Internal Mammary Lymph Nodes: Pathologic Correlation by Ultrasound-Guided Fine-Needle Aspiration and Assessment of Associated Risk Factors

Transcription

1 Nuclear Medicine and Molecular Imaging Original Research Wang et al. 18 F-FDG PET/CT Positive IM lymph nodes Nuclear Medicine and Molecular Imaging Original Research Carolyn L. Wang 1,2 Marna J. Eissa 2,3 James V. Rogers 1 Aleksandr Y. Aravkin 2,4 ruce A. Porter 1 J. David eatty 1 Wang CL, Eissa MJ, Rogers JV, Aravkin AY, Porter A, eatty JD Keywords: breast cancer, FDG PET/CT, internal mammary lymph nodes DOI: /AJR Received February 16, 2012; accepted after revision September 17, Swedish Cancer Institute, 1221 Madison St, Seattle, WA Address correspondence to C. L. Wang (wangcl@uw.edu). 2 Department of Radiology, University of Washington, Seattle, WA. 3 Present address: Southern Louisiana Veterans Healthcare System, New Orleans, LA. 4 RSM, IM T. J. Watson Center, Yorktown, NY. AJR 2013; 200: X/13/ American Roentgen Ray Society 18 F-FDG PET/CT Positive Internal Mammary Lymph Nodes: Pathologic Correlation by Ultrasound-Guided Fine-Needle Aspiration and Assessment of Associated Risk Factors OJECTIVE. Metastatic breast cancer in internal mammary (IM) lymph nodes is associated with a poor prognosis. This study correlates 18 F-FDG PET/CT positive IM lymph nodes with ultrasound-guided fine-needle aspiration (FNA) cytopathologic results and determines risk factors for IM node positivity on PET/CT. MATERIALS AND METHODS. For this retrospective study, a database search was performed to identify patients referred for whole-body 18 F-FDG PET/CT for initial staging or restaging of breast cancer from January 1, 2005, through December 31, The radiology reports and images were reviewed for patients with 18 F-FDG avid IM lymph nodes on PET/ CT and correlated with the cytopathologic results from FNA of selected PET/CT-positive IM lymph nodes. The patients with positive IM nodes on PET/CT who underwent PET/CT for initial staging were compared against age-matched and tumor size matched patients to identify risk factors for IM node positivity on PET/CT. RESULTS. One hundred ten of 1259 patients (9%) had an 18 F-FDG avid IM lymph node on PET/CT. Twenty-five patients underwent ultrasound-guided FNA of a suspicious IM node, and 20 IM lymph nodes (80%) were cytologically proven metastases from the primary breast malignancy. High tumor grade, the presence of lymphovascular invasion (LVI), and triple receptor negative hormonal receptor status were found to be significant risk factors for IM node positivity on PET/CT (p < 0.05). CONCLUSION. Although fewer than 10% of breast cancer patients have positive IM nodes on 18 F-FDG PET/CT performed for initial staging or restaging, a positive IM node indicates a very high likelihood of malignant involvement on ultrasound-guided FNA. The presences of high tumor grade, LVI, or triple receptor negative status are risk factors for IM node positivity on 18 F-FDG PET/CT. reast cancer is the most frequent malignancy in women and is the second leading cause of cancer deaths in Western countries [1]. Many standard risk factors for poor prognosis include tumor size, axillary node status, hormone receptor status, and human epidermal growth receptor 2 (HER2 [also known as ERR2]) status. However, some studies have also shown that the presence of internal mammary (IM) metastasis is a useful prognostic indicator and that IM metastasis has been associated with higher rates of distant disease and lower overall survival rates [2 4]. Although the gold standard for establishing the status of IM metastases is surgical, most surgeons do not routinely perform IM lymph node dissections or evaluate the IM nodes. Two recent studies from European groups who do routinely biopsy IM nodes surgically have shown that improved accuracy of staging results in more appropriately tailored therapy and therefore improvement in survival [5, 6]. Other studies have shown that 18 F-FDG PET could be used as a noninvasive method of evaluating the IM nodes [7, 8]; however, those studies are limited because there was not pathologic correlation of the positive IM nodes [7, 8]. A single study presented as an exhibit at a meeting of the Society of Nuclear Medicine and Molecular Imaging (Cho I, et al., presented at the 2010 annual meeting of the Society of Nuclear Medicine and Molecular Imaging) correlated pathologic biopsy results of nine patients with IM lymph node uptake on PET/CT. Of the 13 FDG-avid nodes, 12 were pathologically positive and there were nine of 11 IM nodes that were falsely negative in a single patient. Our goal in this study was 1138 AJR:200, May 2013

