The Treatment Efficacy for Patients Undergoing Combined Transanal‐Transabdominal Endoscopic Resection of Rectal Anastomosis Stenosis

ABSTRACT

Background

Combined transanal and transabdominal resection for anastomotic stenosis may provide an opportunity to restore bowel continuity in patients with colorectal anastomotic stenosis. This study aimed to evaluate the therapeutic efficacy of combined transanal‐transabdominal resection in patients with rectal anastomotic stenosis.

Methods

We retrospectively analyzed a consecutive cohort of patients who underwent combined transanal‐transabdominal endoscopic resection for rectal anastomotic stenosis between August 2019 and March 2023. Data on intraoperative variables, mortality, postoperative complications, and stoma closure were collected. Functional outcomes were evaluated using the Low Anterior Resection Syndrome (LARS) score and the Wexner incontinence score.

Results

A total of 54 patients, including 46 men and 8 women, met the inclusion criteria. In all patients, anastomotic stenosis developed secondary to either anastomotic leakage or preoperative radiotherapy. The median length of hospital stay was 15 days (11–24 days). No postoperative mortality occurred, and the overall morbidity rate was 15%. During a median follow‐up of 48 months, stoma closure was achieved in 49 patients after a median interval of 3.8 months (2.5–11 months). At the end of follow‐up, stoma closure had failed in 2 patients because of poor anastomotic functional outcomes, and 3 patients developed recurrent anastomotic stenosis. Among the 49 patients available for functional assessment, 35 (71.4%) reported no or minor LARS. The median Wexner incontinence score was 8 (0–18), and 13 patients had a score of 0. Erectile function was evaluated in 28 male patients, of whom 18 reported normal postoperative erectile function.

Conclusions

Combined transanal‐transabdominal endoscopic resection appears to be an effective treatment for rectal anastomotic stenosis, with a high rate of stoma closure, low morbidity, and acceptable long‐term functional outcomes.

Clinical Trial Registration Number

NCT06036862.

1. Introduction

Rectal cancer ranks among the most prevalent malignancies worldwide, and its incidence continues to rise [1]. Currently, surgical resection remains the predominant therapeutic modality for advanced rectal cancer [2]. Despite the adoption of stapling devices and advancements in anastomotic techniques, the management of postoperative anastomotic complications remains a significant challenge for surgeons [3]. Anastomotic leakage is one of the most severe complications, and it has the potential to evolve into anastomotic stenosis (AS), especially in the context of middle and low rectal cancer patients who have undergone neoadjuvant radiotherapy [4]. Moreover, the incidence of AS varies between 3% and 30% [5]. AS can be categorized into two main types based on the degree and extent of stenosis: long‐segment stenosis and membranous stenosis. In cases of long‐segment severe AS, the current treatment strategy is either redo‐surgery or the establishment of a permanent stoma. Previous studies have explored the application of the transabdominal approach for redo‐surgery. However, conventional transabdominal methods encounter challenges in dealing with the complete resection of scar tissue around the stenotic intestinal segment due to limited visibility [6]. This limitation hinders the descent of the neorectum into the pelvis, potentially resulting in failed re‐anastomosis and the imposition of a permanent stoma [7, 8].

In an effort to address these issues, our team firstly reported the utilization of transanal endoscopy technique for the treatment of long segment AS in 2019 [9]. This method offers several advantages including minimizing trauma and an expanded field of view of the pelvic bottom in the treatment of rectal AS. We conducted a retrospective analysis of 17 patients who underwent resection of anastomotic stenosis through combined transabdominal‐transanal endoscopic surgery. The ileostomy or colostomy was successfully reversed in 15 patients, underscoring the safety and feasibility of this surgical method [9, 10, 11].

Our prior study exclusively reported perioperative outcomes, but the entire assessment of treatment efficacy of this innovative combined transanal‐ transabdominal approach in the treatment of rectal AS is still unknown. Therefore, this study aimed to assess the treatment efficacy for patients who have undergone resection of AS through the combined transanal‐ transabdominal method.

