Single axillary incision for minimal access (endoscopic or robotic-assisted) nipple sparing mastectomy: axillary crease or lateral chest wall incision? Authors: Cher-Hui Koh, Wen-Pin Kao, Dineshwary Periasammy, Yi-Yuan Lee, Hung-Wen Lai Published: 2026 ABSTRACT Minimal access nipple sparing mastectomy (MA-NSM), including endoscopic- (E-NSM) and robotic-assisted (R-NSM) techniques, enables NSM to be performed through aesthetically favourable incisions without compromising oncologic safety. While early MA-NSM used dual incisions, current practice favours a single axillary incision (SAI), most commonly placed at the axillary crease or along the lateral chest wall corresponding to the level of the nipple–areolar complex (NAC). This is a retrospective study of all patients who underwent MA-NSM via a SAI either the axillary or lateral chest wall approach between January 2012 to June 2025 at a single institute. Clinicopathologic characteristics, surgical outcomes, postoperative complications and patient reported outcome measures (PROMs) were analyzed. The study included 371 MA-NSM cases, of which 222 (59.8%) were performed via an axillary crease incision and 149 (40.2%) via a lateral chest wall incision. For E-NSM, operative time was significantly shorter with the lateral chest wall approach than with the axillary approach (132.2 ± 46.4 vs. 188.3 ± 97.2 min, p < 0.001), whereas no such difference was seen in R-NSM. Delayed wound healing is significantly higher in the lateral chest wall group compared to the axillary crease group (6.7% vs. 0.5%, p < 0.001). PROMs including post-operative sensory outcomes were comparable between the two groups. Shorter operative time with the lateral chest wall approach was noted in E-NSM but not in R-NSM. Both the axillary crease and lateral chest wall approaches remain viable options for MA-NSM with similar postoperative complications and patient satisfaction. Introduction With surgical advancements in recent decades, mastectomy techniques for breast cancer have evolved from the highly invasive Halsted radical mastectomy to contemporary skin- and nipple-sparing mastectomy (NSM), when oncologically feasible. In NSM, the nipple–areolar complex (NAC) is preserved while the underlying breast tissue is completely removed without compromising oncological safety [1]. Among the various incisions used for conventional NSM (C-NSM), the inframammary fold is the most commonly preferred, as it is well concealed and associated with a lower risk of NAC necrosis [2]. Minimal-access nipple-sparing mastectomy (MA-NSM), enables NSM to be performed through smaller, hidden and more aesthetically favorable incisions compared to C-NSM [3]. Multiple studies have demonstrated that MA-NSM is associated with superior aesthetic satisfaction and lower postoperative complication rates compared with C-NSM [4–6], without significant differences in oncologic margin status, disease-free survival, or overall survival [7, 8]. MA-NSM may be carried out via an endoscopic approach or, more recently, with the use of robotic assistance. Even within MA-NSM, there has been a progressive evolution in both the number and placement of surgical incisions. Early techniques employed multiple small port incisions [9–11] followed by dual-incision approaches incorporating a peri-areolar incision for skin-flap development [12, 13]. More recently, the technique has evolved toward a single axillary incision (SAI) [14–16], enabling NSM to be performed through a one hidden incision without any visible scarring on the breast. In addition to NSM, axillary procedures routinely performed in breast cancer surgery may be carried out through the same incision. This single access can also be utilized for breast reconstruction [17]. Although numerous studies have compared C-NSM with MA-NSM, there remains a paucity of studies directly comparing techniques within MA-NSM itself and evaluating whether these technical variations influence patient outcomes [8]. One of the differences in technique of MA-NSM is the placement of the single port. There are 2 common placements for the single port used in MA-NSM which is at the axillary crease (Fig. 1a) or placed slightly inferior on the anterior axillary line at the level of the NAC, referred to as the lateral chest wall incision (Fig. 1f) [18, 19]. Several authors have described performing endoscopic assisted nipple-sparing mastectomy (E-NSM) via an axillary crease incision [14, 16, 20] or a lateral chest wall incision [4, 15, 17]. Similarly, in robotic-assisted nipple-sparing mastectomy (R-NSM), both the axillary crease incisions [21, 22] and lateral chest wall incisions [23–25] have been adopted as preferred sites for single-port placement. We hypothesized that the placement of the incision may influence the technical aspects of the procedure, as well as sensory preservation and aesthetic outcomes. Fig. 1Representative clinical photographs associated with MA-NSM via the axillary approach on the left and lateral chest wall approach on the right. Rows depict (from top to bottom): diagrammatic representation of the incision, intraoperative views, pre-operative appearance of patient, immediate post-operative outcomes, and follow-up results. (a - e) Axillary crease approach: Diagrammatic representation of axillary crease incision, intra-operative views, pre-operative appearance of patient, immediate post-operative outcomes, and follow-up results. (f - j) Lateral chest wall approach: Diagrammatic representation of lateral chest wall incision, intra-operative views, pre-operative appearance of patient, immediate post-operative outcomes, and follow-up results Representative clinical photographs associated with MA-NSM via the axillary approach on the left and lateral chest wall approach on the right. Rows depict (from top to bottom): diagrammatic representation of the incision, intraoperative views, pre-operative appearance of patient, immediate post-operative