CHAPTER 10 Thoracic Metastases, Mimics, and Treatment Effects

Most breast cancer patients with intrathoracic metastases are asymptomatic, the exception being those with significant pleural effusions. Therefore, unless pleural, parenchymal, or nodal metastases are demonstrated at initial staging, thoracic metastases are most commonly found at the time of follow-up imaging. A significant number are found on chest x-rays done for other reasons. Isolated involvement of the lung or pleural space occurs in 15% to 25% of women with metastatic breast cancer.¹

Pulmonary metastases are typically peripheral,² and, if large enough, may be discovered on preoperative chest radiographs. Because of its low cost, chest radiography may be a reasonable choice as a baseline study or in follow-up; however, the extremely low yield, false-positive results, inferior sensitivity compared with CT, and lack of data supporting an outcome benefit limit routine use in follow-up.

CT is the preferred modality for investigation of thoracic metastases. CT has proven efficacy for detecting local, regional, and distant disease. Normal-sized internal mammary nodes are not visualized on CT, so a patient in whom CT identifies a node or nodes larger than 6 mm is at risk for malignant involvement.³ Because of the recognized problem of relying on size only to diagnose disease, positron emission tomography (PET)/CT is presently considered the modality of choice for assessing internal mammary nodes (Figure 1). In women who die from disseminated breast cancer, intrathoracic nodal metastases may occur in more than 70% of patients.⁴ As with internal mammary nodes, using size as the major criterion for disease identification reduces both sensitivity and specificity. Magnetic resonance imaging (MRI) can detect most lesions larger than 5 mm, does not expose the patient to ionizing radiation, and has acceptable imaging times; however, it is generally viewed as complementary to CT, rather than as a replacement for CT (Figure 2).⁵^–⁹ Cardiac and respiratory motion, reduced sensitivity for lesions smaller than 5 mm (compared with CT), calcified metastases, and suboptimal ability to assess lymphangitic spread are recognized limitations of MRI in the thorax.¹⁰

FIGURE 1 PET/CT images from an initial staging study of a 77-year-old woman with newly identified locally advanced infiltrating ductal carcinoma (IDC). Internal mammary nodal involvement is seen on the right. Normal internal mammary lymph nodes are not usually visualized on chest CT. This enlarged node, measuring 8 mm, is readily seen on CT and is hypermetabolic on PET. Although not biopsy proven, these imaging features are strong presumptive evidence of internal mammary nodal involvement. Seven of 12 axillary nodes were involved with tumor at surgery.

FIGURE 2 A, Breast MRI of a neglected, locally advanced left breast cancer with chest wall, skin, and nipple involvement shows a lung nodule on the right, suggesting possible stage IV disease (arrow). B, Chest CT confirmed a lung nodule, with an additional tiny subpleural nodule seen on this slice as well (arrowhead).

PET using fluorodeoxyglucose (FDG), an analogue of glucose, has assumed a primary role as a whole-body survey examination in many types of cancer, including breast. PET, and more recently PET/CT, excels at identifying the metabolic status of intrathoracic nodes, including internal mammary nodes, and solitary pulmonary nodules. Metabolic changes, resulting in pathologic or increased uptake on PET scan, can occur months before changes in size or shape. Additionally, it is now well recognized that lymph nodes larger than 10 mm may be benign or reactive, and lymph nodes smaller than 10 mm may represent metastases. With one important caveat, PET does not rely on the size of a lymph node or nodule to characterize it as benign or malignant. The caveat relates to the sensitivity of the PET/CT system based on lesion size. Presently, most scanners in clinical practice are able to reliably identify hypermetabolic foci down to about 8 mm. The addition of CT capability in PET/CT combination units will in many cases further improve sensitivity by allowing relatively precise image registration. The ability to confirm hypermetabolic activity in very small lesions may complicate reliance on the standardized uptake value (SUV). This calculated value is a measure of radioactivity in a lesion divided by the average value of radioactivity in the body, corrected for body weight. Many centers use a cut-off value of 2 to 2.5 or higher to indicate malignancy, with values less than 2 as evidence for benignity. The SUV is dependent on a multitude of factors, including serum glucose at the time of FDG injection, time between injection and imaging, and lesion size. Many investigators believe that any uptake above background in small (i.e., 5 to 7 mm) lesions should be regarded as suspicious for malignancy.

