Mammogram Analysis and Interpretation


In the United States, statistics indicate that one in eight American women will develop breast cancer if women live a 90-year life span. The incidence of breast cancer in women in the United States is rising, and although the rate of increase has slowed recently, the rate of in situ breast cancer continues to increase. The United States breast cancer death rates have decreased since the early 1990s, with decreases of 2.5% per year among white women. These decreased breast cancer deaths are attributed to both improved breast cancer treatments and mammography screening. Randomized, controlled population trials of women invited to breast cancer screening using x-ray mammography showed an approximate 30% reduction in breast cancer deaths in the invited group compared with women the control group. The National Comprehensive Cancer Network (NCCN) recommends annual screening mammography for women aged 40 years and older (V1.2015). Societies such as the American Cancer Society (ACS) and the U.S. Preventive Services Task Force recommend annual screening after ages 45 and 50, respectively, whereas other societies differ widely from every other year to no screening mammography at all.

This chapter reviews breast cancer risk factors, signs, and symptoms of breast cancer; the normal mammogram; mammographic findings of breast cancer; basic interpretation of screening mammograms; and workup of findings detected at screening with additional mammographic views and tomosynthesis.

Breast Cancer Risk Factors

Risk factors for breast cancer are important to consider when reading mammograms because they indicate a pretest probability of breast cancer. Compiling risk information on a breast history sheet or compiled by the technologist on a computerized form provides interpreting radiologists quick access to this important information ( Fig. 2.1 ). Breast cancer risk factors are listed in Box 2.1 . The most important risk factors are personal or family history of breast cancer/ovarian cancer, genetic mutations predisposing to breast cancer such as BRCA 1/2 , older age, and female gender. Men also develop breast cancer, but only 1% of all breast cancers occur in men in the United States.

FIG. 2.1, Breast history form. It includes a diagrammatic breast template and places to record the patient’s history and current problems or complaints.

BOX 2.1
Breast Cancer Risk Factors

  • Female

  • Older age

  • Personal history of breast cancer

  • First-degree relative with breast cancer (see also models for breast cancer risk based on family history of breast and ovarian cancer, such as Tyrer–Cuzick and modified Gail model)

  • Early menarche

  • Late menopause

  • Nulliparity

  • First birth after age 30

  • Atypical ductal or lobular hyperplasia

  • BRCA1, BRCA2, and other genetic predispositions

  • Radiation exposure

  • Lobular carcinoma in situ

Breast cancer risk increases with increasing age and drops off at 80 years old. Women with a personal history of breast cancer have a higher risk of developing breast cancer in the ipsilateral or contralateral breast than does the general population. After breast cancer surgery, the conservatively treated breast has a 1% per year risk of developing recurrent or new cancer.

Breast cancer risk assessment uses breast cancer risk assessment models based on a detailed family history of breast or ovarian cancer looking for genetic predispositions for breast cancer, and calculates lifetime risks for breast cancer. The BRCAPRO, BOADICEA, modified Gail model, and Tyrer–Cuzick model take into account the age, number, and cancer types in affected relatives, as well as other risk factors, to estimate a lifetime risk for breast cancer. Both the NCCN and ACS consider a >20% to 25% lifetime risk to be high risk. Women with a first-degree relative (mother, daughter, or sister) with breast cancer have approximately double the risk of the general population and are at particularly high risk if the cancer was premenopausal or bilateral. If many relatives had breast or ovarian cancer, the woman may be a carrier of BRCA1 or BRCA2 , the autosomal dominant breast cancer susceptibility genes. Genetic testing is most appropriately performed by genetic counseling professionals who evaluate, counsel, and support women because of untoward social effects of either positive or negative results. Carriers of the breast cancer susceptibility gene BRCA1 on chromosome 17 have a breast cancer risk of 85% and an ovarian cancer risk of 63% by age 70. Women with BRCA2 on chromosome 15 have a high risk of breast cancer and a low risk of ovarian cancer. These genes account for 5% of all breast cancers in the United States and for 25% of breast cancers in women younger than age 30 years old. Women of Ashkenazi (Eastern European) Jewish heritage have a slightly higher risk of breast cancer than does the general population ( Box 2.2 ), but additional work is being done to determine whether this population has a higher rate of breast and ovarian cancer related to BRCA1 and BRCA2 mutations. Other genetic syndromes that have a higher risk of breast cancer include the Li-Fraumeni, Cowden, and ataxia-telangiectasia syndromes.

