The Incidence and Risk of Early Postoperative Small Bowel Obstruction After Laparoscopic Resection for Colorectal Cancer

Introduction

Early postoperative small bowel obstruction (EPSBO) generally occurs within 30 days of abdominal surgery and is most commonly caused by adhesions. ¹ As patients who undergo surgery for colorectal cancer (CRC) typically experience some degree of postoperative ileus, this can be difficult to differentiate from EPSBO, presenting both diagnostic and therapeutic challenges. EPSBO is a major contributor to morbidity and mortality after colectomy for patients with CRC ² and is associated with considerable healthcare costs. Furthermore, in patients with advanced CRC, prolonged recovery may delay adjuvant chemotherapy, consequently affecting the long-term prognosis of the patients. Strouch et al. ³ recently reported that postoperative complications occurred more often in patients undergoing open surgery than in those undergoing laparoscopic treatment and demonstrated that patients who underwent laparoscopic surgery received postoperative chemotherapy an average of 25 days earlier than patients who underwent open surgery.

Although several studies have shown that laparoscopic surgery for CRC offers several advantages over open surgery in terms of short-term outcomes,^4–10 the incidence of EPSBO has not been shown to be significantly different.^11,12 Lack of a uniform definition and the study of a relatively small number of patients to sufficiently address this rare complication have impaired the understanding of the circumstances associated with EPSBO. In this study, we assessed the incidence and identified risk factors for EPSBO in an attempt to provide clinicians with clinically relevant information concerning EPSBO following laparoscopic surgery for CRC.

Subjects and Methods

In total, 1787 consecutive CRC patients who underwent laparoscopic colorectal surgery (CRS) with curative intent from June 2008 to May 2012 were enrolled in this study. Patients who underwent stoma/bypass surgery, those with recurrent CRC, and those who underwent laparoscopic colectomy under emergent circumstances were excluded. Institutional review board approval was obtained, and all patients gave informed written consent.

EPSBO was clinically defined as nausea, vomiting, and abdominal distension developing within 30 days of abdominal surgery ¹³ and was confirmed by abdominal computed tomography and/or a contrast study. Because postoperative small bowel ileus occurs immediately following surgery and is mainly dependent on the return of colonic motility (usually 48–72 hours), two criteria to exclude ileus were added: (1) patients with signs and symptoms of obstruction occurring any time within 30 days after operation and lasting more than 7 days or (2) patients with signs and symptoms of bowel obstruction occurring from 7 to 30 days after operation and lasting any number of days. ¹⁴ All patients were managed according to identical practice standards. Operative treatment was offered to patients who failed to improve, despite conservative treatment or clinically aggravated. Patient variables, including patient characteristics, operative time, estimated blood loss, need for transfusion, conversion to open surgery, postoperative complications, length of hospital stay, previous operative history, combined surgical procedures, neoadjuvant chemoradiotherapy, and histopathologic tumor/node/metastasis (TNM) staging were recorded in a prospectively maintained database.

All patients with CRC received CRS with curative intent according to surgical oncologic principles. ¹⁵ Neither intracorporeal nor extracorporeal mesenteric defect closure was performed. Rectal cancer was defined as the lower margin of the tumor being located with 15 cm of the anal verge and was categorized as occurring in the middle (6–10 cm) or upper (11–15 cm) rectum. For patients with mid- to lower rectal cancer, total mesorectal excision was performed. Patients with locally advanced disease received neoadjuvant chemotherapy before surgery. ¹⁶ In patients with left-sided colon and rectal cancer, splenic flexure mobilization is performed after high ligation of the inferior mesenteric vessels. Since 2011, we have preferred an extended medial-to-lateral approach. ¹⁷ For clinical T4 CRC with suspected invasion into adjacent structures, when possible, laparoscopic inspection was attempted, and en bloc resection was performed. If laparoscopic resection was impossible because of invasion, the procedure was converted to open surgery. ¹⁸ A diverting stoma was created at the surgeon's discretion. Anti-adhesives were not routinely used.

Postoperative complications were classified as recommended by Dindo et al. ¹⁹ Conversion to open surgery was defined as an incision made sooner than initially planned to delivery a specimen. The completeness of resection was based on operative and pathologic reports. Resections were classified as follows: R0, negative gross and pathologic margins; R1, negative gross margins with positive pathologic margins; or R2, positive gross margins.

Statistical analysis was performed using SPSS for Windows software (SPSS version 20.0; SPSS, Inc., Chicago, IL). Differences between groups were determined using Student's t test and the chi-squared test to compare continuous and categorical variables as appropriate. Logistic regression analysis identified factors that were significantly associated with EPSBO. Variables with a univariate P value of <.10 were included in the multivariate analysis. A P value of <.05 was considered statistically significant.

