CHAPTER 5 Locally Advanced Breast Cancer (LABC) and Neoadjuvant Chemotherapy

DEFINITION OF LOCALLY ADVANCED BREAST CANCER

The definition of locally advanced breast cancer (LABC) has continued to evolve since Haagensen and Stout outlined their criteria for operability more than 60 years ago.¹ Their conclusions remain a part of the current American Joint Committee on Cancer Classification System for LABC, as depicted in Table 1. LABC includes large tumors (>5 cm), those (of any size) with involvement of the skin or chest wall, and those with clinically apparent axillary nodal involvement (matted or fixed) or ipsilateral internal mammary, infraclavicular, or supraclavicular nodal disease.

Table 1 Primary Tumor (T) and Clinical Regional Lymph Node (N) Categories Comprising LABC in the Current American Joint Committee on Cancer Classification System

T3	Tumor >5 cm in greatest diameter
T4	Tumor of any size, with direct extension to chest wall or skin, as described below
T4a	Extension to chest wall, not including pectoralis muscle
T4b	Edema (including peau d’orange) or ulceration of the skin of the breast or satellite nodules confined to same breast
T4c	Both Ta and Tb
T4d	Inflammatory carcinoma
N2	Metastases in ipsilateral axillary nodes fixed or matted, or in clinically apparent ipsilateral internal mammary nodes in the absence of clinically evident axillary node metastases.
N3	Metastases in ipsilateral infraclavicular or supraclavicular lymph nodes or in clinically apparent internal mammary nodes in the presence of clinically evident axillary node metastases.

Data from Greene FL, Page DL, Fleming ID, et al. AJCC Cancer Staging Manual, 6th ed. New York, Springer Verlag, 2002.

By definition, LABC is nonmetastatic, other than to locoregional nodal basins. This group of patients needs to be carefully evaluated at initial diagnosis because some apparent LABC patients actually have systemic metastases, or stage IV disease.

ASSESSMENT OF PATIENTS SUSPECTED TO HAVE LABC

In recent years, the use of neoadjuvant chemotherapy has played an important role in reducing the number of “inoperable” patients with LABC, as well as providing access to clinical trials. This group of patients, similar to other breast cancer patients, requires an initial diagnosis and assessment of the locoregional extent of disease (typically, with mammography, ultrasound, tissue sampling, and MRI), as well as systemic staging. Systemic assessment generally includes blood studies (e.g., liver function tests, complete blood count, serum tumor markers, lactate dehydrogenase level, and alkaline phosphatase level) and may incorporate imaging modalities such as CT, nuclear medicine bone scintigraphy, MRI of the brain (in symptomatic patients or those with neurologic findings on physical examination), positron emission tomography (PET), and PET/CT. Most often, staging should be completed before initiation of neoadjuvant therapy; it is known that PET may turn negative soon after beginning chemotherapy.

Continued refinement of protocols used to treat LABC has broadened the role of imaging in both neoadjuvant and post-treatment management. Initial diagnosis is routinely accomplished with some combination of physical examination, mammography, ultrasound, and core biopsy. MRI is assuming an ever-increasing role in newly diagnosed patients. Systemic staging is typically performed with CT, bone scintigraphy, and PET. An increasingly important application of imaging in the management of LABC is the monitoring of response to neoadjuvant chemotherapy. This is discussed later with special attention to the emerging role of PET in assessing primary tumor response.

Locoregional staging of breast cancer is covered in depth in Chapter 3, including assessment with mammography, ultrasound, and MRI. The principles outlined in that chapter are identical for the patient who meets the definition of LABC. MRI has proved extremely valuable in accurately estimating tumor size and extent, as well as determining multifocality, multicentricity, and bilaterality (Figure 1). MRI more accurately correlates with tumor size and number of tumors at pathology than mammography or ultrasound. It is the modality of choice for assessment of chest wall invasion (Figure 2).

FIGURE 1 Axial maximal intensity projection (MIP) view from contrast-enhanced, subtracted breast MRI of a 52-year-old woman with newly diagnosed left LABC, a multifocal infiltrating ductal carcinoma (IDC), with involvement of 10 of 11 nodes on axillary dissection. Multiple adjacent left lateral breast masses form a large confluent mass. A satellite nodule projects medially. Several enlarged involved left axillary lymph nodes are demonstrated as well.

FIGURE 2 A, Enhanced, subtracted axial breast MRI demonstrates a neglected left LABC. This tumor would be classified as LABC based on size alone, but it demonstrates other criteria as well, including infiltration of the skin and chest wall (note enhancing, thickened skin and chest wall). B, Enhanced chest CT of the same patient, for comparison. The bulky left breast tumor masses are well seen, and extension into the thickened skin is shown as well. The bulky chest wall involvement in this case is well seen on CT, but lesser degrees of infiltration are less reliably demonstrated on CT than MRI, owing to the lower inherent soft tissue contrast of CT.

