Abstract
Background/Aim: Sentinel lymph node biopsy (SLNB) has become the standard procedure for axillary staging in clinically node-negative breast cancer. Traditionally, axillary lymph node dissection (ALND) has been performed when intraoperative pathological assessment revealed sentinel lymph node (SLN) metastasis. However, growing evidence suggests that postoperative radiotherapy and systemic therapy may safely replace ALND in selected patients, challenging the clinical necessity of intraoperative SLNB assessment.
Patients and Methods: We retrospectively reviewed 540 patients with stage I-III primary breast cancer who underwent curative surgery and SLNB without intraoperative pathological evaluation between January 2018 and December 2021. Clinicopathological characteristics, adjuvant treatment strategies, and survival outcomes were analyzed, with a focus on patients with SLN metastases.
Results: SLN metastases were identified in 87 patients. Postoperative management consisted of ALND (n=19), axillary radiotherapy (n=39), systemic therapy alone (n=28), or no further treatment (n=1). Patients undergoing ALND showed a significantly higher recurrence rate compared with other treatment groups (p=0.028, log-rank), though ALND was more commonly performed in those with ≥3 positive SLNs or after mastectomy (p<0.001). In a high-risk subgroup of 31 patients not fulfilling ALND omission criteria, recurrence-free survival did not significantly differ between ALND and non-ALND groups (p=0.209, log-rank). Multivariate analysis confirmed that omission of ALND was not an independent prognostic factor for recurrence (hazard ratio=0.29; 95% confidence interval=0.03-2.39; p=0.248).
Conclusion: In the era of effective adjuvant radiotherapy and systemic therapy, the indications for omitting ALND may be further extended even to higher-risk patients. These results highlight the diminishing role of intraoperative pathological assessment of SLNB in contemporary breast cancer management.
Introduction
Surgical evaluation of the axilla has long been considered essential in the management of breast cancer, as patients with clinically node-negative (cN0) disease may still harbor pathological axillary lymph node metastases (pN+). Sentinel lymph node biopsy (SLNB) has become the standard procedure for axillary staging, demonstrating comparable accuracy to axillary lymph node dissection (ALND) while significantly reducing morbidity in node-negative patients (1).
Traditionally, ALND was performed when SLNB revealed metastasis. However, pivotal trials such as ACOSOG Z0011 and AMAROS demonstrated that, in patients with T1-T2 breast cancer undergoing breast-conserving surgery followed by axillary radiotherapy, ALND could be safely omitted when metastases were limited to one or two SLNs (2, 3). These findings have been incorporated into clinical practice guidelines worldwide (4).
More recently, the SENOMAC trial expanded the criteria for omission of ALND to include patients with T3 tumors and those undergoing mastectomy, underscoring the efficacy of systemic therapy and radiotherapy in controlling residual nodal disease (5). Furthermore, the SOUND and INSEMA trials suggested that even SLNB itself may be safely omitted in carefully selected low-risk cases (6, 7).
Despite these advances, intraoperative frozen section analysis of SLNB is still frequently performed to guide immediate ALND. However, this approach prolongs operative time and may increase the risk of overtreatment (8). Since 2015, our institution has discontinued intraoperative SLNB assessment, instead relying on postoperative strategies to guide axillary management.
The present study therefore aimed to evaluate axillary involvement and postoperative treatment outcomes in patients with SLN metastases who underwent surgery without intraoperative pathological assessment. By analyzing real-world data from a Japanese single-institution cohort, we sought to clarify the prognostic implications of postoperative management and reconsider the current role of intraoperative SLNB evaluation in contemporary breast cancer care.
Patients and Methods
Patient selection. We retrospectively reviewed the medical records of 540 patients with clinically node-negative (cN0) stage I-III primary breast cancer who underwent curative surgery and SLNB without intraoperative frozen section analysis at Osaka Metropolitan University between January 2018 and December 2021. Patients were classified as either SLNB-negative (pN0) or SLNB-positive (pN+), and clinicopathological characteristics and outcomes were compared. Among those with SLN metastases, patients were further stratified according to postoperative management, including ALND, axillary radiotherapy, systemic therapy alone, or no additional treatment.
Pathological and immunohistochemical assessment. Tumor stage and T/N classification were defined according to the seventh edition of the TNM Classification of Malignant Tumors by the Union for International Cancer Control (UICC) (9). Tumors were classified into molecular subtypes based on immunohistochemistry (IHC) for estrogen receptor (ER), progesterone receptor (PgR), HER2, and Ki-67. ER and PgR positivity were defined as nuclear staining in more than 1% of tumor cells. HER2 positivity was determined as an IHC score of 3+ or an IHC score of 2+ with HER2/CEP17 ratio ≥2.0 using fluorescence in situ hybridization (FISH) (10). A Ki-67 index of ≥14% was defined as high proliferative activity (11).
