Fax: (713) 745-1462
Low locoregional failure rates in selected breast cancer patients with tumor-positive sentinel lymph nodes who do not undergo completion axillary dissection†
Article first published online: 22 JUN 2007
Copyright © 2007 American Cancer Society
Volume 110, Issue 4, pages 723–730, 15 August 2007
How to Cite
Hwang, R. F., Gonzalez-Angulo, A. M., Yi, M., Buchholz, T. A., Meric-Bernstam, F., Kuerer, H. M., Babiera, G. V., Tereffe, W., Liu, D. D. and Hunt, K. K. (2007), Low locoregional failure rates in selected breast cancer patients with tumor-positive sentinel lymph nodes who do not undergo completion axillary dissection. Cancer, 110: 723–730. doi: 10.1002/cncr.22847
Presented in part at the 28th Annual San Antonio Breast Cancer Symposium, San Antonio, Texas, December 8–11, 2005.
- Issue published online: 2 AUG 2007
- Article first published online: 22 JUN 2007
- Manuscript Accepted: 27 APR 2007
- Manuscript Revised: 20 APR 2007
- Manuscript Received: 19 MAR 2007
- breast cancer;
- sentinel lymph node biopsy;
- lymph node dissection;
The role for completion axillary dissection (CLND) in patients with breast cancer who have tumor-positive sentinel lymph nodes (SLN) has been questioned. The objective of this study was to examine the long-term safety of avoiding CLND in selected patients with positive SLNs.
Patients with invasive breast cancer who underwent SLN biopsy at the authors' institution between 1993 and July 2005 were reviewed. Of 3366 total patients, 750 patients had a positive SLN. There were 196 patients with a positive SLN who did not undergo CLND based on clinician and patient preference. Clinicopathologic variables and treatment patterns were analyzed along with locoregional, distant recurrence, and survival.
Most tumors were infiltrating ductal carcinomas (74%), estrogen receptor-positive tumors (82%), progesterone receptor-positive tumors (70%), HER-2/neu-negative tumors (78.6%), and tumors were classified predominantly as either T1 or T2 (95.4%). The median number of SLNs removed was 3, and the median number of positive SLNs was 1. The median size of the tumor deposit in the SLN was 1.0 mm (range, 0.1–12.9 mm). Most SLNs were positive by on hematoxylin and eosin staining (64.3%), whereas 35.7% of SLNs were positive only by immunohistochemistry. Most patients underwent breast conservation (68.9%), radiation (58.2%), and chemotherapy (neoadjuvant in 14.3%, adjuvant in 55.6%). With a median follow-up of 29.5 months, no patients had an axillary recurrence, 1 patient had a supraclavicular lymph node recurrence, and 3 patients developed distant metastases. The median time to recurrence was 32 months.
In selected patients who had positive SLNs, the locoregional failure rate was low without CLND. Prospective studies will be valuable to corroborate these results and to refine further the optimal selection criteria for this approach. Cancer 2007. © 2007 American Cancer Society.
The advent of sentinel lymph node (SLN) biopsy (SLNB) for patients with cancer has made a significant impact on the field of surgical oncology.1, 2 For patients with breast cancer, SLNB affords decreased morbidity compared with axillary dissection,3, 4 including a lower rate of lymphedema, seromas, pain, and sensory changes.5 More detailed analysis of the SLN increases the accuracy of staging,6, 7 and this has been adopted into the most recent edition of the American Joint Committee on Cancer (AJCC) staging handbook.8, 9
Current guidelines from the American Society of Clinical Oncology recommend completion axillary lymph node dissection (CLND) for patients who have metastases (>0.2 mm) identified in an SLN by any method of detection.10 However, CLND potentially may be beneficial for only half of those patients, because, in 40% to 60% of patients, the SLN is the only positive lymph node.11 Studies from our group as well as others have indicated that the primary tumor size, the presence of lymphovascular invasion (LVI), the size of the SLN metastasis, and the number of positive SLNs are associated with an increased risk of having positive nonsentinel lymph nodes (NSLNs).12–18 In our previous study, the number of SLNs harvested also was related inversely to the risk of additional positive NSLNs.16 Van Zee and colleagues at Memorial Sloan-Kettering Cancer Center have developed a web-based nomogram to help clinicians predict a patient's risk of additional positive lymph nodes19 that has been validated by other institutions.20, 21 The increasing use of this nomogram and other models will influence clinicians in their recommendations for CLND. Furthermore, widespread use of chemotherapy, radiation therapy (RT), and endocrine therapy for many patients also may diminish the added benefit of CLND.
