Alice P. Chung

Preoperative Breast MRI in the Surgical Treatment of Ductal Carcinoma In Situ

Abstract:  Accurate determination of the size or extent of ductal carcinoma in situ (DCIS) by imaging is uncertain, and incomplete resection of tumor results in involved margins in up to 81% of cases. This study examined the accuracy of magnetic resonance imaging (MRI) for assessment of DCIS size, and evaluated the effect of preoperative breast MRI on achievement of tumor‐free surgical margins after breast‐conserving surgery (BCS). One‐hundred and fifty‐eight female patients with DCIS were identified from a prospective database: 60 patients (62 cases) had preoperative breast MRI, and 98 patients did not have MRI. The accuracy of tumor size assessed by MRI was determined by comparison with histopathologic size. All patients underwent BCS initially. The rate of involved margins after resection was compared in MRI and no‐MRI groups. The overall correlation between MRI size and histopathologic size was high (p < 0.0001). MRI assessment of size was significantly more accurate when DCIS was high grade (p < 0.0001) or intermediate grade (p = 0.005) versus low grade (p = 0.187). The rate of tumor‐involved margins was not significantly different in MRI and no‐MRI groups (30.7% and 24.7%, respectively; p = 0.414). The rate of mastectomy was significantly higher in the MRI group than the no‐MRI group (17.7% versus 4.1%; p = 0.004). These findings indicate that MRI can detect DCIS, especially when lesions are high or intermediate grade, but that MRI does not accurately predict the size of DCIS. In this study, MRI did not improve the surgeon’s ability to achieve clear margins following BCS.

Long-term follow-up confirms the oncologic safety of sentinel node biopsy without axillary dissection in node-negative breast cancer patients.

The sentinel node biopsy (SNB) technique for breast cancer was introduced in the early 1990s as a less invasive means of axillary staging than standard axillary dissection (ALND). There have been a multitude of studies performed to evaluate the accuracy of SNB, and a recent meta-analysis demonstrated that the status of the sentinel lymph node (SLN) can accurately predict the status of the axillary basin with an average false-negative rate of 7.3%. Experienced surgeons achieve even lower false-negative rates, ranging from 0% to 5%. Before definitive evidence from randomized controlled trials was available to demonstrate its long-term safety, SNB was rapidly accepted by the surgical and oncological communities. This occurred because the procedure proved to be an accurate diagnostic test to assess the axillary lymph nodes. Each surgeon could determine his or her own accuracy by performing SNB and comparing the results to ALND in the same patient. A randomized trial is not necessary to test the accuracy of a diagnostic test, and SNB has now replaced ALND as the standard of care for axillary staging in patients with early breast cancer. A randomized trial is necessary to determine the safety and therapeutic implications of a new procedure. A number of randomized trials specifically addressed shortand long-term morbidity of SNB compared with ALND. These trials showed that the morbidity of SNB is less than that of ALND. However, there are only a few studies that report data on the long-term outcome of SNB, and the follow-up in these reports is limited. In this issue of Annals of Surgery the study by the excellent group led by Veronesi, is the first randomized trial to report 10-year follow-up for SNB with or without routine completion ALND. The trial was a single institution study involving 516 patients with breast tumors of 2 cm or less in size, who were randomly assigned to either SNB followed by ALND or SNB with ALND only if the SLN contained metastatic disease. The initial results from this trial were published in 2003, where the efficacy and safety of the SNB procedure were validated. In their earlier report, they demonstrated the false-negative rate of the SNB procedure to be 8.8%, with an overall accuracy of 96.9%. They found that there was less pain and greater arm mobility in patients who had SNB alone compared with those who had ALND. These results are similar to the findings reported from other randomized trials that specifically evaluated efficacy and side effects. With a mean follow-up of 46 months, the 2003 report demonstrated that short-term survival was not compromised in those who had a negative SNB without ALND. The authors then provided an update on this trial in 2006, where after a median follow-up of 79 months, there were 34 breast cancer-related events, 16 in the SNB arm with one axillary recurrence and 18 in the ALND arm with no axillary recurrences (P 0.6). The 5-year overall survival was 98.4% in the SNB group and 96.4% in the ALND group. These findings are similar to results of the other randomized trials with shorter follow-up that evaluated the disease-free and overall survival of SNB with and without ALND. Zavagno et al conducted a multicenter trial of 697 patients with tumors less than or equal to 3 cm in diameter. After a mean of 56 months of follow-up, there were 29 locoregional events in the SNB group and 22 events in the ALND group. They found no difference between the 2 groups with respect to disease-free or overall survival. Canavese et alcompared SNB with and without ALND in a much smaller cohort of 115 patients followed up for a median of 66 months. There were 12 events in the ALND arm with one axillary recurrence and 10 events in the SNB arm with no axillary recurrences. The authors did not demonstrate a difference in disease-free or overall survival between the 2 groups. In the current report by Veronesi et al, an update of their randomized trial is provided with a median follow-up of 102 months. There were a total of 49 breast cancer-related events, 23 in the SNB arm and 26 in the ALND arm (P 0.52). There was no difference between the 2 groups with respect to disease-free survival (89.9% in the SNB arm vs. 88.8% in the ALND