Endpoints and statistical analysis. Recurrence-free survival (RFS) was calculated from the date of surgery to the date of disease recurrence, while overall survival (OS) was calculated from surgery to death from any cause. Associations between clinicopathological variables and SLNB results or adjuvant treatment strategies were assessed using χ2 or Fisher’s exact test. Survival curves were estimated using the Kaplan–Meier method and compared with the log-rank test. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using Cox proportional hazards regression. Statistical analyses were performed using JMP version 16 (SAS Institute, Cary, NC, USA), and a p-value <0.05 was considered statistically significant.
Ethics statement. This study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Ethics Committee of Osaka Metropolitan University (approval number #2025-045). Written informed consent was obtained from all patients or their legal guardians before treatment.
Results
SLNB outcomes and baseline characteristics. Among the 540 patients analyzed, SLN metastases were detected in 87 cases (16.1%), while 453 patients (83.9%) were negative for SLN involvement. Clinicopathological comparisons between these two groups are summarized in Table I. High Ki-67 expression was significantly associated with the presence of SLN metastases (p=0.003). In survival analyses, patients with SLN metastases exhibited significantly worse RFS compared with SLN-negative patients (p<0.001, log-rank), whereas OS did not differ significantly (p=0.810, log-rank) (Figure 1).
Correlation between clinicopathological features and the result of sentinel lymph node biopsy (SLNB).
Kaplan–Meier curves of recurrence-free survival (RFS) and overall survival (OS) according to sentinel lymph node (SLN) status in all patients with early breast cancer.
Postoperative management in SLN-positive patients. Of the 87 SLN-positive patients, 19 (21.8%) underwent ALND, 39 (44.8%) received axillary radiotherapy, 28 (32.2%) were treated with systemic therapy alone, and one patient received no further treatment. Clinicopathological features according to treatment strategies are presented in Table II. ALND was significantly more frequent among patients who underwent mastectomy (p=0.001) or those with three or more positive SLNs (p=0.012). Conversely, ALND was significantly less common in patients with HR-positive/HER2-negative tumors (p=0.010), and there was also a tendency to avoid ALND in cases with low Ki-67 expression (p=0.064).
Correlation between clinicopathological features and adjuvant axillary treatment in all patients with sentinel lymph node (SLN) metastasis.
Survival outcomes by treatment strategy. Patients treated with ALND demonstrated significantly worse RFS compared with those treated with non-ALND strategies (p= 0.028, log-rank), although OS did not differ significantly (p=0.083, log-rank) (Figure 2A and B). In univariate analysis, ALND was significantly associated with an increased risk of recurrence (HR=2.83; 95%CI=1.02-7.39; p=0.046). In multivariate analysis, premenopausal status (HR=8.17; 95%CI=1.29-165.71; p=0.023) and triple-negative breast cancer (HR=8.55; 95%CI=1.03-58.57; p=0.048) emerged as independent predictors of recurrence, while ALND itself did not remain significant (Table III).
Kaplan–Meier curves in patients with sentinel lymph node (SLN) metastases, stratified by postoperative management. (A) Recurrence-free survival (RFS). Patients treated with axillary lymph node dissection (ALND) show significantly worse RFS compared with those treated with non-ALND strategies (p=0.028, log-rank). (B) Overall survival (OS). No significant difference is observed between ALND and non-ALND groups (p=0.083, log-rank). (C) High-risk subgroup analysis of RFS (patients with ≥3 positive SLNs or mastectomy). RFS does not significantly differ between ALND and non-ALND groups (p=0.209, log-rank).
Univariate and multivariate analyses with respect to recurrence-free survival (RFS) in all patients with sentinel lymph node (SLN) metastasis.
High-risk subgroup analysis. A subgroup of 31 patients was defined as high-risk, comprising those with three or more SLN metastases or those who underwent mastectomy. Fourteen of these patients (45.2%) underwent ALND, whereas 17 (54.8%) received other management strategies. Clinicopathological features of this subgroup according to treatment are summarized in Table IV. RFS did not significantly differ between patients who underwent ALND and those who did not (p=0.209, log-rank) (Figure 2C). One death was recorded in the ALND group during follow-up.
Correlation between clinicopathological features and adjuvant axillary treatment in high malignant patients with sentinel lymph node (SLN) metastasis.
In univariate analysis, premenopausal status was significantly associated with worse RFS (HR=5.72; 95%CI=1.31-39.18; p=0.020). Multivariate analysis revealed premenopausal status (HR=1.05×1019; 95%CI=8.28-∞; p=0.002) and tumor size greater than 2 cm (HR=11.21; 95%CI=1.14-313.19; p=0.037) as independent predictors of recurrence in this subgroup, whereas ALND was not identified as a significant prognostic factor (Table V).
