Efficacy comparison of optimal natural orifice specimen extraction for robotic middle rectal cancer resection in women: transanal or transvaginal orifice

Abstract

Purpose

This study aimed to determine the optimal natural orifice specimen extraction (NOSE) method for robotic-assisted mid-rectal cancer resection in women.

Methods

This retrospective propensity score-matched (PSM) study was to analyze the clinical data prospectively collected from female rectal cancer patients who underwent either robotic-assisted transvaginal specimen extraction (RATV) or robotic-assisted transanal specimen extraction (RATA) at our center between June 2016 and December 2022. The main outcome measures were urinary, anal, and sexual function. Disease-free survival (DFS), and overall survival (OS) were also included .

Results

Anal function, assessed by the Wexner score, was better in the RATV group than in the RATA group (P = 0.034). Additionally, pre-menopausal women in RATV group exhibited superior anal function over those in RATA group (P = 0.031). There was no statistically significant difference in urinary function between the groups for both pre-menopausal and peri-menopausal patients (P = 0.711, P = 0.106). No difference was observed in sexual function between the two groups (P = 0.351); however, pre-menopausal patients in RATA group had better sexual function than those in RATV group (P = 0.045). Univariate logistic regression analysis showed surgical procedure was not a significant factor for the occurrence of sexual dysfunction. There were no significant difference in DFS (P = 0.845)and OS (P = 0.642) between the two groups.

Conclusion

Though the postoperative efficacy of the RATA and RATV was equivalent on urinary and sexual function, RATV is an optimal natural orifice specimen extraction for robotic middle rectal cancer resection in women based on anal function.

Introduction

Rectal cancer (RC) has become a public health problem that is too important to be ignored [1]. Though neoadjuvant therapy with achieving a complete pathological response also play vital roles in improving outcomes and reducing recurrence, radical rectal resection is still a crucial treatment for rectal cancer [2]. With the increasing screen popularity of gastrointestinal endoscopy, the percentage of patients diagnosed with early rectal cancer has grown dramatically [3]. The number of patients with RC reaching cure has increased substantially by surgical resection over the years [4]. Data from our institution’s study showed that the 5-year overall survival and disease-free survival for robot-assisted rectal cancer surgery were 87.7% and 73.5%, respectively [5]. Therefore, consideration must be given to thefunctional results of treatment, including anal, urinary and sexual function. In particular, the simultaneous achievement of the goals of cure and of minimal impact on quality of life remain be challenging [6].

Robotic surgical systems have been widely used in the treatment of rectal cancer [7]. These systems offer enhanced visibility and precision in the narrow pelvic cavity, reducing conversion rates to open surgery [8]. Robotic surgery improves surgical precision and suturing quality, significantly lowering the incidence of postoperative complications such as anastomotic leakage [9]. Additionally, it reduces intraoperative blood loss, infection rates, and hospital stays, and enhances the negative margin rate in low tumor resections [10, 11].For traditional robot-assisted radical rectal cancer surgery, specimen extraction usually requires an abdominal incision, which may lead to wound infections, incisional hernia, and increased postoperative pain and scarring, affecting patient recovery and quality of life [12]. To mitigate these issues, natural orifice specimen extraction surgery (NOSES) has been developed [13].

NOSES, with minimal or no abdominal incisions, results in smaller scars, reducing psychological stress and enhancing patient satisfaction [14]. This technique allows specimen removal through the anus or vagina, avoiding the adverse effects of large abdominal incisions. Studies show that NOSES [15] has advantages over traditional specimen extraction in terms of postoperative complications, recovery time, and hospital stay. Patients experience less postoperative pain and require fewer analgesics, indicating a reduction in surgical trauma and faster recovery [16]. Most research has focused on long-term tumor prognosis, often neglecting postoperative quality of life and physiological functions, particularly anal, urinary, and sexual function. Studies indicate that RATV has limited impact on sexual function post-surgery in women [17]. Research on RATA shows a little impact on anal function [18]. Further studies comparing postoperative physiological function and quality of life between these surgical approaches are still needed.

In mid-rectal cancer NOSES, both transanal and transvaginal routes are used for specimen extraction. In this study, we retrospectively determine the optimal natural orifice specimen extraction (NOSE) method for robotic-assisted mid-rectal cancer resection in women based on anal, urinary, and sexual function, providing evidence to support optimal natural orifice specimen extraction for robotic middle rectal cancer resection.

