Observational study with prospective recording of data from consecutive patients with rectal cancer at two hospitals who underwent low anterior resection (LAR) of the rectum with colorectal anastomosis using CMS measuring 28–29 mm or 31–33 mm in diameter (the choice of stapler used was based on the subjectivity and experience of the surgeon).

The present study was approved by both local institutional ethics committees (CEIC: 2024/5002 and 02-24-102-019), and complied with the criteria of the Declaration of Helsinki. The STROBE guidelines for observational studies were followed [13].

All patients were operated on by surgeons from the Coloproctology Units of the Universitari Parc Taulí Hospital in Sabadell (5) and the Consorci Sanitari in Terrassa (4) between June 2012 and June 2022. The selected patients followed the protocols of the international guidelines for rectal cancer [14, 15] and their cases were discussed by the multidisciplinary colorectal cancer committees of both hospitals. The follow-up of the patients was based on the NCCN rectal cancer guidelines, which includes endoscopic evaluation, allowing the diagnosis of stenosis that do not present clinical symptoms for the patient [14].

Inclusion criteria. Adult patients (over 18 years of age) who underwent scheduled surgery for non-disseminated middle or lower rectal cancer (lower margin less than 10 cm from the anal verge) by total mesorectal excision (TME) with or without protective ileostomy at the discretion of the surgeon; circular mechanical anastomosis; open or minimally invasive approach (laparoscopy or robot-assisted); follow-up of at least 1 year after rectal surgery.

Exclusion criteria. Pregnancy; patients with disseminated disease (M1); palliative surgery; emergency surgery; absence of colorectal anastomosis (Hartmann procedure or abdomino-perineal resection); manual anastomosis, or mechanical anastomosis other than CMS.

All patients underwent mechanical colon preparation. According to the protocol of each center, antibiotic prophylaxis was administered during anesthetic induction and thromboembolic prophylaxis.

Techniques performed: LAR: open, laparoscopic LAR was performed applying the classic standards described by the COLOR II study [16] and Heald [17]. TaTME: performed using transanal TEO TME, as described by Serra-Aracil [18].

Presence of BAS with a minimum follow-up of 1 year and data on the diameter of the CMS used.

BAS was divided into three groups according to the Truong classification [19]: grade I, anastomosis diameter between 10 and 20 mm and/or the presence of occasional abdominal symptoms; grade II, anastomosis diameter between 5 and 9 mm and/or in the presence of frequent abdominal symptoms; grade III, anastomosis diameter less than 5 mm and/or the presence of abdominal symptoms of intestinal occlusion.

Patient-dependent and demographic variables: sex, age, comorbidities [smoking, diabetes, vasculopathy, BMI (body mass index), ASA, preoperative albumin, chronic use of corticosteroids or immunosuppressants], surgical indication and neoadjuvant therapy.

Surgical variables: date of intervention, technique performed, approach, anastomosis diameter, stapler used, type and location of the anastomosis, surgical time and evidence of intraoperative anastomotic leak.

Immediate postoperative variables (30 days): postoperative complications defined according to the Clavien-Dindo classification [20], clinically relevant when greater than II (IIIa, IIIb, IVa, IVb and V) and the Comprehensive Complication Index (CCI) [21]. Postoperative anastomotic leak was classified according to the definition of the International Study Group of Rectal Cancer [22] in three groups: A, not requiring active treatment; B, requiring active non-surgical treatment; C requiring active surgical treatment.

Postoperative variables at medium term (from 30 days to at least 1 year): presence of ileostomy, ileostomy closure and date (if applicable), presence of BAS, BAS grade according to Truong classification [19] and date of last follow-up.

To establish the power of the sample, the sample size was calculated for a proportion in an infinite population. To do this, the formula n = (Z1−α)2*p*q/d2was applied, where 1 − α is the confidence level (0.95), Z1 − α is 1.96, p the approximate proportion of the phenomenon under study in the reference population, q = proportion of the reference population that does not present the phenomenon under study (1 − p). d = level of precision (0.05). For an estimated proportion of stenosis in the population of 15% [23] and a loss of 10% the sample size calculated is 210 patients [24].

SPSS version 26 (SPSS, Inc., Chicago, IL) was used. Due to the prospective nature of data collection, there were few missing values. Quantitative variables were described using mean and standard deviation or median and interquartile range (IQR) values when normality conditions were not met. Categorical variables were presented as absolute numbers and percentages.

The univariate statistical analysis of the quantitative variables, with independent groups, was performed using the Student t test whenever its application conditions were met; otherwise, the Mann–Whitney or Kruskal–Wallis U test was applied. For categorical variables, the Pearson χ2 test or Fisher’s exact test was used, depending on the conditions of application. A value of p < 0.05 and a confidence interval of 95% were considered statistically significant whenever the conditions were met.

A multivariate logistic regression analysis was performed with anastomotic stenosis as dependent variable and introducing the statistically significant variables or those with a trend towards significance with p < 0.2.