Combined Clinical and Imaging Workup of Breast Abnormalities

Once the referring physician finds a suspicious breast mass or receives a suspicious mammographic report, the patient undergoes a thorough history and a focused breast examination. Usually a breast cancer specialist (commonly a general surgeon with an interest in breast cancer, a dedicated breast surgeon, or a surgical oncologist) then estimates the patient’s risk of having breast cancer, seeks patterns of familial breast cancer, and helps the patient to make an informed decision about imaging versus immediate intervention.

Most patients are then referred for a thorough diagnostic imaging workup. Patients with suspicious palpable abnormalities undergo ultrasound, with or without mammography, depending on age, family history, and level of concern over the finding. For example, ultrasound would likely be the sole imaging modality in an 18-year-old woman with a new breast lump and no family history of breast cancer. On the other hand, ultrasound and mammography would likely be used for a 25-year-old woman with a new palpable lesion and an extensive family history of breast cancer in young relatives. The final decision as to whether or not to incorporate mammography into a very young patient’s workup is a shared responsibility of the clinician directing the breast workup, the radiologist performing the initial imaging (ultrasound in this case), and the patient. Breast magnetic resonance imaging (MRI) is used sparingly during the initial evaluation of a palpable finding, unless there is an extensive family history, in which case it serves a dual role as both a diagnostic tool on the affected breast and a screening tool on the contralateral breast.

For patients with nonpalpable findings on screening mammography, workup always includes diagnostic mammography. For suspicious calcifications alone, the radiologist usually obtains magnification mammograms, often not needing ultrasound. An exception might be extensive pleomorphic microcalcifications, in which ultrasound might be used to search for masses within the area that could be indicative of invasive cancer, prompting biopsy. However, if there is an image-detected mass, area of architectural distortion, or palpable mass, the radiologist usually uses both mammography and ultrasound to evaluate the abnormality, estimate its size, and direct later biopsy. Breast MRI may be valuable in selected cases, as discussed in Chapter 7. Ideally, the radiologist correlates all physical and imaging findings in the report to form a composite picture of all potential abnormalities and their level of suspicion on mammography, ultrasound, and MRI.

Using the combined report, the directing clinician and radiologist plan percutaneous or open biopsy to sample all areas of concern. This sequence varies from patient to patient. This may be as simple as FNA in a young woman with a single area of fibrocystic nodularity and a normal ultrasound or as complex as numerous core biopsies or surgical biopsies in one or both breasts using palpation or image guidance for localization.

Although there are no hard and fast rules about what defines a “suspicious” palpable abnormality, in general cancers are firm or hard, asymmetric compared with the opposite breast, irregular in shape, and feel as if they are rising up out of the breast tissue, rather than spreading out in the substance of the breast. Physical examination, ultrasound/mammography, and FNA are generally considered the “minimum” intervention for a suspicious palpable finding and in combination are referred to as the triple test. For suspicious palpable findings in which all components of the triple test are negative, the risk of malignancy is considered approximately 3% or less. Even if all components of the triple test are normal, it is extremely important to inform the patient that there is a low, but measurable, false-negative rate for the triple test and that surgical excision can be performed to completely exclude the possibility of malignancy. This discussion is ideally documented in the medical record. Patients with likely benign palpable findings, unremarkable imaging, and normal percutaneous sampling with FNA (i.e., a negative triple test) usually undergo a single follow-up visit 3 to 6 months later with the referring physician. Patients who undergo image-guided core biopsy usually undergo repeat imaging 6, 12, and 24 months later to assess stability of any residual findings. Progressive findings on repeat palpation or breast imaging at follow-up prompt surgical excision.

For suspicious image-detected nonpalpable lesions, image-guided FNA, percutaneous core biopsy, or wire-localized excisional biopsy is generally considered the “minimum” intervention. Percutaneous needle core biopsy has become the more common choice. Here, too, it is important to inform the patient of the limitations of percutaneous core biopsy—specifically, that there is a small false-negative rate with needle biopsy. Wire-localized surgical biopsy is usually recommended to exclude malignancy if percutaneous biopsy is indeterminate or discordant with imaging findings. For an anxious patient, wire-localized surgical excision may be a better option initially; the surgeon will usually document this discussion in the medical record, too. For some patients with lesions close to the chest wall or nipple, wire-localized excisional biopsy may also be the safer initial approach.

