Sentinel Node Identification Rate and Nodal Involvement in the EORTC 10981-22023 AMAROS Trial
- First Online:
- Cite this article as:
- Straver, M.E., Meijnen, P., van Tienhoven, G. et al. Ann Surg Oncol (2010) 17: 1854. doi:10.1245/s10434-010-0945-z
- 2.5k Downloads
The randomized EORTC 10981-22023 AMAROS trial investigates whether breast cancer patients with a tumor-positive sentinel node biopsy (SNB) are best treated with an axillary lymph node dissection (ALND) or axillary radiotherapy (ART). The aim of the current substudy was to evaluate the identification rate and the nodal involvement.
The first 2,000 patients participating in the AMAROS trial were evaluated. Associations between the identification rate and technical, patient-, and tumor-related factors were evaluated. The outcome of the SNB procedure and potential further nodal involvement was assessed.
In 65 patients, the sentinel node could not be identified. As a result, the sentinel node identification rate was 97% (1,888 of 1,953). Variables affecting the success rate were age, pathological tumor size, histology, year of accrual, and method of detection. The SNB results of 65% of the patients (n = 1,220) were negative and the patients underwent no further axillary treatment. The SNB results were positive in 34% of the patients (n = 647), including macrometastases (n = 409, 63%), micrometastases (n = 161, 25%), and isolated tumor cells (n = 77, 12%). Further nodal involvement in patients with macrometastases, micrometastases, and isolated tumor cells undergoing an ALND was 41, 18, and 18%, respectively.
With a 97% detection rate in this prospective international multicenter study, the SNB procedure is highly effective, especially when the combined method is used. Further nodal involvement in patients with micrometastases and isolated tumor cells in the sentinel node was similar—both were 18%.
The concept of sentinel node biopsy (SNB) is based on an orderly pattern of lymphatic drainage from the primary tumor to regional lymph node basins.1 The first lymph node to which a tumor drains, the sentinel node, is detected with the aid of blue dye and/or a radioactive tracer and subsequently removed. The pathological status of the sentinel node is used to decide whether an axillary lymph node dissection (ALND) should be performed. Patients with a tumor-negative sentinel node can be spared a completion ALND and the associated side effects. In 1994, Giuliano et al. first reported the SNB procedure in breast cancer.2 The SNB procedure was followed by routine ALND in the learning phase. In this period, the median false-negative rate of the SNB procedure was 7%.3 At present, the sentinel node procedure is the standard treatment in patients with a clinically negative axilla. The axillary recurrence rate in patients with a negative sentinel node is low, 0.3% after a median of 34 months.4 Nevertheless, whether omitting an ALND in patients with tumor-negative sentinel nodes will affect the survival is still subject of research in the National Surgical Adjuvant Breast and Bowel Project B-32 trial.5
Patients with a tumor-positive sentinel node are generally treated with an ALND. Severe side effects of ALND include lymph edema and decreased arm and shoulder function and are observed in 5% to 39% of the patients after axillary clearance.6–8 In 2001, the European Organization of Research and Treatment of Cancer (EORTC) initiated a trial to evaluate the role of axillary radiation. The EORTC 10981 AMAROS (After Mapping of the Axilla, Radiotherapy or Surgery?) trial is a phase III study comparing ALND with axillary radiotherapy (ART) in patients with a tumor-positive sentinel node.9 The main objective of the trial is to prove equivalent locoregional control and reduced morbidity for ART. A secondary aim is to assess the survival in patients with tumor-negative sentinel node without additional axillary clearance. Therefore, patients with a tumor-negative sentinel node are also included in the AMAROS trial. The AMAROS trial is ongoing. The accrual of the required 4,767 patients is expected to be finalized in the year 2010.
The aim of the current substudy was to analyze the technical aspects and outcome of the SNB procedure in the first 2,000 patients, mainly to control for validity of future trial results. We analyzed the identification rate of the SNB procedure and the association between the identification rate and several technical, clinical, and pathological variables. With regard to the outcome of the SNB procedure, research was focused on the relation between the size of the sentinel node metastases and the further nodal involvement.
Patients and Methods
All participating surgeons had to have previously performed a minimum of 30 prequalifying cases of the SNB procedure for breast cancer before being allowed to participate in the trial. In 1,744 patients, the SNB procedure was performed by the combined method of blue dye and isotope with intraoperatively detection with a gamma probe. A minority of SNB procedures were performed with isotope (n = 181) or blue dye (n = 19) only. Lymphoscintigrams were recommended, although not mandatory. Radioactive and blue-stained nodes were removed, and if present, nonsentinel nodes that were suspicious for metastatic cancer on palpation but that were not radioactive or blue were also removed. Subsequently, mastectomy or breast-conserving surgery was carried out. Patients with tumor-positive sentinel nodes who were allocated to the ALND arm underwent a level I and II ALND within 12 weeks. In that case, removal of at least eight lymph nodes was mandatory.
