Efficacy and safety of robotic-assisted versus endoscopic-assisted axillary lymph node dissection in node-positive breast cancer: a retrospective comparative study

Abstract

Background

Robotic surgery is increasingly being adopted for breast cancer treatment. However, robust clinical evidence regarding its effectiveness and safety remains limited. This retrospective cohort study aimed to compare the surgical quality and short-term outcomes of robotic-assisted axillary lymph node dissection (R-ALND) and endoscopic-assisted axillary lymph node dissection (E-ALND) in patients with node-positive breast cancer. Here, we report the short-term outcomes of this trial.

Methods

This single-center retrospective study compared the short-term efficacy and safety of R-ALND and E-ALND in patients with node-positive breast cancer. Patients who underwent surgery at the Sixth Affiliated Hospital of Sun Yat-sen University between January 2022 and October 2024 were included. Clinical and pathological characteristics, surgical outcomes, and postoperative complications were analyzed.

Results

A total of 56 patients were included, with 29 undergoing E-ALND and 27 undergoing R-ALND. The R-ALND group demonstrated significantly shorter operative times (43.37 ± 12.40 min vs. 60.10 ± 19.37 min, p < 0.001) and lower mean intraoperative blood loss (3.26 ± 2.40 ml vs. 9.24 ± 4.29 ml, p < 0.001). Postoperatively, the R-ALND group exhibited better upper limb function and sensation, as evidenced by significantly lower DASH scores at 1-month (10.87 ± 1.35 vs. 14.64 ± 3.49, p < 0.001) and 3-month (6.68 ± 1.86 vs. 9.24 ± 2.74, p < 0.001) follow-ups. Additionally, the R-ALND group had fewer postoperative complications, including a reduced incidence of sensory disturbances, burning sensations, and numbness in the upper limb.

Conclusion

Compared with E-ALND, R-ALND significantly reduces intraoperative blood loss and postoperative complications, with less impact on upper limb function and sensory outcomes. These findings indicate that R-ALND may provide better clinical benefits for patients requiring axillary lymph node dissection in the management of breast cancer.

Introduction

Breast cancer is one of the most prevalent malignant tumors among women, with high incidence and mortality rates globally [1]. For patients initially diagnosed with breast cancer without systemic metastasis, treatment planning must consider both tumor size (T stage) and the presence of lymph node metastasis (N stage) [2]. When axillary lymph node metastasis is present, axillary lymph node dissection (ALND) is a critical component of surgical management. ALND not only aids in accurate tumor staging but also serves as a cornerstone for guiding treatment strategies and predicting prognosis [3].

However, ALND is associated with several complications, including lymphedema, wound infection, fat necrosis, sensory disturbances in the arm, and restricted range of motion [4]. With the advancement of minimally invasive surgical techniques, endoscopic technology has been increasingly adopted in breast cancer treatment [5]. This approach provides an enlarged and more precise visual field, thereby minimizing damage to microstructures. Although several studies have demonstrated the efficacy and safety of endoscopic axillary lymph node dissection (E-ALND) [6], the current evidence is largely derived from low-quality, small-scale randomized controlled trials. Notably, there is a significant lack of multicenter prospective clinical studies. Therefore, the effectiveness and safety of E-ALND remain controversial.

Compared with endoscopic surgery, robotic surgery offers distinct advantages, including a magnified three-dimensional surgical view and enhanced stability during surgical manipulation. The robotic arms, equipped with wrist-like joints that provide seven degrees of freedom, enabling 540° rotation and precise movements within confined spaces. Robotic surgery has been successfully applied in various lymph node dissections, including retroperitoneal, cervical, and inguinal regions [10]. The robotic system allows surgeons to access narrower surgical fields and more clearly identify microstructures, thereby improving surgical precision and safety [11]. Theoretically, these features render robotic surgery particularly suitable for axillary lymph node dissection. However, given the higher cost associated with robotic surgery compared to endoscopic techniques, robust evidence is needed to justify its clinical application [13]. Currently, there is a paucity of studies focusing on robotic axillary lymph node dissection (R-ALND), and comparative analyses between endoscopic and robotic approaches are lacking. Therefore, this study aims to compare the short-term efficacy and safety of endoscopic and robotic axillary lymph node dissection in patients with breast cancer.