BOX 2.2
Family History Suggesting an Increased Risk of Breast Cancer

  • >2 relatives with breast or ovarian cancer

  • Breast cancer in relative age <50 years

  • Relatives with breast and ovarian cancer

  • Relatives with two independent breast cancers or breast plus ovarian cancer

  • Male relative with breast cancer

  • Family history of breast or ovarian cancer and Ashkenazi Jewish heritage

  • Li-Fraumeni syndrome

  • Cowden syndrome

  • Ataxia-telangiectasia

Factors such as early menarche (before age 12), late menopause (after age 55), nulliparity, and first live birth after age 30 bestow a slightly higher risk for breast cancer, as a result of having more menstrual cycles and longer exposure to estrogen and progesterone. Data from a 2003 study, part of the Women’s Health Initiative, which is a randomized, controlled trial of the effects of estrogen plus progestin (combination hormone replacement therapy [CHRT]) versus placebo, showed a 24% greater incidence of breast cancer in women receiving CHRT compared with the control group. Whereas previous data showed an adjusted relative risk of 1.46 for the development of breast cancer in women receiving CHRT for more than 5 years, the 2003 analysis showed the risk for breast cancer rising within 5 years of starting CHRT; in addition, it showed more difficulty in detecting cancers by mammography.

A breast biopsy showing atypical ductal hyperplasia (ADH) histology increases the risk for breast cancer to four to five times that of the general population. The presence of lobular carcinoma in situ (LCIS) also increases the risk for breast cancer, but at a much higher rate than ADH (about 10 times that of the normal population). The acronym LCIS is a misnomer and not a cancer at all; rather, LCIS is a high-risk marker for developing breast cancer. A woman with LCIS has a 27% to 30% chance of developing invasive ductal or lobular cancer in the ipsilateral or contralateral breast over a 10-year period. Thus a biopsy showing LCIS results in patient management of either “watchful waiting” with increased surveillance by frequent imaging and physical examination, or bilateral mastectomy.

Women who had an early exposure to radiation also have an increased risk for breast cancer. A medical history of radiation therapy to the mediastinum for Hodgkin disease, multiple fluoroscopic examinations for tuberculosis, ablation of the thymus, or treatment of acne with radiation infers scattered radiation to the breasts at an early age, which may induce breast cancer. The risk for developing breast cancer is so high in women treated for Hodgkin disease that in 2007 the ACS recommended magnetic resonance breast cancer screening for Hodgkin disease survivors as well as BRCA1/2 genetic mutation carriers or women with a high lifetime risk of breast cancer >20% to 25% lifetime risk.

Extensive mammographic breast density, defined as a large amount of fibroglandular tissue within the breast by volume as measured on the mammogram, is associated with the risk of breast cancer. However, the association and the reasons for this finding, as well as its relative association among different ethnicities, are still being studied.

Other lifestyle choices also affect breast cancer risk. One of these is drinking alcohol. One drink per day bestows a very small risk, but two to five drinks per day increases the risk to 15 times that of women who do not drink. Being overweight or obese also increases the risk of cancer, especially if the weight gain happens after menopause and the fat is around the abdomen. A woman with an “apple-shaped” body is at higher risk than one with a “pear-shaped” body. Exercise has been shown to reduce breast cancer risk after menopause, with one study suggesting that cancer risk was reduced at least in part via hormonal pathways. However, more study of these changeable risk factors is needed.

Despite all that is known about breast cancer risk factors, 70% of all women with breast cancer have none of these risk factors other than older age and female gender. What can women do to prevent breast cancer? The ACS recommends adopting a healthy lifestyle including exercise, maintenance of an appropriate body mass index, and decreased alcohol consumption. Women at very high risk also may take antiestrogen medications to prevent breast cancer.

Signs and Symptoms of Breast Cancer

A breast lump is one of the most common symptoms for which women seek advice ( Box 2.3 ), but most lumps are benign, often found by the woman herself or her partner, and most commonly caused by breast cysts or solid fibroadenomas (the most common solid benign breast masses in women). Breast lumps are worrisome for cancer if they are new, growing, hard, stuck to the skin or chest wall, causing skin dimpling or nipple retraction, or are associated with bloody nipple discharge.

