Donna Plecha, MD 1, Shiyu Bai, BS 2, Helen Patterson 3, Cheryl Thompson, PhD 4, and Robert Shenk, MD 5

Ann Surg Oncol DOI 10.1245/s10434-015-4527-y ORIGINAL ARTICLE BREAST ONCOLOGY Improving the Accuracy of Axillary Lymph Node Surgery in Breast Cancer with Ultrasound-Guided Wire Localization of Biopsy Proven Metastatic Lymph Nodes Donna Plecha, MD 1, Shiyu Bai, BS 2, Helen Patterson 3, Cheryl Thompson, PhD 4, and Robert Shenk, MD 5 1 Department of Radiology, University Hospitals Case Medical Center, Cleveland; 2 Case Western University School of Medicine, Cleveland; 3 Amherst College, Amherst; 4 Case Western Reserve University, Cleveland; 5 Department of Surgery, University Hospitals Case Medical Center, Cleveland ABSTRACT Background. This study aimed to evaluate whether the use of preoperative ultrasound (US)-guided wire localization of metastatic axillary lymph nodes (LN) assessed previously by core needle biopsy (CNB) and clip placement in breast cancer patients improves successful surgical removal. Methods. A retrospective review examined breast cancer patients who underwent US-guided CNB of an axillary LN and biopsy clip placement as well as axillary lymph node dissection (ALND) or sentinel node lymph node biopsy (SLNB) from 1 January 2010 to 30 September 2013. Preoperative needle localization status, neoadjuvant chemotherapy, and type of axillary LN surgery were reviewed. Confirmation that the metastatic LN had been surgically removed was determined on the specimen image, by pathologic report confirmation, or by pre-radiation therapy computed tomography (CT) scan. Results. Preoperative US-guided needle localization was performed for 68.2 % (73/107) of the patients, with 97.3 % (n = 71) demonstrating confirmation of biopsy clip and LN removal versus 79.4 % (n = 27) of the 34 patients showing no performance of needle localization (p = 0.0043). Subgroup analysis showed a significant difference in removal of metastatic LN between the patients who received neoadjuvant chemotherapy [97 % of LNs removed with wire localization (n = 65/67) vs. 83.3 % of LNs removed without wire localization (n = 20/24; p = 0.04)] Ó Society of Surgical Oncology 2015 First Received: 5 June 2014 D. Plecha, MD e-mail: donna.plecha@uhhospitals.org and the patients who had ALND, [96.3 % of LNs removed with wire localization (n = 52/54) vs. 77.8 % of LNs removed without wire localization (n = 21/27; p = 0.015)]. Conclusion. US-guided wire localization of metastatic axillary LNs that have had biopsy with clip placement significantly improves the success rate of surgical removal, allowing more accurate staging and decreasing the falsenegative rates of SLNB after neoadjuvant therapy. Identifying the presence of axillary lymph node (LN) metastasis in patients with invasive breast cancer is a critical prognostic factor that guides local, regional, and systemic treatment decisionsc. LN involvement can be reliably determined preoperatively with focused ultrasound (US) and US-guided biopsy techniques. 1 3 This is one of the factors used in deciding whether to use neoadjuvant therapy. Neoadjuvant therapy is used with increasing frequency and has several advantages. It can reduce the size of large primary tumors, improving the chances for a lumpectomy. As opposed to adjuvant treatment, it allows observation of the response to chemotherapy. A pathologic complete response (pcr) to neoadjuvant chemotherapy is a prognostic factor that occurs for 30 40 % of patients. 4 Among HER2neu-positive patients receiving pertusamab or lapitinab with herceptin, the cpr is as high as 60 %. 5 Among patients who would have undergone an axillary dissection, such as those with grossly involved LNs or those who had a mastectomy with a positive node, neoadjuvant therapy may allow an axillary dissection to be avoided if a pcr was obtained. This would decrease the risk of associated morbidities such as lymphedema, pain, numbness, and decreased arm mobility. 6