ASSESSMENT OF RESPONSE TO NEOADJUVANT CHEMOTHERAPY

Assessment of response to neoadjuvant chemotherapy with anatomic imaging modalities can be problematic. On mammography, breast cancers that initially manifest as masses generally decrease in size in response to chemotherapy, but may not completely resolve. Breast cancer manifesting on mammography as microcalcifications is even more difficult to accurately assess in terms of responsiveness to neoadjuvant therapy because microcalcifications may not resolve, even in responders. On ultrasound, responding tumors show a decrease in size and, occasionally, resolution. MRI appears to be the most reliable anatomic imaging modality in common use today for monitoring patients being treated preoperatively.³^–⁸ In addition to showing decreased tumor size, the enhancement pattern changes in responders. Responding tumors showing intense enhancement and washout initially often show decreased intensity of peak enhancement, as well as more benign patterns of progressive or persistent enhancement (flattening of the enhancement curve). Responding tumors may shrink, retaining a smaller, but still mass-like, morphology, or they can “break apart,” manifesting as less intense, smaller foci of enhancement within the distribution of the initial abnormality. Complete response by MRI (complete resolution of all tumor-associated enhancement) does not confirm a complete pathologic response because some patients with complete imaging responses have microscopic disease at pathology. Conversely, some patients with good, but incomplete, responses by MRI (some residual enhancement in the distribution of the original tumor) prove at pathology to have had complete pathologic responses.

Neoadjuvant chemotherapy is being used increasingly to convert some LABC patients with large tumors into breast conservation candidates. Because the degree of response is unknown at the outset of therapy and some patients will have complete imaging resolution of their tumors, it is important to anticipate the need for tumor marking with clip placement. This should be considered at the time of initial diagnostic biopsy of large, highly suspicious abnormalities that appear to be likely candidates for preoperative chemotherapy. Placement of a clip at the time of initial biopsy may save the patient from having to undergo a separate procedure during chemotherapy for just that purpose.

There has been great interest in monitoring the response to chemotherapy in the neoadjuvant setting using molecular imaging with the hope of separate responders from nonresponders. This allows the oncologist to continue effective therapy or change to alternate agents. Although MRI appears to show the best correlation, anatomic imaging modalities (including ultrasound, mammography, and MRI) rely predominantly on change in tumor size or volume to suggest responses. It is known that there may be a significant delay between the initiation of therapy and tumor shrinkage.

Patients may undergo several rounds of chemotherapy to find that they have not responded and need other chemotherapeutic agents. Methods such as MRI and molecular and optical imaging are being studied to determine their ability to predict tumor response earlier, based on functional and metabolic properties other than tumor size. Earlier prediction of response may allow selection of an appropriate chemotherapy regimen sooner, potentially avoiding unnecessary chemotherapy. An ACRIN trial, 6657, is currently under way to study MRI parameters, such as tumor kinetics and choline spectroscopy, to predict breast cancer response to neoadjuvant therapy.

PET and PET/CT are assuming an increasing role in this assessment, and a larger role for breast-specific gamma imaging and positron emission mammography can be anticipated in the future. Quantitative PET assessment, using the standardized uptake value (SUV), has shown early success in separating responders from nonresponders.¹⁰^–¹² Careful attention must be paid to all the factors influencing SUV measurement, especially region-of-interest assignment and patient glucose level. Flare responses (increased uptake as a manifestation of response), which can be seen on PET with initiation of tamoxifen therapy, are not seen with chemotherapy.

ASSESSMENT OF SENTINEL NODE STATUS IN LABC TREATED NEOADJUVANTLY

A controversial issue in the use of neoadjuvant chemotherapy is whether to assess preoperatively the status of the axilla with sentinel lymph node (SLN) sampling. It has been demonstrated that information from SLN biopsy or axillary dissection performed after completion of chemotherapy correlates well with overall prognosis. The role of SLN sampling and axillary dissection in the postchemotherapy setting was reviewed in a meta-analysis, with an overall accuracy of 95%.¹³

In an attempt to reduce false-negative SLN analysis after neoadjuvant chemotherapy, some surgeons advocate routine SLN sampling before initiation of chemotherapy. Positive SLN biopsy would lead to completion axillary dissection, before or after chemotherapy. Pathology information from prechemotherapy SLN biopsy is used by radiation oncologists in planning radiation field coverage.

ASSESSMENT FOR SYSTEMIC METASTASES (EXCLUSION OF STAGE IV DISEASE)

