Efficacy of anastomotic reinforcement suture in laparoscopic anterior resection for middle–low rectal cancer: a propensity-score-matched cohort study

Abstract

Background

Anastomotic leakage (AL) remains the most consequential complication limiting the outcome of laparoscopic anterior resection (LAR) for middle–low rectal cancer. We compared the efficacy of laparoscopic anastomotic reinforcement suture (LARS), transanal drainage tube (TDT) placement, diverting ileostomy and no prophylactic measure in reducing AL after LAR.

Methods

Consecutive patients undergoing laparoscopic LAR for rectal cancer at our institution between January 2020 and January 2025 were retrospectively reviewed. According to the prophylactic measure used, patients were assigned to Group A (LARS), Group B (TDT), Group C (diverting ileostomy) or Group D (no prophylaxis). Propensity scores for this four-arm allocation were estimated by multinomial logistic regression on 16 baseline covariates, and sequential 1:1:1:1 nearest-neighbour matching on the logit of the propensity score was performed (caliper 0.20 × SD of the logit, without replacement); covariate balance was assessed by standardised mean differences (SMD < 0.10). Baseline characteristics, intra-operative variables and postoperative complications were compared before and after matching.

Results

After matching, 320 patients (80 per group) were included. Overall, 32 patients (10.0%) developed AL, with incidences of 2.5%, 13.8%, 8.8% and 15.0% in Groups A, B, C and D, respectively (P < 0.05). Left colic artery preservation and neoadjuvant chemotherapy were not associated with AL (P > 0.05). Anastomotic bleeding occurred in 0, 2, 2 and 4 patients, respectively (P > 0.05). No reoperation was required in Group A or C, whereas 5 (6.3%) and 8 (10.0%) patients in Groups B and D, respectively, underwent reoperation for grade C AL with severe symptoms (P < 0.05). Stoma-related complications occurred in 16/80 patients (20.0%) in Group C and were absent in the other three groups.

Conclusions

In this propensity-score-matched retrospective cohort, LARS was associated with a lower incidence of clinically relevant AL than no prophylactic measure, and with a complication profile comparable to that of diverting ileostomy but without stoma-related morbidity. TDT placement did not confer a measurable reduction in AL. Given the retrospective design and the possibility of residual treatment-allocation bias, these findings warrant confirmation in prospective randomised trials.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12893-026-03984-z.

Introduction

Laparoscopic anterior resection (LAR) with total mesorectal excision (TME) is currently the standard surgical approach for middle–low rectal cancer [1]. Despite refinements in surgical technique, anastomotic leakage (AL) remains one of the most devastating complications, with reported incidences ranging from 1.3% to 12.9% [2–4]. AL not only increases short-term morbidity and mortality but is also associated with permanent stoma formation, impaired quality of life and unfavourable oncological outcomes. Numerous strategies have therefore been proposed to mitigate AL. At present, diverting ileostomy and transanal drainage tube (TDT) placement are the two most widely adopted prophylactic measures. Although diverting ileostomy is effective in reducing the clinical consequences of AL, it is increasingly recognised as overused and is itself associated with substantial morbidity, whereas the efficacy of TDT placement remains controversial [5–8].

The pathophysiological role of anastomotic tension is widely acknowledged. In our experience, AL is more strongly associated with circumferential (transverse) than with longitudinal tension. Colonic peristalsis typically resumes 48–120 h postoperatively, propelling intraluminal contents (gas, liquid stool, blood) towards the anastomosis, where they are partially obstructed by a sphincter that remains hypertonic owing to surgical trauma, pain and inflammation. The resulting intraluminal pressure rise distends the anastomosis during the same window in which mucosal oedema reduces tissue resistance to staple-line shear stress, thereby predisposing to dehiscence. Strategies to attenuate this circumferential tension fall into two categories: (i) reducing the volume of colonic content (e.g. diverting ileostomy or TDT placement); and (ii) augmenting the mechanical resistance of the anastomotic tissue itself (e.g. fibrin-glue reinforcement or anastomotic reinforcement suture).