Sentinel node–positive patients allocated to the ART arm were irradiated within 12 weeks after surgery. All three levels of the axilla together with the medial part of the supraclavicular fossa were considered clinical the target volume. The prescribed dose to the axilla as a whole was 50 Gy in 25 fractions of 2 Gy in 5 weeks. All institutions documented their techniques for axillary irradiation on a dummy run that was evaluated by the radiotherapy quality assurance team before their participation was allowed.10 Postoperative axillary irradiation in patients undergoing ALND was allowed in patients with four or more tumor-positive nodes (pN2 or pN3) and was applied according to institutional protocols.
As a minimal requirement, three histological levels (500-μm distance) for each sentinel node were examined. On each level, two parallel sections were performed, one for immunohistochemistry and one for hematoxylin and eosin staining. Immunohistochemical staining was performed for markers containing at least cytokeratin 8 and 18 (e.g., CAM 5.2). Immunohistochemical staining was required only when hematoxylin and eosin staining was negative. The sentinel node was considered tumor positive if any tumor deposit in the node or in the afferent or efferent lymph vessels was found. Tumor deposits were categorized as isolated tumor cells (<0.2 mm), micrometastases (0.2–2 mm), or macrometastases (>2 mm).
Associations between the identification rate and patient and tumor related factors were evaluated by Cochran–Mantel–Haenszel (χ2) type tests for overall association. All P values were two tailed, with P = 0.05 or lower considered significant.
Patient and tumor characteristics of eligible patients (n = 1,953)
Menopausal status, n (%)
Pathological tumor size, n (%)
Histology, n (%)
Grade, n (%)
Variables affecting the SNB identification rate
Not identified (n = 65), n (%)
Identified (n = 1,888), n (%)
Pathological tumor size (cm)
Year of accrual
Blue dye only
Radioactive tracer only
Blue dye and radioactive tracer
Outcome of Sentinel Node Procedure
Sixty-five percent of patients (n = 1,220) were sentinel node negative and underwent no further axillary treatment. The sentinel node was positive in 34% (n = 647). In this group, 409 patients (63%) had macrometastases, 161 patients (25%) had micrometastases, and 77 patients (12%) had isolated tumor cells. Ten patients (0.5%) had positive sentinel nodes only in an extra-axillary region, i.e., internal mammary chain or supraclavicular. In these patients, no further axillary treatment was performed. In 357 (19%) of 1,888 patients, nonsentinel nodes were removed (nonblue or nonradioactive lymph nodes) during the SNB procedure. These nonsentinel nodes were tumor positive in 12% (n = 44). Predictive variables for tumor-positive nonsentinel nodes were tumor size (P < 0.001) and tumor grade (P < 0.001). Nonsentinel nodes were more frequently tumor positive in large and high-grade tumors. In 11 patients (3%), these nonsentinel nodes were the only tumor-positive sentinel nodes. Four of these patients had lack of drainage seen on preoperative lymphoscintigraphy. These SNB procedures can actually be considered falsely negative. Because axillary clearance is omitted in sentinel node–negative patients, the actual false-negative rate in this study will remain unknown.
We did not separately collect data about the outcome of the sentinel nodes in the internal mammary chain in patients in whom these were removed.
Further Nodal Involvement
Further nodal involvement in ALND specimena
Macro (n = 200)
No further involvement
Micro (n = 84)
No further involvement
ITC (n = 33)
No further involvement
In this prospective multicenter study, the sentinel node identification rate was 97%. Lymphatic mapping with the sentinel node procedure was first reported in breast cancer in 1994 and is therefore a relatively young procedure. The patients described in this study were included in the period from 2001 and 2004, and were thus in the relatively early days of the general introduction of this procedure. Nevertheless, the success rate was high. The identification rate was influenced by several variables that we will discuss separately.
The reduced identification rate in older patients is consistently reported.11–13 The increased fatty tissue in the breast in elderly patients might cause an decreased lymphatic flow.14 It is also suggested that the replacement of lymph nodes by fatty tissue decreases the capacity of lymph nodes to retain the radioactive colloid.15
Tumor size was also associated with the detection rate, showing a low identification rate in tumors >5 cm in size. It must be noted that only 13 patients in this study had tumors of >5 cm. Patients with large tumors have a greater risk of extensive axillary tumor burden, which decreases the lymphoscintigraphic visualization.16,17 Patients with more than four tumor-positive lymph nodes have nonvisualization in >50%.18 Performance of ultrasound-guided fine-needle aspiration before the SNB procedure will identify at least some of these patients and thus increase the identification rate.