D. Plecha et al. Currently, the standard of care for patients with preoperatively determined metastatic LNs after neoadjuvant therapy is an axillary dissection, partly because sentinel node biopsy (SLNB) is not as accurate after neoadjuvant therapy. 7 A possible reason for this is blocked lymphatics from tumor or fibrosis of the positive nodes as a response to chemotherapy. Ability to identify and remove the nodes that were positive preoperatively is critical in measuring the response to therapy and should decrease the falsenegative rate of SLNB. Accurate SLNB would allow it to become standard treatment after neoadjuvant therapy. To ensure that a complete response has occurred, the known positive node needs to be removed. It is assumed that this node would be removed with an axillary dissection after preoperative chemotherapy. Placement of clips in nodes after core needle biopsy (CNB) has shown that the biopsied LN was not always removed after surgery. The missed clips were identified on planning computed tomography (CT) scans for radiation therapy (Fig. 1). Missed nodes were found even after axillary dissection. Missing a positive node could lead to undertreatment of the axilla, possibly affecting regional recurrence. The patient with a positive node may be eligible for further systemic treatment, so accurate assessment of the axilla after neoadjuvant therapy is important for systemic recurrence as well. Because of this, we started using wire localization of the preoperatively biopsied positive node to ensure its removal. We examined the use of US-guided wire localization of biopsy proven metastatic axillary LNs marked with a clip to improve success of surgical removal. MATERIALS AND METHODS A retrospective review was performed after approval by the University Hospitals Case Medical Center Institutional Review Board. Our study was performed in compliance with the Health Insurance Portability and Accountability Act. Study data were collected and managed using RED- Cap electronic data capture tools (Vaderbilt University, Nashville, TN) hosted at University Hospitals of Cleveland Case Medical Center. 8 A retrospective medical record review of breast cancer patients from 1 January 2010 to 30 September 2013 who underwent US-guided CNB of an axillary LN with placement of a biopsy marker was performed. Our institution routinely scans the ipsilateral axilla in patients highly suspected to have malignancy with a breast imaging-reporting and data system (BI-RADS) category 4 or 5 lesion(s). Radiologists with 3 20 years of experience prospectively evaluated the axillary LNs and determined whether they were suspicious enough to warrant biopsy based on cortical thickening and previously described morphology characteristics. 2,3 FIG. 1 Status of patient with a history of invasive ductal carcinoma grade 3 after preoperative chemotherapy and right ALND, with 14 LNs removed at surgery. Axial CT scan image performed for radiation therapy planning purposes demonstrates right axillary LN clip remaining after surgery (arrow) Biopsies were performed by radiologists using a nonthrowing 14-gauge Vacora needle (C.R. Bard, Inc., Tempe, AZ, USA). One or two samples per LN were obtained. Specimens were closely inspected in a search for nonfatty tissue with a solid tissue component. The specimens were interpreted using hematoxylin-eosin staining by pathologists, and the pathology reports were reviewed. The vast majority of the tissue markers (98 %, 105/107) placed were HydroMARK breast biopsy-site markers (Biopsy Sciences, Clearwater, FL, USA). All wire localizations were performed using US guidance and a 7-cm 20-gauge needle with a 20-cm GHIATAS Beaded Breast Localization Wire (C.R. Bard, Inc.). All patients were treated by three surgeons specializing in breast surgery, each of whom had more than 20 years of experience. All the breast surgeons performing SLNB had more than 10 years of experience. Technetium-99 m microfiltrated sulfur colloid was injected preoperatively into all the patients undergoing SLNB. Isosulfan blue dye was used in addition by surgeon preference if poor radioactive uptake was detected in the axilla of patients in the operating room. Patients who had an axillary LN CNB consistent with metastatic disease underwent clip placement at the time of biopsy, and those who received axillary lymph node dissection (ALND) or SLNB surgery were included in our study. A review of medical records provided demographic information, body mass index (BMI), tumor size, tumor grade, tumor histology, receptor status, clinical and pathologic stage, final pathology, and number of LNs involved at surgery. Preoperative wire localization status, neoadjuvant chemotherapy, type of axillary LN surgery, surgical specimen imaging result, surgical pathologic report, and preradiation oncology planning CT scan images were reviewed. Surgical removal of the LN marked with a biopsy clip was confirmed on the surgical specimen image

