Oncological outcome of endoscopic nipple-sparing mastectomy versus conventional open mastectomy in early breast cancer: a multicenter cohort study

Abstract

Background

In recent years, endoscopic nipple-sparing mastectomy (ENM) has been applied in breast cancer treatment, whereas conventional open mastectomy (COM) remains widely adopted in clinical practice. The data about the oncological outcomes of ENM vs. COM was rare, especially the multicenter data.

Methods

This multicenter retrospective cohort study enrolled 2,314 early breast cancer patients diagnosed with stage 0–III, and treated with either ENM or COM from September 1st, 2021 to January 30th, 2025 at three institutions. After exclusions, the final cohort consisted of 1,722 patients, of whom 136 were in the ENM group and 1,586 in the conventional group. Demographic data, clinicopathological features, and oncological outcomes were collected and analyzed. Propensity score matching (PSM) was performed to minimize selection bias and balance confounders.

Results

Before PSM, the patients in ENM group were significant different from COM group in age, menopausal status, hormone receptor status, HER2 status and molecular subtypes. After PSM, the final analysis cohort comprised 136 patients in the ENM group and 408 patients in the COM group. The median follow-up time was 22 months. No significant differences were observed between the two groups in any of the oncological outcome parameters: local recurrence-free survival between COM and ENM groups (97.8% vs. 97.8%; HR, 0.63; 95%CI, 0.14–2.90; P = 0.50); regional recurrence-free survival between COM and ENM groups (98.8% vs. 99.3%; HR, 1.24; 95%CI, 0.17–9.22; P = 0.84); distant metastasis-free survival between COM and ENM groups(97.3% vs. 98.5%; HR, 1.70; 95%CI, 0.47–6.11; P = 0.48); disease free survival between COM and ENM groups(94.9% vs. 96.3%; HR, 1.24; 95%CI, 0.50–3.10; P = 0.66); overall survival between COM and ENM groups(99.3% vs. 99.3%; HR, 0.76; 95%CI, 0.07–8.72; P = 0.81).

Conclusions

In this multicenter cohort study, ENM group showed no different oncological outcomes from COM group. ENM could be used as a viable alternative to COM for eligible breast cancer patients.

Introduction

Breast cancer persists as the most frequent malignancy in women across the globe, contributing substantially to healthcare system burdens and compromising patient well-being [1–6]. Considerable progress has been made in the surgical management of breast cancer over the past several decades, with an increasing emphasis on balancing oncological efficacy, procedural safety, and patient-centered outcomes, including cosmetic satisfaction and quality of life [7–11]. Conventional open mastectomy (COM), involving the removal of the nipple-areolar complex (NAC) and a significant portion of breast skin, has long been a well-established treatment for eligible patients [12]. However, this procedure is associated with inherent limitations, including prominent scarring, potential restrictions in upper extremity function, and adverse effects on body image and psychological wellbeing [13, 14], prompting the development of more refined surgical approaches.

In response to these unmet needs, nipple-sparing mastectomy (NSM) has arisen as an important alternative for selected patients, preserving the NAC and breast skin without compromising oncological safety compared to conventional mastectomy [15–17]. Building upon this advancement, minimally invasive techniques, particularly endoscopic nipple-sparing mastectomy (ENM), have gained increasing attention. ENM offers the potential to conceal incisions (e.g., within the axillary fold), minimize tissue trauma, and further enhance cosmetic outcomes compared to conventional open NSM [18, 19]. By leveraging magnified visualization and precise dissection, ENM aims to achieve adequate tumor resection while preserving critical vascular and neural networks, which may reduce complications such as NAC necrosis [20].

Notably, most existing research on ENM has focused on comparisons with conventional NSM [21–23], demonstrating comparable short-term complication rates and superior cosmetic satisfaction for the minimally invasive approach. Nevertheless, a notable deficiency remains in literature: the absence of direct comparative data between ENM and COM, particularly regarding long-term oncological outcomes. Actually, COM continues to be widely used in clinical practice, especially for patients with specific disease characteristics or preferences [24], making it essential to evaluate the performance of ENM against this benchmark to inform clinical decision-making.

Several studies are carrying out or underway to address this gap, however, the existing research on ENM is predominantly based on single-center experiences [25–30]. While these investigations provide valuable initial insights, they may be influenced by institutional surgical practices, surgeon expertise, and homogeneous patient demographics, potentially limiting the generalizability of findings to diverse clinical settings. Therefore, there is a clear need for data from multicenter cohorts, reflecting real-world variability, strengthen the external validity of comparative analyses, and, most importantly, provide critical insights into long-term oncological prognosis - a core endpoint for evaluating cancer treatment efficacy.