outcomes, and follow-up results. (a - e) Axillary crease approach: Diagrammatic representation of axillary crease incision, intra-operative views, pre-operative appearance of patient, immediate post-operative outcomes, and follow-up results. (f - j) Lateral chest wall approach: Diagrammatic representation of lateral chest wall incision, intra-operative views, pre-operative appearance of patient, immediate post-operative outcomes, and follow-up results One study which compared the axillary crease and lateral chest wall approaches in performing E-NSM demonstrated that the axillary approach was associated with a longer operative time and a steeper learning curve [26]. Although both incisions are located off the breast, unlike in C-NSM, and are generally non visible from the front, opinions differ regarding their aesthetic advantages. Those who prefer the axillary crease incision consider it superior due to its concealment within the natural axillary fold [27, 28]. Others who favor the lateral approach thinks despite not being within a skin crease, the scar remains hidden by the patient’s arm and undergarments [417, 29]. Farr et al. [21] and Toesca et al. [30], who performed R-NSM via axillary crease and a lower inferior-lateral approach respectively reported better sensory preservation compared to the C-NSM [31]. Despite these described advantages, they are largely based on single-approach reports, with few studies providing a direct comparison between the two techniques. Therefore, we conducted a retrospective analysis to compare these two approaches, focusing on clinical outcomes, post-operative complications and patient-reported outcome measures (PROMs). Methods Study design Patients A retrospective review was conducted of 440 patients who underwent MA-NSM (both E-NSM and R-NSM), at Changhua Christian Hospital (CCH), a tertiary referral center in central Taiwan, between 1 January 2012 and 30 June 2025. Among these, 59 patients underwent bilateral MA-NSM in the same operative setting, while 5 patients underwent contralateral MA-NSM for metachronous breast cancer. In total, 504 MA-NSM procedures were performed. After excluding cases which are non-SAI (N = 115), history of prior surgery on the same breast (N = 10) and cases with ipsilateral recurrent breast cancer(N = 8), a total of 371 cases of MA-NSM were included in the analysis. This included 264 (71.2%) cases of E-NSM and 107 (28.8%) cases of R-NSM. There were 222(59.8%) cases of MA-NSM performed via the axillary incision and 149 (40.2%) cases via the lateral chest wall approach (Fig. 2). Fig. 2Flow chart of patients who underwent MA –NSM. NSM: Nipple sparing mastectomy Flow chart of patients who underwent MA –NSM. NSM: Nipple sparing mastectomy A prospectively maintained comprehensive breast surgery database served as the source of data. This included clinicopathological variables such as patient age, body mass index (BMI), tumor size, disease stage, surgical procedures performed, histological findings, use of neoadjuvant and adjuvant therapies. Outcome measurements The primary outcomes of this study included perioperative parameters and post-operative complications. Peri-operative variables evaluated comprised operative duration for MA-NSM, intraoperative blood loss, duration of hospital stay, and rates of malignant involvement of surgical margins and the sub-nipple region. Surgical margins were considered clear if no tumor cells were present at the inked margin in cases of invasive carcinoma. For pure ductal carcinoma in situ (DCIS), a margin of at least 2 mm was defined as a clear resection margin [32]. Sub-nipple tissue biopsy for intra-operative frozen section were sent for all the patients who underwent MA-NSM. Nipple excision was performed in patients with positive sub-nipple biopsy results, and these patients were subsequently excluded from the study. Complications occurring within 3 months post-surgery were considered related to the initial procedure and were recorded. These included wound-related complications such as infection and delayed wound healing, hematoma and seroma formation, blistering, ischemia of the skin flap and NAC, as well as implant loss. The viability of the NAC was graded into 3 groups: full recovery without loss of nipple volume (transient ischaemia), partial loss of nipple volume (partial recovery), and total loss of nipple volume (total necrosis). The Clavien–Dindo (CD) classification was used to grade and report postoperative complications [33]. Indications for MA-NSM Patients with tumors measuring around 5 cm, without evidence of skin or chest wall invasion or multiple axillary lymph nodes involvement, were considered eligible for MA-NSM [16, 34]. Contraindications for MA-NSM included T4 tumors, apparent NAC involvement, inflammatory and locally advanced breast cancer [16, 34]. Single axillary incision (SAI) In our institution, we started performing E-NSM using the dual incisions with one incision at the axillary crease and another at the peri-areolar region before gradual transition to the single-axillary-incision hybrid technique of E-NSM(SAI-E-NSM) in May 2014. The techniques for this SAI-E-NSM have been described in details previously [16, 35]. We initially performed SAI-E-NSM predominantly through an oblique incision made near the axillary crease (axillary crease incision) (Fig. 1a). From 2018 onwards, an increasing number of cases were carried out using a more inferior incision along the anterior axillary line at the level of the NAC (referred to as the lateral chest wall incision) (Fig. 1f). We started performing R-NSM in 2017 using the axillary crease approach, followed by adoption of the lateral chest wall incision in 2018 [22]. There were no strict criteria guiding the choice of SAI. Both the incisions measured approximately 2.5 to 5 cm, with the exact length determined by the volume of breast tissue to be excised. Patient’s satisfaction score