BOX 2.3
Signs and Symptoms of Breast Cancer

  • Breast lump

  • Nipple discharge (new and spontaneous, bloody, serosanguineous or serous but copious)

  • New nipple inversion

  • Skin retraction or skin tethering

  • Peau d’orange

  • Nothing (cancer detected on screening mammography)

Nipple discharge is another finding for which women often seek advice. Nipple discharge is usually benign, especially if it is whitish, green, or yellow or produced from several ducts, and it is usually caused by fibrocystic change. Nipple discharge is suspicious for cancer if it is new, expressed from only one duct, bloody or serosanguineous, spontaneous, copious, or serous. An example of a suspicious clinical history is a woman describing new bloody or serous nipple discharge on her nightgown or undergarments. Other causes of bloody nipple discharge are benign intraductal papilloma, pregnancy, or trauma.

Nipple change may be caused by benign etiologies or to Paget disease of the nipple. Nipple irritation caused by nursing is common, and nipple irritation with associated mastitis is not uncommon but is benign. However, Paget disease of the nipple is a malignancy, and may present with a bright red nipple initially, can proceed to an eczema-like flakiness simulating infection, and later, frank tumor invasion and tissue destruction. Patients with Paget disease of the nipple are often treated for months with antibiotics until the true diagnosis of malignancy is made by nipple tissue punch biopsy.

Long-standing nipple inversion is not uncommon and is benign if it is present at birth. On the other hand, new nipple inversion is worrisome because retroareolar breast cancers can pull in the nipple by productive fibrosis or invasion, causing the nipple retraction.

Skin retraction or skin dimpling is a sign of breast cancer and is caused by superficial cancers tethering or invading the skin and pulling the skin in toward the breast. On physical examination, skin retraction or tethering might become evident when the woman raises her arms as she inspects her breasts in the mirror. Raising her arms or placing her hands on her hips pulls in the pectoralis muscle, which pulls on the cancer, which pulls on the skin, and dimples the skin in toward the cancer.

Peau d’orange is a French word for “orange peel” skin or pitting of the skin caused by breast edema. The pitting is caused by fluid accumulating in the skin and rising around the bases of tethered hair follicles, resulting in skin pitting. Breast edema is a nonspecific finding and can be caused by mastitis, trauma, inflammatory cancer, edema from radiation therapy, or axillary lymph node obstruction.

Lymphadenopathy can cause large or painful axillary lymph nodes in the armpits for which women seek advice. Bilateral lymphadenopathy may be caused by systemic illnesses such as widespread infection, lymphoma, leukemia, collagen vascular disease, and widespread malignancy, to name a few etiologies. Causes of unilateral lymphadenopathy are infection, tumors that include breast cancer, and granulomatous disease, whereas other causes are rarer. In the breast, the radiologist looks at the ipsilateral breast tissue to exclude breast cancer as the cause for the patient’s complaint of unilateral lymphadenopathy.

Despite all these signs and symptoms of breast cancer, some women have no physical findings or symptoms at all despite having breast cancer. Their breast cancers are detected on screening mammography and are asymptomatic.

Breast pain is not generally caused by cancer, but it deserves special mention because breast pain is very common. If cyclic, breast pain is usually endocrine in nature. Although breast pain is usually caused by benign etiologies, unfortunately, both breast pain and breast cancer are common. Thus the physician’s goals are to reassure patients with breast pain, search for treatable causes of breast pain such as cysts, and exclude coexistent malignancy.

The Normal Mammogram

Anatomy and Image Contrast

A normal breast has glandular breast elements surrounded by fat and breast stroma, which is surrounded by a honeycomb fibrous structure of thin strandlike Cooper’s ligaments. The glandular elements are composed of lactiferous ducts leading from the nipple and branch into excretory ducts, interlobular ducts, and terminal ducts that lead to acini that produce milk ( Fig. 2.2 ). The ducts are lined by epithelium composed of an outer cellular myoepithelial layer and an inner secretory cellular layer. The ducts and glandular tissue extend posteriorly in a fanlike distribution consisting of 15 to 20 lobes draining each of the lactiferous ducts, with most of the glandular tissue found in the upper outer breast near the axilla. Fatty tissue surrounds the glandular tissue. Posterior to the glandular tissue is retroglandular fat, described by Dr. Laszlo Tabar as a “no man’s land,” in which no glandular tissue should be seen. The pectoralis muscle lies behind the fat on top of the chest wall.

