Malignant Large Bowel Obstruction

Abstract

Large bowel obstruction is a serious and potentially life-threatening surgical emergency which is associated with high morbidity and mortality rate. The most common etiology is colorectal cancer which accounts for over 60% of all large bowel obstructions. Proper assessment, thoughtful decision-making and prompt treatment is necessary to decrease the high morbidity and mortality which is associated with this entity. Knowledge of the key elements regarding the presentation of a patient with a large bowel obstruction will help the surgeon in formulating an appropriate treatment plan for the patient. Comprehensive knowledge and understanding of the various treatment options available is necessary when caring for these patients. This chapter will review the presentation of patients with malignant large bowel obstruction, discuss the various diagnostic modalities available, as well as discuss treatment options and the various clinical scenarios in which they are most appropriately utilized.

Large bowel obstruction (LBO) is a serious and potentially life-threatening surgical emergency which is associated with high morbidity and mortality rate, ranging from 23 to 77.6% and 11.9 to 27%, respectively. ^{1 2 3} There are various etiologies of LBOs. The most common is colorectal cancer which accounts for over 60% of all LBOs. Other common etiologies include volvulus of the cecum or sigmoid and diverticular disease which account for 10 to 15% and 5 to 10% of all LBOs, respectively. This study will focus on LBOs caused by malignancy. Regardless of the etiology of the obstruction, proper assessment, thoughtful decision-making, and prompt treatment is necessary to decrease the high morbidity and mortality which is associated with this entity. Each decision is critical, as each decision made will influence the patient's ultimate outcome. It is necessary to think of not only the current situation, but how the decisions made at this time, may affect future treatment options. The surgeon caring for these patients has two challenges to resolve. First and foremost is to relieve the obstruction and then second, if possible, to treat the etiology of the obstruction. To do this, it is necessary to have a comprehensive knowledge of the various treatment options available and when each is best utilized in caring for these patients.

Presentation

Patient presentation may vary greatly in severity depending on the acuity and completeness of the obstruction. Patients who present with an acute complete obstruction typically will be distended and have more abdominal pain. They typically have not passed flatus or had a bowel movement in several days. If the obstruction is partial or more chronic in nature, the patient will present with distention and less abdominal pain. They may give a history of constipation leading up to the time of presentation which spans weeks to months. In these more chronic cases, the bowel may distend slowly over time as the obstruction progresses, thus giving time for the bowel to accommodate. As the obstruction progresses, patients may show signs of dehydration. Nausea and vomiting are usually seen late in the course of the disease. Fever and peritonitis may indicate progression of the obstruction to the point of causing ischemia and possible perforation. Colonic obstruction can have other physiologic effects as well. Colonic distention can cause impaired ventilatory function due to reduced diaphragmatic movement. There is increased risk of infection due to increased microbial proliferation and potential translocation of bacteria leading to septic shock. In addition, renal function can be affected by dehydration secondary to third spacing or vomiting.

Pathophysiology

Understanding the pathophysiology of a LBO is critical to being able to intervene in a timely manner and perform the appropriate intervention to prevent irreversible complications. Obstruction of the large bowel can occur anywhere in the colon however is much more common on the left side. This is because the left colon has a narrower lumen and the stool is more formed. Two factors determine the urgency needed to treat a LBO. The first is the degree of obstruction (partial vs. complete) and the second is competency of the ileocecal valve. Either an incompetent ileocecal valve, which occurs in approximately 25 to 40% of patients, or a partial obstruction will allow for some decompression of the colon and perhaps delay the progression to ischemia and perforation. ^{2 4} Without the ability for decompression, a “closed loop” will occur. The colon will continue to distend as the small bowel continues to pass air and stool into its lumen and the competent ileocecal valve prevents reflux. This increase in intraluminal pressure will diminish total intestinal blood flow. Boley et al studied the effect of distention on blood flow in intestines using a dog model. They showed, as intraluminal pressures increased, total blood flow decreased. They also noted that as blood flow decreased, there was an associated decrease in oxygen extraction. Injection studies done showed filling of only the submucosal and serosal vessels when intraluminal pressures were at or above 60 to 90 mm Hg. Interestingly, they also noted that the external appearance of the bowel remained normal even when the total blood flow was only 20% of control. The exact mechanism as to why this happens it not known. ⁵

The mechanical effect of this distention will affect different aspects of the bowel differently based on the Law of Laplace. The tensile force on the wall of the colon is equal to the intraluminal pressure multiplied by the diameter multiplied by pi. It is not surprising then that the tensile force on the cecum is greater than in other parts of the large bowel as the cecum has the greatest diameter. The sequela of a LBO is therefore determined by the degree of distention, the duration of distention, the amount of tension on the colon wall based on the Law of Laplace, and the underlying vasculature of the patient. As surgeons are making treatment decisions, they must remember that perforation at a distance from the obstruction indicates ischemic changes in the entire segment of distended colon even if the bowel has a normal external appearance, thus making this compromised bowel not suitable for anastomosis. ⁶

