Technical Considerations

Real-time hand-held scanners provide easy and rapid direct visualization of breast lesions for diagnosis or ultrasound-directed breast biopsy. Hand-held units should include a linear array, high-frequency transducer operating at a frequency of 7.5 to 10 MHz or greater, which provides good tissue penetration to 4 or 5 cm. All scanners should also include a marking system to document and annotate ultrasound images.

Technical issues must be overcome to obtain good sonographic image quality. Superficial lesions in the near field of the transducer may be distorted, but they can be imaged by using a high-frequency transducer or a soft fluid offset. The heterogeneity of breast tissue results in absorption of the ultrasound beam with increasing distance from the transducer. Extensive diffraction of the ultrasound beam leads to beam defocusing and poor image quality the further the lesion is from the transducer. Thus, accurate diagnosis of cysts and deep lesions depends on appropriate power, gain, and focal zone settings. Improper adjustments of any of these parameters can result in misdiagnosis by producing suboptimal images or artifactual echoes within simple cysts. Routine calibration of the unit and evaluation of the unit’s performance with a breast phantom help prevent these technical errors.

The amount of shadowing at normal breast tissue interfaces depends on the transducer’s diameter and the distance of the tissue from the transducer. Artificial shadowing can be caused by Cooper ligaments and normal breast structures, which usually angle up toward the skin and the transducer.

Flattening or compressing the breast tissue decreases the amount of tissue penetrated by the ultrasound beam and diminishes edge artifacts by straightening Cooper ligaments parallel to the transducer. To flatten the breast tissue in the upper outer quadrant, the patient is scanned supine with her hand behind her head in a posterior oblique position, with her back supported by a wedge. Facilities often use the sponge wedges used for positioning posterior oblique lumbar spine radiographs to help the patient achieve this position. For medial lesions, the patient lies flat on her back, which flattens the medial breast tissue. Sonographers use moderate compression during scanning by lightly but firmly pressing the breast with the transducer. This decreases the thickness of the tissue to be scanned, reduces beam absorption and defocusing, allows better penetration, and decreases shadowing from ligaments and glandular elements.

To ensure that the field of view includes all the breast tissue from the skin surface to the chest wall, the sonographer makes sure that the image includes the pectoralis muscle and chest wall at the bottom of the screen. The time-compensated gain (TCG) curve should be adjusted so that fat is uniformly gray from the subcutaneous tissues to the chest wall. This adjustment enables accurate evaluation of masses as cystic or solid at any depth in the breast. Incorrect settings that make the fat look anechoic may also make a solid mass look like an anechoic cyst.

Once the sonographer identifies an area of interest or a mass, he or she magnifies the finding to fill the monitor or screen appropriately because it is hard to see and analyze a lesion if it is too small on the screen. The sonographer then resets the focal zone, TCG curve, and depth-compensated gain (DCG) curves on the lesion to evaluate the finding’s shape, margins, and internal characteristics.

When scanning a breast with palpable findings, the sonographer asks the patient to point out the mass or symptomatic area to ensure evaluation of the area prompting investigation. This ensures that the patient’s area of concern is addressed and that the patient is more confident that her questioned area was investigated (because she pointed it out). If the patient is unsure of the location of the mass, the sonographer scans the quadrant or area requested by the referring physician on the order or requisition.

To scan palpable masses, the sonographer scans the palpable finding, then places a finger over the mass. The sonographer scans over the finger on the mass, and then removes the finger to scan only the palpable finding. This ensures that the palpable finding is in the field of view. Alternatively, the palpable finding can be trapped between two fingers. The sonographer scans the trapped mass between two fingers so that the mass does not roll out of the field of view from under the transducer.

The American College of Radiology (ACR) has made specific recommendations for ultrasound labeling. The sonographer labels each finding according to its location in right or left breast, quadrant or clock position, scan plane (radial or antiradial, longitudinal or transverse), and number of centimeters from the nipple, along with the sonographer’s initials (Box 5-1). The sonographer takes images of the mass with and without measuring calipers. Any other pertinent clinical information, such as whether the lesion is palpable, may also be helpful to note.

Normal Sonographic Breast Anatomy

The breast is composed of fibrous connective tissue (Cooper ligaments) arranged in a honeycomb-like structure surrounding the breast ducts and fat (Fig. 5-1A and B). The proportion of supporting stroma to glandular tissue varies widely in the normal population and depends on the patient’s age, parity, and hormonal status. In young women, breast tissue is composed of mostly dense fibroglandular tissue. In later years, dense tissue involutes into fat in varying degrees, producing a mixed fatty and dense breast or an all-fatty breast (see Fig. 5-1C to F).

Figure 5-1 Normal breast ultrasound images. A and B, Normal breast ultrasound scans in fatty, mixed, and dense breasts. Unlabeled (A) and labeled (B) ultrasounds of a normal fatty breast show the thin, white, superficial skin line (arrow); dark subcutaneous fat; dark fatty lobules separated by sharp thin Cooper ligaments; and the muscle and chest wall at the posterior aspect of the image. Note that the fat is uniformly gray throughout the image, and multiple fatty lobules are interspersed between the thin, linear Cooper ligaments. Unlabeled (C) and labeled (D) normal breast ultrasounds of fibroglandular and fatty breast tissue show the thin, white, superficial skin line; dark subcutaneous fat; and white glandular tissue interspersed by dark hypoechoic fatty lobules. Note how the white glandular tissue is thicker than the Cooper ligaments seen in A and how the fatty islands might be mistaken for breast masses. Unlabeled (E) and labeled (F) ultrasounds of a normal dense breast show mostly white glandular tissue with scant amounts of fat and hypoechoic ducts over an area of thickening in a young patient. Note how the ducts appear as small, round, dark structures interspersed in the dense glandular tissue when caught in cross-section and appear like long tubes when caught in longitudinal section. Unlabeled (G) and labeled (H) ultrasounds of a normal lymph node show a lobulated hypoechoic mass with an echogenic center that represents the fatty hilum. Note that the fat outside the lymph node in the adjacent breast tissue is hypoechoic (dark), whereas fat inside the lymph node is echogenic (white). This is the typical appearance of a normal lymph node. I and J, Cooper ligaments and normal fatty tissue in a thin fatty breast. Unlabeled (I) and labeled (J) normal breast ultrasounds show the thin echogenic skin line at the top of the image; dark hypoechoic subcutaneous fat; thin, gently curving Cooper ligaments coursing through the fat; and the thin, parallel, tightly packed lines of muscle just above the chest wall and the rib. K and L, Landscape ultrasound view of normal dense tissue in a thin dense breast. Unlabeled (K) and labeled (L) landscape ultrasounds over a thin dense breast show the echogenic skin line, dark subcutaneous fat, dense white glandular tissue, and the pectoralis muscle and chest wall overlying the periodic round shadows of the ribs and intercostal muscles. The ribs cause acoustic shadowing. However, the ribs should not be mistaken for breast masses because they are located behind the muscle and chest wall just above the lung and are not within the breast.

Breast tissues are either echogenic (white) or hypoechoic (black) on ultrasound. The skin is an echogenic line immediately under the transducer in the near field. It is normally about 2 to 3 mm thick and has a hypoechoic layer of dark subcutaneous fat immediately beneath it (Box 5-2). Unlike echogenic or white-appearing fat around the superior mesenteric artery in the abdomen, fat in the breast appears dark or hypoechoic. The only exception to hypoechoic fat in the breast is the echogenic fat in the middle of a lymph node. The normal lymph node is an oval, well-circumscribed mass with a hypoechoic cortex and fatty echogenic hilum, often seen in the upper outer quadrant and axilla, and often near an artery (see Fig. 5-1G and H).

Breast glandular tissue and connective tissue are echogenic or white. Connective tissue has the highest acoustic impedance, fat has the lowest, and glandular parenchyma is of intermediate echogenicity. The Cooper ligaments are thin, sharply defined linear structures that support the surrounding fat and glandular elements (see Fig. 5-1I and J). Cooper ligaments in a fatty breast look like thin, white, gently curving lines surrounding hypoechoic fat. Normally, Cooper ligaments are thin and sharply demarcated. In breast edema, the fat becomes gray and the normally sharp Cooper ligaments become blurred.

Subareolar ducts are dark, hypoechoic, tubular structures leading to the nipple. The glandular tissue elements are echogenic (white), so normal hypoechoic ducts appear like dark tubes against the normal white background when imaged along their long axis. In cross-section, the ducts are dark, hypoechoic, round or oval circles seen against the white echogenic normal glandular tissue.

The pectoralis muscle is a hypoechoic structure of varying thickness that contains thin lines of supporting stroma coursing along its long axis at the chest wall near the bottom of the image. The pectoralis muscle abuts the intercostal muscles and fascia of the chest wall (see Fig. 5-1K and L). Ribs in between the intercostal muscles are round or oval in cross-section, shadow intensely, and are seen at regular intervals along the chest wall. High-resolution transducers may display calcifications in the anterior portions of the cartilaginous elements of the ribs. Newcomers to breast ultrasound may mistake the ribs for masses, but their periodicity along the chest wall and the fact that one can palpate the ribs along their course will help prevent newcomers from making this mistake.