Materials and methods

Study population and comparator group

A retrospective analysis of clincal data, which were prospectively collected from female patients diagnosed with resectable mid-rectal cancer at the First Affiliated Hospital of Nanchang University, was performed with approval from the institutional review board (IRB) of our center. All female patients diagnosed with resectable mid-rectal cancer without distant metastases were underwent robotic-assisted radical resection with NOSES between June 2016 and December 2022. All patients included in this study met the following criteria: 1.Postoperative pathology confirming adenocarcinoma of the rectum. 2.Tumor location in the mid-rectum, as determined by imaging, preoperative colonoscopy, and intraoperative anuscopy. 3.Body mass index (BMI) < 35 kg/m². 4. Absence of distant metastasis as confirmed by preoperative evaluation and intraoperative findings. The exclusion criteria were as follows:1.Presence of other concurrent malignancies. 2.Requirement for emergency surgery due to bleeding, obstruction, or perforation. 3.Receipt of neoadjuvant radiotherapy. 4. Lack of preoperative urinary, anal, and sexual functions. 5.Incomplete follow-up data or missing data.

A total of 315 patients were initially screened, and 277 patients with rectal adenocarcinoma were finally enrolled in this study; 192 patients were underwent robotic-assisted radical resection with RATA, while 85 patients underwent robotic-assisted radical resection with RATV. After PSM, 67 patients were included in the RATA and RATV groups, respectively (Figure S1). All procedures adhered to the ethical standards of the institution and the Declaration of Helsinki. Informed consent was obtained from all patients prior to inclusion in the study.

Surgical technique

All operations were performed by the same surgical team, who had undergone standardized training in robotic surgery and had experience with over 200 colorectal cancer surgeries. Only the tumor size or local invasion meeting the international consensus on NOSES for colorectal cancer was adopt for NOSES [19]. For both RATA and RATV groups, anesthesia was successfully administered, and the absence of distant metastases was confirmed intraoperatively. The surgeries were then conducted following the principles of Total Mesorectal Excision (TME). There were no differences between the RATA and RATV groups in terms of mesenteric mobilization within the abdominal cavity, lymph node dissection, or tumor resection. In both groups, the rectum was transected with a linear stapler at the lower margin of the tumor, maintaining distal margin of 2–3 cm. The primary distinction between the groups was the method of specimen extraction.

In the RATA group, NOSES II was utilized for transanal specimen extraction for mid-rectal cancer. The procedure involved careful anal dilation followed by transanal irrigation of the rectal lumen with iodophor saline. An incision was made at the end of the rectum, and a protective sleeve was introduced through the abdominal assistant’s port, which was then extended out through the anus. A toothed grasping forceps was inserted through the protective sleeve to secure the proximal rectal stump under direct vision. The rectum and specimen were then gently extracted to the anal verge via the protective sleeve. The rectum and tumor were excised at the pre-determined site above the tumor 10 cm. The anvil of a circular stapler was placed into the distal sigmoid colon, secured with a purse-string suture, and then returned to the abdominal cavity.

In the RATV group, NOSES III was employed for transvaginal specimen extraction. The procedure involved careful vaginal expansion and irrigation with iodophor saline. Using the inner valve of a 50 ml syringe as a guide, a transverse incision was made on the posterior vaginal wall with an ultrasonic scalpel. A protective sleeve was inserted through the abdominal assistant’s port and extended out through the incision in the posterior vaginal vault. A toothed grasping forceps was used to clamp the proximal rectal stump tightly under direct vision, and the rectum and specimen were extracted to the vulva via the protective sleeve. The rectum and tumor were excised at the predetermined site above the tumor 10 cm. The anvil of the circular stapler was then placed into the distal sigmoid colon, secured with a purse-string suture, and returned to the abdominal cavity. The posterior vaginal wall was closed using barbed sutures under direct robotic laparoscopic visualization.

After specimen extraction in both the RATA and RATV groups, the pelvic cavity was irrigated with iodophor saline to ensure hemostasis. The digestive tract was reconstructed by performing an end-to-end anastomosis of the sigmoid colon and rectum using a circular stapler. The stapler was introduced transanally, and end-to-end anastomosis was completed after confirming the absence of sigmoid mesenteric torsion. A gas injection test was performed to ensure the integrity of the anastomosis, and any potential weak points in the anastomotic triangle were reinforced with robot-assisted suturing. Key steps in the specimen extraction process for both RATA and RATV groups are depicted in Figure S2.

Data collection and follow-up

Each patient was brought into the clinical database established by gastrointestinal surgical institute, Nanchang university. Clinical data such as demographic information, surgical details, postoperative pathology, postoperative complications, rehabilitation data, and long-term follow-up information were prospectively collected. Overall survival (OS) was defined as the time from surgery to death from any cause. Disease-free survival (DFS) was defined as the time from surgery to disease recurrence or death from any cause.