To address these unresolved questions, we conducted a retrospective multicenter cohort study comparing ENM and COM in patients with early breast cancer. By integrating data from multiple institutions, we aimed to enhance the generalizability of our results.

Method

Patients and design

This multicenter retrospective cohort study enrolled 2,314 patients who had undergone mastectomy for early-stage breast cancer between September 1, 2021, and January 30, 2025, at three institutions: Shanghai General Hospital affiliated Shanghai Jiao Tong University School of Medicine, International Peace Maternity & Child Health Hospital (Shanghai Jiao Tong University School of Medicine), and Jinshan Hospital of Fudan University. Eligible patients were those aged 18–75 years, diagnosed with stage 0–III breast cancer, and treated with either endoscopic nipple-sparing mastectomy (ENM) or conventional open mastectomy (COM). Patients were excluded if they had incomplete baseline data, underwent bilateral surgery, had a history of other malignancies within the preceding five years, presented with nipple or skin involvement on physical examination or MRI, experienced severe complications within the past six months, were receiving immunosuppressive therapy post-organ transplantation, were pregnant or breastfeeding, or required long-term steroid use. After applying the exclusion criteria, 1,722 patients remained in the cohort, with 136 in ENM group and 1,586 in COM group. In COM group watch you Baby Oh, patients received either modified radical mastectomy or total mastectomy combined with sentinel lymph node biopsy. Axillary lymph node dissection was indicated in cases of positive sentinel lymph node findings. In ENM group, patients first underwent sentinel lymph node biopsy or direct axillary lymph node dissection via an axillary skin-crease incision and those with positive sentinel lymph nodes then received complementary axillary lymph node dissection. Subsequently, endoscopic subcutaneous mastectomy was performed through the same axillary incision, and biopsy of the retroareolar tissue was conducted to ensure no residual tumor. Most patients in the ENM group underwent immediate implant reconstruction (134/136).

Pathological staging for each patient was defined as the stage recorded at the time of initial diagnosis, in accordance with the eighth edition of the American Joint Committee on Cancer (AJCC) TNM staging system. The systemic therapy was based on NCCN Guidelines for Breast Cancer, which is determined by the tumor’s molecular subtypes. For Hormone Receptor-Positive (HR+) patients, endocrine therapy is received for nearly all patients. Ovarian function suppression (OFS) combined with either tamoxifen or an aromatase inhibitor (AI) was recommended for higher-risk patients. For patients with high-risk, node-positive disease, a 2-year course of a CDK4/6 inhibitor was recommended as an adjunct. Chemotherapy was added based on the risk of recurrence, which was assessed using validated genomic assays in node-negative or select node-positive cases, along with traditional clinicopathologic factors. For HER2-Positive patients, anti-HER2 targeted therapy was used for all patients, combined with chemotherapy. The standard regimen is 1 year of adjuvant trastuzumab, typically given concurrently with chemotherapy. For higher-risk lymph node-positive patients, pertuzumab was added to trastuzumab and chemotherapy. For those Non-PCR after neoadjuvant therapy patients, adjuvant ado-trastuzumab emtansine (T-DM1) was recommended. For triple-negative patients, chemotherapy formed the mainstay of systemic treatment. For stage II-III disease, anthracycline- and taxane-based regimens are standard. For high-risk patients, the addition of carboplatin was taken for one year. As to adjuvant radiotherapy, chest wall and nodal radiotherapy were recommended for lymph node positive patients or those with tumors size ≥ 5 cm.

Follow up

Postoperative follow-up commences from the date of surgery and end on October 30th, 2025. For the first two years, evaluations were conducted every three months, followed by semi-annual visits thereafter. Assessment included physical examination, blood tests, tumor marker assays, ultrasound, mammography, magnetic resonance imaging (MRI), and bone scan. Additional investigations such as cranial MRI, PET-CT, or biopsy were performed if necessary. Data collection occurred either at the time of clinical visits or by means of telephone review. The median follow-up time (interquartile range) for the entire cohort was 22 months (10–50), with 23 months (10–45) in COM group and 20 months (10–50) in ENM group. No loss to follow-up occurred among the finally enrolled patients.