The nipple is a hypoechoic structure at the skin surface that occasionally produces an intense acoustic shadow as a result of the dense connective tissue within it (Fig. 5-2A and B). Because of the presence of retroareolar ducts and blood vessels, there may be marked vascularity in the retroareolar region on color or power Doppler imaging. Newcomers to breast ultrasound may mistake the nipple for a breast mass because of its hypoechoic appearance, shadowing, and the intense vascularity beneath it. However, knowledge of the shadowing, vascularity, and the ability to correlate the mass with the nipple on physical examination will help prevent newcomers from making this mistake.

Figure 5-2 The nipple and adjacent breast mass. A, Ultrasound shows a hypoechoic, shadowing nipple in the near field (arrow) separated from an irregular hypoechoic mass (double arrows). Note that the nipple is at the skin surface and can be easily distinguished from a breast mass once the normal nipple appearance is recognized. The mass was an invasive ductal cancer. B, Ultrasound of a normal nipple shows that the flattened nipple appears as a hypoechoic mass, sometimes with intense shadowing. C, This 5-year-old girl had a breast lump under the nipple, representing the normal breast bud, that was asymmetric on the right side. D, Breast ultrasound shows the normal hypoechoic nipple–breast bud complex on the enlarged side and the smaller nipple–breast bud complex on the normal side.

In children, the breast bud that develops into the adult breast is right underneath the nipple. The breast bud may produce an asymmetric lump under the nipple that may be mistaken for a mass rather than a normal developing structure (see Fig. 5-2C and D). This normal structure should be left alone because surgical removal of the breast bud results in no breast formation on the ipsilateral side.

Ultrasound Evaluation of Mammographically Detected Findings

The ACR Breast Imaging Reporting and Data System (BI-RADS®) committee developed an ultrasound lexicon to provide descriptors for findings seen by ultrasound and recommended specific descriptors for breast masses (Table 5-1). Use of the words in the ACR BI-RADS® lexicon helps clarify one’s impression of the finding, improves communication between the radiologist and referring physician, and may trigger specific patient managements. This is because specific ultrasound features described by the lexicon suggest either benign masses or cancer. Although there is some overlap in benign versus malignant ultrasound features, the radiologist can use the lexicon to be reminded of what features should be searched for on the image (Table 5-2).

The BI-RADS® ultrasound lexicon descriptors for breast masses and their effect on the surrounding breast tissue are illustrated in Figures 5-3 to 5-6. Mass shapes are reported as oval, round, or irregular. Mass margins are circumscribed, angular, indistinct, microlobulated, or spiculated. The internal echo pattern is described as anechoic (all black inside), hyperechoic (white), complex (mixed black and white), isoechoic (equal), or hypoechoic (dark). Posterior acoustic features are described as no posterior acoustic features, enhancement (white), shadowing (dark), or a combined pattern. The boundary between the mass and the surrounding tissue is described as having an abrupt interface or as containing an echogenic halo (a white blurry band surrounding the mass). Calcifications are described as no calcifications, macrocalcifications (>0.5 mm), microcalcifications within the mass, or microcalcifications outside the mass. Effects of the mass on surrounding breast tissue are described using the terms no effect, duct changes, changes in Cooper ligaments, edema, architectural distortion, skin thickening, skin retraction, and skin irregularity.

Figure 5-3 Examples of BI-RADS® ultrasound lexicon descriptors for mass shape and margins. A, An oval, deeply situated, well-circumscribed mass has homogeneous internal characteristics and no posterior acoustic enhancement. This was a complicated cyst at biopsy. B, A suspicious, round, microlobulated mass with well-circumscribed borders has complex hypoechoic internal characteristics and acoustic enhancement; this is an unusual appearance of a breast metastasis. C, Irregularly shaped invasive ductal cancer with angular margins (arrows) and an echogenic halo (double arrows) in its anterior surface. D, Round, irregular, hypoechoic invasive ductal cancer (double arrows) with acoustic spiculations at its anterior surface (arrows). Note that when the mass is spiculated and displayed against the dark fat, the acoustic spiculations are white. E, Round mass displayed against the white glandular tissue with indistinct margins (arrows). Contrast the indistinct margins to the well-circumscribed margins in part A; part E represents invasive ductal cancer.

Figure 5-4 Examples of BI-RADS® ultrasound lexicon descriptors for internal echo pattern. A, This cyst is anechoic, or completely black inside, and has the typical appearance of a cyst; that is, it is well-circumscribed, oval, and has an imperceptible back wall and enhanced through-sound transmission. B, Oval, well-circumscribed, hyperechoic mass in the near field (arrows) is whiter than the surrounding dark hypoechoic fat. Biopsy showed angiolipoma. C, A taller than wide, irregularly shaped invasive ductal cancer has a well-circumscribed superficial border; the lateral and posterior borders are indistinct. Internally there are bright and dark shadows compatible with a complex internal echo pattern. D and E, An oval, wider than tall, isoechoic breast mass has well-circumscribed borders with a tapering lateral margin; note that the internal echogenicity is the same as the surrounding breast tissue, making this mass isoechoic. Biopsy showed fibroadenoma. F, This hypoechoic invasive ductal cancer has mixed internal echogenicity within it, angulated anterior margins, and a combined posterior acoustic pattern comprised of both acoustic shadowing and enhancement.

Figure 5-5 Examples of BI-RADS® ultrasound lexicon descriptors for posterior acoustic features. A, An oval, lobular, hypoechoic mass with thin septations deep within the breast has no posterior acoustic features. This was a cyst at aspiration. B, Two cysts adjacent to each other with a thick septation between them have posterior acoustic enhancement, which is the white enhancement of sound deep to the cysts (arrows). C, A taller than wide, not parallel, hypoechoic invasive ductal cancer shows as a mass (arrow) that has an irregular shape, indistinct margins, an echogenic halo, and marked acoustic shadowing (double arrows). Note that the mass itself is difficult to distinguish from the acoustic shadowing because both are so dark. Note also that the acoustic shadowing, which is the dark, black band posterior to the mass, can be caused by productive fibrosis from the tumor, and is often seen in masses that appear spiculated on the mammogram. D, Magnified craniocaudal mammogram shows a spiculated mass (arrow) corresponding to the hypoechoic shadowing cancer. Note that the cancer is far from the linear scar marker (double arrows) and cannot represent a postbiopsy scar.

Figure 5-6 Examples of BI-RADS® ultrasound lexicon descriptors for tissue interface and growth pattern. A, Benign masses usually have an abrupt interface with the surrounding tissue. This benign fibroadenoma is oval, well-circumscribed, and fairly homogeneous. It is parallel, wider than tall, and has an abrupt interface with the surrounding glandular tissue. Note the sharp margins compared to part B. B, An indistinct white band surrounding a mass is an echogenic halo in the BI-RADS® lexicon and is a sign of cancer. There is an echogenic halo in the superior portion of this invasive ductal cancer (arrow). The cancer itself is hypoechoic, taller than wide, and has indistinct margins. C, This fibroadenoma is a well-circumscribed, oval, parallel (wider than tall) mass and has homogeneous internal characteristics. Note that the parallel mass respects adjacent tissue planes and does not cross Cooper ligaments, which is typical for benign masses. D, This hypoechoic, spiculated invasive ductal cancer with an echogenic halo is taller than wide (not parallel), meaning that it is growing through Cooper ligaments perpendicular to the transducer surface. The mass is most likely malignant because it is growing through tissue planes. The not parallel or taller than wide sign on ultrasound is always worrisome for cancer. E, Schematic of taller than wide malignant versus wider than tall benign mass growth patterns on ultrasound.

The terms parallel or not parallel relate to tumor growth patterns with respect to normal tissue planes. They are important because they indicate if the mass is growing along or in between tissue planes versus growing through them. A parallel growth pattern indicates a benign finding (wider than tall, as described by Stavros and colleagues) because it indicates a growth pattern along tissue planes. Not parallel or taller than wide indicates that the mass is growing through the normal tissue planes, which is not normal and indicates cancer (see Fig. 5-6E).

Finally, the ultrasound BI-RADS® lexicon suggests standard reporting for masses, as in Box 5-3.

Breast Cysts, Intracystic Tumors, and Cystic-Appearing Masses

The most frequent clinical application of breast ultrasonography is to characterize masses initially detected by mammography as cystic or solid. Cysts are the most common breast mass and occur in an estimated 7% to 10% of all women. Cysts are lined by apocrine cells that actively secrete material, predisposing these types of cysts to recur after aspiration. Sometimes, cysts are lined by a flat epithelial lining that is less active. The accuracy of ultrasound in distinguishing cystic from solid masses can be as high as 98% to 100%, as reported by Hilton and colleagues.

Strict ultrasound criteria for a simple cyst include a mass with well-circumscribed margins, sharp imperceptible anterior and posterior walls, a round or oval contour, absence of internal echoes, and posterior acoustic enhancement (Box 5-4 and Fig. 5-7A). Cysts may be single or multiple, gathered into small clusters, or contain thin septations (see Fig. 5-7B to D). Cysts are not malignant or premalignant, but examination of them is important because they may cause lumps that mimic round cancer on physical examination or mammography. When palpable, a cyst is a smooth, mobile mass on physical examination. Occasionally cysts appear as a visible mass if the patient is supine and the cyst is large. Cysts may be painful and may wax and wane with the patient’s menstrual cycle. If a mass is proven to be a cyst by ultrasound, the patient can be monitored by screening mammography because cysts are not cancer. Symptomatic cysts that are painful or cause a lump that disturbs the patient can be treated by aspiration. Cysts may be simple or “complicated,” meaning that the cyst contains sloughed debris. These complicated cysts contain material within them rather than being anechoic. Some complicated cysts require aspiration to confirm that they are cysts rather than solid masses (see Fig. 5-7E).