In this study, the identification rate was slightly lower in patients with other than ductal and lobular carcinomas. This group includes mainly tubular, mucinous, and mixed ductolobular tumors. Because of the heterogeneity of this group, this finding is difficult to explain. Others have not found a correlation with histology.19,20
The association between method of detection and the sentinel node identification is notable, showing a higher detection rate with the combined method (radioactive tracer and blue dye). With blue dye, the lymphatic channel can be identified until it enters and stains the first node. The value of the gamma detection probe is that the location of the node can be determined through the intact skin, and the sentinel can be identified once the lymphatic channel is accidentally damaged. Relying on the gamma detection probe only and omitting blue dye leads to a situation where relevant nodes are left behind, because 5–17% of the sentinel nodes are only blue.5,21,22 Some investigators, however, still use only one of the detection methods. On the basis of our results, which have been confirmed by others, we recommend the use of the combined method.3,11,13
The year of accrual also influenced the identification rate and reflected a learning curve, although surgeons participating in this trial were past their learning period of 30 SNB procedures with a completion axillary clearance. Remarkably, the detection rate was lowest in 2002 and not at the start of the study in 2001. This might be explained by the relatively high number of patients accrued by highly experienced academic institutes at the beginning of the AMAROS trial.
In this study, lymphoscintigraphy did not visualize sentinel nodes in a small number of patients. Despite the nonvisualization, the sentinel node could still be identified in 77% of these patients, particularly by blue dye. This suggests that SNB procedure ought to be pursued in these patients, and that nonvisualization is not an absolute indication for ALND.
Finally, we observed that in 3% of the patients in whom nonsentinel nodes were removed, these nonsentinel nodes were the only tumor-positive nodes. In these patients, the SNB procedure can be considered as falsely negative. Extensive tumor burden may obstruct the lymph flow and thus lead to falsely negative sentinel nodes. This finding emphasizes the importance of palpation after removing the sentinel node. The false-negative rate could also be diminished by performing preoperative axillary ultrasound.
In the patient population of the AMAROS trial—that is, patients with a clinically negative axilla and breast tumor of <3 cm—34% of the patients had a tumor-positive sentinel node. Isolated tumor cells within the sentinel node were classified as having node-positive disease at this time, according to the previous edition (5th edition) of the American Joint Committee on Cancer staging manual.23 Yet in the current substudy of this prospective clinical trial, we were able to assess the further nodal involvement in macrometastases, micrometastases, and isolated tumor cells separately. Further nodal involvement was seen in 18% of the patients with both micrometastases (15 of 84) and isolated tumor cells (6 of 33), whereas extensive nodal involvement was slightly higher in patients with micrometastases. In the current staging classification systems, isolated tumor cells are classified as node-negative (pN0(i+)) disease.24,25 Since 2008, in the AMAROS trial, isolated tumor cells are also considered to be sentinel node negative and do not require further axillary treatment.
In the most recent meta-analysis, the overall pooled risk for additional nodal involvement in patients with isolated tumors cells was 12.3% (95% confidence interval 9.5–15.7).26 Omitting ALND is probably acceptable for most patients whose sentinel node contains isolated tumor cells, especially given a marginally lower false-negative rate of 7% of the SNB procedure.3 Additional microscopic axillary metastases might be eradicated by adjuvant radiotherapy to the breast, including the caudal half of the axilla, or by adjuvant chemotherapy or endocrine therapy. These hypotheses are reflected by a low axillary recurrence rate in sentinel node–positive patients treated with a SNB procedure only (i.e., without axillary clearance).27–32
The retrospective Dutch MIRROR trial recently showed that omission of further axillary treatment in patients with micrometastases resulted in a far higher 5-year axillary recurrence rate (1.2 vs. 6.2%, hazard ratio 4.45; 95% confidence interval 1.46–13.54).33 The International Breast Cancer Study Group trial IBCSG-23-01 is currently addressing whether differences in survival exist between patients with micrometastases who have a SNB procedure compared with ALND.34 However, the size of the sentinel node metastases is not the only predictor for further nodal involvement. Increased tumor size, more than one positive sentinel lymph node, lymphovascular invasion in the primary tumor, and lobular histology also have statistically significant predictive value.35 Several nomograms and models including these variables have been established to estimate the risk of further nodal involvement and might be used to select those patients with micrometastases and isolated tumor cells who may benefit from axillary treatment and those who can be safely spared axillary clearance.36 Further studies addressing which patients with microscopic metastases in the sentinel node can be safely withhold axillary treatment are warranted.
In conclusion, this study indicates that with a 97% detection rate in this prospective international multicenter study the sentinel node procedure is highly effective. The success rate is influenced by the method of the SNB procedure and by several patient and tumor characteristics. In patients with micrometastases and isolated tumor cells in the sentinel node, further nodal involvement was low (18%) but not negligible. The final analysis of the AMAROS trial will show whether patients with a tumor-positive sentinel node will be adequately treated with ART compared to ALND in terms of axillary control and arm and shoulder morbidity.
This publication was supported by grants 2U10 CA11488-28 through 5U10 CA011488-38 from the National Cancer Institute (Bethesda, MD) and by a donation from the Kankerbestrijding/KWF from the Netherlands through the EORTC Charitable Trust. Its content is solely the responsibility of the authors and does not necessarily reflect the official views of the National Cancer Institute. We are indebted to the women who participated in this study and to the doctors, nurses, and data managers from the participating hospitals in the Europe who enrolled patients onto the AMAROS trial.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.