The most common sites of breast cancer metastasis are bone, lung, and liver, in that order.¹⁴ Evaluation of bone should begin with a careful history to elicit any symptoms or history of recent trauma. Serum alkaline phosphatase and calcium measurements may be helpful if positive and heighten suspicion of bony involvement. Imaging options, discussed in greater depth in Chapter 8, include bone scanning, CT, MRI, and PET. Bone scintigraphy, using a technetium-based agent (e.g., methylene diphosphonate), is exquisitely sensitive to changes in bone metabolism. Localization of these agents into bone is dependent on many factors, with the most important two being blood flow and osteoblastic activity. This results in a very predictable concentration in osteoblastic metastases (Figure 3). Positive findings on scintigraphy antedate findings on plain radiography by weeks to months. Drawbacks of bone scanning include suboptimal specificity, reduced sensitivity in osteolytic metastases, and persistent positivity at sites where active tumor is no longer present. Additionally, the well-recognized flare phenomenon, resulting in apparent worsening on scintigraphy due to treatment response, may mislead the unaware clinician or imager. Specificity is arguably the most problematic feature of whole-body skeletal scintigraphy. Any process that results in an increase in bone remodeling will demonstrate increased uptake of radiopharmaceutical. For this reason and because of the implications of labeling a patient as having bony metastases, correlative imaging (or even tissue sampling) is required. Correlation can be accomplished with plain radiography, CT, or MRI.

FIGURE 3 Anterior and posterior images from whole-body bone scan in a patient with known breast cancer. Typical bone scan findings of extensive breast cancer bone metastases are demonstrated: multiple lesions in a random distribution, with several rib lesions displaying a particularly suggestive pattern of elongated uptake.

Plain radiography is of minimal value as a screening modality, requiring 30% to 50% loss of bone mineral for a metastasis to become visible.¹⁵ Plain radiographic correlation with a scintigraphic abnormality may be of benefit; however, the increased sensitivity and improved anatomic detail (including surrounding soft tissues) are factors favoring CT (Figure 4) or MRI. Although MRI has a higher rate of detection of skeletal metastases than scintigraphy in the spine, pelvis, limbs, sternum, scapulae, and clavicles,¹⁶ the logistics and cost of whole-body MRI give scintigraphy the edge as an initial screening examination.

FIGURE 4 Chest CT image, displayed with bone windowing, from the same patient in Figure 3, shows abnormal, mottled, partially blastic bone mineral texture involving a long segment of a right posterior rib, corresponding to the bone scan.

PET and, more recently, PET/CT have been used to evaluate the entire body for metastases, including bone. An emerging consensus is that PET and scintigraphy have a similar sensitivity for detection of metastases, whereas PET shows a definite in-crease in specificity.¹⁷^–²⁰ There also appears to be a significantly higher sensitivity with fluorodeoxyglucose (FDG) PET for osteolytic metastases. Conversely, bone scintigraphy has shown superiority for demonstration of osteoblastic lesions. Currently, PET and bone scintigraphy are viewed as complementary imaging modalities for the detection of skeletal metastases.

Lung metastases are relatively common in patients with metastatic disease and in those who die from breast cancer. However, lung metastases are very uncommon at initial diagnosis of breast cancer.²¹ Many centers still recommend a chest x-ray at initial screening (in part, because of the age of their breast cancer population); however, positive findings on chest x-ray generally necessitate CT correlation. CT is the modality of choice for chest evaluation. PET/CT offers advantages over CT alone in evaluating the mediastinum and as a whole-body survey.²²

Evaluation of the liver is covered in depth in Chapter 9. In addition to CT, MRI, and PET, ultrasound may be appropriate in selected patients. The number of patients with hepatic metastases at initial presentation is extremely low. Screening, whether using liver enzymes or imaging, is nonspecific and of low diagnostic yield. For symptomatic patients and those with clinical evidence of liver involvement, CT and MRI are considered the imaging modalities of choice.²³ Ultrasound may be useful to help characterize small lesions identified on CT.²⁴ Finally, FDG PET is best viewed as complementary in the liver; it carries an excellent specificity and will occasionally find lesions not appreciated prospectively on CT or MRI, but it has limited sensitivity for small lesions (<1 cm) and can be difficult to interpret when there is heterogeneous FDG uptake, which can be exacerbated by attenuation correction (Table 2).

Table 2 Role of Imaging Modalities in Diagnosis (D), Staging (S), Restaging (R), and Monitoring Response to Therapy (M)

REFERENCES

1 Haagensen C, Stout A. Carcinoma of the breast: Criteria of operability. Ann Surg. 1943;118:859-868.

2 Greene FL, Page DL, Fleming ID, et al. AJCC Cancer Staging Manual, 6th ed. New York: Springer Verlag, 2002.

3 Rieber A, Brambs HJ, Gabelmann A, et al. Breast MRI for monitoring response of primary breast cancer to neo-adjuvant chemotherapy. Eur Radiol. 2002;12(7):1711-1719.

4 Partridge SC, Gibbs JE, Lu Y, et al. Accuracy of MR imaging for revealing residual breast cancer in patients who have undergone neoadjuvant chemotherapy. AJR Am J Roentgenol. 2002;179:1193-1199.

5 Rosen EL, Blackwell KL, Baker JA, et al. Accuracy of MRI in the detection of residual breast cancer after neoadjuvant chemotherapy. AJR Am J Roentgenol. 2003;181:1275-1282.

6 Londero V, Bazzocchi M, Del Frate C, et al. Locally advanced breast cancer: comparison of mammography, sonography and MR imaging in evaluation of residual disease in women receiving neoadjuvant chemotherapy. Eur Radiol. 2004;14(8):1371-1379.