Initial Evaluation

A complete LBO, if not treated promptly can lead to bowel ischemia and perforation. A partial LBO may give the evaluating team additional time to do a more thorough evaluation, possibly determine the cause of the obstruction, and consider other treatment options including endoscopic stenting as definitive treatment or as a bridge to surgery. The patient should be questioned regarding the onset of pain, the severity of pain, and any changes in the pain since its presentation. Sudden improvement in pain followed by sudden increase in pain may signify a perforation. Other constitutional symptoms should be acquired as they may also shed light on the etiology of the obstruction. Ongoing fatigue and weight loss may suggest a malignancy. Past episodes of left lower quadrant pain and fever may suggest diverticular disease. Past medical history of inflammatory bowel disease, prior radiation, prior episodes of diverticulitis, narcotic usage as well as any prior surgeries are all important pieces of information which should be obtained from the patient.

On physical examination, patients may be tachycardic secondary to either hypovolemia or sepsis. Hypovolemia may be due to fluid losses into the distended bowel as well as from emesis. Abdominal pain is typically present. Pain out of proportion to the patient's physical exam is a classic sign suggesting ischemic bowel. If bowel ischemia and/or perforation has occurred, the patient may have signs of sepsis including fever, tachycardia, and hypotension. The degree of abdominal distension as well as any signs of peritonitis should be evaluated. Digital rectal exam should always be performed as a rectal mass or fecal impaction may be found. Initial blood work should include a CBC as well as a complete metabolic panel to assess for anemia, leukocytosis as well as any electrolyte abnormalities that may be present. If ischemic bowel is a concern, a lactic acid should also be obtained.

Imaging

Plain Radiographs

Plain radiographs are inexpensive, easy to obtain, and can quickly identify pneumoperitoneum. It has a sensitivity of 84% for detecting LBO with a specificity of 72%. ⁷ Plain radiographs can give some helpful pieces of information such as the diameter of the bowel which is used as an indicator of possible impending perforation. Normal caliber of the bowel ranges from 3 to 8 cm with the larger diameter seen in the cecum. Risk of ischemia and perforation increases as the cecum continues to dilate to a diameter of 12 cm or greater. ^{8 9 10 11} One must be cautious though, as the location of the LBO can mimic a small bowel obstruction. A left-sided LBO would cause proximal large bowel dilatation, however, a right-sided LBO with an incompetent ileocecal valve could mimic a small bowel obstruction with a gasless colon and dilated small bowel. Identification of the cecum or ascending colon with fecal material without gas, known as the “dilated cecal sign” suggests a right-sided LBO. ¹²

Contrasted Enema

A contrasted enema may be indicated in the work up for LBO for (1) accurate determination of the location of obstruction, (2) differentiation between pseudo-obstruction from a mechanical LBO, and (3) evaluation of atypical cases in which a LBO is on the differential. Contrasted enema can also be therapeutic in the management of sigmoid volvulus and intussusception. Ericksen et al studied 335 consecutive patients admitted with diagnosis of acute small or LBO. Twenty-three patients (6.3%) were diagnosed with a LBO. Four patients were operated on immediately based on their clinical presentation with the remainder having a barium enema performed. Sixteen of the 19 patients required surgery. One patient had resolution of their intussusception and two patients had resolution of their sigmoid volvulus and did not undergo surgery. In this study, barium enema was 100% accurate in localizing the site of colonic obstruction as well as predicting the need for surgery. It should be noted that in this study two patients were diagnosed with perforation after their barium enema. Both patients died from complications of barium peritonitis. They concluded that barium should not be used in cases of suspected perforation or in those cases where the cecum diameter was 10 cm or greater. ¹³ Water soluble-contrast material should be used for these studies as it is easily absorbed in the peritoneum should there be a perforation. Limitations of contrasted enemas include the need for rectal contrast and insufflation which may increase risk of perforation and the need for patients to be able to move freely to assist with the study.

Computed Tomography

Computed tomography (CT) is the most widely used diagnostic tool and is the imaging modality of choice to evaluate LBO. This imaging modality cannot only diagnose LBO but can evaluate complications related to the obstruction as well as shed light on the etiology of the obstruction. CT has the advantage of being fast, readily available, well tolerated, and without the need for rectal contrast or rectal insufflation. Frager et al performed a prospective study to determine the success of CT in the evaluation of suspected colonic obstruction. CT has a sensitivity and specificity of 96 and 93%, respectively. CT correctly localized the point of obstruction 94% of the time. ¹⁴ CT can be used to diagnose intraluminal, intrinsic, and extrinsic causes of LBO as well as can give additional information regarding other abnormalities which might be present, such as local or distant metastatic disease. It can identify inflammation of the bowel wall which may suggest inflammatory bowel disease or diverticular disease. The addition of intravenous contrast can help in identifying the presence of masses, inflammation as well as bowel wall ischemia. ⁴

False negatives can occur when the obstruction is partial with no proximal dilation or when there is air or fecal material remaining distal to the obstruction. False positives of LBO can occur due to spasm of the bowel wall which can mimic a fixed obstruction or proximal dilation of the small bowel and right colon and decompression of the left colon which though appearing to be a LBO in reality is either ileus or pseudo-obstruction. ¹⁵