Box 5-4 Simple Cyst Criteria

Oval or round shape with circumscribed margins

Anechoic

Imperceptible back wall

Enhanced transmission of sound

Figure 5-7 Examples of breast cysts. A, This simple cyst has an anechoic interior, an imperceptible wall, sharply marginated smooth borders, and enhanced transmission of sound. B, This simple cyst is lobulated and contains thin septations and imperceptible walls. C, A typical anechoic simple cyst with imperceptible walls and enhanced through-sound transmission. D, Two typical anechoic simple cysts that are sharply marginated with imperceptible walls and enhanced through-sound transmission. E, Two cysts adjacent to each other include a simple cyst (arrow) with lobulated borders that is anechoic and has an imperceptible wall and enhanced through-sound transmission. Adjacent to the simple cyst are complicated cysts (double arrows) that contain debris and have indistinct margins and enhanced through-sound transmission; needle aspiration showed simple cyst fluid. F, This simple cyst does not show enhanced through-sound transmission because of focal zones that are located too deep for proper evaluation of the cyst. G, Acorn cyst. An oval, well-circumscribed mass has a lower cystic component and a possible solid upper component, comprising the “acorn cyst,” which has a fluid/fluid level (arrow). Aspiration showed a cyst with debris in it.

Attention to technical detail is especially important because increasing the TCG curve may produce artifactual echoes in benign cysts that suggest a solid mass. An improperly set DCG curve may inaccurately evaluate the internal matrix of the mass and make a cyst look solid and may make a solid mass appear to be a cyst. On real-time imaging, cyst contours may be flattened with compression, whereas solid masses are less compressible. Alternatively, small, clustered, or deeply located cysts may be at the technical limits of ultrasound to distinguish the usually anechoic cyst from a solid mass.

Deeply located cysts may not show enhanced through-sound transmission because of their location close to the chest wall, and lateral cyst walls may be obscured by refractive shadows (see Fig. 5-7F). These problems may be resolved by repositioning the patient or the transducer to scan from a different angle. This permits visualization of distal acoustic enhancement or eliminates the refractive shadows obscuring the sharp cyst walls. Acorn cysts contain a fluid/fluid level, with the dark dependent portion of the cyst representing clear fluid (the acorn) and the lighter top representing layering fluid above it (the acorn cap) (see Fig. 5-7G). Changing the patient’s position may cause the layer to move dependently, clinching the diagnosis of an acorn cyst.

The internal characteristics of cysts must be analyzed to exclude mural masses or irregular thick walls, which indicate complex masses. Complex masses contain cystic and solid components; intracystic tumors and necrotic neoplasms are in the differential diagnosis. Complex masses are different from complicated cysts, which contain debris. Complex masses might be cancer, but complicated cysts are benign (Table 5-3 and Boxes 5-5 and 5-6).

Table 5-3 BI-RADS® Ultrasound Special Cases (Cystic)

Cystic Mass Type	Description	Differential Diagnosis
Clustered microcysts	2- to 3-mm cysts Thin (<0.5 mm) septations No solid component

Benign Complicated cysts

Cyst with debris or fluid/debris level

Debris may shift at real-time

Not pus or blood

Modified from American College of Radiology: ACR BI-RADS®—ultrasound, In ACR Breast Imaging and Reporting and Data System, breast imaging atlas, Reston, VA, 2003, American College of Radiology.

Real-time ultrasound imaging can help distinguish speckle artifact from debris in cyst fluid from a solid mass. Real-time ultrasound can show particulate matter slowly moving inside the cyst. On real-time imaging, the debris causes speckle artifact, which swirls in the cyst like fake snow in a snow globe (Fig. 5-8A to C). Placing the patient in the decubitus position can cause a difference in the sedimentation pattern in the complicated cyst, but not always. Color Doppler or power Doppler ultrasound can detect movement of particulate matter within complicated breast cysts or blood vessels in solid masses (see Fig. 5-8D and E). Doppler imaging will show no blood vessels in breast cysts. Unfortunately, the absence of blood flow in a mass is not diagnostic of a cyst because Doppler imaging does not always detect blood flow in solid masses or even in cancers.

Figure 5-8 Complicated cysts and Doppler ultrasound. A, This complicated cyst had debris within it that did not move at real-time imaging. Color Doppler ultrasound shows pulsating blood vessels around the cyst but not within it. Aspiration yielded cloudy fluid. B, Ultrasound of a palpable mass showed speckle artifact within a complicated cyst that moved at real-time imaging. C, Color Doppler ultrasound detected movement of the debris in real time. In contradistinction, color Doppler (D) and power Doppler (E) imaging show a pulsating vessel flowing into a round cancer, thus differentiating it from complicated cysts.

In everyday clinical practice, cysts do not always fulfill all the strict sonographic criteria for cysts because of a variety of technical factors, or they may contain echoes from debris within the fluid. Posterior enhanced through-transmission of sound was not seen on all images in 25% of 80 cysts reported by Hilton and colleagues. Internal cyst echoes may be produced by reverberation artifacts, although the near-field reverberations may be reduced by scanning through an offset. However, Berg and colleagues (2003, 2005) and other researchers have shown that complicated cysts with low-level internal echoes, no mural masses, thin walls, and thin septations rarely represent cancer and can either be monitored or aspirated with little or no morbidity.

Differentiation of complicated cysts from benign or malignant cystic masses can be tricky (see Table 5-3). Some cysts contain true internal echoes as a result of thick tenacious fluid or hemorrhage from previous aspirations. Some cysts have thick walls as a result of inflammation from cyst fluid leaking into the surrounding tissues. In cases in which all the sonographic criteria of a simple cyst have not been met, fine-needle aspiration may obviate the need for core needle biopsy or surgical biopsy. Once the needle is within the mass, the presence of cyst fluid rather than solid tissue can be confirmed by moving the needle, as suggested by Stavros and colleagues (Fig. 5-9). Cysts that do not fulfill all criteria for simple cysts, in the right clinical setting, require aspiration (see Fig. 5-9C to E). Cyst fluid should be sent for cytologic analysis if it is bloody, if there is an intracystic mass on ultrasound or pneumocystography, or if the patient has had prior intracystic carcinoma. Clear cyst fluid can be discarded if there are no clinical factors that would require cytologic examination.

Figure 5-9 Simple cyst fluid compared with cyst fluid containing debris. A, The complicated cyst fluid in the left test tube, which contains debris consisting of particulate matter, causes speckles within the cyst on ultrasound. The simple cyst fluid in the right test tube is clear and contains no debris. Both fluids are normal and may be discarded after aspiration. B, A simple way to tell whether a mass is cystic at needle biopsy is to move the needle tip within the mass. In a complicated cyst, the needle will move within the mass. In a solid mass, the needle will move the mass up and down. C, Complicated cyst. This patient felt a palpable mass; ultrasound shows an oval, hypoechoic mass with sharp walls and enhanced through-sound transmission, suggesting cystic or solid material. Doppler ultrasound showed no flow. Aspiration showed yellow fluid with debris within it compatible with a cyst. D and E, Deeply located complicated cyst and simple cyst. Ultrasound shows an oval, well-circumscribed hypoechoic mass with an imperceptible wall anterior to a simple cyst near the chest wall. Despite scanning from several angles, the hypoechoic mass could not be shown to be a simple cyst and Doppler imaging showed no flow. Fine-needle aspiration under ultrasound guidance showed cystic fluid from both masses. The anterior mass represented a complicated cyst and the deeper mass represented a simple cyst.

On the other hand, complex cystic masses (i.e., fluid-filled masses with thick walls or mural projections) require biopsy to exclude the rare intracystic papilloma, intracystic carcinoma, phyllodes tumor with a marked cystic component, or solid cancers with central necrosis. Other complex masses include hematoma, abscess, galactocele, and seroma; management of these masses is based on their appearance and the clinical situation (Fig. 5-10).

Figure 5-10 Complex cystic masses that contain both anechoic (cystic) and echogenic (solid) components. This large palpable mass, in a patient with diabetes, has an anechoic center and a solid, irregular rim in transverse (A) and longitudinal (B) scans. C, Doppler ultrasound showed pulsation in the rim. Another complex cystic mass is oval and has small cystic structures with some solid components on transverse (D) and longitudinal (E) scans. This was a phyllodes tumor. Phyllodes tumors have solid leaflike components floating in small cystic structures, which give the tumor its name and account for the cystic and solid appearance in this image. The cystic versus solid components vary according to each tumor.

Intracystic carcinomas are a rare subgroup of tumors that arise from the walls of a cyst; they represent 0.5% to 1.3% of all breast cancers (Fig. 5-11A). These tumors have a better prognosis than other malignant breast neoplasms do. On ultrasound, intracystic carcinomas often appear as solid mural excrescences projecting into the cyst fluid. Differentiation of intracystic carcinoma from benign intracystic papilloma is not possible, and surgical biopsy is thus necessary. The finding of a mural nodule within a cyst has the differential of an intracystic carcinoma, papilloma, a cyst with debris, and reverberations in a simple cyst produced by high gain settings (see Fig. 5-11B to D). Color or power Doppler imaging may be helpful if a blood vessel can be identified in the intracystic mass.

