deep in the medial aspect of the breast on the CC view not seen on the MLO view.

Spot compression and magnification are used to analyze mass borders, areas of distortion, and microcalcifications. Special views, such as a magnification and spot compression, have been shown to increase the specificity of mammography in one half of patients recalled from screening (14). If a mass or focal area of distortion is found, spot compression views will displace the surrounding tissue and aid in identifying the presence of an underlying lesion. Berkowitz et al. (15) found spot compression was of help in the analysis of 75 lesions in rendering them more or less suspicious than was evident on the routine view. The spot-compression view is useful to determine if (a) a focal density is superimposed tissue or a true lesion, (b) the border of a relatively circumscribed mass identified on the routine view is actually circumscribed, and (c) a focal area of architectural distortion is being produced by a spiculated lesion (Fig. 3.9).

The rolled CC view is performed to determine the relative inferior or superior position of a lesion seen on the CC view but not demonstrated on the MLO view. If the superior aspect of the breast is rolled laterally and the lesion moves laterally from its position on the CC view, it is located in the superior aspect of the breast (Fig. 3.10). If the lesion moves in the opposite direction of the roll, it is located inferiorly.

The radiologist must choose the additional views carefully in order to correctly evaluate a potential lesion. The proper choice of views will answer two questions; (a) Is there a true lesion? and (b) Where is the lesion located? Based on these considerations, further evaluation with ultrasound (Fig. 3.11) or biopsy may be considered.

Evaluation of Breast Masses

In the evaluation of a mass lesion, an assessment of its margins (16,17,18), shape, density, location, orientation, presence of a fatty halo, and size is made. Masses may be divided into four groups based on their density relative to the parenchyma: (a) fat containing, (b) low density, (c) isodense, or (d) high density. Fat-containing radiolucent and mixed-density circumscribed lesions are benign (Fig. 3.12), whereas isodense to high-density masses may be of benign or malignant origin. Benign lesions tend to be isodense or of low density, with very-well-defined margins, and surrounded by a fatty halo (Fig. 3.13), but this is certainly not diagnostic of benignancy (19). The halo sign is a fine radiolucent line that surrounds circumscribed masses and is highly predictive that the mass is benign (19). The halo sign is different from the broad band of lucency described by Gordenne and Malchair (20). The broad band is a Mach band effect creating a hyperlucent zone around malignant lesions, but this hyperlucency had an average diameter of 5 to 10 mm.

A spot compression view is important to define the borders of a seemingly circumscribed mass as either completely or partially well defined. Lack of a complete halo should warrant further investigation. The likelihood of a circumscribed mass being benign is great, but various series (21,22,23) report significant numbers of occult carcinomas presenting as at least partially defined masses. Sickles (22) found that in a series of 300 nonpalpable cancers, only one half of the noncalcified lesions were spiculated, and the remainder were well-defined or poorly defined masses or areas of architectural distortion. Swann et al. (19) found at least a partial halo sign to be associated with 25 of 1,000 breast cancers, and in 60% of these cancers, the halo was complete.

An algorithm that describes the management of a patient with a nonpalpable circumscribed mass is shown in Figure 3.14. Ultrasound plays an integral role in the assessment of a circumscribed isodense mass (24,25,26,27,28,29). If ultrasound demonstrates a cyst to correspond to a circumscribed mammographic mass (Fig. 3.15), the lesion is benign and is followed annually. If ultrasound does not confirm a cyst, mammographic follow-up is necessary. It is important that in the follow-up of such lesions, the subsequent examinations should be tailored to allow accurate comparisons with the initial study. Noncystic masses that are completely well defined and nonpalpable are considered probably benign and are followed mammographically at 6-month intervals for 2 years (Figs. 3.16 and 3.17) (30).

Although some radiologists routinely biopsy circumscribed solid masses larger than a certain size, others do not. Sickles (31) showed no significant effect of patient age or lesion size on the likelihood of cancer for nonpalpable circumscribed breast masses. Therefore, the decision to




biopsy a circumscribed nonpalpable mass should not be made based on its diameter. Palpable masses or dominant masses are usually evaluated by biopsy even if they are circumscribed. The presence of a new palpable mass implies that the mass has developed or changed, so early mammographic follow-up is not the best management plan.


Figure 3.8 HISTORY: A 53-year-old gravida 3, para 0, abortus 3 woman with a positive family history of breast cancer for screening mammography.

MAMMOGRAPHY: Left CC (A), MLO (B), ML (C), and enlarged spot-compression (D) views. The breast is dense for the age of the patient. Deep in the central aspect of the breast on the CC view (A) is a small, high-density irregular mass (arrow). On the MLO view (B), several nodular densities are seen (arrows), any of which could correspond to the lesion. An ML view (C) was obtained but again does not clearly identify the lesion. Spot-compression views were obtained in the ML position, along the plane of the various densities. The spot ML (D) view of the lowermost density revealed the lesion.

IMPRESSION: Indistinct mass at the 6 o'clock position of the left breast, highly suspicious for carcinoma.

HISTOPATHOLOGY: Well-differentiated infiltrating ductal carcinoma.


Figure 3.9 HISTORY: A 61-year-old woman for routine screening.

MAMMOGRAPHY: Left CC view (A) from 1989 and left CC view (B) from 1986. The breast is mildly glandular. In the subareolar area, there is a relatively well-circumscribed 5-mm isodense mass (A, arrow). The nodule was not present on the mammogram (B) 3 years earlier. On spot compression, (C) the mass is high density and indistinct, which are features suspicious for malignancy. Ultrasound was performed but did not demonstrate the mass. A nodule of this size and in a superficial location would probably be seen on ultrasound if it were cystic; it therefore is presumed to be solid. Because of the interval change from the prior study, the mass should be considered moderately suspicious for malignancy. In this case, the lesion was excised following a needle localization procedure.

IMPRESSION: New indistinct mass, left breast, BI-RADS® 4.

HISTOPATHOLOGY: Infiltrating ductal carcinoma, comedocarcinoma.

A lesion that is ill-defined or spiculated and in which there is no clear history of trauma to suggest hematoma or fat necrosis suggests a malignant process (Figs. 3.18,3.19,3.20,3.21). The presence of a central high density with surrounding fine spiculation is a highly suspicious appearance. Spot compression may be of help in elucidating such an appearance within dense parenchyma. Stomper et al. (23), in the evaluation of specimen films and histology of noncalcified nonpalpable carcinomas, found that gross spiculation of >2 mm was microscopically found to represent islands of neoplastic cells surrounded by dense collagenous stroma.




Figure 3.10 HISTORY: A 70-year-old woman with a palpable mass in the medial aspect of the right breast.

MAMMOGRAPHY: Right ML (A) and CC (B) views show a small mass medially (arrow) that corresponded to the palpable lesion. A second area of increased density is present centrally (arrowheads) and is associated with nipple retraction. On rolled CC medial (C) and lateral (D) views, the palpable mass (arrow) and the central indistinct mass (arrowheads) are confirmed. Both lesions are displaced eccentrically on the rolled view (moving medially on the lateral roll), indicating that they are located slightly inferiorly.

IMPRESSION: Findings suspicious for multiple carcinomas.

HISTOPATHOLOGY: Multicentric, invasive lobular carcinoma.




Figure 3.11 HISTORY: A 78-year-old woman who had been followed at 6-month intervals for 5 years for a right breast axillary tail density. She requests second opinion.

MAMMOGRAPHY: Right MLO (A) and CC (B) views show scattered fibroglandular densities. A BB marks a skin lesion. An indistinct mass is seen on the MLO view only, overlying the pectoralis muscle, but the lesion is not seen on the CC view. This lesion had been considered to be a lymph node at the outside facility and was followed at short intervals. No additional mammographic views had been obtained. Because the lesion is not seen on the CC view, its position on the transverse plane is only speculative. An ML view was obtained showing that the lesion appeared higher than on the MLO, suggesting a medial location. A directed ultrasound (C) of the right upper-inner quadrant was performed based on the suspected position of the mass. Ultrasound demonstrates that the lesion is solid and slightly irregular in contour.

