Breast Ultrasonography

The first known clinical use of breast ultrasound (US) was reported in 1951 by Wild and Neal, ^[1] who used A-mode sonography to describe the features of one benign and one malignant breast mass. Several attempts were subsequently made to develop automated or multitransducer scanners to evaluate the whole breast, both to detect subclinical disease (screening) as well as for lesion characterization. ^[2]  With increasing resolution and quality of US imaging, the use of ultrasonography for breast scanning has increased. ^[3]  Ultrasonography has been playing an increasingly important role in the evaluation of breast cancer. US is useful in the evaluation of palpable masses that are mammographically occult, of clinically suspected breast lesions in women younger than 30 years, and of many abnormalities seen on mammograms. ^{[4, 5, 6, 7]}  

(An image depicting the breast structures visible on ultrasound can be seen below.)

Indications

Ultrasound is primarily used in the diagnostic setting in the presence of a palpable abnormality or when a previous mammogram has demonstrated a focal lesion. To its advantage, both dense and fat-replaced breasts can be well evaluated.

Major indications include the following:

• Palpable abnormality

• Focal abnormality on mammography

• Breast pain

• Nipple discharge

• Follow-up of lesions not biopsied (mostly BIRADS-3 lesions)

• Determination of lesion extent in patients with suspicious or malignant nodules

• Assessment of regional lymph nodes in patients with suspicious or malignant lesions

• Guiding interventional procedures

Attempts have been made to establish a place for ultrasound in screening. The American College of Radiology Imaging Network has published data on whole breast ultrasound screening (ACRIN6666), showing that the cancer detection rate with ultrasonography can be comparable to that with mammography in some cases. ^[5]  Given the laborious protocol of screening whole breasts, widespread screening applications are expected to appear only when automated breast scanners have demonstrated ease of use and a sensitivity and specificity comparable to handheld breast screening. ^{[6, 7]}

Ultrasound is inherently operator dependent, and breast ultrasound even more so; thus, a major contraindication is inadequate operator experience. Ultrasound equipment with a high-resolution linear transducer is necessary. Transducers with 7-MHz to 12-MHz frequency are preferred because of better near-field resolution. ^[6]  The American College of Radiology (ACR) recommends that breast ultrasound be performed with a high-resolution, real-time, linear-array, broad-bandwidth transducer operating at a center frequency of at least 12-MHz and preferably higher. Other transducers may be utilized in special circumstances. Focal zones should be electronically adjustable. In general, the highest frequency capable of adequate penetration to the depth of interest should be used. ^[8]

Two large studies of screening ultrasonography for breast cancer are the Japan Strategic Anti-Cancer Randomized Trial (J-START) (72,998 women; 36 139 women in the mammography plus ultrasonography arm) and a report from an Austrian population–based screening program (66,680 women overall and 31,918 women with dense breasts). In the J-START study, the cancer detection rate was 3.3 per 1000 screens in the mammography arm and 3.9 per 1000 screens in the mammography plus ultrasonography arm (increase of 0.6 per 1000 screens). ^{[9, 10]}  In the Austrian study, the cancer detection rate with mammography alone was 3.5 per 1000 screens, which increased to 4.0 per 1000 screens when ultrasonography was added. In those women with dense breasts, the cancer detection rate with mammography alone was 1.8 per 1000 screens, which increased to 2.4 per 1000 screens when ultrasonography was added. ^{[9, 11]}

 A systematic review conducted for the U.S. Preventive Services Task Force (USPSTF) found that ultrasound after a negative mammogram had a sensitivity of 80-83%, specificity of 86-95%, and positive predictive value (PPV) of 3-8%. ^{[12, 13]}

Equipment and settings

Too deep a focus causes volume-averaging artifacts and gives erroneous results about tissue consistency. Matrix probes that can also be focused in the short axis offer better resolution.

Even with high-resolution probes, the focal zone (indicated by moving small arrowheads at the side of the screen) must be kept either at the lesion or just deep to it for best resolution. When surveying the breast, the focal zone should be placed at the level of the mammary zone, improving resolution of this region, where most breast disease occurs. When a lesion is identified, positioning of the focal zone at or just below the lesion helps optimize the resolution of the lesion and its posterior features. ^[14]

Machines with panoramic view (extended field of view) options are often useful because this allows very large areas to be collated into a single image. Some machines offer an extended imaging field (trapezoid imaging), while in others, the 2 split fields can be appended to image more of a larger lesion. However, note that one must avoid using the panoramic images for measurements. 