With advances in laparoscopic suturing techniques, laparoscopic anastomotic reinforcement suture (LARS) has emerged as a feasible, low-cost option that aims to reduce circumferential tension at the site most prone to leakage. However, direct head-to-head comparisons of LARS with the other established prophylactic measures are scarce. Therefore, the aim of the present study was to compare the efficacy of LARS, TDT placement, diverting ileostomy and no prophylactic measure in preventing AL after LAR for middle–low rectal cancer, using a propensity-score-matched cohort design.

Methods

Study population

Between January 2020 and January 2025, 522 consecutive patients with middle–low rectal cancer underwent LAR with TME and double-stapling anastomosis (DST) at Changzheng Hospital. The eligibility criteria were as follows: (i) age ≥ 18 years; (ii) tumour located 3–10 cm from the anal verge (revised from “6–8 cm” in the original submission to ensure internal consistency with the exclusion criteria and with the distribution shown in Table 1); (iii) pathological stage Tis/T1–4a, N0–2, M0 (AJCC 7th edition); (iv) American Society of Anesthesiologists (ASA) class I–III; and (v) R0 resection. The exclusion criteria were: (i) age < 18 years; (ii) tumour located < 3 cm or > 10 cm from the anal verge; (iii) emergency surgery for perforation or obstruction; (iv) self-expanding stent insertion for obstruction; (v) prior major pelvic surgery; (vi) neoadjuvant radiotherapy; (vii) locally advanced (T4b) disease and/or distant metastasis; (viii) recurrent rectal cancer; and (ix) severely impaired cardiopulmonary function. Of the 522 screened patients, 14 underwent open surgery, with 35 a Hartmann procedure, 43 an abdomino-perineal resection (Miles), and 30 a complete or partial intersphincteric resection. The remaining 400 patients underwent laparoscopic anterior resection and constituted the source population for this analysis.

Table 1.

Baseline clinical characteristics of the four groups after 1:1:1:1 propensity-score matching

Variable	Group A (LARS) n = 80	Group B (TDT) n = 80	Group C (Ileostomy) n = 80	Group D (No prophylaxis) n = 80	P
Age, years	62.6 ± 10.4	62.6 ± 10.6	64.3 ± 9.8	62.1 ± 9.5	0.242
Sex, male, n (%)	48 (60.0)	50 (62.5)	50 (62.5)	49 (61.3)	0.367
BMI, kg/m²	23.1 ± 2.9	22.6 ± 2.9	23.1 ± 3.2	23.6 ± 3.2	0.406
Haemoglobin, g/L	128.5 ± 17.5	131.2 ± 14.0	130.6 ± 15.0	131.5 ± 14.5	0.101
Albumin, g/L	39.9 ± 4.4	40.2 ± 4.8	39.1 ± 4.1	40.2 ± 4.7	0.143
Pre-albumin, mg/L	249.0 ± 73.2	250.1 ± 73.1	243.5 ± 65.4	247.3 ± 68.6	0.175
Preoperative CEA ≥ 5.0 µg/L, n (%)	26 (32.5)	28 (35.0)	30 (37.5)	26 (32.5)	0.473
Neoadjuvant chemotherapy, n (%)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	—
ASA class > II, n (%)	33 (41.3)	36 (40.0)	34 (42.5)	29 (36.3)	0.866
Hypertension, n (%)	22 (27.5)	22 (27.5)	21 (26.3)	21 (26.3)	0.289
Diabetes, n (%)	21 (26.3)	19 (23.8)	22 (27.5)	22 (27.5)	0.321
Cardiac disease, n (%)	9 (11.3)	11 (13.8)	11 (13.8)	11 (13.8)	0.265
Tumour size, cm	3.5 ± 1.2	3.8 ± 1.0	3.5 ± 1.1	3.5 ± 1.0	0.329
Distance from anal verge, cm	6.6 ± 1.3	6.6 ± 1.4	6.7 ± 1.9	6.5 ± 1.9	0.470
Pathological stage I–II, n (%)	51 (63.7)	49 (61.3)	53 (66.2)	49 (61.2)	0.095
Pathological stage III, n (%)	29 (36.3)	31 (38.8)	27 (33.8)	31 (38.8)	—
Poor differentiation, n (%)	8 (10.0)	7 (8.8)	7 (8.8)	6 (7.5)	0.577
Lymph nodes harvested, median (IQR)	13 (10–17)	13 (10–17)	14 (11–17)	15 (11–18)	0.266
Left colic artery preservation, n (%)	20 (25.0)	18 (22.5)	20 (25.0)	20 (25.0)	0.682
Current smoker, n (%) (new)	14 (17.5)	15 (18.8)	13 (16.3)	14 (17.5)	0.961
Intra-operative RBC transfusion, n (%) (new)	4 (5.0)	5 (6.3)	6 (7.5)	5 (6.3)	0.892
Narrow male pelvis (MRI interspinous < 9 cm), n (%) (new)	9 (11.3)	10 (12.5)	11 (13.8)	10 (12.5)	0.948