In one series of 73 patients who underwent FDG PET and CT for recurrent or metastatic disease, PET proved more sensitive (85% versus 50%) and specific (90% versus 83%) than CT. Also, PET uptake in mediastinal and internal mammary nodes was 2 times more prevalent than enlarged nodes on CT.¹¹ FDG PET has been shown to be significantly more accurate in staging the mediastinum in patients with non–small cell lung cancer, using histology as the gold standard.^12–¹³ Extrapolation of these data to patients with breast cancer may be problematic. It has been recognized that certain subtypes of breast cancer (e.g., invasive lobular, tubular, carcinoma in situ) show relatively low FDG uptake, resulting in reduced sensitivity compared with infiltrating ductal carcinoma (IDC) primary tumors.^14–¹⁵ Theoretically, patients whose initial staging FDG PET is negative in these tumor subtypes (especially if the primary has not been excised) may demonstrate reduced sensitivity for detection of metastases, including within the chest, on restaging studies.

Pleural involvement may be suspected based on clinical signs on physical examination or symptoms (e.g., cough, shortness of breath, pleuritic-type chest pain, orthopnea), the discovery of a pleural effusion on plain radiography or CT, or the identification of pleural abnormalities on CT or FDG PET (Figure 3). Pleural effusion is often unilateral and on the same side as the breast primary.¹⁶ The effusion can be confirmed by chest radiography, CT, or even ultrasound. Thoracentesis may be performed for both diagnostic and therapeutic purposes. If the result is nondiagnostic, or negative with a strong clinical suspicion of malignancy, consideration may be given to FDG PET imaging (preferably, PET/CT).

FIGURE 3 PET/CT axial images of a breast cancer patient with bilateral pleural metastases, more extensive on the left than the right. The pleural spaces extend far inferiorly, so that hypermetabolic pleural metastases appear to project in the periphery of the abdomen, overlapping the liver and spleen. The coronal projection image also shows a large, hypermetabolic left axillary lymph node.

FDG PET commonly identifies more pleural involvement than other imaging modalities and may be especially helpful in delineating the caudal extent of disease. The effusion itself tends to be only mildly positive on PET or frankly negative, presumably owing to the paucity of malignant cells relative to the volume of the effusion.

Many special circumstances exist with respect to the thoracic manifestations of breast cancer, reviewed in depth by Jung and associates.¹⁷ One important consideration is radiation therapy–related complications. Radiation pneumonitis typically occurs within 3 months of radiation therapy and progresses from diffuse haziness to coalescence of areas of consolidation to fibrous changes on chest x-ray or CT. Findings in the irradiated field may persist for months to years (Figure 4). Although PET/CT has demonstrated efficacy in separating scar from active tumor in many malignancies, its use following radiation therapy in the irradiated field may be problematic for months to 1 year or more. The heightened metabolic activity of activated monocytes at sites of radiation pneumonitis can last for a variable period of time, especially in the lung and with head and neck tumors. A negative FDG PET study at a site of postirradiation change on chest x-ray or CT is reassuring; however, a positive study must be interpreted with caution and reassessed or further investigated before initiation of additional tumor-targeted therapy.

FIGURE 4 Axial lung window chest CT image from a 56-year-old woman, 4 years after treatment for a high-risk IDC involving 19 of 26 lymph nodes with lumpectomy, axillary dissection, systemic chemotherapy, and radiation to the breast, axillary, and supraclavicular nodal basins, shows asymmetrical right apical radiation fibrosis. This was stable on imaging, corresponded to the radiation field, and showed stable, low-level activity on FDG PET.

The development of a solitary pulmonary nodule in patients who have been treated for breast cancer should not be assumed to be a breast recurrence without histologic proof. About half of these patients will have lung cancer, and a small percentage of cases will be benign.¹⁸ These patients are excellent candidates for PET/CT. Hypermetabolic uptake in the nodule directs the patient to tissue sampling (sometimes obviated by the finding of widespread metastases). Absent FDG uptake, as well as an otherwise negative study, may allow watchful waiting with short-term CT follow-up (e.g., 3 to 6 months). Recall the limitations of PET with respect to histology and lesion size.

Lymphangitic metastases are extremely common in women who die from breast cancer.¹⁹ CT is the imaging modality of choice for the identification of this process, which is most often bilateral (Figure 5). Findings include irregular thickening of the interlobar septa and peribronchovascular sheaths as well as thickening of the structures in the central regions of the secondary pulmonary lobules.²⁰

FIGURE 5 Axial lung window chest CT image from a 63-year-old woman with bone, thoracic, and brain metastases shows a large right pleural effusion. The lung parenchyma, particularly on the right, shows bronchovascular thickening and a finely nodular appearance of the interstitium, highly suggestive of lymphangitic tumor spread.