• 1. Body Image Questionnaire (BIQ): This tool was used to assess patients’ attitudes towards their physical appearance and their satisfaction with the appearance of surgical scars at 3 months postoperatively. The BIQ has been validated in previous studies on cholecystectomy [20].

• 2. Wexner Incontinence Score: This score was employed to evaluate patients’ anal function at 3 months postoperatively [21].

• 3. International Consultation on Incontinence Questionnaire-Short Form (ICIQ-SF): This questionnaire was utilized to assess urinary incontinence in patients at 3 months postoperatively [22].

• 4. Female Sexual Function Index (FSFI): This index was used to assess patients’ sexual feelings and reactions over the past 4 weeks, administered at 4 months postoperatively [23]. A total score of ≤ 19 was indicative of possible sexual dysfunction.

Routine follow-up was scheduled in accordance with the National Comprehensive Cancer Network guidelines: one month after surgery, then every three months for the first two years, and every six months up to five years, or until patient death or study termination. Follow-up information was collected via email or telephone if patients were unable to return for observation. All patients were monitored until death or until December 31, 2022.

Statistical analysis

To balance the baseline characteristics between the two groups and minimize selection bias, the Propensity Score Matching (PSM) method was utilized. PSM was performed in a 1:1 ratio based on baseline information using a logistic regression model, with a caliper value set at 0.05. Kolmogorov-Smirnov test was used for normal distribution detection, P > 0.05 indicates that the data follow a normal distribution. Continuous variables following a normal distribution were expressed as mean ± standard deviation (SD) and analyzed using the independent samples t-test. For variables not following a normal distribution, the Mann-Whitney U test was employed for comparisons. Categorical data were presented as frequencies and percentages and were compared using theχ² test or Fisher’s exact test, as appropriate. Disease-free survival (DFS) and overall survival (OS) were estimated using the Kaplan-Meier method and compared between groups using the Log-rank test. P < 0.05 was considered statistically significant. All statistical analyses were conducted using IBM SPSS Statistics software (version 26.0, Chicago, IL).

Results

Patients and characteristics between RATV group and RATA group

Firstly, 277 cases with NOSES were selected from our clinical data. Patients in RATV group were older and had smaller tumors than patients in RATA group. PSM was performed in this study for age, BMI, ASA score, maximum tumor diameter, histological differentiation, pathological TNM stage, T stage, and N stage. Finally, 67 pairs of patients were matched successfully. After PSM, the baseline information was balanced and comparable (P > 0.05) (Table S1).

Comparison of perioperative indexes between RATV group and RATA group

Before PSM, the operation time (P<0.001), intraoperative blood loss (P = 0.016), postoperative hospital stay (P = 0.006) and harvested lymph nodes (P = 0.005) of the two groups were statistically significant. As for postoperative complications, 12 complications occurred in the RATV group. 30 complications occurred in the RATA group. Among the postoperative complications in the two groups, anastomotic leakage was not significantly different (Table 1).

Table 1 Comparison of postoperative conditions between RATV group and RATA group before and after PSM

After PSM, the operation time was shorter in the RATA group than in the RATV group (149 ± 33.5 min in the RATA group vs. 158.1 ± 28.7 min in the RATV group, P = 0.027). However, intraoperative blood loss was similar (61.7 ± 27.9 ml in the RATA group vs. 67.3 ± 28.2 ml in the RATV group, P = 0.251). In terms of postoperative recovery, the time to recovery of gastrointestinal function was comparable in both groups (51.0 ± 19.0 h in the RATA group vs. 55.5 ± 16.6 h in the RATV group, P = 0.154). There was also no difference in the postoperative hospital stay between the two groups (8.4 ± 3.8 day in the RATA group vs. 8.1 ± 3.5 day in the RATV group, P = 0.635). In addition, in terms of postoperative complications, 10 complications occurred in the RATV group. Among them, there was 1 rectal anastomotic fistula and 2 rectovaginal fistula. In contrast, 11 complications occurred in the RATA group. Among them, there were 2 cases of rectal anastomotic fistula. There was no significant difference in anastomotic leakage between the two groups (Table 1). In terms of surgical stress, this study compared white blood cell and C-reactive protein levels of the two groups at 1, 3, and 5 days postoperatively. Before PSM, There was also no difference in the inflammatory markers and VAS (Visual Analogue Scale) score between the two groups. After PSM, the RATA group had lower inflammatory markers than the RATV group (P = 0.015, P = 0.019). There was also no significant advantage in terms of VAS scores or additional analgesics in either group (Fig. 1, Table S2).