2. Materials and Methods

2.1. Study Population

The study protocol was approved by the Institutional Review Board. This study included 54 patients who presented with inflammatory scar hyperplasia due to anastomotic leakage and underwent combined transanal‐ transabdominal endoscopic resection for rectal stenosis and subsequent bowel reconstruction at our institution between August 2019 and March 2023, part of the patients in this study were from article published by our team in 2019 [9]. Patients who initially underwent surgery for conditions other than rectal cancer were excluded from this study. AS was diagnosed based on clinical symptoms, including constipation and bowel obstruction, along with endoscopic evidence of a tight narrowing that hindered the passage of a 12‐mm diameter sigmoidoscope through the anastomosis. Patient characteristics, encompassing age, gender, history of pelvic radiotherapy, and stenosis characteristics, were collected as part of the study. Pertaining to the resection and reconstruction surgery, detailed information regarding the surgical procedure, morbidity, the rate of temporary stomas, conversion rates, intraoperative events, and the duration of hospitalization were systematically documented. Functional outcomes were obtained through a combination of telephone follow‐up and on‐site questionnaire administration.

2.2. Inclusion and Exclusion Criteria

The inclusion criteria were the following: (1) The length of anastomotic stenosis was more than 5 mm, and a 12 mm diameter sigmoidoscope cannot through the anastomosis. (2) Anal digital examination, defecography, magnetic resonance imaging (MR), and CT suggesting the presence of anastomotic stenosis. (3) Previous treatments such as anal dilatation, balloon dilatation, and narrow incision that showed poor effectiveness.

The exclusion criteria were the following: (1) patients with unstable primary tumors or combined with tumors at other sites. (2) Primary surgery without sphincter preservation.

2.3. Surgical Techniques

Both laparoscopy and laparotomy were used as abdominal approaches for adhesiolysis and intra‐abdominal exploration, depending on the individual clinical scenario. In patients with inflammatory stenosis, adhesions between the mesentery and surrounding tissues were carefully dissected, with particular attention to preservation of the marginal vascular arcade. The proximal colon was mobilized toward the splenic flexure, and additional splenic flexure mobilization was performed when necessary to obtain sufficient bowel length for reconstruction. For the transanal approach, a purse‐string suture was placed 0.5–1 cm distal to the stenotic segment to achieve secure closure of the rectal lumen, followed by a circumferential incision of the bowel wall. When the stenosis was close to the anal canal, and a transanal single‐port device could not be inserted directly, full‐thickness circumferential division of the bowel wall was first performed approximately 0.5 cm distal to the purse‐string suture. After dissection of the bowel wall to approximately 4 cm from the anal verge, the transanal port was inserted. Pneumoperitoneum was then established using the AirSeal system at a constant pressure of 15 mmHg. The transanal phase was performed under endoscopic visualization using a TaTME platform to facilitate access and exposure; however, the procedure was not a redo TaTME or repeat transanal total mesorectal excision. Rather, the aim of the transanal dissection was to resect the stenotic bowel segment together with the surrounding extramural scar tissue and thereby facilitate redo reconstruction. Dissection proceeded in the correct anatomical plane, usually beginning laterally. During anterior dissection, careful attention was paid to identifying the plane between the rectum and the vagina in women or the prostate in men. During posterior dissection, care was taken to avoid injury to the presacral vessels. After completion of both the transabdominal and transanal dissections, the specimen was extracted transanally. Reconstruction was performed using either stapled or hand‐sewn anastomosis, depending on the distance between the distal bowel and the anal verge. For hand‐sewn anastomosis, the muscular layer was approximated with interrupted 2‐0 VICRYL sutures, followed by continuous full‐thickness closure with a 3‐0 barbed suture. In patients considered at high risk at the time of reconstruction, a Bacon procedure was performed. In such cases, the bowel was exteriorized through the anal canal to ~4 cm beyond the anal verge, and delayed definitive anastomosis was completed 2 weeks later. The key intraoperative steps are shown in Figure 1.

Figure 1.

Iconic procedures. (A) Stenosis under direct vision. (B) Circular excision of anal canal. (C) Exposure of conjoined longitudinal muscle. (D) Exposure of the anastomotic nail for the primary anastomosis. (E) Separation leaves room for re‐anastomosis. (F) anastomosis completion.