REFERENCES

1 Patanaphan V, Salazar OM, Ricco R. Breast cancer: metastatic patterns and their prognosis. South Med J. 1998;81(9):1109-1112.

2 Shaw JP, Glassman LR. Thoracic metastases from breast cancer. In: Breast Cancer. Philadelphia: Elsevier Churchill Livingston; 2005:661-666.

3 Meyer JE, Munzenrider JE. Computed tomographic demonstration of internal mammary lymph-node metastases in patients with locally recurrent breast carcinoma. Radiology. 1981;139:661-663. (A)

4 Thomas JM, Redding WH, Sloane JP. The spread of breast cancer: importance of intrathoracic lymphatic route and its relevance to treatment. Br J Cancer. 1979;40:540-547.

5 Feirerstein IM, Jicha DL, Pass HL, et al. Pulmonary metastases: MR imaging with surgical correlation—a prospective study. Radiology. 1992;181:123-129.

6 Müller NL, Gamsu G, Webb WR. Pulmonary nodules: detection using magnetic resonance and computed tomography. Radiology. 1985;155:687-690.

7 Panicek DM. MR imaging for pulmonary metastases? Radiology. 1992;182:10-11.

8 Webb WR, Sustman HD. MR imaging of thoracic disease: clinical uses. Radiology. 1992;182:621-630.

9 Ohno Y, Sugimura K, Hatabu H. MR imaging of lung cancer. Eur J Radiol. 2002;44(3):172-181.

10 Vogt FM, Herborn CU, Hunold P, et al. HASTE MRI versus chest radiography in the detection of pulmonary nodules: comparison with MDCT. AJR Am J Roentgenol. 2004;183:71-78.

11 Eubank WB, Mankoff DA, Takasugi J, et al. ¹⁸Fluorodeoxyglucose positron emission tomography to detect mediastinal or internal mammary metastases in breast cancer. J Clin Oncol. 2001;19:3516-3523.

12 Vansteenkiste JF, Stroobants SG, De Leyn PR, et al. Lymph node staging in non-small-cell lung cancer with FDG-PET scan: a prospective study on 690 lymph node stations from 68 patients. J Clin Oncol. 1998;16(6):2142-2149.

13 Scott WJ, Gobar LS, Terry JD, et al. Mediastinal lymph node staging of non-small-cell lung cancer: a prospective comparison of computer tomography and positron emission tomography. J Thorac Cardiovasc Surg. 1996;111(3):642-648.

14 Crippa F, Seregni E, Agresti R, et al. Association between [¹⁸F]-fluorodeoxyglucose uptake and postoperative histology, hormone receptor status, thymidine labeling index and p53 in primary breast cancer: a preliminary observation. Eur J Nucl Med. 1998;25:1429-1434.

15 Avril N, Menzel M, Dose J, et al. Glucose metabolism of breast cancer assessed by ¹⁸F-FDG PET: histologic and immunohistochemical tissue analysis. J Nucl Med. 2001;42:9-16.

16 Connolly JEJr, Erasmus JJ, Patz EFJr. Thoracic manifestations of breast carcinoma: metastatic disease and complications of treatment. Clin Radiol. 1999;54:487-494.

17 Jung JI, Kim HH, Park SH, et al. Thoracic manifestations of breast cancer and its therapy. RadioGraphics. 2004;24:1269-1285.

18 Casey JJ, Stempel BG, Scanlon EF, et al. The solitary pulmonary nodule in the patient with breast cancer. Surgery. 1984;96:801-805.

19 Kreisman H, Wolkove N, Finkelstein HS, et al. Breast cancer and thoracic metastases: review of 119 patients. Thorax. 1983;38:175-179.

20 Webb WR, Müller NL, Naidich NP. High-Resolution CT of the Lung, 3rd ed. Philadelphia: Lippincott Williams & Wilkins, 2001.

CASE 1 Solitary pulmonary nodule in the breast cancer patient: Primary lung cancer versus breast cancer metastasis

An 80-year-old woman with emphysema was found to have a left upper lobe (LUL) mass on chest x-ray. This was confirmed on chest CT, which also showed a similar-appearing, suspicious right lower lobe (RLL) lesion (Figure 1). Fine-needle aspiration (FNA) of the LUL lesion was consistent with a non–small cell lung carcinoma. The patient’s past medical history was significant for prior left mastectomy for breast cancer nearly 20 years earlier. Lymph nodes were reportedly negative, and the patient did not undergo additional therapy at that time. She also had a significant past smoking history, on the order of 1.5 packs per day.