Breast Cancer Diagnosis and Treatment

When a combined clinical and imaging workup leads to a breast cancer diagnosis, treatment planning usually involves a consideration of surgery, chemotherapy, and radiation therapy, with the goal to remove all the cancer from the breast, optimize chances for locoregional control, and eradicate occult foci of metastatic disease via systemic treatment (e.g., hormone therapy, chemotherapy), if indicated. The team of breast imagers, surgeons, medical oncologists, pathologists, radiation oncologists, and breast reconstruction surgeons plan the sequence in which surgery, chemotherapy, and radiation occurs. The pathology report is a key component on which treatment is based. The report states tumor histology; size; estrogen, progesterone, and her2neu receptor status; and lymph node involvement. Traditionally, breast tumors are staged using the TNM (tumor, lymph node, metastasis) Classification on Breast Cancer from the American Joint Committee on Cancer (currently in the 7th edition) (Table 8-1). The treatment plan is based on this classification. A clinical decision algorithm is also available from the National Comprehensive Cancer Network regarding the full spectrum of care; Adjuvant! Online is an Internet-based tool that provides guidance regarding prognosis and the potential benefit of different chemotherapy protocols. Additional tests based on tumor gene signatures are emerging (OncoType DX and MammaPrint) and are the subject of two large randomized trials, one in the United States and the other in Europe. Gene expression profiling may play an increasingly important role in the future; preliminary data suggest improvement in separating high- and low-risk patients.

Locoregional control of the cancer means that the patient undergoes surgical removal of the cancer with a margin of normal breast tissue. (Although the definition of an acceptable margin varies from institution to institution, the more commonly followed models range from simple nontransection [tumor not on inked margin] to 2 mm of normal tissue at the margin.) The patient achieves this locoregional control by either breast-conserving surgery, usually followed by whole-breast irradiation, or mastectomy. As shown by Protocol B-06 conducted by the National Surgical Adjuvant Breast and Bowel Project (NSABP), both approaches yield equivalent local control and identical survival rates in women with tumors 4 cm or smaller in diameter whether the axillary lymph nodes are positive or negative for metastatic disease.

The radiologist helps the team select candidates for breast-conserving surgery or mastectomy by estimating the location and extent of disease. The critical information the surgeon requests relates to lesion location and size. This allows the surgeon to form a three-dimensional (3-D) representation of normal versus malignant tissue, develop a mental image of the tumor within the breast, estimate the amount of additional tissue needed to obtain tumor-free margins, and plan the incision (surgical approach) with the goal of maximizing probability of tumor removal while preserving cosmesis as best as possible. For example, it is difficult to remove an extensive ductal carcinoma in situ (DCIS) completely with microscopically clear margins, and these patients are usually treated with mastectomy. Increasingly, there is interest (though no good randomized data to support) in removing multiple lesions from the same breast while preserving the breast. Multifocal disease refers to lesions in the same quadrant; multicentric disease refers to lesions in separate quadrants. As a straightforward example, a 3-mm satellite lesion is almost always amenable to resection with a primary lesion using breast-conserving techniques with an acceptable cosmetic outcome. In contradistinction, a pair of 3- to 4-cm lesions on opposite sides of the breast are usually treated with mastectomy. There are no hard and fast rules for excising multiple lesions with breast conservation, and excellent clinical judgment must be used. For this reason, in the setting of multiple lesions, the surgeon requests information regarding the number and size of the lesions, as well as their geographic relationship to each other. If too many foci of invasive cancer or extensive DCIS are present, the patient is not a candidate for breast-conserving surgery because the surgeon would have a hard time excising all the cancer and because of concern over an elevated risk of IBTR.

In general, surgeons perform mastectomy when the entire cancer cannot be excised with a good cosmetic result (as just discussed), if the woman has a contraindication to radiotherapy, or if it is the patient’s desire. Usually, patients are offered ipsilateral breast reconstruction with an autologous tissue flap or a tissue expander after mastectomy, unless there is a medical contraindication to reconstruction (e.g., multiple co-morbidities). Because the contralateral breast is often larger than the reconstructed breast, patients may also need reduction mammoplasty on the contralateral side. Characteristic appearances of reduction mammoplasty and breast reconstruction are discussed in Chapter 9.

If the patient has breast-conserving surgery, she usually undergoes postsurgical whole-breast irradiation to achieve control of residual microscopic disease. Relative contraindications to radiation therapy include pregnancy, previous radiation therapy, and collagen vascular disease (Box 8-1). Axillary nodal involvement is not a contraindication. Six randomized trials of lumpectomy and radiation therapy showed that the frequency of local recurrence and overall survival rates are generally comparable to mastectomy. However, IBTRs are reported in 5% of patients at 5 years and in 10% to 15% at 10 years after completion of therapy. Treatment failures (i.e., IBTR) usually undergo salvage mastectomy.