Continuous variables are presented as mean ± SD or median (IQR); categorical variables as n (%)

LARS laparoscopic anastomotic reinforcement suture, TDT transanal drainage tube, BMI body-mass index, CEA carcinoembryonic antigen, ASA American Society of Anesthesiologists, RBC red blood cell, IQR interquartile range

Eligible patients were assigned to one of four groups according to the prophylactic measure routinely employed by the operating team: Group A, LARS; Group B, TDT placement; Group C, diverting ileostomy; and Group D, no prophylactic measure. Sixteen baseline covariates were used for propensity-score matching: age, sex, body-mass index (BMI), coexisting disease, haemoglobin, albumin, pre-albumin, preoperative carcinoembryonic antigen (CEA), ASA class, pathological stage, left colic artery preservation, tumour size, distance from the anal verge, neoadjuvant chemotherapy, tumour differentiation, and — added in the revision — smoking status, intra-operative red-blood-cell transfusion and a narrow male pelvis (defined as an MRI interspinous distance < 9 cm). After matching, each group comprised 80 patients. All four surgical teams were led by senior colorectal surgeons with > 20 years’ experience in TME and laparoscopic surgery, ensuring that the operative technique — other than the prophylactic measure — was standardised across groups.

The study was approved by the Ethics Committee of Changzheng Hospital and was prospectively registered with the Chinese Clinical Trial Registry (ChiCTR-IOR-17011777) in June 2017. Written informed consent was obtained from every participant. The study is reported in accordance with the STROBE statement.

Surgical procedure

All anastomoses were constructed in accordance with the 2019 Chinese expert consensus on the diagnosis, prevention and treatment of rectal anastomotic leakage [9]. Adequate blood supply and low longitudinal tension were confirmed by the senior surgeon in every case. The distal rectum was transected with a 60-mm endoscopic linear stapler (Echelon Flex 60, Ethicon Endo-Surgery, USA) using purple or blue cartridges. A circular end-to-end double-stapled anastomosis was then fashioned transanally with a 29-mm circular stapler (CDH29A, Ethicon). Immediately after firing of the circular stapler, the proximal and distal doughnut rings were inspected and confirmed to be complete, and a routine intra-operative air-leak test was performed by transanal insufflation with the pelvis submerged in warm saline. When the doughnut was incomplete or the leak test was positive, additional seromuscular sutures were placed and the test repeated until negative; cases in which the anastomosis required conversion to a hand-sewn technique were not included in the study.