FIGURE 1 Two chest CT sections show similar-appearing lung masses, one in the anterior segment of the LUL (A), and the other in the superior segment of the RLL (B). Both display suspicious morphology, with irregular spiculated margins, suggesting primary bronchogenic cancers in appearance.

The PET scan showed both the LUL and RLL lesions to be hypermetabolic (Figure 2). Based on the information available, this result seemed most consistent with synchronous lung cancers. However, subsequent immunocytochemical stains performed on the material obtained from FNA suggested the cause to be metastatic breast carcinoma.

FIGURE 2 A and B, PET shows both lesions to be hypermetabolic, consistent with malignancy at both sites.

TEACHING POINTS

This is not an uncommon clinical scenario. Identification of solitary pulmonary nodules in patients with a history of breast cancer represents new diagnoses of lung cancer rather than metastatic breast cancer in about 50% of cases. Differentiation between these two diagnoses generally cannot be reliably made on morphologic grounds, as seen in this case, and requires histologic sampling. The distinction between primary lung adenocarcinoma (60% to 70% of lung cancers) and metastatic adenocarcinoma of breast is particularly difficult. Differentiation generally comes down to immunohistochemical stains.

In the rare patient who has a true solitary pulmonary breast cancer metastasis, there are data to suggest survival benefit from metastasectomy. However, consideration of surgical treatment requires careful imaging exclusion of other metastases. Today, such an evaluation should include bone scan; enhanced chest, abdomen, and pelvis CT scans; and positron emission tomography (PET) scan.

CASE 2 Lung metastases, progression to pleural metastases

A 56-year-old woman with biopsy-proven lung metastases, taking exemestane (Aromasin), was restaged after transferring her care. She was 8 years post diagnosis of a palpable node-negative, 3.5-cm left breast cancer, estrogen receptor positive, HER-2/neu negative. She had been treated with lumpectomy, radiation, and chemotherapy (four cycles of doxorubicin [Adriamycin] plus cyclophosphamide [Cytoxan] (AC)), but no hormone therapy.

Four years after her initial diagnosis, a chest x-ray showed lung nodules. Malignancy was confirmed with fine-needle aspiration. She was started on tamoxifen. Subsequent progression prompted a change in therapy to Aromasin.

At the time of restaging, the patient was essentially asymptomatic. Chest CT and positron emission tomography (PET) showed four lung nodules on CT, three of which were hypermetabolic on PET (a 3 × 7-mm nodule was probably too small) (Figures 1, 2, 3, 4, and 5).

FIGURE 1 Contrast-enhanced chest CT (A, soft tissue window; B, lung window) images show two right lung nodules. The superior segment right lower lobe lesion measures 3 cm, whereas the right upper lobe lesion measures 1 cm. Fibrosis of the left anterior subpleural lung is from prior radiation therapy for breast cancer.

FIGURE 2 Lung window through the lower lobes shows a 3 × 7-mm nodule in the right lower lobe, which showed no correlate on PET, probably because of its size.

FIGURE 3 Axial PET/CT images show two hypermetabolic lung metastases.

FIGURE 4 A more inferior PET/CT image shows hypermetabolism in the two right lung lesions depicted in Figure 1.

FIGURE 5 A to D, Coronal PET/CT images show uptake in bilateral upper lobe metastases.

Seven months later, follow-up CT scans showed new right pleural and mediastinal disease, confirmed as fluorodeoxyglucose avid on PET (Figures 6 and 7). The patient was not particularly symptomatic, and was reluctant to have chemotherapy. She began a trial of androgenic hormone therapy (Halotestin).

FIGURE 6 Repeat chest CT, 7 months later. A, New, plaque-like right pleural soft tissue masses are now seen, in addition to the known left upper lobe (LUL) metastasis. B, Lung window from the same level as in A shows the LUL metastasis, left anterior subpleural radiation therapy change (arrowheads), and right pleural metastases (arrows). C, New nodal disease is seen at this level (azygoesophageal recess and right internal mammary), in addition to new right pleural metastases and the preexisting right lung nodules. A new pleural effusion is also noted. D, Lung window from the same level as C. E, A right paraspinal pleural based mass is outlined by pleural effusion at this level. F, Pleural metastases are seen in major and minor fissures on the right on this slice. G, Pleural metastases stud the right diaphragmatic pleura. H, Lung window from the same level as G.