IMPRESSION: Indistinct mass in the upper inner quadrant, suspicious for carcinoma.

HISTOPATHOLOGY: Medullary carcinoma.

NOTE: The error made by the initial radiologist was not obtaining additional views to locate the lesion and assuming that it was in the axillary tail. The incomplete workup led to the false assumption that the lesion could be a node. By identifying the position of the lesion with the ML view, the further workup with ultrasound is performed, identifying the mass as likely malignant.




Figure 3.12 A large, heterogeneous fat-containing mass is outlined by a very-well-defined margin with a fatty halo (arrows). The appearance of the lesion is typical of a hamartoma because of its mixed density and its well-circumscribed border.


Figure 3.13 Multiple well-circumscribed round masses are outlined by fatty haloes, suggestive of a benign nature, likely cysts. Also noted are scattered benign calcifications.

Ultrasound is used increasingly for masses that are indistinct or even somewhat spiculated. With ultrasound, the level of suspicion is further defined. In addition, sonography is an excellent method for guidance of percutaneous breast biopsy, so ultrasound is performed to determine if this method for biopsy is feasible (Figs. 3.22,3.23,3.24).

The additional findings of suspicious microcalcifications also are important in defining the likelihood of malignancy. An indistinct mass that is noncalcified may represent a breast cancer, fibrocystic change, abscess, hematoma, inflamed cyst, lymphoma, or desmoid tumor. If fine pleomorphic microcalcifications are associated with the lesion, the differential no longer includes most of the lesions, other than cancer and fibrocystic change. Also, secondary signs, such as skin thickening or retraction, nipple retraction, disruption of Cooper's ligaments, or fixation of the pectoralis muscle are all very suspicious features that may be associated with a malignant mass. Biopsy is usually necessary in the evaluation of a poorly defined or spiculated lesion identified on two views.




Figure 3.14 Algorithm for management of a nonpalpable circumscribed mass.


Figure 3.15 HISTORY: A 69-year-old woman for screening mammography.

MAMMOGRAPHY: Right MLO (A) and CC (B) views show an isodense oval mass at 8 o'clock. On spot compression (C), the borders of the lesion are very well defined. Ultrasound (D) shows the mass to be a small simple cyst.

IMPRESSION: Simple cyst.




Figure 3.16 Well-defined, isodense round mass. If this mass is seen on a baseline mammogram, ultrasound is performed. If the mass is not a cyst and not suspicious on ultrasound, then it is classified as a probably benign lesion (BI-RADS® 3) and is followed at an early interval.


Figure 3.17 HISTORY: A 42-year-old gravida 2, para 2 woman for screening mammography.

MAMMOGRAPHY: Left MLO (A) and CC (B) views show the breast to be heterogeneously dense. There is a lobular isodense mass with circumscribed margins in the left breast at 9 o'clock. Mammographic features suggest that this is most likely benign. On ultrasound (C), the very-well-defined thin margins of the solid hypoechoic mass are seen. The features are most suggestive of a fibroadenoma. Because the lesion was not palpable, and because of its benign features, 6-month follow-up was performed.

IMPRESSION: Well-defined solid mass, probably benign, BI-RADS® 3; recommend 6-month follow-up.

NOTE: This lesion meets the criteria for a probably benign mass. Its borders are very defined. It is not palpable or enlarging. The sonographic features suggest a benign etiology. Therefore, the likelihood that it is benign is at least 98%, and early mammographic follow-up is performed.




Figure 3.18 HISTORY: A 57-year-old woman for screening.

MAMMOGRAPHY: Left MLO (A) and CC (B) views show a focal asymmetry located posteriorly. On spot compression (C), this is a true mass that is lobulated with indistinct margins. In this case, spot compression helps with assessment of the patient in predicting the likelihood of malignancy.

IMPRESSION: Indistinct mass, suspicious for carcinoma.

HISTOPATHOLOGY: Invasive ductal carcinoma.

Evaluation of Asymmetric Densities

In analyzing all ill-defined area of asymmetric tissue without other associated findings, one should first determine if, in fact, the asymmetry is present on two views or if it merely represents overlapping glandular tissue on one view. Global asymmetry is a large area of parenchymal density in one breast compared with the other, and this finding usually represents asymmetric breast tissue. A focal asymmetry is seen on two views and has a similar configuration, but it lacks the borders and conspicuity of a mass. Architectural distortion is focal spiculation and disruption of the orientation of the normal parenchyma without a central mass.

Focal asymmetries are often benign but occasionally may be a sign of breast cancer. Kopans et al. (32) found that asymmetric glandularity occurred in 3% of patients on mammography; carcinoma was found in only three patients with breast asymmetry without other signs of malignancy but in whom there were concurrent palpable findings. Asymmetries of concern are those that are changing or enlarging or new, those that are palpable, and those that are associated with other findings, such as microcalcifications or architectural distortion. If the patient has a scar or has had trauma to the area, the density may be fat necrosis. If a palpable thickening or mass corresponds to an asymmetric density, the density is regarded with a greater degree of suspicion for malignancy.

If the patient is premenopausal, particularly if she has other fibrocystic changes on mammography, a nonpalpable asymmetry may often be followed in 2 to 3 months, immediately after a menstrual cycle, and may diminish in size or disappear completely. In a postmenopausal patient, if there are no old films for comparison, a focal nonpalpable asymmetric density may be followed or biopsied, depending on the degree of asymmetry and other glandularity in the breast, the risk factors, and the clinical examination. A developing asymmetry in a postmenopausal woman who is not on hormones should be regarded with suspicion and biopsied. However, in the patient who has been placed on




hormone replacement therapy and who develops a focal asymmetry, discontinuation of the hormones and a repeat mammogram after 3 to 4 weeks may demonstrate that the area has diminished in size and is, therefore, parenchymal tissue.


Figure 3.19 HISTORY: A 58-year-old woman for screening mammography.

MAMMOGRAPHY: Left MLO view (A) shows heterogeneously dense tissue. There is a small apparent mass located far posteriorly(arrow). On spot-compression CC (B), the mass is confirmed (arrow), and it has a lobular contour with indistinct margins. Sonography of the mass (C) confirms that it has suspicious characteristics. On ultrasound, the mass is hypoechoic and irregular with central shadowing, consistent with malignancy.

IMPRESSION: Small mass, suspicious for carcinoma.

HISTOPATHOLOGY: Invasive ductal carcinoma.

NOTE: The workup of this potential mass includes spot compression and ultrasound. These are performed to determine if the density is a true mass and to further characterize it.


Figure 3.20 HISTORY: A 68-year-old gravida 2, para 2 woman for screening.

MAMMOGRAPHY: Left MLO (A), CC (B), and spot-compression magnification (1.5÷) (C) views. The breast shows fatty replacement. In the upper outer quadrant, there is a 5-mm, relatively well-circumscribed, isodense mass (A and B, arrow). The spot-compression view (C) shows the nodule to be smoothly marginated posteriorly, but to have a linear extensions anteriorly (arrow). Because of this finding, the mass is of suspicion for malignancy.

HISTOPATHOLOGY: Infiltrating ductal carcinoma.

NOTE: In the evaluation of a small circumscribed mass, it is important to perform spot compression to evaluate the margins. If the margins are not completely smooth and round, the lesion should be biopsied.

Architectural distortion may be the only manifestation of breast cancer and is a sign that is easily missed. By comparing images using the search patterns described, distortion and asymmetry are more easily identified. Architectural distortion is a finding commonly seen after surgery; however, in the absence of a surgical scar to correspond to the mammographic finding, distortion should be regarded as suspicious for malignancy (Fig. 3.25).