7 Martincich L, Montemurro F, De Rosa G, et al. Monitoring response to primary chemotherapy in breast cancer using dynamic contrast-enhanced magnetic resonance imaging. Breast Cancer Res Treat. 2004;83(1):67-76.

8 Yeh E, Slanetz P, Kopans DB, et al. Prospective comparison of mammography, sonography, and MRI in patients undergoing neoadjuvant chemotherapy for palpable breast cancer. AJR Am J Roentgenol. 2005;184(3):868-877.

9 Bassa P, Kim EE, Inoue T, et al. Evaluation of preoperative chemotherapy using PET with fluorine-18-fluorodeoxyglucose in breast cancer. J Nucl Med. 1996;37(6):931-938.

10 Schelling M, Avril N, Nahrig J, et al. Positron emission tomography using [(18)F]-fluorodeoxyglucose for monitoring primary chemotherapy in breast cancer. J Clin Oncol. 2000;18:1689-1695.

11 Biersack HJ, Palmedo H. Locally advanced breast cancer: is PET useful for monitoring primary chemotherapy? J Nucl Med. 2003;44(11):1815-1817.

12 Rosen EL, Eubank WB, Mankoff DA. FDG PET, PET/CT, and breast cancer imaging. RadioGraphics. 2007;27:S215-S229.

13 Xing Y, Ding M, Ross M, et al. Meta-analysis of sentinel lymph node biopsy following preoperative chemotherapy in patients with operable breast cancer. ASCO Annual Meeting, 2004, New Orleans, LA abstract 561.

14 Huston TL, Osborne MP. Evaluating and staging the patient with breast cancer. In: Ross D, editor. Breast Cancer. 2nd ed. Philadelphia: Elsevier Churchill Livingstone; 2005:309-318.

15 Schirrmeister H. Detection of bone metastases in breast cancer by positron emission tomography. PET Clin. 2006;1(1):25-32.

16 Chom Y, Chan K, Lam W, et al. Comparison of whole body MRI and radioisotope bone scintigram for skeletal metastases detection. Chin Med J (Eng1). 1997;110(6):485-489.

17 Cook GJ, Houston S, Rubens R, et al. Detection of bone metastases in breast cancer by 18 FDG PET: differing metabolic activity in osteoblastic and osteolytic lesions. J Clin Oncol. 1998;16(10):375-379.

18 Ohta M, Tokuda Y, Suzuki Y, et al. Whole body PET for the evaluation of bony metastases in patients with breast cancer: comparison with 99 m Tc-MDP bone scintigraphy. Nucl Med Commun. 2001;22(8):875-879.

19 Yang SN, Liang JA, Lin FJ, et al. Comparing whole body ¹⁸F-2 deoxyglucose positron emission tomography and technetium-99 m methylene diphosphonate bone scan to detect bone metastases in patients with breast cancer. J Cancer Res Clin Oncol. 2002;128(6):325-328.

20 Uematsu T, Yuen S, Yukisawa S, et al. Comparison of FDG PET and SPECT for detection of bone metastases in breast cancer. AJR Am J Roentgenol. 2005;184(4):1266-1273.

21 Ciatto S, Pacini P, Azzini V, et al. Preoperative staging of primary breast cancer: a multicentric study. Cancer. 1988;61(5):1038-1040.

22 Dose J, Bleckmann C, Bachmann S, et al. Comparison of fluorodeoxyglucose positron emission tomography and “conventional diagnostic procedures” for the detection of distant metastases in breast cancer patients. Nucl Med Commun. 2002;23(9):857-864.

23 Reinig JW, Dwyer AJ, Miller DL, et al. Liver metastasis detection: comparative sensitivities of MR imaging and CT scanning. Radiology. 1987;162:43-47.

24 Eberhardt SC, Choi PH, Bach AM, et al. Utility of sonography for small hepatic lesions found on computed tomography in patients with cancer. J Ultrasound Med. 2003;22(4):335-343.

CASE 1 LABC (large tumor size)

A 45-year-old woman was noted on routine physical examination by her physician to have an abnormal left breast examination. Although the patient had not noted any discrete masses, an area of induration and irregularity was palpated in the left upper breast, extending from upper inner to upper outer quadrant.

Breast imaging evaluation showed extremely dense breast parenchyma. New clustered microcalcifications were identified in the upper inner quadrant (UIQ). Sonography showed irregular, hypoechoic parenchymal echotexture in the left upper breast at 11 to 12 o’clock with cysts and shadowing (Figure 1), and the region of the microcalcifications was visualized as well. Ultrasound-guided biopsy confirmed infiltrating ductal carcinoma (IDC) with ductal carcinoma in situ (DCIS).

FIGURE 1 A and B, Representative ultrasound images of the 11 to 12 o’clock region show dense shadowing, with irregular anterior margination. The extent of the process is difficult to define and to differentiate from extreme fibrocystic changes.

