Introduction

Colorectal carcinoma is one of the leading cancer diagnoses and cause for cancer-related death worldwide [1, 2]. Depending on tumor location and stage regarding the Union international contre le cancer (UICC) criteria, surgery is the crucial element of modern tumor therapy both as a single treatment and as the key element in multimodal treatment approaches [3]. Especially for rectal carcinoma, primary resection in early or limited stage (UICC stage I) or multimodal therapy with surgery after neoadjuvant radio-chemotherapy in locally advanced tumor stages (UICC stages II and III) can be considered as curative treatment options [4]. Depending on location of the carcinoma, a primary anastomosis by descendo-rectostomy should be applied to the patients. The protective diverting/defunctioning loop ileostomy is one routine surgical procedure, which can be considered during low anterior rectal resection, depending on intraoperative findings (e.g., height of the anastomosis ab ano, perfusion, and tension of the anastomosis), preoperative induction therapy, and at least patient’s characteristics (e.g., immunosuppression, vascular status, and other comorbidities). There is striking evidence that a defunctioning ileostomy is basically useful to prevent patients from septic complications in cases of anastomotic leakages [5, 6]. Nevertheless, the defunctioning loop ileostomy by itself can be responsible for postoperative complications like high fluid and electrolyte loss, parastomal hernia, prolapse, and skin irritations [7,8,9]. In the current literature, some weak evidence exists regarding a higher rate of prolonged postoperative gut paralysis and paralytic ileus in patients, who underwent large bowel and rectal resection with primary anastomosis with a defunctioning loop ileostomy versus patients without the defunctioning ileostomy [10, 11]. Thereby, postoperative gut dysmotility, paralysis, and ileus are quite a physiologic, temporary reaction due to the tissue trauma after abdominal surgery [12]. Typically, the entire gut recovers completely within 3–5 days—the small intestine returns to normal function at first within 24 h after surgery, and function of the stomach and colon recovers within 24–48 h and 48–72 h after surgery, respectively [12, 13]. However, a delayed postoperative return to normal gut function can frequently be responsible for reduced patient comfort, worse postoperative outcome, and longer hospital stay and thus affects patient’s recovery and causes higher public health costs [13,14,15,16,17].

We herein describe our institutional experience with prolonged postoperative paralytic ileus in a selected patient collective, who underwent oncologic open anterior rectal resection with complete mesorectal excision for rectal carcinoma in patients with or without a protective diverting ileostomy within a period of 43 months. Contributing factors for postoperative gut paralysis were excluded to the greatest extent due to formally focusing on isolated rectal resections as well as due to evaluating possible confounders in univariate analysis for a logistic regression model.

Material and methods

One hundred one patients were conducted to elective open anterior rectal resection with complete mesorectal excision and primary anastomosis by descendo-rectostomy with or without protective defunctioning loop ileostomy for rectal carcinoma within a period of 43 months. Patient data were retrospectively analyzed from the prospectively maintained institutional database. Each patient was treated by the institutional standard of care. The acquisition of data was approved by the local ethics committee (AZ 131/14).

Patients, who underwent primary intended and isolated open sphincter-preserving anterior rectal resection with complete mesorectal excision for rectal carcinoma with primary anastomosis by descendo-rectostomy, were included into this study. To obtain a homogeneous study population, exclusion criteria from this retrospective analysis were (1) severe cardiac insufficiency (≥ NYHA 3), (2) laparoscopic or converted-to-open rectal resection as well as Hartman procedure, (3) extended adhesiolysis and/or multi-visceral resections, (4) peritoneal carcinomatosis, and (5) postoperative major complications including anastomotic leakage, intraabdominal abscess, massive bleeding, and postoperative mechanical ileus.

Patients were divided into two groups: anterior rectal resection (1) with and (2) without a protective diverting loop ileostomy and the rate of postoperative paralytic ileus, determined by the intravenous administration of neostigmine during the first 10 days after surgery, was assessed as the primary endpoint of the study. The institutional standard of care was to stimulate gut motility stepwise, initially by conservative therapy (e.g., maintenance of fluid and electrolyte balance) and enteral laxatives on postoperative days 2 and 3. Neostigmine was given intravenously, when evidence for adequate gastrointestinal motility was lacking thereafter. The intraoperative decision for or against a diverting ileostomy was usually made by the surgeon during the operation as usual. Patients were basically treated using an “enhanced recovery after surgery” (ERAS) protocol postoperatively, including early mobilization, oral nutrition as soon as possible, and multimodal pain therapy.

