Feasibility and outcomes of single-incision robotic nipple-sparing mastectomy: a systematic review and meta-analysis.

Introduction
Nipple-sparing mastectomy (NSM) is a surgical approach that aims to remove the breast tissue while preserving the overlying skin and nipple-areola complex (NAC). The NSM is considered oncologically safe for selected breast cancer patients, as well as for prophylactic mastectomy candidates [1, 2]. Preservation of the NAC in NSM allows for improved cosmetic satisfaction and psychological outcomes compared to conventional approaches, such as skin-sparing mastectomy or total mastectomy [3, 4]. However, NSM is often associated with drawbacks, including risks of ischemia and necrosis of the skin flap or NAC, along with technical limitations, such as ergonomic challenges for surgeons, especially in patients with large and ptotic breasts [5, 6].
Robotic-assisted nipple sparing mastectomy (RNSM) appeared as a minimally invasive method, enabling surgeons to conduct complex dissections through small, concealed incisions and using high definition, three-dimensional visualization and articulated instruments [7, 8]. The RNSM approach optimizes the patients’ post-operative complications and the surgeons’ workload burden [8–10]. Despite its favorable outcomes and ergonomic advantage, RNSM is still time-intensive, technically demanding, and costly when compared with conventional approaches or other minimally invasive techniques, such as endoscopic mastectomy [11–13].
The RNSM often employs a multi-port approach, permitting the passage of a camera and dissection instruments [14, 15]. This technique can be performed either with gas insufflation or gasless (using retractors) [14, 16]. Although these techniques are efficient, the requirement for multiple incisions may compromise cosmetic benefits [15]. Single-port and single-incision approaches are recent advancements in RNSM, which offer a theoretical advantage of reduced surgical trauma and improved esthetic outcome (single incision), with the maintenance of oncological and reconstructive goals of NSM [17].
Despite attractiveness and growing interest, perioperative and postoperative outcomes for single-incision RNSM remain unclear. Current evidence comes mainly from small sample-sized single-center reports with variable documentation of complications [15, 18–21]. A detailed pooling of presently available data is required to refine a better definition of perioperative variables, rates of complications, and oncological outcomes. To this end, a systematic review with meta-analysis was conducted to examine the feasibility and outcomes of single-incision RNSM by pooling results from the existing literature, offering a comprehensive and up-to-date synthesis of evidence to enhance clinical practice and direct future research endeavors.

Methods

Search strategy and registration
This comprehensive systematic review and single-arm meta-analysis study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines [22] (Fig. 1). A comprehensive search was performed to identify eligible studies that reported on perioperative parameters and postoperative outcomes of single-incision RNSM. Four electronic databases, including PubMed, Medline, Scopus, and Web of Science, were searched (Fig. 1). Search strategy comprised relevant keywords linked by either of the Boolean operators (Table S1). No time restriction was applied, and a literature search was conducted up to August 12, 2025, “search date”. This study is registered on PROSPERO, an international systematic review registry platform (CRD420251157574).

Research question and eligibility criteria
This study aimed to assess the feasibility, perioperative parameters, complication rates, and oncologic outcomes of the single-incision approach in robotic nipple-sparing mastectomy. The following inclusion criteria were used to identify eligible studies: i) peer-reviewed clinical studies including clinical trials, cohort studies, or case series with ≥ 10 patients; ii) studies reporting outcomes of NSM performed using single-port or single-incision robotic systems; iii) studies published in English; and iv) studies providing either comparative or descriptive outcome data. The excluded studies were case reports or series reporting on < 10 patients, review articles, editorial comments, letters to the editor, abstracts presented as part of a conference, studies involving cadaveric or non-human subjects, studies that did not clearly describe the single-port robotic technique, and duplicate patient cohorts. The research question and eligibility criteria were defined using the PICO (Population, Intervention, Comparator, and Outcomes) framework and are summarized in Table S1.

