INTRODUCTION — Bowel obstruction occurs when there is reduced or absent progress of the forward flow of gastric and intestinal contents through the gastrointestinal tract. Bowel obstruction caused by cancer is referred to as malignant bowel obstruction [1]. In everyday practice, however, it is often difficult to differentiate malignant from benign bowel obstruction, as cancer patients, especially those who have a history of abdominal surgery, can develop bowel obstruction from benign etiologies (eg, adhesions). Regardless of etiology, management of bowel obstruction in patients with advanced stage cancer can be difficult; a decision to proceed to surgical intervention requires careful weighing of risks and benefits, including an assessment of the estimated life expectancy and the patient's goals and preferences for care.
In this topic, we discuss the etiologies, clinical features, diagnostic evaluation, and management of bowel obstruction in cancer patients. General discussions of small and large obstruction can be found in other topics. (See "Management of small bowel obstruction in adults" and "Large bowel obstruction".)
EPIDEMIOLOGY — Malignant bowel obstruction (MBO) is common in patients with abdominal or pelvic cancers. It is most prevalent in ovarian cancer (5.5 to 42.0 percent), colorectal cancer (4.4 to 24.0 percent), and gastric cancer [2]. Uncommonly, metastases from extra-abdominal cancers, including breast cancer, lung cancer, and melanoma, can also cause MBO. MBO presents most frequently in advanced stage cancer patients. In a large cohort study of 490 cancer patients from MD Anderson, the anatomic sites of obstruction were classified as gastric outlet, small bowel, and large bowel in 16, 64, and 20 percent, respectively [3].
ETIOLOGY AND PATHOPHYSIOLOGY — Intestinal obstruction may be caused by benign or malignant etiologies (table 1). Distinguishing benign causes from malignant bowel obstruction in a cancer patient can be challenging but has profound impacts on management decisions and prognosis [4]. Among cancer patients, large bowel obstruction is four to five times less frequent than small bowel obstruction, and the causes differ substantially [3]. Most small bowel obstructions are due to benign causes, whereas malignant causes predominate in large bowel obstructions.
Benign etiologies — Bowel obstruction can be caused by non-malignant causes, including postsurgical adhesions, radiation enteritis, or other infectious or inflammatory gastrointestinal diseases (eg, abscess, phlegmon) (table 1). Benign conditions are more commonly observed when the obstruction is in the small bowel, where primary tumors are rare.
Postsurgical adhesion — Among cancer patients with small bowel obstructions, adhesions are the main cause in 50 to 80 percent of cases [5-7]. Almost all adhesions are postoperative, with a minority being secondary to peritonitis or other causes [8]. (See "Postoperative peritoneal adhesions in adults and their prevention".)
Radiation enteritis — Radiation therapy induces adhesive and fibrotic changes in the small bowel mesentery, and this is a late complication, typically developing one year or more after treatment is completed. Radiation therapy can also cause serositis, which can lead to luminal narrowing and bowel dysmotility. The ileal loops in the pelvis are most susceptible to developing radiation enteritis. (See "Overview of gastrointestinal toxicity of radiation therapy", section on 'Enteritis' and "Diagnosis and management of chronic radiation enteritis".)
Malignant etiologies — Malignant etiologies of bowel obstruction include intraluminal obstruction by a primary tumor or a local tumor recurrence, and external compression by peritoneal metastases involving the bowel and/or adjacent mesentery. In contrast to small bowel obstruction, malignant etiologies predominate in large bowel obstruction.
Intraluminal obstruction — An obstructing intraluminal lesion, most commonly a colorectal malignancy, is the most common cause of large bowel obstruction (>60 percent) [9]. (See "Large bowel obstruction", section on 'Etiology'.)
Extrinsic compression — Extrinsic compression causing a bowel obstruction can occur by a number of mechanisms. Peritoneal deposits or tumor masses may cause external compression of loops of small bowel. Additionally, pericancer inflammation can lead to adhesions. Mesenteric and omental tumor involvement may angulate the bowel and provoke an extramural bowel occlusion.
CLINICAL FEATURES — Symptoms of intestinal obstruction typically include crampy, paroxysmal abdominal pain, loss of appetite, nausea, vomiting, inability to have a bowel movement or pass flatus, and abdominal bloating. Physical findings include a distended, tender abdomen with or without a palpable mass. The patterns of clinical presentation are further determined by the level of obstruction (table 2).
Proximal bowel obstruction — Patients with an obstruction at the level of the stomach or proximal small bowel typically have abrupt onset of abdominal pain and nausea/vomiting in large volumes, with nausea temporarily relieved after each episode of vomiting. The loss of gastric secretions due to repeated vomiting may lead to metabolic alkalosis and hypokalemia. Abdominal distension may be absent in proximal obstruction. (See "Etiologies, clinical manifestations, and diagnosis of mechanical small bowel obstruction in adults", section on 'Clinical presentations'.)
Distal bowel obstruction — Patients with an obstruction at the level of the distal small bowel, colon, or rectum often have insidious and progressive onset of symptoms, including abdominal pain, constipation or obstipation, and abdominal distension [10]. Abdominal distention is more pronounced than that typically seen with proximal bowel obstruction. Bowel sounds are usually hypoactive in patients with distal bowel obstruction due to the cessation of peristalsis. (See "Large bowel obstruction", section on 'Clinical presentations'.)
The clinical features of malignant and benign bowel obstructions may be indistinguishable without imaging studies [4]. (See 'Radiographic studies' below.)
Natural history — In addition to being caused by benign or malignant etiologies, intestinal obstruction in cancer patients may be partial or complete, acute or chronic, reversible (especially when caused by postsurgical adhesions) or irreversible, and single or multiple, and they may occur at any level of the gastrointestinal tract.
Intestinal obstruction may progress rapidly, causing serious, life-threatening complications such as intestinal ischemia, infarction, and perforation, or peritonitis. (See "Management of small bowel obstruction in adults", section on 'Indications for immediate surgery'.)
More frequently, however, the progression of bowel obstruction is slow and insidious. With a partial obstruction, the intestine goes through cycles of luminal distension and secretion, leading to local inflammation and luminal accumulation of fluid, gas, and solid waste. As the obstruction progresses from partial to complete, symptoms gradually shift from nausea and colicky pain with occasional vomiting to less colicky pain but greater gastrointestinal symptoms and abdominal distention. If left untreated, bowel obstruction will eventually lead to intestinal ischemia, when patients present with severe abdominal pain and tenderness.
DIAGNOSTIC EVALUATION — For patients with known or suspected cancer who present with signs and symptoms that are consistent with a bowel obstruction, a diagnostic workup should be performed not only to establish the diagnosis and the level of obstruction, but also to assess tumor burden, which may influence the goals of care. Depending on the extent of cancer spread, many patients will not be candidates for aggressive (surgical) treatment of their bowel obstruction. (See 'Surgical candidates' below.)