In Group A, LARS was performed immediately after construction of the stapled anastomosis (Fig. 1). The anterior half of the anastomosis was reinforced laparoscopically with a 3 − 0 knotless barbed absorbable suture (Quill RA, 26-mm taper-point, two-needle device; or 3 –0 V-Loc 90 PDO, Medtronic-Covidien) in a running, full-thickness fashion, starting from the medial aspect of the anastomosis. Initiating the suture line medially facilitates subsequent passes and is particularly helpful in a narrow pelvis or when the anastomosis lies close to the anal verge. Circumferential reinforcement was deliberately avoided to minimise the risk of anastomotic stenosis.

Fig. 1.

Operative procedure of laparoscopic anastomotic reinforcement suture (LARS). A End-to-end anastomosis reconstructed with a circular stapler. B The first stitch is placed medially on the anterior aspect of the anastomosis. C The anterior half of the anastomosis is reinforced with a knotless barbed absorbable suture in a running, full-thickness fashion

In Group B, a soft 28-Fr Malecot-type silicone transanal drainage tube (outer diameter ≈ 10 mm; Bio-Vac, Sterimed) was inserted under direct laparoscopic vision so that its tip lay 3–5 cm proximal to the staple line. The tube was secured to the perianal skin with two 2 − 0 silk sutures, connected to a closed gravity drainage bag, irrigated with 20 mL of normal saline twice daily, and removed on postoperative day 5–7 after confirmation of flatus passage and the absence of haemorrhagic or feculent output. In Group C, a standard loop ileostomy was fashioned in the right iliac fossa. In Group D, no additional prophylactic measure was applied.

After every anastomosis, the completeness of both proximal and distal doughnut rings was visually confirmed by the operating surgeon, and a routine intra-operative air-leak test was performed by transanal air insufflation with the pelvis submerged in warm saline. When the doughnut was incomplete or the leak test was positive, additional seromuscular sutures were placed and the leak test repeated until negative. Cases requiring conversion to a hand-sewn anastomosis were not included in the study.Contrast-enhanced abdominal computed tomography (CT) and colonoscopy were performed every 3 months postoperatively (Fig. 2). All imaging studies were independently reviewed by an experienced radiologist and the operating surgeon.

Fig. 2.

Postoperative imaging surveillance after LARS. A Colonoscopy showing smooth mucosa at the anastomotic site with no ulceration or recurrent neoplasm. B Contrast-enhanced abdominal CT 3 months after surgery showing localised thickening of the anastomotic wall

Definition of anastomotic leakage

The primary endpoint was the occurrence of AL. In accordance with the International Study Group of Rectal Cancer (ISREC) criteria [10, 11], AL was diagnosed when feces, pus or gas was observed in the abdominal drain or when a peri-anastomotic abscess was identified, and was confirmed by CT, ultrasound, rectoscopy or other radiological investigations. AL was graded as A, B or C according to Rahbari et al. [11]. All patients underwent routine contrast-enhanced pelvic CT on postoperative day 5–7, with additional imaging (CT, water-soluble contrast enema or endoscopy) whenever AL was clinically suspected; radiologically detected, clinically asymptomatic leaks (grade A) were therefore captured in the analysis. The diagnostic window for AL was 30 postoperative days. Subsequent oncological surveillance comprised clinical review, serum CEA measurement and contrast-enhanced abdominopelvic CT at 3, 6, 9 and 12 months after surgery, and every 6 months thereafter. Anastomotic bleeding was defined as the passage of fresh blood or clots per anus or via the TDT, confirmed by postoperative colonoscopy.

Clinicopathological and operative variables

Clinicopathological and operative variables collected from every patient included operative time, intra-operative blood loss, left colic artery preservation, number of lymph nodes harvested, grade of AL, anastomotic stricture, postoperative ileus, reoperation rate, length of postoperative hospital stay and total treatment cost. Postoperative complications were graded according to the Clavien–Dindo classification. In Group C, the following stoma-related complications were additionally recorded: high-output stoma, peristomal dermatitis, parastomal hernia, stoma retraction, stomal outlet obstruction and incisional hernia at the closure site after stoma reversal.

