By definition, a screening mammogram is for the routine surveillance of breast cancer in an asymptomatic patient. It is comprised of two (nearly) perpendicular low-dose x-rays of each breast: a craniocaudal (CC or top-to-bottom) and a medial-lateral oblique (MLO or side-to-side) view. Two images of each breast are the standard of care to be able to include as much tissue in the image as possible; studies have shown a 25–45% increase in cancer detection when two views are obtained compared with one [8]. This is because inherent in how the patient is positioned, each view has a “blind spot” that has the potential of excluding some tissue—even with superb technique, for example, a CC view may exclude some superior tissue, and an MLO view may miss far medial tissue.

The specifics of frequency and age of initiation of screening mammography are controversial. As with any screening program, the goal is early detection: to find early stage cancer that is more susceptible to less aggressive treatments and therefore has the best prognosis and survival . With breast cancer, the 5-year survival rate for localized disease at diagnosis can be up to 98.8%, and 12-year survival up to 95% with tumors smaller than 1cm at diagnosis [9, 10].

The national benchmark for overall sensitivity of mammography is 86.9%, and specificity 88.9% [11, 12]. Mammography is not a perfect test (is there a perfect test in medicine?); however, it is one of the most studied tests in medicine, resulting in a large volume of data, from but not limited to randomized control trials, that consistently shows a significant (at least 30%) decrease in mortality when done annually in average-risk women over 40, through early detection, with decades of follow-up [13–17]. Part of the advantage of mammography is its ability to detect microcalcifications associated with stage 0 disease (intraductal cancer or ductal carcinoma in situ) better than any other imaging modalities, even in women with dense breast tissue. This is because DCIS often manifests as microcalcifications, the detection of which is optimized by the radiographic technique used in mammography (low-dose, high-resolution imaging). Starting at age 40 is recommended because the incidence of breast cancer doubles in women ages 35–39 to ages 40–45 years old, and cancers grow more aggressively in these women than in postmenopausal patients [9, 18, 19]. Although cancers in postmenopausal women may be more indolent, causing some to favor less frequent screening in these patients, increasing age is a risk factor, and so earlier detection may allow for less aggressive treatment options that are better tolerated in the setting of other health problems that may come with age.

Recent changes in screening guidelines are not based on new or refuting data, but rather on a shift of attention away from the benefits of mammography to its potential harms (see below). Simply put, for women of average risk, the US Preventive Services Task Force (USPSTF) recommends biannual screening mammography for women ages 50–74, stating that the net benefit is moderate. For women younger than 50, the position of the USPSTF is that the decision to screen should be an individual one based on values of potential benefits versus potential harms and that the net benefit for women in this group is small. And for women older than 74, the current evidence is deemed “insufficient” to assess the balance of benefits and harms [20].

The American Cancer Society now advises that women should have the opportunity to begin screening at age 40 if they choose, and that mammography be done annually between 45 and 54; women over 54 can be screened every other year or annually, depending on personal preferences [21].

The changes in guidelines, and the controversies, stem from new attention on issues such as patient anxiety , radiation, false positives, and invasive procedures and their complications [22]. These are important topics; however, as discussed below, they do not outweigh the proven benefits of mammography when it comes to early detection and survival.

There is conflicting data regarding short term and long term effects of the anxiety related to mammography, however, there is evidence that women are still willing to return yearly even after a false positive; the apprehension is not incapacitating, nor does it outweigh the desire to “know” one’s status regarding breast cancer [23, 24]. Direct, radiologist-led patient education about screening and breast cancer has also been shown to decrease anxiety through increased patient knowledge and feelings of empowerment [25]. Lastly, significant portion of this anxiety is attributed to waiting for results, and the fear of the unknown. But rather than discourage mammograms for this reason, there are ways to shorten the interval between screening, recall, diagnosis, and biopsy that serve to alleviate this component of anxiety [26]. As outlined above, radiologists are legally required to provide the patient with her imaging results, which is routinely done the same day in the case of a diagnostic workup. But practices may opt to do this even with screening studies, even if only at the specific request of the patient. Our practice routinely accepts add-ons and walk-in patients for screening and diagnostic studies and offers same-day ultrasounds and biopsies. We have an agreement with our pathologists to receive biopsy results the next day, and the radiologist provides those results directly to the patient at that time. If cancer is diagnosed, we immediately schedule the patient for a surgical consultation, and the patient is seen within the week. While this can make for unpredictable workflow, efforts by the entire team to streamline the process are seen as being in the best interest of the patients.