2 18 F-FDG PET/CT Positive IM lymph nodes to present a larger series of cytopathologically correlated 18 F-FDG PET/CT positive IM nodes and to assess possible risk factors for IM node positivity on 18 F-FDG PET/CT. Materials and Methods This retrospective study was approved by our institution s institutional review board and was HIPAA compliant. Our outpatient facility for PET/CT is embedded in a large urban multidisciplinary community cancer treatment center that is part of an inpatient and outpatient complex of buildings covering 12 city blocks. Patients We retrospectively reviewed our database using codes from the International Classification of Diseases, ninth revision [9] to identify patients with breast cancer who were referred for whole-body 18 F-FDG PET/CT from January 1, 2005, through December 31, The reports and images were reviewed. All patients who had 18 F-FDG PET/CT positive IM lymph nodes, which was defined as IM lymph node activity greater than adjacent pectoralis muscle background activity, were included in the study. The medical records, pathology reports, and ultrasound-guided IM lymph node fine-needle aspiration (FNA) reports of all patients were reviewed. A cohort of patients who were referred for whole-body 18 F-FDG PET/CT for initial staging of breast cancer during the same time period but did not have PET/CT-positive IM nodes was also analyzed for comparison to identify risk factors. Whole-ody PET/CT After patients had fasted for a minimum of 6 hours, whole-body scanning was performed on a PET/CT system (DST PET/16 slice LightSpeed CT, GE Healthcare). A CT examination was performed for attenuation correction and lesion localization using the automatic exposure control A Fig year-old woman with right breast infiltrating ductal carcinoma with suspicious right internal mammary (IM) nodes on PET/CT. Ultrasound-guided fine-needle aspiration (FNA) of abnormal 8-mm node in right third intercostal space yielded malignant cells. A, Static gray-scale ultrasound image of right third intercostal space shows suspicious-appearing 8-mm IM lymph node., Static gray-scale ultrasound image obtained during ultrasound-guided FNA shows tip of needle within IM node. technique (Smart ma, GE Healthcare) with the following parameters: ma, 20.0-mm collimation, 140 kvp, and 70-cm display FOV. The patients were scanned in the supine position from the skull base to the mid thigh for the PET/CT examination. Image Analysis The PET/CT images were reviewed on a workstation (Advantage Workstation, GE Healthcare). The attenuation-corrected images were converted to show standardized uptake value (SUV) by normalizing for injected dose and body weight. The maximum SUV (SUV max ) of the lesion was determined by drawing a region of interest around the lesion and viewing it in more than one plane to ensure that the entire lesion and only the lesion was included. Technique for Ultrasound-Guided Fine-Needle Aspiration of Internal Mammary Lymph Nodes The 18 F-FDG PET/CT images were first reviewed by the radiologist performing the procedure to identify and localize the suspicious IM node. The node was then confirmed to be abnormal on the basis of the following ultrasound features: oval to rounded shape, hypoechoic appearance with absence of the fatty hilum, and diameter greater than 4 5 mm (Fig. 1A). Written informed consent was obtained from the patient for the FNA biopsy. Aseptic technique was used for FNA, and infiltration of the skin and subcutaneous tissues was performed using 2 3 ml of buffered 1% lidocaine. Doppler imaging guidance was used to avoid vessels, and a craniocaudal or, occasionally, a lateral medial approach was taken depending on the location of the node, access to the node, and the presence of adjacent vessels. Aspirations were performed with 22-gauge needles on a 10-mL syringe using mild vacuum and needle rotation during aspiration (Fig. 1). A total of three passes with three separate needles were performed to ensure adequate tissue sampling. Three smears are made and the slides are placed in containers filled with alcohol. The needles and syringes are flushed with a preservative collection fluid (methanol-water solution [CytoLyt solution, Cytyc]) for preparation of a cell block. The samples were promptly transported to the pathology facility located on the same campus in a separate building. Postbiopsy ultrasound was performed to assess for possible hematoma. There were no complications from any of the IM node FNAs performed in this series. Histopathologic Evaluation Our institution s pathology department used their standard clinical protocol to evaluate all resection specimens. In addition, pathology reports of prior lesion biopsies were reviewed. The following pathologic information was included in the study analysis: histologic type; histologic grade; and estrogen receptor (ER), progesterone receptor (PR), and HER2 status. The histologic grade was assessed using the Nottingham modification of the loom- Richardson score index [10]. The expression of ER, PR, and HER2 was assessed by immunohistochemical staining. According to our institutional protocol, any expression of ER or PR greater than 1% was considered positive [11, 12]. An HER2 value of 0 or 1 was considered negative (HER2-negative) and 3 was considered positive (HER2-positive). An HER2 value of 2 was considered indeterminate; for indeterminate cases, fluorescent in situ hybridization was performed and values of more than 2.0 or 2.2 were considered positive (HER2-positive) [13]. Statistical Analysis A proportions test was performed using the R program (R Project for Statistical Computing, R version 2.9.2) to compare the proportion of triple receptor negative breast tumors in our study with that in a large series study [14]. To evaluate potential risk factors for metastatic IM lymph nodes, we compared the subset of pa- AJR:200, May