CT diagnosis of a LBO is based on the finding of proximal bowel dilation leading to a transition point with decompressed distal bowel. A colon diameter greater than 9 cm in the cecum and 6 cm in the left colon is considered dilated. ¹⁶ Maximum dilation of the cecum which corresponds to increased risk of perforation has been reported at between 9 and 12 cm. ^{9 17} The continence of the ileocecal valve correlates with cecal distention and is therefore a risk factor for the progression of a LBO to perforation and as such, should be a part of the final reading of the CT scan. ¹⁸

Intrinsic causes of LBO include cancer and diverticulitis; with inflammatory bowel disease being a rare cause. Colon cancer is the most common cause of LBO accounting for >60% of cases, with nearly 20% of colon cancer cases being complicated by some degree of obstruction. ¹⁴ CT findings of a colonic malignancy may include an asymmetric, short segment of wall thickening with shouldering or an enhancing soft tissue intraluminal mass that narrows the colon lumen. Additional finding that would favor the diagnosis of colon cancer may include irregularity of the contours, eccentric lumen, ulcerated mucosa, and the presence of pericolonic lymph nodes. CT has the added benefit of being able to assess tumor extension, involvement of adjacent structures, lymphadenopathy, as well as presence of intraabdominal metastasis. ^{12 16}

Management

Initial Management

Management of malignant LBOs can be challenging and the risk of complications and poor outcomes daunting. Thoughtful decision making, both preoperatively and intraoperatively can significantly change the long-term outcome for the patient. Some treatment options will depend on the acuity of the situation, condition of the patient, location of the obstruction, etiology of the obstruction, and presence of complications due to the obstruction.

LBO is not the same as a small bowel obstruction and should not be viewed in the same light. It is imperative to make the distinction between a complete LBO and a partial bowel obstruction as well as the time frame of the obstruction. Gradual dilation of the bowel over weeks to months allows the bowel wall to accommodate and is less likely to cause ischemia and perforation compared with rapid dilation of the bowel. Thus, a cecum measuring 12 cm in the former case will be less likely to perforate compared with a cecum that has very quickly dilated to 12 cm over a very short amount of time. One must quickly ascertain if the ileocecal valve is competent or not, as this will determine if a “closed loop” is present. A closed loop obstruction, created by a competent ileocecal valve with a complete distal obstruction, requires urgent treatment. Delays can allow ongoing distention, further vascular compromise to the more proximal bowel leading to ischemia with progression to perforation and patient instability, thus potentially converting a stable straightforward bowel resection into an unstable damage control scenario requiring total colectomy.

Initial management includes fluid resuscitation and correction of abnormal electrolytes. Placement of a nasogastric tube may be considered if the patient is exhibiting nausea and vomiting. This will only be helpful if the small bowel is also dilated signifying an incompetent ileocecal valve. Nasogastric decompression will not assist in depression of the colon.

The patient's presentation will dictate the need for immediate surgical intervention or if further testing and nonsurgical treatment options can be considered. If on initial assessment, the patient has peritonitis with evidence of sepsis, and/or imaging showing evidence of perforation, imminent impending perforation, or ischemic bowel then immediate surgical intervention is needed after initial resuscitation is complete. Informed consent should be obtained with discussion of possible surgical treatments including possible bowel resection (partial vs. total colectomy), possible need for stoma (temporary or permanent), and possible need for further surgery (open abdomen, second look, future ostomy reversal). Broad spectrum antibiotics are administered. Deep vein thrombosis prophylaxis with anticoagulation and sequential compression devices should be applied if there is no contraindication to their use. Preoperative stoma marking is ideal if possible. It is also helpful, if possible, to identify abdominal wall creases with the patient in a more upright position, as these may not be apparent in the supine position. The patient should be positioned in the modified lithotomy position to allow for access to the rectum should this be necessary. A midline laparotomy incision is typically used to allow for adequate exposure of all quadrants of the abdomen. Performing a flexible sigmoidoscopy, preferably with CO ₂ insufflation in the operating room, can also be a helpful adjunct to help discern the etiology.

Emergent Surgery—Overview of Treatment Considerations

The surgeon needs to not only consider what is before them, but also how the decision of the day will impact the future care of the patient. Exploration is undertaken to identify the point of obstruction, presence of ischemic bowel, location of perforation if present, and determine the etiology of the obstruction. If signs of perforation are encountered upon entering the abdomen, identification of the site of perforation and subsequent control of this will limit continued contamination throughout the remainder of the procedure.

If the lesion is deemed resectable, proper oncologic guidelines should be followed which include adequate margins, high ligation of vessels to assure appropriate lymphadenectomy, as well as en bloc resection. The condition of the more proximal bowel will dictate if the resection can be a segmental resection or if a subtotal colectomy is indicated. If there is presence of ischemia, perforation, or concern for bowel wall integrity of the proximal bowel, a subtotal colectomy may be indicated. If the patient is unstable or there is presence of carcinomatosis, resection should be deferred, and proximal diversion performed. In patients with large obstructing rectal cancers consideration of diversion at the index operation should be made so that adjuvant therapy may be given prior to definitive resection.

If resection is performed, careful consideration must be given to the prospect of restoring continuity at the index operation. Factors that will affect this decision include the stability of the patient, degree of contamination, the risk of anastomotic dehiscence based on location, (right-sided vs. left-sided anastomosis) and other factors that will affect healing including significant weight loss prior to surgery, steroid use, immunosuppression, and malnutrition. Options at this point include primary anastomosis alone, primary anastomosis with protective diverting loop ileostomy, or leaving the patient in discontinuity with an end stoma and Hartman's pouch. Primary anastomosis should only be performed in low-risk patients who are stable. This has typically been reserved for right-sided lesions. However, current literature supports this to be possible for left-sided lesions under appropriate conditions. Consideration of the use of on-table lavage or manual decompression to clear the proximal colon prior to anastomosis may increase the safety of this option. If a primary anastomosis is performed and there is increased concern for potential leak, a diverting loop ileostomy may be considered to protect the anastomosis. These various treatment options as well as their pros and cons will be considered in detail in the later portion of this chapter.

If the lesion is found to be unresectable at the index operation, careful consideration needs to be given to what proximal diversion is best for the patient. The goal of the diversion should be to allow decompression of the colon without adversely affecting the possibility of resection and anastomosis in the future should that be possible. A diverting loop ileostomy, in the setting of a competent ileocecal valve, would not relieve the obstruction and has no role in this situation. Ideally the diversion should be as close to the point of obstruction as can easily be created so as to preserve as much conduit length for possible later anastomosis. The diversion itself can be either a loop colostomy or an end colostomy with a mucous fistula. It is important that the bowel distal to the stoma and proximal to the point of obstruction is also decompressed. End colostomy with a small mucous fistula at the same stoma site allows for decompression of both the proximal and distal bowel with only one stoma site which is smaller than those created for a loop colostomy. If the diversion cannot be done close to the obstruction, a right-sided colostomy should be considered as this also saves the conduit for later anastomosis. In addition, if a proximal diversion was needed at the time of resection and anastomosis, this stoma could be left in place and taken down at a third surgery.

As a last resort, a loop transverse colostomy can be created. Though this may seem like an easy option at the time of the index surgery, it has many disadvantages and should only be created if no other option is available. Loop transverse colostomies are typically large and difficulty to pouch, have a high complication rate (38%), and are prone to prolapse (17%). ^{19 20} They potentially jeopardize the conduit which would be used should resection and anastomosis be attempted in the future.

Nonemergent Surgery—Overview of Treatment Considerations

In the stable patient, without peritonitis, and who has a partial obstruction, or an incompetent ileocecal valve additional time can be taken to determine the etiology and thus consider other nonsurgical treatment options. Typically, right-sided lesions, in stable patients without perforation and feculent peritonitis, will be managed with resection and primary anastomosis. Nonemergent obstructions distal to the splenic flexure are more complex and have more options to consider. Colonoscopy may be considered to further localize the obstruction, confirm the diagnosis, and obtain biopsies. Typically, this is not needed for right-sided lesions as these are usually treated with surgical resection regardless of their malignant potential. Preoperative knowledge of the etiology of left-sided lesions may significantly impact treatment options and should be considered. CO ₂ insufflation should be used if possible. Endoscopic procedures by their very nature will introduce air into the colon and can precipitate further cecal dilation which may lead to ischemia and perforation. Close monitoring of the patient is needed. Pre- and post- procedure plain radiographs can be helpful in following these patients and can identify sudden increases in cecal diameter post procedure. In the setting of malignancy, endoscopic stent placement for either palliation or as a bridge to surgery, can be considered. Some of the same questions that were raised in the emergent setting are again encountered, including resectability, safety of creating an anastomosis, need and type of proximal diversion, and feasibility of on-table lavage.

Obstructing Colorectal Cancers—Treatment Options

The location of the cancer is important as right-sided and left-sided cancers are treated somewhat differently and have been shown by some to have a different prognosis so will be considered separately.

Proximal Obstructing Colon Cancer

The preferred surgical treatment for obstructing colon cancers proximal to the splenic flexure is resection and primary anastomosis at the index operation. Studies confirm this with evidence of this procedure being done in 75 to 98.8% of cases when the cancer is in this location. ^{3 21 22 23 24 25} This allows for resection of the obstructing cancer as well as the dilated proximal colon, leaving normal healthy distal ileum to be anastomosed to the distal colon. ³ The anastomotic leak rate quoted in the literature for elective resection for right-sided colon cancer ranges from 2.8 to 8.4%. ^{26 27 28} Anastomotic leaks are particularly troublesome in this group of patients as it will delay the start of any systemic chemotherapy which can significantly alter the patient's prognosis.

Biondo et al studied patients who had been treated for obstructing colon cancers and found the mortality rate for those with a proximal obstructing lesion which suffered a leak was 38% compared with 20% for those with a leak associated with a distal obstructing lesion. The authors were surprised by this finding and suggested that the morbidity of proximal obstructing lesions may be underestimated. ²⁹ Frago et al echoed the same sentiment when they found in their study that their anastomotic leak rate was 16.4% in the proximal obstructing cancer group compared with 7.7% in the distal obstructing cancer group. ²² One should keep in mind that the leak rates reported are primarily retrospective studies and will vary based on patient selection.

Left-Sided Obstructing Colon Cancer

The majority of obstructing colon cancers present distal to the splenic flexure (54–80%). ^{2 22 23 29 30} A range of options are available in managing these cancers.

Three-Stage Procedures

Treatment of left-sided obstructing colon cancers is much more controversial and has evolved over time. Originally, a three-staged approach was thought to be the safest approach for patients. The first stage manages the obstruction by diversion without resection of the tumor. At the second stage the cancer is resected, and the bowel left in discontinuity. During the third stage, the colostomy or ileostomy is taken down and an anastomosis is performed to re-establish continuity of the bowel. This approach is known to have high morbidity and mortality not only with the index surgery but recognizing the cumulative risk with multiple surgeries to accomplish the goals of treatment. Surgeons began to look for other approaches to this problem including consideration of resection at the index case. Studies done comparing patients who presented with obstructing left colon cancers who were treated with resection at the index case versus those who had a colostomy with plan for staged resection found the mortality rate ranged from 11 to 21.1% and 25 to 27.7%, respectively. They noted that 34 to 55.5% of patients who were in the stages resection group never underwent resection. ^{31 32} Survival of greater than 1 year for each group was 67 and 32%, respectively. ³² Based on these and similar studies, it was generally agreed upon that a more aggressive approach should be taken in these cases and if possible, surgical resection should be performed at the index operation. Currently the three-stage approach is reserved for those who are severely ill and cannot tolerate a resection at the index operation, those who are unresectable or those diagnosed with rectal cancer who would benefit from neoadjuvant treatment prior to resection.

Two-Stage Procedures

Surgeons moved away from the three-stage approach and adopted a two-stage approach, thus performing the resection at the index operation with creation of an end stoma and Hartmann's pouch with subsequent surgery to restore continuity. This is the most common procedure for left-sided obstructing lesions in high-risk patients in which it is not safe to perform an anastomosis. This has the advantage of relieving the obstruction as well as removal of the cancer at a single operation. The disadvantage is the patient is left with a stoma and a second surgery is needed for reversal. Of patients who have this procedure, it has been shown that 35% of patients never get their stoma taken down for a variety of reasons. ³³ If there is ischemia or perforation in the proximal bowel, then a subtotal colectomy is performed with end ileostomy and Hartmann's pouch. Zorcolo et al reviewed 323 patients who underwent emergent colorectal surgery for colorectal cancer or diverticular disease and found that those undergoing a Hartmann's procedure had a systemic morbidity rate of 39.5%, surgical morbidity of 24.3%, and mortality rate of 20.4%. They also found that those who had survived the index surgery and were alive 1 year later, only 39.2% had their stoma closed and more specifically only 8.7% of those with colorectal cancer had their stomas closed. ³⁴

Segmental Resection versus Subtotal Resection

For many years, no one even considered doing an anastomosis in the setting of an obstructing left colon cancer. It was thought that the risk of leak was too great, and the increased morbidity and mortality made it too dangerous of a proposition. The increased risk was thought to be related to not only the dilated proximal bowel but the fact that the bowel was unprepped. This led some to consider a subtotal colectomy as a means of lessening the complications and allowing anastomosis. This approach would remove the cancer as well as the proximal stool filled dilated bowel to the level of the ileocecal valve. In removal of the proximal dilated bowel any potential ischemic or weakened segments would also be removed. It had the added benefit of resecting any possible undiagnosed proximal synchronous lesions which occur in 3.7 to 11% of cases. ^{23 35 36 37 38} This resection enables a healthy terminal ileum of normal size to be anastomosed to a healthy distal bowel of normal size allowing for a single operation without the need for diversion or subsequent takedown. This does have the disadvantage of longer operative times as well as potential for chronic diarrhea. Hennekinne-Mucci et al looked at the results of patients treated with subtotal or total colectomy with primary anastomosis without diversion in the emergent setting. Of the 156 patients who presented with left colonic obstruction and received emergent surgery, 72 underwent either subtotal or total colectomy with primary anastomosis. Postoperative mortality and morbidity was 9.7 and 15%, respectively. In the immediate postoperative period, 20% of the patients complained of diarrhea but recovered with the use of antidiarrheal medications. ³⁸ This is the procedure of choice for those with distal obstructing cancers who present with ischemia or perforation of the cecum.

Colonic Irrigation and Manual Decompression

Ideally, if one could replicate elective procedure conditions, it would follow that perhaps the same surgery could be done safely with low morbidity and mortality even in the setting of an obstructing colon cancer. This led those who were considering a resection with primary anastomosis at the index operation to consider methods which could relieve the dilation of the proximal bowel as well as clear the proximal bowel of stool. Two options emerged which include colonic irrigation and manual decompression.

Colonic irrigation was first described by Muir in 1968 and later modified in 1980 by Dudley et al. ³⁹ The purpose of the irrigation is to decompress the proximal colon as well as decrease the amount of stool present so that a safe anastomosis can be performed during the index surgery. Dudley proposed a mechanism which included mobilization of the left colon with placement of a corrugated plastic tubing above the obstructing lesion and secured with nylon ties. A foley catheter is then placed through a small enterotomy in the terminal ileum and fed past the ileocecal valve into the cecum. Three liters of saline is irrigated through the colon with drainage into a separate collection basin outside the operative field. This keeps a completely closed system to irrigate the more proximal colon. ⁴⁰ Overall, studies have shown that this colonic irrigation is safe and feasible and allows for primary anastomosis in the setting of obstructing colon cancer with no increased risk of mortality or anastomotic leak rate, with ranges of 0 to 4.9% and 1.6 to 5.5%, respectively. There have been concerns for increased risk of wound infection with rates ranging from 18.4 to 24.5%. ^{41 42 43} Though this shows promising results, colonic irrigation can be cumbersome and adds time to the surgical case, limiting widespread adoption. Nyam et al suggested that this process might be simplified by performing only manual decompression without irrigation. They found no difference between those who received colonic irrigation versus those who had only colonic decompression with respect to mortality, leak rate, or wound infection rate and concluded that colonic irrigation was not necessary and thus one could avoid additional time as well as fluid shifts associated with colonic irrigation. ⁴⁴ Other studies performed comparing colonic irrigation to manual decompression showed similar results supporting no difference between these two with respect to mortality, leak rate, or infection. ^{45 46 47}

Comparison of Subtotal to Segmental Resection

Studies have compared subtotal colectomy to segmental resection. You et al examined short-term morbidity and long-term function as well as quality of life after segmental resection with primary anastomosis (SEG), subtotal resection with ileosigmoid anastomosis (STG-ISA) or total abdominal colectomy with ileorectal anastomosis (TC-IRA). Median stool frequency was two after SEC, four after STG-ISA, and five after TC-IRA even after dietary adjustments and antidiarrheal usage. A significant number felt restricted from preoperative social activity, housework, recreation, and travel after an ileorectal anastomosis. They concluded that there is significant change in body function as well as quality of life for more extensive resections, so careful consideration should be taken when choosing this as the treatment option for patients. ⁴⁸ Käser et al also reviewed their data regarding morbidity associated with segmental colectomy versus subtotal colectomy in the setting of malignant left-sided colon obstruction and found the anastomotic leak rate was lower in the subtotal colectomy group (0%) compared with the segmental colectomy group (20%), ( p  = 0.042) though this did not reach statistical significance. ⁴⁹ Torralba et al also looked at subtotal colectomy versus intraoperative colonic irrigation with segmental resection in the setting of obstructing left colon cancer and found the complication rate to be significantly higher in the colonic irrigation group (41.9%) compared with the subtotal colectomy group (14.2%). Specifically, the leak rate and wound infection rate was 16.1% compared with 0 and 22.5% compared with 0%, respectively. ⁵⁰ Both of these authors conclude that subtotal colectomy should be the procedure of choice for obstructing left colon cancers. Finally, the SCOTIA study group performed a multicenter prospective randomized trial comparing subtotal colectomy to segmental resection with anastomosis after colonic irrigation. They found that hospital mortality and complication rates were not significantly different, however, they did find that increased bowel function was significantly more common in the subtotal colectomy group, 40% compared with 11%. They concluded that segmental resection following intraoperative irrigation is the preferred treatment for left-sided malignant colonic obstruction with the exception of those who present with a perforated cecum or have a known synchronous cancer in the proximal colon. ⁵¹ The final decision regarding segmental versus subtotal resection needs to be based on the integrity of the proximal bowel. If the proximal bowel is heathy and can be preserved, studies have shown that segmental resection can be performed safely, with low morbidity, and provide the patient with improved functional outcome with the option of proximal diversion if there is concern for potential anastomotic leak.

Single-Stage Procedures

A single-stage procedure not only relieves the obstruction as well as eradicates the cancer, but also has the additional benefits of avoidance of stomal complications, averting the risk of a second surgery as well as potential of improved quality of life for the patient. It is known that those given a stoma at the index operation have a very high chance that it will remain. There are numerous studies comparing the outcomes of single- versus two-stage procedures for the treatment of obstructing left-sided cancers. Overall, the resection rate of left-sided obstructing cancers ranges from 90 to 97.8%. Of those patients who did undergo resection, 46.6 to 83% underwent a primary anastomosis at the index operation while 17 to 53.4% received resection with a creation of a stoma. Mortality rate and leak rates in these studies ranged from 5.7 to 14.7% and 3 to 18%, respectively. ^{2 3 22 23 34} These ranges reflect the fact that these procedures are not randomized and that there is variation of criteria used to determine who might get a primary anastomosis and who might be left in discontinuity. Single-stage surgery with resection and primary anastomosis is safe in the properly selected patient. Aslar et al stressed the importance of the condition of patient as the prime driver of the outcome for the patient instead of the procedure performed. They also found no significant difference in surgical outcomes when the data was stratified by procedure, but a significant difference when the results were stratified by outcome. Sicker patients were much more likely to undergo resection with stoma formation. ³ Zorcolo et al studied the outcome of patients following resection and primary anastomosis requiring emergent surgery for left-sided colonic disease and also found that high ASA in elderly patients predicted higher leak rates and also concluded that the decision to perform a primary anastomosis should be based on the overall condition of the patient. ³⁴ Resection with primary anastomosis without diversion should be reserved for only those patients who have no significant comorbidities, are stable and have healthy bowel after resection suitable for anastomosis.