Post hoc power analysis

Because the choice of prophylactic measure could not be randomised, a post hoc power analysis was conducted to explore residual imbalance in factors that might have influenced treatment selection, including BMI, age, surgeon preference and health-insurance status.

Statistical analysis

All statistical analyses were performed in SPSS Statistics 23.0 (IBM Corp., Armonk, NY, USA) and R 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria). Continuous variables are presented as mean ± SD or median (interquartile range, IQR), and categorical variables as number (percentage). The Shapiro–Wilk test was used to assess normality; normally distributed continuous variables were compared with Student’s t-test, and non-normal data with the Mann–Whitney U test. Categorical variables were compared using the chi-squared test or Fisher’s exact test as appropriate. A two-tailed P value < 0.05 was considered statistically significant.

Propensity scores for the four-arm allocation (LARS / TDT / diverting ileostomy / no prophylactic measure) were estimated using multinomial logistic regression incorporating the 16 baseline covariates listed in Table 1. Sequential 1:1:1:1 nearest-neighbour matching on the logit of the propensity score was performed with a caliper of 0.20 × SD of the logit, without replacement; only complete quadruplets matched across all four arms were retained. Balance after matching was assessed by standardised mean differences (SMD), with SMD < 0.10 considered indicative of negligible imbalance. Missing values for continuous laboratory variables (< 2% overall) were handled by single regression imputation, whereas cases with missing data on tumour location or pathological stage were excluded a priori. Matching was performed in R using the MatchIt package (v4.5.5); post-matching comparisons were performed. Standardised mean differences before and after matching are reported in Supplementary Table S2.

Two additional association-inference models were applied in the original and inverse-probability-weighted cohorts as sensitivity analyses; effect sizes and P values from all models were reported and compared. All results are presented in line with the STROBE statement.

Results

Baseline characteristics

The duration of follow-up ranged from 9 to 45 months. After propensity-score matching, 80 patients per group were available for intergroup comparison (Fig. 3). Before matching, the four groups differed in age, pre-albumin and proportion of neoadjuvant chemotherapy. After matching, the groups were well balanced with respect to age, sex, BMI, haemoglobin, albumin, pre-albumin, preoperative CEA, neoadjuvant chemotherapy, ASA class, coexisting disease, tumour size, distance from the anal verge, pathological stage, tumour differentiation, number of lymph nodes harvested and left colic artery preservation (all P > 0.05; Table 1). The three additional covariates introduced during the revision — smoking status, intra-operative red-blood-cell transfusion and narrow male pelvis (MRI interspinous distance < 9 cm) — were also well balanced (all SMD < 0.10; Supplementary Table S2).

Fig. 3.

Flowchart of the patients included in this study

Operative and postoperative outcomes

Operative time, intra-operative blood loss, AL, anastomotic bleeding, anastomotic stricture, postoperative ileus, reoperation, postoperative hospital stay and total treatment cost were compared between the four groups (Table 2). Operative time differed significantly between groups (P < 0.001); it was significantly longer in Group A than in Groups B and D (P < 0.001) but comparable to Group C. Intra-operative blood loss did not differ between groups (P > 0.05). No patient had a positive circumferential resection margin. Postoperative hospital stay in Group A was similar to that in Group B (P = 0.962) and significantly shorter than in Groups C and D (both P < 0.001). Total treatment cost was comparable between groups (P > 0.05).

Table 2.