Invasive IBTR usually occurs in the lumpectomy site or quadrant within the first 7 years, but rarely earlier than 18 months after treatment. IBTR after 7 years will more likely occur in any quadrant, not necessarily at the original site, and is usually considered a new cancer. IBTR near the original lumpectomy site is associated more frequently with systemic relapse than IBTR in other quadrants, which more often reflect a new primary tumor. IBTR is considered more likely in women who have invasive ductal cancer with an extensive intraductal component, residual disease in the breast, extensive DCIS, lymphatic or vascular invasion, or multicentricity, and is more common in younger women (Box 8-2).

Evaluation of Axillary Lymph Nodes

The treatment of invasive breast cancer has historically involved removal of ipsilateral axillary lymph nodes. This was natural, because most women receiving treatment for breast cancer 100 years ago had nodal involvement. With earlier detection of breast cancer, nodal involvement is no longer the norm. In fact, approximately 65% to 70% of women with newly diagnosed invasive breast cancer have normal lymph nodes and therefore will not derive any benefit from axillary lymph node dissection (ALND).

ALND is also problematic from the standpoint of side effects. It exposes patients to the risk of major complications such as lymphedema, shoulder dysfunction, and sensory changes in and around the axilla. To address this problem, routine level I/level II ALND (Table 8-2) has evolved to use the SLN biopsy as an initial screen for nodal involvement in patients who are clinically node-negative.

Table 8-2 Location of Lymph Nodes Draining the Breast

Level	Location
I	Infralateral to lateral edge of the pectoralis minor muscle
II	Behind the pectoralis minor muscle
III	Between the pectoralis minor and subclavius muscles (Halsted ligament)

SLN biopsy was initially described for patients with penile cancer, but did not attract much attention until it was broadly adopted for use in melanoma patients. SLN biopsy is performed by injecting a tracer material, either a radionuclide, blue dye, or both into the breast either preoperatively or perioperatively and by looking for evidence of the tracer in one or more sentinel nodes (Box 8-3).

SLN biopsy alone does not eliminate, but does significantly decrease, the risk of developing the common complications of lymphedema. A level I/level II ALND is now most commonly performed contingent on identification of tumor in one of the sentinel lymph nodes.

The role of the radiologist is to understand the rationale for SLN biopsy and to facilitate its performance. First, the radiologist should not inject tracer into the biopsy cavity or the tumor; tracer injected into a biopsy site cavity is likely to remain in the cavity rather than be transported into the lymphatics. The most common tracers are technetium-99 sulfur colloid and lymphazurin blue; some also use methylene blue dye.

Preoperative lymphoscintigraphy is used in some facilities to assist preoperative localization of sentinel lymph nodes in the axilla or in extra-axillary sites (Fig. 8-1A to C). Most commonly these extra-axillary sites will be in the supraclavicular, infraclavicular, or internal mammary groups. If tracer does not identify an axillary SLN, the surgeon may choose to harvest an SLN from one of these other sites. Some facilities do not remove an internal mammary SLN or other nonaxillary SLN due to the very low frequency of isolated positive biopsies (usually <3%) and the relatively few cases that would result in meaningful changes in prognosis or therapy. Perhaps not surprisingly, institutions that harvest both axillary and internal mammary sentinel lymph nodes have demonstrated a poorer prognosis when lymph nodes at both sites are involved.

Although there are differences of opinion as to the “optimal” location of tracer injection, as well as “optimal” tracer modality, there is general agreement from randomized studies that the technique is sensitive and specific enough to obviate the need for a full ALND in patients whose sentinel nodes test negative for tumor. In general, the SLN is harvested at the time of surgery and tested with touch preparation or frozen section intraoperatively. If there are tumor cells in the SLN, the surgeon proceeds to a completion level I/level II ALND. Nonvisualization of the SLN on lymphoscintigraphy does not preclude SLN identification by the surgeon in the operating room. The SLN may be within thick adipose tissue that can only be identified by the gamma probe in the operating room. The yield for SLN identification in the operating room when it cannot be visualized on lymphoscintigraphy can be increased if blue dye is also used.