Patient data (gender, age, body mass index (BMI), obesity, diabetes mellitus, peridural anesthesia, ileus-related re-intervention rate, and rate of neoadjuvant radio-chemotherapy) were proceeded to statistical analysis. Statistical analysis was performed using SAS V9.2 (SAS Institute, Cary, NC, USA). For univariate analysis, Fisher’s exact test was applied for two-group comparisons of categorical data in cross-tabulation as well as Wilcoxon rank-sum test for continuous variables. Spearman’s rho test was used for correlation analysis. Finally, a logistic regression model was used for estimating the odds ratio for prolonged postoperative paralytic ileus in patients after open oncologic rectal resection with protective defunctioning loop ileostomy.

Results

One hundred one patients (62 male, 39 female) were included into the study. The median (range) age and BMI of the study cohort were 63 (35–85) years and 25.65 (17.01–37.20) kg/m2, respectively. The overall incidence of postoperative paralytic ileus, defined retrospectively through the intravenous administration of neostigmine, was 13.86% (n = 14 patients). Figure 1 shows the remaining patient characteristics of the study cohort.

Fig. 1
figure 1

Characteristics of the patient cohort of this study. BMI, body mass index

Full size image

Sixty-two (61.39%) of the patients received protective defunctioning loop ileostomy during open rectal resection. No differences were observed in gender (male, 37 (59.68%) in the group with ileostomy and 25 (61.1%) in the group without ileostomy, p = 0.68), age (patients with ileostomy, 63.00 (39.00–85.00) years, patients without ileostomy, 64.00 (35.00–85.00) years, p = 0.83), and BMI (of patients with ileostomy, 25.09 (18.75–32.85) kg/m2, of patients without ileostomy, 26.42 (17.01–37.20) kg/m2, p = 0.4) between both groups. Significant differences were obtained regarding the rate of neoadjuvant radio-chemotherapy prior to surgery between patients with and without defunctioning loop ileostomy: more patients of group 1 (with ileostomy) received neoadjuvant radio-chemotherapy (19 (30.65%) patients versus 1 (2.56%) patient, p = 0.0005—see Table 1). Notably, male patients developed prolonged gut paralysis postoperatively more frequently by tendency (p = 0.07—see Table 2). Results of univariate analysis are shown in Table 2 and the supplemental material.

Table 1 Characteristics of patients of the two groups: (1) with and (2) without defunctioning loop ileostomy. BMI, body mass index
Full size table
Table 2 Results of univariate analysis. p values of continuous variables of both age and body mass index (BMI) were assessed by Spearman’s rho correlation
Full size table

Finally, a logistic regression model was used to estimate the odds ratio for the development of prolonged postoperative paralytic ileus in patients with a defunctioning loop ileostomy after oncologic open rectal resection. Due to an influence by tendency of the independent variable “gender” on the dependent variable “postoperative paralytic ileus” in univariate analysis (Table 2), the variable “gender” of the patients was used as a confounder for the regression model. For all other clinically relevant possible confounders, no influence was detected and no evidence for correlation among them was found. Logistic regression analysis showed that a defunctioning loop ileostomy during oncologic open rectal resection influences the development of prolonged postoperative paralytic ileus (p = 0.047). Odds ratio for prolonged postoperative paralytic ileus in patients after oncologic open rectal resection with a protective defunctioning loop ileostomy was 4.96 (95% confidence interval 1.02–24.03).

Discussion

A high number of patients suffer from rectal carcinoma worldwide [2] and undergo rectal surgery. Thus, it is of great importance to optimize surgical therapy and postoperative care to preserve those numerous patients from morbidity, worse postoperative outcome, and subsequently from mortality. Postoperative complications after major abdominal surgery, especially colorectal surgery, are not only relevant for the patients during the immediate postoperative hospital stay, but also delay the onset of adjuvant therapies and/or rehabilitation of the patients and thereby impair clinical and oncological outcomes as well as increase hospital and consecutively public health costs [13,14,15, 18]. The role of a defunctioning or diverting temporary loop ileostomy in reducing the rate of postoperative anastomotic failure after rectal resection with primary anastomosis is discussed diversely in the current literature [5,6,7]. Nevertheless, it has been well described that fecal diversion by a defunctioning loop ileostomy during anterior rectal resection with primary anastomosis per se is a safe surgical procedure and prevents patients from severe septic complications and re-operation in cases of anastomotic leakages [5,6,7]. Compared to colostomy, the reconstruction of the intestinal continuity in patients with a defunctioning loop ileostomy is less complicated, and the rate of surgical site infections, prolapse, and incisional hernias is reduced [7,8,9]. A whole bundle of studies suggests that a temporary defunctioning loop ileostomy is a considerable uncomplicated adjunctive surgical procedure following oncologic rectal resection [7]. Well-known and frequently observed complications of loop ileostomies are impairments in kidney function through high stomal output with consecutive high fluid and electrolyte loss [7]. Nevertheless, some studies raise the evidence for an increased risk of the development of prolonged postoperative paralytic ileus following bowel resection with ileostomy [10, 11].

