It is widely acknowledged that breast cancer is the most highly occurring malignant disease among females, posing significant threats to their health (1, 2). Over the past few years, sentinel lymph node biopsy (SLNB) has become as a prevalent technique for diagnosing the status in the axillary region, aiming to reduce trauma and enhance treatment efficacy (3). SLNB may be a substitute for axillary lymph node dissection (ALND) in patients who have a negative sentinel lymph node (SLN), as evidenced by the NSABP-B32 trial (4–6). The current research agenda in the field of breast cancer surgery includes the question of how to manage the axillary region in cases where the SLN is positive (7). The ACOSOG Z0011 trial has shown that patients who proceed breast-conserving in conjunction with postoperative radiotherapy and present with one or two metastatic SLNs are not likely to benefit from additional axillary surgical intervention (8, 9). The AMAROS and IBCSG23-01 studies have furnished corroborating evidence for the strategic intervention of the axilla in patients exhibiting SLN micrometastases (10). Nevertheless, the conduct of axillary surgery in mastectomy patients with 1-2 metastatic SLNs remains a topic of contention (11).

As a component of breast cancer surgical therapy, ALND diminishes the risk of axillary recurrence and provides valuable insights for guiding subsequent treatment decisions (12, 13). The rates of axillary recurrence and survival outcomes were similar between patients who had ALND and those who had no axillary surgery, as reported by the IBCSG 10-93 trial (14). The findings of the study by Gao et al. indicate the survival advantage associated with ALND in individuals who have received a mastectomy and one or two metastatic SLNs is not statistically significant (15–17). Chen et al. also reported similar findings (18, 19). What methods do surgeons employ to ascertain whether a patient with one or two metastatic SLNs would derive benefit from an ALND?

In addition to survival benefits, the locoregional recurrence rate (LRR) of metastases is an important consideration for surgeons when deciding on the operation. The study conducted by Giuliano et al. indicated that SLN dissection alone (but not ALND) is an effective approach for achieving regional control in those with clinically early-stage breast cancer and receiving breast-conserving surgery followed by adjuvant systemic therapy (20). But in patients who undergo mastectomy, can SLNB achieve the same effect?

Our study aims to explore the effect of ALND on prognosis based on molecular subtypes. Meanwhile, by analyzing the differences in LRR, DFS, and OS in different age groups, we investigated the necessity of ALND.

A tendency of de-escalation has been noticed in the axillary surgery of breast cancer. Whether ALND can be omitted in mastectomy remains controversial. The percentage of ALNDs has been gradually reducing in recent years (15). Additionally, patients who had breast-conserving were more often to avoid ALND than those who had a mastectomy (22). In our research, stratified according to the age and molecular subtypes, we examined the impacts of different treatment strategies of axillary on the prognosis of 1-2 metastatic SLNs patients who had a mastectomy. We found that patients in the SLNB cohort had comparable OS and DFS with those in the SLNB + ALND cohort in four molecular subtypes. In the ≤ 40 years group, the ALND cohort had a lower LRR. No differences in survival and LRR between the SLNB + ALND and SLNB were found in other age groups.

In the whole population, the proportion of patients receiving chemotherapy and radiotherapy was significantly higher in ALND cohort than SLNB cohort. Using adjuvant chemotherapy and radiotherapy may affect outcomes when analyzing patients with or without ALND (23, 24). Matching factors were ER status, PR status, HER2 status, Ki-67, and the presence or absence of chemotherapy, radiotherapy, and endocrinetherapy. It should be noted that assessing the precise effect of adjuvant treatments on axillary surgery remains challenging, despite the outlined efforts. Due to the inherent limitations of PSM, despite matching based on these indicators, it is undeniable that there may still be unmatched variables that could lead to biased results.

For the whole population, no significant differences in LRR, DFS and OS were observed between the ALND and SLNB cohorts. However, subgroup analysis by age revealed that in individuals younger than 40 years, the ALND cohort exhibited lower LRR rates compared to the SLNB cohort, with no significant differences in DFS and OS. This suggests that locoregional recurrence may be more likely to occur in certain subgroups, such as younger individuals, which is corroborated by the findings of our multivariate analysis. In this analysis, younger individuals (≤ 40 years) were identified as an independent contributing factor to LRR.

We successfully matched 234 patients in each of the two cohorts by PSM. The effects of treatment factors, including chemotherapy, radiotherapy, and endocrinetherapy on LRR, DFS, and OS were negated by PSM. Following this adjustment, no distinction was observed in the percentage of patients treated with radiotherapy among the two cohorts. The results indicate that for the whole population, patients without ALND did not experience worse LRR, DFS, or OS. The multivariate analysis showed that being younger (≤ 40 years), progesterone receptor-negative, and undergoing SLNB alone were independent risk factors for LRR. Considering this result, we recommend that for patients without high-risk factors, axillary interventions such as ALND or radiotherapy may be avoided.