3 Wang et al. TALE 1: Characteristics of Study Patients and Primary reast Cancer tients with positive IM nodes on 18 F-FDG PET/ CT who underwent PET/CT for initial staging only with an age-matched, tumor size matched group of breast cancer patients who presented for 18 F- FDG PET/CT for initial staging during the same study period. Logistic regression was used to evaluate the effect of potential risk factors on the odds of having an outcome of a positive IM node on 18 F- FDG PET/CT. The risks factors evaluated were medial location of the primary tumor, SUV max of the primary tumor, tumor grade, hormone receptor status and combinations, the presence of ductal carcinoma in situ (DCIS), the presence of lymphovascular invasion (LVI), and the presence of a positive axillary node on either imaging or pathology. Results During the study period from January 1, 2005, through December 31, 2010, 1259 patients were referred to our institution for whole-body 18 F-FDG PET/CT for initial staging or restaging of breast cancer. Of those patients, 110 (9%) had positive IM lymph nodes on whole-body 18 F-FDG PET/CT, which was defined as IM lymph nodes showing greater activity than adjacent pectoralis muscle background. Seven patients were excluded because data about the primary cancer including receptor status or histologic grade were incomplete, and one patient was excluded Characteristic No. (%) of Patients (n = 102) Medial location of breast tumor 32 (31) Pathologically positive axillary lymph node on initial biopsy or lymph 62 (61) node dissection Axillary lymph node FDG-avid on PET/CT 70 (69) DCIS present in breast tumor 54 (53) LVI present in breast tumor 45 (44) IDC breast tumor 89 (87) ILC breast tumor 3 (3) Mixed IDC-ILC breast tumor 2 (2) Other histology a 8 (8) Receptor status ER-positive, PR-positive, HER2-positive 11 (11) ER-positive, PR-positive, HER2-negative 37 (36) ER-positive, PR-negative, HER2-positive 3 (3) ER-positive, PR-negative, HER2-negative 4 (4) ER-negative, PR-negative, HER2-positive 15 (15) ER-negative, PR-negative, HER2-negative 32 (31) Note DCIS = ductal carcinoma in situ, LVI = lymphovascular invasion, IDC = invasive ductal carcinoma, ILC = invasive lobular carcinoma, ER = estrogen receptor, PR = progesterone receptor, HER2 = human epidermal growth receptor 2. a Metaplastic, n = 2; micropapillary, n = 1; medullary, n = 1; squamous, n = 1; small cell, n = 1; unknown, n = 2. because the histologic diagnosis was angiosarcoma. Therefore, only 102 patients were analyzed: 72 underwent imaging for initial staging and 30, restaging. All patients were female (mean age, 52.6 ± 11.2 [SD] years). The mean size of the IM lymph nodes was 7.7 mm (SD, 6.8 mm; range, mm) and the mean SUV max was 3.7 (SD, 3.4; range, ). The mean SUV max of the associated primary breast tumor was 11.0 (SD, 5.7; range, ). The mean histologic grade as expressed by the Nottingham modification of the loom-richardson score was 7.9 ± 1.4. Forty-one (40%) of the primary breast cancers with positive IM nodes were on the right. Table 1 lists the characteristics of the study patients and primary breast tumors. Of the 102 patients who had PET/CT-positive IM nodes, 25 patients (25%) underwent ultrasound-guided FNA of the suspicious IM nodes. In this subset of patients, a cytopathologically proven positive IM node would have been the only site of disease that would have upstaged disease in the patient and changed management, which is why PET/CT-positive IM nodes warranted biopsy. Twenty of these 25 patients (80%) had positive cytopathologic findings for malignancy (Fig. 2). In five of the 25 IM nodes (20%), the FNA biopsy was negative. In two of the 16 patients (13%) who underwent imaging for initial staging and had positive IM nodes on cytology, the IM lymph node was the only site of metastasis (negative axillary sentinel lymph node biopsy) (Fig. 3). One of the coauthors, a breast surgeon with 34 years of surgical oncology experience with a breast cancer emphasis (i.e., 13 years exclusively breast cancer), reviewed the clinical charts of the 77 patients with PET/CT-positive IM nodes that did not undergo FNA. He concluded that it would have been appropriate