Evaluation of Calcifications

Calcifications of some type are present on a majority of mammograms, and it necessary to exclude those that are characteristically benign to avoid unnecessary biopsies. Microcalcifications rather than macrocalcifications are the form most often presenting as, or associated with, a carcinoma and represent a greater diagnostic dilemma. An analysis of the calcifications as to their distribution, size, morphology, variability, and the presence of associated findings, such as ductal dilatation or a mass, will assist one in deciding which are benign, which should be








followed carefully, and which should be biopsied. In conjunction with the analysis of the pattern of calcifications, the radiologist must keep in mind the patient's history and risk factors. A woman with a synchronous contralateral breast cancer or who is otherwise at high risk may be biopsied more readily for clustered microcalcifications of indeterminate nature.


Figure 3.21 HISTORY: A 67-year-old gravida 0 woman with a left breast lump. No nipple retraction was noted on clinical examination.

MAMMOGRAPHY: Left MLO (A) and CC (B) views. There is a 2.5-cm spiculated high-density mass in the left subareolar area. Retraction of the left nipple and skin retraction of the areola (arrow) are noted with compression of the breast during mammography. It is not unusual for a central carcinoma to produce nipple retraction during compression of the breast, even when it is not evident during clinical examination.

HISTOPATHOLOGY: Infiltrating ductal carcinoma, with macrometastases in 1 of 16 nodes.


Figure 3.22 HISTORY: A 54-year-old gravida 1, para 1 woman with a palpable left breast mass.

MAMMOGRAPHY: Left MLO view (A) shows the palpable mass to be lobular and isodense. On the spot-compression CC (B) and MLO (C) views, the borders of the mass are indistinct, features that are highly suggestive of malignancy. Ultrasound confirmed that the lesion was solid.

IMPRESSION: Highly suspicious for carcinoma.

HISTOPATHOLOGY: Poorly differentiated carcinoma with extensive necrosis.

NOTE: The margination of a mass is the most important feature to predict its nature. Malignant masses may be round, oval, lobular, or irregular in shape, but the indistinctness of the margins suggests the infiltrative nature of the lesion.


Figure 3.23 HISTORY: A 48-year-old woman for screening mammography.

MAMMOGRAPHY: Left CC view (A) shows a focal indistinct density located lateral to the areola (arrow). The mass persists on rolled CC medial (B) and lateral (C) views (arrows). On the rolled views, the lesion is displaced in an opposite direction from the direction of displacement, indicating that it is located inferiorly in the breast. On spot compression (D), the lesion appears spiculated. Ultrasound (E) shows the mass to be solid, irregular, and associated with surrounding edema, all of which are features highly suggestive of malignancy. The lesion was biopsied using a core needle and ultrasound guidance.

IMPRESSION: Highly suspicious mass, left breast.

HISTOPATHOLOGY: Invasive ductal carcinoma.


Figure 3.24 HISTORY: A 45-year-old woman after mastectomy for breast cancer on the right and treatment for colon cancer, for routine screening of the left breast.

MAMMOGRAPHY: Left MLO (A) and CC (B) views and ultrasound (C). There is a high-density mass with indistinct margins in the upper outer quadrant of the left breast (A and B, arrows). The mass has a slightly different shape on the two views but nonetheless persists as a high-density lesion. On ultrasound (C), characteristically malignant features are present (arrow): hypoechogenicity, ill-defined borders, and faint shadowing.

IMPRESSION: Carcinoma, left breast.

HISTOPATHOLOGY: Infiltrating ductal carcinoma with 13 negative axillary nodes.


Figure 3.25 Architectural distortion (arrow) in a palpable area at the edge of the dense parenchyma causes a fixed, tethered, spiculated appearance of the tissue. This appearance should prompt further assessment with spot compression and ultrasound. In this case, the pathology showed invasive lobular carcinoma.

Homogeneous round, smooth microcalcifications, which may be clustered, diffuse, or in tiny florets, usually represent lobular calcifications (33) (Fig. 3.26). Often, such calcifications occur in multiple quadrants and are bilateral. Most often, these represent benign fibrocystic disease—adenosis, sclerosing adenosis, lobular hyperplasia—but they may also occur in lobular neoplasia or lobular carcinoma in situ. For these reasons, such calcifications, when focal or clustered, should be at least followed up carefully with mammography. There are a variety of opinions concerning the recommendation for follow-up about this finding. If a decision is made to follow a patient with mammography rather than to biopsy calcifications, mammograms including magnification views should be obtained at 6-month intervals for at least 2 years and annually thereafter to assess for any increase in number.

Calcifications that are associated with carcinoma tend to lie in the abnormal ducts and assume shapes that are casts of the irregular malignant epithelial lining of the duct (33). Malignant calcifications can be identified as such when they are pleomorphic, linear, or branching, with irregular, jagged, sharp margins. When this morphology is identified on the mammogram—whether in a tight cluster or distributed in several groups, or even throughout an entire quadrant or an entire breast—biopsy is indicated (Fig. 3.27). Many malignant calcifications, however, may not have these classic features and may be more amorphous in appearance, although they do tend to have variability of size and shape.

Magnification views are particularly useful in the evaluation of the morphology of microcalcifications as well as in the more accurate determination of their distribution. The analysis of calcifications should include careful attention to all areas in each breast. There may be a variety of benign calcifications as well as malignant calcifications occurring synchronously in the same breast, and one must inspect carefully all areas to avoid overlooking an occult calcified malignancy (Fig. 3.28).

Ultrasound has a role in the evaluation of some patients with calcifications. Some calcifications are evident on ultrasound but most are not, so sonography really has no role in diagnosing or analyzing calcifications. The importance of ultrasound is to search for an underlying mass that may be indicative of an invasive cancer when malignant-appearing calcifications are identified in a dense breast (Fig. 3.29). Before biopsy, the identification of a suspicious mass on ultrasound affects the assessment of the case and the plan for the interventional procedure. The presence of a suspicious solid mass implies that an invasive cancer may be present, and the identification of an invasive component rather than pure ductal carcinoma in situ (DCIS) is critical to surgical management. When an invasive cancer is present, pathologic assessment of the axilla is necessary, whereas with pure DCIS, the axillary nodes are usually not sampled.

The Changing Mammogram

Comparison with a baseline mammogram is of great help in the decision about a focal mammography abnormality. The development of a new mass, an area of asymmetric soft tissue density, an area of architectural distortion, or calcifications should alert the radiologist that there is activity in the area identified and that further evaluation is necessary. Similarly, the change in size, density, or margination of a mass or density or an increase in the number of microcalcifications focally are of concern. Comparison with multiple previous studies, not just the most recent, is important in the determination of any change in a region that is being followed.




Figure 3.26 HISTORY: A 48-year-old woman recalled from screening mammography for calcifications.

MAMMOGRAPHY: Right ML (A) and enlarged CC (B) magnification views show dense parenchyma and scattered punctuate calcifications. There are clustered microcalcifications (arrows) at 10 o'clock that are a mixture of punctuate and amorphous forms. The morphology of these calcifications is more likely benign, but because of the amorphous type and the clustering, biopsy was performed.

IMPRESSION: Right breast clustered microcalcifications that are indeterminate; recommend biopsy, BI-RADS® 4.

HISTOPATHOLOGY: Fibrocystic changes with calcifications.