Intraoperative evaluation of sentinel lymph nodes occasionally yields false-positive findings. More commonly, false-negative findings occur. This can precipitate return of the patient to the operating room weeks after the original SLN biopsy for completion ALND.

Based on current American Joint Committee on Cancer (AJCC) guidelines, nodal staging is based on the maximal size of the single largest tumor deposit in an SLN (if the SLN is the only involved node) as well as the number of involved lymph nodes. The descriptive category for the smallest extent of disease, isolated tumor cells, means that no single tumor deposit in an axillary node is larger than 0.2 mm. Patients with isolated tumor cells are considered to have normal nodes and are usually not treated with a completion ALND. Proceeding from SLN biopsy alone to the wider axillary node clearance typically requires micrometastatic (>0.2- to 2-mm tumor cell cluster in an SLN) or macrometastatic (>2-mm focus) disease within one SLN. Management of the axilla is performed independent of the decision to pursue lumpectomy or mastectomy.

Not all patients are candidates for SLN biopsy. For example, patients who present with clinically involved axillary nodes usually proceed directly to ALND. However, it is important to exercise caution in declaring an axillary lymph node as clinically positive. With the increased frequency of percutaneous core biopsy, more and more patients are presenting to breast cancer specialists with enlarged “reactive” nodes. A recent study by experienced breast surgeons demonstrated that clinical examination in this setting often overestimates the probability that lymph nodes are involved, which in turn could overestimate the number of patients who proceed directly to ALND. Although SLN biopsy has been widely adopted as a precursor to a full ALND for most patients, many have sought to use imaging studies to obviate the need for SLN biopsy or ALND. Toward this end, investigators have assessed the preoperative appearance of nodes on mammography, ultrasound, MRI, and even positron emission tomography. Among these, only positron emission tomography with a high standardized uptake value may provide near-definitive proof of nodal involvement preoperatively in the absence of percutaneous sampling. Here, too, one must be careful to distinguish between a reactive node versus an uninvolved node.

One preoperative axillary imaging method that has gained a following is axillary lymph node ultrasound with percutaneous FNA of suspicious nodes (see Fig. 8-1D to G). Although this test is not a routine part of the initial breast imaging evaluation, there is a new appreciation for preoperative evaluation of ipsilateral axillary lymph nodes in the setting of breast cancer. Axillary ultrasound is particularly helpful when the results of clinical examination of the axilla are suspicious for cancer. Several studies have recently been published using ultrasound-guided FNA or core biopsy to document nodal involvement preoperatively, thus allowing the surgeon to bypass SLN biopsy. This can obviate several known issues with intraoperative assessment of sentinel lymph nodes, such as the time needed to harvest one or more nodes, the intraoperative time needed for pathology to evaluate the node and, most important, the potential for false-negative touch preparation or frozen section at the time of surgery, which can lead to reoperation at a later date.

Clinical and Breast Imaging Factors in Determining Appropriate Local Therapy: Lumpectomy or Mastectomy

The therapeutic options for local control of a breast malignancy are lumpectomy (almost always followed by radiotherapy) and mastectomy. Lumpectomy (followed by whole-breast radiotherapy) was introduced approximately 40 years ago and offers equivalent survival to mastectomy. Mastectomy has a slightly lower risk of local recurrence than lumpectomy and obviates the need for radiotherapy in most patients. The use of postmastectomy radiotherapy is controversial in premenopausal women with one to three involved nodes (see the meta-analysis of randomized studies with and without radiotherapy by the Early Breast Cancer Trialists’ Collaborative Group, Lancet 2005), but it is a common recommendation for women with tumors larger than 5 cm or with four or more involved nodes. The equivalence in overall survival between lumpectomy with radiotherapy and mastectomy was shown in Protocol B-06 conducted by the NSABP and the Milan I trial conducted in Italy.

The breast imager plays a critical role in aiding the surgeon to make the right therapeutic choice by showing how much cancer is in the breast. There is virtually no disagreement that patients with a unifocal DCIS or invasive cancer may be treated with breast conservation therapy if the entire tumor can be removed with a good cosmetic result and if there are no relative contraindications to radiation therapy (i.e., pregnancy, collagen vascular disease, poorly defined or multicentric disease) or prior radiotherapy involving the breast (Fig. 8-2).