Figure 5-11 Intracystic tumors. A, An intracystic carcinoma is shown as a mural mass projecting into the fluid-filled center of a cyst on ultrasound. B, In another patient, an intracystic papilloma is shown as a mass surrounded by air after fluid aspiration and pneumocystography. C, In a third patient, an apparent intracystic mass was shown to represent debris stuck to the side wall of a complicated cyst at aspiration. D, This oval, well-circumscribed cystic mass has a round, solid mural nodule. Because it is not completely anechoic, it cannot be considered a simple cyst and is a complex mass; the differential is intracystic cancer papilloma or debris in a cyst. Aspiration was performed, showing cystic fluid and debris without atypia or cancer on cytology.

Benign Solid Masses: Fibroadenoma and Fatty Pseudolesions

Fibroadenomas arise from breast lobules and are the most common solid benign masses in women younger than age 30 years. Once diagnosed, fibroadenomas may remain stable in 80% of cases, regress in about 15%, and grow in 5% to 10%. Fibroadenomas are benign, although cancer can occur within a fibroadenoma. Women with a specific histologic diagnosis of complex fibroadenomas have cysts or histologic elements other than the fibroadenoma and have a small increased risk of future breast cancer, as described by DuPont and colleagues. Fibroadenomas may be single or multiple and are called giant fibroadenomas if larger than 8 cm.

On ultrasound, Cole-Beuglet and colleagues describe typical fibroadenomas as solid masses with well-circumscribed, round or oval borders and containing weak low-level homogeneous internal echoes with enhanced, decreased, or unchanged sound transmission. Stavros and colleagues and Fornage and colleagues have described fibroadenomas as smooth, wider than tall solid masses. Stavros and colleagues further characterize fibroadenomas as having at most four gentle lobulations and homogeneous internal echo texture (Box 5-7 and Fig. 5-12A to C).

Figure 5-12 Typical fibroadenomas. A, Ultrasound of a typical fibroadenoma shows a mass that is wider than tall, smooth, and homogeneous with edge refraction. B, Nonpalpable, oval, homogeneous, sharply defined typical fibroadenoma. C, Oval, homogeneous, well-circumscribed fibroadenoma in a third patient is well-displayed against the echogenic glandular tissue. D, Ultrasound of another fibroadenoma shows an atypical round shape containing calcifications. E, The corresponding mammogram shows a well-circumscribed, dense smooth mass with peripheral coarse calcifications. Biopsy revealed fibroadenoma. F, Ultrasound of a papillary cancer mimicking a typical fibroadenoma shows a mass that is wider than tall, smooth, and homogeneous with a few gentle lobulations. G, The corresponding mammogram shows the cancer as a dense, well-circumscribed, lobulated mass. H, Ultrasound of another papillary cancer shows a wider than tall homogeneous mass with more than three gentle lobulations, findings suggestive of cancer and not easily mistaken for fibroadenoma. I, Importance of correlating ultrasound with the mammogram. Ultrasound of a calcified fibroadenoma mimicking a typical cancer shows a calcified, intensely shadowing mass that is suspicious for carcinoma. J, The corresponding mammogram shows that the mass in I has typical peripheral popcorn-like calcifications in a well-circumscribed lobulated fibroadenoma, among other typical calcifying fibroadenomas in the same breast, and biopsy was avoided. K and L, Importance of evaluating a mass in orthogonal planes. A longitudinal ultrasound (K) shows a mostly round, well-circumscribed mass. However, note that on the transverse image (L), the mass is oval, wider than tall, and well-circumscribed. Biopsy showed fibroadenoma. M to O, Lipomas on mammography and ultrasound. M, Cropped craniocaudal mammogram shows markers over fatty palpable masses that have thin rims. N, Magnified cropped mammogram shows the masses to greater advantage. O, Ultrasound shows round, well-circumscribed masses with an abrupt interface and fatty echogenicity compatible with lipomas.

Fibroadenoma appearances, however, can be highly variable (see Fig. 5-12D and E). Fibroadenomas may occasionally display irregular margins, inhomogeneous echo texture, lobulated borders, or posterior acoustic shadowing. These atypical features result in biopsy of the fibroadenoma to exclude cancer.

Stavros and colleagues also described specific ultrasound features of benign solid masses (see Box 5-7): smooth margins with fewer than four gentle lobulations, intense homogeneous hyperechogenicity, thin echogenic pseudocapsule, wider than tall elongated appearance, and no malignant sonographic signs. Suspicious ultrasound findings include acoustic shadowing, microlobulation, microcalcifications, ductal extension, angulated margins, or a very hypoechoic pattern (Box 5-8). In their 1995 study, Stavros and colleagues compared large-core needle or surgical biopsy pathology to prospectively determined ultrasound features of solid breast masses to see if the ultrasound criteria could predict malignancy. When the sonographic findings were benign by their criteria, the results yielded 424 true negatives and 2 false negatives. The negative predictive value was 99.5% and the sensitivity was 98.4% with strict adherence to their benign ultrasound features. However, it is also known that some well-circumscribed carcinomas may simulate fibroadenomas and should undergo biopsy if new or in the appropriate clinical setting (see Fig. 5-12H to J) because not all round or oval solid masses are benign (Box 5-9).

After finding a mass by ultrasound, it is important to re-evaluate the mammogram to make sure the ultrasound findings represent the mammographic findings on the film. The mammogram may provide important clues to the correct diagnosis. For example, if there are calcifications in a mass on ultrasound, the mass may be a typical calcifying fibroadenoma, benign fat necrosis, or calcifying cancer. By placing a skin marker over the ultrasound finding and retaking a mammogram, one can analyze the calcifications further and make a diagnosis. This is especially true if the calcifications found by ultrasound show up on the mammogram as typical for fibroadenoma (popcorn-type) or cancer (pleomorphic) (see Fig. 5-9K and L).

It is also important to scan masses in orthogonal planes and at multiple angles to see all the borders. A mass that appears round in one plane might actually be the cross-section of an oval mass (see Fig 5-9M and N). Scanning carefully in multiple planes allows the sonographer to evaluate both the true mass shape and all the margins of the mass.

Palpable pseudolesions can be produced by fatty deposits or oil cysts in the breast. A mammogram taken with a skin marker on the palpable mass should clarify that the mass is actually a fatty deposit or an oil cyst by showing only fat under the marker. When a fatty pseudomass is scanned from various projections, the masslike appearance of the fatty lobule should blend into the surrounding tissue and lack three-dimensional features. Oil cysts are well-circumscribed cystic or fat echogenic masses on ultrasound. If unsure if an ultrasound finding represents a fatty lobule, the sonographer can place a skin marker over the ultrasound finding and repeat the mammogram (see Fig. 5-9O to Q). If the finding is a fatty mass, there should be only fat under the skin marker.

Other benign solid breast masses include papillomas, hamartomas, lymph nodes, and healed postsurgical scars. In these various benign conditions, the clinical setting and mammographic appearance usually help identify the true nature of the lesion because sonographic features alone are rarely diagnostic (except in the case of lymph nodes). Examples of these masses are shown in Chapter 4.

Malignant Solid Masses

Cancers are generally hypoechoic relative to the brightly echogenic normal fibroglandular tissue. They often have irregular or round shapes and angulated or spiculated margins. Cancers often can show invasion by extending through normal breast planes (“taller than wide”) and may have an echogenic halo. They can show posterior acoustic shadowing, which is a dark band posterior to the mass that is reported to occur in 60% to 97% of spiculated carcinomas. Posterior acoustic shadowing is extremely suspicious for cancer and is thought to relate to fibrosis or collagen associated with the tumor. Acoustic shadowing is different from edge shadowing, which is an artifact caused by the edge of the mass against normal breast tissue. To distinguish edge shadowing from true acoustic shadowing, the sonographer scans from different planes. Edge shadowing will not persist in all planes, but true acoustic shadowing will persist. Examples of breast cancers on ultrasound are shown in Figures 5-13 to 5-15.

Figure 5-13 Typical breast cancers on ultrasound. Transverse (A) and longitudinal (B) scans show a typical invasive ductal cancer: very hypoechoic relative to normal breast tissue and surrounded by a fuzzy echogenic halo. The irregular shape, angulated margins, invasive growth through normal tissue planes, and taller than wide configuration are all suspicious for cancer. Note that acoustic shadowing is absent in this cancer and that shadowing is not necessary to make the diagnosis. C, Ultrasound of a grade II invasive ductal carcinoma shows a very hypoechoic indistinct irregular mass with marked acoustic shadowing. D, Invasive ductal carcinoma as irregular hypoechoic mass with enhanced through-sound transmission.

Figure 5-14 A, Schematic drawing of acoustic shadowing on the left with variations of shadowing and acoustic enhancement on the right. B, In comparison, the artifact of edge refraction will be seen only at the edge of a mass, where it meets the surrounding breast tissue, and not in the middle of the mass. C to F, Acoustic shadowing in cancer. C, Radio ultrasound shows a taller than wide, very hypoechoic, spiculated mass with acoustic shadowing, representing invasive ductal cancer. D, Doppler ultrasound shows a pulsating vessel within the cancer. To correlate the ultrasound mass with mammographic findings, the technologist placed a metallic marker over the mass and repeated the mammogram. Right digital mammograms show the marker in close proximity to a spiculated mass in an otherwise fatty breast on the lateral view (E), but it is at least 2 cm away from the mass on the craniocaudal view (F). This is because the skin containing the BB compressed away from mass in a projection, which is not uncommon when using skin markers to correlate mammograms to ultrasound findings. The mass was invasive ductal cancer.