Comparison of Outcomes of Breast Conserving Therapy in Multifocal and Unifocal Invasive Breast Cancer

BACKGROUND There is controversy about whether breast conserving therapy (BCT) should be contraindicated in multifocal (MF) breast cancer. Few studies have reported on the oncologic safety of BCT in MF breast cancer. STUDY DESIGN We reviewed a prospective database of 1,169 women with invasive breast cancer who were treated with segmentectomy and whole breast irradiation from 1991 through 2009 and followed at our institution. Multifocal breast cancer was defined as 2 or more distinct tumors excised with a single incision or segmentectomy. We compared 2 groups, MF and unifocal breast cancer patients, with respect to demographics, tumor characteristics, adjuvant systemic therapy, local recurrence (LR), disease-free survival (DFS), and overall survival (OS). RESULTS One hundred sixty-four patients with MF and 999 with unifocal invasive breast cancer were treated with BCT. Median follow-up was 112 months. Compared with the unifocal group, patients in the MF group had higher 10-year LR (0.6% vs 6.1%, p < 0.001) and lower 10-year DFS (97.7% vs 89.3%, p < 0.001) and OS (98.4% vs 85.8%, p < 0.001). On multivariable analysis, multifocality was independently significantly associated with local recurrence-free survival (LRFS), DFS, and OS. CONCLUSIONS Our data suggest that BCT in MF breast cancer is oncologically safe but may result in a slightly inferior outcome compared with BCT in unifocal breast cancer.

Reducing Complications and Margin Issues with Nipple-Sparing Mastectomy

Nipple-sparing mastectomy (NSM) followed by immediate reconstruction is associated with low morbidity and mortality rates. The most concerning complication is skin and nipple necrosis which can lead to patient dissatisfaction and ultimately loss of reconstruction. Patient selection plays a key role in minimizing complications. This chapter describes the most common complications associated with NSM and patient as well as technical factors that may be predictive of these complications. The various techniques for minimizing and managing nipple or flap necrosis are discussed. In addition, positive margin status after NSM presents concerns regarding need for reoperation with possible removal of the NAC as well as risk for local recurrence. This chapter will address incidence of close or positive margins following NSM, techniques to accurately assess retro-areolar margin status, management of positive margins, the potential role of post-mastectomy radiation in patients with close or positive margins and rates of local recurrence following NSM.

The basal phenotype as a clinically relevant indicator of trastuzumab resistance in HER2+ breast cancer.

