Chapter 41: BASIC BREAST SONOGRAPHY

Sonography has emerged as an essential tool in breast evaluation. Although originally used as an adjunct to mammography to distinguish solid from cystic masses, advances in instrumentation and extensive clinical experience have allowed expansion of its roles to include characterization of solid masses, guidance for interventional procedures, and screening.

Sonography is very operator dependent by nature,¹ and this is especially true in the breast where standard anatomic references are limited. To achieve acceptable levels of sensitivity and specificity, it is important to optimize technical parameters, develop standard exam protocols, and carefully integrate sonographic findings with other imaging modalities and the clinical setting.

Diagnostic

Diagnostic evaluation remains the main indication for breast sonography and can be performed as either the primary imaging test or as an adjunct to other imaging procedures. In young women and women who are pregnant, sonography is generally the first test performed for a new mass or other focal breast abnormality.² If the sonographic evaluation satisfactorily explains the clinical abnormality, no further testing may be required. If sonography does not explain a persistent focal finding, additional evaluation with mammography would be needed.

In other women, diagnostic evaluation of a focal clinical abnormality would generally begin with mammography, reserving sonography for supplemental assessment, either to characterize a mammographic abnormality or to increase the sensitivity of the evaluation should no mammographic abnormality be found.

Implant assessment can be performed with mammography, sonography, or magnetic resonance imaging (MRI). Although less accurate than MRI, sonography is often a simple first test when rupture of a silicone implant is suspected.

In asymptomatic women, sonography may be used to characterize abnormalities on screening mammography, and increasingly, sonography is the next modality to be used for evaluation of MRI-detected abnormalities that are not seen on mammography (so-called "second look" sonogram).

Breast sonography is not indicated for generalized breast complaints such as diffuse background nodularity or nonfocal mastodynia.

Screening

Sonography as a screening method for breast cancer detection remains controversial.³ There is certainly evidence to support its use, especially in high-risk patients with dense breast tissue,^4,5 but the issue of operator dependency has hindered its expansion into routine clinical practice. Recently published information from the American College of Radiology Imaging Network⁶ confirms an increase in sensitivity for this group, but at the expense of a large number of false-positive studies and unnecessary biopsies. If performed, screening breast sonography should be limited to tertiary-level facilities with well-trained breast sonologists and experienced breast imagers. Because it does not reliably depict calcifications or a variety of other mammographic signs of malignancy it should not be used as a substitute for screening mammography. As breast MRI emerges as a reliable screening tool for high-risk women, the role of screening sonography remains in question.

Mammography remains the standard of care for screening of breast disease in women over 30. The standard bilateral CC (craniocaudal) and MLO (mediolateral oblique) views are evaluated for interval changes or suspicious baseline findings. If further evaluation is warranted, the patient then enters the diagnostic realm where additional mammographic views are obtained, including, but not limited to, spot compression views and magnification views. Depending on mammographic findings, supplemental evaluation with sonography may then be appropriate.

Intervention

Most breast biopsies are now performed with imaging guidance. For those lesions that are only seen on mammography, biopsy is usually performed with stereotactic guidance. For those things well seen with sonography, however, ultrasound guidance offers the advantage of more comfortable patient positioning and simple targeting.⁷ Sonography can be used to direct fine-needle aspiration biopsies, large-gauge core biopsies, and vacuum-assisted core biopsies.⁸ The large-gauge (8- to 12-gauge) vacuum-assisted core biopsies are well suited for percutaneous removal of small fibroadenomas, which may obviate the need for follow-up. Sonographic guidance is especially useful for the aspiration of cysts, seromas, abscesses, and other fluid collections in the breast, and it is increasingly used for deposition of marking clips in patients undergoing neoadjuvant therapy for breast cancer.⁹

Much of the expansion of sonography in breast imaging is related to technical advances in instrumentation. Breast sonography requires high-resolution near-field capability that is not available on most standard transducers. Transducers for the breast typically use linear arrays and operate at bandwidth frequencies ranging from 5 to 10 MHz up to 8 to 17 MHz. Color and power Doppler are useful for assessment of vascularity, but detailed Doppler characterization is less well established than in other areas of ultrasound.^{10,11, and 12}

Spatial compounding and tissue harmonic imaging have allowed us to take advantage of the greater detail by helping manage the artifacts generated by higher frequencies and bandwidth. Compound imaging is performed by processing information from multiple sound beams that are generated from across the transducer face. Harmonic imaging takes advantage of echoes of harmonic frequencies that can be detected and separated from the fundamental frequency and associated artifacts.