Figure 5-15 Features of breast cancers without acoustic shadowing on ultrasound. A, Ultrasound of an invasive ductal carcinoma shows a very hypoechoic irregular mass with an echogenic halo near an implant. B, Ultrasound of a grade II invasive ductal cancer shows a very hypoechoic microlobulated mass with a small blood vessel adjacent to it. C, The corresponding mammogram shows a round mass, perhaps with microlobulations. D, Ultrasound of an extensive breast cancer, shown as multiple nodules on the mammogram, reveals an irregular, hypoechoic, oval mass corresponding to an extensive multifocal invasive ductal cancer without acoustic shadowing. E, In the same patient, another irregular, hypoechoic mass without acoustic shadowing represents a different invasive ductal cancer. F, Doppler ultrasound shows a large pulsating artery within the cancer. G, Contrast two foci of invasive ductal cancer on the same image; one is an irregular mass without acoustic shadowing (double arrows) and the other is shown only as acoustic shadowing (arrows). H, Suspicious mass on longitudinal scan shows a tiny irregular mass with an echogenic halo and no acoustic shadowing; this was invasive ductal cancer.

The ACR BI-RADS® ultrasound lexicon and terms developed by Stavros and colleagues include features that suggest cancer, such as margins that are angulated, indistinct, microlobulated, or spiculated; acoustic shadowing; microcalcifications; ductal extension; an echogenic halo; and a taller than wide configuration (see Box 5-8 and Table 5-2).

However, one cannot always be sure that round, well-circumscribed masses are benign. Unfortunately, round circumscribed solid cancers simulate benign breast masses on ultrasound. The most common round breast cancer is invasive ductal cancer (Box 5-10). The round form of invasive ductal cancer is an uncommon form of the most common cancer. Although the round, circumscribed form is uncommon for invasive ductal cancer, invasive ductal cancers are so common that a round cancer is statistically likely to be invasive ductal cancer (Fig. 5-16A). Other round, circumscribed malignancies include medullary, solid papillary, and colloid (mucinous) carcinoma. These cancers can simulate benign fibroadenomas on ultrasound by appearing round or oval in shape with enhanced transmission of sound (see Fig. 5-16B and C).

Figure 5-16 Breast cancer mimicking benign masses on ultrasound. A, A well-circumscribed, oval, hypoechoic mass on ultrasound is an invasive ductal cancer that was new on the mammogram from the previous year. Although medullary and mucinous cancers are often round, invasive ductal cancer is more common (accounting for about 90% of all breast cancers) and is the most common round cancer. B, Mucinous cancer mimicking a complicated cyst on ultrasound. Ultrasound of a palpable mass shows an oval, circumscribed, complex cystic mass with rather thick septa. Correlation with the mammogram (C) shows a dense, slightly irregular mass that would be unusual for a multilobulated cyst. Biopsy revealed mucinous cancer. D, Medullary cancer. Ultrasound of a palpable mass in a young woman shows a round mass with acoustic enhancement. Note that some of the borders are circumscribed on one scan plane, but one side of the mass has three spiculations, thus demonstrating the importance of scanning all the margins of the mass in more than one plane.

Careful attention to the margins of a mass may prompt biopsy and suggest cancer when, at first glance, the margins appear to be circumscribed. The borders of some tumors may appear circumscribed in one scan plane but irregular in another plane, so it is important to scan in multiple planes to evaluate all mass borders for margin irregularity (see Fig. 5-16D).

Necrosis or mucin within cancers may produce anechoic regions, resulting in posterior acoustic enhancement of sound, thereby mimicking the enhanced transmission of sound seen in cysts. Thus, some benign sonographic features of fibroadenomas and benign complex cystic lesions can also be seen in round malignancies (Table 5-4).

Table 5-4 Unfortunate Ultrasound Look-Alikes

Some of the larger calcifications in breast cancer may be seen by ultrasonography, but this important diagnostic sign is not generally visualized with regularity by ultrasound (Fig. 5-17A and B).

Figure 5-17 Features of malignancy on ultrasound. A, This invasive ductal cancer has microlobulated borders and bright internal echoes, representing calcifications within the tumor. B, A landscape overview shows the invasive ductal cancer of part A with its internal calcifications and a smaller second tumor slightly inferior and lateral to it. C to G, Axillary metastasis. C, A mediolateral oblique (MLO) mammogram shows a dense breast with an abnormal dense lymph node in the axilla. D, Ultrasound of a palpable mass in the breast tissue on the ipsilateral side, not seen on the mammogram, shows a circumscribed lobulated mass with enhanced transmission of sound that represents an invasive ductal cancer. E, Ultrasound of the abnormal lymph node in the axilla shows a hypoechoic mass without a fatty hilum that represents lymphadenopathy from the metastasis. Contrast the abnormal lymph node with the normal lymph node shown in Figure 5-1G. F, On the MLO mammogram, a huge dense abnormal lymph node is evident in the axilla. G, Ultrasound shows a hypoechoic mass so homogeneous that it almost looks like a cyst. Biopsy findings were consistent with lymphoma. H, A hypoechoic, irregular invasive ductal cancer with angulated margins is not parallel and has indistinct borders. I to K, Inflammatory cancer in a male with red skin nodules. I, MLO mammograms show bilateral gynecomastia with an irregular right retroareolar mass, skin and nipple thickening, and a second mass in the axilla. J, Ultrasound shows a hypoechoic spiculated and angulated retroareolar mass with acoustic shadowing and thickening of the areola–nipple complex. K, Ultrasound of the skin changes and red nodules shows marked skin thickening and a hypoechoic mass in the skin from inflammatory cancer.

Abnormal lymph nodes in the axilla are hypoechoic oval masses without a fatty hilum (see Fig. 5-17C to E). Lobulated or irregular abnormal lymph node borders or a thickened lymph node cortex can also indicate metastatic disease. Benign reactive lymph nodes cannot be reliably differentiated from lymph nodes containing metastases or lymphoma (see Fig. 5-17F and G). On the other hand, one cannot tell if a normal-appearing lymph node contains occult metastases. In the setting of known cancer, referring physicians may request needle biopsy of lymph nodes to make a diagnosis of metastatic disease before surgery or neoadjuvant chemotherapy.

Secondary signs of breast cancer include skin thickening, architectural distortion, breast edema, and retraction of Cooper ligaments (Box 5-11; see also Fig. 5-17H). Inflammatory carcinomas may have all of these signs, as well as marked attenuation of the sound beam due to breast edema (see Fig. 5-17I to K).

Breast Calcifications

Breast calcifications may be the only indication of breast cancer on mammography and therefore it is important to understand them. Calcifications are often seen in ductal carcinoma in situ (DCIS) but may also be the only indication of invasive ductal cancer. If ultrasound is done in the region of the abnormal calcifications, ultrasound may show an associated mass that was obscured or hidden by dense breast tissue on the mammogram. The purpose of detecting these associated masses is to further analyze the finding and direct subsequent ultrasound-guided biopsy. If no suspicious mass or calcifications are discovered on targeted ultrasound, the decision to biopsy the calcifications is based solely on mammographic analysis of the calcifications.

Although mammography effectively detects calcifications in DCIS, ultrasound is limited in finding calcifications because the calcifications may be lost in the normal speckle artifact of normal breast tissue (Fig. 5-18A). On ultrasound, DCIS looks like multiple dark hypoechoic nodules or enlarged knobby ducts that may conglomerate into a mass. This is because DCIS expands normal ducts. Ultrasound is much better at finding invasive breast cancer as breast masses than in detecting breast calcifications, unless the calcifications are associated with a mass. In a 2004 study of 111 women with breast cancer by Berg, mammography was compared with ultrasound for detection of invasive and noninvasive breast cancer. Ultrasound showed 12% (129/139) of invasive breast cancers and 47% (18/38) of DCIS, whereas mammography depicted 71% (99/139) of invasive cancers and 55% (21/38) of DCIS. Thus, mammography is better at showing calcifications than ultrasound.

Figure 5-18 A, Ultrasound of recurrent ductal carcinoma in situ (DCIS) shows a few calcifications as bright echoes near hypoechoic dilated ducts, but they might easily be lost in the normal speckled background of breast tissue. B, Ultrasound of DCIS in another patient shows clumped hypoechoic nodular ducts on the chest wall and a few bright echoes within ducts, possibly representing the microcalcifications easily seen on mammography.

On occasion, highly aggressive DCIS shows up on ultrasound as hypoechoic masses, as reported by Soo and colleagues, Huang and colleagues, and Moon and colleagues (see Fig. 5-18B). Nondetection of DCIS calcifications is reported by others when ultrasound is used as a primary and stand-alone method of breast cancer screening. The nondetection of calcifications alone may limit ultrasound’s use as a stand-alone screening modality, discussed in this chapter in the section, “Breast Cancer Screening with Ultrasound.”