The outcome of patients with a positive SLN who do not undergo CLND is unclear. In the current study, we analyzed our prospective database of patients with breast cancer who had SLN metastases but who declined CLND to determine their locoregional recurrence and survival.
MATERIALS AND METHODS
We reviewed the records of patients who underwent SLNB for invasive breast cancer who were included in the prospective database at The University of Texas M. D. Anderson Cancer Center (UT-MDACC). Approval was obtained from the Institutional Review Board prior to conducting this study. Between January 1993 and July 2005, 3366 patients underwent SLNB for invasive breast cancer. We identified the patients within that group who had a positive SLN identified either by hematoxylin and eosin (H&E) staining or by immunohistochemistry (IHC) for cytokeratin. We also selected the patients who did not undergo CLND for additional study.
Technique of Intraoperative Lymphatic Mapping and SLNB
Our technique of SLNB has been described elsewhere.22 Briefly, intraoperative lymphatic mapping was performed with injections of blue dye alone, technetium 99m-labeled sulfur colloid, or a combination of the 2 agents. The injection of technetium 99m sulfur colloid was performed under mammographic or sonographic guidance when the tumor was not palpable. On the day of surgery, 5 mL of 1% isosulfan blue (Lymphazurin; US Surgical, Norwalk, Conn) were injected peritumorally, and the breast was massaged for 5 minutes. SLNs were identified as the lymph nodes with uptake of blue dye, radiotracer, or both.
During the SLNB technique learning phase, each surgeon performed immediate level I and II CLND for the first 30 SLN biopsies. Patients with a positive SLN on final pathology are offered CLND as our institution's standard practice. However, patients in this study did not undergo CLND either because of their participation in a clinical trial or because of patient preference after consultation with their physicians.
Pathologic evaluation of SLNs within our institution has evolved with changes in technology. Prior to April 2000, SLNs were sectioned serially along the short axis at 2-mm to 3-mm intervals; sections were embedded in paraffin blocks, and 1 level from each block was stained with H&E. Beginning in April 2000, each paraffin block was sectioned serially at 5-μm intervals to produce 10 levels. Levels 1, 2, and 4 were evaluated by routine H&E staining, and level 3 was analyzed for cytokeratin by IHC.23
The maximum dimension of the metastasis in each SLN was measured. If >1 deposit of tumor was observed in an SLN or if multiple SLNs were positive, then the size of the largest tumor deposit was recorded. All SLNs were examined for extranodal extension. The 6th edition of the AJCC Cancer Staging Manual was used for staging.9
Primary tumors were evaluated for the size of the invasive component, histologic type, Black nuclear grade, estrogen receptor status, progesterone receptor status, HER-2/neu status, and presence of LVI. For hormone receptors, >10% staining of the cells by IHC was considered positive. HER-2/neu was considered positive for 3+ overexpression by IHC or gene amplification by fluorescence in situ hybridization.
Most patients received an anthracycline-containing regimen. Adjuvant chemotherapy was offered to patients with lymph node involvement and to patients with lymph node-negative disease and tumors that measured ≥1 cm in greatest dimension. Preoperative chemotherapy was used on clinical protocols that were designed to compare the efficacy of 2 chemotherapy regimens or for patients who had operable, large breast cancers to achieve tumor shrinkage and to facilitate breast conservation. Endocrine therapy was considered for patients with estrogen receptor-positive and/or progesterone receptor-positive tumors after chemotherapy or, for patients who did not receive chemotherapy, after surgery.
Adjuvant external-beam RT to the whole breast was offered to all patients who underwent breast conservation. The majority of these patients also received a radiation boost to the tumor bed. Radiation to the regional lymph nodes and postmastectomy radiation were used at the discretion of the radiation oncologist.
Patients were followed with history and physical examinations at least every 6 months for the first 5 years and annually every year thereafter. Mammograms were obtained 6 months after the completion of RT for breast-conserving treatment and annually thereafter. Repeat imaging studies of the lymph node basins and other potential sites of distant disease were obtained if there was clinical suspicion based on history, physical examination, or abnormal laboratory values.
Descriptive statistics were used to assess the frequency distribution among the study population. The endpoints for analysis were rates of locoregional and distant recurrence as well as overall survival. Survival analyses were calculated by using the Kaplan-Meier method. The SPSS 10.1 software package (SPSS Inc., Chicago, Ill) was used for statistical analyses.