Improving the Accuracy of Axillary LN Surgery in Breast Cancer or pathologic report, whereas absence of the clip was confirmed on the preradiation therapy CT scan. Statistical Analysis A Fisher s exact test was used to test the hypothesis that the success rate for surgical removal of the biopsied and marked axillary LN was higher in the group that underwent needle localization than in the group that did not. A twotailed unpaired Student s t test was used to compare the average number of axillary LNs removed between the different cohorts. The threshold for statistical significance was set at a p value lower than 0.05. RESULTS The inclusion criteria was met by 107 patients. The mean age of our patient population was 54.9 years. In 80 % of our patients, the disease was stage 2B or higher, and for 84.1 % of our patients, the diagnosis was invasive ductal carcinoma. Our patient population is summarized in Table 1. The average BMI was 28 kg/m 2. The BMI did not differ between the group that had the clip successfully removed (BMI = 27.9 kg/m 2 ) and the group that did not have the clip removed (BMI = 28.2 kg/m 2 ) nor between the ALND group (BMI = 28.1 kg/m 2 ) and the SLNB group (BMI = 27.2 kg/m 2 ). All the patients who did not receive neoadjuvant chemotherapy had ALND for a biopsy-proven metastatic axillary LN. In the neoadjuvant group, 34 patients (37.4 %) received ALND and 57 patients (62.6 %) had SLNB. Of these patients, 54.4 % were converted to ALND either because of failure to find the sentinel node or due to a positive sentinel node. The total number of axillary LNs removed with ALND or SLNB did not differ between the patients who had wire localization and those who did not (Table 2). Table 3 presents the details for the types of axillary LN surgery and shows the neoadjuvant chemotherapy status of the patient cohort. Preoperative US-guided wire localization was performed for 68.2 % of the patients (n = 73). The biopsy clip was removed in 97.3 % of the patients (n = 71), a rate significantly higher than 79.4 % of the 34 patients (n = 27) who did not have wire localization (p = 0.0043). Neoadjuvant chemotherapy was administered to 91 patients. The patients who underwent a US-guided wire localization had a 97 % (65/67) success rate of clip/ln removal versus 83.3 % for the non-wire localization group (n = 20/24; p = 0.04). In the group that did not receive neoadjuvant chemotherapy, six patients underwent wire localization with TABLE 1 Characteristics of the patient population Mean age: years (range) 54.9 (49 89) Breast cancer stage n Patients (%) 1A 1 0.9 1B 1 0.9 2A 19 17.8 2B 46 43.0 3A 26 24.3 3B 5 4.7 3C 6 5.6 4 2 1.9 Unknown 1 0.9 Primary cancer type Invasive ductal carcinoma 90 84.1 Invasive lobular carcinoma 12 11.2 Invasive mammary carcinoma 4 3.7 Other a 1 0.9 Average BMI (kg/m 2 ) 28 BMI range Normal (18.5 25) 32 29.9 Overweight (25 30) 42 39.3 Class 1 obesity (30 35) 22 20.6 Class 2 obesity (35 40) 8 7.5 Class 3 obesity ([40) 3 2.8 BMI body mass index a Adenocarcinoma of the LN, no primary breast tumor found 100 % clip/ln removal compared with 70 % (n = 7) of those who did not have needle localization (n = 10; p = 0.25). Among the patients who underwent ALND (n = 81), 54 underwent wire localization, with a 96.3 % (n = 52) success rate of clip/ln removal, versus 77.8 % of the patients (n = 21) who did not receive wire localization (n = 27, p = 0.015). In the SLNB group (n = 26), the success rate of clip/ln removal in the wire localization group (n = 19) was 100 %, compared with 86 % (n = 6) in the group that did not have wire localization (n = 7; p = 0.27). The results are summarized in Table 4. DISCUSSION Axillary LN, an important prognostic and staging factor in breast cancer patients, is used to guide surgical, chemotherapy, and radiation treatment decisions. Findings have shown US-guided CNB to be a reliable way to document ipsilateral axillary LN involvement in breast cancer patients. 1 3 Whether US core biopsies should be performed for patients with clinically negative LNs has been controversial. In the American College of Surgeons Oncology Group

D. Plecha et al. TABLE 2 Total number of axillary lymph nodes (LNs) removed and type of axillary LN surgery Type of axillary LN surgery (no. of patients) Wire localization status (no. of patients) Average number of axillary LNs removed ALND (81) Wire localization (54) 16.85 Non-wire localization (27) 13.8 p = 0.092 Overall mean number of LNs removed in the ALND group (15 ± 7.5) SLNB (26) Wire localization (19) 5.7 Non-wire localization (7) 4.7 p = 0.36 Overall mean number of LNs removed in the SLNB group (5.5 ± 2.5) ALND axillary lymph node dissection, SLNB sentinel node lymph node biopsy TABLE 3 Types of axillary LN surgery Type of surgery No neoadjuvant chemotherapy Neoadjuvant chemotherapy SLNB only 0 26 26 ALND only 16 34 50 SLNB converted 0 31 31 to ALND Total 16 91 Total SLNB sentinel node lymph node biopsy, ALND axillary lymph