Local recurrence was defined as tumor recurrence in the ipsilateral chest wall or breast at the original surgical site. Regional recurrence was defined as recurrence involving the ipsilateral internal mammary, axillary, or infraclavicular/supraclavicular lymph node drainage areas. Distant metastasis was defined as metastatic disease at any site other than those specified for local or regional recurrence. Disease-free survival was calculated from the date of surgery to the date of death from any cause, disease recurrence, or metastasis. Overall survival was calculated from the date of surgery to the date of death from any cause, the date of last follow-up, or the end of the follow-up period. In cases where two or more events occurred on the same date, each event was counted separately.

Propensity score matching

To mitigate selection bias and balance confounders, propensity score matching (PSM) was performed using R software with a 1:3 optimal matching algorithm. The propensity score was estimated based on age, menopausal status, pathological invasive type, hormone receptor status, HER2 status, lymph node status, and pathological stage. Following PSM, the final analysis cohort comprised 544 patients: 136 in ENM group and 408 in COM group (Fig. 1).

Fig. 1.

Flow diagram of patients selection

Statistical analysis

Statistical analyses were performed using R version 4.2.0 (R Foundation for Statistical Computing, Vienna, Austria) and SPSS 25.0 (IBM Corp, Armonk, NY, USA). Kaplan-Meier method was used to plot local recurrence-free survival, regional recurrence-free survival, distant metastasis-free survival, recurrence-free survival, and overall survival curves, which were compared using the log-rank test. Quantitative data are presented as median (IQR) and were compared using the Mann-Whitney U test. Categorical variables were compared using the χ² test or Fisher’s exact test. P < 0.05 indicated significant differences.

Ethical review

This study was performed in compliance with the Declaration of Helsinki and was prospectively registered with the Chinese Clinical Trial Registry under registration number ChiCTR2400094450. The study was approved by institutional ethics board of Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine (Approval number: 2024KS502) and informed consent was taken from all the patients.

Results

Table 1 showed the baseline characteristics of two groups before and after PSM. As shown in Table 1, baseline characteristics were similar between the two groups before PSM with respect to pathology types, pathological stage, and lymph node status. In contrast, patients in the ENM group were significantly younger than those in the COM group, and the ENM group also had a significantly higher proportion of premenopausal patients. Furthermore, the two groups differed significantly with respect to hormone receptor status, HER2 status and molecular subtypes. The number of patients included during each year of this study was also differed. After PSM, the two groups demonstrated balanced baseline characteristics, with no significant differences observed.

Table 1.

The baseline characteristics of COM group and ENM group before and after PSM

	Before PSM, No. (%)			After PSM, No. (%)
Characteristic	COM	ENM	P value	COM	ENM	P value
No.	1727	136		408	136
Age, median (IQR), y	59(49–67)	45(40–51)	< 0.001	45(39.25-50)	45(40–51)	0.565
Year of enrollment
2022	289(16.7)	16(11.8)		43(10.5)	16(11.8)
2023	688(39.8)	37(27.2)	< 0.001	136(33.3)	37(27.2)	0.472
2024	678(39.3)	69(50.7)		199(48.8)	69(50.7)
2025	72(4.2)	14(10.3)		30(7.4)	14(10.3)
Menopause
No	596(34.5)	99(72.8)	< 0.001	295(72.3)	99(72.8)	1
Yes	1131(65.5)	37(27.2)		113(27.7)	37(27.2)
Pathology
Noninvasive	332(19.2)	19(14.0)	0.163	65(15.9)	19(14.0)	0.681
Invasive	1395(80.8)	117(86.0)		343(84.1)	117(86.0)
Hormone receptor status
Negative	256(14.8)	34(25.0)	0.002	77(18.9)	34(25.0)	0.158
Positive	1471(85.2)	102(75.0)		331(81.1)	102(75.0)
HER2 status
Negative	1118(64.7)	76(55.9)		221(54.2)	76(55.9)
Positive	280(16.2)	37(27.2)	0.005	103(25.2)	37(27.2)	0.635
Unsure	329(19.1)	23(16.9)		84(20.6)	23(16.9)
Molecular subtype
HR+HER2-	952(55.1)	66(48.5)		183(44.9)	66(48.5)
HER2+	280(16.2)	37(27.2)	0.012	103(25.2)	37(27.2)	0.660
Triple negative	166(9.6)	10(7.4)		38(9.3)	10(7.4)
Unknown	329(19.1)	23(16.9)		84(20.6)	23(16.9)
Stage
0-I	871(50.4)	60(44.1)		118(46.1)	60(44.1)
II	690(40.0)	65(47.8)	0.199	191(46.8)	65(47.8)	0.887
III	166(9.6)	11(8.1)		29(7.1)	11(8.1)
Lymph node status
Negative	1245(72.1)	98(72.1)	1	300(73.5)	98(72.1)	0.823
Positive	482(27.9)	38(27.9)		108(26.5)	38(27.9)