Current-generation ultrasound machines offer additional image processing facilities, including spatial compounding, speckle reduction, harmonic imaging, and speed-of-sound imaging that offer enhanced resolution. These modes should be used and optimized for each patient. ^[15]

Additional features include elastography that color-codes tissue hardness, adding information that can be useful in certain conditions, as well as microcalcification imaging that enhances the presence of calcium particles.

An acoustic stand-off pad is helpful in imaging very superficial lesions.

Machine settings should be optimized for breast ultrasound. This includes appropriate focal depth, an optimal gain, time gain compression (TGC; usually the default, without any curve), a high dynamic range, and contrast and high edge enhancement settings. The depth of field should extend to the pleural echoes.

When optimized, the fat should appear medium gray in the premammary as well as the retromammary space.

Positioning

Generally the following positions are used:

• Medial lesions: The patient is supine, with the ipsilateral arm over patient’s head.

• Lateral lesions: The patient is opposite posterior oblique, with the ipsilateral arm over the patient’s head. The degree of obliquity depends on the breast size, pendulousness, and the location of the lesion within the breast. With large breasts, a complete decubitus position might be the best.

• If a lesion is superficially located and is palpable in the erect position, the patient should be examined, at least for a part of the procedure in this position. Remember that the breast appears thicker in the erect position, so this is not appropriate for posterior lesions.

• Superior lesions: The patient is supine, opposite posterior oblique or sitting, with the ipsilateral arm over the patient’s head.

• Inferior lesions: The patient is supine, and the breast can be held superiorly.

Varying degrees of pressure are used to flatten the breast parenchyma as well as bring the conical surface of the glandular tissue into a more orthogonal plane.

Given the real-time dynamic nature of ultrasound imaging and the tendency of the breast to change its shape with gravity, the patient position should be optimized to the examination. ^[16] This may be done several times to allow the acquisition of high-quality, reproducible images. Different positions are especially useful in large breasts. Inadequate positioning may cause portions of the breast to become inaccessible, the breast to fall over the transducer, or the lesions to appear displaced on follow-up examinations. Small breasts with little soft tissue to glide over the chest wall can be examined in the supine position.

Anatomy

The breast is a modified sweat gland located between 2 layers of the superficial pectoral fascia on the anterior thorax, between the second and sixth ribs (see the image below). ^[17] The 2 layers can sometimes be seen behind the skin echo and in front of the pectoral muscle echoes.

Twelve to 20 ducts exist, and their lobules, along with fat and stroma, make up the parenchyma of the breast. The breast ducts converge at the nipple at one end, and at the other end, they branch to terminate in lobules. The duct size varies from about 1 mm at the lobules to up to 5-8 mm near the nipple. The lobules are on an average 3 mm in the greatest diameter. The ratio of the parenchyma to the fat varies by the individual and even in the same woman at different ages and during pregnancy and lactation.

Before moving on to breast ultrasound anatomy, understand that the reference tissue in the breast is the fat, and structures are labeled as hypoechoic, isoechoic, or hyperechoic with reference to fat.

Ultrasound shows certain discrete structures that can be recognized. From superficial to deep, they are as follows:

• Skin: This is recognized as a thin hyperechoic to isoechoic zone between 2 thinner hyperechoic lines. Usually, this is 2 mm or less in thickness but is slightly thicker over the areolae.

• Subcutaneous fat (premammary fat): This is the reference tissue. Premammary fat is more hypoechoic than fat elsewhere and has a lobular structure, with the fat lobules surrounded by a very thin echogenic layer.

• Parenchyma: This layer appears as an echogenic plate; this can be homogeneous in some women, but can be heterogeneous, almost tigroid in appearance in others, especially in young women. This can also vary in echogenicity from echogenic in usual cases to almost isoechoic in lactating breasts. Within the parenchyma ducts, the terminal ductal lobular unit (TDLU) can be seen occasionally.

The ducts can be seen as thin echogenic lines representing collapsed and apposed duct walls, or lines surrounded by a hypoechoic zone that is the loose stromal tissue surrounding the ducts. Frequently, only the isoechoic periductal stromal tissue is seen, giving the almost striped (tigroid) appearance. With progressive accumulation of intraductal fluid, the ducts can appear as tubes that can be up to 5 mm across. Duct ectasia becomes more frequent with age, and almost 50% of women have ectatic ducts by 50 years of age. Duct ectasia is mostly an incidental finding but may be associated with discharge and periductal mastitis.