All x-rays use radiation. However, there is no data showing that the radiation from yearly mammograms is a cause of breast cancer. In the spectrum of medical tests, mammography is considered low dose [27, 28]. A standard four-view mammogram (two views of each breast) is approximately 1/7 of the radiation received from natural background sources annually, such as the air, water, and soil in our environment [29]. Moreover, as outlined above, all mammography facilities in the United States are required to undergo routine inspection and accreditation by entities such as the FDA and American College of Radiology: practices are regularly and systematically monitored with regard to equipment, safety, quality, radiation dose, and technologist and physician training. So even though mammography uses radiation, the vast, proven benefits of early detection outweigh the theoretical risks from the relatively small dose of radiation, a dose which is kept in check.

With regard to false positives in mammography, in addition to monitoring data such as equipment and dose, mammography facilities are also required to track measures such as patient outcomes and physician performance. Among many national benchmarks, the acceptable “abnormal interpretation rate” for screening mammography is 5–12% [12]. While this may seem high, the number of patients receiving a normal or benign result is, by definition, substantially higher: about 90 percent of screening patients get a clean bill of health. Of the 5–12% of patients recalled, the vast majority will also be “cleared” with additional mammographic views and/or ultrasound. Of the remaining, biopsy rates are less than 2 percent; benchmarks for acceptable PPV for biopsies performed are 20–45% for screening-detected abnormalities and 30–55% for palpable findings [12]. Many breast imagers work hard to minimize recalling patients, and even lower recall rates and higher positive predictive values can be acheived through better technologist and physician training [30]. For example, for about a decade, the collective callback rate in our practice was less than 10% and our PPV for biopsies 50–60%. Double reading has also been shown to decrease call backs [31, 32]. In addition to better training and collaboration, advances in technology can contribute to lowering recall rates and false positives. 3-D mammography, or tomosynthesis, is an example of this. Instead of a single, “flat,” 2-dimensional image, these 3-D studies provide the radiologist with multiple separate thin (1 mm) “slices” through the entire thickness of the breast that can be evaluated layer by layer. This can decrease the effect of summation or superimposition of normal structures encountered more frequently with the traditional (2-D) mammogram, thereby decreasing false positives, and allow for the increased detection of invasive cancers [33–35].

With regard to the potential harm of invasive procedures, when a biopsy is necessary, there are opportunities to minimize local trauma and, therefore, associated discomfort and possible complications. Imaging-guided core needle biopsies are less invasive than surgical biopsies, can be done in the same exam room as the mammogram or ultrasound, and require no IV or general anesthesia, nor advanced preparation (such as fasting or stopping anticoagulation) by the patient. The complication rate of imaging-guided needle biopsy is <1%, which includes bleeding, infection, and tissue damage [36, 37]. Even then, there are opportunities for improvement. For example, smaller gauge, spring-loaded biopsy needles may be used instead of larger, vacuum-assisted ones in certain circumstances. Physician skill and more precise techniques can allow for taking fewer, “high-yield” samples (my mentor once said that theoretically, you only need one core to make the diagnosis—which drives me to make each pass count to this day). Using a patient-centered and specific approach to decide which patients may or may not benefit from placement of a marker clip, which is routinely placed at the biopsy site by almost all radiologists and requires a two-view post-procedure mammogram to confirm its location, may obviate the cost, extra time, and radiation associated with this part of the procedure.

In these ways, the potential harms of screening can be addressed and overcome, rather than being used as excuses to discourage annual mammography. To discourage or possibly limit access to this lifesaving test going forward may undo all of the gains in early detection and survival made previously.

It is important to reiterate here that the controversies in screening guidelines are with regard to the average-risk patient. Most organizations still agree that women who are high risk (see below) should start annual screening at 40 or earlier based on their risk factors such as age of onset of cancer in a first-degree relative.