Survival analysis was conducted and compared between the two groups after PSM. The COM and ENM groups demonstrated no significant differences in local recurrence rate(2.2% vs. 2.2%; P = 0.50), regional recurrence rate(1.2% vs. 0.7%; P = 0.84) and distant metastasis rate(2.7% v 1.5%s; P = 0.48). Figures 2, 3, 4, 5 and 6 presented the Kaplan–Meier curves after PSM between two groups. Figure 2 showed that there was no statistically significant difference in Local recurrence-free survival between COM and ENM groups (97.8% vs. 97.8%; HR, 0.63; 95% CI, 0.14–2.90; P = 0.50). As shown in Fig. 3, no statistically significant difference in regional recurrence-free survival was demonstrated between COM and ENM groups (98.8% vs. 99.3%; HR, 1.24; 95% CI, 0.17–9.22; P = 0.84). Figure 4 reveled distant metastasis-free survival did not differ significantly between COM and ENM groups(97.3% vs. 98.5%; HR, 1.70; 95% CI, 0.47–6.11; P = 0.48) and Fig. 5 showed there was no significatively difference in disease free survival between COM and ENM groups(94.9% vs. 96.3%; HR, 1.24; 95% CI, 0.50–3.10; P = 0.66), although distant metastasis rate and disease-free survival showed numerically favorable trend in the ENM group. Figure 6 demonstrated it showed consistency in overall survival between COM and ENM groups(99.3% vs. 99.3%; HR, 0.76; 95% CI, 0.07–8.72; P = 0.81).

Fig. 2.

The local recurrence-free survival between COM and ENM groups

Fig. 3.

The regional recurrence-free survival between COM and ENM groups

Fig. 4.

The distant metastasis-free survival between COM and ENM groups

Fig. 5.

The disease-free survival between COM and ENM groups

Fig. 6.

The overall survival between COM and ENM groups

Additionally, a post hoc subgroup analysis of DFS was performed on the propensity score-matched cohort according to menopause status, pathology, HR status, HER2 status, stage, and lymph node status. As shown in the forest plot (Fig. 7), no statistically significant differences were observed across any of the subgroups.

Fig. 7.

The forest plot for post hoc subgroup analysis of DFS

Discussion

The present retrospective multicenter study evaluated the oncological outcomes of ENM versus COM in 1722 breast cancer patients, with a 1:3 propensity score-matched design ensuring balanced clinicopathological characteristics between the two groups. After a median follow-up of 22 months, our key findings demonstrated no significant differences in local or regional recurrence rates between ENM and COM. While distant metastasis rate and disease-free survival were numerically favorable in the ENM group and overall survival was comparable across both groups, subgroup analyses failed to identify significant survival disparities by surgical approach across clinicopathological subgroups. These results address a critical gap in the literature by providing real-world multicenter evidence comparing ENM to the traditional surgical benchmark of COM.

The comparable local and regional recurrence rates between ENM and COM align with the core oncological safety profile of NSM established in previous studies. Headon et al.’s systematic review of 12,358 NSM procedures confirmed low recurrence rates comparable to conventional mastectomy [15], and our findings extend this evidence by specifically validating that the minimally invasive endoscopic NSM approach does not compromise local disease control relative to COM. This is particularly reassuring given that ENM preserves the nipple-areolar complex(NAC) and relies on endoscopic visualization for dissection, which have previously raised concerns that this approach might lead to incomplete tumor resection [25]. Our results support the notion that precise endoscopic dissection, combined with strict patient selection, ensures adequate oncological clearance, as reflected by recurrence rates consistent with COM. In comparing endoscopic surgery with traditional open surgery, unlike previous studies that included both breast-conserving surgery and mastectomy in their comparisons, some studies although included relatively large sample sizes, however, they mixed mastectomy and breast-conserving surgery together in their comparisons [9, 27], our study focused on a direct comparison between endoscopic NSM and COM, thereby minimizing the confounding effect of breast-conserving surgery on the outcomes.

In addition to conventional endoscopic approaches, robotic nipple-sparing mastectomy (RNSM) has emerged as another important form of minimal-access NSM. The robotic platform may offer three-dimensional magnified visualization, wristed instrumentation and improved ergonomics, and access through a remote axillary or lateral incision [31]. Although both ENM and RNSM represent promising minimal-access extensions of NSM, their oncological safety should be interpreted within the context of patient selection and institutional expertise [32]. Based on the above reasons, current study focused exclusively on ENM and did not include RNSM, even though both procedures are performed at our center. The oncological safety of RNSM will be addressed in future studies.