for 19 patients (25%) to have had their IM nodes further evaluated by ultrasound, MRI, or ultrasound-guided FNA. The principal clinical rationales for not performing ultrasound-guided FNA of the IM node in the remaining 58 patients (75%) are listed in Table 2. We used a proportions test to compare the percentage of triple receptor negative tumors (32/102, 31%) in this series with an incidence rate of 13% for triple receptor negative tumors in a large series consisting of 61,309 women with breast cancer [14] and found that this difference is highly statistically significant (p = 8.3e 8 ). Of the 102 patients in the study with positive IM nodes on 18 F-FDG PET/CT, the 69 patients who underwent imaging for initial staging were compared with 86 agematched and tumor size matched breast cancer patients without positive IM nodes on initial staging 18 F-FDG PET/CT. The three additional initial staging patients with IM-positive nodes were also in the agematched and tumor size matched database and had to be deleted from this analysis in both arms. The mean age of the study population was 53.5 ± 12.5 (SD) years and of the control group, 51.9 ± 12.0 years (p = 0.4). The mean size of the primary breast tumor of the study group was 4.3 ± 3.0 cm versus that of the control group, 4.1 ± 3.1 cm (p = 0.8). The two groups are well matched by age and tumor size. Table 3 summarizes the odds ratio (OR) for each of the risk factors with the 95% CIs for the ORs and p values. For example, a patient with LVI is almost four times more likely to have a positive IM node on 18 F-FDG PET/CT than a patient without LVI with all other things being equal. Positive ER status was protective, with an OR of This result means that, on average, a patient with negative ER status is 2.8 times more likely to have a positive IM node on 18 F-FDG PET/CT than a comparable patient with positive ER 1140 AJR:200, May 2013

4 18 F-FDG PET/CT Positive IM lymph nodes status. The SUV max of the primary tumor is a continuous variable, so for each unit increase of SUV max, there is on average a 1.13 multiplicative increase of risk of having a positive IM node on 18 F-FDG PET/CT. Other factors that were analyzed but were not statistically significant (p > 0.05) were DCIS, medial location, and positive HER2 status. There were 16 of 86 patients (19%) with an HER2-positive status in the control group. The other receptor combinations of ER-positive, PRpositive, and HER2-positive and of ER-negative, PR-negative, and HER2-positive were not statistically significant. Fig year-old woman with 3.1-cm invasive ductal carcinoma of medial right breast (maximum standardized uptake value [SUV max ] = 3.1), cytopathologically proven 5-mm short-axis internal mammary (IM) lymph node metastasis (SUV max = 1.2), and no axillary lymph nodes. Primary neoplasm was estrogen receptor positive, progesterone receptor positive, and human epidermal growth receptor 2 negative. A and, CT (left), PET (center), and fusion (right) images of breast primary tumor (A) and cytopathologically proven IM metastasis (). Discussion reast cancer is the leading cancer in women, and appropriate staging affects prognosis, treatment, and survival. Several studies have also shown that the presence of IM lymph node metastasis is a useful prognostic indicator and that IM metastasis has been associated with higher rates of distant disease and lower overall survival rates [2 4]. Although surgical evaluation is the gold standard for determining IM node metastases, most breast surgeons do not routinely sample the IM nodes surgically because studies in the 1970s and 1980s comparing Halsted mastectomy versus extended mastectomies (including IM node dissection) did not show improved overall survival [15]. Recent data have suggested that lymphoscintigraphy, particularly when performed using peritumoral, intratumoral, or subtumoral injections, was a useful method for identifying a significant proportion of tumors that had primary IM node drainage A C Fig year-old woman with 3.3-cm invasive ductal carcinoma of medial left breast (maximum standardized uptake value [SUV max ] = 7.2), cytopathologically proven 6-mm short-axis internal mammary (IM) lymph node metastasis (SUV max = 1.2), and additional positive axillary nodes. Primary neoplasm was estrogen receptor negative, progesterone receptor negative, and human epidermal growth receptor 2 negative. A C, CT (left), PET (center), and fusion (right) images of breast primary tumor (A), cytopathologically proven IM metastasis (), and positive axillary lymph nodes (C). [16 22] because anatomic studies have shown that IM nodes are supplied primarily by retromammary lymphatic [17, 20]. A recent study that compiled and analyzed data from six prospective studies of lymphoscintigraphy for axillary sentinel lymph node biopsy showed that 25% had IM node drainage but only 17% were pathologically positive [18]. Many centers including ours perform only subareolar injections, which improve axillary node uptake but rarely show IM uptake because of the superficial lymphatic channels. Studies have clearly shown that patients with IM nodal involvement have a worse prognosis, and many investigators theorize that unrecognized, untreated IM node metastasis explains why patients with medial tumors have a worse prognosis than those with tumors elsewhere in the breast [23]. Many centers including ours rely on imaging to identify IM nodal involvement. Eubank et al. [7] reported that PET was a useful noninvasive method for evaluating IM and mediastinal lymph nodes with an 85% sensitivity and 90% specificity based on biopsy or follow-up CT, although none of the biopsies was of an IM lymph node. Segaert et al. [24] found that PET/CT had a 100% sensitivity versus 85% for FDG PET and 67% for CT in a small series of eight patients, but biopsy confirmed the imaging findings in only A AJR:200, May

5 Wang et al. TALE 2: Reasons 58 Patients Did Not Undergo Ultrasound-Guided Fine-Needle Aspiration (FNA) for Evaluation of 18 F-FDG PET/CT Positive Internal Mammary (IM) Lymph Node one case. To our knowledge, our study is one of the largest series of PET/CT-positive IM nodes with pathologic correlation. Nine percent of our patients who underwent 18 F-FDG PET/CT for initial staging or restaging of breast cancer had at least one FDG-avid IM lymph node, which was defined as IM node uptake greater than background pectoralis muscle uptake. Studies by ellon et al. [8] and Segaert et al. [24] showed a higher rate of IM-positive nodes on 18 F-FDG PET/CT (25% and 13%, respectively) than our study; this difference likely relates to the fact that they included only patients undergoing initial staging before initiation of neoadjuvant chemotherapy. Patients who have neoadjuvant chemotherapy tend to have larger tumors with a much higher rate of node positivity [25]. Our inclusion of patients who were undergoing imaging for restaging in addition to those undergoing imaging for initial staging likely brought down the rate because recurrence in IM nodes is relatively low, ranging from 1% to 5% [20, 26 28]. Our oncologists and surgeons typically use PET/CT for initial staging of breast cancer before initiation of therapy for patients with large (> 5 cm) primary tumors (T3) or axillary nodes with metastatic disease confirmed by FNA or core biopsy or for patients for whom neoadjuvant chemotherapy is planned [29]. ecause PET/ CT is not used routinely for staging for clinical stage I disease by imaging or pathology, the tumors will be larger and higher in stage (IIb and III). Our results show that an FDG-avid IM lymph node on PET/CT has a very high likelihood of being malignant because 80% of the 25 patients who underwent FNA had positive results for neoplasm. Metastasis in an IM lymph node has the greatest effect on staging Reason No. of Patients Patient had known stage IV disease at time of PET/CT 25 Patient had another hypermetabolic site on PET/CT that was biopsied to confirm stage IV disease 12 Patient had widespread disease on PET/CT that was assumed to be stage IV without pathologic confirmation 3 Planned radiation treatment field extended to include PET/CT-positive IM node, which would have been performed even if FNA of IM node 8 yielded negative cytologic result IM node also appeared suspicious on another imaging modality (ultrasound, MRI, or both); therefore, clinicians thought imaging results were 7 adequate for diagnosis of metastasis Patient had positive biopsy result of a PET/CT-positive axillary node; therefore, PET/CT-positive IM node was presumed to be positive also 1 Negative result for sentinel lymph node dissection of PET/CT-positive axillary node; therefore, PET/CT-positive IM node was presumed to be 1 false-positive as well No corresponding suspicious IM node seen on chest wall ultrasound; therefore, ultrasound-guided FNA of IM node was not performed 1 when the IM node is the only site of regional lymph node metastasis (negative axillary nodes) in the patient. This finding would upstage disease from N0 to pn2b (from stage IIa or IIb to stage IIIa or IIIb) with an associated drop in predicted survival from 81% or 74% to 67% or 41%, respectively [30]. In older studies, IM nodal involvement without axillary involvement was seen in up to 11% of patients and the presence of IM node metastasis would have upstaged disease and altered therapy including initiating adjuvant chemotherapy for patients with negative axillary lymph nodes or initiating or modifying radiation therapy [2]. Previous studies have shown that patients with IM node metastasis had a worse prognosis than those who did not but that patients with IM lymph node metastases had a better-than-predicted 5-year survival after IM radiotherapy and chemotherapy [4]. Heuts et al. [31] also suggested that tailored adjuvant systemic therapy and additional parasternal radiotherapy have a beneficial effect on the prognosis of these patients. Although prior studies have suggested that medial tumors and tumors with positive axillary nodes are more likely to have IM nodal involvement, Heuts et al. [31] reported that tumor location was not as good a predictor of IM nodal involvement as previously suspected. They found that 44% of lateral tumors had a parasternal hot spot on lymphoscintigraphy and that biopsy of 76% of IM nodes with medial tumors were negative for malignancy. Similar to their findings, we found that only 31% of the IM nodes with positive findings on 18 F-FDG PET/CT were associated with medial tumors and that medial location was not a statistically significant risk factor in our cohort analysis. Positive axillary nodes are also known to be a risk factor for IM node disease, which was also confirmed in our cohort study. However, surgical data have shown that 6 16% of patients with negative axillary nodes had positive IM nodes [20, 26, 27, 32, 33]. Therefore a more accurate way of assessing for IM node metastasis should be used to identify patients who might benefit the most from adjuvant parasternal radiation because it is associated with risks. There is widespread controversy in the radiation literature about the use of radiation therapy for IM nodes and the risk of cardiac toxicity particularly with the advent of newer chemotherapeutic agents such as trastuzumab [20]. It is worth noting that 31% of IM node positive lesions on 18 F-FDG PET/CT were associated with triple receptor negative tumors. This finding is a statistically higher percentage of that less common subtype compared with a large series (13%) [14] and also with our dataset of patients with breast cancer who underwent 18 F-FDG PET/CT as part of their workup for initial staging (11%) [34]. The risk factor of triple receptor negative status had a 3.41 OR for positive IM nodes on 18 F-FDG PET/CT. This OR suggests that this particularly aggressive biologic subtype might have a prevalence for IM node metastases and a person with triple receptor negative breast cancer is at an increased risk for IM nodal involvement. Large prospective studies of this subpopulation will need to be performed in the future to assess whether triple receptor negative breast cancer may be an independent risk factor for IM node metastasis, distant disease, or both and may be a reason to consider 18 F-FDG PET/ CT for staging. At our institution, if IM node positivity is the only factor that will upstage disease to stage III, our clinicians will generally refer the patient for FNA. If there is evi AJR:200, May 2013

6 18 F-FDG PET/CT Positive IM lymph nodes TALE 3: Odds Ratio (OR) Results for Risk Factors for 18 F-FDG PET/CT Positive Internal Mammary Lymph Nodes Risk Factor OR (95% CI) p SUV max of primary breast tumor 1.13 ( ) Nottingham modification of the loom-richardson score 1.48 ( ) Lymphovascular invasion 3.73 ( ) Positive axillary lymph node 6.66 ( ) ER-positive 0.31 ( ) PR-positive 0.32 ( ) ER-negative, PR-negative, HER2-negative 3.41 ( ) ER-positive, PR-positive, HER2-negative 0.34 ( ) Note SUV max = maximum standardized uptake value, ER = estrogen receptor, PR = progesterone receptor, HER2 = human epidermal growth receptor 2. dence of stage III disease or higher, radiation oncologists have been including the positive IM node in the treatment field without biopsy. The ultrasound-guided FNA technique for IM nodes was originally developed at our institution to evaluate enlarged IM lymph nodes identified on MRI in lymphoma patients undergoing staging or restaging before, during, or after treatment. The two major risks of percutaneous or open surgical biopsy of IM lymph nodes are bleeding and pneumothorax, neither of which occurred in any patient in this study. The ultrasound guidance provides an opportunity to identify the IM lymph node of concern detected on MRI or PET/CT. If the specific node of concern on PET/CT is not identified or if it is identified but is small and normal in appearance, FNA is not performed. When FNA of the IM lymph node is performed, the ultrasound guidance provides an opportunity to select a safe approach and undertake the procedure while avoiding the subcostal and IM vessels and the underlying pleura. In this study, all FNA samples contained cytologic evidence of lymphocytes, malignant cells, or both. In our experience, ultrasound-guided FNA of IM lymph nodes is simple, safe, and accurate in targeting the node of suspicion. The alternative of open surgical biopsy guided by a PET probe is complex, has much greater patient morbidity, and has a greater risk of complications. We recommend that other centers use ultrasound-guided FNA for PET/CT-positive IM lymph nodes and gain experience with it. This procedure not only provides important information in the staging and clinical care