The results of previous studies have reported similar DFS, OS, and LRR among patients with SLNB alone (25). The results of our investigation revealed no meaningful statistical differences in either OS or DFS between the two cohorts. These findings align with those reported by the study carried out by Susan B. Kesmodel et al. (13, 16, 26). The subsequent analysis demonstrated that patients aged ≤ 40 years in the SLNB + ALND cohort experienced a considerably lower rate of LRR than those in the SLNB cohort. However, no statistical difference was found in the ten-year regional recurrence between ALND and SLNB cohorts (8).

Studies have reported associations between HER2 poitive and TNBC breast cancer subtypes with worse survival and higher rates of LRR (27, 28). However, this feature was not found in our study. Therefore, further investigation is necessary to validate these conclusions.

The use of molecular subtypes as the basis for selecting treatments and predicting prognosis is becoming a more prevalent practice in oncology. It is generally believed that Luminal type breast cancer is sensitive to endocrinetherapy and has a better prognosis than other types (29). HER2 positive patients are sensitive to target therapy, while patients with triple-negative breast cancer lack individualized treatment plans. HER2 positive and TNBC patients have a more unfavorable prognosis. Patients with luminal A and TNBC have a less significant likelihood of axillary lymph node involvement, whereas luminal B and HER2 positive demonstrate a higher incidence of axillary lymph node positive. Although our analysis did not show a survival benefit of ALND for patients with HER2 positive or luminal B, considering their unique biological behavior, we still recommend that clinicians adopt a more aggressive approach to axillary management for these patients.

The biology of breast cancer, rather than the scope of surgical intervention, has been acknowledged as a significant determinant of systemic and locoregional recurrence risk (8). We found that cT1, receipt of endocrinotherapy, and negative perineural invasion were correlated with better OS and DFS. Studies have shown that tumors exhibiting low expression of PR demonstrated more aggressive behaviors compared to those with high expression (30). A study by Arisio et al. demonstrated that hormone therapy significantly impacted OS and DFS (31). Additionally, research by Giuliano et al. has indicated that hormone receptor status, age at disease diagnosis, and the management of adjuvant therapy were related to OS, which is consistent with our findings (8, 32). When perineural invasion is detected in mastectomy patients, it is recommended to administer aggressive post-surgical treatment and conduct an intensive follow-up program (26).

After stratifying patients by age, it was observed that for those aged 41 years and above, ALND did not provide significant survival benefits. This is consistent with the trend of de-escalated treatment in elderly patients (33–35). Despite our population being considered younger than the criteria in previous studies, which classified patients aged 70 years and above as elderly (36). The study by Gu et al. suggested that the exclusion of axillary surgery was not correlated with inferior OS but was correlated with an increased incidence of LRR (37). Interestingly, a similar phenomenon was observed in patients aged ≤ 40 years in our study. However, age did not exhibit a significant association with LRR after adjusting for other factors in ACOSOG Z0011 (20). It has been documented that LRR events occur within the initial five-year period following the commencement of treatment, a critical factor in the management of breast cancer (38, 39).

In our research, there were 10 cases (1.2%) of invasive lobular carcinoma (ILC) and 7 cases (0.8%) of invasive ductal carcinoma (IDC) concomitant with ILC. Given the limited sample size and the fact that, clinically, systemic treatments rarely distinguish between these two histological types, we have not provided a detailed analysis of this category. However, it is undeniable that ILC and IDC have some differences in their biological behavior. For example, ILC tends to have larger tumor sizes, a higher incidence of multifocality, and is more common in Luminal A, which are characteristics associated with a favorable prognosis. There is inconsistent evidence in the literature regarding the outcomes of ILC compared to those of IDC (40). In our future research, we plan to expand the sample size and discuss the differences in survival prognosis and other aspects between this type of breast cancer and IDC.

It should be noted that the study is not without its inherent limitations. Firstly, given the retrospective data collection, there is a potential for selection bias that cannot be overlooked. Secondly, as complete data on molecular targeted therapy were not available, it was impossible to analyze how anti-HER2 therapy impact clinical outcomes in these patients. Thirdly, a relatively short follow-up duration may introduce bias in the survival analysis of axillary surgical management.

Our study found that the younger (≤ 40 years), PR-negative, and SLNB alone were independent risk factors for LRR. Perineural invasion was an independent risk factor, while endocrinotherapy was a beneficial prognostic indicator for DFS and OS among patients with positive hormone receptor. ALND does not impact DFS and OS in patients with 1-2 metastatic SLNs who have completed a mastectomy and may be considered to be omitted in patients without risk factors.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

The studies involving humans were approved by The Ethics Committee of The Fourth Hospital of Hebei Medical University (2022KY054). The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements.

XZ: Conceptualization, Formal analysis, Investigation, Methodology, Resources, Visualization, Writing – original draft, Writing – review & editing. LY: Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. CC: Investigation, Writing – review & editing, Methodology. ZS: Investigation, Writing – review & editing, Conceptualization, Funding acquisition, Project administration, Supervision, Validation.