The patient was evaluated by medical and radiation oncology and surgery for possible breast conservation. All three examiners had difficulty quantifying the size of the palpable abnormality, partly because of the presence of large coexisting cysts. One examiner noted a large area of firm dense tissue extending over 5 to 6 cm at the 12-o’clock level, and was concerned about the true extent of disease, because mammography had shown only a small UIQ microcalcification cluster. A breast MRI was, therefore, ordered to assess the extent of disease.

Breast MRI showed striking asymmetry between the two sides, with findings of a very large, diffusely infiltrating cancer on the left (Figures 2 and 3). The intensely enhancing process, centered at 12 o’clock but spanning 11 to 2 o’clock, showed architectural distortion, with additional extensive clumped enhancement diffusely involving the medial breast. By MRI, the abnormality measured at least 6 cm and involved at least two quadrants extensively, indicating that the patient was not a conservation candidate.

FIGURE 2 Axial enhanced, maximal intensity projection MRI shows extreme asymmetry between sides. The right side shows diffuse low-level, extensive, scattered, tiny fibrocystic foci of enhancement. The left side is dominated by a central mass with architectural distortion, and there is geographic enhancement in the medial breast.

FIGURE 3 Representative axial, subtracted, enhanced images, from above to below. A, Clumped enhancement medially with larger central tumor nodules involves both the upper inner and upper outer quadrants at this level. B, There is intense central enhancement with distortion and segmental clumped medial enhancement at this level. C, The spiculation and distortion of the central enhancing component are well demonstrated on this section. The oval defect within the enhancing spiculated mass represents a cyst, which was surrounded by the enhancing tumor. Clumped segmental medial enhancement, suggesting DCIS, is again seen. D, At the level of the nipple, extensive clumped segmental enhancement is seen medially and centrally, suggesting an extensive intraductal component, which was subsequently confirmed. E, Sagittal, subtracted, enhanced breast MRI shows how extensively the entire upper half of the breast is infiltrated.

Unfortunately, the surgeon did not become aware of this result until the day of the surgery because the study had been ordered by another physician. Fortunately, the radiologist who scanned the patient for performance of ultrasound-guided needle localization on the day of surgery recognized that the surgeon could not have been aware of the results of the MRI. The patient was subsequently apprised of the results and the extremely high likelihood of positive margins if lumpectomy was attempted, and therefore underwent mastectomy.

The mastectomy specimen showed an 8-cm IDC of the central breast extending into the upper inner and lower inner quadrants, with a high-grade DCIS component involving 50% of the lesion (extensive intraductal component). Multifocal satellite tumor nodules were noted within the large tumor. Extensive angiolymphatic involvement was seen, and four sentinel lymph nodes showed metastatic carcinoma (Figure 4). Completion axillary dissection showed no additional axillary disease, for a total of 4 of 20 lymph nodes positive for metastatic disease. The deep mastectomy margin was negative.

FIGURE 4 Coronal short tau inversion recovery (STIR) image of the thorax (obtained with the body coil at the time of breast MRI) shows three adjacent mildly prominent left axillary lymph nodes. Although not prospectively interpreted as suspicious because no one lymph node is particularly large, in retrospect the asymmetry in number from the opposite side is a clue to the axillary involvement.

The patient underwent imaging staging postoperatively, with bone scan, CT, and positron emission tomography (PET) (Figures 5 and 6), which showed only postsurgical changes of the axilla and chest wall and no evidence of distant metastatic disease. Final stage was stage IIIA, with T3N1M0 disease, and the tumor was estrogen receptor and progesterone receptor positive.

FIGURE 5 PET scan, obtained 2 weeks after left mastectomy, shows a left chest wall photopenic collection with a metabolically active rim, consistent with a postoperative seroma.

FIGURE 6 Enhanced chest CT image for correlation with PET shows a left chest wall seroma after mastectomy.

TEACHING POINTS

The extensive nature of this locally advanced, very large IDC became apparent relatively late in this patient’s evaluation and was accurately delineated only by MRI. The clinical examination in this patient confounded multiple examiners. The patient had extremely dense, fibrocystic breasts, with multiple cysts, and examiners found it difficult to delineate the extent of the tumor by palpation. Fortunately, one examiner sensed a discrepancy between the imaging findings and the apparent clinical extent, and breast MRI was ordered. It is not surprising that mammography in a patient with very dense breast tissue and multiple cysts would underestimate the extent of disease. Much of this patient’s extensive intraductal disease was noncalcified, with only a small microcalcification cluster noted in the UIQ at mammographic evaluation. The sonogram identified abnormal echotexture in the involved area, accurately guiding core biopsy for a diagnosis, but the size of the process and the full significance of the sonographic findings, with geographic shadowing and poor margin delineation, were not fully appreciated prospectively.