FIG. 2.2, Schematic of a normal breast. (A) Schematic of a normal breast showing the nipple, ducts, and lobes that contain terminal ductal lobular units (TDLUs). (B) A TDLU consists of acini, ductules, and terminal ducts. Breast milk is produced in TDLUs.

Mammography provides image contrast by using the differences in the x-ray attenuation among the different breast tissue types, such as fat, fibroglandular tissue, and carcinoma. With lower energy x-rays used by mammography (approximately 25–28 keV), the attenuation difference between the fibroglandular tissue and carcinoma is more pronounced than on standard radiographs, which use a 50-keV x-ray. Fatty tissue is the least dense and most translucent to x-rays and appears dark on mammography. Fibroglandular tissue, muscle, and lymph nodes are more dense and radiopaque (whiter) than fatty tissue, and are white on mammography. Cancers and fluid-filled cysts may be denser and whiter than normal surrounding fibroglandular tissue. Calcifications and metals are the brightest and whitest of all structures on mammography.

On the normal mediolateral oblique (MLO) mammogram, the pectoralis muscle is a concave white structure posterior to the retroglandular fat near the chest wall ( Fig. 2.3 ). Normal lymph nodes high in the axilla overlie the pectoralis muscle. Normal lymph nodes are sharply marginated, oval, or lobulated dense white masses with a radiolucent fatty hilum. They are commonly found in the axilla and in the upper outer quadrant of the breast along blood vessels. Lymph nodes can also occur normally but less commonly in any quadrant of the breast and are called normal intramammary lymph nodes. A normal lymph node has a typical, sharply marginated kidney bean shape; a white outer cortex; and a dark, fatty hilum on the mammogram and should be left alone. If one is uncertain about whether a mass represents an intramammary lymph node, a tomosynthesis slice or mammographic magnification views may help display the fatty hilum, or an ultrasound may show the typical hypoechoic appearance of the lymph node and the echogenic fatty hilum.

FIG. 2.3, Normal breast anatomy and correlative mammograms. (A) Schematic of a normal mediolateral oblique (MLO) mammogram. Note the normal scalloped edge of glandular tissue, retromammary fat, concave pectoralis muscle, and normal lymph nodes. (B) Normal MLO mammogram.

On the normal craniocaudal (CC) projection, the pectoralis muscle produces a half-moon–shaped or bandlike density near the chest wall ( Fig. 2.4 ). Fat lies anterior to the muscle, and the white glandular tissue lies anterior to the fat. In older women, most of the glandular tissue in the medial breast undergoes fatty involution, and most of the residual dense glandular tissue persists in the upper outer breast.

FIG. 2.4, (A) Schematic of a normal craniocaudal (CC) mammogram. Note the normal fat in the medial and retroglandular regions and the location of the pectoralis muscles. Most of the residual glandular tissue and the sternalis muscle remain in the upper outer quadrants. (B) Normal CC mammogram.

There should be only fatty tissue in the medial breast near the chest wall on the CC view. The only normal exception is the sternalis muscle, which is a white slip of muscle hugging the medial chest wall on the CC view that should not be mistaken for a mass ( Fig. 2.5 ). If there is a question that a density in the medial breast is a mass instead of the sternalis muscle, a tomosynthesis slice should show that it is sternalis muscle. If there is no tomosynthesis, a cleavage view (CV) mammogram or ultrasound can prove that the density is a just a sternalis muscle, is normal, and can be ignored.

FIG. 2.5, Sternalis muscle. (A) The breast is composed of scattered fibroglandular density. A muscle-like density seen in the right breast medial to the half-moon or triangular shape of the pectoralis muscle ( arrow ) near the chest wall on the craniocaudal (CC) view represents the sternalis muscle. (B) The sternalis muscle is not seen on the mediolateral oblique (MLO) view.