FIGURE 7 Concurrent PET/CT, 7 months after PET scan in Figures 3 to 5, shows new peripheral linear and nodular right hemithoracic hypermetabolism, which corresponds to the new pleural masses on CT and represents pleural metastases.

When she was re-evaluated 3 months later, she noted right posterior chest pain and dry cough. Imaging showed further progression. She was restaged in conjunction with the entrance requirements of a chemotherapy clinical trial. These studies showed progression of right pleural metastases and new liver metastases (Figure 8).

FIGURE 8 Coronal volumetric projection image, 5 months later, documents disease progression. The right hemithorax is encased in hypermetabolic tumor. Small metabolically active foci now are seen in the liver as well.

TEACHING POINTS

Malignant pleural effusions developing in breast cancer patients occur more frequently on the side ipsilateral to the breast cancer. This is thought to be lymphatic in etiology, due to transthoracic lymphatic communications, such as between internal mammary and mediastinal lymph nodes. Malignant pleural effusions can also result from hematogenous spread. Malignant pleural effusions developing contralateral to breast cancer may be more likely hematogenous in origin than ipsilateral pleural recurrences. In this case, this patient had known pulmonary metastases for more than 4 years, with little symptomatology on a variety of hormonal regimens, before progressing rapidly to pleural metastases and malignant effusion contralateral to her breast cancer.

This case reminds us of the resolution limits of whole-body PET scanners. With current technology, the lower limit of whole-body PET resolution is about 6 mm. Accordingly, it is not surprising that the smallest lung nodule found in this patient (3 × 7 mm) was not visualized on PET.

CASE 3 Chest wall, pleural, and thoracic nodal recurrence

A 55-year-old woman noted a palpable abnormality in her right breast, which had been treated for breast cancer 8 years before with lumpectomy, axillary lymph node dissection, radiation, and chemotherapy. Physical examination showed an obvious nodular skin recurrence. Her prior tumor was a poorly differentiated, 6.5-cm primary, estrogen receptor and progesterone receptor negative, HER-2/neu negative, with negative lymph nodes.

Mammography confirmed a corresponding 2.5-cm spiculated mass with suspicious pleomorphic microcalcifications, new from a prior study 4 years earlier. Palpation-guided fine-needle aspiration (FNA) was positive for malignant cells, and a concurrent skin punch biopsy showed adenocarcinoma involving the dermis, estrogen receptor negative, progesterone receptor weakly positive, and HER-2/neu negative.

Initial evaluations suggested local recurrence only, and mastectomy was planned. However, her CA-125 was noted to be elevated, leading to restaging with positron emission tomography (PET)/CT, diagnostic enhanced body CT scans, and bone scan. PET and CT findings suggested thoracic recurrence (Figures 1, 2, 3, 4, and 5). CT scans showed a moderate-sized right pleural effusion with modest associated metabolic activity on PET. A 1-cm nodule in the right major fissure was metabolically active, as was a left hilar lymph node. The known right breast local recurrence was intensely hypermetabolic and could be faintly visualized as soft tissue activity on bone scan (Figure 6).

FIGURE 1 Enhanced chest CT images (A, lung window; B, soft tissue window) show a right pleural effusion. Right anterolateral subpleural lung consolidation is seen, in a characteristic distribution for radiation fibrosis. Note the straight linear edge of the lung changes. Right breast skin thickening is also seen on this slice.

FIGURE 2 A more inferior lung window section, at the level of the aortopulmonary window, again shows right anterolateral subpleural radiation fibrosis changes and a right pleural effusion. Nodules up to 1 cm are seen in the right major fissure, suggesting that the etiology of the pleural fluid is neoplastic.

FIGURE 3 A more inferior slice, soft tissue window, shows a right breast mass, extending to the skin, which is locally thickened. A left hilar lymph node seen at this level was metabolically active on PET.

FIGURE 4 Axial PET/CT images, corresponding in level to Figure 2, show the right pleural nodules to be metabolically active. The associated activity is better seen on the NAC image (upper right) than on the AC image (lower right). A linear band of right anterolateral subpleural activity corresponds with the radiation fibrosis changes on CT, and was 8 years old. Activity corresponding with the left hilar lymph node is also seen. Note also subtle increased dependent activity in the right hemothorax compared with the left, seen on fused PET/CT (lower left) and AC PET (lower right), correlating with the pleural effusion. This, coupled with the pleural nodularity and activity, suggests a malignant etiology of the effusion, which was subsequently confirmed histologically.