Patients and Primary Tumor Characteristics
We evaluated the records of 3366 consecutive patients with breast cancer who underwent SLNB at our institution between 1993 and July 2005. The total number of patients who had a positive SLN and invasive breast cancer was 750. There were 196 patients who did not undergo CLND, and they formed the study group for the current analysis (Table 1). Six of the patients in the study group were enrolled in the American College of Surgeons Oncology Group Z0011 trial and were randomized to the observation arm. The remainder of the patients did not undergo CLND because of patient preference after consultation with their physician.
|Variable||No. of patients (%)*|
|Total no. of patients||196 (100)|
|Type of operation|
|Primary tumor characteristics|
|Invasive ductal||145 (74)|
|Invasive lobular||19 (9.7)|
|Mixed invasive ductal and lobular||19 (9.7)|
|Initial tumor stage|
|Estrogen receptor status†|
|Progesterone receptor status†|
The median patient age was 56 years (range, 29–88 years). The majority of patients underwent breast conservation concurrently with or before SLNB (135 patients; 68.9%). The predominant primary tumor histology was invasive ductal carcinoma (145 patients; 74%). The initial stage of the primary tumor was T1 in 72.4% of patients, and most tumors had a modified Black nuclear grade of 2. Of the 158 tumors for which information on LVI was available, 21.9% were positive. Eighty-two percent of tumors were estrogen receptor-positive, 69.9% were positive for progesterone receptor, and 12.8% were positive for HER-2/neu.
The median number of SLNs identified was 3 per patient (range, 1–14 SLNs per patient) (Table 2), with a median of 1 positive SLN identified per patient (range, 1–6 positive SLN per patient). The majority of SLN metastases were detected by serial sectioning and H&E staining (126 patients; 64.3%), rather than IHC (70 patients; 35.7%). The amount of metastatic disease in the SLN was small, and the median size of the largest SLN tumor deposit was 1.0 mm (range, 0.1–12.9 mm). Sixty-seven patients had isolated tumor cells that measured no greater than 0.2 mm (pN0[i+]), 90 patients had micrometastasis (>0.2 mm, none; >2 mm; pN1mi), and 39 patients had macrometastasis that measured >2 mm. Information regarding extracapsular extension of the SLN metastasis was available for 193 patients, and only 10 patients (5.1%) had SLNs with extracapsular extension. There were 23 patients who had NSLNs recovered, 1 of whom had an NSLN that was positive for metastasis.
|Variable||No. of patients (%)*|
|No. of SLNs identified|
|No. of positive SLNs|
|Total size of SLN metastasis, mm|
|No. of patients with NSLNs recovered||23|
|Ratio of all positive LNs to total LNs, %|
|Extracapsular extension†||10 (5.1)|
|Method of detecting SLN metastasis|
|IHC only||70 (35.7)|
|Pathologic lymph node status|
|pN0 (i+)||67 (34.2)|
Table 3 summarizes the patterns of treatment. Twenty-eight patients received neoadjuvant chemotherapy, primarily for primary tumors that were clinical stage ≥T2 on presentation with negative axillae by ultrasound. The majority of patients received adjuvant chemotherapy (109 patients; 55.6%) and RT (114 patients; 55.6%). Adjuvant chemotherapy regimens consisted of anthracycline-based therapy in 49 patients, taxane alone in 5 patients, and anthracycline with taxane in 55 patients. Fifty-three patients received endocrine therapy alone.
|Treatment||No. of patients (%)|
|Neoadjuvant chemotherapy||28 (14.3)|
|Adjuvant chemotherapy||109 (55.6)|
|Anthracycline and taxane||55|
|Adjuvant RT||114 (58.2)|
|Surgery preceding RT|
|Breast and chest wall||32 (36)|
|Breast and axilla||30 (34)|
|Comprehensive RT†||26 (30)|
|Endocrine therapy alone||53 (27)|
Of the patients who received adjuvant RT, 6 patients had undergone mastectomy, and 108 patients had undergone a lumpectomy. Details regarding radiation treatment fields were available for the 89 patients who underwent RT at our institution. Thirty-two of these patients received RT to the breast and chest wall (36%), 30 patients received RT to the breast and axilla (34%), and the remaining 26 patients (30%) received comprehensive RT, which included the axillary apex, internal mammary, infraclavicular, and/or supraclavicular lymph node basins along with the breast, chest wall, and low axilla.
Patient Follow-up and Outcome
The median follow-up was 29.5 months (range, 1.3–62.3 months). Two patients died of unrelated causes (myocardial infarction and advanced lymphoma). Of the remaining 194 patients, none developed an axillary recurrence, 1 patient developed a supraclavicular lymph node recurrence, and 3 patients developed metastatic disease to the lung (2 patients) or bone (1 patient). The median time to recurrence was 32 months (range, 8–44 months).
