Chapter 28: Small Intestine

1. The small intestine performs a diverse set of functions.

2. Small bowel obstruction is one of the most common surgical diagnoses.

3. Most cases of small bowel obstruction are due to adhesions from previous surgery and resolve with conservative management.

4. Tumors and malignancies of the small bowel are rare and difficult to diagnose.

5. Small bowel may be the source of gastrointestinal bleeding, which may be difficult to diagnose.

6. If following surgical resection, less than 200 cm of small bowel remains, patients are at risk of developing short bowel syndrome.

The small intestine is the raison d’être of the gastrointestinal tract as it is the principle site of nutrient digestion and absorption.¹ The small intestine is also the body’s largest reservoir of immunologically active and hormone-producing cells and hence can be conceptualized as the largest organ of the immune and endocrine systems, respectively. It achieves this diversity of action through unique anatomic features, which provide it with a massive surface area, a diversity of cell types, and a complex neural network to coordinate these functions.

Despite its size and importance, diseases of the small intestine are relatively infrequent and present diagnostic and therapeutic challenges. Treatments for common conditions such as postoperative ileus are hardly more effective than those used at the dawn of the last century. Mortality rates associated with acute mesenteric ischemia have not improved during the past 50 years.

Despite introduction of novel imaging techniques such as capsule endoscopy and double balloon endoscopy, diagnostic tests lack sufficient predictive power to definitively guide clinical decision making for individual patients. Furthermore, few high-quality, controlled data on the efficacy of surgical therapies for small bowel diseases are available.

Therefore, sound clinical judgment and a thorough understanding of anatomy, physiology, and pathophysiology remain essential to the care of patients with intestinal disorders.

The small intestine is a tubular structure that extends from the pylorus to the cecum. The estimated length of this structure varies depending on whether radiologic, surgical, or autopsy measurements are made. In the living, it is thought to measure 4 to 6 m.² The small intestine consists of three segments lying in series: the duodenum, the jejunum, and the ileum. The ­duodenum, the most proximal segment, lies in the retroperitoneum immediately adjacent to the head and inferior border of the body of the pancreas. The duodenum is demarcated from the stomach by the pylorus and from the jejunum by the ligament of Treitz. The jejunum and ileum lie within the peritoneal cavity and are tethered to the retroperitoneum by a broad-based mesentery. No distinct anatomic landmark demarcates the jejunum from the ileum; the proximal 40% of the jejunoileal segment is arbitrarily defined as the jejunum and the distal 60% as the ileum. The ileum is demarcated from the cecum by the ileocecal valve.

The small intestine contains internal mucosal folds known as plicae circulares or valvulae conniventes that are visible upon gross inspection. These folds are also visible radiographically and help in the distinction between small intestine and colon, which does not contain them, on abdominal radiographs. These folds are more prominent in the proximal intestine than in the distal small intestine. Other features evident on gross inspection that are more characteristic of the proximal than distal small intestine include larger circumference, thicker wall, less fatty mesentery, and longer vasa recta (Fig. 28-1). Gross examination of the small-intestinal mucosa also reveals aggregates of lymphoid follicles. Those follicles, located in the ileum, are the most prominent and are designated Peyer’s patches.

Most of the duodenum derives its arterial blood from branches of both the celiac and the superior mesenteric arteries. The distal duodenum, the jejunum, and the ileum derive their arterial blood from the superior mesenteric artery. Their venous drainage occurs via the superior mesenteric vein. Lymph drainage occurs through lymphatic vessels coursing parallel to corresponding arteries. This lymph drains through mesenteric lymph nodes to the cisterna chyli, then through the thoracic duct, and ultimately into the left subclavian vein. The parasympathetic and sympathetic innervation of the small intestine is derived from the vagus and splanchnic nerves, respectively.

The wall of the small intestine consists of four distinct layers: mucosa, submucosa, muscularis propria, and serosa (Fig. 28-2).

The mucosa is the innermost layer, and it consists of three layers: epithelium, lamina propria, and muscularis mucosae. The epithelium is exposed to the intestinal lumen and is the surface through which absorption from and secretion into the lumen occurs. The lamina propria is located immediately external to the epithelium and consists of connective tissue and a heterogeneous population of cells. It is demarcated from the more external submucosa by the muscularis mucosae, a thin sheet of smooth muscle cells.

The mucosa is organized into villi and crypts (crypts of Lieberkuhn). Villi are finger-like projections of epithelium and underlying lamina propria that contain blood and lymphatic (lacteals) vessels that extend into the intestinal lumen. Intestinal, epithelial cellular proliferation is confined to the crypts, each of which carries 250 to 300 cells. All epithelial cells in each crypt are derived from an unknown number of multipotent stem cells located at or near the crypt’s base. Our understanding of these crypt cells is rapidly expanding. It appears that there are two subgroups of intestinal stem cells, with specific cell markers. Bmi1-positive cells are usually quiescent, radiation-resistant cells that are induced by injury, whereas LGR5-positive cells facilitate homeostatic vs. injury-induced regeneration and are radiation sensitive.³

The stem cells can differentiate along one of four pathways that ultimately yield enterocytes and goblet, enteroendocrine, and Paneth cells. With the exception of Paneth cells, these lineages complete their terminal differentiation during an upward migration from each crypt to adjacent villi. The journey from the crypt to the villus tip is completed in 2 to 5 days and terminates with cells being removed by apoptosis and/or exfoliation. Thus, the small-intestinal epithelium undergoes continuous renewal, making it one of the body’s most dynamic tissues. The high cellular turnover rate contributes to mucosal resiliency but also makes the intestine uniquely susceptible to certain forms of injury such as that induced by radiation and chemotherapy.

Enterocytes are the predominant absorptive cell of the intestinal epithelium. Their apical (lumen-facing) cell membrane contains specialized digestive enzymes, transporter mechanisms, and microvilli that are estimated to increase the absorptive surface area of the small intestine by up to 40-fold. Goblet cells produce mucin believed to play a role in mucosal defense against pathogens. Enteroendocrine cells are characterized by secretory granules containing regulatory agents and are discussed in greater detail later in the Endocrine Function section. Paneth cells are located at the base of the crypt and contain secretory granules containing growth factors, digestive enzymes, and antimicrobial peptides. In addition, the intestinal epithelium contains M cells and intraepithelial lymphocytes. These two components of the immune system are discussed later.