Materials and methods

Patient enrollment

Patients who underwent single-port endoscopic-assisted axillary lymph node dissection (E-ALND) or robotic-assisted axillary lymph node dissection (R-ALND) for breast cancer at the Sixth Affiliated Hospital of Sun Yat-sen University between January 2022 and October 2024 were recruited for this study. A total of 56 patients were included, with 29 undergoing E-ALND and 27 undergoing R-ALND (Fig. 1). The study protocol was approved by the Ethics Committee of the Sixth Affiliated Hospital of Sun Yat-sen University (approval number: 2024ZSLYEC-467). Written informed consent was obtained from all participants prior to enrollment.

Fig. 1

Flow chart of patient selection. Abbreviations: E-ALND,endoscopic axillary lymph node dissection; R-ALND,robotic axillary lymph node dissection;C-ALND, conventional axillary lymph node dissection

Clinical and pathological characteristics, including age, estrogen receptor (ER) status, progesterone receptor (PR) status, HER2 status, Ki-67 index, histological grade, tumor size, TN staging, and the use of neoadjuvant chemotherapy, were included in the analysis. Additionally, surgical outcomes, such as the number of axillary lymph nodes dissected, the number of positive axillary lymph nodes, operative time, intraoperative blood loss, and postoperative complications (including upper limb dysfunction, wound infection, seroma, and lymphedema), were also evaluated.

The inclusion and exclusion criteria were determined based on demographic and tumor characteristics. Inclusion criteria were as follows: 1) age between 18 and 75 years; 2) pathologically confirmed breast cancer; and 3) presence of positive axillary lymph nodes. Exclusion criteria included: 1) presence of distant metastasis; 2) inability to tolerate surgery due to cardiopulmonary dysfunction; and 3) concurrent diagnosis of other malignant tumors.

Surgical procedure

Robotic-Assisted Axillary Lymph Node Dissection (R-ALND) (Fig. 2A-B)

Fig. 2

Representative surgical images show robot-assisted axillary lymph node dissection (R-ALND) and endoscopy-assisted axillary lymph node dissection (E-ALND). A-B R-ALND surgical images; C-D E-ALND surgical images

For R-ALND, a small incision is made on the lateral chest wall with minimal dissection to create a surgical cavity. A single-port four-channel trocar is inserted, followed by the placement of three 8-mm trocars. The da Vinci Xi robotic arms are then connected to these trocars, and the cavity is insufflated to maintain a pressure of 8–10 mmHg.

The R-ALND procedure follows a systematic "bottom-up, back-to-front" approach. The dissection begins at the lower posterior wall of the axilla and progresses superiorly, followed by the medial wall, lateral wall, central axilla, and finally the apex of the axilla. The specific steps are as follows: 1) Axillary Floor Dissection: The procedure starts with the dissection of the axillary floor, where the reticulated fascia is separated to identify and protect the lowest intercostobrachial nerve. 2) Posterior Wall Dissection: The dissection proceeds medially to isolate the anterior branch of the thoracodorsal vessels, continuing until the main trunk of the thoracodorsal vessels is reached. Subsequently, the dissection moves laterally to separate the circumflex scapular vessels, extending to the point where the thoracodorsal vein joins the axillary vein. 3) Medial Wall Dissection: The medial wall is carefully dissected to protect the thoracodorsal nerve anterior to the serratus anterior muscle. CO₂ insufflation is utilized to dissect along the membrane space superiorly toward the axillary vein angle, which represents the uppermost part of the axilla. 4) Lateral Wall Dissection: The lateral wall is dissected along the outer side of the intercostobrachial nerve to visualize the axillary vein. 5) Anterior Wall Dissection: The anterior wall is dissected with attention to protecting the medial cutaneous nerve of the arm. The dissection follows the axillary vein from lateral to medial. 6) Intercostobrachial Nerve Dissection: The intercostobrachial nerve at the level of the axillary vein is dissected from medial to lateral, separating the upper and lower layers of axillary tissue to preserve the nerve. The lowest intercostal nerve, being finer, may be transected during this process. 7) Lymph Node Clearance and Irrigation: After thorough dissection and complete clearance of the axillary lymph nodes, the cavity is irrigated with warm distilled water, and drains are placed to ensure proper postoperative management.