Study selection and data extraction
A total of 454 records were identified through the systematic database searches. After deduplication, 238 records remained. Two reviewers independently screened titles and abstracts, resulting in 48 reports sought for full-text assessment. Following full-text review, 35 studies were excluded for the following reasons: mixed surgical approaches (n = 15), overlapping patient cohorts (n = 14), sample size < 10 (n = 5), cadaveric study (n = 1). Thirteen studies met the eligibility criteria, which were divided into two groups: prosthesis-based reconstruction cohorts (n = 8) and mixed cohorts (n = 5) (Fig. 1). Discrepancies during the selection process were resolved through discussion or, when necessary, consultation with a senior third reviewer.
Two independent reviewers extracted data pertaining to the characteristics of the included studies, like study design, sample size, follow-up duration and patients’ characteristics (Table 1 and Table 2), peri-operative variables (Table 3), and postoperative complications (including skin flap necrosis, NAC necrosis, surgical site infection (SSI), deep infection, delayed wound healing, hematoma, seroma, and cancer recurrence). Discrepancies were resolved by consensus or senior author arbitration. Zotero software (version 7.0.27) was used for reference management and duplicate removal. No artificial intelligence tools were used in the selection or extraction process.

Quality assessment
Assessment of methodological quality of the included studies was performed using the Joanna Briggs Institute (JBI) Checklist for Prevalence Studies [23]. This quality assessment method was used since it is more suitable for studies reporting proportional outcomes and prevalence data. The JBI checklist evaluates nine domains, and each study was rated independently across these domains. Judgments were made by two independent reviewers based on information provided in the original publications (Tables S3 and S4). Overall, the quality assessment of the included studies was considered acceptable for this analysis.

Statistical analysis
The primary outcome was the pooled proportion of the binary outcomes (prevalence). Event rates were transformed using the Freeman–Tukey double-arcsine method, which stabilizes variances [24]. Pooled estimates were calculated under a random-effects model (REML) due to anticipated heterogeneity [24]. Subgroup analyses or meta-regression were not performed because the number of studies reporting each outcome was small, which could result in unreliable or spurious findings [25]. Continuous outcomes (e.g., operative time and blood loss) were calculated using inverse-variance weighting.
Forest plots were generated to visually display individual study proportions and pooled estimates. Study-level 95% confidence intervals, with statistical significance defined as p < 0.05, were calculated using the Wilson score, a method that provides accurate coverage and balanced interval width compared to other methods, especially for small samples or extreme proportions [26]. Heterogeneity was quantified via the I2 statistic (low: 25%, moderate: 50%, and high: 75%).
Publication bias was visually demonstrated as funnel plots; however, the small number of included studies limits the ability to exclude it [27]. Formal statistical tests for funnel plot asymmetry (e.g., Egger’s test) were not performed because the pooled outcomes were reported from a small number of studies, which may yield unreliable results [28]. Statistical analyses were performed using Stata/BE 19.0 software.

Results

Study characteristics
Thirteen studies published between 2019 and 2025 were included, comprising prospective (n = 5) and retrospective (n = 8) designs and presenting 982 patients who underwent 1135 single-incision RNSM procedures across four countries (Italy, Taiwan, South Korea, and the USA). The single-incision RNSM was performed either for therapeutic (87%) or prophylactic (13%) purposes. Approximately 15% of cases were bilateral NSMs. Average patient age ranged from 40 to 49 years, and body mass index (BMI) ranged from 20.5 to 26.3 kg/m2. Following the NSM procedure, the majority of cases (92%) underwent prosthesis-based reconstruction (including tissue expander or direct-to-implant), while the remaining smaller proportion involved autologous tissue (flap) reconstruction. The reported single incision was 4–5 cm in length, while the average breast tissue weight ranged between 279 and 350 g. Follow-up durations varied significantly, ranging from 9 to 49 months. Detailed per-study characteristics are summarized in Table 1 and Table 2.

Peri-operative parameters
Operative time was reported by most of the included studies. Pooled analysis revealed a mean operative time of 273 min (95% CI: 201.5–344.5, I2 = 98.99%, p < 0.001). Substantial variation existed among the studies, as the statistical analysis revealed significant heterogeneity (Fig. 2A). This implies that the significant disparities in operative time could be attributed to the diversity in the techniques during surgery or the post-mastectomy reconstruction procedure. Intraoperative blood loss was inconsistently reported, with only five studies quantitatively reporting it. Generally, the estimated blood loss was low, with a pooled result of 82 mL (95% CI: 19.0–144.7, I2 = 99.62%, p < 0.001) (Fig. 2B).
No conversions to open surgery and no perioperative mortality were reported in the included studies. Reoperation rates were generally low, 3.3% (95% CI: 1.74–5.30, I2 = 21.06%, p = 0.12) (Fig. 2C), and were mainly performed due to immediate post-operative complications including haematoma, infection, or skin necrosis. Pooled result of positive surgical margins was < 1% (95% CI: 0.00–3.25, I2 = 59.60%, p = 0.01) (Fig. 2D).
Length of hospital stay (LOS) was reported in 6 out of 13 studies and ranged from 2 to 13.3 days. Shorter LOS was reported in cohorts of patients who underwent post-mastectomy implant/expander placement [13, 29–31], while longer LOS was observed in cohorts of patients who, at least part of them, underwent flap reconstruction [18, 32]. A summary of the perioperative parameters for each study is presented in Table 3.