At a minimum, one should use a good linear array transducer operating at least at 7 to 10 MHz¹³ and capable of focusing well in the near field.

Considerable attention is given to this section as many errors are directly related to improper technique.^13,14

Technique is adapted to the purpose of the examination. Most breast sonographic evaluations will be performed in the diagnostic setting as focused examinations targeted to a specific clinical or imaging abnormality. If a persistent focal clinical abnormality is being assessed, targeting of the sonographic exam is usually straightforward. For subtle or inconsistently reproducible clinical findings, targeting may be more difficult, but in all circumstances every attempt should be made to determine whether a sonographic finding does or does not correlate with a given clinical finding.

If sonography is performed for an abnormality identified on another imaging test, it is equally important to determine if a sonographic abnormality corresponds to the imaging abnormality for which ultrasound was undertaken. This correlation can be difficult given differences in breast position and orientation on sonography, mammography, and MRI. It is helpful to adopt a reproducible system for three-dimensional localization in the breast to facilitate correlation with other modalities, as well as to allow more accurate follow-up of sonographic findings placed into imaging surveillance. The use of clock-face position is an almost universal way of providing some degree of localization in the breast, but this does not provide radial or depth information. Radial coordinates can be given as distance from the nipple or in zones (eg, 3 concentric rings from the nipple). Depth coordinates are best given as 3 zones or layers from the skin to the chest wall, as the actual distance may vary considerably depending on patient position.

Position of the patient is usually supine, although it may occasionally be useful to scan the patient in upright position, especially if it is difficult for her to reproduce a palpable finding when supine. The ipsilateral arm is raised to stretch out the breast tissue and decrease tissue thickness. Initially, the breast should be positioned in such a way that it falls symmetrically on the chest wall with the nipple in the center of the breast. Multiplanar coordinates should be derived from this position as radial and depth coordinates may change significantly in other positions. Patient position can then be changed to maximize scanning detail. For example, when evaluating lesions deep in the upper outer quadrant of the breast, it may be helpful to have the patient roll farther away from the examiner so that most of the breast tissue falls medially, thinning out the tissue laterally, reducing depth, and improving detail.

All focal abnormalities should be recorded in 2 planes. While transverse and sagittal scan planes are common in most areas of the body, the breast and its ductal system are arranged radially. The use of radial and antiradial planes is useful for orientation, and likewise, radial scanning may help identify subtle ductal pathology such as duct extension, which may be seen in invasive malignancy. Regardless of the scan planes chosen, all lesions should be measured in 3 dimensions, and all images should have appropriate identifying information.

Gain should be adjusted so that there is even echogenicity throughout the depth of the tissue being imaged and so that mammary fat is mid-gray.¹⁵ Tissue is then characterized as hypoechoic or hyperechoic relative to the mammary fat. Doppler evaluation is suggested for all masses that are not completely anechoic; this is not so much to characterize the nature of the blood flow, but simply to determine if there is blood flow within the mass, which would indicate that it is solid. For superficial lesions, one should use a standoff pad or a thick layer of gel to improve near-field detail.

Transducer pressure and angle can dramatically alter the appearance of the underlying tissue. In general, it is best to use light to moderate pressure on the transducer. Increasing pressure may help resolve edge shadows and other artifacts, and decreasing pressure may help identify subtle vascularity that can be obscured by too much pressure. Angling or rocking the transducer back and forth can improve margin assessment, and rotation of the transducer helps confirm that a finding is three-dimensional.

When performing screening breast sonography, it is necessary to develop a systematic approach to be certain that the entire breast has been scanned in 2 planes. This may be very time consuming, especially with large breasts. Benign sonographic findings are frequently encountered in screening and should not distract the examiner from the primary goal of searching for suspicious sonographic lesions. Any value of screening sonography can easily be lost by under-calling subtle cancers or by over-calling benign findings.

Just as it is important to generate images to completely characterize a sonographic finding, it is important to accurately describe that finding in the breast imaging report. The American College of Radiology's Breast Imaging-Reporting and Data System (BI-RADS) lexicon has been developed to allow standardized reporting for mammography, breast sonography, and breast MRI. Use of standard lexicon terminology is strongly recommended to facilitate communication and understanding of results. Although BI-RADS outcome coding is only mandated for mammography, its use in sonography and other breast imaging modalities is encouraged.