Results

In total, 1787 patients (105 with and 1682 without EPSBO) with CRC requiring laparoscopic CRS were evaluated in this study. Patient characteristics and differences between groups are summarized in Table 1. Patients with EPSBO were more often male (80.9% versus 63.8%; P<.001). The frequency of EPSBO was higher in patients with rectal cancer than in patients with colon cancer (P<.001) and in patients who received preoperative chemoradiation treatment than in those who did not (P<.001). No difference between groups was found with regard to age, body mass index, American Society of Anesthesiologists score, previous abdominal surgery history, maximal tumor size, and pathological tumor category or stage.

The types of previous abdominal surgeries (348 procedures among 334 patients) are summarized in Table 2. Appendectomy was the most frequent previous operation (n=145). Procedures that were performed concurrently with CRS are summarized in Table 3. Hepatic resection for metastatic lesions was the most frequent concurrent procedure (n=14). In addition, 72.6% of concurrent procedures were performed with intent for oncologic curative resection. Ten patients (0.56% among the total patient population, 9.5% among patients who experienced EPSBO) who did not respond to conservative treatment for more than 14 days required surgical intervention (Table 4). Table 5 summarizes the operative parameters between patients with and without EPSBO. The highest rate of EPSBO was observed in patients undergoing intersphincteric resection (17.4%), and EPSBO was more frequent in patients who underwent combined operations (17.8% versus 5.4%; P<.001) and diverting stoma (17.3% versus 3.5%; P<.001). Patients who underwent open conversion had a significantly higher EPSBO rate (5.2% versus 25.4%; P<.001). In terms of surgical difficulty, EPSBO developed more frequently in patients who underwent longer operations (P<.001), and therefore received a blood transfusion (P<.001), which might be caused by profound blood loss.

In this study, six parameters were determined as independent risk factors for postoperative EPSBO in the multivariate analysis (Table 6). Regarding the patients' characteristics, male sex (adjusted odds ratio [AOR]=2.27), combined operation (AOR=2.23), and diverting stoma (AOR=4.79) were associated with a higher EPSBO rate. For factors related to surgical difficulty, open conversion (AOR=2.85), blood transfusion (AOR=3.51), and an operation time longer than 180 minutes (AOR=1.91) were independent factors associated with an increased EPSBO rate. When analyzed according to the tumor location, cases of combined operation were excluded for colon cancer. During rectal cancer surgery, neither combined operation nor operation time longer than 180 minutes showed significant predictive values (Table 6).

Discussion

The occurrence rate of EPSBO in this study was 5.9%, which is comparable with previous studies that have investigated EPSBO after CRS (range, 2%–10.3%).^5,20–23 Surgical intervention was performed in 0.56% of the total patient population. Patients' characteristics (sex and diverting stoma) and perioperative variables (conversion to open surgery, combined operation, blood transfusion, and long operation time) related to a more advanced tumor stage and increased technical difficulty were found to be independent risk factors for EPSBO after laparoscopic CRS.

EPSBO is predominantly the result of adhesions, which are inevitable outcomes of abdominal and pelvic surgery. More than 90% of patients who undergo laparotomy are estimated to develop adhesions to some extent^24,25; therefore, laparoscopic surgery may reduce adhesion formation following abdominal surgery by minimizing surgical trauma. Some studies suggest that postoperative inflammation occurs less frequently in patients undergoing laparoscopic surgery than in those undergoing open surgery.^26,27 Concurrent with these studies, Dowson et al. ²⁸ demonstrated that adhesion formation may be reduced following laparoscopic colectomy. Although multicenter randomized controlled trials^4–10 have found that laparoscopic CRS has benefits compared with open surgery, including reduced blood loss, earlier recovery of bowel function, less pain, and shorter hospital stay, EPSBO rates were not different between the two surgery groups, in both colon ¹¹ and rectal ¹² cancer studies. As indicated by Wilson et al. ²⁹ for the Surgical and Clinical Adhesions Research Group, larger numbers of patients will be needed to confirm this issue in a prospective study, implying that such studies may be unrealistic. Therefore, we believe that the findings of our study could help estimate the rate of and the risk factors for EPSBO following laparoscopic CRS.