Complication profile
The overall complication rates were found to be generally low. Skin flap necrosis was observed in 2.4% of the cases (95% CI: 0.62–4.95) with moderate heterogeneity across the studies (I2 = 62.87%, p = 0.01). The highest skin flap necrosis rate (18.2%) was recorded in the only study that reported a cohort of patients who exclusively underwent post-mastectomy flap reconstruction [18], which appears to be the main contributor to heterogeneity (Fig. 3). On the other hand, necrosis of the nipple-areolar complex (NAC) was observed in a much lower rate, 0.4% of the cases (95% CI: 0.03–1.10; I2 = 4.6%, p = 0.16) (Fig. 3B).
The SSI occurred among 3.8% of the patients (95% CI: 1.59–6.64; I2 = 47.54%, p = 0.06) (Fig. 4A), while deep infection rate, including cases of implant-related infection, was relatively lower at 1.2% (95% CI: 0.29–2.63, I2 = 28.7%, p = 0.19) (Fig. 4B). Delayed wound healing showed a pooled rate at 4.7% (95% CI: 2.38–7.64; I2 = 0.00%, p = 0.75) (Fig. 4C).
Hematoma was recorded in 2.4% (95% CI: 0.80–4.72; I2 = 53.57%, p = 0.02). The highest hematoma occurrence rate (22.7%) was recorded in the only cohort of patients who exclusively underwent post-mastectomy flap reconstruction [18], which significantly contributed to heterogeneity (Fig. 5A). On the other hand, the occurrence of seroma was relatively higher, with a rate of 5.8% (95% CI: 2.49–10.19; I2 = 65.17%, p = 0.01). The moderate heterogeneity can be explained, at least in part, by the outlier rate (17%) reported in a cohort of patients in which 41.5% underwent flap reconstruction (Fig. 5B) [33].
Locoregional recurrence (LRR) of cancer occurred in 0.8% (95% CI: 0.02–2.29; I2 = 9.5%, p = 0.37) (Fig. 6 A), whereas distant recurrence of cancer occurred at a relatively lower rate, 0.4% (95% CI: 0.00–2.05; I2 = 35.07%, p = 0.25) (Fig. 6B). No study reported procedure-related mortality or unexpected oncologic compromise during follow-up. Funnel plots showed no significant asymmetry for most endpoints, indicating small risks of bias (Figure S1).