A diagnostic workup for suspected bowel obstruction typically consists of one or more radiographic studies and laboratory evaluation.
Radiographic studies
Plain abdominal radiography — Plain abdominal radiography is the initial study of choice due to its wide availability and low cost. Plain abdominal films (supine and nondependent [either upright or left lateral decubitus to detect pneumoperitoneum]) have a high sensitivity for diagnosing high-grade small bowel obstructions, which are characterized by the presence of bowel distention and at least two air fluid levels (image 1) [11]. Low-grade small bowel obstruction is more accurately diagnosed by computed tomography (CT). (See "Etiologies, clinical manifestations, and diagnosis of mechanical small bowel obstruction in adults", section on 'Plain radiography'.)
Supine and nondependent abdominal films are also helpful in patients suspected of having a large bowel obstruction [12]. In such patients, plain abdominal radiography can diagnose a large bowel obstruction while excluding a small bowel obstruction or a pneumoperitoneum. Large bowel obstruction is characterized by dilated colon proximal and no air distal to the site of obstruction. Additionally, air-fluid levels are often visible in the dilated colon on the upright or decubitus radiographs [12]. The presence of air-fluid levels suggests a more acute obstruction with which the fluid in the colon has not been absorbed. (See "Large bowel obstruction", section on 'Diagnosis'.)
Complications of large bowel obstruction can sometimes be diagnosed using plain abdominal radiography (image 2). Pneumoperitoneum (ie, "free air") indicates perforated viscus, for which urgent surgical exploration is mandated. The presence of intramural gas (ie, pneumatosis intestinalis (image 3)) is a sign of bowel necrosis and impending perforation [13]. Toxic megacolon, a complication that can occur with several diseases of the colon, such as infections, ischemia, and inflammation, is characterized by its hallmark features of marked bowel wall thickening, loss of haustration, and segmental colonic parietal wall thinning (image 4) [14]. (See "Pneumatosis intestinalis", section on 'Evaluation' and "Toxic megacolon", section on 'Imaging studies'.)
Abdominopelvic computed tomography — Abdominopelvic CT is commonly used to diagnose clinically suspected bowel obstructions. Compared with plain abdominal radiography, contrast-enhanced helical multidetector row CT with multiplanar reconstructions can provide more clinically relevant information, which may lead to changes in management [15]. CT can determine if bowel obstruction is present, localize the obstructive site and the degree of obstruction, diagnose the presence of closed-loop obstruction, and identify ischemia or perforation of the involved bowel (image 5 and image 6 and image 7 and image 8). Owing to its ability to demonstrate strangulation, CT is considered the best modality for determining which patients would benefit from immediate surgical intervention versus conservative management [16]. Furthermore, in patients with malignant bowel obstruction (MBO), CT can also detect local and regional metastases.
Primary cancer is rare in the small bowel. Most intraluminal causes of small bowel obstruction manifest as an endoluminal mass with imaging characteristics that are distinct from those of the remaining enteric content (ie, stool) [17]. When a primary small bowel cancer manifests as an obstruction, it usually occurs in an advanced stage of disease and is characterized by a pronounced, asymmetric, and irregular mural thickening at the transition point. Small bowel involvement by metastatic cancer is more commonly seen in the form of peritoneal carcinomatosis [18]. (See "Diagnosis and staging of small bowel neoplasms", section on 'Computed tomography scan'.)
In radiation enteritis, CT findings may include a narrowing of the lumen secondary to mural thickening, an angular bowel wall due to adhesions, and retraction of the mesentery. There may also be abnormal enhancement of a segment of thickened bowel wall if that segment was present within the radiation field [19]. (See "Diagnosis and management of chronic radiation enteritis", section on 'Imaging'.)
Small bowel obstruction due to adhesions is characterized by the presence of dilated small bowel loops proximally and normal-caliber or collapsed loops distally. It is primarily a diagnosis of exclusion because adhesive bands are not typically visible on conventional CT. However, it is possible to see an abrupt change in the caliber of the bowel without any associated mass lesions, significant inflammation, or bowel wall thickening at the transition point (image 9 and image 10). A history of abdominal surgery, the kinking and tethering of the adjacent nonobstructed bowel loops, and the absence of a mass lesion either within or outside of the bowel may support the diagnosis. (See "Etiologies, clinical manifestations, and diagnosis of mechanical small bowel obstruction in adults", section on 'Abdominal CT'.)
It is preferable to give oral as well as intravenous contrast, as this will allow the grade of small bowel obstruction to be determined by CT imaging. A low-grade partial small bowel obstruction is characterized by flow of contrast material through the point of obstruction. On the other hand, the partial small bowel obstruction is characterized as high-grade when there is some stasis and delay in the passage of the contrast so that diluted oral contrast material appears in the distended proximal bowel and minimal contrast material appears in the collapsed distal loops. Finally, a complete obstruction is diagnosed when there is no passage of contrast medium beyond the point of obstruction, even on delayed repeat scans [20].
Large bowel obstruction is characterized by dilated large bowel proximal to a transition point and decompressed bowel distal to the obstruction. The presence of a transition point is considered a reliable finding for the diagnosis of large bowel obstruction on CT [21]. (See "Large bowel obstruction", section on 'Diagnosis'.)
Abdominopelvic magnetic resonance imaging — Although it is rarely used as the first-line imaging modality, abdominopelvic magnetic resonance imaging (MRI) can be used to diagnose bowel obstruction in select patients. Abdominopelvic MRI has the advantage of being better able than CT to detect and characterize peritoneal and hepatic disease. However, CT scans can be performed more quickly, and the equipment is typically more readily available.
In one study, benign and MBO were distinguishable by their MRI characteristics after administration of gadolinium contrast [4]. Bowel obstruction due to benign causes (most often adhesions) showed mild peritoneal enhancement and mesenteric infiltration, often with segmental or diffuse mural thickening without an obstructing mass. By contrast, malignant obstructions (most often due to carcinomatosis) produced moderate or bulky peritoneal thickening and tumor masses, often with focal or segmental mural thickening, serosal metastases, and disseminated abdominal tumor. Segmental mural thickening can occur in either benign or MBO.
Abdominal ultrasound — Abdominal ultrasound is of limited value in diagnosing bowel obstruction due to interference from gas-filled bowel loops. Adhesions, the most common cause of mechanical small bowel obstruction, are not detected with this technique. However, when the obstructed bowel segments are dilated and filled with fluid, it is possible to detect the level of obstruction and the cause of the obstruction by using the fluid-filled bowel as a sonic window [22].