Without getting into the technical aspects of diagnostic workups, some of the additional mammographic images include spot compression views at the area of radiographic concern and full paddle views done at different angles than standard screening views. Altering the angle at which tissue is seen, and further compressing the breast in the specified area, can reduce the potential masking effect of normal overlapping tissue and allows for confirmation, improved visualization, and localization (for possible ultrasound or biopsy) of the finding in question. In the case of a mass, confirming it in two (perpendicular) planes is integral to the BI-RADS definition of a mass. If the suspected finding “disappears,” it may be attributed to tissue overlap at the time of screening, and no further workup may be needed. As an analogy, for example, when we take our laundry out of the washer, it is often “balled up.” When spread on the clothesline, we see that there was never really a “ball” in it. The radiologist may perceive a “mass” or “lump” at screening that “spreads out” as normal tissue once viewed from a different angle, and further, focal compression is applied. If a finding persists/is confirmed on spot compression views, ultrasound may be undertaken, particularly in the evaluation of a mass, architectural distortion, or focal asymmetry.

Breast tissue is composed of fatty, fibrous, and glandular tissue. These elements vary in proportion from person to person and sometimes between breasts in the same patient (see discussion on asymmetry below). The amount and combination of types of tissue are under genetic and hormonal influences, and though it can change somewhat during a lifetime, it is simply how an individual woman’s breast is made.

Fatty tissue is radiolucent, translating to a higher sensitivity for mammography to detect small cancers, since invasive cancers (most often, masses) are typically equal or higher density than breast tissue. Fibrous and glandular tissue, sometimes termed together as fibroglandular tissue, is more opaque or “dense” radiographically. So it follows that the more fibrous and glandular the tissue, the denser the breast tissue appears, and the more challenging it is to detect a small cancer. It is the old polar bear in a snowstorm analogy, hence the decreased sensitivity of mammography in dense breasts (Figs. 1.1 and 1.2). To add to this limitation in detection, studies also show a mild to moderate increase in cancer risk incurred by having predominantly dense tissue, the mechanism of which is still unclear [38].

The lowered sensitivity of mammography combined with increased cancer risk associated with dense tissue has driven patient advocates and politicians to establish state laws requiring radiologists to directly inform patients about their breast density in more than half of the United States—known as breast density notification legislation. The goal of this is to provide women information to allow them to make more informed decisions regarding screening and breast health. At our institution, the statement added to result letters reads:

Your mammogram demonstrates you have dense breast tissue. Dense breast tissue is very common and is not abnormal. However, dense breast tissue can make it harder to find cancer on a mammogram and may also be associated with an increased risk of breast cancer. This information about the result of your mammogram is given to you to raise your awareness. Use this information to talk to your doctor about your own risks for breast cancer. At that time, ask your doctor if more screening tests might be useful, based on your risk.

Of note, by MQSA requirements, density information has always been included in the radiology report that is sent to the ordering clinician; it is only in recent years that individual states are mandating this information be included in the patient result letter as well.

However, the benefit of this information is not as clear cut as it seems and has given pause to radiologists and patients alike, especially amidst the controversies about screening guidelines. Some radiologists’ threshold for classifying tissue as “dense” is when the mammogram is > 50% dense. This was further divided into categories of “heterogeneously dense” (51–75%) or “extremely dense” (>75%) by older editions of the BI-RADS lexicon; the current edition does not provide percentage guidelines, reflecting that the majority of radiologists make this assessment subjectively [39]. As would be expected, there is much (documented) inter- and intraobserver variability in breast density assessments [40]. Software programs exist to objectively quantify dense tissue; however, these are in large part investigational, and currently not in routine clinical practice. Secondly, density is also affected by radiographic technique. For example, if the image is undercompressed or underpenetrated, breast tissue can appear artificially dense. Making a determination that tissue is dense has serious ramifications for the patient, not the least including anxiety and the possibility of additional tests, and therefore this assessment should be as accurate and reproducible as possible.

Another issue is that breast density does not tend to affect the ability to detect microcalcifications associated with early breast cancer, DCIS, to the same degree as it does for small masses. Calcifications are denser, or “whiter,” than dense breast tissue and can therefore still be seen in a background of dense tissue. Therefore, discouraging women from mammography because of dense breast tissue may result in missing the opportunity to find an early-stage intraductal cancer, before it has a chance to progress to invasive disease and form a mass obscured by overlying tissue (Fig. 1.3).

In a similar vein, one of the myths amidst the screening controversies is that mammograms are ineffective in young women because they mostly have dense breast tissue. This is untrue for two reasons. That all young women have dense tissue is a myth; and as indicated above, microcalcifications of DCIS are still apparent in dense tissue.