Breast Density, Breast Density Notification Legislation, and Breast Density Classification Based on the Breast Imaging Reporting and Data System

Breast density is an important feature of the mammogram that describes how much of the breast volume is filled with white glandular tissue. Breast fibroglandular tissue is white on the mammogram, and fat is black on the mammogram. Women have varying ratios of glandular and fatty tissue in their breasts genetically. The definition of a dense breast is one that contains a lot of glandular breast tissue and looks mostly white on the mammogram. The opposite of a dense breast is a fatty breast, which looks mostly black on the mammogram. A dense breast does not mean the breast is hard to the touch. Breast density on the mammogram has little correlation to how hard or soft the breast feels on physical examination; that is, you cannot predict how soft a breast will feel to the touch by looking at the mammogram density. Conversely, there is no correlation of how white the mammogram will be based on how hard, lumpy, or soft the breast feels on physical examination.

The American College of Radiology’s (ACR) Breast Imaging Reporting and Data System (BI-RADS) lexicon classifies breast density on mammograms into four groups: extremely dense, heterogeneously dense, scattered areas of fibroglandular density, and almost entirely fatty ( Fig. 2.6 ; Box 2.4 ). From 2004 to 2013, extremely dense, heterogeneous dense, scattered areas of fibroglandular density, and almost entirely fatty defined, in quartiles, how dense or how white the breast looked on the mammogram. Dense was defined as >75% to 100% dense tissue by volume, heterogeneously dense was 51% to 75% dense, scattered was 26% to 75% dense, and fatty was <25% dense by volume. Dense and heterogeneously dense glandular tissue lowered the sensitivity of mammography, because breast cancer is also white on the mammogram, and the white dense normal background of glandular tissue can hide a cancer, just like a polar bear can hide in a snowstorm.

FIG. 2.6, Mammograms of normal breast density. BI-RADS 2013 lexicon classifies breast density into four types. (A) “Dense” glandular tissue, which lowers the sensitivity of mammography. A mammogram of a young woman is shown. (B) “Heterogeneously dense” breast tissue, which may obscure small masses. (C) “Scattered” densities with scattered areas of fibroglandular tissue. (D) A “fatty” breast composed of almost entirely fatty tissue. A mammogram of an older women is presented.

BOX 2.4
ACR BI-RADS Terms for Breast Density
From ACR BI-RADS Mammography, In ACR BI-RADS atlas, breast imaging reporting and data system , Reston, VA, 2013, American College of Radiology.

  • Breast is almost entirely fat

  • There are scattered fibroglandular densities

  • Breast tissue is heterogeneously dense, which could obscure detection of small masses

  • Breast tissue is extremely dense

The fifth edition of BI-RADS, published in 2013, no longer defines breast density in the quartile system ( Fig. 2.7 ). Instead, dense and heterogeneously dense describe whether there is enough breast tissue on the mammogram to obscure a cancer. Thus a heterogeneously dense normal background of glandular tissue may have less than 50% dense breast tissue by volume on the mammogram but contains enough dense tissue to obscure a cancer. In the mammographic report, radiologists describe breast density so that referring doctors will know how white the breast looks and how confident the radiologist is in excluding cancer.

FIG. 2.7, The fifth edition of BI-RADS no longer indicates the ranges of percentage dense tissue of the four density categories. In this case, the breast density is classified as heterogeneously dense based on BI-RADS 2013 because the fibroglandular tissue could mask small cancers, even though far less than 50% of volume of the breast contains fibroglandular-density tissue. Mediolateral oblique (MLO) view (A) and craniocaudal (CC) view (B).

Breast density notification legislation in the United States began in 2009 in the state of Connecticut, where the law mandated that women with dense and heterogeneously dense breast tissue on mammograms receive a letter informing them that they have dense breast tissue, that dense breast tissue is normal, that dense breast tissue carries a higher risk of breast cancer, that dense breast tissue may hide breast cancer, and that they may wish to discuss supplemental screening tests with their doctor. BI-RADS 2013 shows that in the United States about 50% of all women have heterogeneously dense or dense breast tissue on mammograms. As of May 2016, 27 states have proposed or enacted dense breast notification federal legislation.

It is important to know that the breast is normally dense at a young age and decreases in density with time as glandular tissue normally involutes into fat. Young women have mostly glandular breasts, and their mammograms are usually dense or white on mammograms (see Fig. 2.6A ). As women age, their fibroglandular tissue involutes into fat, which is black. Therefore the natural progression of the mammogram is mostly white (dense) at a young age when the breasts are filled with glandular tissue and it becomes progressively darker (fatty) as the woman ages. The amount of remaining glandular tissue varies from woman to woman and depends on genetics, parity, and exogenous hormone replacement therapy. Some older women have surprisingly large amounts of dense white tissue on the mammogram, but generally as women age there are greater amounts of fat and less dense glandular tissue, which usually remains in the upper outer quadrants of the breast (see Fig. 2.6D ), producing a darker mammogram. Normal increases in breast density occur only in pregnant and lactating women, or in women starting exogenous hormone replacement therapy, because of the increase in fibroglandular tissue in response to female hormones. The breast density does not increase normally otherwise. Any unexplained increases in breast density should be viewed with suspicion and worked up.