Laboratory evaluation — In patients with suspected or confirmed bowel obstruction, laboratory studies should be performed to evaluate for the presence of metabolic derangements (eg, acidosis, electrolyte abnormality, renal injury) and leukocytosis. Severe abnormalities (extreme leucocytosis, metabolic or lactic acidosis) may suggest bowel ischemia, albeit with a low specificity [15].
A role for circulating tumor markers in the distinction between benign and malignant causes of bowel obstruction is not established. However, for patients in whom there has been a correlation between levels of a serum tumor marker (eg, carcinoembryonic antigen [CEA] for colorectal cancer, cancer antigen 125 [CA-125] for ovarian cancer) and their disease burden, assay of that specific tumor marker may help to clarify the tumor burden. (See "Serum biomarkers for evaluation of an adnexal mass for epithelial carcinoma of the ovary, fallopian tube, or peritoneum", section on 'Cancer antigen 125' and "Overview of epithelial carcinoma of the ovary, fallopian tube, and peritoneum", section on 'CA 125 surveillance' and "Clinical presentation, diagnosis, and staging of colorectal cancer", section on 'Tumor markers'.)
DIFFERENTIAL DIAGNOSES — In addition to postsurgical adhesions and radiation enteritis, the differential diagnosis of a malignant bowel obstruction includes nonobstructive intestinal disorders (ileus, intestinal pseudo-obstruction) and severe constipation.
Ileus or pseudo-obstruction — Nonobstructive intestinal motility disorders can cause signs and symptoms similar to those of bowel obstruction. The process is commonly referred to as paralytic or adynamic ileus when it is acute or self-limited, and as intestinal pseudo-obstruction when it is chronic or recurrent. Ileus or pseudo-obstruction can be caused by recent abdominal or pelvic surgery, intra-abdominal infections (eg, gastroenteritis or appendicitis), medications (eg, opioids, antidepressants, or antimuscarinics), metabolic disturbances (eg, hypokalemia), and peritoneal carcinomatosis without extrinsic bowel compression (by involving the autonomic nerves that innervate the intestine). Intestinal pseudo-obstruction is most frequently caused by neuromuscular disorders such as Parkinson disease. (See "Postoperative ileus" and "Chronic intestinal pseudo-obstruction: Etiology, clinical manifestations, and diagnosis".)
Bowel obstruction and ileus/pseudo-obstruction can be distinguished by radiographic studies. On plain abdominal radiography, paralytic ileus is characterized by small bowel dilation in addition to colonic dilation (image 11) [23]. Also, colonic dilation due to paralytic ileus or colonic pseudo-obstruction usually occurs with minimal colonic fluid. Hence, the presence of air-fluid levels should raise the suspicion of an obstruction. Nevertheless, the differentiation between large bowel obstruction and pseudo-obstruction remains difficult, and computed tomography (CT) may be needed to affirm the diagnosis. On CT, ileus/pseudo-obstruction is characterized by diffuse small and large bowel dilatation without a transition point (image 12) [24].
Constipation — Severe constipation and/or fecal impaction can lead to obstruction of either the large or small intestine. Constipation may be caused by illness or medications such as anticholinergics and opioids. In particular, chronic opioid use without attention to a prophylactic bowel regimen can perpetuate a vicious circle of opioids, constipation, pain, increased doses of opioids, and worse constipation. (See "Approach to symptom assessment in palliative care", section on 'Constipation' and "Prevention and management of side effects in patients receiving opioids for chronic pain", section on 'Constipation'.)
MANAGEMENT — The management of cancer patients with bowel obstruction should be individualized by taking into consideration each patient's clinical condition, cancer stage and estimated life expectancy, response to previous cancer treatment, performance status, comorbidities, and preferences, wishes, and goals of therapy. Any surgical intervention in cancer patients with bowel obstruction should be preceded by a multidisciplinary evaluation and discussion that involves the surgeon, the patient's medical doctor or oncologist, and the patient/family members [25]. A palliative care consultation may be beneficial, particularly if a goals of care discussion has not been undertaken previously with the patient and his or her family.
Management may be surgical or medical. A large fraction are categorically inoperable, and in general, survival is limited. In a large cohort study of 490 cancer patients with bowel obstruction, 49 percent were managed medically, 32 percent surgically, and 17 percent procedurally (radiologic or endoscopic). Median overall survival was 125 days (interquartile range 42 to 399 days), with 42 percent of patients dying within 90 days of surgical consultation. However, compared with medical management, surgery was not associated with an increased mortality (hazard ratio [HR] 0.87, 95% CI 0.66-1.13) [3]. An algorithmic approach to management is provided (algorithm 1).
Initial management — Patients with bowel obstruction who present with generalized peritonitis, clinical deterioration (as evidenced by fever, leukocytosis, tachycardia, metabolic acidosis, or continuous pain), pneumoperitoneum, or imaging signs of ischemia should be offered urgent surgical exploration if it is consistent with their goals and preferences for care. (See 'Surgical candidates' below and "Management of small bowel obstruction in adults", section on 'Indications for immediate surgery'.)
The remainder should be initially managed nonoperatively with nasogastric decompression of the stomach/intestine and intravenous fluids. Initial nonoperative management is successful in relieving bowel obstruction in some patients. Patients whose symptoms have not been relieved with 48 to 72 hours of nonoperative management require definitive surgical intervention if they are deemed suitable surgical candidates. (See "Management of small bowel obstruction in adults", section on 'Initial management'.)
Surgical candidates — In general, patients should be considered potential surgical candidates unless they are critically ill, have a poor performance status (table 3), or refuse surgery [26]. As such, surgical evaluation should be provided for most cancer patients with bowel obstruction. After a thorough surgical evaluation, however, many patients may be deemed unlikely to benefit from surgery for anatomic or clinical reasons [27].
In the published literature, between 6 and 50 percent of cancer patients with bowel obstruction are considered inoperable after surgical evaluation [28]. As expected, there are marked variations in clinical practice with regard to selecting patients who would or would not benefit from surgery. The following clinical or radiographic factors have been associated with poor surgical outcomes in various studies of cancer patients with bowel obstruction [3,28-35]:
●Imaging evidence of residual primary or recurrent tumor was associated with not eating at discharge (odds ratio [OR] 0.45, 95% CI 0.29-0.72) and with death within 90 days of surgical consultation (HR 1.73, 95% CI 1.31-2.27) [3,31].
●Imaging evidence of peritoneal disease (carcinomatosis) was associated with not eating at discharge (OR 0.38, 95% CI 0.24-0.59) and with death within 90 days of surgical consultation (HR 1.98, 95% CI 1.44-2.71) [3,32].
●Imaging evidence of multiple sites, as opposed to a single site, of disease was associated with worse survival (HR 1.4, 95% CI 1.1-1.8) [30-32].
●Patients with ascites were less likely to benefit from surgical procedures [32].
