Introduction

In the year 2018; global cancer statistics reported colorectal cancer (CRC) as the third most common cancer worldwide [1, 2]. CRC is the 6th most common cancer in Egypt. Its crude incidence rate is estimated to be around 4 cases/100.000 population in 2011 [3].

CRC is the leading cause of large bowel obstruction. Approximately 15%–25% of CRC patients manifest with bowel obstruction, which mostly necessitate an emergency operation. Emergency operation is associated with a substantial risk of death and complications in this peculiar set of patients [4, 5, 6]. In Egypt, studies to identify risk factors predisposing to complications following resection and primary anastomosis in case of obstructed CRC are generally lacking. Consequently, this increasingly reported surgical problem requires further study and reporting.

Methods

The study was approved as retrospective, single center, case series study by ethical research committee of faculty of medicine, Alexandria university, Alexandria, Egypt. It was registered under the number 11/6/9/2/2020. Approval obtained before starting the study.

The study was also registered in researchregistry.com with the following Unique Identification Number(UIN):researchregistry7364, the study registration can be found here: https://www.researchregistry.com/browse-the-registry#home/.

Charts of patients presenting with obstructed CRC were retrospectively reviewed. Fifty-eight consecutive adult patients were admitted to our center with a diagnosis obstructed CRC during the period from January 2018 to December 2020. Our center is a tertiary referral university-based teaching hospital serving four governorates located at the north of Egypt.

These patients were diagnosed with contrast abdominal CT study as part of the workup for patients presenting with bowel obstruction. TNM staging was not mandatory at the emergency stage and accurate TNM staging was done prior to patient discharge from the hospital using triphasic abdominal CT and contract chest CT.

All the 58 patients were managed in emergency department intermediate care unit, they underwent NPO regimen, NG tube suction, volume replacement, correction of PH and electrolyte imbalances and blood transfusion if the Hb was <7 g/dl. Central line insertion was mandatory in these cases together with the insertion of urinary catheter.

All of patients were started empirically on Piperacillin/tazobactam (4g/0.5 g) combination every 8 hours and this was routinely upgraded to Imipenem/cilastatin at a dose of 1 gram/8 hours in cases of suspected or diagnosed bowel perforation. Metronidazole at a dose of 500 mg/8 hours was routinely added in all cases. Intravenous route was used for drug administration in all cases.

Pain control and proton pump inhibitors were additional part of preoperative optimization plan. Preoperative anticoagulation was started using Enoxaparin sodium at a prophylactic dose of 0.5 mg/kg/day given on two divided subcutaneous doses. The abovementioned medications were adjusted or substituted according to renal profile of the patients or in case of known drug allergy. Acute kidney injury was defined according to Kidney Disease Improving Global Outcomes (KDIGO) Clinical Practice Guideline for Acute Kidney Injury [7].

Out of Fifty-eight patients presenting with obstructed CRC, nine-patients refused to give a consent for emergency colorectal resections. These patients were managed by emergency colostomy/ileostomy and were excluded from the study. The remaining 49 patients underwent colorectal resections.

According to the location of obstructing tumor, colonic tumors at or distal to splenic flexure were considered as left sided. Patients managed by resection and primary anastomosis were grouped into obstructed right colon cancer (ORCC) and obstructed left colon cancer (OLCC).

Colorectal resections and primary anastomosis were performed by M.D category general surgeon with at least 2 years of surgical experience post M.D. His general surgical training involved the care of patients with bowel obstruction, decision making and extensive training in the performance of different types of bowel resections specially on urgent basis for at least 5 years prior to M.D degree. The operating surgeon had previously performed the intended type of colorectal resection more than 20 times during training period.

Following surgical intervention all patients were managed initially in the setting of intensive care unit on routine basis. Patients were then discharged to a step-down care unit. Postoperative complications within 30 days from operative procedures were reported using the Clavien-Dindo classification. All complications of Clavien-Dindo grade III b or higher were defined as major complications [8]. The results of this study have been reported in line with the PROCESS 2020 criteria [9]. Patients were observed on outpatient basis for at least three months following discharge.

Quality of the data used for the analysis were ensured by obtaining radiology and pathology reports signed by two M.D. consultants with specific experience in the field. Quality and standardization of medical management was ensured for surgical, anesthetic and ICU teams by double checking the privileges of treating physicians.

Continuous variables were described by the mean and standard deviation. Univariate analysis was done using T test in case of continuous variables, while Fisher’s exact test was used to compare categorical variables, p < 0.05 was considered statistically significant. Variables with P value < 0.1 on univariate analysis were included in a multivariate analysis using Cox regression model.

Results

During the period from January 2018 to December 2020, 253 adult patients underwent colorectal resections at Alexandria university hospitals, colorectal resections for benign disease and non-obstructed CRC were excluded from the study, Figure 1. Among 221 of CRC patients presented to our center during the study period, there were fifty-five right sided tumors, seventy- nine left sided tumors and eighty-seven rectal tumors representing (24.9%, 35.7% and 39.4% respectively).