Gross anatomy of the breast is relatively straightforward. From superficial to deep, there is skin, premammary fat, anterior layer of pectoral fascia, glandular tissue, deep layer of pectoral fascia, retromammary fat, and pectoral muscle (Figures 41-1 and 41-2). Cooper's ligaments represent the anterior and posterior extensions of the mammary fascia towards the skin and chest wall, respectively. The nipple-areolar complex consists of the nipple, from which 10 to 20 lactiferous ducts originate, and the surrounding areola, which is variable in pigment and size. The sebaceous glands of Montgomery are found in the areola; they may communicate with adjacent lactiferous ducts and may become inflamed or infected.

The subareolar ducts are variable in size, and the significance of the duct size may depend on the setting. Asymmetric, isolated, or increasing duct dilatation is more significant than bilateral prominence of multiple ducts, which often reflects duct ectasia or even normal variation. As the ducts are followed posteriorly, they become smaller and more difficult to visualize. Finally, they end in small terminal ducts from which the terminal ductal-lobular units arise. If you think of the lactiferous duct at the nipple as the root of a tree, the branching ducts behind it are like the limbs and branches of the tree and the terminal ducts and terminal ductal-lobular units are like the leaves. The terminal ductal-lobular units are important as they are where much of breast pathology arises, including cysts, fibroadenomas, and most breast cancers.

The normal terminal ductal-lobular unit measures about 3 mm in size and, depending on background tissue, can be visualized on high-quality mammography and sonography (Figure 41-3).

Axillary and intramammary lymph nodes represent another benign finding in the breast. Intramammary lymph nodes were previously felt to be less common, but with the use of breast MRI, we have come to realize that they are frequently present. There is tremendous variation in size of lymph nodes in the breast, especially those in the axilla, so they are better assessed by morphology than size. Benign lymph nodes typically maintain a reniform shape with preserved fatty hilum and symmetric cortical thickness (Figure 41-4).

It is important to understand the sonographic appearance of the skin as many superficial lesions are dermal in origin. The skin is seen as 2 thin echogenic lines separated by a thin hypoechoic band. The combined thickness is generally 2 to 3 mm except at the areola where it is thicker. Most dermal lesions will displace the deep echogenic line posteriorly, allowing one to reliably determine that the finding is arising from the skin (Figure 41-5). Detailed assessment of the skin often requires a standoff pad or thick layer of gel to allow better resolution in the near field.

There is marked variation in the sonographic appearance of normal breast tissue, just as there is in the mammographic patterns. In addition, there is considerable variation in the sonographic appearance of breast pathology. These features result in overlapping of the appearances of normal, benign, and malignant processes and present the examiner with significant challenges. Because of this complexity, most ultrasound lesions are assessed based on a combination of features.¹⁶ This section reviews the ultrasound features of some of the more common benign and malignant lesions.

Fibrocystic Condition

Fibrocystic changes are found in up to 60% of women. These changes begin at the level of the terminal duct and ductal-lobular unit and are felt to be the result of cyclical hormonal influences. As the name suggests, fibrocystic condition reflects a combination of periductal fibrosis and cyst formation. Appearances depend on the ratio of fibrous to cystic elements and on the size of the cysts. The simplest presentation is that of a simple macrocyst (Figure 41-6). If a mass is circumscribed, anechoic, and demonstrates good sound transmission, it is a simple cyst and requires no further evaluation unless aspiration is desired for symptomatic relief.² It is important to use proper technique when evaluating cysts to make sure that real echoes within the mass are not suppressed by a low gain setting. Simple cysts may be gently lobulated and may have thin avascular septations.

Simple cysts are straightforward, but cysts may also demonstrate complicated and complex appearances. A complicated cyst is a cyst that meets the criteria for a simple cyst but also contains internal echoes that are not felt to be artifactual (Figure 41-7). Again, proper technique is necessary to make sure that the gain setting is not so high as to suggest internal echoes when none exist.¹⁴ Complicated cysts are managed based on the context. If it is an isolated finding or a finding that corresponds to a new palpable mass, aspiration would be undertaken to confirm its cystic nature and resolve the palpable finding. If it is seen in the presence of multiple other simple and/or complicated cysts, further evaluation may not be required. In this setting it is important to make sure that all of the cysts are circumscribed, that none contain internal blood flow and that there are no malignant features. Following multiple small complicated cysts with ultrasound is technically challenging and usually unnecessary.