The controversy regarding the best surgical approach concerns patients with multifocal disease. Some physicians believe that mastectomy is the proper choice for such patients. This preference may be due to results from the original clinical trials comparing lumpectomy with mastectomy, which involved almost exclusively women with unifocal breast cancers. Hence, the safety of breast conservation with respect to local recurrence, distant metastasis, and survival is not as well documented in women with multifocal disease. Still, surgeons are increasingly offering breast conservation to patients with multifocal disease. Thus, there is no hard and fast rule regarding how many satellite lesions, or what distance between lesions, constitutes an absolute indication for mastectomy. It is the physician’s clinical judgment to avoid predisposing the patient to IBTR; recent data suggest that an IBTR may increase the risk of distant metastasis and death from breast cancer.

Whether the surgeon offers lumpectomy to patients with unifocal disease alone or to patients with multifocal disease, tumor-free margins are a must. For example, offering a woman breast conservation may be reasonable if she has multifocal invasive carcinoma with sub-centimeter lesions 3 mm apart and margins that are tumor-free by several millimeters. On the other hand, breast conservation may not be offered if a patient has multifocal high-grade DCIS scattered over an area of 5 to 6 cm with only a 1-mm margin; in this example, one would be concerned about additional multifocal disease just beyond the surgical margin.

The definition of tumor-free margin varies among institutions, with some accepting the NSABP model of nontransection, and others requiring a 2-mm or greater tumor-free margin. In general, the margin status must be carefully considered in patients with multifocal disease. Ideally, these patients should have the multiple lesions resected in continuity to gain the best histologic understanding of size, extent, and relationship of lesions to one another, and of the true margins.

Proof of multicentric disease has been handled by some surgeons with breast conservation, but the more accepted, and proven, route is with mastectomy as initial treatment. As stated previously, no prospective, randomized study to date has evaluated the safety and effectiveness of breast conservation therapy in the setting of multifocal or multicentric disease. Retrospective studies have been published suggesting that this approach may be safe by demonstrating comparable local recurrence rates in multifocal as well as unifocal disease, whereas others suggest higher IBTR rates. These studies are not powered to draw definitive conclusions but do suggest that such a randomized study in the future may be worthwhile.

Preoperative Imaging

Mammography, ultrasound, and MRI for tumor extent are important tools for selecting appropriate breast conservation therapy candidates and planning surgery (Table 8-3). Mammography is the mainstay for determining extent of disease. Mammography identifies diffuse or multicentric disease by finding suspicious breast masses and pleomorphic calcifications. Mammography also can identify benign, extensive, innumerable bilateral calcifications that could hide early tumor recurrence. Such calcifications are a relative contraindication to breast conservation therapy. Furthermore, mammography finds DCIS that is invisible to MRI. Specifically, approximately 25% of DCIS cases are false-negative on MRI and are discovered only by visualizing pleomorphic calcifications on the mammogram.

Table 8-3 Breast Imaging Relating to Breast-Conserving Therapy

Timing	Reason	Technique(s)
Preoperative	Ipsilateral tumor extent and contralateral tumor	Bilateral mammography US or MRI as warranted

Establish diagnosis Percutaneous biopsy Perioperative Tumor excision SLN identification Preradiation Check for residual tumor Postradiation Baseline/tumor recurrence Ipsilateral unilateral mammogram (initial one at 6 mo, then every 6–12 mo) Evaluate ipsilateral and contralateral breast Bilateral mammogram (12 mo) Clinical problem

MRI, magnetic resonance imaging; SLN, sentinel lymph node; US, ultrasound.

Modified from Dershaw DD: The conservatively treated breast. In Bassett LW, Jackson VP, Fu KL, Fu YS, editors: Diagnosis of diseases of the breast. Philadelphia, 1997, WB Saunders, p. 553.

On the other hand, MRI has been especially useful in predicting tumor extent before the first surgical procedure (Fig. 8-3). Some investigators have claimed particular effectiveness of MRI in women with invasive lobular carcinoma or showing tumor invasion into the pectoralis muscle or chest wall (Fig. 8-4). With respect to invasive lobular carcinoma, several studies have suggested that MRI may be more effective in detecting the extent of disease than physical examination, mammography, and ultrasound. However, false-negative studies in these series have led to mixed opinions regarding the routine use of MRI in staging invasive lobular carcinoma.

Chest wall tumor invasion on MRI was shown by obliteration of the fat plane between the tumor and the pectoralis muscle, with muscle enhancement, and was proven in 5 of 5 cases at surgery (Morris et al, 2000). No muscle involvement was seen at surgery when muscle enhancement was absent in 14 of 14 cases.