●Hypoalbuminemia (<3.5 g/dL) was associated with not eating at discharge (OR 0.57, 95% CI 0.37-0.87) [3]. In another study, patients with hypoalbuminemia had higher surgical mortality (14 versus 11 percent) [29].
●Old age (≥65 years old) was associated with death within 90 days of surgical consultation (HR 0.72, 95% CI 0.54-0.95) [3]. In another study, older patients had higher surgical mortality (18 versus 11 percent) [29].
●Poor performance status is also an important indicator of prognosis in patients with malignant bowel obstruction (MBO). In one study of patients with MBO in the setting of stage IV noncurative cancer, an Eastern Cooperative Oncology Group (ECOG) score of 0 or 1 was the strongest predictor of overall survival [35]. Median survival for those with an ECOG performance status of 0 to 1 was 222 days, while it was 63 and 27 days for those with ECOG performance status 2 and 3 to 4, respectively. In another study, patients with a poor performance status had higher surgical mortality (16 versus 11 percent) [29]. In that study, performance status was the most significant predictor of survival.
●In general, a synchronous presentation of both stage IV malignancy and bowel obstruction carries a better prognosis than a metachronous presentation [34].
Nomograms based on some of these factors have been developed to predict 30-day mortality and select which patients may benefit from surgery [33].
Our absolute and relative contraindications to surgery in patients with MBO are outlined in the table (table 4). However, ultimately the decision of whether or not to offer surgery must be individualized for each patient [36]. A decision to proceed with surgical intervention requires careful weighing of risks and benefits, including an assessment of the estimated life expectancy and patient goals and preferences prior to surgery. A palliative care consultation may be beneficial, particularly if a goals of care discussion has not been undertaken previously with the patient and his or her family. (See "Benefits, services, and models of subspecialty palliative care", section on 'Establishing goals of care' and "Discussing goals of care".)
For patients with advanced cancer, the decision-making process may be particularly difficult due to the limited benefit and the high risks associated with surgical procedures [37]. In particular, patients with MBO due to peritoneal metastases may have only weeks or months to live [38,39], and they are often poor surgical candidates because of malnutrition [40]. Although palliative surgery can benefit some patients [26], it is also associated with high mortality and lengthy hospitalization, especially considering patients' limited life expectancy. Some patients with advanced stage cancer may prefer to forego such burdensome treatment near the end of their lives [41]. For others who choose surgery, surgeons should set realistic goals, explain the limitations of surgery, and clarify the patient's preferences with regard to aggressive postoperative interventions prior to surgery [42].
Surgical outcomes — For patients who undergo surgical intervention for bowel obstruction, there are few data addressing surgical outcomes and the relative value of palliative surgery versus medical management. Studies of bowel obstruction in cancer patients are limited by their retrospective nature, heterogenous patient population, and a lack of definition of what constitutes successful palliation.
While most would consider an intervention successful if it prolongs life by 30 or 60 days, this definition may be too narrow for the advanced cancer population. In patients with advanced cancer, palliative interventions are intended more to relieve symptoms and improve quality of life than to prolong it [43]. Others use discharge from the hospital or the ability to tolerate oral nutrition for 30 or 60 days as criteria for successful outcomes [44]. However, these definitions are quite arbitrary, do not provide enough information about the success or failure of the surgery performed, and do not account for the patient's quality of life.
Systemic reviews on this subject have been inconclusive:
●An updated Cochrane review of surgery for the resolution of symptoms of MBO in advanced gynecological and gastrointestinal cancers reported marked variation in clinical practice [45]. Among patients who underwent surgery, "clinical resolution" of cancer-related bowel obstruction was reported in 27 to over 68 percent of patients, and between 30 and 100 percent of patients were able to resume feeding orally, another popular proxy for symptom resolution. The recurrence rate varied from 0 to 63 percent, although the time to recurrence was often not reported. Postoperative morbidity and mortality also varied widely, with no data reported for quality of life. Additionally, outcomes were not segregated by specific interventions [45]. The authors concluded that the published literature (solely consisting of retrospective case series) was insufficient to draw any meaningful conclusions as to clinical resolution, symptom resolution, ability to feed orally, and rates of re-obstruction, and they called for greater use of validated outcome measures of symptom control and quality of life scores, and standardization of management.
●A second review focusing on ovarian cancer alone found one retrospective study showing a survival benefit with surgery [46], but the review was unable to reach a definite conclusion as to the role of surgery, and overall, found low-quality evidence comparing palliative surgery and medical management [47]. (See "Cancer of the ovary, fallopian tube, and peritoneum: Surgical options for recurrent cancer", section on 'Relief of bowel obstruction'.)
Despite improvements in surgical and perioperative care, postoperative mortality rates range from 0 to 40 percent, with complication rates ranging from 18 to 90 percent in advanced cancer patients [48-51].
Ultimately, the decision of whether or not to undergo surgery must be individualized for each patient, taking into account the potential benefits and risks of the planned procedure, the patient's clinical condition, cancer stage and estimated life expectancy, response to previous cancer treatment, performance status, and comorbidities, and their preferences, wishes, and goals of therapy.
Surgical techniques — The choice of surgical technique depends on each patient's anatomy, comorbidities, risk factors, and cancer prognosis. In general, the surgical technique that will produce the most durable relief of symptoms with a reasonable operative morbidity should be used. Surgical techniques commonly performed to relieve bowel obstruction include bowel resection, bypass, diversion, stenting, and lysis of adhesions.
●Resection – Complete surgical resection of a tumor that causes bowel obstruction is desirable if the tumor can be resected with negative margins. Otherwise, debulking of a tumor is generally of limited benefit as the tumor will likely grow back. The exception is in ovarian cancer, where initial cytoreductive surgery has shown a survival benefit when combined with chemotherapy. (See "Cancer of the ovary, fallopian tube, and peritoneum: Surgical cytoreduction", section on 'Bowel resection'.)
Notably, the role of surgical resection for recurrent ovarian cancer, especially in the setting of bowel obstruction, is less well established and more controversial. The benefit of surgery in this setting is short lived, and it is associated with high rates of morbidity and mortality. (See "Cancer of the ovary, fallopian tube, and peritoneum: Surgical options for recurrent cancer", section on 'Relief of bowel obstruction'.)
●Bypass – If the tumor that causes the bowel obstruction cannot be resected but there is healthy bowel both proximal and distal to the site of obstruction, a side-to-side intestinal bypass can be performed to restore bowel continuity and allow the patient to resume oral nutrition for as long as possible.