Endoscopic-Assisted Axillary Lymph Node Dissection (E-ALND) (Fig. 2C-D)

For E-ALND, a 5-cm incision is made at the lateral edge of the chest wall. Subcutaneous tissue extending 2–3 cm around the incision is separated to facilitate the insertion of a disposable incision fixator. The port of the single-hole cavity mirror is placed over the outer mouth of the notched fixator to maintain the cavity pressure at 8 mmHg. The surgical steps are consistent with those of R-ALND.

Evaluation of postoperative upper limb function

Range of motion (ROM) of the shoulder joint was assessed at baseline and 3 months postoperatively. The shoulder joint ROM comprises three movements: abduction, flexion, and horizontal abduction. Based on our clinical experience, a goniometer was used to measure ROM for all subjects.

The Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire is a widely used [14], standardized tool for assessing upper limb function. It is a patient-reported outcome measure that consists of 30 items, 23 of which evaluate upper limb function and 7 that assess symptoms, including pain (at rest, during activity, and at night), strength, and stiffness.

The DASH score is calculated by summing the scores of the 30 items and then applying the following formula: DASH score = (Total sum of the 30 items—30 [minimum score]) / 1.20. This transformation scales the raw score to a range of 0 to 100, where 0 indicates normal upper limb function and 100 indicates severe impairment.

Statistical analysis

Continuous variables were analyzed using the Mann–Whitney U test and are presented as medians with their respective minimum and maximum ranges. Categorical variables were compared using Fisher's exact test. All statistical analyses were conducted using SPSS software. A p-value of less than 0.05 was considered statistically significant.

Results

Patient characteristics

A retrospective analysis was conducted on the medical records of 136 patients who underwent ALND at our center from January 2022 to October 2024. After applying the inclusion and exclusion criteria, 80 patients from the conventional surgery group were excluded, resulting in the inclusion of 56 patients. Of these, 29 underwent E-ALND, and 27 underwent R-ALND. The demographic and clinical characteristics of the patients are summarized in Table 1. There was no statistically significant difference in age between the two groups (R-ALND group: 55.83 ± 14.89 years vs. E-ALND group: 49.81 ± 10.4 years, P = 0.084). Additionally, no significant differences were observed between the two groups in terms of pathological staging, hormone receptor status, HER2 status, Ki-67 expression, histological grading, preoperative chemotherapy, or type of breast surgery.

Table 1 Clinicopathologic characteristics of E-ALND and R-ALND

Comparison of surgical outcomes

The primary surgical outcomes revealed that the mean operative time for the E-ALND group was significantly longer than that for the R-ALND group (60.10 ± 19.37 min vs. 43.37 ± 12.40 min, p < 0.001). Additionally, the R-ALND group demonstrated significantly lower mean intraoperative blood loss (3.26 ± 2.40 mL vs. 9.24 ± 4.29 mL, p < 0.001).

Secondary surgical outcomes included total axillary drainage volume, postoperative axillary drain retention time, postoperative hospital stay duration, and postoperative axillary pain. The total axillary drainage volume was 247.66 ± 198.42 mL for the E-ALND group and 192.48 ± 139.06 mL for the R-ALND group, with no statistically significant difference between the groups (p = 0.237). The postoperative axillary drain retention time was similar for both the E-ALND group (8.48 ± 3.65 days) and the R-ALND group (7.88 ± 2.51 days, p = 0.485). However, the R-ALND group had a significantly shorter postoperative hospital stay (9.03 ± 2.38 days vs. 10.69 ± 3.20 days, p = 0.033). Postoperative axillary pain occurred in 37.0% of the R-ALND group and 72.4% of the E-ALND group, indicating that the R-ALND group experienced less postoperative pain than the E-ALND group (p = 0.008). There were no significant differences between the two groups in the mean number of axillary lymph nodes dissected or the number of positive nodes (Table 2).