Evaluation regarding PROMs on aesthetics were carried out in the outpatient clinic 3 months post-surgery after the wound has healed using a survey developed in our institution. The survey consisted of nine questions, each evaluated using a four-point Likert scale: 1 (unsatisfied), 2 (fair), 3 (satisfied), and 4 (very satisfied). We also incorporated four scales from The BREAST-Q 2.0 Mastectomy modules questionnaire which were the Satisfaction with Breasts, Psychosocial, Sexual, and Physical Well-being: Chest scales. Equivalent Rasch-transformed scores, ranging from 0 to 100, were derived from the summated scores, with higher values indicating better outcomes. Nipple and skin sensation assessment Patients who underwent MA-NSM were invited to undergo evaluation of nipple and breast skin sensation by trained clinical research nurses during follow-up visits at 3 months postoperatively. The healthy breast was first gently touched to establish a baseline for a normal sensation. Sensation over the operated side was then evaluated by light touch at five specific regions: the NAC and the upper, medial, inferior, and lateral breast skin. Patients were then asked to rate their sensation in each area using a four-point scale (grade 0: absent sensation; grade I: numbness; grade II: fairly sensate and grade III: normal sensation) relative to the normal side. For patients who had bilateral mastectomy, where no normal contralateral control is available, the same assessment was performed and sensation was compared with adjacent non-operated skin, such as the upper sternum. Statistical analyses Continuous data were presented as either mean with standard deviation or median with interquartile range, depending on the distribution of the data. Between-group comparisons were conducted using the independent samples t-test for normally distributed variables and the Mann–Whitney U test for variables that did not follow a normal distribution. Categorical data were reported as frequencies with corresponding percentages, and differences between groups were assessed using the chi-square test. This study received approval from the Institutional Review Board (IRB) of CCH (IRB No. 231217). Written informed consent was obtained from all patients whose photographs are included in this report. Results A total of 371 cases of MA-NSM were included in the analysis. This included 264 (71.2%) cases of E-NSM and 107 (28.8%) cases of R-NSM. There were 222 (59.8%) cases of MA-NSM performed via the axillary incision and 149 (40.2%) cases via the lateral chest wall approach (Fig. 2). There were no significant differences in age or BMI between patients in the axillary crease and lateral chest wall incision groups. Similarly, no significant differences were observed in mean sonographic tumor size (2.5 ± 1.3 cm vs. 2.8 ± 2 cm, p = 0.14) and pathological tumor size (2.6 ± 2.1 cm vs. 2.7 ± 2.5 cm, p = 0.77) in the axillary crease and lateral chest wall incision groups, respectively. Both groups were comparable in terms of tumor grade and immunohistochemistry. In our cohort, HER2 status was unavailable in 114 patients, comprising 73 patients with DCIS, in whom HER2 testing was not routinely performed, 38 non-malignant cases, and 3 patients with unknown HER2 status. A significant proportion of patients in the axillary crease incision group had neoadjuvant chemotherapy compared to the lateral chest wall group (30.3% vs. 20%, p < 0.05) The other baseline characteristics of both groups are summarized in Table 1. Table 1Baseline characteristics of patients who underwent MA-NSM via axillary crease or lateral chest wall approachesTotal MinimalAccess, N = 371(%)Axillary crease incision group,N = 222 (%)Lateral chest wall incision,N = 149 (%)P-valueAge, years (mean ± SD)50.6 ± 10.550.7 ± 10.850.4 ± 10.20.82BMI, kg/m2 (mean ± SD)23.4 ± 3.923.1 ± 3.623.9 ± 4.10.06LocationRight195 (52.6)116 (52.3)79 (53)0.97Left176 (47.4)106 (47.7)70 (47)Sonographic tumor size (cm)(mean ± SD)2.6 ± 1.62.5 ± 1.32.8 ± 2.00.14Pathological tumor size (cm)(mean ± SD)2.6 ± 2.32.6 ± 2.12.7 ± 2.50.77Histopathology, NA = 0In situ76 (20.5)39 (17.6)37 (24.8)0.21IDC237 (63.9)145 (65.3)92 (61.7)ILC19 (5.1)14 (6.3)5 (3.4)Others1 (0.3)0 (0)1 (0.7)No malignancy38 (10.2)24 (10.8)14 (9.4)Tumour size (T), NA = 38is76 (22.8)39 (19.7)37 (27.4)0.10ypT027 (8.1)20 (10.1)7 (5.2)T1110 (33.0)69 (34.8)41 (30.4)T2118 (35.4)70 (35.4)48 (35.6)T32 (0.6)0 (0)2 (1.5)Lymph nodes (N), NA = 38ypN0/N0250 (75.1)146 (73.7)104 (77.0)0.79N166 (19.8)41 (20.7)25 (18.5)N213 (3.9)9 (4.5)4 (3.0)N34 (1.2)2 (1.0)2 (1.5)Metastases (M), NA = 38Yes3 (0.9)2 (1.0)1 (0.7)> 0.99No330 (99.1)196 (99.0)134 (99.3)Stage, NA = 38076 (22.8)39 (19.9)37 (27.4)0.38pCR24 (7.2)17 (8.7)7 (5.2)I85 (25.5)54 (27.6)31 (23.0)II129 (38.7)76 (38.4)53 (39.3)III16 (4.8)10 (5.1)6 (4.4)IV3 (0.9)2 (1.0)1 (0.7)Grade, NA = 50I61 (19.0)37 (19.6)24 (18.2)0.88II194 (60.4)112 (59.3)82 (62.1)III66 (20.6)40 (21.2)26 (19.7)ER status, NA = 42Positive259 (78.7)152 (77.2)107 (81.1)0.48Negative70 (21.3)45 (22.8)25 (18.9)PR status, NA = 42Positive211 (64.1)121 (61.4)90 (68.2)0.26Negative118 (35.9)76 (38.6)42 (31.8)HER2 status, NA = 114Positive45 (13.6)30 (15.3)15 (11.2)0.11Negative212 (64.2)130 (66.3)82 (61.2)Ki67, NA = 129≤ 20174 (71.9)102 (69.9)72 (75.0)0.47> 2068 (28.1)44 (30.1)24 (25.0)Subtype, NA = 114ER+/PR+,Her2-180 (70.0)110 (68.8)70 (72.2)0.89ER+/PR+,Her2+22 (8.6)14 (8.8)8 (8.2)ER-/PR-, Her2+23 (8.9)16 (10.0)7 (7.2)TNBC32 (12.5)20 (12.5)12 (12.4)Neoadjuvant chemotherapy, NA = 38Yes87 (26.1)60 (30.3)27 (20)< 0.05No246 (73.9)138 (69.7)108 (80)Adjuvant chemotherapy, NA = 38Yes143 (42.9)85 (42.9)58 (43)> 0.99No190 (57.1)113 (57.1)77 (57)Adjuvant radiotherapy, NA = 38Yes97 (29.1)62 (31.3)35 (25.9)0.35No236 (70.9)136 (68.7)100 (74.1)Adjuvant hormonal, NA = 38Yes235 (70.6)139 (70.2)96 (71.1)0.96No98 (29.4)59 (29.8)39 (28.9)NA - Not applicable, pCR - Pathological complete response, ER - Estrogen ReceptorPR - Progesterone