Women with dense breasts may undergo supplemental screening; however, the trade-off for the increased cancer detection may be increase in false positives, particularly with ultrasound. Perhaps even more of an obstacle is that while breast density notification legislation obligates radiologists to inform patients of their density, in most states, the legislation does not mandate insurance companies to cover supplemental screening tests such as whole-breast ultrasound or MRI, which cost significantly more than screening mammograms. So while we empower patients with information, their ability to act on it may be limited.

Ultrasound can be used as first line for diagnostic purposes when the risk of even low-dose radiation to the breast tissue outweighs its benefits, mostly when breast tissue is more “active” under strong hormonal influences. This is the case for women under 30, given the continued growth of breast tissue into the 20s, versus the low likelihood of cancer in this age group. Ultrasound is also used preferentially in patients who are or were recently pregnant or breast feeding. Patients who have undergone a mastectomy can also be imaged with ultrasound if presenting with a symptom that requires imaging evaluation.

Otherwise, sonography (ultrasound) should be used as an adjunct to mammography, to further characterize a palpable or radiographic finding and to provide biopsy guidance. Although the utilization of screening ultrasound is increasing with recent breast density notification legislation, even when these are recommended, it is in the setting of concomitant mammography, as the sensitivity and specificity of ultrasound for cancer detection are higher when combined with mammography than when used alone [41, 42]. As mentioned earlier, mammography is not a perfect test; however, we know that some cancers—e.g., intraductal and even some invasive lobular carcinomas—may be occult or at best subtle sonographically. Mammography is the gold standard for cancer detection, and we would be remiss if we start substituting ultrasound for patients who simply don’t want to undergo a mammogram.

Quality assurance programs similar to those established for mammography are also in place for ultrasound. Though not required by law as for mammography, facilities may participate in the voluntary peer-reviewed ultrasound accreditation process by the ACR, comparable to that required for mammography. This accreditation for ultrasound is mandated by some insurance companies for reimbursement. Similar to mammography, the ultrasound accreditation program, which separates diagnostic ultrasound and ultrasound-guided interventions, assesses issues such as radiologist and technologist qualifications and experience, equipment, image quality, documentation, needle positioning, and radiologic-pathologic concordance and patient outcomes with regard to biopsies and fine needle aspirations.

With regard to day-to-day practice of ultrasound, the ACR has also set practice parameters for image acquisition and annotation. These are not merely boxes to be checked for accreditation, but, when followed, allow for optimum image quality and therefore the best possible visualization and categorization of the finding and subsequent management of the patient. Being mindful of and adjusting technical parameters (such as field of view, focal zone, depth) appropriately must be a part of every patient’s scan to obtain accurate diagnostic information. Ensuring that ultrasound findings (with regard to lesion location and imaging characteristics) are concordant with the mammogram is also paramount and is the responsibility of the radiologist even if a sonographer acquires the images. For example, if working up a mammographic finding of a spiculated, solid mass in the upper inner quadrant, a cyst (which is characterized by circumscribed margins and absence of internal echoes) found in the upper outer quadrant would be considered incidental and incongruent with the mammographic finding, which would still need to be identified. Confirmation of findings in perpendicular planes, similar to mammography, is also necessary.

For breast MRI to be useful in the diagnosis of breast cancer, it must be done with the administration of intravenous gadolinium-based contrast, following which multiple “runs” of repeat imaging are done to provide a dynamic set of images over time. This is because most cancers have avid, rapid influx and outflux of contrast due to increased vascularity and vascular permeability compared with normal tissue, related to tumor angiogenesis [43, 44]. In addition to the visual assessment of lesion morphology and contrast uptake, special software is used to process the kinetic curves of this enhancement. For example, lesions that demonstrate “fast” initial enhancement (>100% signal intensity in the first 2 minutes after contrast injection) and “washout” on delayed images (decrease in signal intensity by >10% from peak enhancement) have the highest likelihood of malignancy.

That being said, hormonal influences can cause (sometimes marked) normal background parenchymal enhancement from which the radiologist must tease out possible lesions (Fig. 1.4). Abnormal findings at MRI may be further worked up with targeted ultrasound and/or biopsy (either sonographically or MRI guided).