node dissection (ACOSOG) Z11 trial, patients treated with breast conservation who had only one or two positive LNs not grossly involved did not have improved survival or less local regional recurrence if they underwent ALND. 9 The concern about a preoperative core biopsy with positive results is that it commits patients to an axillary dissection they would have avoided if they had received a SLNB without neoadjuvant therapy. Studies currently are investigating not performing axillary dissection for patients with positive nodes after neoadjuvant therapy. US core biopsies help to stage patients who would potentially benefit from neoadjuvant therapy such as those with grossly involved nodes, large tumors, having mastectomies, as well as HER2neu patients. Once LN involvement is documented, surgical removal of the LN biopsy proven to have metastatic involvement is important. In our study, the most accurate way to ensure removal was to localize the positive node by wire at the time of surgery. Overall, we found that 20.6 % of the patients did not have documented removal of the biopsy-proven metastatic LN without wire localization. Only 2.7 % were missed if wire localization was used (p \ 0.0043). Even among patients who had ALND, 22.2 % of biopsied LNs were not removed without wire localization compared with 3.7 % when wire localization was used (p = 0.015). Currently, the standard of care for patients receiving preoperative chemotherapy for a known positive LN is ALND because of the high false-negative rate for SLNB. The rate for missing the biopsied positive node at surgery in this study seems high. A recent abstract examined a subset of patients from the ACOSOG Z1071 trial who had biopsy-proven metastatic LNs clipped with a biopsy marker and underwent preoperative chemotherapy. The rate for missing the biopsy-proven clipped LN was similarly high, at 17 %, for patients undergoing ALND. 10 After neoadjuvant chemotherapy, it may be challenging to find metastatic nodes during axillary surgery. Some of the reasons include shrinking of the LN after therapy, becoming fibrotic, and not taking up the radiocolloid or dye for an SLNB (Fig. 2). In the current study, the miss rate was not due to differences in the number of LNs removed during axillary surgery between the patients receiving wire localization and those who did not (Table 2). We do not think the high miss rate was due to misplaced clips because they were documented to be in the nodes preoperatively. Potentially, a clip could be dislodged from a node and left in the wound, but when this happens during a lumpectomy, our experience shows that the clip is removed with irrigation and suction. Also, the clips seen on CT scans postoperatively appear to be in LNs. Missed positive nodes account for the high false-negative rate of SLNB after neodjuvant therapy among patients with positive nodes preoperatively. The biopsy-proven positive LN before neodjuvant therapy is most likely to be the one with residual disease because US core biopsies usually are directed at the largest most abnormal node. It can be considered a sentinel node, certainly, when it is the only abnormal node or when only a few nodes are positive. Identifying the preoperatively positive node after neoadjuvant therapy should improve the accuracy of SLNB. The ACOSOG Z1071 trial investigating patients with cn1 disease who underwent chemotherapy followed by both SLNB and ALND found a false-negative rate of 12.6 % for the SLNB procedure. 7 This rate was determined using ALND to define the true-positive rate. Because in our study the positive node was not found in 22.2 % of the

Improving the Accuracy of Axillary LN Surgery in Breast Cancer TABLE 4 Success rates for surgical removal of previously biopsied metastatic axillary LNs in different patient cohorts Patient Cohorts n Wire localization n (%) Non-wire localization n (%) Statistical significance (p value) Neoadjuvant chemotherapy 91 65/67 (97.0) 20/24 (83.3) 0.04 No neoadjuvant chemotherapy 16 6/6 (100) 7/10 (70.0) NS SLNB 26 19/19 (100) 6/7 (85.7) NS ALND 81 52/54 (96.3) 21/27 (77.8) 0.015 Overall 107 71/73 (97.3) 27/34 (79.4) 0.0043 NS not significant, SLNB sentinel node lymph node biopsy, ALND axillary lymph node dissection time of core biopsy and performing wire localization with US guidance at the time of surgery significantly improved the ability to remove the known positive node. This increased removal will improve the accuracy of SLNB after neoadjuvant therapy. Overall, LN removal was successful for 97.3 % of the patients who had needle localizaion. When the patients were subgrouped into those