Receptor, HER2 – Human Epidermal Growth Factor Receptor 2 Table 1Baseline characteristics of patients who underwent MA-NSM via axillary crease or lateral chest wall approachesTotal MinimalAccess, N = 371(%)Axillary crease incision group,N = 222 (%)Lateral chest wall incision,N = 149 (%)P-valueAge, years (mean ± SD)50.6 ± 10.550.7 ± 10.850.4 ± 10.20.82BMI, kg/m2 (mean ± SD)23.4 ± 3.923.1 ± 3.623.9 ± 4.10.06LocationRight195 (52.6)116 (52.3)79 (53)0.97Left176 (47.4)106 (47.7)70 (47)Sonographic tumor size (cm)(mean ± SD)2.6 ± 1.62.5 ± 1.32.8 ± 2.00.14Pathological tumor size (cm)(mean ± SD)2.6 ± 2.32.6 ± 2.12.7 ± 2.50.77Histopathology, NA = 0In situ76 (20.5)39 (17.6)37 (24.8)0.21IDC237 (63.9)145 (65.3)92 (61.7)ILC19 (5.1)14 (6.3)5 (3.4)Others1 (0.3)0 (0)1 (0.7)No malignancy38 (10.2)24 (10.8)14 (9.4)Tumour size (T), NA = 38is76 (22.8)39 (19.7)37 (27.4)0.10ypT027 (8.1)20 (10.1)7 (5.2)T1110 (33.0)69 (34.8)41 (30.4)T2118 (35.4)70 (35.4)48 (35.6)T32 (0.6)0 (0)2 (1.5)Lymph nodes (N), NA = 38ypN0/N0250 (75.1)146 (73.7)104 (77.0)0.79N166 (19.8)41 (20.7)25 (18.5)N213 (3.9)9 (4.5)4 (3.0)N34 (1.2)2 (1.0)2 (1.5)Metastases (M), NA = 38Yes3 (0.9)2 (1.0)1 (0.7)> 0.99No330 (99.1)196 (99.0)134 (99.3)Stage, NA = 38076 (22.8)39 (19.9)37 (27.4)0.38pCR24 (7.2)17 (8.7)7 (5.2)I85 (25.5)54 (27.6)31 (23.0)II129 (38.7)76 (38.4)53 (39.3)III16 (4.8)10 (5.1)6 (4.4)IV3 (0.9)2 (1.0)1 (0.7)Grade, NA = 50I61 (19.0)37 (19.6)24 (18.2)0.88II194 (60.4)112 (59.3)82 (62.1)III66 (20.6)40 (21.2)26 (19.7)ER status, NA = 42Positive259 (78.7)152 (77.2)107 (81.1)0.48Negative70 (21.3)45 (22.8)25 (18.9)PR status, NA = 42Positive211 (64.1)121 (61.4)90 (68.2)0.26Negative118 (35.9)76 (38.6)42 (31.8)HER2 status, NA = 114Positive45 (13.6)30 (15.3)15 (11.2)0.11Negative212 (64.2)130 (66.3)82 (61.2)Ki67, NA = 129≤ 20174 (71.9)102 (69.9)72 (75.0)0.47> 2068 (28.1)44 (30.1)24 (25.0)Subtype, NA = 114ER+/PR+,Her2-180 (70.0)110 (68.8)70 (72.2)0.89ER+/PR+,Her2+22 (8.6)14 (8.8)8 (8.2)ER-/PR-, Her2+23 (8.9)16 (10.0)7 (7.2)TNBC32 (12.5)20 (12.5)12 (12.4)Neoadjuvant chemotherapy, NA = 38Yes87 (26.1)60 (30.3)27 (20)< 0.05No246 (73.9)138 (69.7)108 (80)Adjuvant chemotherapy, NA = 38Yes143 (42.9)85 (42.9)58 (43)> 0.99No190 (57.1)113 (57.1)77 (57)Adjuvant radiotherapy, NA = 38Yes97 (29.1)62 (31.3)35 (25.9)0.35No236 (70.9)136 (68.7)100 (74.1)Adjuvant hormonal, NA = 38Yes235 (70.6)139 (70.2)96 (71.1)0.96No98 (29.4)59 (29.8)39 (28.9)NA - Not applicable, pCR - Pathological complete response, ER - Estrogen ReceptorPR - Progesterone Receptor, HER2 – Human Epidermal Growth Factor Receptor 2 Baseline characteristics of patients who underwent MA-NSM via axillary crease or lateral chest wall approaches Sonographic tumor size (cm) (mean ± SD) Pathological tumor size (cm) (mean ± SD) NA - Not applicable, pCR - Pathological complete response, ER - Estrogen Receptor PR - Progesterone Receptor, HER2 – Human Epidermal Growth Factor Receptor 2 Peri-operative outcome A total of 89.8% of procedures were performed for the therapeutic management of breast cancer, while the remainder were prophylactic. The cases in the lateral chest wall group had a higher, although non-significant breast volume removed compared to the axillary crease incision group (330.7 ± 162 g vs. 299.5 ± 131.8 g, p = 0.05). On univariable analysis, the mean operative time was comparable between the axillary crease and lateral chest wall incision groups (177.2 ± 78.1min vs. 172.3 ± 67.6 min, respectively; p = 0.52). However, after adjusting for type of minimal access surgery, indication for mastectomy, age, BMI, year of surgery and weight of specimen, the lateral chest wall approach was independently associated with a significantly shorter operative time compared to the axillary crease approach (β=-78.15 min, 95% CI: -109.91 to -46.39, p < 0.001) (Table 4). A significant interaction was observed between surgical approach and method (p = 0.011), indicating that the operating time of the different approaches differed between endoscopic and robotic procedures. Subgroup analyses were therefore performed. In the subset analysis of E-NSM cases, operative time was significantly shorter with the lateral chest wall approach than with the axillary approach (132.2 ± 46.4 min vs. 188.3 ± 97.2 min, p < 0.001). However, no significant difference in operative time was found between the two approaches in the R-NSM subgroup (183.7 ± 68.5 min vs. 171.7 ± 66.2 min; p = 0.15) (Table 2). Table 2Comparison of operative parameters and postoperative complications between axillary crease and lateral chest wall approaches in MA-NSMAxillary crease incision group,N = 222 (%)Lateral chest wall incision,N = 149 (%)P-valueType of mastectomyTherapeutic198 (89.2)135 (90.6)0.79Prophylactic24 (10.8)14 (9.4)Type of minimal access surgeryEndoscopic148 (66.7)116 (77.9)0.03Robotic74 (33.3)33 (22.1)Breast weight, g (mean ± SD)299.5 ± 131.8330.7 ± 162.00.05Axillary SurgeryYes193 (97.5)130 (97.0)> 0.99No5 (2.5)4 (3.0)Type of Axillary SurgerySLNB152 (78.8)106 (81.5)0.77ALND5 (2.6)4 (3.1)SLNB followed by ALND36 (18.7)20 (15.4)Operating time (min)MA-NSM (Overall)177.2 ± 78.1172.3 ± 67.60.52Endoscopic188.3 ± 97.2132.2 ± 46.4< 0.001Robotic171.7 ± 66.2183.7 ± 68.50.15Intra-operative blood loss (ml)37.0 ± 26.637.0 ± 27.8> 0.99Post operative hospital stay, days (mean ± SD)6.1 ± 1.86.7 ± 2.80.03Nipple involvementYes7 (3.6)3 (2.2)0.54No189 (96.4)132 (97.8)Margin involvementYes11 (5.6)4 (3.0)0.39No186 (94.4)131 (97.0)Follow up time, months (mean ± SD)36.7 ± 22.942 ± 36.70.11ComplicationsDelayed wound healing1 (0.5)10 (6.7)< 0.001NAC ischaemiaGrade Ⅰ9 (4.1)6 (4.0)0.79Grade Ⅱ10 (4.5)6 (4.0)Grade Ⅲ0 (0)1 (0.7)Seroma42 (18.9)26 (17.6)0.85Blister from thermal injury4 (1.8)3 (2.0)> 0.99Small partial skin flap ischaemia/ necrosis11 (5.0)4 (2.7)0.41Haematoma1 (0.5)4 (2.7)0.16Implant loss2 (0.9)0 (0)0.52Wound infection3 (1.4)2 (1.3)> 0.99Clavien-Dindo Classification Score0157 (71.0)93 (63.3)0.36Ⅰ56 (25.3)43 (29.3)Ⅱ4 (1.8)6 (4.1)Ⅲa3 (1.4)4 (2.7)Ⅲb1 (0.5)1 (0.7)SLNB- Sentinel Lymph Node Biopsy; ALND -Axillary Lymph Node Dissection, MA-NSM-Minimal Access Nipple Sparing Mastectomy Table 2Comparison of operative parameters and postoperative complications between axillary crease and lateral chest wall approaches in MA-NSMAxillary crease incision group,N = 222 (%)Lateral chest wall incision,N = 149 (%)P-valueType of mastectomyTherapeutic198 (89.2)135 (90.6)0.79Prophylactic24 (10.8)14 (9.4)Type of minimal access surgeryEndoscopic148 (66.7)116 (77.9)0.03Robotic74 (33.3)33 (22.1)Breast weight, g (mean ± SD)299.5 ± 131.8330.7 ± 162.00.05Axillary SurgeryYes193 (97.5)130 (97.0)> 0.99No5 (2.5)4 (3.0)Type of Axillary SurgerySLNB152 (78.8)106 (81.5)0.77ALND5 (2.6)4 (3.1)SLNB followed by ALND36 (18.7)20 (15.4)Operating time (min)MA-NSM (Overall)177.2 ± 78.1172.3 ± 67.60.52Endoscopic188.3 ± 97.2132.2 ± 46.4< 0.001Robotic171.7 ± 66.2183.7 ± 68.50.15Intra-operative blood loss (ml)37.0 ± 26.637.0 ± 27.8> 0.99Post operative hospital stay, days (mean ± SD)6.1 ± 1.86.7 ± 2.80.03Nipple involvementYes7 (3.6)3 (2.2)0.54No189 (96.4)132 (97.8)Margin involvementYes11 (5.6)4 (3.0)0.39No186 (94.4)131 (97.0)Follow up time, months (mean ± SD)36.7 ± 22.942 ± 36.70.11ComplicationsDelayed wound healing1 (0.5)10 (6.7)< 0.001NAC ischaemiaGrade Ⅰ9 (4.1)6 (4.0)0.79Grade Ⅱ10 (4.5)6 (4.0)Grade Ⅲ0 (0)1 (0.7)Seroma42 (18.9)26 (17.6)0.85Blister from thermal injury4 (1.8)3 (2.0)> 0.99Small partial skin flap ischaemia/ necrosis11 (5.0)4 (2.7)0.41Haematoma1 (0.5)4 (2.7)0.16Implant loss2 (0.9)0 (0)0.52Wound infection3 (1.4)2 (1.3)> 0.99Clavien-Dindo Classification Score0157 (71.0)93 (63.3)0.36Ⅰ56 (25.3)43 (29.3)Ⅱ4 (1.8)6 (4.1)Ⅲa3 (1.4)4 (2.7)Ⅲb1 (0.5)1 (0.7)SLNB- Sentinel Lymph Node Biopsy; ALND -Axillary Lymph Node Dissection, MA-NSM-Minimal Access Nipple Sparing Mastectomy Comparison of operative parameters and postoperative complications between axillary crease and lateral chest wall approaches in MA-NSM SLNB- Sentinel Lymph Node Biopsy; ALND -Axillary Lymph Node Dissection, MA-NSM-Minimal Access Nipple Sparing Mastectomy Post-operative complications There were no significant differences in nipple or margin involvement between the axillary crease and lateral chest wall incision groups. Nipple involvement occurred in 7/222 (3.6%) versus 3/149 (2.2%) cases (p = 0.54), while margin involvement was observed in 11/222 (5.6%) versus 4/149 (3.0%) cases (p = 0.39), respectively. Cases performed via lateral chest wall incision had a significant rate of delayed wound healing compared to the axillary crease approach (6.7% vs. 0.5%, p < 0.001). Both incision groups have similar rate of NAC ischemia (8.6% vs. 8.7%, p = 0.79) and the only case of Grade III NAC ischemia was observed in the lateral chest wall incision group. Blister from thermal injury and partial skin flap necrosis were comparable between both groups. Most of the postoperative complications were Clavien–Dindo (CD) grade I (25.3% in the axillary group vs. 29.3% in the lateral chest wall group), with one CD grade IIIb complication observed in each group. No other significant differences in post-operative complications were identified between the two groups (Table 2). After adjusting for type of minimal access surgery, indication for mastectomy, age, BMI, year of surgery, and specimen weight, there was no significant difference in overall complications between the groups (OR = 1.11 95% CI: 0.67 to 1.84, p = 0.691) (Table 4). Patient-reported outcome measurements (PROMs) A total of 133 patients who underwent MA-NSM via an axillary crease incision and 102 patients via a lateral chest wall incision responded to PROMs questionnaires derived from the BREAST-Q 2.0 Mastectomy module and a 9-item institutional survey. The two groups demonstrated no significant differences in Satisfaction with Breasts or Physical Well-Being (Chest). Similarly, outcomes for the Psychosocial and Sexual Well-Being scales were comparable between the two approaches. All these findings were summarized in Table 3. Table 3Comparison of patient-reported outcome measures between axillary crease and lateral chest wall approaches in MA-NSMBreast-Q Version 2.0Axillary crease incision group, N = 133Lateral chest wall incision, N = 102P-valueSatisfaction with Breast59.5 ± 14.361.0 ± 17.00.48Psychosocial Well-Being64.0 ± 17.564.4 ± 17.80.86Sexual Well-Being47.3 ± 19.949.3 ± 21.30.47Physical Well-Being: Chest63.4 ± 17.765.9 ± 15.10.24 A questionnaire surveying aesthetic outcomes and satisfaction in our institute Axillary incision group, N = 133Lateral chest wall incision, N = 102P-valuePreoperative Breast Satisfaction[Q1] How you look in the mirror clothed?3.1 ± 0.53.1 ± 0.60.86[Q2] How comfortably your bras fit?3.1 ± 0.53.1 ± 0.50.93[Q3] Being able to wear clothing that is more fitted?2.9 ± 0.63.0 ± 0.60.50[Q4] How you look in the mirror unclothed?2.8 ± 0.62.9 ± 0.60.50Postoperative Breast Satisfaction[Q1] postoperative breast appearance satisfaction–with clothes3.1 ± 0.73.1 ± 0.70.65[Q2] postoperative breast appearance satisfaction–withoutclothes2.8 ± 0.82.8 ± 0.70.91[Q3] postoperative bilateral breast size satisfaction2.8 ± 0.82.8 ± 0.70.76[Q4] postoperative bilateral breast symmetry satisfaction2.8 ± 0.92.9 ± 0.70.69[Q5] postoperative nipple-areola position satisfaction3.2 ± 0.73.0 ± 0.70.04[Q6] scar appearance satisfaction3.2 ± 0.73.1 ± 0.70.54[Q7] scar length satisfaction3.2 ± 0.73.2 ± 0.60.92[Q8] surgical wound position satisfaction3.2 ± 0.73.2 ± 0.60.72[Q9] If you could choose again, would you choose the same operation again Yes68 (89.5)70 (93.3)0.54 No2 (2.6)0 (0.0) Unsure6 (7.9)5 (6.7)Willingness to choose the same surgical procedure under the same medical conditionsBreast Reconstruction Surgery0.07 Yes36 (90)14 (70) No4 (10)6 (30)Nipple-Sparing Mastectomy0.37 Yes46 (85.2)20 (76.9) No8 (14.8)6 (23.1)Minimally Invasive Surgery0.33 Yes49 (87.5)21 (77.8) No7 (12.5)6 (22.2)Pain IndexUpper arm pain index2.0 ± 2.41.8 ± 2.20.42Chest and breast pain index2.2 ± 2.22.1 ± 2.20.65Range of Motion IndexShoulder range of motion6.2 ± 3.47.0 ± 2.90.07Upper arm and limb usage in daily life6.5 ± 3.47.0 ± 2.90.30SensationAxillary incision group, N = 133 (%)Lateral chest wall incision, N = 102 (%)P-valueNipple-Areola Complex1.8 ± 0.91.8 ± 0.90.72Upper breast region2.0 ± 0.92.1 ± 0.80.31Lateral breast region2.0 ± 0.92.0 ± 0.70.74Lower breast region2.1 ± 0.82.0 ± 0.80.51Medial breast region2.1 ± 0.92.1 ± 0.80.48 Table 3Comparison of patient-reported outcome measures between axillary crease and lateral chest wall approaches in MA-NSMBreast-Q Version 2.0Axillary crease incision group, N = 133Lateral chest wall incision, N = 102P-valueSatisfaction with Breast59.5 ± 14.361.0 ± 17.00.48Psychosocial Well-Being64.0 ± 17.564.4 ± 17.80.86Sexual Well-Being47.3 ± 19.949.3 ± 21.30.47Physical Well-Being: Chest63.4 ± 17.765.9 ± 15.10.24 A questionnaire surveying aesthetic outcomes and satisfaction in our institute Axillary incision group, N = 133Lateral chest wall incision, N = 102P-valuePreoperative Breast Satisfaction[Q1] How you look in the mirror clothed?3.1 ± 0.53.1 ± 0.60.86[Q2] How comfortably your bras fit?3.1 ± 0.53.1 ± 0.50.93[Q3] Being able to wear clothing that is more fitted?2.9 ± 0.63.0 ± 0.60.50[Q4] How you look in the mirror unclothed?2.8 ± 0.62.9 ± 0.60.50Postoperative Breast Satisfaction[Q1] postoperative breast appearance satisfaction–with clothes3.1 ± 0.73.1 ± 0.70.65[Q2] postoperative breast appearance satisfaction–withoutclothes2.8 ± 0.82.8 ± 0.70.91[Q3] postoperative bilateral breast size satisfaction2.8 ± 0.82.8 ± 0.70.76[Q4] postoperative bilateral breast symmetry satisfaction2.8 ± 0.92.9 ± 0.70.69[Q5] postoperative nipple-areola position satisfaction3.2 ± 0.73.0 ± 0.70.04[Q6] scar appearance satisfaction3.2 ± 0.73.1 ± 0.70.54[Q7] scar length satisfaction3.2 ± 0.73.2 ± 0.60.92[Q8] surgical wound position satisfaction3.2 ± 0.73.2 ± 0.60.72[Q9] If you could choose again, would you choose the same operation again Yes68 (89.5)70 (93.3)0.54 No2 (2.6)0 (0.0) Unsure6 (7.9)5 (6.7)Willingness to choose the same surgical procedure under the same medical conditionsBreast Reconstruction Surgery0.07 Yes36 (90)14 (70) No4 (10)6 (30)Nipple-Sparing Mastectomy0.37 Yes46 (85.2)20 (76.9) No8 (14.8)6 (23.1)Minimally Invasive Surgery0.33 Yes49 (87.5)21 (77.8) No7 (12.5)6 (22.2)Pain IndexUpper arm pain index2.0 ± 2.41.8 ± 2.20.42Chest and breast pain index2.2 ± 2.22.1 ± 2.20.65Range of Motion IndexShoulder range of motion6.2 ± 3.47.0 ± 2.90.07Upper arm and limb usage in daily life6.5 ± 3.47.0 ± 2.90.30SensationAxillary incision group, N = 133 (%)Lateral chest wall incision, N = 102 (%)P-valueNipple-Areola Complex1.8 ± 0.91.8 ± 0.90.72Upper breast region2.0 ± 0.92.1 ± 0.80.31Lateral breast region2.0 ± 0.92.0 ± 0.70.74Lower breast region2.1 ± 0.82.0 ± 0.80.51Medial breast region2.1 ± 0.92.1 ± 0.80.48 Comparison of patient-reported outcome measures between axillary crease and lateral chest wall approaches in MA-NSM [Q2] postoperative breast appearance satisfaction–without clothes Based on the institutional questionnaire, no statistically significant differences were observed in the mean scores for the three scar-related items. Both the axillary crease and lateral chest wall incision groups reported similar levels of satisfaction with scar appearance and length. Satisfaction with surgical wound position also did not differ significantly (3.2 ± 0.7 for axillary crease vs. 3.2 ± 0.6 for lateral chest wall incision, p = 0.72). Additionally, 89.5% of patients with an axillary crease scar and 93.3% of those with a lateral chest wall scar reported that they would still opt for the same procedure if given the chance to choose again. Both groups reported comparable levels of upper arm pain, with no significant difference observed. Similarly, there were no statistically significant differences in pain experienced in the breast and chest regions. Patients in both groups also reported similar post-operative breast sensory outcomes for the NAC and across all breast regions (Table 3). Discussion MA-NSM has underwent an evolution in the number and location of incisions required for placement of the working instruments and camera. Earlier descriptions of E-NSM used three to four access ports of around 0.5 to 1 cm in length [9–11] which were subsequently combined into one single incision near the axilla [12, 13], but still incorporating a small incision near the peri-areolar incision to reach the more difficult area of dissection especially at the inferomedial part of the breast which would otherwise be technically challenging to reach from the axillary incision alone [18]. However, studies have reported a higher risk of NAC necrosis associated with the peri-areolar incision [2, 36]. Subsequent modifications of the technique focused on the use of a SAI, moving away from the peri-areolar approach. This incorporation was facilitated by the adoption of gas insufflation and incorporation of a 3D video endoscope which overcame the limitation of earlier techniques [37]. This SAI is commonly placed either in the axillary crease or the lateral chest wall [18]. Carroll et al. demonstrated that compared to C-NSM, E-NSM had a significant reduction in NAC necrosis but this difference was only evident after excluding dual-incision E-NSM cases from their analysis [8]. Similarly, since Toesca first described R-NSM in 2015 [38], R-NSM has also utilized a single incision which may be positioned either at the axillary crease or along the lateral chest wall [34, 39]. In this study, no significant difference in NAC necrosis rates was observed between the two single-incision approaches, suggesting that variations in lateral incision placement does not significantly affect the risk of NAC necrosis. On univariable analysis, there was no significant difference in operating time between the axillary crease and lateral chest wall