Accurate and timely communication between specialties is a critical part of the care of the newly diagnosed breast cancer patient, as this case illustrates. Evaluations up to the point of the breast MRI had not shown clear evidence that the patient was not a lumpectomy candidate, and so the treatment planning was proceeding with this expectation. One examiner’s concern that there might be a discrepancy between the imaging findings and the clinical examination led to performance of breast MRI, which confirmed a very large, locally advanced tumor. Unfortunately, this was ordered at the last minute and without the knowledge of the surgeon. Fortunately, the radiologist charged with performing the needle localization on the day of surgery was able to rectify the situation, which was precipitously, but satisfactorily, resolved in favor of the patient undergoing mastectomy.

CASE 2 LABC with axillary and internal mammary involvement; staging with whole-body PET and PEM

A 77-year-old woman was noted on routine physical examination by her primary care physician to have a palpable 3-cm mass on the right at 6 o’clock. Breast imaging evaluations confirmed this mass, which was best seen on ultrasound as a 3.2-cm mass with lobular and irregular margins (Figure 1). On ultrasound, a second suspicious mass, which was not seen mammographically, was noted medial to the dominant mass, at 4 o’clock (Figures 2 and 3). This was a bilobed, hypoechoic mass. A highly suspicious axillary lymph node was also found, measuring 2.3 cm, with complete effacement of the fatty hilus (Figure 4).

FIGURE 1 Ultrasound of the palpable lump on the right at 6 o’clock shows a very hypoechoic, heterogeneous, vascular mass with lobular and irregular margins, with a highly suspicious sonographic appearance. Biopsy confirmed IDC.

FIGURE 2 Ultrasound of the right 4-o’clock level shows a hypoechoic, bilobed, more smoothly marginated mass, resembling an abnormal lymph node in morphology. Biopsy with ultrasound guidance confirmed IDC.

FIGURE 3 Ultrasound shows the relationship between the 4- and 6-o’clock masses, which are 3.5 cm apart by ultrasound. A rounded, sub-centimeter nodule interposed between the two masses (arrow) was not noted prospectively. It corresponds to a smaller satellite lesion, which was FDG avid and seen on both whole-body and dedicated breast PET (PEM).

FIGURE 4 Ultrasound of the right axilla shows a round, completely replaced enlarged axillary lymph node. The fatty hilus is completely effaced.

These three abnormalities were biopsied with ultrasound guidance. Both the 4- and 6-o’clock breast masses were poorly differentiated infiltrating ductal carcinomas (IDC), estrogen receptor and progesterone receptor negative, HER-2/neu negative, grade 8/9. The axillary lymph node fine-needle aspiration showed adenocarcinoma, consistent with metastatic breast carcinoma.

Staging evaluations were obtained because of the patient’s node-positive status, concurrent complaints of back pain, and concern for possible metastatic disease. The patient had a pacemaker, which limited the options for performing MRI. A bone scan and positron emission tomography (PET)/CT scan were obtained. In addition, a positron emission mammography (PEM) scan was obtained of the breasts, immediately after the PET/CT, using the same dose of fluorodeoxyglucose (FDG).

Bone scan suggested an L1 compression fracture as the etiology of back pain (Figure 5). A band of modest increased metabolic activity was seen at the superior end plate of L1 on PET, and CT showed superior end-plate invagination, confirming the bone scan impression of a compression fracture (Figure 6). Three right lower inner quadrant FDG-avid breast masses were seen on whole-body PET/CT, with a small additional FDG-avid nodule seen between the two known IDC masses at 4 and 6 o’clock. The known involved right axillary lymph node was intensely hypermetabolic and was accompanied by several smaller additional metabolically active axillary lymph nodes. Activity was also seen in two internal mammary lymph nodes, which on CT measured 8 mm (Figures 7 and 8).

FIGURE 5 Bone scan images show a band of increased activity at the L1 level. This pattern of activity suggests a compression fracture as the etiology of the patient’s back pain.

FIGURE 6 Coronal CT reconstruction of the lumbar spine shows end-plate invagination of the L1 vertebral body, correlating with the bone scan and consistent with a compression fracture (arrows). Modest PET scan activity was present at this level. Other lumbar levels show discogenic changes and a levoscoliotic curvature.

FIGURE 7 Enhanced chest CT images, from above to below. A, An oval, markedly enlarged axillary lymph node is seen on the right, corresponding to the ultrasound. Scatter artifact from a right chest wall pacemaker is noted. B, A rounded, enlarged right internal mammary lymph node is seen adjacent to the enhanced internal mammary artery (arrow). This was hypermetabolic on PET. The vein is seen over the lateral right sternum. C, A second right internal mammary (IM) lymph node is seen adjacent to the artery and vein (from right to left: IM artery, IM vein, IM lymph node) (arrow). This was also hypermetabolic on PET. D, Through the inferior breasts, two right breast masses are seen. The medial one is the same as the 4-o’clock mass on ultrasound. It has a lobulated contour. No fat plane is seen between it and the muscle. A round smaller mass is seen lateral to it. E, Further inferior, the dominant right 6-o’clock irregularly marginated mass is seen, as well as the inferior aspect of the 4-o’clock mass.