who received neoadjuvant therapy, no neoadjuvant therapy, ALND, or SLNB, preoperative wire localization resulted in successful LN removal for 96.3 100 % of the patients (Table 4). Wire localization always gave a higher rate of positive node removal although this was not always statistically significant. This may have been due to the small number of patients in each subgroup. The biopsy proven metastatic node is a sentinel node, and removing it allows better assessment of response to chemotherapy and should decrease the false-negative rate of SLNB after neoadjuvant therapy. This would allow SLNB to replace ALND as the standard of care for the axilla after neoadjuvant therapy. Because 30 60 % of patients receiving preoperative therapy have a pathologic complete response, 4,5 accurate SLNB with wire localization would result in those patients avoiding ALND. CONCLUSION FIG. 2 a US of metastatic abnormal axillary LN with loss of the normal fatty hilum and cortical thickening (white arrow). b US of the same LN after 6 months of chemotherapy. The LN is not visualized. Only the remaining clip is detectable for needle localization (white arrow) patients who had an ALND without wire localization, the true false-negative rate after neoadjuvant therapy may be higher than 12.6 %. According to a recent abstract from the ACOSOG Z1071 trial, when the clipped LN was removed during SLNB, the false-negative rate fell to 6.6 %. 10 We found that marking biopsy-proven metastatic LNs at the US-guided wire localization of biopsy proven metastatic axillary LNs allows more accurate assessment of the axilla at the time of definitive surgery. Preoperative wire localization with ALND would increase the accuracy of staging and evaluation of response to chemotherapy. Wire localization with SLNB after neoadjuvant therapy should decrease the false-negative rate, potentially approaching the rate for patients with clinically negative LNs and allowing SLNB to be the standard after neoadjuvant therapy. Positive LNs, biopsy proven by US core preoperatively, should be marked with a clip and localized by wire at the time of definitive axillary surgery. CONFLICT OF INTEREST There are no conflict of interest.

D. Plecha et al. REFERENCES 1. Abe H, Schacht D, Sennett CA, Newstead GM, Schmidt RA. Utility of preoperative ultrasound for predicting pn2 or higher stage axillary lymph node involvement in patients with newly diagnosed breast cancer. Am J Roentgenol. 2013;200:696 702. 2. Abe H, Schmidt RA, Kulkarni K, Sennett CA, Mueller JS, Newstead GM. Axillary lymph nodes suspicious for breast cancer metastasis: sampling with US-guided 14-gauge core-needle biopsy clinical experience in 100 patients. Radiology. 2009; 250:41 9. 3. Abe H, Schmidt RA, Sennett CA, Shimauchi A, Newstead GM. US-guided core needle biopsy of axillary lymph nodes in patients with breast cancer: why and how to do it. Radiographics. 2007;27(Suppl 1):S91 9. 4. Hennessy BT, Hortobagyi GN, Rouzier R, et al. Outcome after pathologic complete eradication of cytologically proven breast cancer axillary node metastases following primary chemotherapy. J Clin Oncol. 2005;23:9304 11. 5. Schneeweiss A, Chia S, Hickish T, et al. Pertuzumab plus trastuzumab in combination with standard neoadjuvant anthracycline-containing and anthracycline-free chemotherapy regimens in patients with HER2-positive early breast cancer: a randomized phase II cardiac safety study (TRYPHAENA). Ann Oncol. 2013;24:2278 84. 6. Veronesi U, Paganelli G, Viale G, et al. A randomized comparison of sentinel-node biopsy with routine axillary dissection in breast cancer. N Engl J Med. 2003;349:546 53. 7. Boughey JC, Suman VJ, Mittendorf EA, et al. Sentinel lymph node surgery after neoadjuvant chemotherapy in patients with node-positive breast cancer: the ACOSOG Z1071 (Alliance) clinical trial. JAMA. 2013;310:1455 61. 8. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap): a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377 81. 9. Giuliano AE, Hunt KK, Ballman KV, et al. Axillary dissection vs no axillary dissection in women with invasive breast cancer and sentinel node metastasis: a randomized clinical trial. JAMA. 2011;305:569 75. 10. Boughey JC, Ballman KV, Symmans WF, McCall LM, Mittendorf EA, Ahrendt GM, Wilke LG, Taback B, Hunt KK. Methods impacting the false negative rate of sentinel lymph node surgery in patients presenting with node positive breast cancer (T0 T4, N1 2) who receive neoadjuvant chemotherapy: results from a prospective trial ACOSOG Z1071 (Alliance). Poster session presented at San Antonio Breast Cancer Symposium, 9 13 December 2014, San Antonio, TX.