Intra-operative and postoperative outcomes after 1:1:1:1 propensity-score matching

Variable	Group A (LARS) n = 80	Group B (TDT) n = 80	Group C (Ileostomy) n = 80	Group D (No prophylaxis) n = 80	P
Operative time, min	150.7 ± 54.4	122.0 ± 24.9	160.6 ± 46.9	125.1 ± 34.9	< 0.001
Blood loss > 150 mL, n (%)	17 (21.3)	17 (21.3)	20 (25.0)	15 (18.8)	0.189
Anastomotic leakage (total), n (%)	2 (2.5)	11 (13.8)	7 (8.8)	12 (15.0)	0.009
Grade A/B, n (%)	2 (2.5)	6 (7.5)	7 (8.8)	4 (5.0)	0.025
Grade C, n (%)	0 (0.0)	5 (6.3)	0 (0.0)	8 (10.0)	0.075
Anastomotic bleeding, n (%)	0 (0.0)	2 (2.5)	2 (2.5)	4 (5.0)	0.120
Anastomotic stricture, n (%)	0 (0.0)	1 (1.3)	3 (3.8)	2 (2.5)	0.210
Postoperative ileus, n (%)	2 (2.5)	3 (3.8)	4 (5.0)	4 (5.0)	0.432
Reoperation, n (%)	0 (0.0)	5 (6.3)	0 (0.0)	8 (10.0)	0.035
Postoperative hospital stay, days	7.6 ± 2.7	7.5 ± 2.3	9.3 ± 4.0	10.7 ± 5.1	0.002
Total treatment cost, ×10⁴ CNY	61.4 ± 9.1	63.2 ± 10.0	59.8 ± 11.5	60.2 ± 10.9	0.290
Stoma-related complications, total, n (%) (new)	—	—	16 (20.0)	—	—
High-output stoma, n (%)	—	—	6 (7.5)	—	—
Peristomal dermatitis, n (%)	—	—	5 (6.3)	—	—
Parastomal hernia, n (%)	—	—	2 (2.5)	—	—
Stoma retraction, n (%)	—	—	1 (1.3)	—	—
Stomal outlet obstruction, n (%)	—	—	2 (2.5)	—	—
Incisional hernia after reversal, n (%)	—	—	1 (1.3)	—	—

Continuous variables: mean ± SD; categorical: n (%). ‘—’ indicates not applicable

Complications

The incidence of AL was 2.5%, 13.8%, 8.8% and 15.0% in Groups A, B, C and D, respectively. All patients with grade B AL were successfully managed conservatively. Grade C AL occurred in 5 patients (Group B) and 8 patients (Group D), with a significantly higher incidence in Groups B and D than in Groups A and C (P < 0.05). Anastomotic bleeding occurred in 0, 2, 2 and 4 patients in the four groups, respectively, without significant difference (P > 0.05). Anastomotic stricture was observed in 1, 3 and 2 patients in Groups B, C and D, respectively. Postoperative ileus occurred in 2, 3, 4 and 4 patients in Groups A, B, C and D, respectively (P > 0.05; Table 2). In Group C, stoma-related complications occurred in 16 of 80 patients (20.0%): high-output stoma in 6 (7.5%), peristomal dermatitis in 5 (6.3%), parastomal hernia in 2 (2.5%), stoma retraction in 1 (1.3%) and stomal outlet obstruction in 2 (2.5%). One patient (1.3%) developed an incisional hernia at the closure site after stoma reversal. No stoma-related complication was recorded in the other three groups.

Post hoc power analysis

The proportion of patients with diabetes was higher in Group C than in the other groups, and surgeon preference influenced the choice of prophylactic measure. Age and health-insurance status were comparable between groups (P > 0.05; Table 3).

Table 3.

Post hoc power analysis of factors potentially influencing the choice of prophylactic measure (post-matching cohort)

Variable	Group A (LARS) n = 80	Group B (TDT) n = 80	Group C (Ileostomy) n = 80	Group D (No prophylaxis) n = 80	P
BMI ≥ 30 kg/m², n (%)	7 (8.8)	8 (10.0)	17 (21.3)	8 (10.0)	0.008
Age ≥ 65 years, n (%)	46 (57.5)	44 (55.0)	47 (58.8)	44 (55.0)	0.321
Operating team A, n (%)	7 (8.8)	15 (18.8)	28 (35.0)	19 (23.8)	< 0.001
Operating team B, n (%)	33 (41.3)	18 (22.5)	15 (18.8)	14 (17.5)	< 0.001
Operating team C, n (%)	7 (8.8)	18 (22.5)	27 (33.8)	29 (36.3)	< 0.001
Operating team D, n (%)	33 (41.3)	29 (36.3)	10 (12.5)	18 (22.5)	< 0.001
No health insurance, n (%)	5 (6.3)	5 (6.3)	6 (7.5)	5 (6.3)	0.298