approach. However, after adjustment for type of minimal access surgery, indication for mastectomy, age, BMI, year of surgery, and specimen weight, multivariable analysis demonstrated that the lateral chest wall approach was independently associated with a shorter operating time (Table 4). The association between surgical approach and operating time differed between endoscopic and robotic procedures, indicating a significant interaction between approach and method. Table 4Multivariable Analyses of Surgical Outcomes (Operating Time and Complications) Adjusting for Baseline DifferencesOperating Time (min)Overall ComplicationsVariablesCoefficient (95% CI)P-valueAdjusted OR (95% CI)P-valueIncision Approach Axillary CreaseReference-Reference- Lateral Chest Wall-78.15 (-109.91 to -46.39)< 0.0011.11 (0.67 to 1.84)0.691 Type of minimal access surgery Endoscopic (E-NSM)Reference-Reference- Robotic (R-NSM)-30.47 (-52.93 to -8.02)0.0110.82 (0.44 to 1.53)0.534 Interaction Term Approach × Method75.80 (40.83 to 110.77)< 0.001 Not included - Multivariable Covariates Type of mastectomy ProphylacticReference-Reference- Therapeutic-17.56 (-29.82 to -5.31)0.0080.60 (0.27 to 1.37)0.227 Age-0.08 (-0.84 to 0.68)0.8401.01 (0.99 to 1.04)0.230 BMI (kg/m²)2.08 (-0.08 to 4.25)0.0660.97 (0.89 to 1.05)0.392 Specimen weight (per 100 g)8.49 (2.45 to 14.53)0.0091.32 (1.06 to 1.65)0.012 Year of surgery1.11 (-1.53 to 3.76)0.4141.23 (1.11 to 1.37)< 0.001 Table 4Multivariable Analyses of Surgical Outcomes (Operating Time and Complications) Adjusting for Baseline DifferencesOperating Time (min)Overall ComplicationsVariablesCoefficient (95% CI)P-valueAdjusted OR (95% CI)P-valueIncision Approach Axillary CreaseReference-Reference- Lateral Chest Wall-78.15 (-109.91 to -46.39)< 0.0011.11 (0.67 to 1.84)0.691 Type of minimal access surgery Endoscopic (E-NSM)Reference-Reference- Robotic (R-NSM)-30.47 (-52.93 to -8.02)0.0110.82 (0.44 to 1.53)0.534 Interaction Term Approach × Method75.80 (40.83 to 110.77)< 0.001 Not included - Multivariable Covariates Type of mastectomy ProphylacticReference-Reference- Therapeutic-17.56 (-29.82 to -5.31)0.0080.60 (0.27 to 1.37)0.227 Age-0.08 (-0.84 to 0.68)0.8401.01 (0.99 to 1.04)0.230 BMI (kg/m²)2.08 (-0.08 to 4.25)0.0660.97 (0.89 to 1.05)0.392 Specimen weight (per 100 g)8.49 (2.45 to 14.53)0.0091.32 (1.06 to 1.65)0.012 Year of surgery1.11 (-1.53 to 3.76)0.4141.23 (1.11 to 1.37)< 0.001 Multivariable Analyses of Surgical Outcomes (Operating Time and Complications) Adjusting for Baseline Differences For E-NSM, axillary crease approach seemed to have a significantly longer operating time compared to the lateral chest wall approach (188.3 ± 97.2 versus 132.2 ± 46.4 min, P < 0.001, Table 2). This difference may be partly be attributed to the approach via the axillary crease being primarily performed during the early phase of this study, such that the surgeon has achieved familiarity with E-NSM by the time the lateral approach was adopted. Although an axillary approach facilitates easier access to the upper breast pole, the angulation with narrower field of view and longer viewing distance makes it more technically challenging to dissect the inframammary fold [27]. This was also noted by Li et al. who did a study comparing the learning curve of both approaches and demonstrated that axillary crease approach resulted in longer operating time and learning curve [26]. The authors attributed it to the “chopstick effect”, where the camera shaft may obstruct the surgeon’s instruments, requiring the surgeon to direct the assistant to adjust the camera frequently, thereby prolonging operative time. This is particularly evident when dissecting the lower inner quadrant [14]. Compared with the axillary approach, instruments inserted through the single lateral chest wall port are aligned parallel to the breast without angulation, allowing nearly equal access to the upper and lower breast poles. It increases instrument mobility within a shorter optical working distance, thereby improving visual exposure. Thus, beginners in E-NSM may consider adopting the lateral chest wall approach. While instruments with rigid-tip are used in E-NSM, the flexible robotic arms with articulated tips in R-NSM allow the surgeon to easily navigate the curvature of the breast, thereby overcoming the limitations of E-NSM [23]. The ability of the robotic arms to negotiate difficult curvature within a rigid space may explained the reason no difference in operating time was observed for the R-NSM group regardless of the placement of the single port in our study. Although the R-NSM was able to overcome the technical limitations of E-NSM but it may not necessarily translate to shorter operative time [40]. Some authors have noted that a more posterior incision relative to the mid-axillary line may impair optical visualization, especially around the curvature of the inner breast quadrants [4, 23]. An axillary incision is unavoidable in most breast cancer patients, as it provides direct access for standard axillary staging procedures required in breast cancer surgery such as sentinel lymph node biopsy (SLNB) and axillary lymph node dissection (ALND). Although a separate comparative analysis of axillary procedures performed via each approach was not undertaken in this study, we believe that conventional axillary surgery can be performed more easily through an axillary crease incision than through a lateral chest wall approach. Achieving adequate surgical exposure for conventional ALND often requires extension of the lateral chest incision; however, endoscopic- [41, 42] and robotic- [43] approaches allow ALND to be performed via the same incision without extending it. In a study of patients who underwent R-NSM via a lateral incision, most patients who underwent conventional ALND required the initial incision to be extended for better visualization, whereas those who underwent robotic-assisted ALND did not require