2.4. Long‐Term Result Evaluation

Functional assessments were conducted via telephone, mail, or clinic consultations, with patients providing informed consent to complete questionnaires associated with various standardized scores. All patients completed questionnaires pertaining to both the low anterior rectal syndrome (LARS) score and the Wexner Incontinence score.8

The LARS score is a standardized tool for evaluating functional outcomes following rectal surgery. It comprises five questions, each offering three to four possible responses graded on a scale of 0 to 16 points. The sum of these responses yields the LARS score, ranging from 0 to 42. Three distinct scoring levels exist: no LARS (scores between 0 and 20), minor LARS (scores between 21 and 29), and major LARS (scores between 30 and 42). In this study, poor long‐term functional outcomes were defined by the presence of major LARS or necessitating the creation of a definitive stoma due to suboptimal functional recovery after the surgical procedure [12].

The Wexner Incontinence Scale, comprising five questions encompassing various aspects of incontinence and assigning scores from 0 (no incontinence) to 20 (complete incontinence), represents a validated and internationally recognized questionnaire for quantifying the severity of incontinence [13].

Sexual function assessment was specifically conducted among male participants.

2.5. Data Analysis

Quantitative data conforming to a normal distribution are presented as the median and range. Qualitative data were summarized in terms of frequencies and percentages. Operative mortality was delineated as any deaths occurring within 90 days post‐surgery or prior to discharge. Early morbidity was characterized by complications arising during the hospital stay or within 30 days after discharge.

3. Results

3.1. Patient Characteristics

This study encompassed a cohort of 54 patients, comprising 46 men and 8 women, with a median age of 60 (range, 30–76) who underwent transanal and transabdominal AS resection between August 2019 and March 2023, which with a total follow‐up period of 48 months. Patient characteristics are summarized in Table 1. The primary indication for the initial surgery in all patients was rectal cancer. Notably, 32 patients (60%) had received radiotherapy as part of their treatment. The median BMI was 22.2 kg/m² (range, 15–32 kg/m²). The median length of stenosis was 42 mm (range, 19–80), with a distance from the stenosis site to the anal verge of 3.5 cm (range, 2.5–7). Additionally, 14 patients had a history of hypertension, and 13 patients had a history of diabetes mellitus (Table 1).

Table 1.

General information of the study cohort.

	Present cohort (N = 54)
Follow‐up period	14 (2–48)
Gender
Female	8
Male	46
Age	60 (30–76)
> 60	25
≤ 60	29
BMI	22.2 (15–32)
Primary anastomosis type
Handsewn	14
Stapled	40
Previous surgery history
Laparoscopic Dixon	45
Laparotomic Dixon	9
History of pelvic radiotherapy
Yes	32
No	22
Length of stenosis (mm)	42 (19–80)
Distance from stenosis to anal verge (cm)	3.5 (2.5–7)
Medical history
Hypertension	14
Diabetes mellitus	13

3.2. Intraoperative Results

Out of the 54 patients who underwent transanal and transabdominal endoscopic surgery, 45 patients underwent a laparoscopic procedure, while the remaining 9 patients required a laparotomy. The choice between laparoscopy and laparotomy was primarily influenced by the patients' surgical history and the presence of multiple intra‐abdominal adhesions. The median duration of the operative procedure was 222 min (range, 98–492), and the median blood loss was 110 mL (range, 60–2000). Notably, a 60‐year‐old male patient experienced presacral hemorrhage during the operation and required a blood transfusion. Following standard surgical protocols, 14 patients received stapler anastomosis, 36 patients underwent hand‐sewn anastomosis, and the Bacon procedure was employed in high‐risk cases where anastomosis was prone to failure. The final anastomosis, according to the Bacon technique, was performed after a 2‐week interval. Ultimately, 52 patients received a loop ileostomy, while two men (ages 71 and 69) necessitated a colostomy (Table 2).

Table 2.

Intraoperative and postoperative data for the study cohort.