FIGURE 8 PET scan images show three multifocal, FDG-avid right lower inner quadrant breast masses, as well as FDG-avid axillary lymph nodes and internal mammary nodes (not shown). A, Cursors are positioned at the level of the 6-o’clock IDC. A small adjacent FDG-avid nodule is seen on the coronal slice. B, Cursors are positioned at the level of the 4-o’clock IDC. The small satellite mass is partially seen on the axial image.

A PEM study was also obtained in this patient after completion of PET/CT, using the same FDG dose. The proven multifocality of her tumor would ordinarily have been an indication for breast MRI, but the patient could not readily undergo breast MRI because of her pacemaker. The dominant mass on the right at 6 o’clock was seen with exquisite detail on PEM, with intense heterogeneous uptake of FDG (Figure 9). The 4-o’clock mass was not seen on either view. This result was anticipated because of its far posterior position on the chest wall on CT. The small, intermediately positioned satellite tumor mass was visualized on the mediolateral oblique (MLO) projection, which shows more posterior breast tissue. The PEM also permitted more thorough screening of the left breast.

FIGURE 9 PEM scan images [MLO (A) and CC (B, next page)] show heterogeneous FDG uptake within the 6-o’clock IDC. The small, adjacent FDG-avid nodule is seen posterior to the 6-o’clock IDC only on the MLO view, which shows more posterior tissue.

Modified radical mastectomy and axillary lymph node dissection were performed. At surgery, abutment and adherence of tumor focally to the pectoralis muscle was noted, requiring some excision of muscle.

Pathology showed three IDC tumor masses in the right lower inner quadrant, measuring 4.5 cm each at 4 and 6 o’clock and 1 cm in between. There was angiolymphatic invasion, with the 4-o’clock IDC close (0.09 mm) to the deep margin. Metastatic carcinoma was identified in 7 of 12 axillary lymph nodes, with extreme matting noted by pathology.

TEACHING POINTS

This is a locally advanced breast cancer, based on axillary and internal mammary lymph node involvement. This degree of nodal involvement is classified as N3b, making the patient stage IIIC. The internal mammary lymph node involvement does not have to be histologically proven to be considered positive for involvement. In this case, the ease of visualizing these enlarged internal mammary lymph nodes on CT and the FDG avidity seen on PET leave little doubt that they are involved.

This patient’s concurrent back pain raised the specter of metastatic disease. Fortunately, the bone scan abnormality is characteristic of a compression fracture and the PET and CT were entirely consistent. Of course, this does not entirely exclude the possibility of a pathologic compression fracture due to a bone metastasis, but without other evidence of osseous metastases, this seems unlikely. An osteoporotic compression fracture is a commonly encountered benign cause of symptoms and bone scan findings in older women.

It is interesting to note the node-like morphology of the 4-o’clock lesion in this case. It strongly resembles an abnormal lymph node and was positioned on the chest wall, suggesting it may be a completely replaced external mammary lymph node.

Normally, with proven multifocal breast cancer, local staging would often be supplemented with breast MRI, which would effectively screen the other breast. In this case, a pacemaker precluded performance of breast MRI. PEM was utilized to screen the other breast. It is interesting to see some of the strengths and weaknesses of PEM displayed by this case. As in mammography, only tissue that is imaged can be evaluated. Far posterior lesions that cannot be positioned between the detector arrays will not be seen, even sizable ones like the 4-o’clock lesion in this case. Design improvements of future PEM devices may improve detectability of posterior lesions. On the other hand, the detail of the heterogeneous FDG uptake of the 6-o’clock IDC is striking in contrast to the whole-body PET depiction.

The role of PEM in the breast imaging armamentarium is currently being assessed by a prospective, multi-institutional clinical trial comparing the performance of PEM to breast MRI in local staging of newly diagnosed breast cancers. Essentially, PEM is a small-field-of-view, high-resolution PET scanner, with in plane resolution on the order of 2 mm. Resembling a mammogram unit, the compression “plates” consist of an array of detectors. Compression is applied to immobilize the breast, but not to the same degree as in mammography. As PEM is a tomographic technique, there is no need to thin the breast to the same degree as in mammography.

CASE 3 LABC with nipple skin involvement

A 56-year-old woman noted new right nipple inversion 4 months before presenting for breast imaging evaluation for a newly developed right palpable axillary lump. Mammography and ultrasound showed a dominant 12-o’clock mass with nipple retraction and localized skin thickening, as well as axillary lymphadenopathy (Figures 1, 2, and 3). Ultrasound-guided biopsy of the dominant mass confirmed invasive ductal carcinoma (IDC), and ultrasound-guided axillary node fine-needle aspiration confirmed metastatic disease. Periareolar dusky erythema was noted, and skin punch biopsy was performed, which was negative. Local staging was completed with breast MRI (Figures 4, 5, 6, 7, and 8).

FIGURE 1 Ultrasound of the right upper outer quadrant shows a hypoechoic, taller-than-wide, sonographically suspicious mass with irregular and angular margins.

FIGURE 2 Ultrasound of the retracted right nipple shows unusual increased periareolar vascularity.