Palpable or Mammographically Detected Findings Undetected by Ultrasound

Benign fibrofatty nodules, areas of dense glandular tissue, benign breast tissue, or breast cancer all may be felt as a mass or lump by the patient or her physician. Ultrasound can be normal when scanning directly over the palpable finding (Fig. 5-19). If the mammogram is normal, the patient is directed back to her referring physician for management of the palpable mass. In these cases, management of the palpable mass is based on clinical grounds alone. The decision may be made to perform biopsy based solely on the physical examination because some cancers are missed by both mammography and ultrasound, and the only indication of cancer is the palpable breast lump.

Figure 5-19 Ultrasound of a palpable mass felt by the patient’s physician shows normal breast tissue and no discrete mass. This image was taken after the patient pointed out the mass prompting this examination; the mass was palpated by the sonographer who scanned over it. With a negative mammogram and ultrasound, management of the palpable finding is based on clinical grounds alone, and the patient is encouraged to follow up with her referring physician for the palpable finding. In this case, an image of the normal breast tissue was obtained; however, some facilities take no images of normal breast tissue and only take an image of specific areas described by breast side, clock or quadrant position of the finding, number of centimeters from the nipple, and “palpable finding.”

For patients with suspicious physical findings and both a normal mammogram and ultrasound, the decision to perform biopsy is based on the physical findings and the clinical situation alone. Invasive lobular cancer is especially notorious for producing extremely subtle or no mammographic findings and yet still producing a palpable mass that causes the patient to seek advice. Invasive lobular cancer also may be missed by ultrasound.

Some breast cancers and solid benign breast masses discovered by mammography are ultrasonographically “invisible.” In the setting of a mammographically suspicious lesion and a “normal” ultrasound examination, further investigation or biopsy should be based on the mammogram. If the suspicious mammographic finding is not found by targeted ultrasound, the mass is presumed to be composed of solid tissue and could be benign or malignant. In these cases, the decision to perform biopsy is based solely on the mammographic impression, clinical situation, and physical findings.

Young Patients and Palpable Findings in Dense Breasts

Ultrasound is commonly used as the first modality to evaluate palpable findings in young patients. This is particularly true in women younger than age 30 because breasts in younger women are composed of mostly glandular tissue, which may hide a cancer. Similarly, ultrasound may be used to evaluate palpable masses in women with radiographically dense breasts on mammography at any age.

Invasive cancers can hide in the dense breast tissue on some views and may initially be seen as a “one-view-only” finding. Some invasive cancers may be palpable but not seen at all by mammography. Ultrasound can be extremely helpful when one suspects a mass that can be seen on only one view or if it is palpable and suspicious. However, ultrasound should be used only after the full mammographic workup is complete (Fig. 5-20).

Figure 5-20 Correlating a palpable finding to mammograms. A, Right mediolateral oblique mammogram in a dense breast shows a marker in the upper right over the palpable mass. B, Spot compression in the right mammogram does not show the mass, which is obscured by adjacent glandular tissue. C, Ultrasound shows an oval, well-circumscribed complex mass corresponding to the palpable finding. D, Color Doppler ultrasound shows marked vascularity in the mass. E, Sagittal 3-D spectral-spatial excitation magnetization transfer (3DSSMT) postcontrast breast magnetic resonance imaging (MRI) shows an enhancing upper right breast mass with a small satellite lesion inferiorly. F, Sagittal ROI over the mass under spiral dynamic postcontrast MRI shows a rapid initial rise in signal intensity and late washout (G). Biopsy showed invasive ductal cancer.

Ultrasound may confirm that a suspicious finding on one view only is actually overlapping breast tissue rather than a real mass (see Fig. 5-20H). If no mass is seen on breast ultrasound, the decision for biopsy of the palpable finding is based on clinical grounds alone because some cancers are “invisible” to ultrasound.

Correlating Mammographic Findings with Ultrasound Findings

When ultrasound finds a mass that was initially discovered by mammography, the next step is to determine whether the ultrasound finding and the mammographic finding represent the same thing. To accomplish this, the sonographer places a metallic skin marker over the ultrasound finding and repeats the mammogram. To determine where to place the skin marker, the sonographer scans directly over the finding and slides a finger, jumbo paper clip, or cotton-tipped swab under the transducer so that it overlies the ultrasound finding (Fig. 5-21A). This produces a ring-down shadow over the finding (see Fig. 5-21B). Once the ring-down shadow overlies the finding, the sonographer removes the transducer, leaving the fingertip, paper clip, or cotton-tipped swab on top of the skin over the mass. The sonographer marks that spot on the skin with a permanent ink marker, then places a radiopaque skin marker on the ink spot and takes a craniocaudal and mediolateral mammogram. The skin marker should correlate with the mammographic finding if the ultrasound and mammography findings are one and the same. Because the lesion may be deep in the breast, the skin marker may be compressed a few centimeters away from the lesion on the follow-up mammogram (see Fig. 5-21C to E).

Figure 5-21 Correlating the ultrasound finding with the mammogram. A, A large open paper clip is placed under the transducer over a finding. B, The ring-down shadow from the paper clip produces a bright shadow on the image. Superimposing this ring-down shadow over the lesion triangulates its location directly under the skin. C, Ultrasound shows a palpable, well-circumscribed, homogeneous mass in the outer right breast. Using the ring-down shadow technique, a metallic BB was placed on the skin directly over the mass. Cropped magnified lateral (D) and craniocaudal (E) mammograms show the BB (arrows) directly over a well-circumscribed, equal-density round mass, showing that the ultrasound finding corresponds to the mammographic finding. Biopsy showed fibroadenoma.

If the radiologist is still unsure if the findings correlate, the radiologist can place a radiopaque marker in the finding through a needle under ultrasound guidance. Follow-up orthogonal mammograms will determine whether the marker lies in the mammographic finding, showing whether the mammographic finding and the ultrasound findings are one and the same.

Breast Edema

Breast edema occurs in women with mastitis, inflammatory cancer, radiation therapy, postbiopsy trauma, or blockage of lymphatic drainage by various etiologies (Box 5-12). On physical examination an edematous breast is heavy and boggy. The skin may show peau d’orange, or orange-peel skin, in which skin pitting occurs where hair follicles hold the skin down and the surrounding tissues rise up with edema. On ultrasound, breast edema shows skin thickening greater than 2 to 3 mm. When compared with the contralateral breast, the edematous subcutaneous fat is gray as a result of fluid leakage, the normally sharp Cooper ligaments are less well defined, and breast tissue loses the sharp clarity of individual structures because the edema blurs normal breast landmarks. On occasion, dermal and subdermal lymphatics may become engorged and filled with fluid, producing fluid-filled branching tubules simulating blood vessels or ducts (Fig. 5-22A). Fluid-filled lymphatics can be distinguished from blood vessels by Doppler ultrasound; blood vessels will show pulsation, but fluid-filled lymphatics will show no flow (see Fig. 5-22B). Fluid-filled lymphatics will parallel the skin surface and branch quickly along the superficial layers of the breast. Fluid-filled lymphatics are easily distinguished from normal fluid-filled breast because breast ducts are larger than lymphatics and should branch and converge on the nipple.

Figure 5-22 Breast edema. A, Ultrasound shows marked skin thickening and a gray tinge to the normally markedly hypoechoic subcutaneous fat in this patient with marked breast edema. Just below the skin surface are numerous tubular fluid-filled structures representing blood vessels and fluid-filled lymphatics. Cooper ligaments are not seen distinctly. B, Power Doppler ultrasound shows that some of the structures represent blood vessels from hyperemia, whereas others show no flow and represent fluid-filled dilated lymphatics. C, Breast edema in another patient shows gray-appearing fat, a marked hypoechoic streak through the normal breast tissue, and loss of the usual speckled appearance of Cooper ligaments. In this case, there is no skin thickening or dermal lymphatic engorgement. D, For comparison, the normal opposite side shows normal dark hypoechoic fat, the normal thin lines of Cooper ligaments, and a thinner breast. E, Left mediolateral oblique view shows marked skin thickening and an extensible breast cancer and axillary adenopathy. Note the marked skin thickening and breast edema. F, Left breast ultrasound shows the marked skin thickening in the near field and a markedly shadowing spiculated invasive ductal cancer. G, Image during a core biopsy of a lower invasive ductal cancer in the same patient shows the skin thickening in the near field to greater advantage. The fat is gray, compatible with breast edema.

The key to identifying breast edema is to look for fat that is grayer and skin that is thicker in comparison to the normal contralateral breast (see Fig. 5-22C and D). In a normal breast the fat is dark, and Cooper ligaments and breast structures are sharply demarcated. In an edematous breast the fat is gray, and normal breast structures are less well defined.

Breast edema in inflammatory breast cancer can cause tremendous skin thickening and breast enlargement that attenuates both the x-ray and ultrasound beams because of the enlarged, dense breast tissue filled with fluid and tumor (see Fig. 5-22 E to G).

Mastitis and Breast Abscess

Mastitis is a breast infection commonly caused by Staphylococcus aureus or Streptococcus. It is common in nursing mothers, in whom the pathogen enters the breast through a cracked nipple. Mastitis also occurs de novo in adolescents from sexual contact and in diabetics and other immunocompromised patients. Mastitis also can occur in the postoperative period or after percutaneous biopsy.

Mastitis can be hard to scan because the breast is painful and enlarged. On physical examination the skin may be reddened, with focal peau d’orange. On ultrasound the normally black fat turns gray, the skin becomes thick, and the normal breast structures are blurred.