With very high resolution images, TDLUs are sometimes seen as small hypoechoic elongated or rounded structures. These are usually located peripherally in the parenchymal plate—more numerous in the anterior part than the posterior part. ^[18]

Cooper ligaments attach to the parenchymal surface from the anterior mammary fascia and can be seen as thin echogenic bands, widening as these insert into the anterior parenchymal surface. A few Cooper ligaments are seen posteriorly as well.

The breast is examined in a defined pattern to ensure that the whole organ is visualized; this usually means scanning the breast in a raster pattern in both the longitudinal and horizontal planes. Scanning then proceeds radially and finally in oblique planes, for measuring the long and short axes of lesions and also along ducts.

Special attention must be given to the nipple and the perinipple subareolar region. The ducts within the nipple and immediately near it cannot be seen with the usual technique. Usually the nipple is compressed when scanned orthogonally and appears as a poorly shadowing nodule within or even deep to the skin as it is pushed into the breast substance by probe pressure. The nipple and its ducts achieve a more favorable geometry with the ultrasound beam if it is rolled to one side by the probe with varying degrees of peripheral pressure. The breast can be supported by the other hand to really flatten the breast contour against the probe to give additional information.

Ultrasound guidance for biopsies is both safe and effective. Taking a biopsy of any lesion that is visible on ultrasound is technically possible. One must keep in mind that the breast flattens against the chest wall during ultrasound, so an oblique needle path carries a risk of injury to underlying muscles or even the pleura and lungs. Therefore, the needle path should be as parallel to the chest wall as possible to minimize the risk of injury to deeper tissues.

Annotation

Follow-up imaging, biopsy, or surgery must have a reproducible system of lesion localization. At the least, one must include the side (right or left breast), the clock face with the center at the nipple, the distance from the nipple or the zone (typically, the breast is divided into 5 zones comprising the subareolar zone [SA], 3 circular concentric zones outside the SA, and the axillary zone [AX]). The probe orientation is also mentioned, which could be horizontal, vertical, radial, or antiradial (at right angle to the radial section at that point). Additional oblique sections are often required to assess the longest diameter and the short axis of lesions. Including the depth of the lesion from the skin may be helpful, but this is not always necessary.

All lesions should be recorded in at least the long axis and short axis, appropriately labeled with the annotation information.

Image interpretation

The Breast Imaging Reporting and Data System (BI-RADS) Atlas, published by the American College of Radiology, provides standardized breast imaging terminology, report organization, assessment structure, and a classification system for mammography, ultrasound, and MRI of the breast. The BI-RADS Atlas, 5th Edition, includes complete sections on breast US (ACR BI-RADS–US) and MRI (ACR BI-RADS–MRI). ACR BI-RADS–US may help standardize the terms used for characterizing and reporting lesions, thereby facilitating patient care, the characterization of lesions, and the development of possible screening applications. ^[19]

ACR BI-RADS–US provides terms that describe the following features or findings on breast US examinations: shape, orientation, margin, boundary, echo pattern, posterior acoustic features, and surrounding tissue for masses; breast calcifications (which are poorly characterized by US); special cases, such as complicated cysts and intramammary lymph nodes; vascularity; and assessment categories. ^[19]

ACR BI-RADS–US describes 7 assessment categories, as follows ^[19] :

• Category 0: Incomplete — Need Additional Imaging Evaluation and/or Prior Images for Comparison
• Category 1: Negative
• Category 2: Benign
• Category 3: Probably Benign
• Category 4: Suspicious. This category is reserved for findings that do not have the classic appearance of malignancy but are sufficiently suspicious to justify a recommendation for biopsy. By subdividing category 4 into 4A, 4B, and 4C, it is hoped that patients and referring clinicians will more readily make informed decisions on the ultimate course of action.
• Category 5: Highly Suggestive of Malignancy. These assessments carry a very high probability (≥ 95%) of malignancy.
• Category 6: Known Biopsy-Proven Malignancy. This category is reserved for examinations performed after biopsy proof of malignancy (imaging performed after percutaneous biopsy but prior to surgical excision), in which there are no abnormalities other than the known cancer that might need additional evaluation.