FIGURE 3 Ultrasound of the right axilla shows two adjacent, highly abnormal lymph nodes: both are rounded, with very hypoechoic, thickened cortices. Nodular mass effect is exerted on the echogenic fatty hila, which are partially effaced. No increased vascularity is identified.

FIGURE 4 Maximal intensity projection of enhanced, subtracted, axial gradient echo data set shows an intensely enhancing, dominant mass in the central right breast. Numerous linear spicules extend anteriorly toward the nipple. There are enlarged, draining veins on the right, and enlarged, enhancing right axillary lymph nodes.

FIGURE 5 Axial breast MR images through the dominant right IDC (A, enhanced, subtracted, T1-weighted gradient echo; B, STIR) show rim enhancement and spiculation of the mass. Linear spicules extend anteriorly toward the nipple and posteriorly toward the pectoral muscle. The periareolar skin is thickened and edematous, and enhances. A rounded, low-lying axillary or intramammary lymph node enhances intensely in the posterolateral right breast.

FIGURE 6 Sagittal, subtracted, contrast enhanced MRI of the right breast showing the enhancing linear tendrils extending to the nipple, as well as the periareolar skin enhancement.

FIGURE 7 Kinetic information (A, DynaCAD color map; B, enhancement curve) showing washout from portions of the IDC mass.

FIGURE 8 Axial STIR images of the axilla. A, An enlarged right axillary level I lymph node is seen lateral to the pectoralis major and minor muscles (arrow). B, An enlarged interpectoral (Rotter’s) lymph node is seen on the right, between the pectoralis major muscle in front and pectoralis minor muscle behind (arrow).

Systemic staging with positron emission tomography (PET)/CT showed fluorodeoxyglucose (FDG) avidity of the known right breast cancer and axillary lymphadenopathy, but no additional disease. Neoadjuvant chemotherapy was given, consisting of four cycles each of doxorubicin (Adriamycin) and cyclophosphamide (Cytoxan) (AC) and paclitaxel (Taxol). The nipple inversion resolved, and there was marked clinical regression of the dominant mass. Modified radical mastectomy and axillary dissection were performed. A 4.5-cm residual lesion of admixed IDC and normal breast tissue remained. Margins were negative, and 5 of 13 lymph nodes showed metastatic tumor. Chest wall, supraclavicular, and posterior axillary boost radiation therapy were given, and the patient was placed on anastrozole (Arimidex).

TEACHING POINTS

In this case, the localized periareolar skin thickening, edema, and enhancement probably represented secondary inflammation and involvement by a locally advanced breast cancer (LABC), rather than being a manifestation of primary inflammatory breast cancer.

CASE 4 Natural history of untreated inflammatory breast cancer

A 57-year-old woman noted her right breast to be enlarged and heavier feeling. Mammography showed concerning right upper outer quadrant (UOQ) microcalcifications (Figures 1 and 2). Right breast sonogram showed periareolar skin thickening (Figure 3) and an UOQ 1.5-cm shadowing hypoechoic mass with irregular margins (Figure 4), as well as hypoechoic, rounded, axillary lymph nodes (Figures 5 and 6). Ultrasound-guided core needle biopsy of the UOQ mass confirmed infiltrating ductal carcinoma (IDC), estrogen receptor and progesterone receptor negative, HER-2/neu negative. Ultrasound-guided fine-needle aspiration (FNA) of an axillary lymph node confirmed metastatic carcinoma, consistent with breast primary origin. Two skin punch biopsies were performed because of clinical suspicion of inflammatory carcinoma: both were negative. Skin biopsies were subsequently repeated because of persistent high suspicion of inflammatory cancer, and confirmed intralymphatic carcinoma.

FIGURE 1 MLO (A) and CC (B) mammograms show heterogeneous breast parenchymal density. No parenchymal changes were detected compared with the prior study, 4 years before. An area containing concerning microcalcifications is circled in the UOQ.

FIGURE 2 Lateral magnification view of the right upper breast shows concerning new microcalcifications.

FIGURE 3 Sonographic image of the right periareolar skin is thickened to nearly 4 mm. An edematous appearance of the subjacent tissues is seen, without a discrete mass.

FIGURE 4 A very hypoechoic, highly suspicious mass with angular margins was found in the right UOQ. Ultrasound-guided biopsy confirmed IDC.

FIGURE 5 Right axillary ultrasound shows two oval, adjacent, very hypoechoic lymph nodes, with no identifiable fatty hila.

FIGURE 6 Another right axillary lymph node shows cortical mantle thickening, effaced hilus, and abnormally increased vascularity. FNA confirmed metastatic disease.

Breast MRI showed multiple intensely enhancing right breast masses, as well as enhancement of the skin (Figures 7, 8, 9, 10, and 11). Systemic staging consisted of positron emission tomography (PET)/CT and enhanced body CT scans. Hypermetabolism was identified in the right breast and axillary lymph nodes, but no evidence of distant metastatic disease was found (Figures 12, 13, 14, 15).