additional extension of the incision [43]. In our opinion, the axillary crease incision offers superior direct access for SLNB and, when necessary, ALND compared to the lateral chest wall incision. Apart from preservation of the NAC in NSM, aim of sensory preservation is equally important. Patients who had NSM with preserved sensation of the NAC reported better physical, mental, and sexual well-being [44]. It is important to avoid injuring the lateral cutaneous branch (LCB) and anterior cutaneous branch (ACB) of the fourth intercostal nerve (ICN) during breast surgery as they provide sensory innervation across most of the breast skin and NAC with the LCB being the most consistent contributor to NAC sensation [45, 46]. The LCB of the fourth ICN consistently emerges approximately 8 mm above the fifth rib, medial to the anterior axillary line, and about 33 mm lateral to the pectoral border [47]. Based on this, we suggest that this nerve is at risk of injury in a lateral chest wall approach. Farr et al. noted that, as nipple sensation is primarily derived from the fourth ICN, limiting the lateral incision and relocating it to the axilla may help preserve nipple sensation. In their series of R-NSM performed via an axillary crease incision approach, they reported preservation up to 55% of nipple sensation and almost 100% breast skin sensation post-surgery [21]. Similarly, Wang et al. showed that patients who had E-NSM from an axillary crease incision were significantly less likely to have decreased pressure, temperature and sensation compared to patients who had C-NSM [48]. In our study, there was no significant difference in the post-operative sensory outcomes when we compare patients who underwent surgery via an axillary crease incision and those via lateral wall incision (Table 3). Apart from the LCB of the ICN, the ACB of the fourth ICN which arise medially and traverse in the superficial plane has been demonstrated to provides significant sensory innervation to the breast skin and NAC [46]. Hence, we suggest that although the lateral chest wall incision may risk injuring the LCB, sensory input from the ACB of the fourth ICN may provide compensatory innervation. Therefore, preservation of NAC sensation is likely achievable as long as the incision is placed away from NAC. In a meta-analysis by Carroll et al., E-NSM was associated with a significantly higher patient-reported cosmetic satisfaction compared with C-NSM and this improvement may be attributable to incision-related factors as E-NSM typically employs smaller, more concealed incisions that are aesthetically favorable [8]. In an Asian population with a higher propensity for hyperpigmentation and hypertrophic scarring, lateral incisions were associated with an increased risk of scar hypertrophy [49]. Incision which follows the low-tension skin lines such as the axillary crease is at lower risk of developing hypertrophic scar and keloid [50]. The axillary crease incision can easily be concealed within a natural skin fold. In comparison, the lateral chest wall incision may result in a more noticeable and less aesthetically pleasing scar as it crosses two anatomical regions from the axilla into the outer quadrant of the breast [28]. These findings may support the notion that the axillary crease incision is favored by patients for its aesthetic advantage. In contrast, our findings showed no statistically significant differences were observed between the groups in terms of patient satisfaction with scar appearance, length and location (Table 3). This may be because the lateral chest wall scar, though more visible, can be concealed by undergarments [4, 17], whereas the axillary crease scar may become visible when a bra is worn [43]. Consequently, aesthetic outcomes are subjective and dependent on individual patient preferences. Although delayed wound healing is likely multifactorial, the placement of the axillary crease incision along a low-tension skin line may partly explain the lower rate of delayed wound healing observed in this group (Table 2). In addition, the lateral chest wall is also subjected to more stretching with shoulder movement which may further affect wound healing [49]. Several limitations should be acknowledged. In particular, the retrospective nature of the study may predispose to selection bias. Secondly, a single surgeon (HWL) performed all the cases in this study, and hence, the operative outcomes may not be fully applicable to other surgeons with varying levels of experience in MA-NSM. In addition, this study included the earliest MA-NSM cases, most of which were performed via the axillary crease approach during the initial study period. As the lateral chest wall approach was adopted later, the results may have been influenced by a learning curve effect. We also acknowledge that objective assessment tools were not incorporated in evaluating breast and NAC sensation, which may have influenced the interpretation of the results. The results were only taken at a single time point in time and may not be a true reflection as NAC sensation has been shown to improve over time. Therefore, additional studies are warranted before definitive conclusions regarding sensory outcomes can be established, and future research should incorporate objective assessment methods with longitudinal follow-up. Although this study has certain limitations, it is, to our knowledge, the first to compare axillary crease and lateral chest wall incisions in terms of postoperative complications, sensory outcomes, and patient satisfaction. These findings provide valuable insights for surgeons when selecting between the two most used approaches in MA-NSM. Conclusion Shorter operative time with the lateral chest wall approach was noted in E-NSM but not in R-NSM. Both the axillary crease and lateral chest wall approaches remain viable options for MA-NSM with similar postoperative complications and patient satisfaction.