Introduction

Today, the temporary protection of ileoanal and coloanal/rectal anastomoses is often achieved by the creation of a loop ileostomy [1]. Although defunctioning of the colon does not provide true prophylaxis against anastomotic leakage, it is meant to mitigate potential complications after anastomosis formation. However, the possible morbidity/mortality associated with stoma formation and takedown is still a concern [2]. Common complications of stoma reversal range from wound infection, small bowel obstruction (SBO) or ileus to life-threatening leakage and peritonitis. Therefore, it is important to achieve the lowest possible rates of complications after stoma closure. The side-to-side stapler, or more precisely, the functional end-to-end anastomosis for ileostomy reversal, was introduced into practice in the 1980s [3] and has since gained popularity among surgeons. The advantages of the wider anastomosis achievable by stapler include decreased rates of ileus and SBO, as well as minimization of the so-called “human factor”, i.e., the influence of a particular surgeon’s skill and experience on the quality of the small bowel anastomosis [4, 5]. However, some controversy exists concerning the advantages of manually suturing anastomoses, e.g., cost effectiveness, reduction of operative time, postoperative hospital stay, and other parameters [6, 7].

The results presented in this article are from a prospective, randomized trial designed to compare these two methods of stoma closure.

Materials and methods

The protocol of this prospective, randomized controlled trial was approved in January 2005 by the local ethics committee at the State Research Center of Coloproctology in Moscow, Russia. Operating time was chosen as the primary endpoint. Based on our previous experience and data from a previous meta-analysis [8], we expected application of the stapler technique to reduce operating time by at least 10–15 min. The secondary endpoint was the rate of complications, in particular SBO. Proponents of the stapled anastomosis [4, 7] reported a 2–7-fold decrease in the rate of SBO. Thus, we estimated that a total of 50 patients in each group would provide sufficient statistical power. Treatment assignment sequence was computer generated and randomization occurred in the operating room prior to incision. Patients and clinicians filling the study forms were blinded to treatment’s allocation.

Between 2005 and 2008, 248 patients underwent two-stage surgery for rectal carcinoma with creation of a protective ileostomy or colostomy. Informed consent was given by 121 eligible patients with a defunctioning ileostomy, of whom 119 were then randomized. Patients were randomized into two groups: 56 patients had conventional ileostomy takedown, and 63 had a functional end-to-end anastomosis created by linear stapler. The only stratification criterion was obesity, i.e., body mass index (BMI) ≥ 30. Patient and surgical characteristics are listed in Table 1.

Table 1 Patient characteristics and description of surgical procedure
Full size table

At the time of the first operation, a loop ileostomy was formed in the right lower quadrant of the abdomen. The segment of ileum used for stoma creation was taken at 15–20 cm from the ileocecal junction. A 180° rotation of the ileum was performed to orient the efferent limb below the afferent limb, and a ~3-cm nipple was formed on the functioning limb to facilitate maturation of the ileostomy.

The median time to ileostomy closure did not differ significantly between groups. This time depended primarily on the need for a consolidating chemotherapy (usually capecitabine/oxaliplatin or capecitabine alone). If adjuvant treatment was necessary, stoma closure was performed between the 2nd and 3rd cycles or after definitive completion of chemotherapy.

All patients, including those who received adjuvant chemotherapy, were in suitable condition for surgery, without significant changes in blood work or signs of cancer progression. The integrity of the preexisting colonic anastomoses and pouches was documented before stoma closure using contrast enema and radiographic evaluation. No particular preparation of the ileum was undertaken before stoma closure; however, on the day before surgery, one or two water enemas were used to check for and clean out any impacted barium remaining in the defunctioned colon after contrast enema. Therapeutic antibiotics (cephalosporin + metronidazole or fluoroquinolone + metronidazole) were given at the time of operation and for 2–3 days thereafter.

Most of the operations were performed under a combination of spinal anesthesia and intravenous sedation (Table 1). All closures were performed by consultant surgeons or by senior registrars under direct supervision of the consultant.

Postoperative monitoring was continued for 30 days. Simple charts were given to all patients, on which they could mark time of first postoperative flatus, stool, and sensation of peristalsis if they were able to feel it. Bowel sounds were also checked by the registrar or consultant according to routine practice (three times per day, during morning (7–8 a.m.), afternoon (4–5 p.m.) and evening (8–9 p.m.) ward rounds).

All data were entered into a computerized database. Statistical analysis was performed by SPSS 10.0 for Windows™. The median or mean, depending on normality of distribution, and the standard deviation were chosen for descriptive statistics. Comparison between groups was performed by unpaired t-test. A two-tailed Fisher’s exact test was used to compare dichotomous variables. A P-value less than 0.05 was considered to be significant.

In both groups, surgery began with local circumstomal access. The ileostomy-bearing segment of ileum was sharply dissected out through all layers of the abdominal wall. If inadvertent enterotomy or myotomy occurred, the gap was closed with interrupted seromuscular sutures of 3-0 Vicryl® (Ethicon, Int.).