Surgeon preference, captured by the four operating teams, exerted a significant influence on the choice of prophylactic measure, whereas patient age and health-insurance status did not

Discussion

Diverting ileostomy is widely employed to mitigate the clinical consequences of AL and, in several randomised and observational studies, has been shown to reduce the incidence of symptomatic AL [12, 13]. This benefit, however, is offset by a substantial morbidity burden, including stoma prolapse, retraction, stenosis, peristomal skin disease, parastomal and incisional hernia, stoma re-creation and permanent stoma formation, all of which impair postoperative quality of life [14–16]. In this cohort, stoma-related complications occurred in 20.0% of Group C patients, a figure that must be weighed against the absolute reduction in AL afforded by ileostomy. Strategies that lower AL without the morbidity of diversion are therefore of considerable clinical interest.

The efficacy of TDT placement after LAR is similarly debated. Several reports have suggested a protective effect [8, 17–19], whereas Lee et al. found no measurable benefit [20]. With our protocol — a 28-Fr Malecot-type silicone tube with the tip 3–5 cm proximal to the staple line, removed on postoperative day 5–7 — TDT placement was not associated with a significant reduction in AL compared with no prophylactic measure. This observation is in line with the negative findings of Lee et al. [20] and with the heterogeneous results of recent meta-analyses, and may reflect differences in tube material, calibre, depth of insertion and duration of drainage. We therefore refrain from concluding that TDT is ‘ineffective’ in general; we report only that it conferred no measurable benefit in our cohort.

Our findings are concordant with the most recent systematic evidence. Tamura et al., pooling one randomised controlled trial and eight observational studies, reported that reinforcement sutures significantly reduced AL (OR 0.29, 95% CI 0.18–0.46) and reoperation (OR 0.32, 95% CI 0.14–0.72) without prolonging operative time or affecting mortality [21]. An umbrella review by Jayne D, et al. [22] and an earlier meta-analysis by Wei P et al. [23] reached similar conclusions while emphasising the heterogeneity of suturing techniques. The present study extends this body of evidence by directly comparing LARS with TDT, diverting ileostomy and no prophylactic measure within a single propensity-score-matched cohort.

In the present series, LARS achieved an AL rate comparable to that of diverting ileostomy and lower than that observed with TDT or no prophylactic measure. LARS permitted the safe completion of laparoscopic resections within 6.5 cm of the anal verge in men and within 5 cm in women. Given anatomical differences in the distance between the levator ani and the anal skin, we recommend that LARS be limited to anastomoses situated above the level of the levator ani. The anastomosis is also subject to extrinsic compression by surrounding musculature, and the suture line of LARS may help to counteract this compressive force.

Several mechanisms may account for the favourable performance of LARS. First, the reinforcement suture distributes circumferential tension across the anterior half of the anastomosis, thereby protecting the remaining circumference from sudden rises in intraluminal pressure. Second, LARS provides reliable reinforcement at the anterior triangle of the anastomosis, an area in which AL preferentially originates. Third, the use of a running barbed suture equalises tension along the suture line and minimises tissue cut-through. Consistent with these mechanisms, LARS was also associated with a numerically lower incidence of postoperative anastomotic bleeding and stenosis in our cohort. Importantly, LARS is technically straightforward and cost-effective. The post hoc power analysis confirmed a higher prevalence of diabetes in Group C compared with the other groups. Furthermore, it revealed that surgeon preference significantly influenced the choice of stapling and prophylactic technique; specifically, in the presence of known AL risk factors (e.g., obesity), surgeons at our institution tended to favour diverting ileostomy. Furthermore, it is important to note that in real-world clinical practice, these prophylactic strategies are not necessarily mutually exclusive. While our study compared them as four independent interventions based on institutional protocols, they can be viewed as potentially complementary measures. For example, LARS aims to augment the mechanical resistance of the anastomosis, whereas TDT or diverting ileostomy aims to reduce intraluminal pressure and colonic content volume. In selected high-risk cases, surgeons may combine these approaches to maximize protection.