Colonic Stents—Bridge to Surgery

Physicians continued to search for ways to decrease the high mortality and morbidity associated with emergent operation for malignant bowel obstruction. A potential breakthrough occurred in the early 1990s, with the introduction of the self-expanding metallic stents (SEMS). Stenting can be used as a bridge to surgery, for palliation and for obstruction from extracolonic malignancies. The only absolute contraindication for stent placement is perforation. Relative contraindications include septic shock, as this may suggest the presence of a perforation, and low rectal obstructions. Stenting of low rectal lesions had been avoided due to the fear of severe anal pain and tenesmus. Lee et al showed that stenting of low rectal lesions (<5cm from anal verge) is possible with a technical success rate of 87% and clinical success rate of 69.2%. Only 4 of 27 (14.8%) patients had anal pain and all were well controlled with oral analgesia. ⁵²

Several factors should be considered in deciding if a stent should be placed as a bridge to surgery. If the patient is unstable or there is concern for perforation or impending perforation, a stent should not be placed, and the patient should be taken for surgery. If on the other hand, the patient is stable, without evidence of perforation, stent placement may be entertained. One must have a favorable lesion to stent (<4 cm in length, partial obstruction, and no significant acute angles) as well as the expertise necessary to successfully place the stent to minimize stent complications. Proper selection and execution of stent placement can improve patient outcomes by decompression of the obstruction allowing preoperative bowel preparation, thus avoiding the need for intraoperative colonic irrigation, and allowing resection and primary anastomosis without diversion at the index case. An in-depth discussion of colonic stents is covered in a separate article so will not be covered here.

Primary Anastomosis with Proximal Diversion

If there is some concern for anastomotic dehiscence, proximal diversion can be done to protect the anastomosis. However, one must keep in mind that if this concern is significant, that perhaps it is safer not to perform an anastomosis at all. Proximal diversion with a loop ileostomy has some advantages over a colostomy. An ileostomy has the advantage of typically being easier to create and has less likelihood of damage to the arcade of Riolan either at its creation or its takedown. The reversal surgery of a diverting loop ileostomy can often be done through the stoma site itself thus avoiding another midline incision and its associated morbidities. Overall morbidity and mortality are higher with end colostomy takedown versus takedown a loop ileostomy. The disadvantage of a loop ileostomy is it has the higher risk of dehydration and electrolyte abnormalities associated with higher stoma output. The final decision as to which is better for a particular patient needs to be made at the time of surgery and is dependent not only on patient factors but also on the experience and knowledge of the operating surgeon.

Special Consideration for Palliation

The goals of palliation are very different from those of curative surgery. The main objective should be to improve quality of life by relieving symptoms, shortening length of stay, avoiding stoma creation if possible, and returning to activities of daily living as soon as possible with the least possible morbidity. The goal to resume chemotherapy as quickly as possible should also be considered. There are several palliative treatment options for patients presenting with malignant LBO in the setting of metastatic disease. Operative options might include palliative resection, bypass, or diverting stoma. A nonoperative option might be palliative stent placement. Dalal et al reviewed 141 patients who were treated for malignant bowel obstruction in the setting of stage IV disease. Of those with LBO, palliative treatment was successful in resolving pain in 92% of the cases. Palliative treatment for LBO included placement of a g-tube (2%), stent placement (45%), resection (27%), bypass (3%), and stoma (22%). Morbidity and mortality rates were 25 and 4%, respectively. Of the patients treated with stents, 100% were discharged home compared with 88% of those treated operatively. A key point was that those patient with ascites or carcinomatosis were at higher risk of unsuccessful surgery. They concluded that with careful patient selection, palliative treatment could lead to most patients being discharged home with resolution of their symptoms. ⁵³