Although bowel resection offered better overall survival and fewer complications than intestinal bypass [26], the outcomes were likely skewed by the burden and location of disease. Patients who can be treated with bowel resection are more likely to have minimal intra-abdominal diseases that have caused only one level of obstruction; patients who were treated with bypass (and/or diversion) were more likely to have peritoneal carcinomatosis or extensive intra-abdominal disease with multiple levels of obstruction, which likely contributed to their poor prognosis and higher rates of mortality and morbidity [40]. Nevertheless, in a series of 55 patients, the majority of patients were discharged to home (85.5 percent) and were tolerating an oral diet (74.6 percent) after intestinal bypass for malignant obstruction [52].
●Ileostomy – When patients have an unresectable small bowel obstruction, an ileostomy can be created from the most distal unaffected bowel segment to relieve the obstruction. The surgeon must maintain a minimum of 100 cm of small intestine to permit sufficient nutrient absorption. Additionally, a proximal ileostomy may be associated with high outputs, which may result in dehydration and electrolyte imbalance. (See "Ileostomy or colostomy care and complications", section on 'Ileostomy patients'.)
●Adhesiolysis – Cancer patients can also have bowel obstruction due to benign etiologies (eg, adhesions), especially if they have had previous abdominal surgery. In such patients, obstruction can be relieved by adhesiolysis without any bowel resection, bypass, or diversion, just as in noncancer patients. Given that no imaging study can diagnose adhesive bowel obstruction with certainty, the decision to proceed with adhesiolysis is generally an intraoperative decision. (See 'Postsurgical adhesion' above.)
●Colostomy or colonic stenting as a bridge to surgery – For patients who present acutely with a cancer-related colorectal obstruction, a temporary diverting colostomy can serve as a bridge to definitive surgery [53]. Alternatively, self-expanding metallic stents can also temporarily relieve bowel obstruction and permit definitive surgery after the patient has been optimized [54]. In a systematic review of 14 retrospective studies of 2873 patients with right-sided malignant colonic obstruction, primary resection was associated with higher mortality and major morbidity rates compared with stent placement followed by elective resection [55]. (See "Enteral stents for the management of malignant colorectal obstruction", section on 'Stenting as a bridge to surgery' and "Large bowel obstruction", section on 'Malignant obstruction'.)
Patients who are not surgical candidates — Options for patients who are not candidates for surgery include placement of a self-expanding colorectal stent, decompression using a nasogastric or endoscopically placed gastrostomy tube, and medical management with intravenous fluids, antiemetics, steroids, anticholinergics, and somatostatin analogs.
Self-expanding stents for malignant colorectal obstruction — The indications for stent placement in patients with malignant colonic obstruction include (see "Enteral stents for the management of malignant colorectal obstruction", section on 'Indications'):
●Palliation of surgically incurable colorectal cancer.
●Stenting may allow for optimization of medical status in patients who might eventually be surgical candidates.
●Management of some patients with extracolonic pelvic tumors (eg, ovarian cancer) that are causing extrinsic compression.
For patients with MBO who are inoperable or unwilling to undergo palliative resection, placement of a self-expanding stent is an alternative approach that is successful in over 90 percent in the short term. Long-term success rates are less favorable, and as many as one-third may need reintervention at some point [56]. Stenting is predominantly used for left-sided lesions. Although there have been reports that stenting of right-sided disease is often unsuccessful, specialized centers report reasonable success rates. (See "Large bowel obstruction", section on 'Malignant obstruction'.)
There is less consensus as to the benefit of stenting over palliative surgery for patients who are potential surgical candidates. A meta-analysis comparing palliative surgery versus stent placement for MBO in patients with advanced gynecologic or gastrointestinal cancer found that stents were less effective at relieving obstruction and did not improve overall survival; however, stented patients had lower 30-day mortality and a shorter interval to initiating chemotherapy [57].
Complications of stenting include bowel perforation, stent migration, and recurrent bowel obstruction [58-61]. (See "Enteral stents for the management of malignant colorectal obstruction", section on 'Adverse events'.)
Absolute contraindications to colorectal stenting include existing perforation, very distal rectal lesions within 5 cm of the dentate line, disseminated peritoneal carcinomatosis, and multifocal enteral stenotic segments. Relative contraindications are represented by anatomical difficulties such as a long stricture segment or strictures positioned in tortuous colorectal segments and bowel ischemia. High perforation rates have been reported in patients receiving the antiangiogenic agent bevacizumab as a component of chemotherapy, and colonic stenting should be avoided in patients who are or who will be receiving antiangiogenic agents. (See "Enteral stents for the management of malignant colorectal obstruction", section on 'Perforation'.)
Enteric tube decompression — Tube decompression can alleviate symptoms of MBO but the complication rate is high, these procedures may provide only incomplete symptom relief, and the ongoing presence of these tubes can be uncomfortable and distressing for patients and their caregivers. Pharmacologic management is generally preferred, at least initially. (See 'Pharmacologic management' below.)
Tube decompression can successfully alleviate symptoms of MBO [62-64]. There are several types of decompressive tubes, which all have advantages and disadvantages:
●Nasogastric tube – Temporary decompression via the nasogastric tube can evacuate a large amount of secretions and reduce nausea, vomiting, and pain. However, because the nasogastric tube often becomes occluded and requires flushing and/or replacement, it is not a realistic long-term solution. During long-term drainage, the tube interferes with coughing to clear pulmonary secretions and may be associated with nasal cartilage erosion, otitis media, aspiration pneumonia, esophagitis, and bleeding. This treatment can also create considerable discomfort. Thus, nasogastric tubes should be used only as a temporizing measure and should be removed as soon as feasible.
●Gastrostomy tubes – There are several types of gastrostomy tubes. (See "Gastrostomy tubes: Uses, patient selection, and efficacy in adults" and "Gastrostomy tubes: Placement and routine care" and "Gastrostomy tubes: Complications and their management".)
Percutaneous approaches are generally less expensive and saves time compared with surgical approaches, and they are preferred, unless endoscopic or radiologic placement is not possible. Nonsurgical decompression of gastric contents is most often achieved through placement of a percutaneous gastrostomy tube, which is referred to as a "venting percutaneous endoscopic gastrostomy (PEG)" [62,65-67]. For those who cannot tolerate a standard PEG tube placement (eg, patients with tense ascites), percutaneous trans-esophageal gastrostomy (PTEG) is another alternative to long-term nasogastric tube drainage [68]. PTEGs are decompressive catheters that can be percutaneously inserted through the cervical esophagus with an extension into the stomach. However, PTEGs are not used widely outside of Japan [69].
Regardless of insertion techniques, enteral tube drainage may be recommended if pharmacologic management is unsuccessful. However, the decision to proceed with venting gastrostomy rather than pharmacologic management may also be influenced by patient preferences (eg, desire to taste liquids), the availability and cost of drugs (eg, expense of octreotide in United States hospices that utilize the Medicare capitation system of reimbursement), or unacceptable side effects (delirium and dry mouth with anticholinergics). (See 'Pharmacologic management' below.)
