Imaging Breast Disease:Mammography and Breast UltrasoundE.A.Sickles

(1)

Introduction

The gold standard among breast imaging techniques is mammography, which is a plain-film X-ray examination.

The routine use of mammographic screening for clinical- ly occult breast cancer is widespread, primarily due to fa- vorable results from multiple randomized controlled tri- als and the development of improved methods of preop- erative needle biopsy and localization. A large percentage of radiologists interpret mammograms, and daily case- loads in many practices involve 50 or more such examinations per day. Screening examinations comprise more than 75% of all mammography performed in most devel- oped countries. As a result, the majority of detected lesions, both benign and malignant, are small and nonpalpable. In most general radiology practices, mammography accounts for at least 10% (sometimes as much as 20%) of all examinations performed.

In many countries, the widespread use of mammographic screening among asymptomatic women, coupled with a high level of general awareness of breast cancer as a public health problem, has prompted a series of gov- ernmental regulations affecting mammography. These regulations, although costly, also have resulted in an overall improvement in the quality of mammographic images and, due to the imposition of initial education, continuing education, and continuing experience requirements for radiologists, probably an improvement in the quality of mammographic interpretation as well.

Breast ultrasonography (US) also is widely used to aid in the diagnosis and management of either mammographically detected or palpable breast lesions. US reliably characterizes simple cysts, if rigorous interpretive criteria are used, thereby averting tissue diagnosis for these common, invariably benign lesions (Fig. 1). In addition, several sonographic features are sufficiently suggestive of malignancy to prompt biopsy (often resulting in a diagnosis of cancer), even in the absence of suspicious findings at mammography or clinical breast examination (Fig. 2). These are major reasons why US has be- come an integral part of modern breast imaging practice, although it has not yet been proved to be effective for screening asymptomatic women.

E.A. Sickles¹, R.A. Kubik-Huch²

1Department of Radiology, UCSF Medical Center, San Francisco, CA, USA

2Institut für Radiologie, Kantonsspital Baden, AG, Switzerland

Fig. 1a, b. Simple cysts (benign, BI-RADS category 2 at US) in two different patients. Breast US reliably characterizes simple cysts.

Criteria for this diagnosis include circumscribed margins, oval (or round) shape with parallel (wider-than-tall) orientation, anechoic center, and posterior enhancement, as well as thin echogenic cap- sule and thin edge shadows. Note that reverberation echoes pro- jecting within the cyst lumen (arrows, b) may create an appearance of complexity. Sonographic scanning from different angles and op- timizing the focal zone may eliminate much of the artifact a

b

¥ ¥ ¥

(2)

Fig. 2a, b. Multifocal carcinoma of the left breast (highly suggestive of malignancy, BI-RADS category 5 at US). Breast US demonstrates an irregularly shaped, hypoechoic mass with indistinct margins and some posterior shadowing, as well as focal hyperechoic spots proba- bly representing calcifications. The normal tissue architecture is disrupted. A second smaller lesion (arrows) is seen lateral (a) and supe- rior (b) to the larger mass

a b

Fig. 3a-d.A 40-year-old patient with multifocal invasive ductal carcinoma of the right breast, pT2N2M0) a Mammography, mediolateral oblique views; b mammography, craniocaudal views; c US of the right breast; d US of the right axilla. Assessment: highly suggestive of malignancy (BI-RADS category 5). A large, partially circumscribed and partially irregular/indistinct mass is seen in the right breast at the 6 o’clock position (c, d). On craniocaudal view, a second lesion is seen in the central aspect of the breast close to the chest wall (arrow).

In addition, three smaller circumscribed lesions are present in the axilla, two of which represent enlarged lymph nodes. US confirmed the presence of a partially circumscribed and partially irregular/indistinct hypoechoic mass, as well as axillary lesions that were subsequently found to represent metastatic lymph nodes

a

c

b

d

¥ ¥

¥

(3)

curate procedures for performing percutaneous tissue diagnosis also are in widespread use, based on either sonographic or stereotactic mammographic guidance.

Coupled with either f ine-needle aspiration or core biopsy, these approaches allow reliable diagnosis of most benign lesions without the need for surgical re- section, thereby reducing morbidity and cost. Further- more, the percutaneous diagnosis of malignancy prior to open surgery often permits the delivery of definitive surgical treatment in a streamlined and also less costly manner.

Circumscribed Noncalcified Mass

The major differential diagnostic possibilities for a circumscribed noncalcified mass seen at mammography are cyst, fibroadenoma, and circumscribed carcinoma (Fig. 3). Most circumscribed noncalcified masses identified at screening, are nonpalpable and <1 cm. These types of masses are initially evaluated with US, which can reliably diagnose a simple cyst, thereby eliminating the need for any further work-up. However, if the mass is solid but displays no sonographic features of malignancy, and if it continues to appear circumscribed at fine-detail mammography (using spot-compression with or without magnification), then it usually is assessed as being probably benign and managed with periodic mammographic surveillance. Tissue diagnosis is performed only for (1) those masses with suspicious imaging features at US or fine-detail mammography, (2) masses with suspicious features at palpation, and (3) masses that are seen to have enlarged since a prior examination.

Thus, the few circumscribed masses that prove to be malignant are diagnosed readily but the great majority of benign masses are managed simply with mammography and US. A relatively small percentage of such masses are now evaluated by percutaneous biopsy, and only very few benign lesions are actually excised.

Grouped Microcalcifications

Most cases of grouped microcalcifications are detected at mammographic screening, and most screening-detected cancers are treated with breast preservation surgery. When grouped microcalcifications are detected (or even suspected) at screening, additional fine-detail imaging is usually obtained, using spot-compression

magnification mammography, to portray with greater clarity the shapes and extent of the calcifications (Figs.

4, 5). This permits classification of some otherwise

‘suspicious’ cases into more benign categories, such that truly benign lesions are managed by mammographic surveillance rather than by tissue diagnosis, thereby reducing morbidity and cost. This approach also pro- vides additional information about the extent of disease prior to an appropriately indicated interventional proce- dure, which is very useful in planning exactly what type of intervention(s) to perform, especially if carcinoma is later found at biopsy.

Architectural Distortion

The mammographic feature of architectural distortion most often involves thin radiopaque lines (spiculation) ra- diating from a central point, with no definite mass visible at the center of the lesion. Another, less common manifestation of architectural distortion is focal retrac- tion or distortion of the edge of the parenchyma at one of its interfaces with either the subcutaneous fat or retro- mammary fat/fascia. In the absence of an appropriate his- tory of trauma or surgery, architectural distortion is sus- Fig. 4. Spot-compression magnification mammogram of the right breast. A group of microcalcifications is seen within dense breast tissue. Additional scattered calcifications were distributed throughout much of the right breast parenchyma. The calcifications are round and punctate, some larger and some smaller than 0.5 mm.

Assessment: probably benign, BI-RADS category 3. The patient declined periodic mammographic surveillance; instead, stereotactic core biopsy was performed. Histology confirmed the suspected diagnosis of sclerosing adenosis

(4)

picious for malignancy or radial scar, and biopsy is appropriate (Fig. 6). In a retrospective study of 300 consecutive nonpalpable breast cancers, 9% were identified as architectural distortion. The likelihood of malignancy among findings characterized as architectural distortion is estimated to range from 10 to 35%, depending on case selection, patient age, and other factors.

Asymmetry

There are four types of asymmetry, all of which represent areas of fibroglandular-density tissue that is more exten- sive in one breast than the other. Asymmetry is usually distinguished from mass in that it demonstrates concave- outward rather than the convex-outward contours typical of a mass, and it is interspersed with fat rather than ap-

pearing denser in the center than at the periphery (typical of a mass). Summation artifact is caused by superimpo- sition of normal fibroglandular breast structures on a giv- en mammographic projection. As such, this finding will be visible on only one of the two standard views and will not be seen on additional views taken in different projections. The clinician must distinguish summation artifact from the other types of asymmetry described below, be- cause it never represents breast cancer and should not be subjected to biopsy. A one-view-only mammographic finding is called an asymmetry prior to diagnostic imag- ing evaluation. Most (>75%) asymmetries represent summation artifacts; the rest are other types of asymmetry or masses. Global asymmetry (formerly called asymmetric breast tissue) represents a greater volume of fibroglandular tissue in one breast than in the corresponding location in the opposite breast, without an associated mass, mi- Fig. 5a-d. First-screening mammography study in a 52-year-old asymptomatic woman. a Spot-compression magnification view of the right breast. Note the grouped, fine-linear and branching (casting) microcalcifications, highly suggestive of malignancy (BI-RADS category 5).

b Preoperatively, mammographically guided hook-wire localization was performed. c Complete excision of the lesion was confirmed by a specimen radiograph. d Histology showed ductal carcinoma in situ (DCIS)

a

c

b

d

¥

(5)

crocalcifications, or architectural distortion. It is found in approximately 3% of screening mammography examinations. It almost always represents a normal variant, but occasionally it may indicate the presence of an underly- ing breast cancer if it corresponds to a palpable abnormality. In this latter clinical setting, the radiologist usually will recall the patient for additional imaging evalua- tion. Focal asymmetry (formerly called focal asymmetric density) is an asymmetry of fibroglandular-tissue density seen on two different mammographic projections but lacking the convex-outward contours and conspicuity of a mass. It usually represents an island of normal dense breast tissue, but its lack of specific benign characteris- tics may warrant further evaluation, especially if it is not interspersed with fat. In a retrospective study of 300 consecutive nonpalpable breast cancers, 3% were identified as focal asymmetries. The likelihood of malignancy among findings characterized as focal asymmetry (without associated mass, calcifications, architectural distortion, sonographic abnormality, or palpable correlate) is

<1%. Developing asymmetry (formerly called developing

density or neodensity) is a focal asymmetry that is new, larger, or denser on current examination than on a previ- ous one. To identify such a lesion, prior mammograms must be available for comparison. In a retrospective study of 300 consecutive nonpalpable breast cancers, 6% were identified as developing asymmetries. The likelihood of malignancy among findings characterized as developing asymmetry is estimated to be 10-15%.

Typical Breast Cancer

Mass

The great majority of cancers presenting as masses are nonpalpable and detected at mammographic screening, often smaller than 1 cm in greatest dimension and only rarely with signs of locally advanced disease. At the University of California – San Francisco (UCSF), 85% of the invasive cancers detected at screening have negative axillary lymph nodes, and 71% of invasive cancers are perolateral and deep to the lesion. An irregular/indistinct hypoechoic lesion, non-parallel (taller-than-wide) in orientation, was confirmed at US (highly suggestive of malignancy, BI-RADS category 5). Histology showed invasive ductal carcinoma, pT2N0 a

c

b

(6)

stage 1 lesions. Most malignant masses found at screening have indistinct rather than spiculated margins.

Presumably this is due to earlier detection, before suffi- cient desmoplastic reaction has formed to be visible as spiculation on standard screening images.

Calcifications

Due to the fairly widespread use of mammographic screening, a substantial percentage of cancers present with calcifications alone. These lesions generally are nonpalpable, many are ≤1 cm, and the great majority are stage 0 lesions (ductal carcinoma in situ, DCIS). At the UCSF, during each year over the past 15 years, DCIS has accounted for between 25 and 30% of all cancers detected at mammographic screening. These highly cur- able malignancies are now treated with breast preservation therapy, so long as they do not occupy a large volume of the breast in relation to overall breast size, to the extent that complete tumor excision would compromise cosmesis.

Breast Implants

Implant imaging with film-screen mammography does not include in the image field the most posterior por- tions of a breast implant, but the development of implant-displaced views has greatly improved visualiza- tion of native breast tissues anterior to the implant. This occasionally permits detection of a nonpalpable breast cancer that otherwise would not have been visible.

Breast MRI also adds to the capabilities of modern mammography, by readily demonstrating extracapsular posterior-surface rupture and intracapsular rupture of a silicone-filled implant.

Reporting Breast Imaging Findings

Effective reporting of breast imaging findings requires the use of a logical structure, consistent terminology,

straightforward assessment and associated management recommendations, all packaged in a clear and concise document. The American College of Radiology devel- oped the Breast Imaging Reporting and Data System (BI-RADS) in the early 1990s in order to achieve these goals. BI-RADS is now widely used throughout North and South America, and is being used more extensively in other countries as well. A major advantage of BI- RADS is that it employs standardized terminology for imaging features, associated with specific evidence- based clinical assessment categories and management recommendations (Table 1). The most recent update of BI-RADS (4th edition, 2003) includes not only mammography but also breast US and MRI findings. When a patient has more than one breast examination on the same day, a single report should be produced, describing the imaging findings of the component examinations and including an integrated assessment that takes into ac- count all findings and makes a single set of management recommendations.