The present study focused on patients, who underwent isolated open oncologic rectal resection with a primary rectal—not anal—anastomosis with or without a diverting ileostomy, showing a widely uneventful postoperative hospital stay: patients with anastomotic leakages as well as re-do procedures were primarily excluded from the data analysis. The primary endpoint of the study was the intravenous administration of the parasympathomimetic drug neostigmine as an objective measure for giving the hardest evidence for intestinal paralysis in the present retrospective study. Basically, temporary gut dysmotility after uncomplicated colorectal surgery due to the surgical trauma is frequently observed with an average duration to functional recovery of 3–5 days [10, 13, 14]. Although the duration, when a physiologic postoperative gut dysmotility becomes a pathologic prolonged postoperative paralytic ileus, is not clearly defined [19]. However, a prolonged dysmotility or postoperative paralytic ileus following bowel resection has to be discriminated from the former physiologic condition. Postoperative paralytic ileus causes nausea, vomiting, abdominal distension, and pain, delay in intestinal passage as well as in physiologic enteral nutrition, and thus impairs wound healing, immune response, surgical outcome, as well as patient comfort, and prolongs hospital stay [10, 14, 17,18,19,20]. These circumstances lead to a significant increase in ileus-related morbidity and subsequently mortality for the patients following open bowel resection [10, 16,17,18, 20]. The incidence for prolonged postoperative paralytic ileus after bowel resection given in the current literature ranges between 3 and 32% [17, 21]. Thereby, the progress from physiologic postoperative gut dysmotility to pathologic prolonged dysmotility or postoperative paralytic ileus is fluent and the definition of postoperative paralytic ileus after major colorectal surgery is significantly heterogeneous and weak in the current literature [21]: the absence of adequate bowel function on a certain postoperative day [10], the presence of nausea and vomiting [19], prolonged use of nasogastric tubes, and intolerability of oral nutrition [11, 19, 22]. These criteria for defining postoperative paralytic ileus after bowel surgery are subjective, unspecific [19], and not useful for retrospective patient data analysis. Therefore, in the present study, not these clinical and subjective data but the intravenous administration of the prokinetic drug neostigmine during the first 10 days after surgery was used to record the incidence of prolonged postoperative paralytic ileus retrospectively in this patient cohort.

The pathogenesis of prolonged postoperative gut dysmotility and paralytic ileus is multifactorial and complex [19]. In the past, risk factors for the development of postoperative paralytic ileus after bowel resection had been identified and concepts had been evaluated for preventing or even reducing the duration of (prolonged) postoperative gut dysmotility and postoperative paralytic ileus, e.g., minimally invasive laparoscopic surgery, multimodal analgesic concepts including spinal anesthesia, early postoperative mobilization, and early enteral nutrition of the patients [17, 21,22,23,24,25]. However, postoperative prolonged gut dysmotility and paralytic ileus following bowel resection is still a clinical problem and still affects patients even when applying the above-mentioned measures [10]. In the present study, the application of epidural anesthesia did not influence significantly the frequency of prolonged postoperative paralytic ileus (see Table 2). Previously known influencing factors of postoperative paralytic ileus after bowel resection were widely excluded due to the inclusion criteria of the present retrospective analysis, e.g., high preoperative ASA score (≥ III) [10, 18], cardiac disorders [10], minimally invasive surgery [17, 21, 22, 24, 25], or intraoperative conversion-to-open surgery [24, 25], postoperative anastomotic failure [17], or otherwise paralysis-nonrelated higher-grade complications [26]. Subsequently, gender seems to be a contributing factor (p = 0.07) for the development of postoperative paralytic ileus after open oncologic rectal resection in the present study in univariate statistical analysis (see Table 2). Male gender had been described as a risk factor for postoperative paralytic ileus in previous studies [11, 18]. Furthermore, the construction of a protective defunctioning loop ileostomy during oncologic rectal resection with primary anastomosis showed a tendency (p = 0.07) to increase the incidence of postoperative paralytic ileus in the present study in univariate analysis. This was continuously proven by a logistic regression model, which resulted in a significant influence of the development of prolonged postoperative paralytic ileus after oncologic rectal resection by defunctioning loop ileostomy (p = 0.047) and a high odds ratio (= 4.96) for prolonged postoperative paralytic ileus in patients with loop ileostomy after oncologic rectal resection.