Few studies have reported risk factors for EPSBO after CRC surgery using a multivariate analysis.^21,22 Shin et al. ²² concluded that patients with American Society of Anesthesiologists grade ≥III and presence of local remnant tumor were independent predictive factors for determining the risk of EPSBO. However, Nakajima et al. ²¹ found no significant association between tumor-related factors (pT, pN, or pM) and the incidence of EPSBO. Instead of tumor depth and size, Nakajima et al. ²¹ found that open colectomy and rectal cancer had greater associations with EPSBO than did other variables and suggested that the degree of surgical invasiveness resulting from the expanding of resection for advanced disease is lower than from open colectomy and rectal surgery. In the present study, the completeness of resection, pT category, and maximal tumor size showed no significant association with the incidence of EPSBO. However, these results may be biased because of the small number of patients with R1 and R2 resection or the limited indications for laparoscopic CRS among patients with advanced disease.

The role of anatomical location in the development of EPSBO has been examined in numerous studies of CRS and malignant disease. Recently, Masoomi et al. ³⁰ reported that transverse colectomy (AOR=1.11) and left hemicolectomy (AOR=1.06) were associated with a higher risk of EPSBO when compared with right hemicolectomy. It is interesting that proctectomy (AOR=0.59) was not a risk factor for EPSBO. With regard to rectal cancer surgery, total mesorectal excision has been widely accepted as a standard surgery, and we almost always performed high ligation of the inferior mesenteric artery, which is related with more extended surgical dissection than benign surgery. In this regard, the conclusion of two studies of patients with CRC was different.^21,22 Shin et al. ²² divided their patients into two groups: a pelvic surgery group, who underwent colorectal anastomosis situated beneath the sacral promontory, and a colonic surgery group. No significant difference in the incidence of EPSBO according to anatomic location was found between these two groups. In our study, although EPSBO was significantly associated with rectal cancer surgery in the univariate analysis, we found no statistically significant difference in the multivariate analysis. This result suggests that various factors (long operation time, blood transfusion, combined operation, and open conversion) related to the inflammatory response induced by excessive surgical manipulation may play a critical role in EPSBO, rather than anatomical location. In this present study, diverting ileostomy was performed at the surgeons' discretion based on a difficult dissection, a very low-lying tumor, bleeding, radiation history, or incomplete donuts after anastomosis. As a result, we propose that the relationship between increased incidence of EPSBO in patients with construction of a diverting ileostomy is not only a reflection of a more challenging surgical procedure but that complications of the ileostomy itself also contributed to the development of EPSBO (Table 4). Similarly, Poon et al. ²³ determined that diverting ileostomy was associated with an increased incidence of early obstruction in patients with rectal cancer.

Carrying out mesenteric defect closure and adhesiolysis from previous operations with laparoscopic instruments is technically challenging and may cause mesenteric vascular or adjacent organ injury. The relationship between leaving the mesenteric defect open and increasing the incidence of internal hernia and subsequent EPSBO is controversial. Regarding this issue, Cabot et al. ³¹ reported the long-term outcomes of not performing mesenteric defect closure after laparoscopic right hemicolectomy and showed that, of 530 patients, only 4 (0.8%) had complications attributed to a mesenteric defect. These authors concluded that routine mesenteric defect closure was not supported after laparoscopic right hemicolectomy. As mentioned above, we did not perform mesenteric defect closure. We also determined that our policy not to perform mesenteric defect closure is not associated with a higher incidence of EPSBO, which was confirmed when no case needed relaparotomy for the treatment of internal hernia, although there was no mesenteric closure control group. In addition, if adhesions from previous surgery did not interfere with inserting trocars or accessing the surgical field, we did not perform routine laparoscopic adhesiolysis. Yamamoto et al. ³² recently showed that previous abdominal incisions were a significant risk factor for conversion to open surgery. In their study, ileus occurred more frequently in patients with previous abdominal surgery than in those without a history of abdominal surgery (3.8% versus 2.1%; P=.04). However, these authors did not define the ileus and only used univariate analysis. We suggest that the specific type of incision or abdominal surgery history, which could induce extended dissection, is associated with an increased occurrence of EPSBO, rather than a history of previous surgery itself.

Limitations of this study include its retrospective nature and patient selection. Selection of the best operative method based on the surgeon's discretion and skill level is also a limiting factor. Furthermore, some variables (combined operation, previous surgery) were considered as existing or not, and more clinically relevant information and meaningful analysis will require a grading scale of these variables in future studies.

Despite the limitation of this study, we conclude that factors for predicting the development of EPSBO in patients with CRC are variables reflective of a more difficult surgery, rather than pathologic disease severity or anatomical location. In addition, most patients with EPSBO improved with conservative treatment, and surgical treatment was rarely needed. However, we could not determine factors associated with surgical treatment from patients with EPSBO due to the relatively small sample size. Future studies are needed to fully address this issue.