A preliminary bench experiment was conducted to provide a qualitative mechanical illustration of LARS (Supplementary Figure S1; Supplementary Table S1). Two latex balloons were used as a simplified analogue of the rectum: the anterior half of one balloon’s ‘anastomosis’ was reinforced with a running suture, whereas the control model received no reinforcement. Both balloons were filled with water, and the circumferences at a constant height were measured. When the total water column was reduced from 18 to 15 cm to simulate a rise in intraluminal pressure, the control balloon’s circumference increased from 14.5 to 17.0 cm, whereas the reinforced balloon’s circumference increased only marginally from 11.4 to 11.5 cm. We acknowledge that this bench model is a simplified mechanical analogue and does not reproduce an actual stapled colorectal anastomosis. It is therefore retained in the Supplementary Material only, as a qualitative illustration of how a half-circumference reinforcement may redistribute radial tension, and should not be regarded as biological or surgical evidence of efficacy.

Several limitations of this study warrant comment. First, the design was retrospective, single-centre and not randomised. Because the choice of prophylactic measure reflected each surgical team’s routine practice, treatment-allocation bias cannot be excluded. Surgeons may have intuitively selected higher-risk patients (e.g., those with a higher BMI or unfavorable intraoperative findings not fully captured by the 16 matching covariates) for diverting ileostomy. Therefore, the anastomotic leakage rate in the ileostomy group might partially reflect the inherently higher baseline surgical risk of these patients rather than the true preventive efficacy of the stoma itself. Propensity-score matching mitigates but cannot eliminate such bias; a pragmatic multicentre randomised controlled trial comparing LARS with selective ileostomy is therefore warranted. Second, because operations were performed by several surgeons, a residual surgeon effect on AL cannot be excluded, and multicentre validation is required. Third, the low overall incidence of AL and the limited sample size precluded meaningful subgroup analyses of anastomotic bleeding and stricture. Fourth, patients who had received neoadjuvant radiotherapy were excluded a priori in order to avoid the confounding effect of radiation-induced tissue changes on anastomotic healing; this restriction limits the generalisability of our findings to that subgroup, in which AL rates are typically higher, and a dedicated study of LARS in the post-radiotherapy setting is planned. To our knowledge, this is the first study to compare LARS, TDT and diverting ileostomy within a single propensity-score-matched cohort.

Conclusions

In this propensity-score-matched retrospective cohort, LARS was associated with a lower incidence of clinically relevant anastomotic leakage than no prophylactic measure. Although its complication profile was comparable to that of diverting ileostomy without stoma-related morbidity, we emphasize that these interventions can be complementary rather than strictly alternative strategies in high-risk patients. TDT placement did not confer a measurable reduction in AL in our cohort. Given the retrospective design and the possibility of residual confounding from surgeon-driven allocation, these findings should be interpreted with caution and confirmed in prospective multicentre randomised trials.

Authors’ contributions

Binbin Zheng and Qiang Sun wrote the manuscript. Binbin Zheng, Zunqi Hu and Dejun Yang collected and analyzed the data. Weijun Wang and Qiang Sun revised the manuscript critically. All authors read and approved the final manuscript.

Funding

Not applicable.

Data availability

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the Second Affiliated Hospital of Naval Medical University and was conducted in accordance with the Declaration of Helsinki.

Consent for publication

Written informed consent for participation and for publication of anonymised clinical details and images was obtained from all individual participants included in the study.

Competing interests

The authors declare no competing interests.