Notably, the numerical advantages in distant metastasis rate and DFS observed in the ENM group, while not statistically significant, are clinically meaningful and warrant further exploration. These findings were similar to Zhou et al.’s study [33, 29]. These trends may be attributed to several potential factors. First, ENM’s minimally invasive nature reduces tissue trauma and inflammatory responses and endoscopic operation may reduce tumor compression, which was linked to cancer progression and metastasis. Li et al. previously demonstrated that endoscopic breast surgery is associated with a lower risk of circulating tumor cell-related micrometastasis compared with conventional open surgery [31], suggesting a potential biological mechanism underlying improved metastatic outcomes. Second, the majority of patients(134/136) who underwent ENM in this study received immediate implant-based reconstruction. This approach, combined with the inherent advantage of ENM in preserving NAC and skin, contributed to superior cosmetic outcomes and reduced psychological distress [13, 14]. Patients who experience better aesthetic results and quality of life are more likely to maintain treatment adherence and complete adjuvant therapies [24]. While our study did not directly assess treatment adherence or recovery timelines, this synergistic effect of enhanced cosmesis and streamlined reconstruction could contribute to the numerically favorable outcomes observed in the ENM group - though this hypothesis requires confirmation in future studies.

In the subgroup analysis, although no statistically significant differences were observed in any subgroup, ENM group showed a numerically more pronounced advantage in the HR-positive and HER2-negative subgroups. Notably, in the subgroup analysis based on axillary lymph node status, ENM group demonstrated numerical advantages in both node-positive and node-negative subgroups. In this study, both ENM group and COM group underwent open lymph node surgery. The key difference was the surgical sequence: ENM group first underwent lymph node surgery, followed by endoscopic breast surgery through the axillary incision, whereas COM group underwent breast surgery first, followed by lymph node surgery. Whether this difference in surgical sequence contributed to the observed outcomes remains to be investigated in future studies.

Several limitations of the present study should be acknowledged. First, as a retrospective cohort study, it is inherently susceptible to selection bias, despite the use of propensity score matching to balance groups. Additionally, several potentially relevant criteria, including systemic treatment types and molecular subtypes, were not subjected to PSM in current study. Lastly, the median follow-up of 22 months is still relatively short for assessing long-term oncological outcomes such as late recurrence and OS. Future studies with 5-year or longer follow-up are needed to establish more convincing oncological safety of ENM on early breast cancer.

Despite these limitations, our study has several strengths. To our knowledge, it is the first multicenter retrospective study to compare ENM with COM focusing on oncological outcomes, filling a unique gap in the literature. The large sample size (1722 patients) and 1:3 propensity score matching ensure statistical power and balance between groups, enhancing the reliability of our findings. Additionally, the inclusion of diverse clinicopathological subgroups allows for broader generalizability to real-world clinical settings.

Conclusion

In conclusion, our study demonstrates that ENM achieves comparable local and regional recurrence rates and numerically favorable distant metastasis and DFS outcomes relative to COM, with no compromise in OS. These findings suggest the oncological safety of ENM and support its use as a viable alternative to COM for eligible breast cancer patients. ENM’s potential advantages in metastatic outcomes, combined with its established cosmetic benefits, make it a preferred option for patients seeking NAC preservation without sacrificing oncological efficacy. Future prospective multicenter studies with longer follow-up are needed to validate these findings, assess patient-reported outcomes, and compare different minimal-access platforms, including endoscopic and robotic approaches, under standardized patient selection and surgeon training frameworks.

Authors' contributions

Li Zhu, Wei Wang and Min Wang conceived and designed the study and drafted the main manuscript. Xin Ye, Zhaoqi Zhang, Ye Fang and Yanyan Zhu collected the multicenter's data and follow up the patients. Deyue Liu, Shuning Ding and Qi Wang performed the statistical analysis. Deyue Liu prepared figure 1. Shuning Ding prepared figures 2, 3, 4, 5 and 6. Qi Wang prepared table 1 and figure 7. All authors reviewed the final manuscript.

Funding

This study was supported by Clinical Research Data Sharing and Simulated RCT Project under the Three-Year Action Plan for Promoting Clinical Skills and Clinical Innovation in Municipal Hospitals by Shanghai Shenkang Hospital Development Center (SHDC2024CRI084).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The study was approved by institutional ethics board of Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine (Approval number: 2024KS502) and informed consent was taken from all the patients.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.