of breast cancer patients, but also provides a calibration of the reliability of the PET/CT reporting of IM lymph node metastasis. We consider the 80% identification of IM lymph node metastasis in PET/CT-positive nodes as confirmation of the excellent reliability of our PET/CT techniques and images. Our risk analysis suggests that certain factors may confer increased risk of IM nodal disease including LVI, high tumor grade, positive axillary lymph nodes, and ER-negative receptor status, PR-negative receptor status, and triple receptor-negative status. ecause the presence of IM nodal disease could alter not only staging but also therapy, radiologists should consider these risk factors during initial staging evaluation and should consider using 18 F-FDG PET/CT for the detection of IM node disease given its improved sensitivity and specificity over contrast-enhanced CT. Studies have shown that 18 F-FDG PET/CT detected extraaxillary lymph node involvement in almost one third of patients with stage II III breast cancer and that PET/CT findings upgraded the TNM stage in 17% of patients, which changed the radiation therapy fields in 12% [35]. Champion et al. [36] recently reported that 18 F-FDG PET/CT had a sensitivity of 93.6%, specificity of 85.4%, and accuracy of 92.1% for the detection of breast cancer recurrence in the setting of rising tumor marker levels. Although SUV max is a continuous variable and for each unit increase of SUV max, there is on average a 1.13 multiplicative increase of risk of having a positive IM node on 18 F-FDG PET/CT, we do not recommend cutoff values for SUV max because there are so many physical effects that can affect the PET image values. Cutoff values are not reliable until uniform standards are set for PET image calibration. The limitations of the study include that as a retrospective study we cannot determine how many negative IM nodes on 18 F-FDG PET/CT would have been pathologically positive and we cannot identify how many FNA-negative IM nodes would be positive for metastasis if the node had been excised and serially sectioned. Only 25% of our patients had biopsies of the 18 F-FDG PET/CT positive IM node, which is a small sample size; however, this series is still the largest one in the literature to date because of the risks inherent to biopsy including extension into the pleural space and adjacent vascular injury. This subset of patients were chosen for biopsy by their clinicians prospectively because a positive IM node would have been the only site of disease, and this finding would have upstaged disease and changed management. Many of the other patients had imaging or pathologic proof of stage IV disease or met the criteria established by the American Joint Committee on Cancer [30] for clinically detected IM node disease and were treated accordingly without pathologic proof. In conclusion, breast cancer patients with IM lymph node metastasis are known to have a worse prognosis than those without IM node metastasis. Visualization of FDG-avid IM nodes on 18 F-FDG PET/CT is best confirmed by biopsy. Our study shows that an FDG-avid IM node on PET/CT has a high likelihood of malignancy as confirmed by ultrasound-guided FNA. Risk factors for IM node positivity on 18 F-FDG PET/CT include high tumor grade, LVI, positive axillary nodes, and triple receptor negative receptor status. Identification of IM node metastasis frequently has a major impact on staging; prognosis; and therapy including surgical, chemotherapeutic, and radiation treatment decisions. References 1. Jemal A, Murray T, Ward E, et al. Cancer statistics, CA Cancer J Clin 2005; 55: Cody HS 3rd, Urban JA. Internal mammary node status: a major prognosticator in axillary node-negative breast cancer. Ann Surg Oncol 1995; 2: Sugg SL, Ferguson DJ, Posner MC, Heimann R. Should internal mammary nodes be sampled in the sentinel lymph node era? Ann Surg Oncol 2000; 7: Veronesi U, Cascinelli N, Greco M, et al. Prognosis of breast cancer patients after mastectomy and AJR:200, May

7 Wang et al. dissection of internal mammary nodes. Ann Surg human epidermal growth factor receptor 2 (HER2) vant chemotherapy of breast cancer. Am J Surg 1985; 202: among women with invasive breast cancer in Cali- 2009; 197: ; discussion, Veronesi U, Amone P, Veronesi P, et al. The value fornia, reast J 2009; 15: Donegan WL. The influence of untreated internal of radiotherapy on metastatic internal mammary 15. Veronesi U, Marubini E, Mariani L, Valagussa P, mammary metastases upon the course of mam- nodes in breast cancer: results on a large series. Zucali R. The dissection of internal mammary mary cancer. Cancer 1977; 39: Ann Oncol 2008; 19: nodes does not improve the survival of breast can- 27. Livingston SF, Arlen M. The extended extrapleural 6. Heuts EM, van der Ent FW, von Meyenfeldt MF, Voogd AC. Internal mammary lymph drainage and sentinel node biopsy in breast cancer: a study on 1008 patients. Eur J Surg Oncol 2009; 35: Eubank W, Mankoff DA, Takasugi J, et al. 18 Fluorodeoxyglucose positron emission tomography to detect mediastinal or internal mammary metastases in breast cancer. J Clin Oncol 2001; 19: ellon JR, Livingston R, Eubank W, et al. Evaluation of the internal mammary lymph nodes by FDG-PET in locally advanced breast cancer (LAC). Am J Clin Oncol 2004; 27: Centers of Disease Control and Prevention Website. International classification of diseases, ninth revision (ICD-9). htm. Updated September 1, Accessed January 24, Ellis IO, Elston CW. Histologic grade. In O Malley FP, Pinder SE, eds. reast pathology. Philadelphia, PA: Elsevier, 2006: Harvey JM, Clark GM, Osborne CK, Allred DC. Estrogen receptor status by immunohistochemistry is superior to the ligand-binding assay for predicting response to adjuvant endocrine therapy in breast cancer. J Clin Oncol 1999; 17: Mohsin SK, Weiss H, Havighurst T, et al. Progesterone receptor by immunohistochemistry and clinical outcome in breast cancer: a validation study. Mod Pathol 2004; 17: Wolff AC, Hammond ME, Schwartz JN, et al.; American Society of Clinical Oncology; College of American Pathologists. American Society of Clinical Oncology/College of American Pathologists guideline recommendations for human epidermal growth factor receptor 2 testing in breast cancer. J Clin Oncol 2007; 25: Parise CA, auer KR, rown MM, Caggiano V. reast cancer subtypes as defined by the estrogen receptor (ER), progesterone receptor (PR), and the cer patients: 30-year results of a randomized trial. Eur J Cancer 1999; 35: Madsen E, Gobardhan P, ongers V, et al. The impact on post-surgical treatment of sentinel lymph node biopsy of internal mammary lymph nodes in patients with breast cancer. Ann Surg Oncol 2007; 14: Estourgie SH, Nieweg OE, Olmos RA, Rutgers EJ, Kroon. Lymphatic drainage patterns from the breast. Ann Surg 2004; 239: Hindié E, Groheux D, Hennequin C, et al. Lymphoscintigraphy can select breast cancer patients for internal mammary chain radiotherapy. Int J Radiat Oncol iol Phys 2012; 83: van der Ent FW, Kengen RA, van der Pol HA, Povel JA, Stroeken HJ, Hoofwijk AG. Halsted revisited: internal mammary sentinel lymph node biopsy in breast cancer. Ann Surg 2001; 234: Chen RC, Lin NU, Golshan M, Harris JR, ellon JR. Internal mammary nodes in breast cancer: diagnosis and implications for patient management a systematic review. J Clin Oncol 2008; 26: Farrús, Vidal-Sicart S, Velasco M, et al. Incidence of internal mammary node metastases after a sentinel lymph node technique in breast cancer and its implication in the radiotherapy plan. Int J Radiat Oncol iol Phys 2004; 60: Dupont E, Cox CE, Nguyen K, et al. Utility of internal mammary lymph node removal when noted by intraoperative gamma probe detection. Ann Surg Oncol 2001; 8: Noushi F, Spillane AJ, Uren RF, Gebski V. Internal mammary lymph node metastasis in breast cancer: predictive models to assist with prognostic influence. reast 2011; 20: Segaert I, Mottaghy F, Ceyssens S, et al. Additional value of PET-CT in staging of clinical stage II and III breast cancer. reast J 2010; 16: eatty JD, Precht LM, Lowe K, Atwood M. Axillary-conserving surgery is facilitated by neoadju- radical mastectomy: its role in the treatment of carcinoma of the breast. Ann Surg 1974; 179: Veronesi U, Valagussa P. Inefficacy of internal mammary nodes dissection in breast cancer surgery. Cancer 1981; 47: Pugliese M, Shivaram G, Rogers J, Haseley D, uchanan C. PET-CT imaging in the initial management of high-risk breast cancer patients: who did it help? reast J 2009; 15: Edge S, yrd DR, Compton CC, Fritz AG, Greene FL, eds. AJCC cancer staging manual, 7th ed. New York, NY: Springer-Verlag, 2009: Heuts EM, van der Ent FW, Hulsewe KW, von Meyenfeldt MF, Voogd AC. Results of tailored treatment for breast cancer patients with internal mammary lymph node metastases. reast 2009; 18: Lacour J, ucalossi P, Cacers E, et al. Radical mastectomy versus radical mastectomy plus internal mammary dissection: five-year results of an international cooperative study. Cancer 1976; 37: Huang O, Wang L, Shen K, et al. reast cancer subpopulation with high risk of internal mammary lymph nodes metastasis: analysis of 2,269 Chinese breast cancer patients treated with extended radical mastectomy. reast Cancer Res Treat 2008; 107: Wang CL, MacDonald LR, Rogers JV, Aravkin A, Haseley DR, eatty JD. Positron emission mammography: correlation of estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2 status and 18 F-FDG. AJR 2011; 197:348; [web]w247 W Aukema TS, Straver ME, Peeters MJ, et al. Detection of extra-axillary lymph node involvement with FDG PET/CT in patients with stage II III breast cancer. Eur J Cancer 2010; 46: Champion L, rain E, Giraudet AL, et al. reast cancer recurrence diagnosis suspected on tumor marker rising. Cancer 2011; 117: AJR:200, May 2013