The patient who developed a supraclavicular lymph node recurrence had a T1 tumor on presentation, a 0.7-cm invasive ductal carcinoma with extensive ductal carcinoma in situ on final pathology, 1 positive SLN with 1.0-mm metastasis, 5 additional SLNs that were negative for metastasis, and no extranodal extension. Surgical margins on the segmental mastectomy specimen were positive for intraductal carcinoma, but the patient declined any further treatment. She refused further follow-up until she presented with palpable supraclavicular lymph node disease 39 months after diagnosis.
Patient 2, who developed bone metastasis, was a woman aged 31 years at initial diagnosis with a strong family history of breast cancer. She underwent mastectomy with SLNB, which revealed a 2.9-cm, poorly differentiated, multifocal, invasive ductal carcinoma with 1 positive SLN (0.5 mm metastasis), 10 negative SLNs, and no extranodal extension. She was treated with adjuvant chemotherapy without RT but developed bone metastasis 44 months after diagnosis.
Patient 3 developed lung metastases at 37 months after diagnosis. Her primary tumor was a 2.3-cm invasive ductal carcinoma that was poorly differentiated, and she had 1 positive SLN (2-mm metastasis) and no extranodal extension. After she underwent segmental mastectomy with SLNB, she received RT to the breast and axilla, chemotherapy, and endocrine therapy. At 37 months after diagnosis, she developed hemoptysis and was diagnosed with pulmonary metastases.
Finally, Patient 4, who also developed lung metastases, initially presented with a T3, poorly differentiated, multifocal, invasive ductal carcinoma. After neoadjuvant chemotherapy, she had 2 positive SLNs (0.4 mm and 1.2 mm), no extranodal extension, and 3 additional negative SLNs. Postmastectomy radiation was given to the chest wall, axillary, supraclavicular, and internal mammary lymph node basins. She developed lung metastases at 8 months after initial diagnosis.
Risk of Positive NSLNs
To determine the probability of additional positive NSLNs in our study population, we used both the nomogram developed by Van Zee and colleagues19 at Memorial Sloan-Kettering Cancer Center (MSKCC) and the model developed by our group16 at UT-MDACC. For both models, the initial tumor size was estimated by imaging studies that were obtained at presentation prior to neoadjuvant treatment. In our patients, SLN metastases usually were detected by H&E staining of serial sections or by IHC for cytokeratin; therefore, the method of detection entered in the MSKCC nomogram typically was not frozen section or routine histopathology. We calculated the risk of positive NSLNs for all patients with complete information and, based on the MSKCC nomogram, the median risk for our study population was 9.79% (range, 2–51%).
In the model developed at UT-MDACC, the actual size of the metastasis in the SLN is used instead of the method of detection. To predict the risk of positive NSLNs, 4 pathologic factors are combined into a total score: the size of the primary tumor, the total number of SLNs removed, the largest size of the SLN metastasis, and the presence of LVI. For the patients with complete information, the median score was −0.6 (range, −2.4–3.21) (Fig. 1), which was associated with a 100% negative predictive value (the ability to predict negative NSLNs) and 43% positive predictive value (the ability to predict additional positive NSLNs). Thus, using either of these models, our study population was at low risk for having additional positive NSLNs and. thus, presumably had a low probability of developing axillary recurrences.
Over the past decade, the standard approach to the evaluation of axillary lymph node status in patients with breast cancer has evolved from axillary lymph node dissection (ALND) to the less invasive alternative, SLNB. The role for SLNB continues to be defined, and several groups have developed guidelines and nomograms to predict the risk of additional positive lymph nodes for a patient with a positive SLN. Many clinicians incorporate the nomogram developed by Van Zee et al.19 into their counseling for women with a positive SLN in the decision regarding CLND. After counseling by their physicians, or for personal reasons, women with a positive SLN may elect not to undergo CLND. The outcome of these patients who have a positive SLN but do not undergo additional surgery to the axilla is not clear.
In our analysis, we observed that there were no axillary recurrences in the study population with a median follow-up of nearly 2.5 years. Other recent studies also have reported similarly low rates of recurrence in smaller groups of patients with positive SLNs who did not undergo CLND. A retrospective study of 31 patients with positive SLNs who declined ALND was reported by Fant et al.24 Their study population also had a small volume of metastatic disease in the SLN, similar to our patients, and a majority had micrometastatic disease (<2 mm). No patients had an axillary recurrence at a median follow-up of 30 months. In their review of 1167 patients undergoing SLNB, Jeruss et al. reported 73 patients who had a positive SLN and did not undergo CLND.25 None of those patients developed an axillary recurrence (median follow-up, 27.4 months).