Enteral tube drainage versus pharmacologic management — There are no trials directly comparing pharmacologic management with enteral tube drainage for symptom management in patients with a MBO. In one retrospective study of 2001 to 2012 linked Surveillance, Epidemiology, and End Results (SEER)-Medicare data from patients 65 years or older with stage IV ovarian or pancreatic cancer who were hospitalized for MBO, median survival was 76 days (interquartile range 26 to 319 days) [70]. Survival was shorter after venting gastrostomy tube (38 days [interquartile range 23 to 69]) than with medical management (72 days [23 to 312 days]) or surgery (128 days [42 to 483 days]). However, as compared with medical management, patients treated with venting gastrostomy tube had fewer readmissions (subdistribution HR 0.41 [0.29 to 0.58]), had increased hospice enrollment (1.65 [1.42 to 1.91]), required less intensive care unit (ICU) care (0.69 [0.52 to 0.93]), and had fewer in-hospital deaths (0.47 [0.36 to 0.63]). Surgery was also associated with fewer readmissions (0.69 [0.59 to 0.80]) and decreased hospice enrollment (0.84 [0.76 to 0.92]), but a higher likelihood of ICU care (1.38 [1.17 to 1.64]).
More invasive initial management might result in lower rates of hospitalization. However, these findings should be interpreted with caution, since it is possible that aggressive procedures were not done in patients too ill to receive them, and medical management might simply be a marker for more advanced disease. Well-powered randomized trials are needed.
Pharmacologic management — All patients with inoperable MBO who are not candidates for a stent or enteric tube decompression should undergo treatment with antisecretory drugs. For most patients, we suggest octreotide rather than an anticholinergic agent. Although published randomized controlled trials for the treatment of nausea and vomiting in palliative care patients with MBO are lacking, intravenous haloperidol is considered an effective antiemetic in this setting. Other options include metoclopramide or olanzapine. However, prokinetic agents should be avoided if there is complete mechanical obstruction or colicky abdominal pain. Glucocorticoids may be added to antisecretory therapy with an antiemetic for short-term use, providing a synergic effect, given the different mechanism of action. Anticholinergic agents such as scopolamine (hyoscine) butylbromide may be added in individual situations, particularly in patients with high level of obstruction or to limit colicky abdominal pain.
Pharmacologic management of inoperable patients focuses on adequate control of pain (opioids), diminishing intraluminal secretions and peristaltic movements (anticholinergic agents and octreotide), and/or lessening peritumoral edema (glucocorticoids). Pharmacologic regimens, which include antiemetics, antisecretory agents, glucocorticoids, and analgesics, provide symptom relief in 60 to 80 percent of patients with MBO [71].
Efficacy of individual treatments — The specific benefits of pharmacologic management for MBO, and dosing recommendations are provided in the following paragraphs:
●Somatostatin analogs – Somatostatin analogs such as octreotide have been increasingly used to relieve distressing gastrointestinal symptoms in patients with advanced cancer and MBO. Somatostatin analogs inhibit gastric, pancreatic, and intestinal secretions, reduce gastrointestinal motility, and may relieve pain and other symptoms of bowel obstruction. A systematic review [72] noted that at least four small randomized trials of high or unclear risk of bias [73-76] concluded that octreotide is superior (in reducing nausea and decreasing secretions more rapidly) to an anticholinergic agent alone (typically scopolamine [hyoscine]) for symptomatic management of MBO. In one of the trials, conducted in 97 patients with MBO from ovarian cancer, octreotide significantly reduced the amount of gastrointestinal secretions compared with scopolamine (hyoscine) butylbromide [76].
On the other hand, two adequately powered multicenter randomized trials with a low risk of bias reported no significant difference between a long-acting somatostatin analog versus placebo:
•A benefit for octreotide in controlling vomiting could not be shown in a small double-blind trial in which 87 patients with advanced cancer presenting with vomiting and inoperable bowel obstruction were randomly assigned to placebo or octreotide (0.6 mg per 24 hours by infusion); both groups also received supportive therapy with ranitidine (an H2 histamine receptor blocker that has been withdrawn from the United States market), parenteral hydration, and dexamethasone [77]. The primary outcome, patient-reported days free of vomiting at 72 hours, was not significantly different with octreotide (1.87 versus 1.69 days), nor was the number of patients free of vomiting at 72 hours (38 versus 33 percent), or health-related quality of life, in a later analysis [78]. One potential reason for the apparent lack of benefit from octreotide is the short endpoint studied. Although many patients have symptom improvement within the first four hours, the full benefit may not be seen for several days [79,80]. Patients receiving octreotide had a significantly smaller number of vomiting episodes. Although they were also twice as likely to have received scopolamine (hyoscine) butylbromide, no specific protocol was given to administer this drug, not necessarily reflecting higher rates of colicky abdominal pain due to octreotide.
•A second negative study involved a single injection of long acting lanreotide (which is administered once monthly), the trial only lasted seven days, and the dropout rate was 39 percent of 80 patients [81].
The only conclusions that can be gathered from these trials are that the anticholinergic secretory drugs are inferior to octreotide for first-line treatment of nausea and vomiting related to MBO.
The usual starting dose is 0.1 mg (100 micrograms) two to three times daily given subcutaneously. Rapid dose titration may be needed, and doses several times the starting dose, and sometimes significantly higher, are needed by some patients. Side effects are rarely a problem, although cost may be a significant issue, especially in United States hospices due to the Medicare capitation system of reimbursement.
Patients who respond could be given a depot injection of long-acting octreotide (Sandostatin LAR) monthly for maintenance therapy. Drug levels are not reliably therapeutic for two weeks. The benefit of long-acting octreotide in conjunction with glucocorticoids was shown in a phase II multicenter double-blind study in which 64 patients with inoperable MBO were randomly assigned to glucocorticoid therapy plus either long-acting octreotide (eg, Sandostatin LAR; in combination with short-acting octreotide for the first six days) or placebo [82]. The primary endpoint, "success at day 14," was defined as the absence of a nasogastric tube, vomiting less than twice daily, and no need for anticholinergic agents. More patients receiving the octreotide arm were successfully treated by day 14 (38 versus 28 percent).
●Antiemetics – The efficacy of individual antiemetic drugs for management of nausea and vomiting in MBO has not been systematically reviewed. Appropriate alternatives include haloperidol, metoclopramide (if there is no complete bowel obstruction), and olanzapine.
•Haloperidol – Haloperidol is an effective antiemetic in the setting of MBO but has not been compared with other drugs in randomized trials specifically in this setting. Although published randomized controlled trials of haloperidol for the treatment of nausea and vomiting in palliative care patients with or without MBO are lacking [83], intravenous (IV) administration of haloperidol, a selective dopamine (D2) receptor antagonist, is an effective antiemetic for patients with MBO. Some guidelines consider it to be a first choice antiemetic, especially in the setting of a complete MBO [84-86]. (See 'Recommendations from expert groups' below.)