Mammographic Findings of Breast Cancer

Radiologists detect breast cancers when the tumor produces findings that are different from the normal fibroglandular/fatty background. Common mammographic findings of breast cancer include pleomorphic calcifications, spiculated masses, masses containing calcifications, round masses, or architectural distortion. Less common mammographic signs of cancer include a focal asymmetry, a developing asymmetry, breast edema, lymphadenopathy, a single dilated duct, or the cancer can be mammographically occult (nothing is seen; Table 2.1 ). The radiologist has to perceive the finding, recognize it as abnormal, and correctly interpret it as “actionable” (something to be acted upon; Box 2.5 ). Calcifications are discussed in detail in Chapter 3 , breast masses in Chapter 4 , and findings associated with clinical problems in Chapter 10 .

TABLE 2.1
Mammographic Findings of Breast Cancer
Finding Differential Diagnosis
Pleomorphic calcifications Cancer (most common), benign disease, fat necrosis
Spiculated mass Cancer, postsurgical scar, radial scar, fat necrosis
Mass with calcifications Cancer, fibroadenoma, papilloma; exclude calcifying oil cyst
Round mass Cyst, fibroadenoma, cancer, papilloma, metastasis
Architectural distortion Postsurgical scarring, cancer
A focal asymmetry Normal asymmetric tissue (3%), cancer (suspicious: new, palpable, a mass containing suspicious calcifications or spiculation)
Developing asymmetry Cancer, hormone effect, focal fibrosis
Breast edema Unilateral: mastitis, postradiation therapy, inflammatory cancer
Bilateral: systemic disease (liver disease, renal failure, and congestive heart failure)
Lymphadenopathy Unilateral: mastitis, cancer
Bilateral: systemic disease (collagen vascular disease, lymphoma, leukemia, infection, and adenocarcinoma of unknown primary)
Single dilated duct Normal variant, papilloma, cancer
Nothing 10% of all cancers are false-negative on mammography

BOX 2.5
Steps in Radiologists Recognizing Cancer on Mammograms

  • Radiologist sees the finding

  • Radiologist recognizes the finding is different from normal tissue

  • Radiologist correctly interprets the finding as abnormal/possibly abnormal

  • Radiologist acts on the finding (recall/biopsy)

Between 10% and 15% of breast cancers are mammographically occult, which means that breast cancer is present but the mammogram is normal. Dense fibroglandular background tissue can hide up to 30% to 50% of cancers. Accordingly, if there are suspicious clinical symptoms or physical findings and the mammogram is negative, the decision for biopsy should be based on clinical grounds alone.

Next is described an approach to the mammogram using a consistent, reproducible viewing display, a systematic search pattern, a list of “danger zones” in which cancers are commonly missed, and detailed methods to find these cancers.

Display of Mammograms

Optimization of Reading Room and Image Display

If the mammograms are screen-film studies, the images are viewed on high-intensity viewboxes with the light parts of the films masked to block extraneous light. For full-field digital mammograms (FFDMs) and tomosynthesis images viewed on soft copy, the images are displayed on high-resolution bright monitors in a dark room with little to no ambient light, comparing old studies with new ones in the display protocol. Additionally, an ergonomic setup for the radiologist as described in our article on repetitive stress injury in breast radiologists will help the radiologist avoid injury ( ).

First Look at Two-Dimensional Views and Older Studies

The standard set of mammograms consists of paired MLO views and paired CC views. Normal breast tissue is usually symmetric, or “mirror image.” To evaluate for mammographic symmetry, the MLO and CC views are displayed back to back, and asymmetries are easily identified using the comparison of the right and left breasts ( Fig. 2.8 ). Look at the whitest part of the mammograms for normal fibroglandular symmetry to see if there is more white tissue on one side than on the other (an asymmetry) or if there are any abnormal spots focally whiter than background (a focal asymmetry).