Table 2 Surgical outcomes of E-ALND and R-ALND

Comparison of postoperative complications

In this study, the severity of postoperative complications was assessed using the Clavien-Dindo classification system [15]. Neither the E-ALND nor the R-ALND group experienced postoperative bleeding or wound infection. One case of axillary hematoma occurred in the E-ALND group, which was successfully managed with aspiration and pressure dressing. Additionally, the E-ALND group had one case of poor wound healing and two cases of lymphatic leakage, which improved with persistent pressure dressing in the axillary area. In contrast, the R-ALND group had no occurrences of hematoma, poor wound healing, or lymphatic leakage (Table 3).

Table 3 Postoperative clinical outcome of E-ALND and R-ALND

Postoperative ROM outcomes

Preoperatively, no significant differences were observed in shoulder abduction, shoulder flexion, or horizontal abduction ROM between the two groups. However, at the 3-month postoperative follow-up, the R-ALND group showed significantly greater ROM in shoulder abduction (174.03 ± 5.35° vs. 165.75 ± 5.53°, p < 0.001) and shoulder flexion (173.04 ± 5.35° vs. 167.03 ± 4.62°, p < 0.001) compared to the E-ALND group. Similarly, the horizontal abduction ROM also demonstrated a significant difference (33.48 ± 3.99° vs. 30.07 ± 3.40°, p = 0.001) (Table 4).

Table 4 Arm morbidities in the two groups

DASH scores and upper limb function

Preoperatively, there was no significant difference in DASH scores between the two groups. However, at the 1-month postoperative follow-up, the R-ALND group had significantly lower DASH scores compared to the E-ALND group (10.87 ± 1.35 vs. 14.64 ± 3.49, p < 0.001). At the 3-month follow-up, upper limb function improved in both groups, but the R-ALND group still exhibited lower DASH scores than the E-ALND group (6.68 ± 1.86 vs. 9.24 ± 2.74, p < 0.001) (Table 4).

Postoperative upper limb sensation and pain

At the 3-month postoperative follow-up, the E-ALND group reported a significantly higher incidence of reduced sensation (8 cases, 27.6%), burning sensation (6 cases, 20.7%), and numbness (8 cases, 27.6%) in the upper limb compared to the R-ALND group, which had only 1 case (3.7%) of reduced sensation and 1 case (3.7%) of numbness, with no cases of burning sensation. These differences were statistically significant. However, there was no significant difference in postoperative upper limb pain between the two groups (3 cases, 10.3% in the E-ALND group vs. 0 cases, 0% in the R-ALND group, p = 0.132). These results suggest that R-ALND had a lesser impact on postoperative upper limb sensation and function, enabling greater activity in the affected upper limb (Table 4).

Discussion

This single-center retrospective study was designed to evaluate the safety and efficacy of R-ALND compared to E-ALND in breast cancer patients.

Endoscopic surgery has been widely utilized in the treatment of breast cancer; however, its efficacy and safety remain subjects of debate. An epidemiological study reported a significantly higher mortality rate for minimally invasive surgery in patients with stage IA2 or IB1 cervical cancer compared to open surgery (P = 0.002) [16]. A meta-analysis also indicated that patients undergoing minimally invasive radical hysterectomy faced a 71% higher risk of recurrence and a 56% higher risk of death [17]. During endoscopic surgery, instruments are introduced into the axilla in a nearly vertical direction, with limited horizontal manipulation angles, which can result in instrument interference in the narrow axillary space. Furthermore, endoscopic surgery often requires an experienced assistant, which may not be available at all medical centers. These factors can negatively impact the quality of endoscopic surgery. Despite the safety of E-ALND being demonstrated in some studies, there remains a lack of large-scale, prospective clinical trials to definitively establish its long-term benefits.