	Our cohort (n = 54)
Operative time (min)	222 (98–492)
Blood loss (mL)	110 (60–2000)
Blood transfusion rate	1/54
Surgical procedure
Laparoscopy	45
Laparotomy	9
Intraoperative complication
Presacral hemorrhage	1
Anastomotic method
Stapler	36
Hand‐sew	14
Bacon	4
Length of hospital (day)	15 (11–24)
Mortality	0
Morbidity	n = 8 (15%)
Anastomotic leakage	5
Pelvic abscess	0
Bowel obstruction	3
Bladder injury	0

3.3. Postoperative Results

The median duration of hospitalization was 15 days (range, 11‐24 days). Importantly, there were no recorded fatalities in the cohort. Complication morbidity was noted in 15% of patients, with five individuals experiencing postoperative anastomotic leakage. An additional three patients encountered episodes of incomplete intestinal obstruction, which notably improved following conservative treatment. Fortunately, there were no reported instances of pelvic abscess or injuries to adjacent organs.

3.4. Stoma Closure and Functional Results

Following a median delay of 3.8 months (range, 2.5–11), stoma closure was successfully performed in 49 out of 54 patients, resulting in a commendable success rate of 91%. (Figure 2.) However, five patients did not undergo stoma closure. Failed stoma closure occurred in four cases due to poor functional outcomes marked by fecal incontinence, while one patient experienced a recurrence of anastomotic stenosis. Notably, the functional outcome assessment was conducted for 49 patients. Among those, 35 patients (71.4%) exhibited either no or minor LARS. The median Wexner incontinence score was 8 (range, 0–18), and thirteen patients achieved complete fecal continence with a Wexner score of 0. It is worth mentioning that sexual function assessment was exclusively conducted among male patients. Out of the 28 male patients who completed the assessment, eighteen reported normal erectile function. Comprehensive scores and detailed long‐term results are summarized in Table 3.

Figure 2.

Preoperative and reviewed during follow‐up images. (A) Colonoscopy revealed a narrow intestinal segment. (B) Colonoscopy showed good anastomotic function during postoperative follow‐up. (C) Preoperative defecography showed low anastomotic stenosis. (D) Postoperative defecography showed reconstruction of intestinal continuity. The Red arrow points to the narrow part.

Table 3.

Long‐term functional results for the study cohort.

	present cohort
Patients	n = 54 (100%)
Stoma closure time	3.8 (2.5–11)
Closure rate	49 (91%)
Overall follow‐up months	14 (5–48)
Functional results
Patients	n = 49 (83%)
LARS score
0–20 NO LARS	18
21–29 Minor LARS	17
30–42 Major LARS	14
Wexner incontinence score
Median	8 (0–18)
Score = 0	13
Patients	n = 28 (76%, 35)
Normal Erection	18

4. Discussion

Redo surgery for failed anastomosis, particularly resection of rectal stenosis followed by bowel reconstruction, remains one of the most technically demanding procedures in colorectal surgery [14]. Previous studies have examined the feasibility and safety of transabdominal redo surgery for failed anastomosis. In the cited study, the reported patients with rectal AS who underwent transabdominal redo surgery ultimately received a permanent stoma [6], although this result should not be generalized to all transabdominal redo procedures. Case reports have also suggested the feasibility of Re‐TaTME for anastomotic complications and laparoscopic redo coloanal anastomosis for rectovaginal fistula [15, 16]. In addition, Burke et al. demonstrated the advantages of transanal endoscopy in this setting [17]. To the best of our knowledge, the present study is the first to evaluate both short‐term and long‐term functional outcomes after combined transanal‐transabdominal resection for AS. Our findings suggest that this approach is associated with a high stoma closure rate and favorable long‐term functional outcomes.