FIG. 2.8, Schematic of viewing normal mammograms to judge the symmetry and change over time. The mediolateral oblique (MLO) and craniocaudal (CC) mammograms are viewed with the right and left sides placed back to back. Older mammograms are placed above to check for change from year to year. R, right; L, left.

Asymmetries can be normal. For example, normal asymmetric glandular tissue is defined as a larger volume of normal fibroglandular tissue in one breast than the other, but with one breast not necessarily being larger than the other; this occurs in about 3% of women. A normal asymmetry should be stable over time.

To detect changes over time, good quality older mammograms are placed above the new ones. Because subtle changes may take longer than a year to become evident, one compares new mammograms with last year’s mammogram, one more than 2 years old, and the oldest images of comparable quality.

Unexplained increases in breast density may indicate breast cancer. An unexplained generalized increase in breast density associated with skin thickening may represent breast edema, which has many etiologies, including inflammatory cancer. An unexplained new asymmetric focal density is called a “developing asymmetry” and should prompt investigation because developing asymmetries represent cancer in 15% of cases.

Comparing old studies with current studies makes it easier to see new or developing changes. A normal mammogram does not usually change from year to year after taking into account the normal involution of glandular tissue ( Fig. 2.9 ). Contrastingly, malignant lesions increase in size and can change the mammographic appearance over time. However, the changes on mammography caused by tumors can be very subtle. Some tumors may infiltrate into the breast tissue without producing an apparent contrast against fat ( Fig. 2.10 ). The doubling time of breast cancer is typically approximately 50 to 200 days, but some can grow more slowly ( Fig. 2.11 ).

FIG. 2.9, Example of normal stable mammograms in viewing scheme. (A) Normal old mediolateral oblique (MLO) and craniocaudal (CC) views are placed back to back above the new views. (B) Normal new MLO and CC views, also placed back to back. Comparing the new and old studies shows no change in dense tissue and a stable benign nodule ( arrow in A and B) in the medial left CC view over a 4-year period.

FIG. 2.10, Developing asymmetry over 3 years. (A) Normal old mediolateral oblique (MLO) and craniocaudal (CC) views are placed back to back above the new views. (B) New MLO and CC views, also placed back to back, showing developing asymmetry ( arrows ) in the medial upper left breast. Although the asymmetry looks subtle, comparison of the new study with the older study performed 3 years before (A) helps to identify this abnormality. (C) Ultrasonography shows two masses in the 11:00 to 11:30 position. (D) A sagittal plane of contrast enhanced magnetic resonance imaging shows masses and nonmass enhancement involving the nipple in the upper left breast. This patient was proven to have invasive ductal cancer.

FIG. 2.11, Natural course of slow-growing breast cancer. Mediolateral oblique (MLO) views obtained 4 years prior (A), 2 years prior (B), and 1 year prior (C) to the current mammogram (D). This patient has an infiltrating carcinoma, but refused treatment. Annual follow-up mammograms show a spiculated mass with skin retraction in the left upper breast. This tumor has been growing over 4 years, but is rather stable. Even invasive cancer can have slow-growing nature over time like this case, although others may grow rapidly. Note the BB skin marker on the palpable mass in A and C.

Systematic Search on Each Mammographic Two-Dimensional View and Tomosynthesis

Once the radiologist judges the mammographic density and asymmetries, they search for masses, calcifications, and distortions on the mammograms. A common search pattern uses zigzags or strips of each image, like mowing a lawn with a lawnmower or searching for a lost boat at sea with a rescue helicopter. For two-dimensional (2D) digital mammography, the radiologist electronically magnifies the mammograms in quartiles, the upper half and the lower half of the 2D MLO views and the inner and outer 2D CC views ( Fig. 2.12 ).

FIG. 2.12, Systemic search on digital mammography. First, display the paired mediolateral oblique (MLO) views and the paired craniocaudal (CC) views with the right and left breasts back to back and do the overall search by comparing the breasts. Look though the entire breast using the zigzag approach. Second, display the paired views of the upper half of the electronically magnified breasts. Focus on the right-side view and search for abnormalities very carefully using the zigzag approach to not miss subtle findings. Then do the zigzag search on the left-side view. Repeat the same process for the lower half of the magnified breasts.