Univariate and multivariate analyses with respect to recurrence-free survival (RFS) in in high malignant patients with sentinel lymph node (SLN) metastasis.
Discussion
In this retrospective study, we examined real-world outcomes of patients with SLN metastases who underwent surgery without intraoperative pathological assessment of SLNB. We found that patients with SLN metastases had significantly worse recurrence-free survival (RFS) compared with SLN-negative patients, although overall survival (OS) was not significantly different. Importantly, the omission of ALND, even in patients with high-risk features such as mastectomy or three or more positive SLNs, did not independently predict worse outcomes, suggesting that postoperative radiotherapy and systemic therapy can effectively compensate for the lack of axillary dissection.
Our results are consistent with recent pivotal trials. The SENOMAC trial demonstrated that omission of ALND was safe in patients with T3 tumors and in those undergoing mastectomy (12). In that study, approximately one-third of patients in the ALND group had additional non-sentinel lymph node metastases, yet no significant difference in survival was observed between ALND and non-ALND groups. Similarly, our multivariate analysis indicated that systemic therapy and radiotherapy may adequately control residual nodal disease, thereby diminishing the survival benefit of ALND.
Avoidance of unnecessary ALND is clinically important because it reduces complications such as lymphedema, shoulder dysfunction, and neuropathic pain. Pride et al. recently reported that intraoperative pathological assessment significantly increased the rate of ALND and advocated for postoperative evaluation to reduce overtreatment (13). These findings reinforce the limited value of intraoperative frozen section analysis, particularly in an era when effective systemic and radiation therapies are widely available.
However, omission of ALND presents certain challenges in tailoring adjuvant therapy. In hormone receptor (HR)-positive/HER2-negative breast cancer, the monarchE trial demonstrated that adjuvant abemaciclib improves outcomes in patients at high risk of recurrence, specifically those with ≥4 positive nodes, or those with 1-3 nodes plus additional risk factors (14, 15). Without ALND, the exact nodal burden may be underestimated, potentially limiting appropriate patient selection for abemaciclib. Furthermore, in postmenopausal HR-positive/HER2-negative patients with 1-3 positive nodes, Oncotype DX has been shown to guide chemotherapy decisions, whereas patients with ≥4 positive nodes are generally recommended chemotherapy irrespective of genomic profiling (16). In the SENOMAC trial, 12.9% of patients who underwent ALND were found to have ≥4 positive nodes (12), while in our cohort this proportion was notably higher at 42.1%. This discrepancy suggests that omission of ALND could have a substantial impact on treatment decisions in clinical practice.
To overcome this limitation, predictive tools have been proposed. Yang et al. developed a nomogram incorporating four clinicopathological variables to identify patients with ≥4 positive nodes with reasonable accuracy (17). Such models may help refine treatment selection in the absence of ALND, especially as precision oncology continues to evolve.
Study limitations. First, it was conducted in a single institution with a relatively small sample size, which limits the power of subgroup analyses. Second, inclusion of patients who had received neoadjuvant chemotherapy may have introduced heterogeneity, and its confounding effects could not be fully excluded. Third, the follow-up period was less than five years, which is relatively short for assessing long-term oncological outcomes in breast cancer. Future multicenter studies with larger sample sizes and longer follow-up are warranted to validate our findings.
Despite these limitations, our results add to the growing body of evidence supporting de-escalation of axillary surgery in breast cancer. In particular, they highlight that omission of ALND – even in patients traditionally considered unsuitable for omission – did not significantly compromise outcomes when combined with appropriate adjuvant radiotherapy and systemic therapy. Collectively, these findings emphasize the diminishing clinical utility of intraoperative pathological assessment of SLNB in contemporary practice.
Conclusion
Omission of ALND did not independently worsen prognosis in patients with SLN metastases when adequate postoperative systemic therapy and radiotherapy were administered. These findings suggest that the clinical role of intraoperative SLNB assessment is diminishing and that criteria for ALND omission may be further expanded in modern breast cancer management.
Acknowledgements
The Authors would like to thank the members of the OBCTRG consortium, for their contribution to the research project.
Footnotes
Authors’ Contributions
All Authors were involved in the preparation of this manuscript. HM collected the data, and wrote the manuscript. SH, MN, CW, KT, YT, KO, MS, HK and TM performed the operation and designed the study. WG and SK made substantial contributions to the study design. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors declare that they have no competing interests in relation to this study.
Funding
This study was funded by grants from the Japan Society for the Promotion of Science (KAKENHI, Nos. 20K08938 and 23K08035) to Shinichiro Kashiwagi.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received October 3, 2025.
- Revision received October 23, 2025.
- Accepted October 24, 2025.
- Copyright © 2026 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
