Axillary Surgery

Retrospective Study Offers Insight Into Possible Overuse of Axillary Surgery

Posted by

0


A retrospective evaluation of I-SPY2 clinical trial data indicates that in appropriately selected patients with breast cancer, sentinel lymph node surgery with adjuvant radiation may provide appropriate oncologic control without the routine use of axillary dissection. Judy C. Boughey, MD, of the Mayo Clinic, Rochester, Minnesota, and colleagues presented their results during the 2024 Society of Surgical Oncology (SSO) Annual Meeting (Abstract 3). Although follow-up in the I-SPY2 cohort thus far is fairly short (median, 3.5 years), they found that compared with selective use of sentinel lymph node surgery alone after neoadjuvant chemotherapy, axillary dissection did not improve axillary recurrence, locoregional recurrence, distant recurrence, or event-free survival—even in patients with node-positive disease.

They evaluated the outcomes of patients—those who did and did not have residual nodal disease—who underwent axillary surgery after neoadjuvant chemotherapy between 2011 and 2022. Axillary surgery was classified as sentinel lymph node surgery alone or as axillary dissection (with or without sentinel lymph node surgery).

Slightly more than 1,500 patients were included: 714 (47.1%) had clinically node-negative disease at diagnosis, of whom 104 (14.6%) had pathologically node-positive disease; 801 (52.9%) had clinically node-positive disease at diagnosis, of whom 396 (49.4%) had pathologically node-positive disease. Sentinel lymph node surgery alone was performed in 805 of 1,015 (79.3%) patients with pathologically node-negative disease and in 126 of 500 patients (25.2%) with pathologically node-positive disease.

Dr. Boughey and co-investigators found that among patients who had pathologically node-negative disease, no significant differences occurred between those who underwent sentinel lymph node surgery alone and those who had axillary dissection in the 5-year estimated rate of axillary recurrence, locoregional recurrence, distant recurrence, and event-free survival. Among patients who had pathologically node-positive disease, there was no difference between sentinel lymph node surgery alone and axillary dissection in the 5-year estimated axillary recurrence or locoregional recurrence.

Sentinel Lymph Node Biopsy vs No Axillary Surgery in Patients With Small Breast Cancer and Negative Results on Ultrasonography of Axillary Lymph Nodes

Posted by

0


Abstract

Importance  Sentinel lymph node biopsy (SLNB) is the standard of care for axillary node staging of patients with early breast cancer (BC), but its necessity can be questioned since surgery for examination of axillary nodes is not performed with curative intent.

Objective  To determine whether the omission of axillary surgery is noninferior to SLNB in patients with small BC and a negative result on preoperative axillary lymph node ultrasonography.

Design, Setting, and Participants  The SOUND (Sentinel Node vs Observation After Axillary Ultra-Sound) trial was a prospective noninferiority phase 3 randomized clinical trial conducted in Italy, Switzerland, Spain, and Chile. A total of 1463 women of any age with BC up to 2 cm and a negative preoperative axillary ultrasonography result were enrolled and randomized between February 6, 2012, and June 30, 2017. Of those, 1405 were included in the intention-to-treat analysis. Data were analyzed from October 10, 2022, to January 13, 2023.

Intervention  Eligible patients were randomized on a 1:1 ratio to receive SLNB (SLNB group) or no axillary surgery (no axillary surgery group).

Main Outcomes and Measures  The primary end point of the study was distant disease–free survival (DDFS) at 5 years, analyzed as intention to treat. Secondary end points were the cumulative incidence of distant recurrences, the cumulative incidence of axillary recurrences, DFS, overall survival (OS), and the adjuvant treatment recommendations.