On tomosynthesis, the radiologist first reviews the 2D or synthesized 2D mammogram initially using the zigzag/strip method, then proceeds to magnify the 2D/synthesized 2D mammogram in quartiles because it is harder to see the overall breast in tomosynthesis slices or slabs alone. Then the radiologist scrolls through tomosynthesis slices or slabs. Some radiologists synchronize the tomosynthesis slabs or slices together, back to back, similar to the 2D display, to look for symmetry and auto scroll the tomosynthesis as a movie to get an overall view of the breast. For detailed tomosynthesis analysis, similar to reading a computed tomography (CT) scan or magnetic resonance imaging (MRI), it is important for the radiologist to keep his or her eyes in one place as the movie is scrolled/played to analyze a specific area. Otherwise, the eye is moving and the images are moving, and a finding could be missed. For example, the radiologist scrolls through all the tomosynthesis slices/slabs keeping his or her eye on the upper right breast MLO throughout the series, the lower right MLO, the upper left MLO, and then the lower left MLO. The radiologist then repeats the procedure for the CC studies. The method to review the tomosynthesis is described in Fig. 2.13 . If the radiologist is reviewing the tomosynthesis on slabs, and there is a suspicious finding seen on a slab, then the findings are reviewed on the tomosynthesis slices.

FIG. 2.13, Systemic search on tomosynthesis. Before evaluating the tomosynthesis slices, review the conventional two-dimensional (2D) mammograms or the tomosynthesis-synthesized 2D view using the method described in Fig. 2.12 . Then display the paired mediolateral oblique and the craniocaudal projections of tomosynthesis images with the right and left breasts back to back. Scroll the linked right- and left-side images together, and look through the entire breast comparing both breasts. Next, display the upper half of the electronically magnified right breast, and scroll tomosynthesis images of this series, looking very carefully for abnormalities. Then display the tomosynthesis slices of the lower half of the magnified right breast and check for the abnormality. Repeat the same process for the left breast using the electronically magnified views.

Location of a Finding

This section defines BI-RADS location terminologies, shows how to track findings on mammograms on different projections, how to predict ultrasound locations from the mammograms, how to know the three-dimensional (3D) position of a finding before surgical resection, and how to correlate palpable findings to the mammogram.

Location Description Based on BI-RADS 2013

Finding locations are described by laterality (right or left breast), quadrant and clock face, depth, and distance from the nipple. The breast quadrants describe the breast as the breast was divided into four areas with the nipple at the crosshairs of quadrants and as if the patient is facing the examiner ( Fig. 2.14 ). The upper outer quadrant is the upper breast quadrant closest to the axilla, the upper inner quadrant is the upper breast near the sternum, the lower inner quadrant is the lower inner breast near the sternum, and the lower outer quadrant is the lower breast not near the sternum nor the axilla. The “clock face” description of a breast finding location imagines a clock superimposed on each breast as the woman faces the examiner (see Fig. 2.14 ). This means that the upper outer quadrant in the right breast is between the 9-o’clock and 12-o’clock positions, but the upper outer quadrant in the left breast is between the 12-o’clock and 3-o’clock positions. Although it is simple to see how clock-face lesion location can be easily mixed up (right upper outer quadrant versus left upper outer quadrant), the clock face allows the radiologist and surgeon to describe lesions that fall between quadrants, for example, a lesion located at the 12 o’clock or the 6 o’clock position of the breast. The breast depth includes the anterior, middle, and posterior third ( Fig. 2.15 ) of the breast. Distance from the nipple for a specific lesion is measured as the distance between the root of the nipple and the anterior edge of the lesion on one of the projections that provide the best visualization of the lesion.

FIG. 2.14, Schematic for breast quadrant and clock face. Breast quadrants are divided by cross lines that have the center at the nipple. The clock face location of breast findings is described by imaging a clock on both the left and the right breast as the woman faces the examiner. Note that the outer portion of the breast on the right is at the 9 o’clock position and the outer portion on the left is at the 3 o’clock position. UIQ, upper inner quadrant; LIQ, lower inner quadrant; UOQ, upper outer quadrant; LOQ, lower outer quadrant.

FIG. 2.15, Schematic for breast depth. (A) Schematic of mediolateral oblique (MLO) views. (B) Schematic of craniocaudal (CC) view. Breast depth is classified as the anterior, middle, and posterior third.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here