154 Background: Trastuzumab (Herceptin) resistance remains a clinical challenge but the mechanism is not well understood. Recently, studies have identified a subset of Her2+ breast cancer called basal HER2 that expresses basal genes. We investigated the effect of basal gene expression on Herceptin response in HER2+ breast cancer cell lines and on prognosis in HER2+ breast cancer patients. METHODS Non-basal (BT474, SKBR3) and basal (HCC1569, HCC1954, JIMT-1) HER2+ cell lines were chosen based on basal cytokeratin expression. Cell proliferation was assessed after treatment with vehicle, Herceptin (H), Paclitaxel (P), H+P, Akt Inhibitor (AI), and H+AI. HER2 signaling was examined using immunoblotting of p-Akt and p-ERK. Because breast cancer stem cells (BCSC) are linked to basal breast tumors and treatment resistance, we assessed BCSC activity using mammosphere formation and aldehyde dehydrogenase (ALDH) positivity. Immunohistochemical staining of HER2+ breast cancers for basal markers CK5/6, CK14, and EGFR was correlated with clinicopathologic features and survival in 88 patients with Stage 1-3 HER2+ breast cancer treated with Herceptin. RESULTS Basal HER2 cells were resistant to Herceptin compared to non-basal HER2 cells but this resistance was overcome by Akt inhibition. Immunoblotting showed that non-basal HER2 cells had decreased p-Akt after Herceptin treatment which was not seen in basal HER2 cells. There was no difference in p-ERK levels after Herceptin therapy in all cell lines. Basal HER2 cells had increased mammosphere formation and ALDH positivity suggesting higher stem cell activity compared to non-basal HER2 cells. Of the HER2+ patients, 33/88 (37.5%) expressed at least one basal marker. Basal Her2 tumors were associated with higher grade (p = 0.04) and more ER/PR negativity (p < 0.01). CK14 expression correlated with worse overall survival by log-rank test (p = 0.02), while EGFR showed a similar trend (p = 0.06). CONCLUSIONS Basal HER2 breast cancer cell lines have Herceptin resistance which may be due to constitutively active Akt signaling and increased stem cell activity. Clinically, basal marker expression predicts Herceptin resistance and worse outcomes in HER2+ breast cancer patients.

Papillary lesions of the breast: Can larger core needle biopsy samples identify patients who may avoid surgical excision?

44 Background: Papillary lesions of the breast are frequently diagnosed on core needle biopsy (CNB). The ability to distinguish benign from atypical/malignant papillary lesions is limited by the representative nature of the biopsy method; thus follow-up excision is usually recommended. We aimed to determine if larger CNB samples can more reliably predict the true benign nature of a papillary lesion, thereby sparing certain patients a formal surgical excision. METHODS We reviewed medical records of 53 female patients diagnosed with histologically benign papillary lesions on CNB from 2000 to 2010, who subsequently underwent surgical excision. Pathology slides of the CNB were reviewed to document the benign histologic features of the papilloma, the number of cores sampled and the area of tissue biopsied (mm2). Statistical analysis was performed to identify the characteristics of the CNB that were associated with retention of benign histology on excision. RESULTS Atypical ductal hyperplasia (ADH) and carcinoma were identified in 6% (3/53) and 8% (4/53) of papillary lesions, respectively, when excised. Clinical and radiographic characteristics did not distinguish the ADH/malignant lesions from benign papillomas. The CNB needle sizes ranged from 9- to 18-gauge (median 14). The number of cores sampled ranged from 3-16 (mean 4.5). Patients with benign excisions had a significantly larger area of tissue sampled by CNB than those found to have ADH/malignant lesions on excision (mean ± SD: 95.6 ± 101.2 vs. 41.7 ± 21.9, respectively, p=0.003). By logistic regression, CNB tissue samples consisting of ≥7 cores, or measuring >96 mm2 in aggregate, had a negative predictive value for ADH/malignancy of 100% (AUC of 0.69 and 0.68, respectively). CONCLUSIONS Although no clinical or radiologic features distinguished benign from pathologically significant papillary lesions, larger sample sizes significantly improved the predictive value of benign histology on CNB. A papilloma sampled by ≥ 7 cores or > 96 mm2 showing benign histology at CNB, retained benign features upon excision. Close surveillance may be a reasonable option for patients whose benign papillomas are generously sampled at the time of CNB.