Mastitis is treated with antibiotics. As the breast responds to treatment, it becomes less red, less painful, and less swollen. The object of ultrasound scanning of mastitis in the acute phase is to search for an abscess (focal collection of pus) within the affected breast because an abscess needs immediate intervention by image-guided or surgical drainage to prevent continuing infection and complications from that infection. Thus, the sonographer carefully searches the breast for a focal, well-defined fluid abscess collection that can be drained by a needle. This can be difficult to do because fluid percolates through the infected area. The goal is to find a focal well-defined abscess collection, not percolating fluid. The other goal is to exclude inflammatory cancer, which can present just like mastitis but will persist despite antibiotics.

If untreated, mastitis can give rise to an abscess, which is a pus collection that forms a thick wall. When an abscess forms, the patient may feel a mass, and the patient’s breast will become even more painful to touch, hot, erythematous, and edematous. The patient may develop a fever and have an elevated white blood cell count. Focal abscesses are frequently subareolar because the infection is often introduced through the nipple. Antibiotics, though indicated for treatment, cannot penetrate the thick walls around abscesses. To allow the antibiotics to treat the infection, the abscess must be percutaneously or surgically drained to remove the pus.

On ultrasound, an abscess is an irregular but focal fluid collection that is ill-defined along its edges in the early phases and may be either irregular or well encapsulated in later phases (Fig. 5-23A). A breast abscess usually does not contain air (unlike abscesses in other parts of the body). The abscess can be hard to see because surrounding breast edema obscures normal breast structures. An abscess may contain only one pocket of pus that can be drained by a needle (see Fig. 5-23B); at other times, they may have multiple septae, contain debris or thick pus that cannot be drained without using a larger needle or leaving a catheter in place (see Fig. 5-23C and D). In larger abscesses, percutaneous drainage may help palliate the patient until surgery can be arranged. Although air does not usually occur in breast abscess, percutaneous abscess drainage can introduce air into the biopsy cavity.

Figure 5-23 A, Ultrasound shows breast edema surrounding an irregular fluid collection with enhanced transmission of sound characteristic of a retroareolar abscess. Note the needle used to drain the abscess. B, After drainage of the pus, a postaspiration ultrasound shows skin thickening and persistent edema, with less fluid in the abscess cavity. C, In another patient with an abscess, the fluid collection is just below the skin at the areolar margin, with marked skin thickening and surrounding edema. D, Unlike the patient in parts A and B, even though the fluid collection was in only one pocket, a 19-gauge needle could not drain the collection shown in part C because of its thick tenacious nature, and the ultrasound image with the needle tip in the superior lateral portion of the collection shows no change in its size. This collection was drained surgically (in some facilities a catheter might be placed percutaneously into the collection).

Large abscesses are drained surgically, with irrigation of the abscess cavity and manual description of the septa, and the abscess is packed and left open to the air to heal by granulation.

Breast Biopsy Scars

In patients undergoing lumpectomy for cancer, the surgeon excises the tumor and closes only the subcutaneous tissues and the skin above the cavity. The lumpectomy cavity fills with fluid afterward. Right after surgery, the biopsy bed is a fluid-filled pocket on ultrasound, is hypoechoic, and may have a sharp or ill-defined edematous rim with or without shadowing. Careful scanning over the skin biopsy scar shows the thickened skin at the incision and distortion of breast tissue along the incision from the skin surface to the biopsy scar (Fig. 5-24A). Later, serous fluid in the biopsy cavity may be totally clear or may contain solid debris or fibrous septa that move during real-time scanning (see Fig. 5-24B). Subsequently, the biopsy scar fills in with granulation tissue and becomes fibrotic, forming a hypoechoic spiculated mass with or without acoustic shadowing that mimics a spiculated breast cancer (see Fig. 5-24C to F).

Figure 5-24 Postbiopsy scars. A, Ultrasound of a postbiopsy scar after lumpectomy and radiation therapy shows a crescent-like fluid collection, representing fluid in the biopsy cavity, with an edge trailing toward the skin, representing the incision site. B, A postbiopsy scar in another patient is associated with an irregular fluid-filled mass and an echogenic ball within it, representing fibrin and debris after biopsy. C, Ultrasound of the postbiopsy scar months later shows a hypoechoic irregular spiculated mass corresponding to the biopsy site. Without the correct history, the scar could easily be mistaken for breast cancer. D, Postbiopsy cavities before radiation therapy. Ultrasound for planning of electron beam boost shows a fluid-filled biopsy cavity with echogenic solid debris within it. E, Imaging of a biopsy cavity in another patient shows the fluid-filled cavity with a thick septation and hypoechoic breast edema around the biopsy site. The measurements show the distance from the skin surface to the bottom of the cavity and from the skin surface to the chest wall for planning of electron beam boost. F, In a third patient, ultrasound shows the cavity as a hypoechoic, spiculated mass with very little fluid within it.

Breast cancers occurring near the biopsy site will have the same malignant characteristics and appearance of other breast cancers on ultrasound, but they are separated from the scar by normal breast parenchyma. The separation of a mass from the scar can help distinguish breast cancer from the biopsy scar. Admittedly, it is hard to distinguish breast cancer recurrences directly in the biopsy bed from the scar itself. Clues to the presence of cancer in a postbiopsy scar include the scar growing larger or more rounded like a mass.

Cancers Undergoing Neoadjuvant Chemotherapy

The term locally advanced breast cancer includes inflammatory breast cancer and tumors larger than 5 cm. Locally advanced breast cancer accounts for a small fraction of all breast cancer in the United States. Locally advanced cancers may have bulky or matted lymph nodes containing metastatic disease. In past times, these women usually underwent mastectomy with poor local control and poor 5-year survival rates.

Investigators have reported that preoperative neoadjuvant chemotherapy improves disease-free and overall survival for women. Preoperative neoadjuvant chemotherapy is defined as combination chemotherapy given before definitive surgical treatment (lumpectomy and mastectomy). It is usually given to breast cancer patients who have large tumor masses (stage T3 or T4) or regional lymph node involvement. Neoadjuvant chemotherapy provides tumor shrinkage, decreases tumor burden, and allows some patients to undergo lumpectomy and radiation for local control, rather than mastectomy. After surgery for local control, patients usually have chemotherapy again. There is about a 50% 5-year survival after neoadjuvant chemotherapy. Poor outcomes in these patients are usually due to distant micrometastatic disease at the time of diagnosis.

In the setting of neoadjuvant chemotherapy, ultrasound can guide percutaneous biopsy to establish a histologic diagnosis as needed, determine initial tumor size and extent, document treatment response, and evaluate for residual tumor after chemotherapy (Fig. 5-25). Magnetic resonance imaging is another tool used for detecting both disease extent and chemotherapy response. After neoadjuvant chemotherapy, the original tumor site is often resected to establish the type and extent of residual tumor. This information is important for predicting prognosis. A complete pathologic response (no residual cancer in the original tumor bed by histology) is a good prognostic indicator. Surgical lumpectomy with negative margins also helps to determine whether breast-conserving therapy is an option.

Figure 5-25 Response to neoadjuvant chemotherapy. A, A spot mammogram shows multiple dense round masses in the left breast, some of which were palpable and marked with a skin marker. B, Ultrasound of the mass in the 12-o’clock position shows an ill-defined hypoechoic round mass that was biopsied under ultrasound guidance; invasive ductal cancer was diagnosed, and a marker was placed in the mass under ultrasound control. C, After neoadjuvant chemotherapy, the mass has disappeared, but the metallic marker remains in the initial tumor site, which was localized under mammographic guidance to establish a pathologic response.

On occasion, neoadjuvant chemotherapy may produce a complete clinical response and the tumor is undetectable by both physical examination and imaging. This is a dilemma because how can the surgeon resect the original tumor bed if no residual tumor exists?

Because some tumors become undetectable by both clinical examination and imaging after neoadjuvant chemotherapy, the radiologist may place a metallic marker in the tumor under imaging guidance before the patient undergoes chemotherapy (Fig. 5-26). Then, if all traces of the tumor fade with chemotherapy, the marker will show the location of the original tumor site and can be used for subsequent preoperative needle localization to excise the now-invisible tumor bed.

Figure 5-26 Placement of a marker under x-ray guidance. Craniocaudal (A) and mediolateral (B) mammograms show a 3-cm invasive ductal cancer in the 12-o’clock position on the breast. Because the patient was to undergo neoadjuvant chemotherapy, a marker was placed in the mass under x-ray guidance. C, An alphanumeric grid is placed over the tumor and a needle inserted into the tumor. D, An orthogonal view shows the tip of the needle in the middle of the tumor; the marker is deployed through the needle, and the needle is removed. After neoadjuvant chemotherapy, the tumor is much smaller and the marker is still in place on craniocaudal (E) and mediolateral (F) mammograms. G, Preoperative needle localization bracket wires are placed around the marker and residual tumor, which is resected in toto, as shown on the specimen (H).

Postbiopsy Breast Markers and Core Biopsy Sites

Ultrasound-guided vacuum-assisted biopsy methods may actually remove an entire lesion. Usually, air or fluid is present in the biopsy track, in the biopsy site, or in a hematoma immediately after vacuum-assisted core biopsy. Air and fluid are absorbed relatively quickly after biopsy. After hematoma resorption, the only ultrasound findings are often residua of the original mass, if any remains (Fig. 5-27A to C). This is a problem if the entire lesion is removed and shows cancer, requiring surgical excisional biopsy. Fluid, air, or blood accumulating in the biopsy cavity may resorb before the surgery date and cannot be relied on to guide the surgeon. To solve this problem, tiny permanent metallic markers were developed to place in the biopsy site during percutaneous needle biopsy. The metallic marker provides a landmark in the biopsy cavity to guide subsequent mammographic or ultrasonographic preoperative needle localization.