The evidence for the influence of (prolonged) postoperative paralytic ileus by defunctioning loop ileostomy in patients, who underwent colorectal surgery, is weak since only a small number of mainly retrospective studies evaluate and focus on the issue. In the current literature, Millan et al. used intolerance to oral nutrition and gut dysmotility on postoperative day 6 for the definition of prolonged postoperative ileus and also observed ileostomy as a significant independent risk factor in univariate as well as multivariate analysis with an odds ratio of 1.97 in their retrospective review of 773 patients that underwent surgery for colorectal cancer [10]. With 20.1%, they observed the highest incidence of prolonged postoperative ileus in the group of patients that underwent low anterior rectal resection [10]. Interestingly, based on the results of their multivariate analysis, Millan et al. calculated that the incidence of prolonged postoperative ileus is increased to 38.3% in male patients with preoperative evidence of chronic obstructive lung disease and ileostomy as a part of surgery [10]. In a retrospective analysis of 2400 patients, also, Chapuis et al. observed an odds ratio of 1.4 for postoperative prolonged ileus in patients that underwent colorectal resections with the construction of any stoma [11]. Although prospectively conducted studies are still lacking, the presence of a defunctioning stoma seems to contribute to the multifactorial pathogenesis of prolonged gut paralysis after bowel resection.

An explanation for this observation is not clearly noticed in the current literature. A higher extent of surgical tissue trauma or inflammation or postoperative pain—all of them being well-known factors contributing to postoperative paralytic ileus [12]—due to the construction of a defunctioning ileostomy during large bowel surgery might not account for these findings as the construction of a defunctioning ileostomy is a relatively atraumatic optional part of the surgical procedure during oncologic rectal resection. Other surgery-related factors—potentially promoting intra- and/or postoperative complications like difficult surgery of the male pelvis as well as the size and infiltration of neighboring organs due to the rectal tumor, height of the anastomosis, and difficulty in splenic flexure mobilization for an adequate extent in colonic lengthening during the reconstructive part of the operation—are quite difficult to investigate retrospectively and, however, had to the best of our knowledge never been evaluated and discussed within the context of prolonged postoperative gut paralysis and paralytic ileus so far. Also, metabolic complications driven from the ileostomy itself like high fluid and electrolyte loss, hypokalemia, and acidosis [12] due to high output through the stoma might not be responsible for postoperative prolonged gut paralysis in those patients with defunctioning ileostomy. Although more patients in the ileostomy group received neoadjuvant radio-chemotherapy (p = 0.0005), preoperative induction therapy had no impact on the rate of postoperative paralytic ileus in univariate analysis of this study cohort (p = 1). Also, the current literature gives no evidence that preoperative pelvic irradiation increases the rate of postoperative paralytic ileus [25, 27]. However, inspired by clinical experiences, compared with left-sided procedures, a slow return to normal bowel function and higher incidence of postoperative paralytic ileus is frequently observed in patients undergoing right hemicolectomy [17, 28,29,30,31,32]; thus, prolonged paralytic ileus in patients with defunctioning loop ileostomy after oncologic rectal resection might be contributed by ileocolic manipulation.

In conclusion, we herein report on the first study evaluating the development of prolonged postoperative paralytic ileus as the primary outcome measure in patients after primary open oncologic anterior rectal resection with the simultaneous construction of a defunctioning loop ileostomy. The group of patients in this study is highly selected, due to presenting a widely uneventful postoperative hospital stay, to exclusively focus on the influence of the defunctioning ileostomy on postoperative prolonged paralytic ileus. For defining postoperative paralytic ileus, we retrospectively evaluated the intravenous application of neostigmine as an objective primary outcome measure within the first 10 days following surgery. A standardized and homogeneous definition as well as consensus for the management of postoperative prolonged paralytic ileus is currently not established in the literature. Endpoints of former studies focusing on postoperative prolonged ileus after bowel resection are inconsistent and almost subjective. Prolonged gut dysmotility after bowel surgery is a frequent postoperative problem and results in increased morbidity and prolonged hospital stay for the patients as well as higher costs for public health. In the present retrospective analysis, defunctioning ileostomy was proven as a contributor for and to dramatically increase the incidence of prolonged postoperative paralytic ileus. The influence of loop ileostomies on prolonged postoperative paralytic ileus after bowel surgery has to be proven in larger retrospective patient cohorts and further prospective trials. Especially, the hypothesis that ileocolic manipulation during the formation of defunctioning ileostomies influences prolonged postoperative gut dysmotility as well as whether the construction of any colostomy would be more beneficial within this context might be further investigated as this could majorly impact on clinical practice in the future. But, nevertheless, surgeons have to accurately consider the advantages—regarding the prevention of septic complications in patients with anastomotic failure—and disadvantages—regarding high fluid and electrolyte loss as well as prolonged postoperative paralytic ileus—of a defunctioning loop ileostomy in patients undergoing oncologic rectal resection.