The submucosa consists of dense connective tissue and a heterogeneous population of cells, including leukocytes and fibroblasts. The submucosa also contains an extensive network of vascular and lymphatic vessels, nerve fibers, and ganglion cells of the submucosal (Meissner’s) plexus.

The muscularis propria consists of an outer, longitudinally oriented layer and an inner, circularly oriented layer of smooth muscle fibers. Located at the interface between these two layers are ganglion cells of the myenteric (Auerbach’s) plexus.

The serosa consists of a single layer of mesothelial cells and is a component of the visceral peritoneum.

The first recognizable precursor of the small intestine is the embryonic gut tube, formed from the endoderm during the fourth week of gestation. The gut tube is divided into forgut, midgut, and hindgut. Other than duodenum, which is a forgut structure, the rest of the small intestine is derived from the midgut. The gut tube initially communicates with the yolk sac; however, the communication between these two structures narrows by the sixth week to form the vitelline duct. The yolk sac and vitelline duct usually undergo obliteration by the end of ­gestation. Incomplete obliteration of the vitelline duct results in the spectrum of defects associated with Meckel’s diverticulum.

Also during the fourth week of gestation, the mesoderm of the embryo splits. The portion of mesoderm that adheres to the endoderm forms the visceral peritoneum, while the portion that adheres to the ectoderm forms the parietal peritoneum. This mesodermal division results in the formation of a coelomic cavity that is the precursor of the peritoneal cavity.

At approximately the fifth week of gestation, the bowel begins to lengthen to an extent greater than that which can be accommodated by the developing abdominal cavity, resulting in the extracoelomic herniation of the developing bowel. The bowel continues to lengthen during the subsequent weeks and is retracted back into the abdominal cavity during the tenth week of gestation. Subsequently, the duodenum becomes a retroperitoneal structure. Coincident with extrusion and retraction, the bowel undergoes a 270° counterclockwise rotation relative to the posterior abdominal wall. This rotation accounts for the usual locations of the cecum in the right lower quadrant and the duodenojejunal junction to the left of midline (Fig. 28-3).

The celiac and superior mesenteric arteries and veins are derived from the vitelline vascular system, which in turn is derived from blood vessels formed within the splanchnopleuric mesoderm during the third week of gestation. Neurons found in the small intestine are derived from neural crest cells that begin to migrate away from the neural tube during the third week of gestation. These neural crest cells enter the mesenchyme of the primitive foregut and subsequently migrate to the remainder of the bowel.

During the sixth week of gestation, the lumen of the developing bowel becomes obliterated as bowel epithelial proliferation accelerates. Vacuoles form within the bowel substance during the subsequent weeks and coalesce to form the intestinal lumen by the ninth week of gestation. Errors in this recanalization may account for defects such as intestinal webs and stenoses. Most intestinal atresias, however, are believed to be related to ischemic episodes occurring after organogenesis has been completed rather than to errors in recanalization.

During the ninth week of gestation, the intestinal epithelium develops intestine-specific features such as crypt-villus architecture. Organogenesis is complete by approximately the twelfth week of gestation.

Digestion and Absorption

The intestinal epithelium is the interface through which absorption and secretion occur. It has features characteristic of absorptive epithelia in general including epithelial cells with cellular membranes possessing distinct apical (luminal) and basolateral (serosal) domains demarcated by intercellular tight junctions and an asymmetric distribution of transmembrane transporter mechanisms that promotes vectorial transport of solutes across the epithelium.

Solutes can traverse the epithelium by active or passive transport. Passive transport of solutes occurs through diffusion or convection and is driven by existing electrochemical gradients. Active transport is the energy-dependent net transferof solutes in the absence of or against an electrochemical ­gradient.

Active transport occurs through transcellular pathways (through the cell), whereas passive transport can occur through either transcellular or paracellular pathways (between cells through the tight junctions). Transcellular transport requires solutes to traverse the cell membranes through specialized membrane proteins, such as channels, carriers, and pumps. The molecular characterization of transporter proteins is evolving rapidly, with different transporter families, each containing many individual genes encoding specific transporters, now identified. Similarly, understanding of the paracellular pathway is evolving. In contrast to what was once believed, it is becoming apparent that paracellular permeability is substrate-specific, dynamic, and subject to regulation by specific tight junction proteins.

Water and Electrolyte Absorption and Secretion

Eight to nine L of fluid enters the small intestine daily. Most of this volume consists of salivary, gastric, biliary, pancreatic, and intestinal secretions. Under normal conditions, the small intestine absorbs over 80% of this fluid, leaving approximately 1.5 L that enters the colon (Fig. 28-4). Small-intestinal absorption and secretion are tightly regulated; derangements in water and electrolyte homeostasis characteristic of many of the disorders discussed in this chapter play an important role in contributing to their associated clinical features.

Gut epithelia have two pathways for water transport: (a) the paracellular route, which involves transport through the spaces between cells, and (b) the transcellular route, through apical and basolateral cell membranes. Although most water absorption was thought to occur through the paracellular pathway, it is now well documented that most intestinal water transport occurs through the transcellular pathway.⁴ The specific transport mechanisms mediating this transcellular transport are not completely characterized and may involve passive diffusion through the phospholipid bilayer, cotransport with other ions and nutrients, or diffusion through water channels called aquaporins. Many different types of aquaporins have been identified; however, their contribution to overall intestinal water absorption appears to be relatively minor.⁵

The prevailing model for intestinal epithelial Na⁺ absorption is shown in Fig. 28-5. Activity of the Na⁺/K⁺ ATPase enzyme, which is located in the basolateral membrane and exchanges three intracellular Na⁺ for every two extracellular K⁺ in an energy-dependent process, generates the electrochemical gradient that drives the transport of Na⁺ from the intestinal lumen into the cytoplasm of enterocytes. Na⁺ ions traverse the apical membrane through several distinct transporter mechanisms including nutrient-coupled sodium transport (e.g., sodium glucose cotransporter-1 [SGLT1]), sodium channels, and sodium-hydrogen exchangers (NHEs). Absorbed Na⁺ ions are then extruded from enterocytes through the Na⁺/K⁺ ATPase located in the basolateral membrane. Similar mechanistic models that account for the transport of other common ions such as K⁺ and HCO₃^– also exist.

Substantial heterogeneity, with respect to both crypt-villus and craniocaudal axes, exists for intestinal epithelial transport mechanisms. This spatial distribution pattern is consistent with a model in which absorptive function resides primarily in the villus and secretory function in the crypt.