Fig. 1

Comparative Perioperative Indexes in the two groups of patients after PSM (A-C). (A). VAS scores between two groups; (B) White blood cell scores between two groups; (C) C-reactive protein scores between two groups. The P-value was calculated by repeated measures statistical analysis. VAS, Visual Analogue Scale; RATV, robotic-assisted transvaginal specimen extraction surgery; RATA, robotic-assisted transvanal specimen extraction surgery

Comparison of short-term quality of life between RATV group and RATA group

In this study, the Body image questionnaire (BIQ) was used to compare the body image of the two groups of patients at 3 months postoperatively. The questionnaire has 8 scoring items. Of these, items 1–5 assessed the Body image scale and items 6–8 assessed the Cosmetic scale. Before PSM, the body image was superior in the RATV group to the RATA group (P<0.001). However after PSM, the body image and cosmetic scores remained at the same level in both groups(P = 0.506, P = 0.115)(Table S3). As for the assessment of anal and urinary function, this study found that the Wexner score and ICIQ-SF score at 3 months postoperatively was not clear advantage between RATV group and RATA group in pre-menopausal and pro-menopausal patients before PSM (Table S4). Yet, after PSM, the Wexner score in the RATV group was better than that in the RATA group (P = 0.034). Meanwhile, further analysis revealed that the anal function of pre-menopausal female patients in the RATV group was superior to that in the RATA group (P = 0.031). The two groups were not statistically significant in the pro-menopausal female patients (P = 0.267). Furthermore, urinary function at 3 months postoperatively was not statistically significant in both groups in pre-menopausal and pro-menopausal patients (P = 0.711, P = 0.106) (Table 2).

Table 2 Comparison of postoperative Wexner and ICIQ-SF scores between RATV group and RATA group after PSM

Regarding the assessment of sexual function, There were 165 patients in this study of all patients, 41 in the RATV group and 124 in the RATA group. After 4 months postoperatively, this study investigated how sexually active the patients were in the last 4 weeks. Sexual function showed no significant differences in both groups in pre-menopausal and pro-menopausal patients before PSM (Table S5). Only 63 patients answered the FSFI questionnaire after PSM. 29 in the RATV group and 34 in the RATA group. There was no difference between the two groups in terms of sexual function assessment (P = 0.351). However, further subgroup analysis of the factor of whether or not they were menopausal found that the RATA group of pre-menopausal female patients had better sexual function than the RATV group (P = 0.045), especially in terms of sexual satisfaction (P = 0.042) (Table 3). Simultaneously, FSFI score ≤ 19 was defined in this study as an indication of possible sexual dysfunction. In this study, a one-way logistic regression analysis of Surgery Procedure, Age, Menopausal status, BMI, Maximum tumor diameter, Pathological TNM stage, and Histological Differentiation was performed, and no significant influence on the occurrence of sexual dysfunction was found (Table 4).

Table 3 Comparison of FSFI scores of patients in RATV group and RATA group after PSM

Table 4 Predictors of sexual dysfunction after surgery – univariate logistic regression analysis

Comparison of Wexner、ICIQ-SF and FSFI scores of patients in left colic artery-preserving (LCA-P) group and left colic artery-no preserving (LCA-NP) group

During the operation, the surgeon divided the patients into LCA-P group and LCA-NP group according to the length of the mesentery. There were no significant differences in urinary and sexual function between the two groups either before or after PSM, but anal function was better in the LCA-P group than in the LCA-NP group after PSM (P = 0.046) (Table S6, S7).

Long-term survival outcomes between RATV group and RATA group

At the last follow-up as of December 31, 2022, the median follow-up was 29 months (6–83) and 29 months (6–83) for the RATV group and the RATA group, respectively. The 3-year overall survival rate was 80.5% and 79.3% in the RATV group and the RATA group, respectively. The Log-rank test revealed no statistically insignificant difference in overall survival between the two groups (P = 0.642). Furthermore, both groups were not significantly different in 3-year disease-free survival (74.8% vs. 74.3%, P = 0.845) (Fig. 2).