Appropriate initial doses are 0.5 to 2 mg IV or subcutaneous every six to eight hours or a continuous subcutaneous infusion at 1 to 5 mg per 24 hours.
•Metoclopramide – It is accepted practice that prokinetic agents (such as metoclopramide) are effective antiemetics for patients with advanced cancer and refractory nausea and vomiting from a variety of causes, including MBO [86,87]. However, the available data demonstrating benefit for nausea, vomiting, and improved intestinal transit are limited to mainly small retrospective series [88-91]. A small randomized trial of 16 patients with partial MBO and nausea who were randomly assigned to olanzapine (5 mg daily) or metoclopramide (20 to 30 mg daily) for three days concluded that both drugs diminished the mean daily nausea score with no significant difference between the two [92].
Metoclopramide should only be used for patients with partial obstructions (intermittent rather than continuous nausea and vomiting, still some passage of flatus or stool, or ileus) and should be avoided if there is complete mechanical obstruction or colicky pain [86,93-95]. (See 'Additional maneuvers for partial bowel obstruction' below.)
•Olanzapine – Olanzapine, an antipsychotic medication that has affinity for multiple neurotransmitters that may play a role in modulating nausea and vomiting [96], is an alternative to metoclopramide [86]. The following data are available:
-A retrospective study of 20 patients found that olanzapine decreased the severity of nausea and the frequency of emesis in 90 percent of patients with advanced cancer and incomplete bowel obstruction [97].
-As noted above, a small randomized trial of 16 patients with partial MBO and nausea concluded that both olanzapine and metoclopramide diminished the mean daily nausea score with no significant difference between the two [92].
Olanzapine can be given orally or intramuscularly, and a preliminary trial suggests that subcutaneous administration is also possible [98].
●Histamine H1 antagonists and phenothiazines – Limited evidence supports benefit from these types of antiemetics for reducing nausea and vomiting in MBO [91,99,100].
●5-HT3 antagonists – The benefit of 5-hydroxytryptamine 3 (5-HT3) antagonists is not well studied in patients with MBO. One prospective trial of granisetron in conjunction with dexamethasone in 24 patients with nonsurgical intestinal obstruction suggests control of nausea and vomiting in 87 percent; however, the contribution of granisetron to these results is unclear [101]. Updated year 2022 Multinational Association for Supportive Care in Cancer (MASCC) guidelines for management of MBO stated that granisetron may reduce nausea and the frequency of vomiting in MBO, but emphasized the weakness of the evidence, and that further studies are needed to assess its use [86].
●Anticholinergics – Anticholinergic agents such as scopolamine (hyoscine) butylbromide may be added to somatostatin analogs in individual situations, particularly in patients with high level of obstruction or to limit colicky abdominal pain. As noted above, a 2016 systematic review of four randomized trials with a high risk of bias found that somatostatin analogs were more effective than hyoscine butylbromide in reducing nausea and vomiting associated with MBO [72].
Scopolamine (hyoscine) base can be administered by a transdermal patch, a convenient route for those with limited gastrointestinal absorption [73,74]. Outside of the United States, scopolamine (hyoscine) butylbromide is widely available and the preferred antisecretory drug for medical treatment of inoperable bowel obstruction when an anticholinergic is indicated. The dose usually prescribed is 60 mg/day as a continuous subcutaneous injection [102].
In some countries, however, scopolamine is available only as hydrobromide salt (scopolamine [hyoscine] hydrobromide) for subcutaneous injection. Scopolamine (hyoscine) hydrobromide is not equivalent to scopolamine butylbromide on a mg per mg basis and may not be interchanged (table 5). Where available, scopolamine (hyoscine) butylbromide is preferred because the hydrobromide salt crosses the blood-brain barrier and may produce central nervous system side effects such as somnolence and confusion.
Glycopyrrolate has a pharmacologic profile that is similar to scopolamine, but there is minimal penetration of the blood-brain barrier. When an anticholinergic is chosen, it is another option and may be preferred in patients who are predisposed to somnolence and mental confusion. Although this drug has never been systematically evaluated as a treatment for symptoms of bowel obstruction, anecdotal evidence supports benefit [103]. The usual dose is 0.1 to 0.2 mg intravenously or subcutaneously once or twice daily.
●Glucocorticoids – While the use of glucocorticoids for MBO is rational, the data to support efficacy are relatively weak. A systematic review of three unpublished randomized placebo-controlled trials and seven uncontrolled (prospective and retrospective) studies concluded that there was a nonstatistically significant trend for resolution of bowel obstruction within 10 days of diagnosis using IV glucocorticoids (daily dose range of 6 to 16 mg dexamethasone) and that the incidence of side effects was low [104]. Notably, resolution of nausea and vomiting were secondary outcomes, and no data were available on improvement or lack of improvement in symptoms.
Glucocorticoids could be added to antisecretory therapy with an antiemetic, providing a synergic effect, given the different mechanism of action. A starting dose of 4 mg twice daily is reasonable.
Additional maneuvers for partial bowel obstruction — For patients with early and incomplete bowel obstruction, a combination of propulsive and antisecretory agents might act synergistically to allow fast recovery of bowel transit without inducing unpleasant colic when an intensive pharmacological approach is used [105]. This treatment might ameliorate gastrointestinal symptoms and reverse MBO when using a combination of drugs acting with different mechanisms, including metoclopramide, octreotide, dexamethasone, with or without an initial bolus of an oral contrast agent such as diatrizoate meglumine-diatrizoate sodium water-soluble radiocontrast media (Gastrografin). We typically use metoclopramide 30 to 40 mg per day (5 to 10 mg every six to eight hours) in combination with dexamethasone 12 mg per day, octreotide 0.3 mg per day, and a one-time oral bolus of Gastrografin 25 mL. For refractory cases, the dose of metoclopramide can be increased to 10 mg every four hours or 20 mg every six hours, as needed and tolerated.
With this active intervention, recovery of intestinal transit may be observed in some patients. This finding suggests that the most important mechanism in these circumstances is functional and can be reversible if an aggressive treatment is initiated early, before fecal impaction and edema render bowel obstruction irreversible [106].
Recommendations from expert groups — Published guidelines from the National Comprehensive Cancer Network (NCCN) [107] and the European Association for Palliative Care (EAPC) [85] for the medical management of MBO are similar, recommending opioids, glucocorticoids, antisecretory drugs, and antiemetics, but in no specific order; enteral tube drainage is recommended if drugs are not successful.