Redo surgery for AS is generally associated with a substantial risk of postoperative complications and anastomotic failure [6]. The transanal endoscope provides magnified visualization in the deep pelvis, which may help reduce intraoperative bleeding and injury to adjacent organs. Presacral hemorrhage is widely recognized as one of the most serious complications during resection of rectal anastomotic stenosis and bowel reconstruction [18]. Teak Woo et al. reported a redo case complicated by presacral venous plexus injury [19]. Similarly, Genser et al. reported 50 patients undergoing transabdominal redo surgery, among whom 12 required intraoperative blood transfusion, with a mean of 2.5 units per patient; five sustained bladder injury, and one required splenectomy because of severe adhesions [7]. Lefevre et al. also reported five bladder repairs and one caudal pancreatectomy [8]. In contrast, no adjacent organ was inadvertently injured or resected in our series. Direct visualization of the stenotic bowel segment also facilitated distal dissection and reconstruction, which is particularly relevant in patients with multiple prior abdominal operations, severe adhesions, and pelvic fibrosis [7]. In our study, one patient developed presacral hemorrhage because of multiple previous abdominal surgeries and extensive pelvic fibrosis, but recovered after appropriate management and maintained a functional anastomosis during follow‐up. Previous studies frequently required laparotomy in this setting. Genser et al. and Pitel et al. performed surgery through laparotomy [7, 20], while Lefevre et al. reported a 50% conversion rate from laparoscopy to laparotomy because of extensive adhesions [8]. Teak Woo et al. likewise used open surgery in 13 patients with severe adhesions or incomplete anastomoses [19]. In our cohort, 45 patients underwent laparoscopy combined with transanal endoscopy, with a mean transanal operating time of 78 min, whereas the remaining 9 underwent open surgery because of a high risk of presacral hemorrhage, with the transanal phase still completed endoscopically and averaging 83 min. The median total operative time was 222 min, which is comparable to previous reports. These findings suggest that transanal endoscopy may reduce the technical difficulty of redo surgery and decrease the need for laparotomy, thereby minimizing surgical trauma and facilitating postoperative recovery. Five patients developed anastomotic leakage because of poor general condition, and three developed bowel obstruction. No acute severe complications were observed (Table 4).

Table 4.

Perioperative and long‐term functional results in patients undergoing redo‐anastomosis and reconstruction between different studies.

		Our cohort (n = 54)	Genser 2013 (n = 50)	Pitel 2012 (n = 66)	Lefevre2011 (n = 33)	TeakWoo2018 (n = 32)
Surgical procedure		Transanal transabdominal combined	Transabdominal	Transabdominal	Transabdominal	Transabdominal
Transabdominal procedure	Laparoscopic	45	0	0	6 (3 convert to open)	19
	open	9	50	66	27	13
Age		60 (30–76)	62 (40–80)	60 (27–79)	57 (39–79)	53 (43–83)
Cause of stenosis/redo‐surgery	Chronic abscess	3	14	21	5	13
	Hartmann	0	8	13	6	0
	Inflammation/AS	47	20	10	17	7
	Fistula	4	3	22	—	12
	Recurrence	0	55	0	5	0
Operative time (min)	Average	231	422	—	279	223.4
	Median	222	435	—	—	255
	Range	98–492	170–660	—	133–480	80–480
Blood loss (mL)	Average	152	—	—	—	81.7
	Median	110	—	—	—	80
	Range	60–2000	—	—	—	30–1000
Complication	Bladder damage	0	5	—	2	—
	Anastomotic leakage	5		5	4	4
	Splenectomy	0	1	—	—	—
	Bowel obstruction	3		—	—	3
Length of hospitalization	Average	15	15	13	16.5	14
	Median	15	13	14	—	—
	Range	(11–24)	7–47	6–31	6–60	6–195
Long‐term outcome
Duration of follow up (months)	Median	14	21	35.7	—	30.3
	Range	(5–48)	1–137	0–122.4	—	5–137
Stoma closure (months)	Closure rate	91%	88%	78.8%	79%	78.1%
	Median closure time	3.8	3	2.2	3.9	—

Protective enterostomy is intended to be temporary and may reduce the consequences of anastomotic leakage while improving the chance of restoring bowel continuity [21, 22], although stoma closure may still fail. A recent review reported an overall bowel continuity restoration rate of 79% after redo surgery for complicated coloanal or colorectal anastomoses [23]. However, colorectal anastomosis accounted for a substantial proportion of those cases, whereas coloanal anastomosis is generally more challenging. Low anastomosis is closely associated with anastomotic leakage, suggesting that coloanal anastomosis may carry a higher risk of failure than colorectal anastomosis [24]. In the present study, all patients underwent coloanal anastomosis because of the low level of reconstruction. Previous studies have also identified radiotherapy as an important risk factor for failed stoma closure [25, 26]. Calmel et al. developed a predictive model showing that a preoperative predictive score of ≥ 3 points was associated with the greatest risk of redo‐anastomosis failure [25]. In our cohort, 60% of patients had received preoperative radiotherapy, and 85% were male, both of which are generally considered unfavorable factors for pelvic redo surgery. Despite these baseline challenges, 49 patients ultimately underwent stoma closure and achieved a functional anastomosis, resulting in a stoma closure rate of 90%, which was at least comparable to previously reported outcomes. Long‐term functional outcomes were assessed in 49 patients, as three declined the questionnaire and two were lost to follow‐up. Thirteen patients (26%) had a Wexner incontinence score of 0, which is slightly higher than that reported by Pitel et al. In addition, 35 patients (71.4%) had no LARS or only minor LARS, whereas Collard et al. reported major LARS in 50% of patients [27]. Erectile function was assessed in 28 male patients, of whom 18 reported normal postoperative erectile function. These findings suggest that, compared with transabdominal resection alone, the transanal approach may facilitate more precise entry into the correct anatomical plane, thereby improving preservation of the pelvic nerves while allowing complete resection of the fibrotic stenotic segment. However, the limited number of patients available for sexual function assessment precludes firm conclusions regarding this outcome. This study is limited by its retrospective design, and we have therefore registered a prospective clinical study (NCT06036862). Nevertheless, the favorable stoma closure rate and functional outcomes observed in this cohort suggest that this approach may provide meaningful long‐term benefit and improve quality of life compared with living with a defunctioning stoma.

5. Conclusion

The combined transanal‐ transabdominal endoscopic resection has been proven to be an effective approach for anastomotic stenosis. This method is distinguished by its high stoma closure success rate, low complication morbidity, and favorable long‐term functional outcomes. It is particularly beneficial in cases of low and long‐segment stenosis that may pose challenges for traditional transabdominal surgical techniques.

Author Contributions

Conceptualization: Liang Huang and Taixuan Wan. Data curation: Haoqi Zheng and Yunxing Shi. Formal analysis: Hao Xie, and Maram Alenzi. Funding acquisition: K.L. and Liang Huang. Investigation: Shuangling Luo and Zhanzhen Liu. Methodology: Liang Huang and Pinzhu Huang. Project administration: Pinzhu Huang and Liang Huang. Resources: Liang Huang and Liang Kang. Software: Taixuan Wan and Fujin Ye. Supervision: Liang Huang and Liang Kang. Validation: Liang Huang and Pinzhu Huang. Visualization: Pinzhu Huang. and Haoqi Zheng Writing – original draft: Taixuan Wan and Yunxing Shi Writing – review and editing: Maram Alenzi and Liang Huang.

Ethics Statement

The study protocol received was approved by the Ethics Committee of the Sixth Affiliated Hospital, Sun Yat‐Sen University (Ethics Approval ID: 2014ZSLYEC‐013). This study was adhered to the Declaration of Helsinki. There were no experiments on humans or the use of human tissue samples in this study.

Consent

All authors agree to the publication.

Conflicts of Interest

The authors declare no conflicts of interest.

Synopsis

This retrospective study evaluated combined transanal‐transabdominal endoscopic resection for rectal anastomotic stenosis in 54 patients. The procedure achieved a high stoma closure rate, low postoperative morbidity, and acceptable long‐term bowel and sexual functional outcomes, suggesting that it may be an effective treatment option for complex low rectal anastomotic stenosis.

Wan T., Shi Y., Zhou Y., et al., “The Treatment Efficacy for Patients Undergoing Combined Transanal‐Transabdominal Endoscopic Resection of Rectal Anastomosis Stenosis,” Journal of Surgical Oncology 134 (2026): 531‐538, 10.1002/jso.70316.

Data Availability Statement

Colorectal Disease (CODI). The datasets used or analyzed during the current study are available from the corresponding author on reasonable request.