Complex cysts in the breast are like complex cysts in the ovary—they contain cystic and solid elements (Figure 41-8). Generally, the concern is over a focal abnormality of the wall such as eccentric thickening (especially if convex inward) or nodularity. They may or may not have internal echoes, but usually have some degree of sound transmission. The concern here is over the possibility of an intracystic neoplasm, such as a papilloma or malignancy.¹⁷ In general, complicated cysts should not undergo aspiration as removal of the fluid component may make it difficult to identify any residual solid component. Complex cysts are usually biopsied, either with image-directed core or vacuum-assisted biopsy, or with excision.¹⁸ If image-directed biopsy is performed, a marking clip should be deposited in case pathology is positive and no sonographic residual remains.

Sonographic features are more variable for microcysts¹⁹ and clusters of cysts, especially when there is associated fibrosis (Figure 41-9). Incidental clusters of simple-appearing micro- and macrocysts can often be followed, but if there is any question of an associated soft tissue mass, biopsy should be considered, as the appearance of early invasive and intraductal malignancy can overlap with the appearance of some clustered microcysts.

Solid Masses

The first step with solid masses is characterization based on margins, echogenicity, sound transmission, and relationship to adjacent structures.

Benign features for solid masses include^2,16:

• Oval, round, or gently (macro) lobulated margins (Figure 41-10)

• Circumscribed (smooth) margins (Figure 41-10)

• Continuous thin capsule surrounding the nodule (Figure 41-10)

• Thin surrounding halo

• Long axis parallel to the skin ("wider than tall") (Figure 41-10)

• Uniform hyperechogenicity

• Adjacent structures displaced rather than distorted

• Dermal origin

• Reniform shape with echogenic hilum (lymph node)

Malignant features for solid masses include^2,16:

• Irregular, angular, indistinct, microlobulated, or spiculated margins (Figure 41-11)

• Long axis non-parallel to the skin ("taller than wide") (Figure 41-12)

• Associated microcalcifications (Figure 41-13)

• Extension of the lesion into adjacent ducts (Figure 41-14)

• Adjacent structures distorted

• Posterior acoustical shadowing (Figure 41-11)

A mass with any malignant feature should be regarded as suspicious and biopsy should be performed. An incidental mass with no malignant features, smooth or gently lobulated margins, and thin continuous capsule has less than a 2% risk of malignancy and can be followed.¹⁶ Most other solid masses are biopsied unless stability has been previously demonstrated.

Although there are particular features that may be associated with different benign solid masses (eg, papillomas, fibroadenomas, phyllodes tumors, etc), it is generally not necessary to predict a precise etiology as most will undergo biopsy. There are also different features that may be seen in different forms of malignancy, but, again, since biopsy will be performed, general predictions regarding cell type are usually not necessary.

Other Findings

Skin changes may be seen in a variety of inflammatory and infiltrative conditions and may consist of thickening, edema, and hypervascularity. The correct diagnosis is often made based on the clinical setting and other mammographic and sonographic findings. If there are inflammatory findings that do not resolve with appropriate antibiotic therapy, the patient should be referred for punch biopsy of the skin to exclude inflammatory breast cancer.

Some degree of duct dilatation is frequently seen in the subareolar regions, and internal echoes are frequently seen within these ducts secondary to proteinaceous material or debris. Findings that may indicate the need for additional evaluation include isolated ductal dilatation, intraductal nodules (Figure 41-15), or Doppler flow associated with any echogenic material within a distended duct.

Implants

There are several sonographic features that may suggest intracapsular or extracapsular rupture of silicone implants. Extracapsular rupture may be manifest as areas of dense irregular echogenicity with posterior shadowing and loss of normal tissue planes. This appearance is referred to as the "snowstorm" pattern. Intracapsular rupture may be seen as layers of implant shell collapsed into the intact collection of intracapsular silicone ("linguine" or "step ladder" signs). Implant characterization can be very complex as there are many different kinds of implants consisting of multiple lumens in various configurations. If there is any concern over implant integrity that cannot be resolved based on ultrasound or mammographic features, breast MRI should be considered.

In addition to standards established by the American College of Radiology, the American Institute of Ultrasound in Medicine (AIUM) has published guidelines for breast sonography. Not only do the AIUM guidelines cover indications, instrumentation, and technique as described above, they also refer to the qualifications and responsibilities of personnel, appropriate orders, exam documentation, and quality control.²⁰

Sonography has established itself as an important part of the breast imaging armamentarium. It is able to characterize a variety of clinical, mammographic, and MRI abnormalities, adding important information that frequently alters management. It is particularly well suited to the guidance of an increasing number of interventional procedures. As technical advances continue, we may see further expansion of breast ultrasound to include three-dimensional techniques, whole breast imaging, elastography,^21,22 and automated ultrasound²³-directed intervention.