MRI also helps exclude candidates for APBI when it finds more than one focus of cancer. Bedrosian and colleagues (2003) reported a 95% tumor detection rate with MRI and a change in surgical management in 26% (69/267) of patients requiring wider/separate excision or mastectomy, with pathologic verification in 71% (49/69).

Overall, these studies show that MRI may be helpful in surgical planning, but they also indicate that MRI prompts a number of unnecessary biopsies because of a relative lack of specificity. MRI also has false-negative results in invasive lobular carcinoma and DCIS. Other data show that MRI may be associated with treatment delay and an increased mastectomy rate and does not decrease the number of fewer positive margins at surgery. The use of pretreatment MRI before definitive breast cancer surgery remains controversial, particularly if one anticipates whole-breast radiotherapy. The literature on this subject is extensive.

When imaging is complete, additional dialogue with the breast imaging team or additional review of imaging studies may be necessary to help the surgeon, medical oncologist, or radiation oncologist properly counsel the patient regarding appropriate treatment options. This involves a review of the original workup to ensure that all potential abnormalities on physical examination have been evaluated and that the breast imaging workup has been completed (such as up-to-date contralateral mammography as well as additional ultrasound or mammographic imaging for lesions previously considered of secondary concern). The thorough combination of abnormalities identified by palpation or on breast imaging helps ensure that any suspicious foci of tumor are evaluated and incorporated into the treatment plan.

Normal Postoperative Imaging Changes after Breast Biopsy or Lumpectomy

To perform a local excision for diagnostic or therapeutic purposes, the surgeon makes a skin incision, removes the mass or wire-localized abnormality, and then closes the subcutaneous tissues and skin. More tissue is excised when removing a cancer to obtain a margin of normal tissue. Usually, the surgeon allows the surgical cavity to fill in with fluid and granulation tissue.

As a rule, mammograms are not often obtained immediately after diagnostic surgical excisional biopsy. However, in the rare cases when a mammogram is obtained within a few days of surgery, mammography shows a round or oval mass in the postoperative site representing a seroma or hematoma, with or without air. This mass represents the biopsy cavity, filled with fluid that should resolve over time (Fig. 8-5A and B). The adjacent breast tissue shows thickening of trabeculae in subcutaneous fat and increased density caused by local edema or hemorrhage. Skin thickening at the incision is usually present. On MRI the biopsy site is filled with blood or seroma. The fluid in the biopsy cavity is high signal intensity on T2-weighted noncontrast fat-suppressed images (see Fig 8-5C to E).

Over the subsequent weeks, the postoperative site resorbs the air and fluid collection; the collection is replaced by fibrosis and scarring, with residual focal skin thickening and breast edema. On MRI the immediate postbiopsy cavity is a fluid-filled structure with surrounding normal healing tissue enhancement for up to 18 months after the biopsy. The biopsy cavity shows high signal intensity, architectural distortion, and a scar that can simulate cancer (Fig. 8-6 and Box 8-4). The biopsy site usually contains fluid from the seroma, which will be bright on T2-weighted images on MRI. Rim enhancement around the biopsy site is normal even if there is no residual tumor and is due to healing. In the ipsilateral axilla, reactive lymph nodes may develop that cannot be distinguished from metastatic disease (Fig. 8-7). MRI after surgery may reveal cancer at the margin edge by showing clumped enhancement or an eccentric residual mass. Although immediate postbiopsy MRI for cancer staging may depict cancer at the biopsy margin, it is more often used to look for cancer elsewhere in the breast away from the biopsy site.

Normal postoperative findings on mammography include architectural distortion, increased density, and parenchymal scarring in at least 50% of patients (Box 8-5). These findings diminish in severity over time (Fig. 8-8A to I). After 3 to 5 years, the findings should be stable on subsequent mammograms. On the mammogram, in 50% to 55% of cases, the biopsy cavity resolves so completely that it leaves no scar or distortion in the underlying breast parenchyma, and only comparison with prebiopsy mammograms indicates that breast tissue is missing. In other cases, the scar appears as a chronic architectural distortion or a spiculated mass more evident on one projection than the other.

The remaining 45% to 50% of patients continue to have variable mammographic findings ranging from spiculated masslike scars to slight architectural distortion (see Fig. 8-8J and K). In still other, more rare cases, seroma cavities persist, appearing as a round or oval mass.