Intestinal absorption and secretion are subject to modulation under physiologic and pathophysiologic conditions by a wide array of hormonal, neural, and immune regulatory mediators (Table 28-1).

Carbohydrate Digestion and Absorption

Approximately 45% of energy consumption in the average Western diet consists of carbohydrates, approximately one half of which is in the form of starch (linear or branched polymers of glucose) derived from cereals and plants. Other major sources of dietary carbohydrates include sugars derived from milk (lactose), fruits and vegetables (fructose, glucose, and sucrose), or purified from sugar cane or beets (sucrose). Processed foods contain a variety of sugars including fructose, oligosaccharides, and polysaccharides. Glycogen derived from meat contributes only a small fraction of dietary carbohydrate.

Pancreatic amylase is the major enzyme of starch digestion, although salivary amylase initiates the process. The terminal products of amylase-mediated starch digestion are oligosaccharides, maltotriose, maltose, and α-limit dextrins (Fig. 28-6). These products, as well as the major disaccharides in the diet (sucrose and lactose), are unable to undergo absorption in this form. They must first undergo hydrolytic cleavage into their constituent monosaccharides; these hydrolytic reactions are catalyzed by specific brush border membrane hydrolases that are expressed most abundantly in the villi of the duodenum and jejunum. The three major monosaccharides that represent the terminal products of carbohydrate digestion are glucose, galactose, and fructose.

Under physiologic conditions, most of these sugars are absorbed through the epithelium via the transcellular route. Glucose and galactose are transported through the enterocyte brush border membrane via intestinal Na⁺-glucose cotransporter, SGLT1 (Fig. 28-7). Fructose is transported through the brush border membrane by facilitated diffusion via GLUT5 (a member of the facilitative glucose transporter family). All three monosaccharides are extruded through the basolateral membrane by facilitated diffusion using GLUT2 and GLUT5 transporters. Extruded monosaccharides diffuse into venules and ultimately enter the portal venous system.

There is evidence of overexpression of hexose transporters, specifically SGLT1, in disease states such as diabetes.⁶ ­Several approaches aimed at downregulation of intestinal ­glucose transporter are being investigated as a novel therapy for disease states such as diabetes and obesity.⁷

Protein Digestion and Absorption

Ten to fifteen percent of energy consumption in the average Western diet consists of proteins. In addition to dietary proteins, approximately one half of the protein load that enters the small intestine is derived from endogenous sources including salivary and gastrointestinal secretions and desquamated intestinal epithelial cells. Protein digestion begins in the stomach with action of pepsins. This is not, however, an essential step, since surgical patients who are acholorhydric or who have lost part or all their stomach are still able to successfully digest proteins. Digestion continues in the duodenum with the actions of a variety of pancreatic peptidases. These enzymes are secreted as inactive proenzymes. This is in contrast to pancreatic amylase and lipase, which are secreted in their active forms. In response to the presence of bile acids, enterokinase is liberated from the intestinal brush border membrane to catalyze the conversion of trypsinogen to active trypsin; trypsin in turn activates itself and other proteases. The final products of intraluminal protein digestion consist of neutral and basic amino acids and peptides two to six amino acids in length (Fig. 28-8). Additional digestion occurs through the actions of peptidases that exist in the enterocyte brush border and cytoplasm. Epithelial absorption occurs for both single amino acids and di- or tripeptides via specific membrane-bound transporters. Absorbed amino acids and peptides then enter the portal venous circulation.

Of all amino acids, glutamine appears to be a unique, major source of energy for enterocytes. Active glutamine uptake into enterocytes occurs through both apical and basolateral transport mechanisms.

Fat Digestion and Absorption

Approximately 40% of the average Western diet consists of fat. Over 95% of dietary fat is in the form of long-chain triglycerides; the remainder includes phospholipids such as lecithin, fatty acids, cholesterol, and fat-soluble vitamins. Over 94% of the ingested fats are absorbed in the proximal jejunum.

Since fats are normally water insoluble, key to successful digestion of ingested fats is solubalization of them into an emulsion by the mechanical actions of mastication and antral peristalsis. Although lipolysis of triglycerides to form fatty acids and monoglyciderides is initiated in the stomach by gastric lipase, its principal site is the proximal intestine, where pancreatic lipase is the catalyst (Fig. 28-9).

Bile acids act as detergents that help in solubalization of the lipolysis by forming mixed micelles. These micelles are polymolecular aggregates with a hydrophobic core of fat and a hydrophilic surface that act as shuttles, delivering the products of lipolysis to the enterocyte brush border membrane, where they are absorbed. The bile salts, however, remain in the bowel lumen and travel to the terminal ileum, where they are actively resorbed. They enter the portal circulation and are resecreted into bile, thus completing the enterohepatic circulation.

Dissociation of lipids from the micelles occurs in a thin layer of water (50–500 μm thick) with an acidic microenvironment immediately adjacent to the brush border called the unstirred water layer. Most lipids are absorbed in the proximal jejunum, whereas bile salts are absorbed in the distal ileum through an active process. Fatty acid binding proteins (FABPs) are a family of proteins located on the brush border membrane, facilitating diffusion of long-chain fatty acids across the brush border membrane. Cholesterol crosses the brush border membrane through an active process that is yet to be completely characterized. Within the enterocytes, triglycerides are resynthesized and incorporated into chylomicrons that are secreted into the intestinal lymphatics and ultimately enter the thoracic duct. In these chylomicrons, lipoproteins serve a detergent-like role similar to that served by bile salts in the mixed micelles.

The steps described above are required for the digestion and absorption of triglycerides containing long-chain fatty acids. However, triglycerides containing short- and medium-chain fatty acids are more hydrophilic and are absorbed without undergoing intraluminal hydrolysis, micellular solubilization, mucosal re-esterification, and chylomicron formation. Instead, they are directly absorbed and enter the portal venous ­circulation rather than the lymphatics. This information provides the rationale for administering nutritional supplements containing medium-chain triglycerides to patients with gastrointestinal diseases associated with impaired digestion and/or malabsorption of long-chain triglycerides.

Vitamin and Mineral Absorption

Vitamin B₁₂ (cobalamin) malabsorption can result from a variety of surgical manipulations. The vitamin is initially bound by saliva-derived R protein. In the duodenum, R protein is hydrolyzed by pancreatic enzymes, allowing free cobalamin to bind to gastric parietal cell–derived intrinsic factor. The cobalamin-intrinsic factor complex is able to escape hydrolysis by pancreatic enzymes, allowing it to reach the terminal ileum, which expresses specific receptors for intrinsic factor. Subsequent events in cobalamin absorption are poorly characterized, but the intact complex probably enters enterocytes through translocation. Because each of these steps is necessary for cobalamin assimilation, gastric resection, gastric bypass, and ileal resection can each result in vitamin B₁₂ insufficiency.

Other water-soluble vitamins for which specific carrier-mediated transport processes have been characterized include ascorbic acid, folate, thiamine, riboflavin, pantothenic acid, and biotin. Fat-soluble vitamins A, D, and E appear to be absorbed through passive diffusion. Vitamin K appears to be absorbed through both passive diffusion and carrier-mediated uptake.

Calcium is absorbed through both transcellular transport and paracellular diffusion. The duodenum is the major site for transcellular transport; paracellular transport occurs throughout the small intestine. A key step in transcellular calcium transport is mediated by calbindin, a calcium-binding protein located in the cytoplasm of enterocytes. Regulation of calbindin synthesis is the principle mechanism by which vitamin D regulates intestinal calcium absorption. Abnormal calcium levels are increasingly seen in surgical patients who have undergone a gastric bypass. Although usual calcium supplementation is often in the form of calcium carbonate, which is cheap, in such patients with low acid exposure, calcium citrate is a better formulation for supplemental therapy.

Iron and magnesium are each absorbed through both transcellular and paracellular routes. A divalent metal transporter capable of transporting Fe²⁺, Zn²⁺, Mn²⁺, Co²⁺, Cd²⁺, Cu²⁺, Ni²⁺, and Pb²⁺ that has recently been localized to the intestinal brush border may account for at least a portion of the transcellular absorption of these ions.⁸

Barrier and Immune Function

Although the intestinal epithelium allows for the efficient absorption of dietary nutrients, it must discriminate between pathogens and harmless antigens such as food proteins and commensal bacteria, and it must resist invasion by pathogens. Factors contributing to epithelial defense include immunoglobulin A (IgA), mucins, and the relative impermeability of the brush border membrane and tight junctions to macromolecules and bacteria. Recently described factors likely to play important roles in intestinal mucosal defense include antimicrobial peptides such as the defensins.⁹ The intestinal component of the immune system, known as the gut-associated lymphoid tissue (GALT), contains over 70% of the body’s immune cells.

The GALT is conceptually divided into inductive and effector sites.¹⁰ Inductive sites include Peyer’s patches, mesenteric lymph nodes, and smaller isolated lymphoid ­follicles scattered throughout the small intestine (Fig. 28-10). Peyer’s patches are macroscopic aggregates of B-cell follicles and intervening T-cell areas found in the lamina propria of the small intestine, primarily the distal ileum. Overlying Peyer’s patches are a specialized epithelium containing microfold (M) cells. These cells possess an apical membrane with microfolds rather than microvilli, which is characteristic of most intestinal epithelial cells. Using transepithelial vesicular transport, M cells transfer microbes to underlying professional antigen-presenting cells (APCs), such as dendritic cells. Dendritic cells, in addition, may sample luminal antigens directly through their dendrite-like processes that extend through epithelial tight junctions. APCs interact with and prime naïve lymphocytes, which then exit through the draining lymphatics to enter the mesenteric lymph nodes where they undergo differentiation. These lymphocytes then migrate into the systemic circulation via the thoracic duct and ultimately accumulate in the intestinal mucosa at effector sites. Alternative induction mechanisms, such as antigen presentation within mesenteric lymph nodes, are also likely to exist.

Effector lymphocytes are distributed into distinct compartments. IgA-producing plasma cells are derived from B cells and are located in the lamina propria. CD4⁺ T cells are also located in the lamina propria. CD8⁺ T cells migrate preferentially to the epithelium, but are also found in the lamina propria. These T cells are central to immune regulation; in addition, the CD8⁺ T cells have potent cytotoxic (CTL) activity. IgA is transported through the intestinal epithelial cells into the lumen, where it exists in the form of a dimer complexed with a secretory component. This configuration renders IgA resistant to proteolysis by digestive enzymes. IgA is believed to both help prevent the entry of microbes through the epithelium and to promote excretion of antigens or microbes that have already penetrated into the lamina propria.

It has been increasingly recognized that the gastrointestinal tract is colonized with many bacteria that are essential for health. Communication between the microbiota and the host defense allows for protective immune responses against pathogens while preventing adverse inflammatory responses to harmless commensal microbes, which could lead to chronic inflammatory disorders such as celiac disease and Crohn’s disease.¹¹

Motility

Myocytes of the intestinal muscle layers are electrically and mechanically coordinated in the form of syncytia. Contractions of the muscularis propria are responsible for small-intestinal peristalsis. Contraction of the outer longitudinal muscle layer results in bowel shortening; contraction of the inner circular layer results in luminal narrowing. Contractions of the muscularis mucosa contribute to mucosal or villus motility, but not to peristalsis.

Several distinctive patterns of muscularis propria activity have been observed to occur in the small intestine. These patterns include ascending excitation and descending inhibition in which muscular contraction occurs proximal to a stimulus, such as the presence of a bolus of ingested food, and muscular relaxation occurs distal to the stimulus (Fig. 28-11). These two reflexes are present even in the absence of any extrinsic innervation to the small intestine and contribute to peristalsis when they are propagated in a coordinated fashion along the length of the intestine. The fed or postprandial pattern begins within 10 to 20 minutes of meal ingestion and abates 4 to 6 hours afterward. Rhythmic segmentations or pressure waves traveling only short distances also are observed. This segmenting pattern is hypothesized to assist in mixing intraluminal contents and in facilitating their contact with the absorptive mucosal surface. The fasting pattern or interdigestive motor cycle (IDMC) consists of three phases. Phase I is characterized by motor quiescence, phase II by seemingly disorganized pressure waves occurring at submaximal rates, and phase III by sustained pressure waves occurring at maximal rates. This pattern is hypothesized to expel residual debris and bacteria from the small intestine. The median duration of the IDMC ranges from 90 to 120 minutes. At any given time, different portions of the small intestine can be in different phases of the IDMC.

The regulatory mechanisms driving small-intestinal motility consist of both pacemakers intrinsic to the small intestine and external neurohumoral modulatory signals. The interstitial cells of Cajal are pleomorphic mesenchymal cells located within the muscularis propria of the intestine that generate the electrical slow wave (basic electrical rhythm or pacesetter potential) that plays a pacemaker role in setting the fundamental rhythmicity of small-intestinal contractions. The frequency of the slow wave varies along the longitudinal axis of the intestine: it ranges from 12 waves per minute in the duodenum to 7 waves per minute in the distal ileum. Smooth muscle contraction occurs only when an electrical action potential (spike burst) is superimposed on the slow wave. Thus, the slow wave determines the maximum frequency of contractions; however, not every slow wave is associated with a contraction.

This intrinsic contractile mechanism is subject to neural and hormonal regulation. The enteric nervous system (ENS) provides both inhibitory and excitatory stimuli. The predominant excitatory transmitters are acetylcholine and substance P, and the inhibitory transmitters include nitric oxide, vasoactive intestinal peptide, and adenosine triphosphate. In general, the sympathetic motor supply is inhibitory to the ENS; therefore, increased sympathetic input into the intestine leads to decreased intestinal smooth muscle activity. The parasympathetic motor supply is more complex, with projections to both inhibitory and excitatory ENS motor neurons. Correspondingly, the effects of parasympathetic inputs into intestinal motility are more difficult to predict.

Endocrine Function

Endocrinology as a discipline was born with the discovery of secretin, an intestinal regulatory peptide that was the first hormone to be identified. Our improving understanding of the physiology of the small intestine has lead to identification of many additional intestinal-derived hormones that make this the largest hormone-producing organ in the body. Over 30 peptide hormone genes have been identified as being expressed in the gastrointestinal tract. Because of differential posttranscriptional and posttranslational processing, over 100 distinct regulatory peptides are produced. In addition, monoamines, such as histamine and dopamine, and eicosanoids with hormone-like activities are produced in the intestine.

“Gut hormones” were previously conceptualized as peptides produced by the enteroendocrine cells of the intestinal mucosa that are released into the systemic circulation to reach receptors in target sites in the gastrointestinal tract. Now it is clear that “gut hormone” genes are widely expressed throughout the body, not only in endocrine cells, but also in central and peripheral neurons. The products of these genes are general intercellular messengers that can act as endocrine, paracrine, autocrine, or neurocrine mediators. Thus, they may act as true blood-borne hormones as well as through local effects.

There are notable homology patterns among individual regulatory peptides found in the gastrointestinal tract. Based on these homologies, approximately one half of the known regulatory peptides can be classified into families.¹² For example, the secretin family includes secretin, glucagon, and glucagon-like peptides, glucose-dependent insulinotropic peptide, vasoactive intestinal polypeptide, peptide histidine isoleucine, growth ­hormone–releasing hormone, and pituitary adenylyl cyclase–activating peptide. Other peptide families include those named for insulin, epidermal growth factor, gastrin, pancreatic polypeptide, tachykinin, and somatostatin.

Receptor subtype multiplicity and cell-specific expression patterns for these receptor subtypes that are characteristic of these regulatory mediators make definition of their actions complex. Detailed description of these actions is beyond the scope of this chapter; however, examples of regulatory peptides produced by enteroendocrine cells of the small-intestinal epithelium and their most commonly ascribed functions are summarized in Table 28-2. Some of these peptides, or their analogues, are used in routine clinical practice. For example, therapeutic applications of octreotide, a long-acting analogue of somatostatin, include the amelioration of symptoms associated with neuroendocrine tumors (e.g., carcinoid syndrome), postgastrectomy dumping syndrome, enterocutaneous fistulas, and the initial treatment of acute hemorrhage due to esophageal varices. The gastrin secretory response to secretin administration forms the basis for the standard test used to establish the diagnosis of Zollinger-Ellison syndrome. Cholecystokinin is used in evaluations of gallbladder ejection fraction, a parameter that may have utility in patients who have symptoms of biliary colic but are not found to have gallstones. Of the peptides listed in Table 28-2, glucagon-like peptide-2 (GLP-2) has been identified as a specific and potent intestinotrophic hormone and is currently under clinical evaluation as an intestinotrophic agent in patients suffering from the short bowel syndrome, as discussed in the later Short Bowel Syndrome section.

Intestinal Adaptation

The small intestine has the capacity to adapt in response to varying demands imposed by physiologic and pathologic conditions. Of particular relevance to many of the diseases discussed in this chapter is the adaptation that occurs in the remnant intestine following surgical resection of a large portion of the small intestine (massive small bowel resection). Postresection intestinal adaptation has been studied extensively using animal models. Within a few hours after bowel resection, the remnant small intestine displays evidence of epithelial cellular hyperplasia. With additional time, villi lengthen, intestinal absorptive surface area increases, and digestive and absorptive functions improve. Postresection intestinal adaptation in human patients is less well studied, but seems to follow similar steps as those seen in experimental models, and takes 1 to 2 years to complete.¹³

The mechanisms responsible for inducing postresection intestinal adaptation are under active investigation. Several classes of effectors that stimulate intestinal growth include specific nutrients, peptide hormones and growth factors, pancreatic secretions, and some cytokines. Nutritional components with intestinal growth-stimulating effects include fiber, fatty acids, triglycerides, glutamine, polyamines, and lectins.

Postresection adaptation serves to compensate for the function of intestine that has been resected. Jejunal resection is generally better tolerated, as ileum shows better capacity to compensate. However, the magnitude of this response is limited. If enough small intestine is resected, a devastating condition known as the short bowel syndrome results. This condition is discussed in the Short Bowel Syndrome section at the end of this chapter.

Epidemiology

Mechanical small bowel obstruction is the most frequently encountered surgical disorder of the small intestine. Although a wide range of etiologies for this condition exists, the obstructing lesion can be conceptualized according to its anatomic relationship to the intestinal wall as:

1. Intraluminal (e.g., foreign bodies, gallstones, or meconium)

2. Intramural (e.g., tumors, Crohn’s disease–associated inflammatory strictures)

3. Extrinsic (e.g., adhesions, hernias, or carcinomatosis)

Intra-abdominal adhesions related to prior abdominal surgery account for up to 75% of cases of small bowel obstruction. Over 300,000 patients are estimated to undergo surgery to treat adhesion-induced small bowel obstruction in the United States annually. A 20-year trend analysis between 1988 and 2007 has documented no decrease in this rate during this time period, highlighting the ongoing problem with this “old” disease.¹⁴

Less prevalent etiologies for small bowel obstruction include hernias, malignant bowel obstruction, and Crohn’s disease. The frequency with which obstruction related to these conditions is encountered varies according to the patient population and practice setting. Cancer-related small bowel obstructions are commonly due to extrinsic compression or invasion by advanced malignancies arising in organs other than the small bowel; few are due to primary small bowel tumors. The most commonly encountered etiologies of small bowel obstruction are summarized in Table 28-3. Although congenital abnormalities capable of causing small bowel obstruction usually become evident during childhood, they sometimes elude detection and are diagnosed for the first time in adult patients presenting with abdominal symptoms. For example, intestinal malrotation and mid-gut volvulus should not be forgotten when considering the differential diagnosis of adult patients with acute or chronic symptoms of small bowel obstruction, especially those without a history of prior abdominal surgery. A rare etiology of obstruction is the superior mesenteric artery syndrome, characterized by compression of the third portion of the duodenum by the superior mesenteric artery as it crosses over this portion of the duodenum. This condition should be considered in young asthenic individuals who have chronic symptoms suggestive of proximal small bowel obstruction.

Pathophysiology

With onset of obstruction, gas and fluid accumulate within the intestinal lumen proximal to the site of obstruction. The intestinal activity increases in an effort to overcome the obstruction, accounting for the colicky pain and the diarrhea that some experience even in the presence of complete bowel obstruction. Most of the gas that accumulates originates from swallowed air, although some is produced within the intestine. The fluid consists of swallowed liquids and gastrointestinal secretions (obstruction stimulates intestinal epithelial water secretion). With ongoing gas and fluid accumulation, the bowel distends and intraluminal and intramural pressures rise. The intestinal motility is eventually reduced with fewer contractions. With obstruction, the luminal flora of the small bowel, which is usually sterile, changes, and a variety of organisms have been cultured from the contents. Translocation of these bacteria to regional lymph nodes has been demonstrated, although the significance of this process is not well understood. If the intramural pressure becomes high enough, intestinal microvascular perfusion is impaired, leading to intestinal ischemia and, ultimately, necrosis. This condition is termed strangulated bowel obstruction.

With partial small bowel obstruction, only a portion of the intestinal lumen is occluded, allowing passage of some gas and fluid. The progression of pathophysiologic events described earlier tends to occur more slowly than with complete small bowel obstruction, and development of strangulation is less likely.

A particularly dangerous form of bowel obstruction is closed-loop obstruction in which a segment of intestine is obstructed both proximally and distally (e.g., with volvulus). In such cases, the accumulating gas and fluid cannot escape either proximally or distally from the obstructed segment, leading to a rapid rise in luminal pressure and a rapid progression to ­strangulation.

Clinical Presentation

The symptoms of small bowel obstruction are colicky abdominal pain, nausea, vomiting, and obstipation. Vomiting is a more prominent symptom with proximal obstructions than distal. The character of vomitus is important because with bacterial overgrowth, the vomitus is more feculent, suggesting a more established obstruction. Continued passage of flatus and/or stool beyond 6 to 12 hours after onset of symptoms is characteristic of partial rather than complete obstruction. The signs of small bowel obstruction include abdominal distention, which is most pronounced if the site of obstruction is in the distal ileum and may be absent if the site of obstruction is in the proximal small intestine. Bowel sounds may be hyperactive initially, but in late stages of bowel obstruction, minimal bowel sounds may be heard. Laboratory findings reflect intravascular volume depletion and consist of hemoconcentration and electrolyte abnormalities. Mild leukocytosis is common.

Features of strangulated obstruction include abdominal pain often disproportionate to the degree of abdominal findings, suggestive of intestinal ischemia. Patients often have tachycardia, localized abdominal tenderness, fever, marked leukocytosis, and acidosis. Any of these findings should alert the clinician to the possibility of strangulation and need for early surgical intervention.

Diagnosis

The diagnostic evaluation should focus on the following goals: (a) distinguish mechanical obstruction from ileus, (b) determine the etiology of the obstruction, (c) discriminate partial from complete obstruction, and (d) discriminate simple from strangulating obstruction.

Important elements to obtain on history include prior abdominal operations (suggesting the presence of adhesions) and the presence of abdominal disorders (e.g., intra-abdominal cancer or inflammatory bowel disease) that may provide insights into the etiology of obstruction. Upon examination, a meticulous search for hernias (particularly in the inguinal and femoral regions) should be conducted.

The diagnosis of small bowel obstruction is usually confirmed with radiographic examination. The abdominal series consists of (a) a radiograph of the abdomen with the patient in a supine position, (b) a radiograph of the abdomen with the patient in an upright position, and (c) a radiograph of the chest with the patient in an upright position. The finding most specific for small bowel obstruction is the triad of dilated small bowel loops (>3 cm in diameter), air-fluid levels seen on upright films, and a paucity of air in the colon. The sensitivity of abdominal radiographs in the detection of small bowel obstruction ranges 70% to 80%.¹⁵ Specificity is low, because ileus and colonic obstruction can be associated with findings that mimic those observed with small bowel obstruction. False-negative findings on radiographs can result when the site of obstruction is located in the proximal small bowel and when the bowel lumen is filled with fluid but no gas, thereby preventing visualization of air-fluid levels or bowel distention. The latter situation is associated with closed-loop obstruction. Despite these limitations, abdominal radiographs remain an important study in patients with suspected small bowel obstruction because of their widespread availability and low cost (Fig. 28-12).

Computed tomography (CT) scanning is 80% to 90% sensitive and 70% to 90% specific in the detection of small bowel obstruction.¹⁵ The findings of small bowel obstruction include a discrete transition zone with dilation of bowel proximally, decompression of bowel distally, intraluminal contrast that does not pass beyond the transition zone, and a colon containing little gas or fluid (Figs. 28-13 and 28-14). CT scanning may also provide evidence for the presence of closed-loop obstruction and strangulation. Closed-loop obstruction is suggested by the presence of a U-shaped or C-shaped dilated bowel loop associated with a radial distribution of mesenteric vessels converging toward a torsion point. Strangulation is suggested by thickening of the bowel wall, pneumatosis intestinalis (air in the bowel wall), portal venous gas, mesenteric haziness, and poor uptake of intravenous contrast into the wall of the affected bowel (Fig. 28-15). CT scanning also offers a global evaluation of the abdomen and may therefore reveal the etiology of obstruction. This feature is important in the acute setting when intestinal obstruction represents only one of many diagnoses in patients presenting with acute abdominal conditions.