The doubling times of breast carcinomas vary greatly, from 44 to 1,869 days (34); therefore, the changes that occur at 6 months or a year may be quite variable. Kusama et al. (35) found the mean doubling time of breast tumors to be 3.5 months but to range from 0.2 to 18 months. The growth rate of breast cancer varies relative to patient age (35), with younger women exhibiting faster growth rates than older women (36). Although not common, it is not unusual to see the development of a 1-cm carcinoma in a 1-year interval, nor is it unusual to identify lack of perceptible change in a cancer over a year (Figs. 3.30 and 3.31). For these reasons, follow-up at 6-month intervals initially, followed by annual mammography after 2 years of stability, is a reasonable approach to management of a circumscribed mass. The lack of interval change of a suspicious lesion after 2 years is not a completely reliable indicator that it is benign, but it is highly suggestive of benignity. Lev-Toaff et al. (37) found that some microcalcifications associated with DCIS did not change for up to 5 years on serial mammograms.




Figure 3.27 HISTORY: A 26-year-old woman with a palpable thickening in the right lower-inner quadrant.

MAMMOGRAPHY: Right MLO (A) and magnified (2÷) (B) views. There are extensive irregular microcalcifications throughout the right lower-inner quadrant (A, arrow). Although the distribution is extensive, the morphology of these is highly malignant. Even if the area were nonpalpable, it should be considered highly suspicious for carcinoma.

IMPRESSION: Extensive calcifications, highly suspicious for carcinoma.

HISTOPATHOLOGY: Comedocarcinoma and infiltrating ductal carcinoma.

The importance of correct use of the information that is gleaned from the prior studies cannot be overemphasized. The documentation of the presence and stability of an equivocal finding on prior studies aids one in avoiding an unnecessary workup for a benign lesion. Any increase in size of a mass requires further evaluation with ultrasound. A noncystic enlarging mass that is circumscribed, indistinct, or spiculated requires biopsy for diagnosis. Microcalcifications that are increasing in number require biopsy unless the morphology is clearly benign. The formation of more benign-appearing calcifications, however, may not necessitate tissue sampling.

When the prior studies show the area in question, routine mammographic follow-up is performed. For many lesions that are cancers, the change in growth is slow, and the management of the lesion is completely affected by this inapparent change in size. For a very suspicious lesion that does not show appreciable change in size on mammography, the lack of change should not dissuade one from performing a biopsy. The radiologist must use the principle of judging a lesion by its worst features when in doubt about its etiology. In this situation, a suspicious morphology overrides a lack of growth or change on mammography.




Figure 3.28 HISTORY: A 75-year-old gravida 0, para 0 woman for routine screening mammography.

MAMMOGRAPHY: Bilateral CC views (A), left ML view (B), and specimen film (C). There are extensive benign calcifications of fat necrosis bilaterally (A and B, arrows). In addition, in the left lower-outer quadrant (A and B, curved arrows), there are irregular linear ductal-type microcalcifications extending from the nipple posteriorly. Because of the pleomorphic morphology and linear distribution, these calcifications have an appearance typical of comedocarcinoma. The area was removed following a needle localization procedure, and the specimen film (C) demonstrates many of the calcifications.

HISTOPATHOLOGY: Extensive intraductal carcinoma of comedo type, with residual carcinoma in the mastectomy specimen.

NOTE: It is important for the radiologist to evaluate all areas of calcifications independently, because there can be two or more etiologies present.




Figure 3.29 HISTORY: A 49-year-old woman for baseline mammography.

MAMMOGRAPHY: Left MLO (A) and CC (B) views show heterogeneously dense tissue. There are fine pleomorphic microcalcifications located at 4 o'clock (arrows). On the magnification CC view (C), the microcalcifications are highly pleomorphic and linearly arranged, features that are consistent with ductal carcinoma. Ultrasound (D) was performed to assess the region for an underlying mass and demonstrates a hypoechoic irregular mass that is highly suspicious for malignancy.

IMPRESSION: Pleomorphic microcalcifications and solid mass, highly suspicious for invasive ductal carcinoma.

HISTOPATHOLOGY: Invasive ductal carcinoma, DCIS.

NOTE: The use of ultrasound in the assessment of an area of calcifications that are highly suspicious for malignancy is helpful to identify a potentially invasive component. The presence of a solid mass within a region of microcalcifications suggests the possibility of invasive ductal carcinoma. Percutaneous biopsy of this mass is important in treatment planning. The preoperative diagnosis of invasive carcinoma prompts assessment of the axillary nodes pathologically, whereas the diagnosis of pure DCIS on needle biopsy does not usually lead to removal of lymph nodes at the time of lumpectomy. If invasive cancer is found at lumpectomy performed for DCIS, then a second surgical procedure is necessary to assess the axilla. The preoperative diagnosis of invasive carcinoma can reduce the need for a second surgical procedure because lumpectomy with sentinel node biopsy can be performed initially.




Figure 3.30 HISTORY: An 84-year-old woman after treatment of left breast cancer, for screening of the right breast.

MAMMOGRAPHY: Right CC view (A) from February 1987, right CC view (B) from December 1988, and right CC view (C) from November 1989. The breast shows fatty replacement. There is a small irregular density located medially (arrow) that was not identified on the original mammogram (A). Eighteen months later (B) it was noted, but because it was unchanged in size and density, biopsy was not recommended. It was followed at 24 months and remained unchanged, but at 30 months (C), it increased in size. The lesion was considered suspicious, and biopsy was recommended.

IMPRESSION: Irregular lesion, suspicious for carcinoma.

HISTOPATHOLOGY: Infiltrating ductal carcinoma, intraductal carcinoma.

NOTE: The lack of change in a nodule over 1 or even 2 years does not confirm a benign nature but suggests it. Carcinomas may grow slowly, and the changes mammographically may be subtle from 6-month interval to 6-month interval. It is important that comparison be made from the most current mammogram and the earliest study to be able to perceive the slight changes that occur in some malignant lesions.

Evaluation of a Palpable Breast Mass

Mammography has a significant role to play preoperatively in the presence of a palpable breast mass. The presence of a neoplasm can be confirmed; a clearly benign lesion, such as a lipoma or an oil cyst, which does not require biopsy, can be identified; an ipsilateral multicentric cancer is demonstrated, and an occult contralateral cancer can be detected; and an occult malignancy in a breast in which a palpable mass is to be excised can be demonstrated (38).

The approach to the patient with a palpable mass begins with mammography unless the patient is very young. An algorithm showing the management of the patient with a palpable mass is shown in Figure 3.32. Often, an additional spot-compression tangential view over the lump is performed to evaluate the area of clinical concern (39). Patel et al. (40), in a study of 496 patients with breast cancer, found that mammography was abnormal in


92%. In all 14 mammograms with missed cancers, the breast tissue was very dense. The authors (40) emphasized the importance of optimal mammographic techniques and the use of nonmammographic imaging modalities when examining women with dense breasts.


Figure 3.31 HISTORY: A 53-year-old gravida 0 woman presenting with a new 1-cm palpable mass in the right breast at the 12 o'clock position.

MAMMOGRAPHY: Right MLO (A) and CC (B) views from April 1990 and a right CC (C) view from March 1989. On the current study (Aand B), there is a high-density irregular mass (arrows) in the upper aspect of the right breast, corresponding in location to the palpable mass. The mass has a highly suspicious mammographic appearance. Comparison with the previous study (C) shows development of the lesion over a 1-year interval.

IMPRESSION: New irregular mass, right breast, suspicious for carcinoma.

HISTOPATHOLOGY: Intraductal and infiltrating ductal carcinoma, grade II of III, with peritumoral lymphatic invasion.

NOTE: There is considerable variability in the doubling times of breast carcinomas, some of which may be followed for several years without definite change, whereas others develop over a period of months. For these reasons, it is important not only to perform mammography regularly to detect interval changes but also to compare with studies before the last one in order to detect subtle changes of slowly growing cancers.