Fig. 2

Comparison of Overall Survival and Disease-Free Survival Between two groups after propensity score matching (A-B). RATV, robotic-assisted transvaginal specimen extraction surgery; RATA, robotic-assisted transanal specimen extraction surgery

Discussion

With the advancement of gastrointestinal endoscopy, an increasing number of early-stage rectal cancers are being detected, and surgical interventions have significantly improved the survival rate of patients with rectal cancer. Nowadays, female patients’ expectations for surgery have shifted from merely improving survival rates to a greater emphasis on both short-term outcomes and long-term quality of life. In traditional robot-assisted radical rectal cancer resection, specimens are typically extracted through an abdominal incision, which can lead to both physical and psychological discomfort due to the incision’s appearance and potential postoperative complications. In cases of women, Natural Orifice Specimen Extraction Surgery (NOSES) allows specimens to be retrieved through the vagina or anus, avoiding large abdominal incisions and thus reducing the associated adverse effects on women. Here we tried to find optimal natural orifice specimen extraction for robotic middle rectal cancer resection in women: RATA and RATV.

Perioperative data is the most commonly used indicator for assessing the quality of these two surgical procedures. Our findings show that robotic transvaginal specimen extraction requires a longer operative time than transanal extraction, possibly due to the more complex surgical steps or vaginal anatomy in the transvaginal group. Unlike the randomized controlled trial by Park et al. [24], this study observed significantly less blood loss and quicker gastrointestinal function recovery for both procedures compared to conventional robotic surgery. Importantly, none of the robotic-assisted NOSES required an additional abdominal incision for specimen extraction, thus eliminating wound-related complications and facilitating early postoperative mobility and gastrointestinal recovery. Furthermore, the upper part of the vagina is mainly innervated by visceral nerves that are less sensitive to cutting pain, and the lower opioid dosage used postoperatively may reduce the incidence of nausea and vomiting, accelerating recovery and shortening hospital stays. This aligns with the principles of enhanced recovery after surgery [25].Surgical stress responses may accelerate pre-existing micrometastases or promote further tumor spread [26]. Our study found that postoperative inflammatory indexes were better in the transanal group, suggesting that this approach may be less disruptive and more conducive to patient recovery. The incidence of postoperative complications, including anastomotic fistula, was 16.4% in the transanal group and 14.9% in the transvaginal group, with no significant difference between them, echoing the results reported by Wolthuis et al. [27].

For female patients with mid-rectal cancer, both transvaginal and transanal specimen extractions can be beneficial, but selecting the appropriate procedure is challenging. Our study found that the cosmetic outcomes of both procedures led to patient satisfaction in the early postoperative period. However, concerns remain regarding their impact on physiological functions. Urinary, anal, and sexual dysfunction are serious complications of rectal surgery and significantly affect patients’ physical, mental health, and quality of life [28].Some studies indicate that robotic surgery systems favor the preservation of pelvic autonomic nerves, particularly for genitourinary function [29]. The lower incidence of postoperative urinary retention or infection in both groups further validates the advantage of robotic surgery.Pre-menopausal female rectal cancer patients with no reproductive requirements were included in this study, comprising 27% of the transanal group and 25% of the transvaginal group. Given the increasing incidence of rectal cancer in younger patients, assessing postoperative physiological function in pre-menopausal women is essential [30]. We found that anal function was better in the transvaginal group, especially among pre-menopausal patients, suggesting a pronounced advantage for transvaginal extraction in this subgroup. There was no difference in sexual function between the two groups; however, the transanal group showed better outcomes in pre-menopausal patients. This suggests a need for careful procedural selection in this population. Prospective studies have found no significant differences in anal or sexual function for peri-menopausal women. Further logistic regression analysis found that the surgical modality did not significantly affect sexual function, indicating that while the choice of surgery might reduce sexual function, it does not necessarily cause sexual dysfunction.

This study also considered the NOSES procedure’s requirement to drag the proximal bowel out of the anus or vagina, necessitating adequate mesenteric length. Adequate free bowel can be obtained without preserving the left colonic artery. However, the impact of preserving or not preserving the left colonic artery on organ function remains controversial. An Italian randomized controlled study [31] concluded that preserving the left colonic artery did not affect oncologic outcomes and better protected genitourinary function. In our study, anal function was found to be better in the group where the left colonic artery was preserved, indicating that mesenteric length considerations during robotic-assisted NOSES and postoperative anal issues warrant attention.

Conclusions

The postoperative outcomes of the robotic-assisted transanal and transvaginal specimen extraction groups were comparable, making both viable options for female rectal cancer patients. While the transanal group is more established and widely used, the potential for anal function problems, especially in pre-menopausal women, merits careful consideration. This study addresses the lack of research on postoperative physiological functions in female rectal cancer patients by comparing the two surgical procedures in terms of treatment outcomes, body image, urinary, anal, and sexual function. This provides a basis for more individualized treatment for female rectal cancer patients.