Abstract
Background/Aim: Early postoperative small bowel obstruction (EPSBO) prolongs hospital stays after surgery. This study aimed to evaluate the risk factors for EPSBO associated with colorectal cancer resection. Patients and Methods: We retrospectively compared the clinical variables of patients with EPSBO (n=37) and those without (n=812) after primary tumor resection for colorectal cancer at our hospital between January 2010 and December 2015. Results: In multivariate analysis, significant differences between the two groups was found in male sex, open surgery, and defunctioning ileostomy (DI) placement (p=0.024, p<0.0001, and p=0.023, respectively), but not for colostomy placement. Of 16 patients with DI who developed EPSBO, 13 (81.3%) cases resulted from obstruction of the stomal outlet. Conclusion: Male sex, open surgery, and DI placement are risk factors for EPSBO after colorectal cancer resection. For patients with placement of DI, obstruction of the stomal outlet should be carefully considered.
Bowel obstruction is a common postoperative morbidity after colectomy performed for colorectal cancer, with an incidence rate of 1.4-13.7% (1-4). Various anti-adhesion barriers have been developed and used to reduce the risk of adhesion formation in the region of the surgical wound (5-8). In this regard, laparoscopic surgery, which requires a smaller incision than for an open approach, may lower the risk of postoperative bowel obstruction. However, this remains an issue of controversy, with some studies having reported a lower rate of bowel obstruction with laparoscopic colectomy (3,9), a finding which was not supported by other studies (1, 4).
In this study, we specifically focused on early postoperative (EP) small bowel obstruction (SBO), which is an important clinical issue as it inhibits early oral intake and early ambulation, thus delaying postoperative recovery and prolonging the length of hospitalization. Moreover, EPSBO can prevent the initiation of postoperative adjuvant chemotherapy, which is recommended at 4-8 weeks post-surgery for patients with colorectal cancer (10, 11). EPSBO also increases the risk of future adhesive SBO (12), which would require subsequent surgery and increase the risk of postoperative morbidity and mortality (3). Therefore, reducing the risk factors for EPSBO is an important surgical goal for patients undergoing colectomy for colorectal cancer, which formed the aim of our study, namely to clarify the risk factors of EPSBO associated with colorectal cancer resection.
Patients and Methods
We conducted a retrospective analysis of 866 patients who underwent colectomy for the treatment of primary colorectal cancer at the Jikei University Hospital between January 2010 and December 2015. Prior to surgery, all patients underwent colonoscopy, chest computed tomography (CT), and abdominal CT or abdominal magnetic resonance imaging (MRI) for tumor staging using the Union for International Cancer Control (UICC)-TNM criteria. From this initial cohort, 17 patients were excluded due to presence of multiple colorectal carcinomas, requiring resection at more than one primary site or subtotal colectomy (Figure 1).
The 849 patients included in our analysis comprised of 536 men and 313 women, with a mean age of 68 years (range=26-95 years). Postoperatively, patients who exhibited symptoms of SBO within 30 days of the procedure underwent abdominal radiographs or CT for assessment. Of these patients, those requiring discontinuation of oral take or insertion of a gastric or intestinal tube for decompression formed the EPSBO group for analysis. Patients who developed postoperative ileus were excluded.
The following variables were compared between the EPSBO group (n=37) and the non-EPSBO group (n=812) to identify the risk factors for EPSBO: age, sex, body mass index (BMI), American Society of Anesthesiologists score (ASA), tumor site (the colon or rectum); stage (UICC-TNM classification), concomitant surgery performed, surgical method (laparoscopic or open surgery), requirement for colostomy or defunctioning loop ileostomy (DI), type of anesthesia (general or epidural), elective or emergency surgery, operative time, intra-operative volume of blood loss, and presence/absence of surgical site infection (SSI).
This study was approved by an Institutional Review Board (27-283 8168).
Statistical analysis. Data are expressed as a median (25-75th percentile) or number (%), as appropriate for the data type. Univariate analyses were performed using the Mann–Whitney U and chi-squared tests, as appropriate for the data distribution. A logistic multivariate regression analysis was performed, using stepwise backward elimination, to identify independent risk factors for EPSBO. All analyses were performed using SPSS (version 22.0; IBM, Tokyo, Japan), and p-values less than 0.05 were considered significant.
Results
Patient demographics and operative variables are summarized in Table I. Among the 849 patients forming the current study cohort, EPSBO developed in 37 patients (incidence rate of 4.4%). On univariate analyses, rectal cancer and male sex were found to be significantly more frequent in the EPSBO than the non-EPSBO group (rectal cancer: 5.6% vs. 2.2%, respectively, p=0.021; male: 7.6% vs. 2.3%, respectively, p=0.001). The operative time was significantly longer in the EPSBO than non-EPSBO group (274 versus 220 min, respectively, p=0.016). Although the incidence rate of colostomy was equivalent between the two groups, the need for placement of a DI was significantly greater in the EPSBO than non-EPSBO group (16.2% versus 2.8%, respectively, p<0.0001).
On multivariate analysis, the following factors were found to be predictive of EPBSO (Table II): male sex [odds ratio (OR)=2.665, 95% confidence interval (CI)=1.1-6.3, p=0.024]; DI (OR=9.113, 95% CI=4.2-19.8, p<0.0001); and open surgical approach (OR=2.497, 95% CI=1.1-5.5, p=0.023).