At times, mammography alone may be the only imaging modality used in the patient with a palpable breast mass. However, most often, ultrasound is performed next in this clinical situation. In situations in which the palpable lump is clearly benign—a lipoma, a calcified fibroadenoma, a hamartoma, or a normal lymph node—ultrasound is not necessary and does not change the management of the lesion. If the mass is clearly malignant on mammography, such as a spiculated mass with pleomorphic microcalcifications, ultrasound is not necessary to plan the management. The patient needs a biopsy no matter what the ultrasound shows. However, a reason to perform ultrasound in this situation may be to plan biopsy guidance if ultrasound rather than stereotaxis is preferred for percutaneous breast biopsy.




Figure 3.32 Algorithm for management of a palpable mass in a patient older than 30 years.

The importance of mammography is also to evaluate the remainder of the breasts for other clinically occult abnormalities. The mammogram may demonstrate additional clinically occult benign or malignant lesions that certainly affect the management of the palpable lesion itself. Before a needle biopsy of a palpable mass by a clinician, it is best to perform mammography, because the trauma from the biopsy procedure itself may alter the mammographic findings.

If the mammogram demonstrates a round non–fat-containing mass, an indistinct mass, asymmetry, or suspicious microcalcifications in the area of palpable concern, ultrasound is extremely important for further evaluation (Figs. 3.33 and 3.34). It is important that when scanning the patient with a clinical finding that the radiologist palpate the lesion and correlate the clinical finding with the mammographic finding. Sonography is used in the setting of a palpable mass to determine of the lesion is solid, cystic, or not seen. A simple cyst, even a palpable cyst, is benign based on sonography and is followed. A solid mass that is palpable is biopsied.

When mammography is negative and a palpable mass is present, ultrasound is the necessary next step. If ultrasound demonstrates a solid mass in the patient with a new palpable lesion, needle biopsy is performed (Figs. 3.35 and 3.36). Similarly, if ultrasound is negative and there is a clinically evident palpable lesion that is seen on mammography a needle biopsy is performed. When both mammography and ultrasound are negative, the management of the patient is based on the clinical findings. In a study of 420 patients with 455 palpable breast lesions that were negative on mammography and sonography, Soo et al. (41) found the negative predictive value of the combined mammographic and sonographic modalities was 99.8%. In another study of 600 palpable breast masses and negative mammography and sonography, Dennis et al. (42) found no cancers to occur or develop within 2 years in the women who were not biopsied. Sixty lumps were biopsied because they were clinically suspicious for cancer or occurred in high-anxiety, high-risk patients. Of these 60 palpable lumps, none were found to be malignant on tissue sampling. In a series of 233 women with negative imaging and a palpable lump, Moy et al. (43) found that 2.6% of patients had breast cancer in








the palpable area. In all six patients with cancer, the breast tissue was radiographically dense. If the palpable area is clinically of concern, even with negative imaging, fine-needle aspiration biopsy (FNAB) or needle biopsy are performed.


Figure 3.33 HISTORY: A 40-year-old woman with a palpable mass in the left breast at 9 o'clock.

MAMMOGRAPHY: Left MLO (A) and CC (B) view shows heterogeneously dense tissue. A BB marks the palpable abnormality; however, no abnormality is seen on these routine views nor on the spot compression view (C). Ultrasound of the palpable lesion (D) shows it to be round and nearly anechoic. However, there are low-level echoes and slightly indistinct margins associated with the mass.

IMPRESSION: Palpable mass, likely complicated cyst; recommend aspiration/biopsy.

HISTOPATHOLOGY: Cyst fluid with proteinaceous debris and histiocytes.

NOTE: The management of the palpable mass includes mammography and ultrasound, particularly when mammography is negative.


Figure 3.34 HISTORY: A 70-year-old woman with two palpable masses in the right periareolar area. She had a history of prior benign biopsy in the right breast years ago.

MAMMOGRAPHY: Right MLO (A) and CC (B) views show a wire marking the prior surgical site and two BBs marking the palpable masses. The mass at 5 o'clock is mammographically evident as a lobulated lesion, but the mass located more superiorly is not visible on mammography. A right MLO spot-magnification view (C) shows the edges of the palpable mass to be somewhat indistinct and therefore suspicious for malignancy. Ultrasound of the 5 o'clock mass (D) shows a hypoechoic, microlobulated mass with angulated margins and some shadowing. Ultrasound of the mammographically occult mass (E) located superiorly at 10 o'clock shows a smaller, densely shadowing lesion.

IMPRESSION: Highly suspicious for multicentric carcinoma.

HISTOPATHOLOGY: Five o'clock mass, infiltrating ductal carcinoma; superior mass, fibrosis.


Figure 3.35 HISTORY: A 36-year-old woman with a palpable left breast mass.

MAMMOGRAPHY: Left MLO (A) and CC (B) views show very dense parenchyma. On spot compression (C) over the palpable lump, slight increased density is seen. Sonography of the lump (D) shows a hypoechoic mass that is wider than tall and that is associated with slight acoustic enhancement. However, the edges are indistinct, and the echotexture is inhomogeneous, findings that are suspicious for malignancy.

IMPRESSION: Palpable mass, suspicious for carcinoma.

HISTOPATHOLOGY: Mucinous carcinoma.


Figure 3.36 HISTORY: A 36-year-old woman with a questionable, small, palpable nodule high in the right upper-inner quadrant.

MAMMOGRAPHY: Right MLO view (A) and ultrasound (B). There is a metallic marker overlying the palpable lump (A). No abnormality is seen on the mammogram. The palpable lesion is probably lying posterior to the edge of the film. Lesions high in the upper inner quadrant are difficult to image with the routine MLO and CC views. Ultrasound is of help demonstrating the location and characteristic of a palpable lesion not seen on mammography. On the sonogram (B), the mass, which lies just anterior to the pectoralis major muscle, is hypoechoic, somewhat irregular, and therefore suspicious in nature (arrow). The area was localized under ultrasound guidance before excisional biopsy.

IMPRESSION: Solid mass, suspicious for carcinoma.

HISTOPATHOLOGY: Infiltrating ductal carcinoma.

The triple test has been described (44,45,46,47,48,49) in the management of the patient who presents with a palpable breast lump. The likelihood that a palpable lesion is benign or malignant is predicted by the triple test, and depending on the test outcome, it may be used in lieu of open biopsy for some masses. The triple test consists of mammography, clinical examination, and FNAB. The addition of ultrasound since the triple test was originally described increases the strength of the combined test reliability. The concept of the triple test is that if all three tests indicate that the palpable mass is benign, the likelihood of malignancy approaches 0%. However, if any of the tests are positive or suspicious for malignancy, further evaluation with open biopsy is necessary.

The accuracy of FNAB is very much related to the technique for biopsy performance, quality of the aspirate, and skill of the pathologist in cytologic diagnosis. The technique includes the vigorous movement of a small-gauge


needle throughout the abnormality, shearing cells from the area of concern. Some cancers may be desmoplastic, so the use of a small-gauge needle (25 g) is helpful to mobilize cells.

The sensitivity of FNAB is in the range of 75% to 95% (50,51,52,53,54,55). The overall accuracy of FNAB for breast masses ranges from 85% to 90% (56,57,58). Depending on how the atypical smears are categorized (positive or negative) in the data analysis, reported sensitivity and specificity are affected. False negatives can occur, particularly with small lesions, low-grade tumors, fibrous tumors, or invasive lobular carcinomas (46). Therefore, a negative cytology alone is not sufficient to exclude cancer in the face of a palpable suspicious mass. A comprehensive approach to the patient with a palpable mass includes imaging, clinical findings, and cytology or biopsy if necessary.

Imaging the Symptomatic Young Patient

The approach to a young (<30 years) patient who is symptomatic, presenting with a palpable mass, is different from that for older patients. Because of the relatively low likelihood of malignancy, mammography is usually not performed first. Ultrasound is performed initially in women younger than 30 years (Fig. 3.37). Typically after age 30, mammography is performed first. If the palpable mass is a cyst, no further imaging evaluation is necessary in the normal-risk patient. If ultrasound demonstrates a solid mass that corresponds to the palpable lump, further management is typically performed. When sonography shows a smoothly marginated, oval, solid lesion, the likelihood is very high that the mass is a fibroadenoma (Fig. 3.38). FNAB or tissue sampling is usually performed to confirm this. Before an intervention, limited mammography can be performed in women older than age 20 to assess the lesion and the remainder of the breast (Fig. 3.39).


Figure 3.37 Algorithm for management of a palpable mass in a woman younger than 30 years of age.

If sonography shows the mass to be solid and suspicious because of irregularity of margins and a more rounded, irregular, or microlobulated shape, bilateral diagnostic mammography should be performed before biopsy. The purpose of mammography in this situation is to assess the extent of disease when a palpable cancer is present. Mammography may depict the suspicious lesion identified clinically and on ultrasound, but more importantly,


mammography may demonstrate other clinically occult foci of malignancy.


Figure 3.38 HISTORY: A 26-year-old gravida 0, para 0 woman with a new palpable mass in the right breast.

MAMMOGRAPHY: Ultrasound (A) was performed first because of the patient's age and demonstrates that the palpable mass is solid, lobular, and hypoechoic with well-defined margins, suggesting that it is most likely a fibroadenoma. Mammography was performed to evaluate this area and to exclude any other lesions before biopsy. Right MLO (B) and CC (C) views show no focal abnormality.

IMPRESSION: Palpable mass is solid, likely fibroadenoma; recommend biopsy.


Lesions that occur in young patients include fibroadenomas most commonly, with cysts, lymph nodes, and cancer less likely. In patients who are pregnant or lactating, particular lesions related to these states include lactating adenomas (59) and galactoceles. In young and in pregnant patients with breast cancer, the prognosis is similar to that of older patients with similar size and stage of disease (60). However, because of the lower frequency of breast cancer in younger women, the diagnosis may be delayed by not imaging and evaluating the patient (61,62). Because a patient is young and not at high risk because of family history, one should not assume that a palpable mass is not cancer. Although the breast tissue tends to be dense in young women, a majority of breast cancers in this age group are visible on mammography (63).




Figure 3.39 HISTORY: A 31-year-old gravida 0, para 0 woman with a palpable thickening in the 9 o'clock position of the right breast.

MAMMOGRAPHY: Right MLO view (A) shows a small mass located superiorly in the axillary tail (arrow) but no abnormality in the region of palpable concern at 9 o'clock. On the CC view (B), the mass is not clearly seen, but it is evident on the ML view (C, arrow). The lower location of the mass on the ML view compared with the MLO view is consistent with the lesion being in a lateral location. A spot exaggerated CC lateral (D) shows the small, high-density indistinct mass (arrow). The BB marking the palpable region, which was located more inferiorly, is also evident. Ultrasound of the palpable region at 9 o'clock (E) shows no abnormality. However, ultrasound of the axillary tail (F) demonstrates a small solid mass that is a taller-than-wide lesion with angulated margins.

IMPRESSION: Palpable region is not demonstrated on imaging; recommend management based on clinical findings. Solid, clinically occult lesion in the axillary tail is highly suspicious for carcinoma.

HISTOPATHOLOGY: Invasive ductal carcinoma axillary tail, negative sentinel lymph node.

NOTE: The assessment of the patient with a palpable mass requires attention to the area of palpable concern as well as the remainder of the breasts. Our protocol for imaging young women who present with a palpable mass is as follows: ultrasound only for those younger than age 20; ultrasound first, then mammography if needed between ages 20 and 30; mammography first, then ultrasound after age 30.



Imaging Extent of Disease

When an obvious mammographic finding is identified, a satisfaction of search phenomenon can occur and can cause a cancer to be missed. The radiologist's attention is drawn by the obvious finding, and a careful review of the remainder of the images does not occur. It is imperative that when a suspicious abnormality is identified, a complete review of the images be performed to search for other cancers (Fig. 3.40). Ipsilateral multiple cancers and contralateral synchronous cancers are identified frequently, and their diagnosis greatly affects patient management (64,65,66,67,68).


Figure 3.40 HISTORY: A 38-year-old gravida 3, para 3 patient for baseline screening.

MAMMOGRAPHY: Left MLO (A) and CC (B) views show a small indistinct mass located posteriorly at 5 o'clock (arrows). On the spot-compression magnification ML (C) and CC (D) views, the mass is dense and indistinct in margination. Within it and extending around it are fine pleomorphic microcalcifications (arrowheads) that are highly suspicious for DCIS.

IMPRESSION: Highly suspicious for malignancy, possible extensive intraductal component.

HISTOPATHOLOGY: Infiltrating ductal carcinoma, grade II; DCIS, solid and cribriform types.

NOTE: The identification of a suspicious mass as well as calcifications requires communication regarding the extent of disease for surgical planning. Pleomorphic calcifications extending beyond the mass suggest the possibility of an extensive intraductal component. Depending on the size of the area, percutaneous biopsy of the various components may be necessary. If lumpectomy is performed or attempted, bracketing the perimeters of the abnormality with several wires is helpful to achieve complete excision.

Multiple breast cancers may occur in the same quadrant, which is termed multifocal disease (Fig. 3.41), or they may occur in different quadrants, which is multicentric disease (Fig. 3.42). The clinical significance of multicentric disease is that this entity is a contraindication to breast conservation therapy, and typically a mastectomy is necessary (Fig. 3.43). Multifocality has been described to occur in 45% of cases of DCIS, and multicentricity has been found in 8% to 23% of cases (69,70). Multiple invasive cancers occur as well in the same and contralateral breasts (Figs. 3.44 and 3.45); in particular, patients with invasive lobular carcinoma are at highest risk of having synchronous contralateral disease (67,68). The incidence of synchronous




bilateral breast cancer has been reported in as many as 2% to 9% of patients (66). Modalities other than mammography, particularly MRI, have proven to be helpful in assessing the extent of disease in a patient with what appears to be a unifocal breast cancer (71,72,73,74,75,76). Regardless, a careful mammographic search pattern when one lesion is identified must be implemented to identify other cancers.


Figure 3.41 HISTORY: A 58-year-old woman with a palpable right breast mass.

MAMMOGRAPHY: Bilateral MLO (A) and CC (B) views show heterogeneously dense breasts. The palpable lump, marked by a BB, is a high-density lobular mass with indistinct margins. Just posterior to this mass is a smaller, similar-appearing lesion, better seen on the enlarged image (C).

IMPRESSION: Multifocal carcinoma.

HISTOPATHOLOGY: Invasive ductal carcinoma, multifocal.

NOTE: When the suspicious lesion is identified, the radiologist must search for other sites of malignancy that indicate multicentric or multifocal disease.


Figure 3.42 HISTORY: A 62-year-old woman who presents with a small palpable lump at the site of a bruise. This is her first mammogram.

MAMMOGRAPHY: Left MLO (A) and CC (B) views show a lobular isodense mass located superficially at 9 o'clock (arrow) and marked by the BB at the site of the palpable lump. In addition, there are multiple, high-density irregular masses with spiculated margins throughout the breast. Spot-compression CC (C) shows the highly spiculated appearance of several of these lesions.

IMPRESSION: Highly suspicious for multicentric carcinoma.

HISTOPATHOLOGY: Invasive ductal carcinoma, multiple foci, including the site of the palpable mass.


Figure 3.43 HISTORY: A 36-year-old woman for a baseline mammogram. She presents with palpable thickening in the right breast at 12 o'clock.

MAMMOGRAPHY: Right MLO (A) and CC (B) views show a dense indistinct mass posteriorly that corresponds to the palpable lesion. Anterior to this are multiple groups of pleomorphic fine linear microcalcifications (arrows). On the magnification ML (C) and CC (D) views, these highly malignant microcalcifications are visible (arrow). In addition, there are two other foci of linear and pleomorphic microcalcifications (arrowheads) located anteriorly, remote from the index lesion. These separate foci are very important in treatment planning. Because of their distance from the index lesion, mastectomy is likely necessary.

IMPRESSION: Multicentric ductal carcinoma.

HISTOPATHOLOGY: Ductal carcinoma, high grade, multiple foci with invasive ductal carcinoma.

Careful attention to subtle areas of asymmetry, nodules, and the presence of architectural distortion or microcalcifications is necessary in order to detect breast cancers at an early stage. Correlation with the history and physical findings is of help in determining recommendations about equivocal mammographic findings. The use of ultrasound and interventional techniques—such as aspiration, percutaneous breast biopsy, and galactography—as an adjunct to mammography allows the radiologist to make a more accurate diagnosis and plan for patient management.




Figure 3.44 HISTORY: A 67-year-old woman who presents with an abnormal screening mammogram.

MAMMOGRAPHY: Left MLO (A) and right MLO (B) views show heterogeneously dense tissue. Multiple abnormalities are present. Focal asymmetry is present in the left axillary tail (arrow), and a small, dense spiculated mass is noted in the superior posterior region(curved arrow) of the left breast. Clustered microcalcifications are present in both breasts (arrowhead). On the left CC view (C), the spiculated mass is noted to contain pleomorphic microcalcifications (curved arrow). On the right CC view (D), the clustered microcalcifications (arrowheads) are suspicious. In addition, there is a spiculated area of distortion laterally (curved arrow) in the right breast (D).

IMPRESSION: Highly suspicious for multicentric bilateral carcinoma.

HISTOPATHOLOGY: Invasive lobular carcinoma, both breasts, and multiple foci of DCIS (calcifications).

NOTE: When one abnormality that is suspicious for carcinoma is identified, the radiologist must search for other foci that could represent multicentric carcinoma in the same breast as well as the contralateral breast. The correct preoperative assessment of the extent of the disease is critical to achieving the best surgical management for the patient.




Figure 3.45 HISTORY: A 58-year-old gravida 5, para 5 woman with a firm tender mass in the right breast.

MAMMOGRAPHY: Right MLO (A) and CC (B) views and left MLO (C) and CC (D) views. There is a high-density spiculated mass with associated microcalcification in the right upper-outer quadrant (A and B). This lesion has an appearance highly suggestive of carcinoma. In the left upper-outer quadrant, there is an 8-mm indistinct mass (arrows) that is isodense (C and D). Although the mass appears less indistinct and dense on the CC view (D), it nonetheless persists and is therefore of moderate suspicion for a contralateral carcinoma.

IMPRESSION: Carcinoma of the right breast, moderately suspicious mass left breast.

HISTOPATHOLOGY: Right breast: infiltrating lobular carcinoma. Left breast: widespread intraductal carcinoma, lobular carcinoma in situ.




  1. ACR Breast Imaging and Reporting Data Systems, Breast Imaging Atlas.2003. Reston, VA: American College of Radiology.
  2. 21 CFR Part 16 and 900: Mammography quality standards: final rule. Federal Register. 1997 Oct 28:62(208): 55851–55994.
  3. Sickles EA. Periodic mammographic follow-up of probably benign lesions: results in 3,184 consecutive cases. Radiology1991;179:463–468.
  4. Vizcaíno I, Gadea L, Andreo L, et al. Short-term follow-up results in 795 nonpalpable probably benign lesions detected at screening mammography. Radiology2001;219: 475–483.
  5. Rosen EL, Baker JA, Soo MS. Malignant lesions initially subjected to short-term mammographic follow-up. Radiology2002;223:221–228.
  6. Leung JWT, Sickles EA. Multiple bilateral masses detected on screening mammography: assessment of need for recall imaging. AJR Am J Roentgenol2000;175:23–29.
  7. Bird RWE, Wallace TW, Yankaskas BC. Analysis of cancers missed at screening mammography. Radiology1992;184: 613–617.
  8. Goergen SK, Evans JE, Cohen GPB, et al. Characteristics of breast carcinomas missed by screening radiologists. Radiology1997;204:131–135.
  9. Roubidoux MA, Helvie MA, Lai NE, et al. Bilateral breast cancer: early detection with mammography. Radiology1995; 196:427–431.
  10. Krupinski EA. Visual scanning patterns of radiologists searching mammograms. Acad Radiol1996;3:137–144.
  11. Barrett JR, de Paredes ES, Dwyer SJ, et al. Unobtrusively tracking mammographers' eye gaze direction and pupil direction. Acad Radiol1994;1:40–45.
  12. Sickles EA. Practical solutions to common mammographic problems: tailoring the examination. AJR Am J Roentgenol1988;151:31–39.
  13. Egan RL. Fundamentals of mammographic diagnoses of benign and malignant diseases. Oncology1969:23:126–148.
  14. Hayes R, Michell M, Nunnerley HB. Evaluation of magnification and paddle compression techniques in the assessment of mammographic screening detected abnormalities. Clin Radiol1991;44:158–160.
  15. Berkowitz JE, Gatewood OMB, Gayler BW. Equivocal mammographic findings: evaluation with spot compression. Radiology1989;171:369–371.
  16. Gold RH, Montgomery CK, Rambo ON. Significance of margination of benign and malignant infiltrative mammary lesions: roentgenographic pathological correlation. AJR Am J Roentgenol1973;118:881–895.
  17. Sickles EA. Evaluation of breast masses. Radiology 1989; 173:297–303.
  18. Feig SA. Breast masses: mammographic and sonographic evaluation. Radiol Clin North Am1992;30(1):67–92.
  19. Swann CA, Kopans DB, Koerner FC, et al. The halo sign of malignant breast lesions. AJR Am J Roentgenol1987;149: 1145–1147.
  20. Gordenne WH, Malchair FL. Mach bands in mammography. Radiology1988;169:55–58.
  21. Marsteller LP, Shaw de Paredes E. Well defined masses in the breast. RadioGraphics1989;9(1):13–17.
  22. Sickles EA. Mammographic features of 300 consecutive nonpalpable breast cancers. AJR Am J Roentgenol1986;146: 661–663.
  23. Stomper PC, Davis SP, Weidner N, et al. Clinically occult, noncalcified breast cancer: serial radiologic-pathologic correlation in 27 cases. Radiology1988;160:621–626.
  24. Rubin E, Miller VE, Berland LL, et al. Hand-held real-time breast sonography. AJR Am J Roentgenol1985;144: 623–627.
  25. Jackson VP. The role of US in breast imaging. Radiology1990;177:305–311.
  26. Hilton SvW, Leopold GR, Olson LK, et al. Real-time breast sonography: application in 300 consecutive patients. AJR Am J Roentgenol1986;145:479–486.
  27. Sickles EA, Filly RA, Callen PW. Benign breast lesions: ultrasound detection and diagnosis. Radiology1984;151:467–470.
  28. Harper AP, Kelly-Fry E, Noe JA, et al. Ultrasound in the evaluation of solid breast masses. Radiology1983;146:731–736.
  29. Fornage BD, Lorigan JG, Andry E. Fibroadenoma of the breast: sonographic appearance. Radiology1989;172:671–675.
  30. Adler DD, Helvie MA, Ikeda DM. Nonpalpable, probably benign breast lesions: follow-up strategies after initial detection on mammography. AJR Am J Roentgenol1990;155:1195–1201.
  31. Sickles EA. Nonpalpable, circumscribed, noncalcified solid breast masses: likelihood of malignancy based on lesion size and age of patient. Radiology1994;192(2):439–442.
  32. Kopans DB, Swann CA, White G, et al. Asymmetric breast tissue. Radiology1989;171:639–643.
  33. Tabar L, Dean PB. Teaching atlas of mammography.Stuttgart: Georg Thieme Verlag, 1983.
  34. Fournier DV, Weber E, Hoeffken W, et al. Growth rate of 147 mammary carcinomas. Cancer1980;45:2198–2207.
  35. Kusama S, Spratt JS, Donegan WL, et al. The gross rates of growth of human mammary carcinoma. Cancer1972;30(2): 594–599.
  36. Moskowitz M. Breast cancer: age-specific growth rates and screening strategies. Radiology1986;161:37–41.
  37. Lev-Toaff AS, Feig SA, Saitas VL, et al. Stability of malignant breast microcalcifications. Radiology1994;192:153–156.
  38. Kopans DB, Meyer JE, Cohen AM, et al. Palpable breast masses: the importance of preoperative mammography. JAMA1981;246:2819–2822.
  39. Faulk RM, Sickles EA. Efficacy of spot compression-magnification and tangential views in mammographic evaluation of palpable breast masses. Radiology1992;185:87–90.
  40. Patel MR, Whitman GJ. Negative mammograms in symptomatic patients with breast cancer. Acad Radiol1998;5(1): 26–33.
  41. Soo MS, Rosen EL, Baker JA, et al. Negative predictive value of sonography with mammography in patients with palpable breast lesions. AJR Am J Roentgenol2001;177:1167–1170.
  42. Dennis MA, Parker SH, Klaus AJ, et al. Breast biopsy avoidance: the value of normal mammograms and normal sonograms in the setting of a palpable lump. Radiology2001;219: 186–191.
  43. Moy L, Slanetz PJ, Moore R, et al. Specificity of mammography and US in the evaluation of a palpable abnormality: retrospective review. Radiology2002;225:176–181.
  44. Hermansen C, Poulsen HS, Jensen J. et al. Diagnostic reliability of combined physical examination, mammography, and fine-needle puncture (“Triple-Test”) in breast tumors. Cancer1987;60:1866–1871.
  45. Butler A, Vargas HI, Worthen N, et al. Accuracy of combined clinical-mammographic-cytologic diagnosis dominant breast masses.Arch Surg1990;125:893–896.
  46. Adye B, Jolly PC, Bauermeister DE. The role of fine-needle aspiration in the management of solid breast masses. Arch Surg1988;123:37–39.
  47. Abele B, Miller TR, Goodson WH, et al. Breast fine-needle aspiration of palpable breast masses. Arch Surg1983;118:859–863.
  48. Frable WJ. Cancer trends: aspiration biopsy of the breast. Virginia Medical Med1982;109:452–455.
  49. Kaufman Z, Shpitz B, Shapiro M, et al. Triple approach in the diagnosis of dominant breast masses: combined physical examination, mammography, and fine-needle aspiration. J Surg Oncol1994;56:254–257.
  50. Stavric GD, Tevcev DT, Kaftandjiev DR, et al. Aspiration biopsy cytologic method in diagnosis of breast lesions: a critical review of 250 cases. Acta Cytol1973;17:188–190.
  51. Franzen S, Zajicek J. Aspiration biopsy in diagnosis of palpable lesions of the breast: critical review of 3,479 consecutive biopsies.Acta Radiol1968;17:241–262.