Robotic surgery offers several advantages over endoscopic techniques by overcoming these limitations. Robotic instruments have multiple degrees of freedom, providing flexibility akin to a surgeon's hand, with enhanced stability. The magnified three-dimensional view offers a detailed surgical field controlled by the surgeon, potentially eliminating the need for an assistant [11]. These technical advantages make robotic surgery a more accurate and efficient option for axillary lymph node dissection [2]. However, few studies have focused on robotic axillary lymph node dissection, and even fewer have directly compared R-ALND with E-ALND [18]. In our study, R-ALND significantly reduced intraoperative bleeding, postoperative pain, and hospital stay duration. Moreover, it resulted in less postoperative impairment of upper limb function and sensation, allowing for greater mobility of the affected limb.

The reduction in intraoperative bleeding associated with robotic surgery can decrease surgical time and accelerate postoperative recovery. In our study, the R-ALND group exhibited significantly shorter surgical times (43.37 ± 12.40 min vs. 60.10 ± 19.37 min, p < 0.001) and lower mean blood loss (3.26 ± 2.40 ml vs. 9.24 ± 4.29 ml, p < 0.001) compared to the E-ALND group. The da Vinci robotic system, which provides up to 10–15 times magnification, offers a clearer and more detailed surgical field. This enhanced visibility facilitates the early identification of small blood vessels in the axillary region, thereby reducing intraoperative bleeding and enhancing the overall quality and safety of the surgery. Additionally, robotic surgery minimizes the shaking of the surgical field, a limitation often observed in endoscopic procedures, which further improves stability and precision. Studies comparing robotic and open axillary dissection have similarly demonstrated that robotic surgery can significantly reduce intraoperative bleeding [18].

During axillary lymph node dissection, there is a risk of damaging the intercostobrachial nerve, which can lead to postoperative sensory and functional impairments in the upper limb. These impairments are often difficult to treat and can negatively affect the patient’s quality of life [19]. The intercostobrachial nerve, originating from the anterior branch of the second intercostal nerve, innervates the skin of the axilla and lateral chest wall, controlling upper limb sensory functions [20]. Due to its unique anatomical location in the axillary space, this nerve is challenging to preserve during surgery. In our study, at the 3-month postoperative follow-up, the R-ALND group showed significantly fewer cases of upper limb sensory reduction (1 case, 3.7%), burning sensation (0 cases, 0%), and numbness (1 case, 3.7%) compared to the E-ALND group, which had higher incidences of these sensory disturbances. These findings suggest that R-ALND may better preserve the intercostobrachial nerve, reducing the risk of postoperative sensory impairments.

The stability of robotic surgery, characterized by its tremor-free instruments, enables more precise dissection of delicate structures like the intercostobrachial nerve. Studies in prostate cancer lymph node dissection have shown that robotic surgery, with its enhanced precision, better preserves nerves and reduces postoperative complications such as urinary retention [21]. The robotic system’s 7 degrees of freedom and 540° rotation capability allow for more efficient and precise dissections around the intercostobrachial nerve. Although the nerve does not directly control joint movements, preserving it can significantly reduce postoperative sensory abnormalities and neuropathic pain, promoting better upper limb functional recovery. Our results also showed that at 1 and 3 months postoperatively, patients in the robotic group had significantly better upper limb function, as evidenced by lower DASH scores (10.87 ± 1.35 vs. 14.64 ± 3.49, p < 0.001; 6.68 ± 1.86 vs. 9.24 ± 2.74, p < 0.001), emphasizing the potential benefits of robotic surgery in improving postoperative outcomes.

This study has several limitations. First, it is a single-center retrospective study with a small sample size. Second, the relatively short follow-up period limits our ability to assess the long-term oncological safety of robotic surgery. Future studies with larger cohorts and longer follow-up periods are needed to confirm the reliability of these findings and further evaluate the long-term outcomes.

Conclusion

Compared to E-ALND, R-ALND significantly reduces intraoperative bleeding, results in less postoperative upper limb impairment, and provides greater upper limb motion on the affected side, suggesting that robotic surgery may offer substantial advantages in the surgical management of breast cancer.