Overall data for patients presenting with obstructed CRC (58 patients)

Obstructed CRC was present in 58 patients corresponding to 22.9% of those presenting with CRC. The rate of presentation with obstruction varied according to the location of the tumor, tumors involving the right colon, left colon, and rectum, presented with obstruction in 29.1%, 49.4%, 3.5% respectively, Figure 2.

Figure 2 

Categorization of CRC cases presented during the study period.

Cases with obstructed CRC were 23 males (38.8%) and 35 females (61.2%), their age ranged from 18–78 years with a mean of 54.4 ± 15.6 years. Eight patients were below 40 years (16.3%), 17 patients (34.7%) were between 40 and 60 years old, 14 patients were between 60 and 70 years old (28.6%) while those above 70 were 10 patients (20.4%). 35 patients were ≤60 years old and 23 were above 60.

The incidence of different comorbid conditions in the 58 patients presenting with obstructed CRC was as follows: DM (Diabetes Mellitus) were detected in 12 patients (20.7%), hypertension in 10 patients (17.2%), ischemic heart disease in in 6 patients (10.3%), BMI > 35 kg/m² in 5 patients (8.6%), serum Albumin <2.8 g/dl in 9 patients (15.5%), and Hb < 7 g/dl in 6 patients (10.3%). ASA score was II in 24 patients (41.4%), III in 22 patients (37.9%) and IV in 12 patients (20.7%). Two-patients (3.4%) presented with stage 1 acute kidney injury according to (KDIGO) Clinical Practice Guideline for Acute Kidney Injury [7]. Another two patients (3.4%) presented with stage 3 acute kidney injury in the context of multiorgan failure following anastomotic leakage.

Out of Fifty-eight patients presenting with obstructed CRC, tumors were right sided in 16 patients (27.6%), left sided in 39 patients (67.2%) and rectal in three patients only (5.2%). Taking in consideration the unavailability of self-expanding metallic stents (SEMS) at our center due to financial reasons, all the fifty-eight patients presented by obstructed CRC were managed surgically. Forty-two resections with primary anastomoses were performed in the acute setting (72.4%), covering stoma was needed in 16 patients of them (38.1%). Following initial resection, primary anastomosis deemed unsafe due to unstable vital signs and was not attempted in another seven patients (12%). The remaining nine-patients (3 cases of right sided tumors, 4 cases of left sided tumors and two cases of rectal tumors) refused to give a consent for emergency colorectal resections. These patients represented 15.5% of all patients presenting with obstructed CRC, they were managed by emergency colostomy/ileostomy and were excluded from the study.

Discussion

The incidence of CRC is still on the rise across the world. Egypt is no exception [3, 10]. Recently, the incidence of CRC in USA is reported to decline by about 2% to 3% annually starting from 2005 [11, 12]. Obstructed CRC is not infrequently reported worldwide. A range of 15% to 25% is frequently quoted with more than 75% of tumors located distal to splenic flexure [4, 13, 14, 15, 16].

The current study showed an incidence of obstructed CRC toward the high end of the reported range (22.3%) with 73.5% at or distal to splenic flexure. CRC presented to our center were left sided or rectal in more than 75% of cases. The rate of obstruction was the highest at left sided tumor (almost 50%) and lowest at the rectal tumors (3.5% only). This might be explained by earlier occurrence of tenesmus and bleeding in rectal cancers.

Furthermore, sigmoid colon and splenic flexure are reported to be the most common locations for obstruction by CRC [15, 16]. Sigmoid colon ranked the first as the most common site of obstruction by CRC in the current study, it was followed by descending colon, ascending colon and splenic flexure comprising 34.7%, 26.5%, 12.2% & 10.2% respectively.

In the current study, the mean age at diagnosis was 54.4 ± 15.6 years. This is more than a decade younger than the mean age in USA (69 in men and 73 in women) [2]. The younger age at diagnosis is also shown in other studies from Egypt. Moreover, these studies reported a relatively more advanced tumors at diagnosis [3, 13, 14].

Anastomotic leak is the most serious and dreadful complication for resection with primary anastomosis. This can be disastrous specially if diagnosis and management were delayed and sepsis ensued. Factors predisposing to anastomotic leak in elective setting were thoroughly reported. It was frequently sub-grouped to preoperative, tumor-related, and intraoperative risk factors. Stomas were reported to minimize consequences but not the incidence of anastomotic leak [17].

In the context of bowel obstruction, many authors indicated the importance of tumor location and stage of intestinal obstruction on the incidence of leak. The more distal the tumor and the more decompensated bowel obstruction; the more is the chance for anastomotic leak [18, 19]. Moreover, it was reported that development of peritonitis is an independent risk for anastomotic leakage in these settings [19].