A 2022 systematic review by MASCC came to the following conclusions regarding antiemetics in patients with MBO [86]:
●The benefit of anticholinergics (hyoscine butylbromide) may be inferior to octreotide to reduce vomiting in MBO.
●Haloperidol may be an effective antiemetic in MBO, particularly for complete MBO.
●Dopamine antagonist prokinetic drugs (eg, metoclopramide, domperidone) may be effective for management of nausea, vomiting, and for restoring bowel transit time in partial MBO; due to the potential risk of a bowel perforation, these should likely be avoided in complete MBO.
●Histamine H1 antagonists (eg, dimenhydrinate, cyclizine) may be an effective antiemetic in complete MBO.
●Phenothiazines (eg, chlorpromazine) may reduce nausea and vomiting in MBO.
●Granisetron may reduce nausea and the frequency of vomiting in MBO.
●Octreotide may reduce vomiting in MBO.
●Thienobenzodiazepine antipsychotics (eg, olanzapine) may reduce nausea and vomiting in MBO.
Role of nutritional support — There is little, if any, justification for the use of parenteral nutritional support in patients with advanced cancer who have MBO, but are terminally ill with estimated life span that is measured in weeks to a few months. An analysis of survival, quality of life, and cost effectiveness of home parenteral nutrition in patients with inoperable bowel obstruction revealed a short survival and high associated costs [25,108]. We recommend against the use of parenteral nutrition for most patients in this setting. This subject is discussed in detail elsewhere. (See "The role of parenteral and enteral/oral nutritional support in patients with cancer", section on 'The perioperative setting'.)
Perioperative parenteral nutrition may be a reasonable short-term option for selected malnourished patients with MBO who are undergoing surgery to enable subsequent chemotherapy and have a postoperative survival that is likely to be more than three months [26,109,110]. This is most likely in women who have advanced ovarian cancer. (See "Cancer of the ovary, fallopian tube, and peritoneum: Surgical cytoreduction", section on 'Bowel resection'.)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Colorectal cancer" and "Society guideline links: Bowel obstruction" and "Society guideline links: Palliative care".)
SUMMARY AND RECOMMENDATIONS
●Etiologies – Malignant bowel obstruction (MBO) is common in patients with abdominal and pelvic tumors; it is most prevalent in ovarian, colorectal, and gastric cancer. (See 'Epidemiology' above.)
Distinguishing benign from MBO can be challenging but has profound impacts on management decisions and prognosis (table 1). Most small bowel obstructions are due to benign causes (most often postsurgical adhesions), whereas malignant causes predominate in large bowel obstructions. (See 'Etiology and pathophysiology' above.)
●Clinical presentation – Symptoms of intestinal obstruction typically include crampy, paroxysmal abdominal pain, loss of appetite, nausea, vomiting, inability to have a bowel movement or pass flatus, and abdominal bloating. Physical findings include a distended, tender abdomen with or without a palpable mass. Clinical presentation varies by the level of obstruction (table 2). (See 'Clinical features' above.)
●Management – An algorithmic approach to management of MBO in palliative care patients is provided (algorithm 1). The following reflects our approach to patients with a known or suspected cancer who present with signs/symptoms consistent with bowel obstruction:
•Diagnostic workup – A diagnostic workup should be performed, if consistent with the goals of care, to establish the diagnosis, the level of obstruction and tumor burden; findings may influence the goals of care. (See 'Diagnostic evaluation' above.)
Plain abdominal radiography is the initial radiographic study of choice. Contrast-enhanced computed tomography (CT) scan provides additional information which may change management. (See 'Radiographic studies' above.)
•Indications for surgery – Patients who present with generalized peritonitis, clinical deterioration, pneumoperitoneum, or imaging signs of ischemia should be offered urgent surgical exploration if it is consistent with their goals of care. The remainder should be initially managed nonoperatively with nasogastric decompression of the stomach/intestine and intravenous fluids. If symptoms are not relieved within 48 to 72 hours of nonoperative management, definitive surgical intervention is indicated if they are deemed suitable surgical candidates. (See 'Initial management' above.)
Management is individualized based on risks and potential benefits of surgery, the clinical condition, estimated life expectancy, performance status, comorbidities, and patient preferences and treatment goals. Patients should be considered potential surgical candidates unless they are critically ill, had a poor performance status prior to presenting with the MBO (table 3), or refuse surgery. After a thorough surgical evaluation, patients who are potential surgical candidates may be deemed unlikely to benefit from surgery for anatomic or clinical reasons. Our absolute and relative contraindications to surgery in patients with MBO are outlined in the table (table 4). (See 'Surgical candidates' above.)
If surgery is chosen, the choice of technique depends on anatomy, comorbidities, risk factors, and cancer prognosis. Surgical techniques commonly performed to relieve bowel obstruction include bowel resection, bypass, diversion, stenting, and lysis of adhesions. (See 'Surgical techniques' above.)
•Patients who are not surgical candidates – Options for patients with inoperable MBO or who are not candidates for surgery include placement of a self-expanding colorectal stent, decompression using an endoscopically placed gastrostomy tube, and medical management with antiemetics, somatostatin analogs, glucocorticoids, and anticholinergic agents.
-Decompression of gastric contents can be achieved through placement of a percutaneous gastrostomy tube. However, pharmacologic management is generally preferred for most patients, at least initially. (See 'Enteric tube decompression' above.)
-All patients with inoperable MBO who are not candidates for a stent should undergo treatment with antisecretory drugs. For most patients, we suggest octreotide rather than an anticholinergic agent (Grade 2B). The usual starting dose is 0.1 mg (100 micrograms) twice daily subcutaneously. Rapid dose titration may be needed, and doses several times the starting dose are needed by some patients. (See 'Efficacy of individual treatments' above.)
Although published randomized controlled trials for the treatment of nausea and vomiting in palliative care patients with MBO are lacking, intravenous haloperidol is considered an effective antiemetic in this setting. Other options include metoclopramide or olanzapine. However, prokinetic agents should be avoided if there is complete mechanical obstruction or colicky abdominal pain.
Short-term glucocorticoids may be added to antisecretory therapy with an antiemetic, providing a synergic effect, given the different mechanism of action.
Anticholinergic agents such as scopolamine (hyoscine) butylbromide may be added in individual situations, particularly in patients with high level of obstruction or to limit colicky abdominal pain.
For patients with early and incomplete MBO, a combination of propulsive and antisecretory agents might act synergistically to allow fast recovery of bowel transit without inducing unpleasant colic when an intensive pharmacologic approach is used. (See 'Additional maneuvers for partial bowel obstruction' above.)
-For patients with advanced cancer who are terminally ill and have an estimated life span of weeks to a few months, we recommend against parenteral nutritional support (Grade 1B). (See 'Role of nutritional support' above.)