  1. Bjurstam N, Hedberg K, Hultborn KA, et al. Diagnosis of breast carcinoma: evaluation of clinical examination, mammography, thermography, and aspiration biopsy in breast disease. Prog Surg1974;13:1–65.
  2. Smith C, Butler J, Cobb C, et al. Fine-needle aspiration cytology in the diagnosis of primary breast cancer. Surgery1988;103:178–183.
  3. Boerner S, Sneige N. Specimen adequacy and false-negative diagnosis rate in fine-needle aspirates of palpable breast masses.Cancer1998;84(6):344–348.
  4. Wanebo HJ, Feldman S, Wilhelm MC, et al. Fine needle aspiration cytology in lieu of open biopsy in management of primary breast cancer. Ann Surg1984;199:569–578.
  5. Norton LW, Davis JR, Wiens JL, et al. Accuracy of aspiration cytology in detecting breast cancer. Surgery1984;96:806–814.
  6. Sheikh FA, Tinkoff GH, Kline TS, et al. Final diagnosis by fine-needle aspiration biopsy for definitive operation in breast cancer. Am Surg1987;154:470–474.
  7. Kline TS, Neal HS. Needle aspiration biopsy: a critical appraisal. JAMA1978;239:36–39.
  8. James K, Bridger J, Anthony PP. Breast tumour of pregnancy (“lactating” adenoma). J Pathol1988;156:37–44.
  9. Petrek JA, Dukoff R, Rogatko A. Prognosis of pregnancy-associated breast cancer. Cancer1991;67:869–872.
  10. Hoover HC. Breast cancer during pregnancy and lactation. Surg Clin North Am1990;70(5):1151–1163.
  11. Donegan WL. Breast cancer and pregnancy. Obstet Gynecol1977;50(2):244–252.
  12. Shaw de Paredes E, Marsteller LP, Eden BV. Breast cancers in women 35 years of age and younger: mammographic findings.Radiology1990;177:117–119.
  13. Kinne DW. Management of the contralateral breast. In: Harris JR, Hellman S, Henderson IC, et al, eds. Breast diseases.Philadelphia: JB Lippincott, 1987:620–621.
  14. Kesseler HJ, Grier WRN, Seidman I, et al. Bilateral primary breast cancer. JAMA1976;236(3):278–280.
  15. Hungness ES, Safa M, Shaughnessy EA, et al. Bilateral synchronous breast cancer: mode of detection and comparison of histologic features between the two breasts. Surgery2000;128:702–707.
  16. Broet P, de la Rochefordiere A, Scholl SM, et al. Contralateral breast cancer: annual incidence and risk parameters. J Clin Oncol1995;13:1578–1583.
  17. Healey EA, Cook EF, Orav EJ, et al. Contralateral breast cancer: clinical characteristics and impact on prognosis. J Clin Oncol1993;11:1545–1552.
  18. Coombs JH, Hubbard E, Hudson K, et al. Ductal carcinoma in situ of the breast: correlation of pathologic and mammographic features with extent of disease. Am Surg1997; 63:1079–1083.
  19. Lagios MD, Westdahl PR, Margolin FR, et al. Ductal carcinoma in situ: relationship of extent of noninvasive disease to the frequency of occult invasion, multicentricity, lymph node metastases, and short-term treatment failure. Cancer1982;50:1309–1314.
  20. Berg WA. Imaging the local extent of disease. Semin Breast Dis2001;4:153–173.
  21. Berg WA, Gutierrez L, NessAirer MS et al. Diagnostic accuracy of mammography, clinical examination, ultrasound, and MR imaging in preoperative assessment of breast cancer. Radiology2004;233:830–849.
  22. Liberman L, Morris EA, Dershaw DD, et al. MR imaging of the ipsilateral breast in women with percutaneously proven breast cancer.AJR Am J Roentgenol2003;180:901–910.
  23. Orel SG, Schnall MD, Powell CM, et al. Staging of suspected breast cancer: effect of MR imaging and MR-guided biopsy. Radiology1995;196:115–122.
  24. Boetes C, Mus RD, Holland R, et al. Breast tumors: comparative accuracy of MRI imaging relative to mammography and US for demonstrating extent. Radiology1995;197(3): 743–747.
  25. Smith M, Allison K, Shaw de Paredes E. Nonmammographic evaluation of the extent of breast carcinoma. Semin Ultrasound CT MR2006;27:308–319.