We performed a specific comparison of the demographic and surgical characteristics between the EPSBO and non-EPSBO groups only for patients who required DI (Table III). The incidence rate of emergency surgery was significantly higher for the EPSBO than non-EPSBO group (univariate, p=0.022). On multivariate analysis, emergency surgery was associated with EPSB, but not to a significant degree (p=0.054). Based on the abdominal radiographs and CTs, performed on patients who required a DI in the EPSBO group (16 patients), rotation of the ileostomy was suspected as the cause of the EPSBO in two patients. Among these 16 patients, 13 (81.3%) presented with an obstruction of the stomal outlet (SOO; Figure 2). When comparing patients with and without a SOO in the EPSBO group only, the proportion of patients who underwent colectomy with concomitant surgery and those who developed a SSI was higher among patients with than without SOO (concomitant surgery: 33.3% versus 0%, respectively, p=0.032; SSI: 66.7% versus 0%, respectively, p=0.002). However, recovery from EPSBO was shorter among patients with than without SOO (5 versus 9 days, respectively, p=0.002). Although 66.7% of patients without SOO required insertion of a nasogastric or long intestinal tube for treatment, 76.9% of patients with SOO underwent treatment with insertion of a tube through the stoma for drainage, with clinical improvement. Re-operation was not required for treatment in any patient with SOO (Table IV).
Patient characteristics and clinical variables.
Comparison between patients with and without early postoperative small bowel obstruction (EPSBO) by univariate and multivariate analyses.
Comparison among patients with defunctioning ileostomy between those with and without early postoperative small bowel obstruction (EPSBO) by univariate and multivariate analyses.
Discussion
Among the current study cohort, we identified male sex as an independent risk factor for EPSBO, although the reason for this is not clear. It might be that the smaller pelvic space and higher amount of visceral fat in men could be predisposing factors for EPSBO. We also identified that a laparoscopic approach lowered the risk for EPSBO as compared to open surgery, which is probably due to the smaller surgical wound area and reduced exposure to ambient air.
Placement of a DI was a significant risk factor for bowel obstruction. Previous studies reported an incidence rate of EPSBO of 4.2% among patients who underwent colon cancer surgery (12). In the current study, placement of DI was a risk factor for EPSBO, whereas colostomy was not. Compared to a colostomy, a DI includes a loop stoma of the small intestine only and does not require a retroperitoneal route for placement. Moreover, compared to a colostomy, the stool formed in patients with an ileostomy contains a significant amount of fluid. This muddy-like stool might be difficult to extrude, against gravity, by peristalsis alone. Moreover, loop stomas carry the risk for rotation of the proximal and distal sides of the intestine. In particular, the small intestine may become easily rotated as it has no specific attachment to the surrounding tissues because of not being placed in a retroperitoneal route. Ihnát et al. reported an incidence rate of semi-rotation of a DI around its longitudinal axis of 3.8% (13).
Comparison among patients with defunctioning ileostomy who developed early postoperative small bowel obstruction (EPSBO) between those with and without stomal outlet obstruction (SOO) by univariate analysis.
Use of a non-retroperitoneal route increases the risk for an internal hernia. From CT imaging of the 16 patients with a DI who developed EPSBO in the current study cohort, signs of rotation of the ileostomy was identified in only two patients, with no evidence of rotation or internal herniation identified in the other 14 patients. However, in 13 patients, including the two with suspected rotation of the ileostomy, a dilation of the small intestine was observed immediately before the rectus muscle, with accumulation of a large quantity of liquid stool (Figure 2). Therefore, we consider that difficulty of passing stool through the stoma, through the rectus abdominis muscle, may have a cause rotation of the ileostomy.
We previously reported an increased incidence of SOO in patients with DI (14). In the current study, the majority of patients with SOO were treated with a stomal tube, with symptoms being improved after drainage of intestinal fluid, without need for re-operation. If the principal cause of SOO is a tight rectus abdominis or insufficient fasciotomy, then the SOO would not improve over an average of 5 days without reoperation. As such, we consider that SOO results from temporary swelling and edema of the ileostomy or of the surrounding tissues after surgery.
Among patients with SOO, the incidence rate of SSI or concomitant operation was lower than those without SOO. Therefore, SOO is not likely to be associated with intra-abdominal inflammation or complexity of the surgical procedures. Ng et al. reported a higher rate of complications for DIs placed by laparoscopic surgery, with a rate of obstructive complications of about 5%. They also reported that rotation of the ileostomy was found in 50% and adhesive kinking proximal to the ileostomy was found in 37.5% of the patients with obstructive complications (15). In the current study, ileostomy rotation was suspected in two patients, both of whom improved within 4 days. Therefore, we consider it unlikely that ileostomy rotation improved spontaneously within 4 days, in patients who developed EPSBO, even with decompression of the intestine with insertion of an intestinal drainage tube. We consider that kinking and rotation developed as a consequence of SOO. Considering that 81.3% of the cases of EPSBO were identified in patients with DI who developed SOO, measures to lower the risk of SOO would be very important.
The selection process for the study.
Avoiding placement of DI would be effective in preventing EPSBO. However, in patients at high risk for anastomotic leakage, placement of DI is necessary. In these cases, inserting a tube into the stoma just after surgery may be effective in lowering the risk of SOO.
Conclusion
Factors associated with the development of EPSBO after colorectal cancer resection include male sex, open surgery and placement of DI. For patients with placement of DI, precautions to prevent SOO should be carefully considered.
Computed tomographic imaging of the two cases showing the outlet obstruction of defunctioning ileostomy. In both cases, dilation on the oral side intestines of the defunctioning ileostomy immediately before the rectus muscle was observed.
Footnotes
Conflicts of Interest
The Authors have no conflicts of interest to declare.
- Received December 12, 2017.
- Revision received January 18, 2018.
- Accepted January 24, 2018.
- Copyright© 2018, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved