Current Management of Small Bowel Obstruction

Although common, small bowel obstruction (SBO) remains one of the most challenging clinical problems treated by surgeons. Responsible for up to 300,000 hospital admissions every year in North America, SBO arises from multiple etiologies and manifests as a diverse panoply of clinical presentations [1]. Initial evaluation should center on differentiating those patients who need urgent exploration from those who may undergo a safe, nonoperative trial. The wide range of etiologies, however, combined with specific, and often unique, patient parameters, renders this decision difficult. Traditionally, the decision between urgent operative intervention and initial nonoperative management has hinged on the distinction between complete and partial obstruction. However, the clinical diagnosis of complete obstruction is imprecise, and the complete/partial dichotomy has not eliminated avoidable obstruction-associated ischemia and necrosis. Rather than trying to predict those patients at risk for ischemic complications, we may do better to define clinical parameters that predict failure of nonoperative management and offer prompt operation to patients demonstrating these parameters. Hopefully, such an approach, codified into practice management guidelines, will minimize both ischemia and hospital length of stay associated with SBO. After reviewing the pathogenesis and pathophysiology of SBO, this article outlines newly developed and refined management and surgical techniques to reach these goals.

Pathogenesis

SBO implies compromised luminal patency and as such is differentiated from the functional abnormalities of ileus and dysmotility disorders. Adhesive disease, neoplasia, and hernias are the 3 most common causes in the western world, collectively accounting for 80% of all obstructions [2,3]. In developing countries, hernias, intussusceptions, and volvulae predominate [4].

SBO most frequently arises from adhesive disease (49%), which in turn is almost always related to prior operation (Fig. 1) [2]. Only a minority of patients with adhesions will experience SBO. Although 93.0% of patients with more than one abdominal operation will have adhesions at autopsy, only 4.6% of patients with a prior abdominal procedure will have an adhesive SBO [5,6]. Adhesive SBO is frequently a relapsing disease: up to 30% of patients who undergo celiotomy for obstruction will require reexploration for recurrence [2,7].

Fig. 1 Adhesive small bowel obstruction.

Barmparas and colleagues [5] reviewed the English literature and collected more than 440,000 reported patients with postceliotomy adhesive SBO to examine risk factors. The likelihood of subsequent SBO varied among different index operative procedures, and was greatest after open adnexal operations (23.9%), followed by ileal pouch anal-anastomosis (19.3%), and open total abdominal hysterectomy (9.5%). Laparoscopic cholecystectomy (0.2% vs 7.1%), laparoscopic total abdominal hysterectomy (0.0% vs 15.6%), and laparoscopic adnexal operations (0.0% vs 23.9%), but not laparoscopic appendectomies (1.3% vs 1.4%), all resulted in fewer adhesive obstructions than their open counterparts. Overall, open procedures were associated with twice as many adhesive obstructions as was laparoscopy, but the investigators cautioned that interpretation of this collective comparison was hampered by limited follow-up, heterogeneity between the primary studies, and selection biases. Indeed, early review of the Conventional versus Laparoscopic-Assisted Surgery In Colorectal Cancer CLASICC (CLASICC) trial data demonstrated no statistical difference between adhesive obstructions after laparoscopic (2.5%) versus open resections (3.1%) for colorectal cancer at 3-year follow-up [8].

Neoplasia, the second leading cause of SBO, is responsible for 16% of SBO admissions (Fig. 2) [2]. Although primary small bowel tumors do cause obstruction, colorectal and ovarian metastases are the most common malignant etiologies (41% and 28%, respectively) [9]. Breast cancer and malignant melanoma are the 2 most common nonabdominal tumors that can cause obstruction [10]. Neoplastic masses causing obstruction may be a result of the primary tumor, peritoneal metastases, or bulky lymph node metastases. The mechanism of obstruction may be direct compression, malignant adhesion with consequent torsion, knuckling, or internal herniation. The median time from the diagnosis of cancer to the first episode of obstruction is about 1 year, but decades can pass from the time of initial diagnosis to obstruction. Median survival is on the order of 3 to 6 months after onset of initial obstructive symptoms and is almost universally fatal [9,10].

Fig. 2 Malignant small bowel obstruction.

Hernias cause 15% of SBOs [2]. There are 2 main categories of hernia: external and internal. Among external hernias, incarceration is encountered most frequently from inguinal and incisional hernias, but is also seen with femoral, umbilical, traumatic, and peristomal hernias (Fig. 3). Whereas external hernias are diagnosed on physical examination, internal hernias are generally diagnosed in the operating room, and on occasion, by preoperative computed tomography (CT). Internal hernias can be further classified into congenital or acquired types. For example, an obturator hernia, resulting from protrusion of intra-abdominal contents through the obturator foramen created by the pubic bones and ischium, is considered a congenital internal hernia (Fig. 4) [11]. Obturator hernias, generally present in elderly women after significant weight loss, are associated with a 25% mortality owing to the difficulty in diagnosis and the comorbidities of the patient population [12]. Omental and paraduodenal defects, as well as the foramen of Winslow are additional, but uncommon, sites of congenital herniation [13].

Fig. 3 Incarcerated inguinal hernia.

Fig. 4 Incarcerated left obturator hernia.

Acquired internal hernias arise after surgical manipulation. With the rapid increase in the rate of laparoscopic bariatric procedures, the rate of internal hernia is rising [14]. Surgeons must be knowledgeable in the diagnosis and treatment of the specific hernias that can result from Roux-En-Y Gastric Bypass including those through mesocolic, Petersen, and mesomesenteric defects (Fig. 5) [13,14]. The incidence has been reported to be as high as 3.1%, but this is likely declining with greater awareness of the need to close the defects at the primary operation [15]. The mesocolic defect is created for the retrocolic passage of the Roux limb. Petersen defect is bordered by the transverse mesocolon superiorly, the Roux limb and its mesentery anteriorly, and the proximal-most jejunum and its mesentery posteriorly [16]. Mesomesenteric defects are created by inadequate apposition of mesentery after bowel resection with anastomosis; after gastric bypass, a large such defect exists between the mesentery of the biliopancreatic limb anteriorly and the mesentery of the alimentary limb/common channel posteriorly. Symptomatic hernias can occur at any time after bypass, especially as the patients lose weight, allowing for widening of defects not closed properly during the initial procedure. As an additional consideration in patients with gastric bypass, a large remnant stomach on CT after gastric bypass must be recognized as an abnormal finding indicative of biliopancreatic limb obstruction.

Fig. 5 Internal hernia defects: (A) mesocolic, (B) Petersen, (C) mesomesenteric.

(From Kendrick ML, Dakin GF. Surgical approaches to obesity. Mayo Clin Proc 2006;81(10 Suppl):S18–24; with permission.)

The risk of strangulation with any symptomatic hernia, internal or external, is significant and has been reported to be 28%, as they tend to be closed loop obstructions; consequently, symptomatic hernias should be repaired [7,17]. Asymptomatic and minimally symptomatic hernias, on the other hand, are safe to observe. In the largest randomized, clinical trial, Fitzgibbons and colleagues [18] studied symptomatic and minimally symptomatic hernias, defined as easily reducible hernias without pain during limited physical activity. Patients in the observation arm had similar symptoms at follow-up as those patients undergoing repair. Of those patients who underwent watchful waiting, the risk of incarceration was minimal (0.0018 hernia-related events per patient-year). Thirty-one percent of those initially observed underwent operative repair for symptom progression within 3.2 years of median follow-up. This study was performed in white men older than 40, so its applicability to other ethnic groups, younger patients, and women is unknown.

The remaining 20% of SBOs have multiple, but much less common, causes. The inflammatory changes of Crohn disease may result in acute, subacute, or chronic bowel obstructions. Although most patients will require surgical resection at some point in their disease process, the aim is to avoid intervention until absolutely necessary and, therefore, these patients are managed primarily with medical treatment [19]. It is imperative, however, to ensure there is no evidence for perforation. Gallstone ileus results from a fistulous connection from the gallbladder to the gastrointestinal tract with passage of a large gallstone that obstructs the bowel lumen [20]. Remaining causes include radiation enteritis, bezoar/foreign body, volvulus, hematomas, and abscesses (Fig. 6).

Fig. 6 Small bowel obstruction resulting from volvulus.

It should be noted that there can be multiple causes of obstruction within the same patient. For instance, patients with colorectal cancer who undergo colectomy and have postoperative radiation therapy for recurrence may have any combination of adhesive disease, hernia, malignancy, and radiation enteritis causing their obstruction. In these patients, diagnosis of the specific etiology is difficult and will usually become apparent only upon operative exploration.

Pathophysiology

A fundamental understanding of the pathophysiology of SBO and strangulation obstruction is necessary for an understanding of the changes in the peritoneal cavity, bowel wall, and mesentery, which are closely associated with the patient’s clinical course and CT findings. Regardless of the cause of obstruction, the pathophysiologic effects of bowel obstruction are similar; the symptoms result from a blockage of normal intestinal transit. As luminal fluids pool proximal to the blockage, the patient may experience nausea and vomiting. Vague, diffuse abdominal pain may result from persistent peristalsis both above and below the obstruction. The patient may continue to have bowel function in the form of flatus and stool early in the process owing to distal peristalsis, even with complete obstructions. Abdominal distention results from increased intestinal volume. As the obstruction progresses, the bowel wall is stimulated to secrete fluid and electrolytes and may result in postobstructive diarrhea [21,22]. If not decompressed, a positive feedback loop of distension and secretion occurs. As the pressure approaches 30 mm Hg, the terminal lacteals become occluded, resulting in bowel wall lymphedema [23]. This causes the luminal pressure to increase further, which, if left untreated, will lead to venous outflow obstruction at the postcapillary venules at pressures greater than 50 mm Hg. The increased venous hydrostatic pressure leads to alterations in Starling forces, causing substantially increased filtration across the capillary bed into the bowel lumen, further exacerbating intravascular volume depletion. Progressive volume depletion will ultimately result in dehydration, hypotension, tachycardia, and metabolic acidoses. Further unchecked, venous hypertension will lead to microvascular or macrovascular arterial compression and, consequently, bowel wall necrosis (Fig. 7). Sepsis from either bacterial translocation through the compromised gut-mucosal barrier or free perforation may result, which, if left untreated, is fatal.

Fig. 7 Strangulation obstruction with bowel necrosis.

Diagnosis and workup

The approach to SBO should be systematic and should focus on (1) securing and confirming the diagnosis, (2) identifying the etiology, and (3) determining likelihood of strangulation. Each step is aided by a thorough history and physical examination, pertinent laboratory evaluation, and judicious radiographic imaging. The information gained will optimize eventual surgical decision making.

History and physical examination

The diagnosis of SBO should be made primarily on history and physical examination. Signs and symptoms include colicky abdominal pain, nausea and vomiting, obstipation, and abdominal distension. The patient with partial obstruction may describe borborygmi, often quite graphically and with the corroboration of a companion. The patient’s history will usually point to a specific cause. The vast majority of patients with adhesive obstructions will have had a prior abdominal operation. Suspicion of malignancy is heightened by an unoperated abdomen in the absence of radiation or inflammatory bowel disease, as well as unintentional weight loss and/or night sweats. Despite the rarity of gallstone ileus, intermittent chronic right upper quadrant pain in an elderly patient with obstructive symptoms should call this entity to mind.

The consulting surgeon must take heed of abnormal vital signs and perform a thorough physical examination. Orthostatic changes in heart rate and blood pressure indicate early volume depletion, whereas more severe hypovolemia will manifest as frank hypotension and tachycardia. Sepsis secondary to advanced bowel ischemia or perforation will amplify hypotension and tachycardia, and patients with neglected SBO may present with combined hypovolemic and septic shock.

Typical abdominal physical findings of SBO include diffuse abdominal tenderness, distension, and tympany. The abdominal tenderness of uncomplicated SBO is mild; more severe tenderness follows development of bowel ischemia culminating in the exquisite percussion tenderness seen with necrotic bowel. The degree of abdominal distension increases with more distal levels of obstruction and will be minimal for proximal jejunal obstructions. Incisional scars indicate previous operations, which may not have been reported. Palpation of incisional scars and the periumbilical, inguinal, and femoral regions may reveal the firm, tender bulge of an incarcerated hernia. The rare palpable mass suggests advanced malignancy or a long-standing inflammatory process. A metastasis in the pouch of Douglas may be manifest as a hard anterior mass on digital rectal examination, a “Blumer shelf.” Rushes and tinkles heard on auscultation of the abdomen add nothing to the signs and symptoms. Indeed, Dr Charles H. Mayo, speaking before the American Surgical Association in 1922, stated, “I should prefer to see a stomach tube rather than a stethoscope hanging around the neck of my surgical intern” [24].

Laboratory testing

Currently available laboratory tests do not contribute to the diagnosis of SBO, but they can confirm clinical suspicion of volume depletion, guide volume resuscitation, and assist recognition of bowel ischemia. Hypovolemia may manifest as azotemia with an elevated blood urea nitrogen to creatinine ratio. Patients with associated vomiting are at risk for hypochloremic, hypokalemic metabolic alkalosis. Leukocytosis has been implicated as a marker of strangulation, although a normal white blood cell count does not rule out ischemia [25]. Last, patients with severe volume depletion or sepsis from perforation or bacterial translocation will present with metabolic acidosis.

Imaging

Abdominal flat and upright radiographs, along with a chest radiograph, can confirm the diagnosis of SBO, but are accurate only in 50% to 60% of patients, and can be normal in 21% of patients with confirmed obstruction [2,26,27]. X-ray interpretation relies on the recognition of bowel gas patterns. As defined by plain radiograph, complete bowel obstruction comprises dilated small bowel (>3 cm), air-fluid levels, and the absence of colonic gas. Although this pattern is diagnostic of obstruction, other findings are possible. For example, the absence of bowel gas shadows is consistent with proximal and closed loop obstructions and/or obstructed bowel completely distended and exclusively distended with fluid.

CT findings of SBO include small bowel dilatation with air-fluid levels, collapsed distal bowel, and, at times, a distinct transition point. CT has replaced upper gastrointestinal series/small bowel follow-through and enteroclysis for confirmation of a clinical diagnosis of SBO [14,28,29]. Although CT has greater sensitivity, its main utility lies in its ability to diagnose the underlying cause of the obstruction [30–32]. Closed loop obstruction, internal hernia, tumor, intussusception, and volvulus all give distinctive CT images. General findings of SBO absent these specific findings suggest adhesive disease in the proper setting.

Beyond making a diagnosis, CT findings can guide the decision to pursue immediate exploration or nonoperative management. CT findings of free intraperitoneal fluid, mesenteric edema, pneumatosis, and portal venous gas correlate with small bowel ischemia [25,33]. As discussed later in this article, CT findings can predict need for operative intervention; further, the water-soluble oral contrast agents used have been shown to be therapeutic and decrease the need for operative intervention [25,33–39]. New techniques in multidetector-row CT with 3-dimensional reconstructing techniques may allow for even greater diagnostic ability. In a limited series, Hong and colleagues [29] were able to demonstrate an improved ability to locate the site of obstruction and diagnose its cause compared with conventional CT. As a final imaging consideration, magnetic resonance imaging (MRI) in the form of enterography and enteroclysis has been used successfully in small studies [40,41]. Issues of cost and accessibility severely limit the current usefulness of MRI as a front-line technique for imaging SBO, however.

Management

Resuscitation

The initial step in management of SBO consists of correcting the fluid and metabolic derangements that result from protracted vomiting and intraluminal fluid sequestration. Aggressive crystalloid resuscitation should be performed simultaneously with the diagnostic workup. The essential end points of resuscitation for noncomplicated SBO should be normal blood pressure and heart rate with a urinary output of 0.5 mL/kg/h monitored with an indwelling bladder catheter. Correction of electrolyte abnormalities should also be undertaken concurrently. In patients presenting with hypovolemic and/or septic shock, central venous pressure measurement may be a useful guide. All patients with SBO require nasogastric tube decompression regardless if operative or nonoperative management is undertaken. Nasogastric decompression improves mortality, ameliorates symptoms, minimizes aspiration, and provides access for radiographic contrast administration.

Deciding between operative or nonoperative management

The challenge of SBO management lies in deciding on the appropriate treatment plan. An optimal approach will (1) provide immediate operative exploration in patients with strangulation obstructions, (2) facilitate early recognition of patients without strangulation who will not resolve without intervention, and (3) minimize unnecessary operations.

Delay to operative exploration for strangulation obstruction will increase mortality and morbidity, including a greater frequency of bowel resection; hence the adage, “Never let the sun rise or set on a small bowel obstruction” [1,25,39,42,43]. Implicit in this quotation is an ability to promptly and precisely recognize the presence of strangulation. Unfortunately, this predictive ability has been elusive, as even the most experienced surgeons are correct only half of the time [17]. The clinical markers of strangulation are numerous, as demonstrated in Box 1 [25,42–44]. Patients who present with signs of intestinal ischemia comprise the minority of patients with SBO, but patients with signs of ischemia will have a strangulation obstruction up to 45% of the time [39]. Therefore, any patient with these features and a concurrent diagnosis of SBO has a strangulation obstruction until proven otherwise and should undergo emergent celiotomy after resuscitation.

Especially vexing are those patients who harbor a strangulated obstruction and fail to display any of the features outlined in Box 1 [25,39]. To combat this discrepancy, the presence of “complete” SBO was traditionally used as a tool to screen for patients with either silent strangulation or at risk of developing strangulation. Used in this way, the diagnosis of complete SBO implies the need for emergent/urgent operation; however, complete SBO is actually defined as the presence of dilated small bowel with air fluid levels in the absence of colonic gas on abdominal radiograph [45]. Further, the differentiation of complete versus partial SBO is fraught with some subjective elements: time to last flatus and the patient’s memory thereof, and absence versus paucity of colonic gas as visualized by plain radiograph versus CT. An additional difficulty with the complete versus partial paradigm follows from the finding that 31% to 43% of patients with complete SBO or peritonitis will not require bowel resection [39,42,43]. Given these concerns, the authors believe the traditional focus of differentiating complete from partial SBO should change to one of predicting failure of nonoperative management with the goal of exploring patients with predicted failure as soon as possible. This approach promises not only early exploration for patients with silent strangulation, but also shortened hospital stay for all. Two potential decision-aids are emerging to assist with such an approach: (1) response to water-soluble oral contrast meals and (2) predictive models.

Water-soluble oral contrast agents such as methylglucamine diatrizoate (Gastrografin) and sodium diatrizoate/meglumine diatrizoate (Urografin) are high-osmolar liquids used in CT and other gastrointestinal radiographic studies. Oral or nasogastric administration of these agents with subsequent imaging has been shown to both predict and reduce the need for operative intervention in patients presenting with SBO without signs of strangulation [35–38,46–50]. The methodological variability between studies has been significant and impedes interpretation of the general technique; different agents, different doses, and different durations from administration to assessment have been used. In the trials studying methylglucamine diatrizoate, a dose of 50 to 100 mL was administered, whereas the single sodium diatrizoate/meglumine diatrizoate trial used 40 mL. Failure of contrast to pass into the colon as detected by plain radiograph predicted need for operation, but this assessment was made at times varying from 4 to 72 hours. Several studies demonstrated an apparent therapeutic effect of the water-soluble agent: patients in the contrast arms underwent fewer operative explorations than those in the control groups [35–38,47]. This therapeutic effect was not demonstrated in all studies, however [48–50]. Both the predictive and therapeutic utility of water-soluble contrast administration is supported by a meta-analysis performed by Branco and colleagues [50]. The investigators showed that if upper gastrointestinal contrast reached the colon within 8 hours of administration, then 99% of patients would not require operation. On the other hand, 90% of those without contrast in the colon at 8 hours required operation. Further, patients receiving contrast had a reduced need for operation compared with controls (20% vs 29%, P = .007). Although 508 patients were included in the meta-analysis, the investigators caution that the primary studies’ significant methodological differences limit the certainty of the conclusions: only 3 of the 9 studies analyzed were randomized, controlled trials. Another limitation was the lack of data regarding missed strangulation obstructions. First studied in 1994, water-soluble contrast administration in SBO management has many proponents, but widespread acceptance awaits accrual of better data demonstrating efficacy.

An alternative, but also complementary, approach involves the development of predictive models based on combinations of readily available clinical parameters. The goal of a predictive model for SBO management is to identify, soon after surgical consultation, those patients who will ultimately require operative exploration. Two early studies attempted to develop predictive systems based on clinical, laboratory, and plain radiographic data; both initiatives failed [3,51]. Newer models incorporate data from state-of-the-art CT imaging. This modality provides remarkable images of the small bowel mesentery and bowel wall, demonstrating such signs as mesenteric edema, mesenteric swirling, and reduced bowel wall enhancement [52–55]. This technology rekindled interest in predicting the need for exploration, leading to recent development of 2 models attempting to predict the need for operative exploration in SBO [25,33,56]. The goal of these models is to identify those patients who will undergo operative exploration for SBO during their hospital stay in an attempt to reduce missed strangulation obstructions and minimize time to operation for all who will require it. An initial attempt by Jones and colleagues [56]

Related posts:

Glucose Elevations and Outcome in Critically Injured Trauma Patients Geographic Variation in Health Care and the Affluence-Poverty Nexus Recent Advances in the Diagnosis and Treatment of Gastrointestinal Carcinoids Immunotherapy for Metastatic Solid Cancers Influence of Surgical Volume on Operative Failures for Hyperparathyroidism Endovascular Approaches to Arteriovenous Fistula

Advances in Surgery - E-Book