Using the BI-RADS approach, a patient is placed in either the screen category or the diagnostic category. A screening mammogram is for a patient with no complaints and a normal examination. If there is a finding or symptom such as pain, a palpable lump, or discharge, the patient is placed in the diagnostic category. ^[19]

The mammographic report suggested by BI-RADS includes breast density, imaging findings, final assessment, and management. The mammographic lexicon includes category descriptions for breast composition or density, masses, calcifications, asymmetries, associated features, and location of the lesion. ^[19]

Breast density is described as fatty, scattered, heterogeneously dense, and extremely dense. If there is a mass, it is described by shape, margin, and density. The shape can be round, oval, or irregular. The margins can be circumscribed, obscured, microlobulated, indistinct, and spiculated. The density is described as high density, equal density, low density, and fat-containing. A mass that is an irregular shape with spiculated margins and is high density is the most concerning for malignancy, whereas a mass that is round with circumscribed margins is more likely to be benign, especially if it is fat-containing. ^[19]

In Bi-RADS 2, the benign findings include secretory calcifications, simple cysts, fat-containing lesions, calcified fibroadenomas, implants, and intramammary lymph nodes.  With BI-RADS 3, the risk of malignancy is below 2%, and a finding is described as a nonpalpable, circumscribed mass on a baseline mammogram; a focal asymmetry, which becomes less dense on spot compression images; or a solitary group of punctate calcifications. ^[19]

The BI-RADS category 4 is subdivided into a, b, and c. Subcategory (a) has a low probability of malignancy, with a 2-10% chance of malignancy. Subcategory (b) probability of malignancy is 10-50%. Subcategory (c) probability of malignancy ranges from 50 to 95%. ^[19]

BI-RADS 5 probabilitiy of malignancy is greater than  95%. At this level, even if  the pathology comes back as benign, the recommendation is still surgical consultation, because the pathology is discordant with the radiographic findings.  BI-RADS 6 is proven malignancy. ^[19]

Normal tissue

The breast parenchyma can appear thin and membranelike in old age or thick and almost isoechoic during lactation. The breast in a young woman who has never been pregnant demonstrates great variation in echogenicity and pattern that ranges from a homogeneously echogenic to a diffusely heterogeneous, almost striped, appearance because of ducts and periductal hypoechoic tissue. The parenchyma may also show focal thickening, which may occasionally cause a palpable abnormality.

Another age-related change is ductal ectasia, which can be completely asymptomatic or may be associated with nipple discharge. The fluid in ectatic ducts can be clear, can have low-level echoes, or can be dependent debris. A solid component is highly suspicious for intraductal papilloma or cancer.

Focal disease

Focal disease includes simple and complex cysts and solid masses. The following imaging characteristics are evaluated:

• Size - Measured in the long and short axes, to document stability, regression, or growth

• Shape - Taller than wide, or wider than tall, extension along ducts

• Surface - Smooth, irregular, macro- or microlobulated, number of lobulations (up to 3 can be considered nonsuspicious), angular margins, spiculations, capsulation

• Internal texture - Isoechoic, hyperechoic, or hypoechoic; heterogeneous or homogeneous; complex; calcification; presence of echogenic, cystic center, or calcification

• Surrounding tissue - Echogenic halo, radiating branches, whether the tissue pulled in toward the mass, and acoustic shadowing

• Fixity - On real time ultrasound, fixity to surrounding tissue and underlying muscles can be assessed.

• Doppler - Vascularity does not help in differentiating between benign and malignant lesions, although benign lesions tend to be more hypovascular. Color Doppler can assist in planning an appropriate needle trajectory to avoid large vessels.

Simple cyst

A simple cyst, if it fulfils the appropriate ultrasound criteria (thin regular walls, no mural nodules, no septations, "clear" anechoic interior, posterior enhancement) is invariably benign. Asymptomatic simple cysts need not be aspirated, and the patient reverts to routine screening mammography.

Complicated cysts

Complicated cysts have echogenic foci in the contained fluid but have other features of simple cysts. These are usually benign, but a short-term follow-up may be indicated to document stability or resolution.

Complex cysts

Cysts with mural nodules or thick septations should be evaluated further using aspiration cytology of the fluid or a needle biopsy from the solid component of the thickened cyst wall or septum, if present. The presence of solid tissue or a detectable linear vascularity consistent with a vascular stalk is concerning for an intracystic mass.

Solid nodules

To be labeled as benign, the lesion should demonstrate only benign features. The presence of any suspicious finding warrants a biopsy. (See Table 2.)

Table 2. Ultrasound Features Suggestive of Benign and Malignant Nodules. (Open Table in a new window)

 

(See the images below.)

Sonographic guidance for interventional procedures offers several advantages: it is less expensive, quicker, available in real time, avoids ionizing radiation, and is portable. However, a learning curve exists, and the operator must be appropriately trained and skilled to undertake the intervention.