Discussion
This systematic review and single-arm meta-analysis study synthesizes an up-to-date quantitative evidence of single-incision RNSM feasibility and outcomes. This study demonstrates acceptable perioperative parameters, low postoperative complication rates, and favorable oncologic outcomes. The findings in this study demonstrate that single-incision RNSM is a technically feasible and safe procedure. The reported low complication rates of this approach were comparable to or superior to the reported outcomes of the conventional or other minimally invasive NSM approaches.[35, 36]. In addition, the single-incision approach offers cosmetic advantages by restricting the number of incisions to a single incision, particularly when placed laterally. This may enhance patient satisfaction and postoperative outcomes [37].
Ouyang et al. (2025) reported a meta-analysis of outcomes from three studies exclusively utilizing the da Vinci SP system, representing true single-port RNSM [34]. However, the present study is more comprehensive, incorporating all available studies reporting single-incision RNSM performed using various da Vinci platforms (Si, Xi, X, and SP). All procedures were performed through a single axillary incision with a mean length of 4–5 cm, comparable to the true single-port RNSM incision length [34]. In addition, this study distinctively investigates oncologic safety outcomes, including clear surgical margins and cancer recurrence rates.
The pooled total operative time in the single-incision NSM was longer than conventional NSM but consistent with early experiences in robotic-assisted breast surgery [38, 39]. High heterogeneity in the reported operative duration reflects variations in surgeon experience, different operative/robotic settings, case complexity, and reconstructive approach. In one study, operation time gradually decreased as the experience improved from the first to the final cases [39]. With growing proficiency and optimized port access, subsequent studies have reported shortening of the robotic console and docking times as surgical experience accumulates, suggesting a progressive learning curve [40].
Intra-operative blood loss, reoperation rate, and LOS in this study were small, which are comparable to the current evidence derived from large cohorts of conventional and robotic NSM patients [41]. Conversion to open surgery was distinctly not reported in any of the included studies, suggesting an improved learning compared to earlier experiences of RNSM, where conversion to open surgery was marginally above zero [38, 39]. The positive surgical margins rate was less than 1%, a remarkably lower rate compared to recent evidence from robotic (4%) and conventional (7%) NSM [42].
The single-incision RNSM has reported a low overall rate of post-operative complications. Skin flap and NAC necrosis rates were significantly lower compared to conventional and other minimally invasive NSM approaches (e.g., endoscopic and multi-port robotic approaches) [35, 36]. Likewise, infection rates, including superficial and deep SSI, and delayed wound healing, were relatively lower [35, 36]. Hematoma and seroma both occurred in single-incision NSM at rates comparable to those recorded in conventional and other robotic NSM [43]. The LRR and distant recurrence of cancer rates are substantially lower than those observed with conventional procedures [44, 45]; however, the heterogeneity in follow-up durations across the included studies warrants cautious interpretation of these findings.
Several limitations associated with this analysis must be acknowledged. This is a single-arm meta-analysis which limits definitive conclusions regarding the relative effectiveness compared to other approaches. Most of the included studies had retrospective designs with small sample sizes [15, 18, 21], with heterogeneity in reporting peri-operative and post-operative outcomes. Long-term oncologic/recurrence outcomes remain underreported in approximately half of the included studies. In addition, most of the included studies reported mixed cohorts without separate complication profile data for different patient populations, such as the indication for surgery (prophylactic versus therapeutic mastectomy), precluding reliable subgroup analysis. Future prospective studies should therefore consider reporting indication- and approach-specific outcomes to enable meaningful comparative assessments. Another limitation is that most of the pooled outcomes in this analysis were derived from less than 10 studies, limiting the reliability of formal statistical testing for publication bias. Consequently, publication bias was assessed through visual inspection of funnel plots, and the results should be interpreted with appropriate caution. The use of an English-language filter may also introduce selection bias. These limitations introduce potential biases related to reporting and selection [46]. Although the individual studies have small sample sizes that limit generalisability, this meta-analysis of proportions is well-suited for synthesizing evidence across these studies, as it stabilizes variance and enables a more accurate estimation of prevalence [47].
The high heterogeneity observed for some outcomes likely reflects different surgeons’ experience, operative and robotic settings, and post-mastectomy immediate reconstruction procedures across the included studies [48]. Given the limited number of studies contributing to each outcome, formal subgroup analyses or meta-regression were not undertaken, and the pooled prevalence estimates should therefore be interpreted with caution [25]. Assessment of publication bias was limited to visual inspection of funnel plots. Quantitative tests, such as Egger’s regression, were not applied due to the small number of studies per outcome and substantial heterogeneity, which may compromise the validity of such tests [28].
Nonetheless, this study offers important insights. The consistency of findings across diverse geographical and temporal settings suggests that single-incision RNSM provides a reproducible and generally acceptable treatment approach for breast cancer treatment or prophylactic intervention. Future research should focus on prospective, comparative, and cost-effectiveness trials comparing single-incision RNSM with multi-port robotic, endoscopic, and conventional (open) NSM approaches. Standardized reporting of cosmetic, functional, and oncologic endpoints is essential to validate the reproducibility and scalability of this innovative method.

Conclusion
Within the limitations of this study, the single-incision approach in RNSM represents a safe and feasible minimally invasive surgical approach with low morbidity and promising reconstructive outcomes. Although current evidence is encouraging, further high-quality studies are required to reliably investigate the subgroup comparison, long-term oncologic safety, economic value, and learning curve dynamics.

Supplementary Information
Below is the link to the electronic supplementary material.