Postbiopsy scars often have a spiculated masslike appearance that can simulate cancer. Spiculated masses should be viewed with suspicion unless one knows that a biopsy was performed in that location. For this reason, it is important to document the date and location in the breast of previous biopsies on the breast history sheet. Some facilities also place a linear metallic scar marker directly on the skin’s biopsy scar before taking the mammogram to show the previous biopsy site. On the mammogram, the linear metallic scar marker on the skin will be near the underlying scar. The skin scar may not be immediately adjacent to the scar inside the breast because the skin is compressed away from the underlying breast parenchyma during the mammogram. If a spiculated mass is seen far from the metallic scar marker, that mass might be cancer rather than a scar. The radiologist reviews the preoperative mammograms to see where the biopsy occurred and correlates the prebiopsy and current mammograms to make this determination (see Fig. 8-8L and M).

Fat necrosis is common after a breast biopsy and usually appears as a radiolucent lipid-filled mass. Mammography is pathognomonic for fat necrosis if it shows lipid cysts or typical calcified eggshell-type rims around a radiolucent center (Fig. 8-9A). The fat necrosis, lipid cyst, and calcifications usually form in the scar, so these findings should be located near any linear metallic scar markers on the skin (see Fig. 8-9B and C).

On ultrasound, the immediate postoperative site shows a seroma or hematoma, breast edema, and focal skin thickening. The fluid collection occasionally contains air. More commonly, the seroma is completely filled with fluid, sometimes containing septa or debris that has varying appearances on ultrasound (Fig. 8-10). Usually the incision can be traced from the biopsy cavity up to the skin and is shown as a linear scar that disturbs the normal breast architecture (Box 8-6).

Later, the fluid in the biopsy cavity resolves and only the fibrotic scar remains. In these cases, ultrasound shows the scar as a hypoechoic spiculated mass that simulates breast cancer, but it should correlate with the postoperative site (Fig. 8-11). Correlating biopsy histories and the physical finding of a scar on the skin distinguishes normal postoperative scarring from cancer. On ultrasound, the spiculated scar often has a “tail” that extends from the scar to the skin, representing the healing biopsy cavity and its adjacent subcutaneous tissue anastomosis (Fig. 8-12).

Breast Imaging before Re-Excision Lumpectomy or Radiotherapy

For patients who have undergone an initial cancer excision but have close or involved tumor margins, additional imaging is warranted before re-excision. If, for some reason, ultrasound was not performed for a palpable lesion, ultrasound may visualize previously unanticipated residual disease around the cavity margins or unsuspected satellite lesions. If a specimen radiograph from the original excision had shown microcalcifications at or near the specimen edge and involved margins were seen histologically, repeat mammography (with wire localization) may be helpful in guiding re-excision. Given the difficulty in identifying microscopic residual disease at the time of reoperation, the treatment team should attempt to utilize any advantage possible to identify residual areas of disease with the intent of targeting them for the surgeon at the time of re-excision.

Patients who have undergone therapeutic excision demonstrating tumor-free margins may still benefit from additional imaging. This is particularly true for women whose lesions presented initially as microcalcifications. Such imaging helps determine the completeness of tumor excision. There is some controversy in the literature as to whether preradiotherapy mammography of the affected breast should become standard to identify possible multifocal disease before committing to radiotherapy. Because the main concern with breast conservation is local recurrence, and because IBTR may affect overall survival, physicians may order postbiopsy mammograms to determine the completeness of tumor excision if there is any question about residual gross disease or residual microcalcifications (Fig. 8-13). These mammograms are particularly appropriate before the initiation of breast irradiation to make sure there is no residual tumor.

When the biopsy shows unexpected cancer, magnification views of the lumpectomy site and remaining breast before re-excision are especially helpful to show unsuspected, suspicious calcifications or masses. Calcifications are important because they could be cancer, but they do not always represent tumors. Dershaw and colleagues found the positive predictive value of residual microcalcifications representing residual tumor to be 69%, with the likelihood that they represent residual tumor being greatest in cases with DCIS and in those with more than five microcalcifications.

When considering radiation therapy, the number and extent of any calcifications is important even when they do not indicate tumor. The radiation oncologist must gauge whether they will be able to detect early recurrences on the post–WB-XRT or APBI mammogram. The presence of extensive innumerable calcifications that limit the radiologist’s ability to find early cancer is a consideration in patient exclusion from breast conservation.

Pre-ABPI MRI aids patient selection by excluding patients with multifocal or multicentric disease. In one study, Horst and colleagues showed that, by using MRI, 9.8% of 51 patients (n = 5) had unsuspected disease, including multifocal (3) or multicentric (1) disease, or pectoral fascia involvement (1), precluding enrollment into APBI studies. It is also important to recognize typical magnetic resonance findings in post-ABPI patients and to distinguish these findings from recurrent or new disease.