Results  Among 1405 women (median [IQR] age, 60 [52-68] years) included in the intention-to-treat analysis, 708 were randomized to the SLNB group, and 697 were randomized to the no axillary surgery group. Overall, the median (IQR) tumor size was 1.1 (0.8-1.5) cm, and 1234 patients (87.8%) had estrogen receptor–positive ERBB2 (formerly HER2 or HER2/neu), nonoverexpressing BC. In the SLNB group, 97 patients (13.7%) had positive axillary nodes. The median (IQR) follow-up for disease assessment was 5.7 (5.0-6.8) years in the SLNB group and 5.7 (5.0-6.6) years in the no axillary surgery group. Five-year distant DDFS was 97.7% in the SLNB group and 98.0% in the no axillary surgery group (log-rank P = .67; hazard ratio, 0.84; 90% CI, 0.45-1.54; noninferiority P = .02). A total of 12 (1.7%) locoregional relapses, 13 (1.8%) distant metastases, and 21 (3.0%) deaths were observed in the SLNB group, and 11 (1.6%) locoregional relapses, 14 (2.0%) distant metastases, and 18 (2.6%) deaths were observed in the no axillary surgery group.

Conclusions and Relevance  In this randomized clinical trial, omission of axillary surgery was noninferior to SLNB in patients with small BC and a negative result on ultrasonography of the axillary lymph nodes. These results suggest that patients with these features can be safely spared any axillary surgery whenever the lack of pathological information does not affect the postoperative treatment plan.

Introduction

Sentinel lymph node biopsy (SLNB) is the standard of care for axillary node staging in patients with early breast cancer (BC). The application of this technique represented a milestone in surgical de-escalation, providing the same outcome as axillary lymph node dissection (ALND).1,2 The results of the American College of Surgeons Oncology Group Z0011 (ACOSOG Z0011) randomized clinical trial3 showed that there is no advantage in performing ALND compared with not performing ALND, even when up to 2 sentinel nodes are positive, for patients receiving breast-conserving surgery, adjuvant radiotherapy, and medical treatment. These data were supported by other randomized clinical trials4,5 and became the basis for a new standard of axillary care.6,7 In the past, SLNB was conceived as a reliable means to distinguish between patients with negative nodes who can be spared the morbidity associated with a complete ALND and those with nodal involvement who might benefit from a more extensive surgical procedure. But the absence of advantages from ALND revealed in the ACOSOG Z0111 trial3 raised 2 questions: first, whether it is really necessary to perform surgical staging of axillary lymph nodes, and second, whether imaging might replace surgery for reliable staging of axillary lymph nodes. The SOUND (Sentinel Node vs Observation After Axillary Ultra-Sound) trial was launched in February 2012 with the aim of evaluating the oncological safety of omitting axillary surgery in patients with BC of a diameter equal to or smaller than 2 cm and a negative result on preoperative axillary lymph node ultrasonography.

Methods

The SOUND trial was a prospective multicenter noninferiority phase 3 randomized clinical trial conducted in 18 hospitals in Italy, Spain, Switzerland, and Chile (eTable in Supplement 2). Participants were recruited and enrolled from February 6, 2012, to June 30, 2017. Data were analyzed between October 10, 2022, and January 13, 2023. The Trial Protocol and Statistical Analysis Plan for this trial are provided in Supplement 1. Of note, the SOUND trial was unintentionally registered late with ClinicalTrials.gov on June 18, 2014, after 492 patients had already been randomized. No interval analysis was conducted, and no data were examined before clinical trial registration. The trial was conducted in accordance with the amended Declaration of Helsinki,8 and the protocol was approved by the institutional review board at each participating center. All patients provided written informed consent. This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline for randomized clinical trials.

Study Population

Eligible patients were women of any age with invasive BC up to 2 cm in diameter, lack of involvement of axillary nodes at clinical evaluation, and a plan to undergo a breast-conserving surgery and radiotherapy. All patients were required to have preoperative axillary ultrasonography showing no lymph node involvement at imaging. In the case of a doubtful finding on ultrasonography concerning an isolated lymph node, fine-needle aspiration was performed to rule out the presence of nodal metastases by cytological examination. Axillary lymph nodes with micrometastases or macrometastases were defined as positive. Exclusion criteria were the preoperative presence of multiple doubtful or suspicious lymph nodes, extensive multifocality or multicentricity, bilateral BC, diagnosis of synchronous distant metastases, previous cancer, ongoing pregnancy or lactation, and obstacles to obtaining informed consent or undergoing regular follow-up.