Ultrasound-guided interventions are used for cyst aspiration, fine-needle biopsy, and core biopsy, sentinel node injections, localizing wire placement, guiding the placement of brachytherapy catheters, and directional vacuum-assisted biopsy. During biopsy, care must be taken to orient the needle as parallel to the chest wall as possible, especially when performing a core biopsy with a spring-loaded biopsy gun or automated needle. The needle "throw," which may be over 2 cm in some cases, should be taken into account, so that when fired, the needle does not injure the underlying tissues.

Ultrasound of breast implants

Breast implants have been used both for purely cosmetic reasons and post mastectomy. Saline implants have replaced silicone implants as the most common breast prosthesis. Silicone implant imaging is important because the clinical signs of rupture are often nonspecific. In contrast, saline implant ruptures are more obvious because of the decrease in breast volume from extravasated saline. ^{[20, 21]}

An intact implant shows a smooth, thin linear membrane with a sonolucent anechoic interior, reverberation artifacts from the proximal wall, radial folds, and linear echoes. With intracapsular rupture, a "stepladder" appearance may be seen. Additionally, one may observe focal irregular bulges, and the margins can become ill-defined. With extracapsular rupture, hyperechoic noise or a snowstorm appearance may be observed. Silicone granulomas may be seen as focal lesions with a snowstorm appearance, hyperechoic, hypoechoic, or even anechoic. ^[22]

As ultrasound imaging technology advances, new methods offer incremental information over what basic scanning can yeild. Some of the newer ways of processing include elastography and microcalcification imaging. ^[15]

The World Federation of Ultrasound in Medicine and Biology (WFUMB) has published guidelines on the use of ultrasound elastography for breast examination to measure the stiffness of breast tissue. According to the WFUMB, ultrasound elastography has been shown to be highly accurate in characterizing breast lesions as benign or malignant. Cancer tissue is stiffer than normal breast tissue, and it is believed that the tissue stiffening begins in the early stage of cancer. The Tsukuba score is a 5-point scale for grading the stiffness of a mass, but other methods of interpretation (eg, lesion-to-fat ratio) have been shown to be effective. ^[23]

Studies have shown that strain elastography and shear wave elastography can have high sensitivity and specificity for characterizing breast lesions as benign or malignant. Strain elastography has displayed the highest sensitivity, and shear wave elastography the highest specificity. ^{[24, 25, 26]}

Strain elastography is a qualitative technique that evaluates the changes in tissues when an external force is applied, with softer tissues deforming more than stiffer tissues. The 2 methods of applying stress in strain elastography are manual compression and release and an acoustic radiation force impulse (ARFI) push pulse. Strain elastography produces an image based on the relative displacement of the tissue. ^{[24, 25, 26]}

Shear wave elastography (SWE), using the acoustic radiation force induced by the ultrasound push pulse generated by the transducer, provides quantitative elasticity parameters and displays a visual color overlay of elastic information in real time. In addition to proving useful for the diagnosis of breast cancer, shear-wave elastography has been shown to be valuable as a preoperative predictor of the prognosis for or the response to chemotherapy. ^{[24, 25, 26]}

A retrospective study of 65 men with palpable breast masses concluded that SWE may improve diagnostic management when clinical examination, mammography, and conventional ultrasound are doubtful. Of the 50 benign lesions and 15 malignant lesions examined, mammography alone or mammography and conventional ultrasound could not determine 10 lesions and 14 lesions, respectively. SWE was able to correctly classify all the tumors retrospectively. If SWE quantitative criteria had been applied to the equivalent ACR BI-RADS 3 and 4, 7 biopsies for mammography alone, and 10 biopsies for the combination mammography and ultrasound with Doppler, could have been avoided, according to Crombe et al. ^[27]

Three-dimensional ultrasound adds additional features to the image and is slowly refining benign and malignant characteristics of breast masses. It is now known that many malignant lesions can have smooth contours, be wider than tall, and show acoustic enhancement, while benign lesions can have marked irregularity of contour. ^[28] Other areas of progress include contrast-enhanced ultrasound, which might find use in sentinel node scanning and complement or replace radionuclide methods. Automated breast ultrasound systems are already commercially available that scan the whole breast volume and increase the cancer pick-up rate by almost 100%, as compared to mammography alone. ^[29] This technique is especially useful in women with dense breasts in whom mammography offers limited information and sensitivity.