Figure 5-27 A, Ultrasound shows a core biopsy needle in the post-fire position in an irregular hypoechoic invasive ductal cancer. B, After biopsy, a second needle was placed in the mass and a marker deployed to a position near the biopsied mass. The marker is the thin, bright echogenic line just beyond the needle tip. C, Ultrasound after marker placement shows the mass and the bright echogenic linear clip, easily seen adjacent to the hypoechoic irregular mass, which is the residua after biopsy. In this case, either the clip or the mass would be used as a target for preoperative needle localization for subsequent excisional biopsy. D, Another portion of the same breast shown in parts A to C had undergone stereotactic vacuum-assisted biopsy 3 weeks before. Ultrasound shows a small oval fluid-filled cavity, and the marker was placed adjacent to the fluid collection by stereotaxis. The clip is a thin, bright echogenic line that might subsequently be lost in the speckle artifact of normal breast tissue once the fluid collection is resorbed. E, In another patient, an echogenic pledget with a metallic marker is placed near a tumor close to the chest wall. The marker is the thick echogenic line to the left of the word gel. In this case, the tumor was difficult to see on the craniocaudal mammogram because of its position in the high inner aspect of the breast, and the residual tumor and the clip were used for bracket localization, as shown in the craniocaudal postlocalization mammogram (F).

On ultrasound, the metallic markers look like tiny bright echogenic lines, but the metallic marker echo can be lost in the speckle artifact of normal breast tissue (see Fig. 5-27D). To overcome problems in imaging the metallic markers by ultrasound, some manufacturers encase the markers in echogenic pledgets composed of various materials. The radiologist places the pledgets and their encased metallic markers through a vacuum-assisted biopsy probe or directly through a separate needle deployment device under ultrasound guidance. The pledgets are radiolucent and invisible to mammography, but they are detectable as echogenic lines or plugs on ultrasound (see Fig. 5-27E). These pledgets are absorbed by the body at a slower rate than blood or seromas and were developed primarily to be targets for subsequent ultrasound-guided preoperative needle localization. If using these pledgets, the radiologist should be familiar with the pledget resorption rate for the specific manufacturer.

Color Doppler, Power Doppler, Ultrasound Contrast Agents, Three-Dimensional Imaging, and Elastography

Color Doppler and power Doppler ultrasound depict the location of blood vessels when planning the trajectory of a percutaneous breast biopsy needle (Fig. 5-28). As a diagnostic tool, color and power Doppler imaging may show swirling debris in cysts or pulsating blood vessels within breast masses (Fig. 5-29).

Figure 5-28 Color Doppler ultrasound shows a blood vessel that was avoided during biopsy of a mass deep in the breast tissue. Because the vessel’s location was known, the needle was guided below it to approach the mass, which proved to be a cyst.

Figure 5-29 A, The craniocaudal mammogram shows a possible round mass in the inner breast (arrow). B, The left mediolateral oblique (MLO) view shows the possible mass in the upper left breast (arrow). C, Spot compression mammogram shows the round mass (arrow) to better advantage. The transverse (D) and longitudinal (E) scans show a hypoechoic round mass with slightly microlobulated borders. F, Color Doppler ultrasound shows that the mass is not a cyst and has a pulsating blood vessel within it; it is invasive ductal cancer. G, Sagittal noncontrast T2-weighted magnetic resonance imaging (MRI) shows a high signal intensity mass in the upper breast (arrow). H, MRI shows a mass in the upper breast (arrow) on the noncontrast 3-D spectral-spatial excitation magnetization transfer (3DSSMT). I, The mass enhances with contrast and is slightly heterogeneous (arrow). The ROI over the mass (J) and spiral dynamic image (K) shows rapid enhancement and washout of the enhancement curve. This was just as rapid as the ROI over the heart (L). M, Contrast enhancement of invasive ductal cancer usually has a rapid initial upstroke, with either a late plateau or washout phase, as seen in this case.

It was hoped that color Doppler imaging would distinguish cancer from benign breast lesions by showing increased blood flow in breast malignancies. The increased flow was thought to arise from tumor angiogenesis. However, color Doppler imaging does not always detect increased flow in breast cancer, and there is overlap between benign and malignant blood flow patterns. Attempts to increase the sensitivity of ultrasound for detecting blood flow with power Doppler improved these results, but not enough to advocate its use as a screening mechanism for breast cancer or to influence the decision to monitor a mass in lieu of biopsy. The use of contrast agents has been proposed as a means of increasing the ability of ultrasound vascular imaging techniques to detect small increases in vascular density. Three-dimensional gray-scale ultrasound, though promising, is also still being developed.

Elastography uses Hook’s law to determine the relative stiffness of breast tissues on ultrasound. Using ultrasound, elastography shows cancers, which are generally stiffer than normal soft breast tissue, as darker and larger than on the B-mode gray-scale ultrasound. Benign masses are soft and less stiff than cancers. The elastogram shows benign masses as smaller on elastography than on B-mode gray-scale images. Cysts are not stiff, and on the elastogram, cysts are smaller than on B-mode gray-scale ultrasound and have a bright target within them when compared to the B-mode ultrasound. Elastography is not in widespread use in the United States.

Breast Cancer Screening with Ultrasound

X-ray mammography is the gold standard for breast cancer screening and diagnosis. It depicts calcifications in DCIS and effectively displays invasive breast cancer masses in fatty breasts, but its limitations in dense breast tissue are well-known. On mammograms, Stomper and colleagues showed that dense breast tissue exists in about one third of women over age 50, but in only about half of women under age 50.

Because breast ultrasound is not limited by dense breast tissue, is relatively easy to use, requires only moderate breast compression, does not use ionizing radiation, and is widely available; there was great hope that screening for breast cancer with ultrasound would replace mammography.

The initial clinical investigations of screening breast ultrasound from the early 1980s were disappointing. An automated whole-breast ultrasound screening study by Kopans and colleagues depicted only 64% of 127 breast cancers in a study of 1140 women; in contrast, mammography detected 94% of the cancers. Another study with a similar number of patients by Sickles and colleagues resulted in sonographic detection of 58% (37 of 64) of cancers and mammographic detection of 97% (62 of 64). Only 8% of the ultrasound-detected tumors were smaller than 1 cm, but mammography detected all tumors less than 1 cm in size. Breast ultrasound had very poor visualization of microcalcifications, which were lost in the normal speckled breast tissue background. Furthermore, the limitations of ultrasound in fatty breasts resulted in an early recommendation that ultrasound not be used for breast cancer screening.

Subsequent improvements in transducer and ultrasound technology resulted in more optimistic results. Hand-held whole-breast ultrasound screening studies found small invasive cancers undetected by mammography in asymptomatic women. A 2002 study of 11,130 asymptomatic dense-breasted women undergoing screening mammography and whole-breast ultrasound screening by Kolb and colleagues showed 246 cancers in 221 women (1.98% of the 11,130 women). Mammography had a sensitivity and specificity of 77.6% and 98.8% versus 75.3% and 96.8%, respectively, for ultrasound. A 1995 study by Gordon and colleagues showed 1575 solid nonpalpable masses invisible by mammography but visualized by ultrasound in 12,706 women. Of these, 279 ultrasound-detected masses underwent biopsy, with 44 cancers found (16% of 279, 2.8% of the 1575 solid masses, 0.35% of 12,706 women).

The early promise and enthusiasm for whole-breast ultrasound screening was dampened by limited availability of trained sonographers, unreliable reproducibility, nondetection of calcifications in DCIS, and lack of specificity, resulting in unnecessary biopsies of incidentally detected benign breast lesions. A follow-up large-scale trial of breast cancer screening with ultrasound was then done under the American College of Radiology Imaging Network (ACRIN) to determine whether the sensitivity and specificity of screening breast ultrasound could be reproduced across multiple facilities.

In her review article on screening breast ultrasound, Berg reviewed mammography screening supplemented with ultrasound compared to mammography screening alone. Berg noted that in seven single institutional trials, the trial conducted by Corsetti and coworkers in Italy, two multicenter trials, and the ACRIN Protocol 6666, supplemental screening ultrasound showed additional breast cancers at a rate of between 2.7 and 4.6 cancers per 1000 women. In the ACRIN trial, most of the cancers were invasive, with a median size of 9 to 11 mm, and were node-negative. However, Berg also noted that false-positive ultrasound findings were common in the ACRIN study, yielding only an 8.8% cancer rate in sonograpically prompted biopsies. This limits the applicability of breast ultrasound in the United States, where there are not enough trained sonographers to perform screening ultrasound and where reimbursement does not cover the examination costs. Her recommendation in this 2009 review article is to retain mammography with digital technique as the mainstay of screening for women with dense breasts in the United States, with the supplement of MRI in high-risk women. In the United States, for high-risk women with dense breasts unable to tolerate MRI, she suggests that ultrasound screening “is an option at facilities with availability of qualified personnel.”

Additional research on women undergoing screening ultrasound is ongoing both in the United States and overseas, particularly in Asia, where women have mostly dense breast tissue.

Key Elements