For the sample size calculation, the 5-year distant disease–free survival (DDFS) in the group randomized to receive no axillary surgery (no axillary surgery group) was assumed to be 96.5%. We calculated a target sample size of 1560 participants (780 per group) to test whether the no axillary surgery group experienced outcomes that were no worse than those of the group randomized to receive SLNB (SLNB group) given a margin of noninferiority for the 5-year DDFS of 2.5%. Between February 6, 2012, and June 30, 2017, 1463 women were recruited and enrolled in the study (93.8% of the planned sample size). The steering committee decided to close the accrual period early because the enrollment became too slow after most of the participating hospitals adopted the ACOSOG Z0011 criteria. Most of the enrolled patients (n = 1406) had negative ultrasonographic results; 57 patients who had a single doubtful node on ultrasonography were randomized after undergoing fine-needle aspiration cytological examination with negative results. Overall, 727 patients were randomized to the SLNB group and 736 to the no axillary surgery group. In the SLNB group, 19 patients discontinued intervention; in the no axillary surgery group, 39 patients discontinued intervention (Figure 1). The remaining 1405 women (708 in the SLNB group and 697 in the no axillary surgery group) were included in the intention-to-treat analysis.

Discussion

In the SOUND trial, the omission of axillary surgery was noninferior to surgical staging performed by SLNB when evaluating DDFS at 5 years in patients with BC up to 2 cm and a negative result on preoperative ultrasonography of axillary lymph nodes. Of note, in the no axillary surgery group, the cumulative incidence of lymph node recurrences in the axilla was very low (0.4% at 5 years), despite a 13.7% rate of nodal involvement in the SLNB group.

Other prospective randomized clinical trials13–15 conducted in the pre-SLNB era have shown that ALND did not improve outcomes compared with no surgery in the axilla, confirming that removal of axillary lymph nodes had no therapeutic effect by itself and was performed mainly as a staging procedure. Thus, a relevant research aim addressed as a secondary end point was to evaluate possible differences in terms of adjuvant treatment recommendations. Data from the current trial indicated that adjuvant treatments were not significantly different in the 2 study groups, regardless of whether the pathological information from SLNB was available. These data confirm the increasing pattern of guiding adjuvant treatment mostly through the use of biological parameters rather than clinicopathological variables.16–18

The data in this trial were consistent with guidelines from the Choosing Wisely campaign of the Society of Surgical Oncology,19 which recommends omitting SLNB in patients older than 70 years with small ER-positive ERBB2-negative BC when the adjuvant treatment plan is clear and does not include the addition of chemotherapy to endocrine treatment. However, the information provided by nodal status is currently not being completely ignored when selecting postoperative treatment for younger patients, and the absence of the pathological information acquired from SLNB might still create challenges in the management of ER-positive ERBB2-negative BC. Even with the wide availability of genomic testing, chemotherapy can be prescribed or at least considered in addition to endocrine treatments for women with endocrine-responsive disease and axillary lymph node involvement, especially for premenopausal patients. In the Rx-PONDER (A Clinical Trial Rx for Positive Node, Endocrine Responsive Breast Cancer) trial,20 the advantage of adding cytotoxic agents could not be ruled out in patients younger than 50 years with nodal metastases, even when the recurrence score was low or intermediate. In addition, the duration of adjuvant endocrine therapy can be adapted according to the risk estimate and usually prolonged after 5 years in patients with node-positive disease.21–23 On the other hand, the absence of pathological nodal involvement might allow de-escalation of the hormonal treatment, both in terms of drug choice (tamoxifen vs aromatase inhibitors or ovarian suppression vs no ovarian suppression) and duration, especially in the case of adverse effects that substantially affect patient quality of life. Furthermore, in the subset of patients with ERBB2-positive disease undergoing upfront surgery, information on nodal status is relevant to properly tailor adjuvant treatment, which in node-negative disease might be restricted to paclitaxel and trastuzumab.24 In addition, in patients with small triple-negative BC undergoing upfront surgery, pathological staging of nodal status might be relevant to modulate the postoperative treatment plan. Moreover, nodal radiation fields are frequently adapted for women with nodal involvement as a complement to breast radiotherapy after breast conservation. In contrast, some patients 65 years and older with node-negative disease might even be spared from undergoing any radiotherapy with a limited number of locoregional events and no detrimental effect on OS.25,26