A stapled, functional, end-to-end anastomosis was created using a linear stapler GIA™-80 (Auto Suture™, USA). After mobilization of the ileal segment which involved the stoma, an enterotomy was created at the base of the nipple valve allowing insertion of both branches of the stapler into the afferent and efferent limbs (Fig. 1a). Alternatively, resection of the ileostomy-bearing bowel segment was performed (Fig. 1b). The anastomotic line was checked for bleeding, and the reloaded stapler or any other available linear stapler, e.g., the TLH-60 (Johnson & Johnson, Inc.) was fired transversely to resect the ileostomy-bearing segment and close the anastomosis (Fig. 1c). In all cases, the distal staple line was oversewn and imbricated with a running suture using the Lembert technique, which added approximately 5 min to anastomosis creation. An additional Lembert stitch was used to reinforce the crotch site.

Fig. 1
figure 1

Creation of functional end-to-end stapler anastomosis for ileostomy closure. a Both limbs of the ileum are mobilized, and the stoma-bearing segment is resected; b side-to-side stapler anastomosis; c the ileum is cross-stapled, and the anastomosis is complete

Full size image

A hand-sewn end-to-end anastomosis was created using two-layered interrupted sutures of 3-0 Vicryl® (Ethicon, Int.). Excision of the ileostomy-bearing segment was performed in all cases.

In both groups, the postoperative wound was closed primarily in layers. An indwelling subcutaneous drain was used according to surgeon preference, usually in obese patients.

Results

No formal laparotomy was required in any case, and all ileostomies were closed via local, circumstomal access. A significant difference was observed in overall operating time (Table 2), which was shorter when a stapling technique was employed for closure (68 ± 7 vs. 92 ± 11 min (P = 0.01)).

Table 2 Results: stapled versus hand-sewn ileostomy closure
Full size table

A trend toward earlier fluid and solid food intake after functional end-to-end anastomosis was also identified; however, this did not reach statistical significance.

There was no perioperative mortality or rehospitalization in either group.

The overall morbidity rate was higher after suture anastomosis than after stapler anastomosis (14.2 vs. 2.4% (P = 0.04)). The main complication observed after suture anastomosis was prolonged ileus and small bowel obstruction. Nevertheless, no additional surgical intervention was necessary, as nasogastric decompression and conservative treatment were successful in all patients. The only case of small bowel obstruction (1.6%) in the stapler group was caused by adhesions and required midline laparotomy; adhesiolysis was followed by an uneventful postoperative period. Another anastomosis-related complication after stapler closure was bleeding from the staple line, which manifested first in bloody stool and next in a decrease in hemoglobin level from 15.0 to 13.5 g/dl. Nevertheless, bleeding was self-limited, and no reoperation was needed. No significant difference in incidence of wound infection was observed after use of stapler closure.

Discussion

Ever since protective ileostomy was introduced as an option for defunctioning the colon, numerous trials comparing closure, complications, cost effectiveness, and other parameters have been published [2, 4, 8–11]. The morbidity rate and severity of complications related to ileostomy formation vary from series to series, but it is widely agreed that the consequences of anastomotic failure are more devastating than those following ileostomy formation, existence, and closure [5]. Systematic reviews [12] have demonstrated overall morbidity rates of ileostomy takedown of 17.3%, with a mortality rate of 0.4% and conversion to formal laparotomy at the time of closure in 3.7% of patients. The most serious problems associated with stoma closure were SBO and leakage, because they can lead to reoperation, reestablishment of an ileostomy, mortality, and emotional and psychological trauma to the patient [11–13]. A better performance by stapled anastomoses with regard to SBO was identified in a prospective randomized trial by Hasegawa et al. [4], but could not be confirmed in a randomized trial from Hull et al. [7], or in other retrospective and prospective series [9–11]. A meta-analysis published by Leung et al. [8] also found no statistically significant differences between stapled and hand-sewn loop ileostomy closures, although it did report a trend favoring stapled closures with regard to lower small bowel obstruction rates and shorter operative time; this study also stressed the necessity of a prospective randomized trial. Our trial used the operating time required for temporary loop ileostomy closure as the primary endpoint and demonstrated shorter times in patients with a stapled anastomosis. In addition, significantly decreased rates of postoperative ileus were achieved when a stapled anastomosis was used. The single case of adhesion-related SBO that occurred in the stapled anastomosis group suggests that there is a greater reduction in intestinal obstruction rates in a stapled versus hand-sewn anastomosis because stapled anastomosis creation usually does not induce adhesion formation. On the other hand, the narrow edematous lumen of the hand-sewn end-to-end anastomosis is the leading cause of SBO [6]. Given these circumstances, the quickly accomplished, wide-mouthed end-to-end stapled anastomosis is an attractive alternative whose superiority has been demonstrated in our randomized trial.

Although cost was not an endpoint for our trial, we calculated possible savings and expenses in stapler versus hand-sewn closures. With regard to suture materials, the stapler anastomosis is far more expensive than the hand-sewn, with an estimated cost of 202€ (with a second cartridge for reloading) vs. 6€, respectively. On the other hand, according to Russian Ministry of Public Health estimates, the cost per diem of reconstructive surgery is approximately 158€. If we assume an average hospital stay in the stapler group is 2 days less than in the hand-sewn group, we can conclude that introduction of stapled ileostomy closure will lead to savings of roughly 100€ per patient.

Conclusion

Use of the stapling technique for ileostomy closure results in decreased operating time. A lower rate of morbidity (primarily due to ileus and SBO) is the main advantage of the procedure.