Naik et al. sought to identify the rate of axillary recurrence after SLNB in a group of 4008 patients with breast cancer.26 In their subgroup analysis of patients who had a positive SLN without CLND, 3 of 210 patients had axillary recurrences, which occurred from 11 months to 46 months after diagnosis. The size of the SLN metastasis was not reported in that study, but it may have varied widely, because the method of detection ranged from IHC, to frozen section, to H&E staining.
The efficacy of adjuvant treatment on the clearance of residual microscopic disease in the axillary lymph nodes is unclear. Nearly 60% of our patients received either adjuvant chemotherapy and/or RT (Table 3), and a portion of the axilla was covered in the radiation field for some of the patients in this study. The majority of level I and II axillae can be included in a breast-conserved radiation field when a high tangent technique is utilized. This technique achieves near-complete axillary coverage by extending the superior field border to within 2 cm of the humeral head and by setting the deep field edge at least 2 cm posterior to the lung-chest wall interface.27 In patients who receive comprehensive postmastectomy RT, axillary coverage commonly is achieved by the combination of tangent fields (treating the low axilla and chest wall) matched superiorly to an appositional field (treating the supraclavicular lymph nodes and the axillary apex). Of the 89 patients who received RT at our institution, 30 patients (15% of the entire cohort) received radiation using the high tangent technique; and, of the entire cohort, at least 56 patients (29%) received axillary irradiation by this or other means.
Pejavar and colleagues retrospectively reviewed the regional lymph node recurrence patterns of a large cohort of patients who underwent breast-conserving surgery and received RT. Of the 1920 patients who were analyzed in that study, 16 patients had a positive SLN without undergoing ALND and received RT with varying degrees of lymph node coverage. At a minimum follow-up of 2 years and a median follow-up of 13 years, none of those 16 patients had a lymph node failure.28
The results from a prospective study of axillary radiation without CLND for patients with breast cancer who had a positive SLN were reported recently by Gadd et al.29 In their study, patients with T1/T2,N0 disease who were planning to undergo breast-conservation prospectively underwent SLNB. If the SLN was positive (as defined by any metastatic deposit by H&E), then the patients received RT to the axilla without CLND. With median follow-up of 32 months, 1 of 73 patients who had a positive SLN without CLND but with axillary RT developed an axillary recurrence. All patients received systemic therapy. Thus, RT to the axilla may be an acceptable alternative to CLND for selected patients with positive SLNs. This approach is being tested in a European Organization for Research and Treatment of Cancer (EORTC) Phase III randomized study, which compares axillary RT with CLND for patients with breast cancer who have a positive SLN (the EORTC 10981/22023 AMAROS trial).30, 31
In addition to the effects of adjuvant treatment on clearance of residual axillary disease, the most likely explanation for the absence of regional recurrences in our study population is their very low risk of harboring additional positive axillary lymph nodes. Based on the MSKCC nomogram,19 their risk of having positive NSLNs was <10%. According to the predictive model published by our group,16 the median score for our study population was −0.6, which was associated with a 100% negative predictive value. Thus, the vast majority of these patients probably did not have residual disease in the axilla and would not have benefited from CLND. In addition, the relatively short follow-up of 29 months in our study is yet another factor contributing to the absence of observed axillary recurrences.
Although the current study and several investigations have reported a very low rate of axillary failure for patients with a positive SLN who did not undergo CLND, it is not possible from those results alone to conclude definitively that CLND should be abandoned for these patients. Because almost all patients with breast cancer undergo chemotherapy, RT, and/or hormone therapy in addition to surgery, it is difficult to clearly dissect out the contribution of each individual treatment modality to the extremely low rate of locoregional recurrence that has been reported. The American College of Surgeons Z0011 trial,11 which closed prematurely, may not have the power to determine whether CLND is necessary for all patients who have a positive SLN; however, the results of the EORTC study at least will shed light on whether RT will be an acceptable alternative to CLND for this group of patients.
- 9GreeneFL,PageDL,FlemingID, et al, eds. AJCC Cancer Staging Manual,6th ed. New York, NY: Springer-Verlag; 2002.
- 29Prospective study of axillary radiation without axillary dissection for breast cancer patients with a positive sentinel node. Presented at the Annual San Antonio Breast Cancer Symposium, San Antonio, Texas, December 8–11, 2005. Abstract 22., , , et al.
- 30European Organization for Research and Treatment of Cancer.Phase III randomized study of complete axillary lymph node dissection versus axillary radiotherapy in sentinel lymph node positive women with operable invasive breast cancer. Available at URL: http://www.clinicaltrials. gov/ct/show/NCT00014612. Accessed June 14, 2007.