We were also interested in evaluating the capacity of ultrasonography to detect nodal involvement to understand whether imaging might eventually replace surgery for reliable staging.9 It is well known that ultrasonography of axillary nodes has several limitations in detecting lymph node involvement, with a sensitivity ranging from 24% to 94%.27 However, in the current study, the use of ultrasonography was able to rule out the presence of relevant nodal burden, which might not have been identified with clinical evaluation alone. In the SLNB group, the presence of micro- and macrometastases was limited (13.7%) and much lower than the rate reported in previous trials,1,2 likely due to the screening effect of the negative preoperative axillary ultrasonography result required to enter the trial. Given the limited number of patients with macrometastases, the very low number of patients with extensive nodal involvement (0.6% with 4 or more positive nodes) in the axillary surgery group, and the extremely low cumulative incidence of axillary lymph node recurrence in the no axillary surgery group (0.4% at 5 years), the performance of ultrasonography can be considered clinically meaningful. Despite the need for further research to improve imaging methods, the multi-institutional nature of our study supported the wide reproducibility of ultrasonography as a simple and inexpensive method that can be routinely applied in the preoperative workup of all patients with BC.

The results of this trial support the safety of omitting axillary surgery in older postmenopausal women with ER-positive ERBB2-negative BC who met the SOUND eligibility criteria. This subset of women represents approximately 25% of the whole population of women with BC.28,29 Considering that an estimated 2.3 million women are diagnosed with BC every year, approximately 500 000 patients might be able to take advantage of the total omission of axillary surgery, which has been shown to improve arm function in the early postoperative period.11,30,31 Data from the SOUND trial should be considered in the multidisciplinary decision-making process of the individual patient to identify those who might be able to omit SLNB without affecting the postoperative treatment plan. Moreover, the incorporation of these data in future guidelines might lead to a substantial decrease in health care costs due to the reduced involvement of human resources and savings in terms of materials and time.

Limitations

This study has several limitations. The inclusion criteria of the study led to enrollment of patients who could be considered to be at low risk of recurrence in the short term. Thus, we cannot exclude the possibility that differences in outcome might appear over a longer follow-up period because the curve of event onset is expected to occur later in patients with ER-positive ERBB2-negative BC than in patients with triple-negative or ERBB2-positive disease.32,33 Therefore, we have planned to continue the follow-up with a formal analysis after 10 years. We also highlight that the analysis of adjuvant treatments was not the primary end point and that sample size calculations were not performed for this purpose. Therefore, this trial might be underpowered to detect small differences in the details of medical or radiotherapy treatment recommendations.

The SOUND trial was designed in 2011, immediately after publication of the ACOSOG Z0011 trial34 and as its natural continuation. At that time, ALND was the standard of care in the presence of sentinel lymph node metastases because data from the ACOSOG Z0011 trial had not yet been incorporated into guidelines. Because the objective of the SOUND trial was to evaluate the oncological safety of omitting axillary surgery, we decided to compare the experimental group with the most standardized and radical approach comprising ALND in the presence of nodal macrometastases. It was only between 2016 and 2017 that most of the participating centers embraced the ACOSOG Z0011 approach, slowing down the recruitment pace of our enrollment. Furthermore, all available options of radiotherapy were allowed, even partial breast radiotherapy. Of note, 114 patients (16.3%) randomized to the no axillary surgery group received ELIOT as a full dose or an intraoperative boost.

We also highlight the unintentional late registration of the SOUND trial in ClinicalTrials.gov even though we specify that no data were examined and no interval analysis was conducted before the trial was registered. Late registration occurred without any intention to bias the reporting, which is confirmed by the fact that study design, inclusion criteria, exclusion criteria, end points, and sample size were published in a peer-reviewed journal9 immediately after the study started. Another article10 was later published reporting the number of patients included in the SOUND trial up to that time, which provided a forecast on the conclusion of patient enrollment.

Conclusions

This randomized clinical trial found that omission of axillary surgery was noninferior to SLNB in women with small BC and negative results on ultrasonography of the axillary lymph nodes. These results suggest that patients with these features can be safely spared any axillary surgery when the lack of pathological information does not affect the postoperative treatment plan.