Outcomes for Malignant Obstructing Colon Cancer

Patients who present with obstructing colorectal cancer are known to have a worse prognosis compared with those without obstruction. This is true even when having potentially curative surgery. Ohman performed a retrospective study including 1,061 patients comparing those presenting with obstructing colon cancer to those undergoing elective surgeries for colorectal cancer. The 5-year survival rate for those with obstruction was 16% compared with 37% in the elective surgery group. This compares to rates found in the literature ranging from 12 to 29% and 27 to 53%, respectively. There were significantly more advanced tumors in the obstructed group, so it follows that a significantly lower number of patients in the obstructed group compared with the elective surgery group were able to be resected for cure, 53 versus 72%, respectively. Survival rates were not significantly different within each stage of cancer. ^{54 55} Lavanchy et al compared outcomes of those undergoing emergent surgery versus elective surgery for the treatment of colon cancer. Twenty two percent of the patients underwent emergent surgery for either obstruction (63.8%), perforation (17.1%), or bleeding (14.3%). With regard to oncologic factors, those undergoing emergent surgery were more likely to have a T4 tumor, 42.9 versus 16.5% ( p ≤ 0.001), had a higher number of lymph nodes retrieved, 26 versus 20 ( p ≤ 0.001), were less likely to undergo chemotherapy, 19 versus 32% ( p  = 0.028), and had a higher mortality rate, 8.4 versus 3%, ( p  = 0.023). Overall survival and median disease-free survival were not statistically different, 72.4 versus 78.9%, ( p  = 0.257), again showing that recurrence was based on stage and not the nature of the case. ⁵⁶

To better understand this negative effect further studies were done to determine prognostic factors that might help direct therapy. Manceau et al studied outcomes of over 2,000 patients treated for obstructive colon cancer. Lymph node involvement and synchronous distant metastasis were found in 62 and 39%, respectively. Excluding patients who died during the first 30 days, the median overall survival was 77.6 months with 3-, 5-, and 10-year overall survival rates of 68, 56, and 45%, respectively. In multivariate analysis, age ≥75, ASA score ≥3, right-sided colon cancer, presence of synchronous metastasis, and absence of postoperative chemotherapy were independent factors of poor overall survival. They found anastomotic leak to be a negative prognostic factor, however, only effecting short-term mortality and increasing the risk of recurrence. ⁵⁷ Carraro et al also compared obstructing colon cancer to nonobstructing cancer and found a significantly higher mean age in the obstructing group compared with the nonobstructing group, 69.3 versus 66 ( p ≤ 0.001). Survival was significantly affected by age >70 years ( p <0.001), Duke's stage ( p <0.001), histologic grade ( p <0.006), and presence of positive lymph nodes ( p <0.001). Local recurrence rates were similar; however, metastatic recurrence rates were higher in the obstructed group, 12.8 versus 10.4% ( p  = 0.44), and 27.6 versus 17.8% ( p  = 0.029), respectively. ⁵⁸

Finally, right- and left-sided cancers behave differently, potentially from differing biological factors. A meta-analysis was performed including 66 studies and 1,437,846 patients and found that left-sided colon cancers were associated with an increased survival compared with right-sided cancers. (HR = 0.82; 95% CI, 0.79–0.84, p <0.001) and was independent of stage, race, or administration of chemotherapy. ⁵⁹ Faucheron et al found that patients presenting with right-sided obstructions were older (>80 years old) and more likely to have advanced locoregional disease compared with those with left-sided obstruction lesions, 58 versus 30%, ( p  = 0.03) and 77.9 versus 51%, ( p  = 0.04). They also found that long-term overall survival was worse in those with right-sided lesions. ⁶⁰ Overall recurrence rates as well as distant metastatic rates were significantly higher in patients with obstructing right-sided colon cancers compared with nonobstructing right-sided lesions. This data suggests that not only tumor stage, but obstruction is an independent prognostic factor. ⁶¹

Conclusion

LBO is a serious condition with high risk of morbidity and mortality. Expedient, thoughtful management of this problem can improve outcomes. Determination of the presence of at-risk bowel will dictate the urgency of intervention. There are many therapeutic options available depending on the presentation of the patient and these must be weighed carefully by the surgeon. One must consider all aspects of the patient in determining what treatment option is best for the patient. Unstable patients are better suited to limited time in the OR with the focus on resolving the obstruction. Choice and location of diversion needs to be carefully planned so that if possible, future therapeutic options are not hindered. If stable but at high risk of anastomotic leak, one should consider resection, ± anastomosis and proximal diversion. For stable patients, one can consider resection and primary anastomosis with manual decompression. If resection is planned, it is imperative that oncologic principles be followed to assure the best outcomes for these patients. Chemotherapy should be considered for all presenting with obstruction as outcomes are worse with higher risk of metastatic disease. Finally, stenting or surgical diversion can be utilized for palliation. Though perioperative morbidity and mortality are higher in this population and survival outcomes decreased, careful attention to detail and timely intervention can improve the outcomes of these patients.

Fig. 1.

( A ) Obstructing transverse colon cancer ( arrow pointing to mass ). ( B ) Obstructing transverse colon cancer: dilated cecum with decompressed left colon. ( C ) Transverse obstructing colon cancer: dilated cecum with decompression of left colon. ( D ) Obstructing transverse colon cancer - gross pathology.

Fig. 2.

( A ) Obstructing sigmoid colon cancer - plain film. ( B ) Obstructing sigmoid colon cancer ( arrow pointing to mass ). ( C ) Obstructing sigmoid cancer CT coronal view. ( D ) Obstructing sigmoid colon cancer - gross pathology.

Fig. 3.

( A ) Obstructing rectal cancer: extreme dilation of the entire colon. ( B ) Obstructing rectal cancer: extreme dilation of the entire colon. ( C ) Obstructing rectal cancer with extreme dilation of the entire colon.