Normal Imaging Changes after Radiation Therapy

Treatment Failure or Ipsilateral Breast Tumor Recurrence

The incidence of treatment failure is approximately 1% per year. Women who are at greatest risk for failure include those younger than age 35 (and especially those younger than age 30); women treated for invasive cancer with an extensive intraductal component or infiltrating ductal carcinoma with a large intraductal component; women with intraductal carcinoma of the comedo type; women with intraductal cancer measuring 2.5 cm or greater in diameter; women with multicentric lesions, as suggested in the studies discussed in this chapter, and those treated for more than one synchronous cancer in the same breast; and women with angiolymphatic invasion. Gross residual tumor also has a poor prognosis, but microscopic residual disease may not infer a greater risk of IBTR. Despite the slightly higher tendency for recurrence in these groups, no risk factor is an absolute contraindication to breast conservation; data on lesions with these features in randomized studies more often than not suggest a trend for increased recurrence rather than a statistically significant association.

For women who choose lumpectomy, IBTR rates are approximately 5% at 5 years and between 10% and 15% at 10 years after therapy. Invasive IBTR is most common between 18 months and 7 years after treatment; during this period IBTR more commonly occurs in or around the lumpectomy cavity. IBTR after 7 years more frequently is a random event in any quadrant of the affected breast, not necessarily at the original site, and is usually unrelated to the original lesion in the breast.

Late ipsilateral breast treatment failures consisting of DCIS or invasive tumors smaller than 2 cm may have a better prognosis; thus, some feel it is important to diagnose recurrences near the original tumor or new cancers elsewhere in the breast as early as possible. However, there is no clinical trial evidence supporting this belief. Treatment failures after lumpectomy and WB-XTR are usually treated by salvage mastectomy.

Treatment failures detected on mammography manifest as new pleomorphic calcifications or masses developing in the biopsy site. At times, a breast cancer recurrence is hard to distinguish from the normal postbiopsy scar, which mimics cancer. However, unlike cancer recurrences, the normal postbiopsy scar becomes smaller and less apparent over time on the mammogram, with stabilization at 2 to 3 years. Central fat necrosis may produce a radiolucent center in the biopsy cavity. Thus, it is not normal if the scar grows in size or becomes denser or more masslike. The radiologist suspects recurrent carcinoma and prompts biopsy if the “scar” develops new pleomorphic calcifications or becomes more dense, if the “scar” edge becomes rounder, or if the “scar” grows (Fig. 8-20).

Recurrent tumor in the irradiated breast may arise at the site of the original tumor or elsewhere in the breast (Fig. 8-21). Recurrences at the original tumor site are usually due to failure to eradicate the original cancer and represent true treatment failures; they occur sooner than a tumor developing elsewhere in the breast. Tumors developing outside the treated area occur at the same rate as tumors forming in the contralateral breast and represent new cancers. Breast irradiation does not lead to an increased incidence of breast cancer in the opposite breast or in the boosted area of the treated breast.

Recurrent disease is diagnosed by mammography or physical examination. About half of the recurrences are detected by mammography and half by physical examination. Those that are mammographically detected usually contain pleomorphic microcalcifications or masses (Box 8-9). Palpable recurrences are usually manifested as masses, are more frequently invasive cancer, and may be displayed on the mammogram as developing densities or masses. On ultrasound, an IBTR shows as a mass separate from or in continuity with the biopsy scar (Fig. 8-22A to D) if it occurs near the original biopsy site.

On mammography, breast cancer recurrences contain pleomorphic calcifications (see Fig. 8-22E and F) or are shown as masses with or without calcifications. This is why radiologists investigate any new mass, because even benign new solid masses may represent a new cancer (Figs. 8-23 to 8-25).

There are no absolute guidelines for management of an IBTR after lumpectomy and radiotherapy. Traditionally, because the breast can only tolerate the doses used for WB-XTR once, most IBTRs are treated with completion mastectomy with or without reconstruction. Some patients and physicians will attempt repeat lumpectomy without additional radiotherapy but little long-term data on the safety and effectiveness of this approach are available. Recently, investigators have treated patients with IBTR using repeat lumpectomy and APBI. Here as well, isolated case reports alone exist and little long-term data on safety or effectiveness are available. Thus, generally, recurrent tumor is usually treated with salvage mastectomy.

Key Elements