The duration and degree of mechanical bowel obstruction caused by CRC is usually shown by different types of sequalae including varying degrees of cecal dilatation, closed loop obstruction and colonic perforation. This is especially true when the ileocecal valve is competent. This usually leads to closed loop obstruction which is believed to be the precursor for cecal perforation. Otherwise, the pressure will be manifested by small bowel dilatation.

Colonic perforation was found in (6.1%) of forty-nine patients managed by surgical resection. All of them were cecal perforations secondary to close loop obstruction, perforation is reported to occur in a range of 3%–8% of obstructed CRC. The most common site of perforation in cases of obstructed CRC is the cecum followed by tumor site [20, 21].

In the current study, incidence of anastomotic leakage was not significantly affected by tumor location (right vs left sided). Conversely, colonic perforation and the degree of bowel decompensation -manifested by cecal dilatation and the presence of closed loop obstruction- were significantly related to leakage rates. This indicates that duration of bowel obstruction and its sequalae were more influential in development of anastomotic leakage than simply the tumor location. Nevertheless, these results should be considered with caution due to small sample size.

Similarly, mortality rates in the current study did not differ significantly between ORCC and OLCC patients (7.7% and 3.4 respectively). The small number of ORCC patients in the current study might explain the higher leakage rate –7.7% of ORCC patients- than what is frequently stated in literature. Recently, few studies have shown a higher-than-expected leakage rate in ORCC, this was reported as high as 16.4% [22, 23]. Consequently; some authors are encouraging the use of protective stoma in high-risk patients [22, 23, 24, 25]. Selection criteria for what to be considered a high-risk group ORCC is yet to be identified [22].

One stage resection with primary anastomosis has been believed to be safe and feasible for ORCC even in the sickest patient. Anastomotic leak rate was reported to be between 2.8% and 4.6% following formal and extended right hemicolectomy, respectively [22].

In the current study, covering ileostomy was performed in only two patients presenting with ORCC, single stage procedure was performed in remaining 11 patients (84.6%). Of these 11 patients only one patient had anastomotic leak with consequent sepsis, multiorgan failure and death.

On the other hand, single stage resection with primary anastomosis was always deemed hazardous in case of OLCC. The patient is usually old age with frail general condition compounded by the presence of bowel obstruction with subsequent dehydration and malnourishment. The picture is further gloomed by the varying occurrence of sepsis. Consequently, operative time is considerably expanded, and the overall burden on of surgery is greatly increased on critically ill patient. Therefore, complication rate is expected to increase substantially [22].

Despite this common belief, nonrandomized retrospective data have reported an anastomotic leak rate between 2.2% and 12% in case of OLCC, a much comparable range to what was documented for left sided colonic resections in elective setting, namely 1.9% to 8% [26]. This led to the endorsement of primary resection and anastomosis for OLCC in properly selected subset of patients by many guidelines and societies [16, 22, 25, 26].

Similarly, the current study showed that single stage procedure was feasible in 15 patients of those presenting with OLCC (51.7%). Our practice to select patients for this approach was based on the intraoperative stability of the patient and the judgment from both surgical and anesthesia team. Otherwise, covering stoma was used in almost half of OLCC patients.

Despite the later need for stoma closure, creation of stoma following resection with primary anastomosis offered an option to decrease the incidence of anastomotic leak while avoiding the need for a major operative intervention as seen in reversal of Hartmann’s procedure or closure of temporary end colostomy

Acute kidney injury was identified as an important prognostic factor to the outcome of patients presenting with bowel obstruction in general and those with obstructed CRC in specific [24]. The limited number of patients complicated with acute kidney injury in the current study prevented its further analysis as prognostic factor.

On the other hand, DM and obesity were frequently reported to predispose to anastomotic leakage and mortality in patients with obstructed CRC [19]. Similar findings are observed in the current study, DM and BMI > 35 Kg/m² were significantly related to anastomotic leakage rate, reoperation rate, ICU readmission and the development of burst abdomen.

The main limitations for the current study are the retrospective design and the small sample size specially the number of cases with ORCC. Future research should be designed in multicenter prospective study to further assess the risk factors predisposing to complications in a larger number of patients with obstructed CRC.

Conclusions

Acute bowel obstruction is an infrequent presentation of CRC in Egypt, this calls for immediate implementation of screening program. Obstructed CRC is associated with a mortality rate of 4.8 % and predisposes to the development of complications which are frequently of major type.

Results of the current study indicated a more significant role of prolonged bowel obstruction sequalae (e.g., closed loop obstruction, and cecal dilatation >10 cm) in the development of anastomotic leakage compared to tumor location. DM and BMI > 35 Kg/m² significantly predisposed to anastomotic leakage rate, reoperation rate, ICU readmission and the development of burst abdomen.