The CT scan is usually performed after administration of oral water-soluble contrast or diluted barium. The water-soluble contrast has been shown to have prognostic and therapeutic value too. Several studies and a subsequent meta-analysis have shown that appearance of the contrast in the colon within 24 hours is predictive of nonsurgical resolution of bowel obstruction.¹⁶ Although use of oral contrast did not alter the rate of surgical intervention, it did reduce the overall length of hospital stay in those presenting with small bowel obstruction.

A limitation of CT scanning is its low sensitivity (<50%) in the detection of low-grade or partial small bowel obstruction. A subtle transition zone may be difficult to identify in the axial images obtained during CT scanning. In such cases, ­contrast examinations of the small bowel, either small bowel series (small bowel follow-through) or enteroclysis, can be helpful. For standard small bowel series, contrast is swallowed or instilled into the stomach through a nasogastric tube. Abdominal radiographs are then taken serially as the contrast travels distally in the intestine. Although barium can be used, water-soluble contrast agents, such as Gastrografin, should be used if the possibility of intestinal perforation exists. These examinations are more labor-intensive and less rapidly performed than CT ­scanning but may offer greater sensitivity in the detection of luminal and mural etiologies of obstruction, such as primary intestinal tumors. For enteroclysis, 200 to 250 mL of barium followed by 1 to 2 L of a solution of methylcellulose in water are instilled into the proximal jejunum via a long nasoenteric catheter. The double-contrast technique used in enteroclysis permits a better assessment of mucosal surface and detection of relatively small lesions, even through overlapping small bowel loops. Enteroclysis is rarely performed in the acute setting but offers greater sensitivity than small bowel series in the detection of lesions that may be causing partial small bowel obstruction. Recently, CT enteroclysis has been used and reported to be superior to plain x-ray small bowel contrast studies.

Therapy

Small bowel obstruction is usually associated with a marked depletion of intravascular volume due to decreased oral intake, vomiting, and sequestration of fluid in bowel lumen and wall. Therefore, fluid resuscitation is integral to treatment. Isotonic fluid should be given intravenously, and an indwelling bladder catheter may be placed to monitor urine output. Central venous or pulmonary artery catheter monitoring may be necessary to assist with fluid management in patients with underlying cardiac disease and severe dehydration. Broad-spectrum antibiotics are given by some because of concerns that bacterial translocation may occur in the setting of small bowel obstruction; however, there are no data to support this approach.

The stomach should be continuously evacuated of air and fluid using a nasogastric (NG) tube. Effective gastric decompression decreases nausea, distention, and the risk of vomiting and aspiration. Longer nasoenteric tubes, with tips placed into the jejunum or ileum, were favored in the past but are rarely used today, as they are associated with higher complication rates than NG tubes, with no proven greater efficacy in several studies.

The standard therapy for complete small bowel obstruction has generally been expeditious surgery, with the dictum that “the sun should never rise and set on a complete bowel obstruction.”Recently, however, some have advocated nonoperative approaches in management of these patients, provided closed-loop obstruction is ruled out and there is no evidence of intestinal ischemia. Such patients need to be observed closely and undergo serial exams. The rationale for those favoring early surgical intervention is to minimize the risk for bowel strangulation, which is associated with an increased risk for morbidity and mortality. Clinical signs and currently available laboratory tests and imaging studies do not reliably permit the distinction between patients with simple obstruction and those with strangulated obstruction prior to the onset of irreversible ischemia. Therefore, the goal is to operate before the onset of irreversible ischemia. Others note, however, that a period of observation and nasogastric decompression, provided no tachycardia, tenderness, or an increase in white cell count is noted, are appropriate (see Fig. 28-16 for a proposed management algorithm).

However, conservative therapy, in the form of NG decompression and fluid resuscitation, is commonly recommended in the initial recommendation for:

1. Partial small bowel obstruction

2. Obstruction occurring in the early postoperative period

3. Intestinal obstruction due to Crohn’s disease

4. Carcinomatosis

In partial obstruction, progression to strangulation is unlikely to occur, and an attempt at nonoperative resolution is warranted. Nonoperative management has been documented to be successful in 65% to 81% of patients with partial small bowel obstruction. Of those successfully treated nonoperatively, only 5% to 15% have been reported to have symptoms that were not substantially improved within 48 hours after initiation of therapy.¹⁷ Therefore, most patients with partial small bowel obstruction whose symptoms do not improve within 48 hours after initiation of nonoperative therapy should undergo surgery. In a recent study, using the National Inpatient Sample, this principle was further highlighted. The authors concluded that a 2-day limit of watchful waiting before surgery is not associated with an increase in mortality or postoperative morbidity, although inpatient costs were higher.¹⁸

Patients undergoing nonoperative therapy should be closely monitored for signs suggestive of peritonitis, the development of which would mandate urgent surgery. As stated before, the administration of hypertonic water-soluble contrast agents, such as Gastrografin used in upper gastrointestinal (GI) and small bowel follow-through examinations, causes a shift of fluid into the intestinal lumen, thereby increasing the pressure gradient across the site of obstruction. This effect may accelerate resolution of partial small bowel obstruction; however, there is less evidence that administration of water-soluble contrast agents increases the probability that an episode of bowel obstruction will be successfully managed nonoperatively.¹⁶

Obstruction presenting in the early postoperative period has been reported to occur in 0.7% patients undergoing laparotomy.¹⁹ Patients undergoing pelvic surgery, especially colorectal procedures, have the greatest risk for developing early postoperative small bowel obstruction. The presence of obstruction should be considered if symptoms of intestinal obstruction occur after the initial return of bowel function or if bowel function fails to return within the expected 3 to 5 days after abdominal surgery. Plain radiographs may demonstrate dilated loops of small intestine with air-fluid levels but are interpreted as normal or nonspecific in up to a third of patients with early postoperative obstruction. CT scanning or a small bowel series is often required to make the diagnosis. Obstruction that occurs in the early postoperative period is usually partial and only rarely is associated with strangulation. Therefore, a period of extended nonoperative therapy (2–3 weeks) consisting of bowel rest, hydration, and total parental nutrition (TPN) administration is usually warranted. However, if complete obstruction is demonstrated or if signs suggestive of peritonitis are detected, expeditious reoperation should be undertaken without delay.

Crohn’s disease as a cause of small bowel obstruction is discussed in more detail later in the Crohn’s Disease section.

Twenty-five to thirty-three percent of patients with a history of cancer who present with small bowel obstruction have adhesions as the etiology of their obstruction and therefore should not be denied appropriate therapy.²⁰ Even in cases in which the obstruction is related to recurrent malignancy, palliative resection or bypass can be performed. Patients with obvious carcinomatosis pose a difficult challenge, given their limited prognosis. Management must be tailored to an individual patient’s prognosis and desires, and relief of the obstruction may be best achieved by a bypass procedure, avoiding a potentially difficult bowel resection.

The operative procedure performed for small bowel obstruction varies according to the etiology of the obstruction. For example, adhesions are lysed, tumors are resected, and hernias are reduced and repaired. Regardless of the etiology, the affected intestine should be examined, and nonviable bowel resected. Criteria suggesting viability are normal color, peristalsis, and marginal arterial pulsations. Usually, visual inspection alone is adequate in judging viability. In borderline cases, a Doppler probe may be used to check for pulsatile flow to the bowel, and arterial perfusion can be verified by visualizing intravenously administered fluorescein dye in the bowel wall under ultraviolet illumination. However, neither technique has been found to be superior to clinical judgment. In general, if the patient is hemodynamically stable, short lengths of bowel of questionable viability should be resected and primary anastomosis of the remaining intestine performed. However, if the viability of a large proportion of the intestine is in question, a concerted effort to preserve intestinal tissue should be made. In such situations, the bowel of uncertain viability should be left intact and the patient re-explored in 24 to 48 hours in a “second-look” operation. At that time, definitive resection of nonviable bowel is completed.

Successful laparoscopic surgery for bowel obstruction is being reported with greater frequency.^22,23 Those who undergo successful laparoscopic procedure have a quicker recovery, less complications, and lower costs. Since distended loops of bowel can interfere with adequate visualization, early cases of proximal small bowel obstruction that are likely due to a single ­adhesive band are best suited for this approach. Presence of bowel distention and multiple adhesions can cause these procedures to be difficult and potentially hazardous. Conversion rate to open surgery is between 17% and 33%.^21,22,23

Outcomes

Prognosis is related to the etiology of obstruction. The majority of patients who are treated conservatively for adhesive small bowel obstruction do not require future readmissions; less than 20% of such patients will have a readmission over the subsequent 5 years with another episode of bowel obstruction.²³

The perioperative mortality rate associated with surgery for nonstrangulating small bowel obstruction is less than 5%, with most deaths occurring in elderly patients with significant comorbidities. Mortality rates associated with surgery for strangulating obstruction range from 8% to 25%.

Considering the frequency of small bowel obstruction and the varied degree of clinical severity and presentation, there is often no consistency as to whether the patient is admitted to the medical or surgical service, and further variability in whether and when surgery is consulted. Recent studies have shown that a standard hospital-wide policy can help improve care of patients with bowel obstruction, reducing their time to surgery and shortening their length of hospital stay.²⁴

Prevention

With adhesive small bowel obstruction representing a large therapeutic burden, prevention of postoperative adhesions has become an area of great interest. Good surgical technique, careful handling of tissue, and minimal use and exposure of peritoneum to foreign bodies form the cornerstone of adhesion prevention. These measures alone are often inadequate. In patients undergoing colorectal or pelvic surgery, hospital readmission rates of greater than 30% over the subsequent 10 years have been reported for adhesive small bowel obstruction.²⁵

Use of laparoscopic surgery, when possible, has been strongly promoted. A recent study using the Swedish Inpatient Register has shown that, compared to laparoscopy, open surgery is associated with a four-fold increase in risk of small bowel obstruction within 5 years of the index procedure even after accounting for other risk factors such as age, comorbidity, and previous abdominal surgery.²⁶

In those undergoing open surgery, several strategies for adhesion prevention have been tried; however, the only therapy that has shown some success has been the use of hyaluronan-based agents, such as Seprafilm. The use of this barrier has been clearly shown to reduce the incidence of postoperative bowel adhesions; however, its effect in actually reducing the incidence of small bowel obstruction remains less well defined.²⁷ The use of these products is often left to the discretion of the surgeon and the clinical context. Wrapping of an intestinal anastomosis with the material may be associated with increased leak rates and is generally discouraged.²⁸

Ileus and intestinal pseudo-obstruction designate clinical syndromes caused by impaired intestinal motility and are characterized by symptoms and signs of intestinal obstruction in the absence of a lesion-causing mechanical obstruction. Ileus is a major cause of morbidity in hospitalized patients. Postoperative ileus is the most frequently implicated cause of delayed discharge following abdominal operations; its economic impact has been estimated to be between $750 million and $1 billion annually in the United States.²⁹

Ileus is a temporary motility disorder that is reversed with time as the inciting factor is corrected. In contrast, chronic intestinal pseudo-obstruction comprises a spectrum of specific disorders associated with irreversible intestinal dysmotility.

Pathophysiology

Numerous factors capable of impairing intestinal motility, and thus inciting ileus, have been described (Table 28-4). The most frequently encountered factors are abdominal operations, infection and inflammation, electrolyte abnormalities, and drugs.

Following most abdominal operations or injuries, the motility of the GI tract is transiently impaired. Among the proposed mechanisms responsible for this dysmotility are surgical stress-induced sympathetic reflexes, inflammatory response mediator release, and anesthetic/analgesic side effects, each of which can inhibit intestinal motility.³⁰ The return of normal motility generally follows a characteristic temporal sequence, with small-intestinal motility returning to normal within the first 24 hours after laparotomy and gastric and colonic motility returning to normal by 48 hours and 3 to 5 days, respectively. Because small bowel motility is returned before colonic and gastric motility, listening for bowel sounds is not a reliable indicator that ileus has fully resolved. Functional evidence of coordinated GI motility in the form of passing flatus or a bowel movement is a more useful indicator. Resolution of ileus may be delayed in the presence of other factors capable of inciting ileus such as the presence of intra-abdominal abscesses or electrolyte abnormalities.

Chronic intestinal pseudo-obstruction can be caused by a large number of specific abnormalities affecting intestinal smooth muscle, the myenteric plexus, or the extraintestinal nervous system (Table 28-5). Visceral myopathies constitute a group of diseases characterized by degeneration and fibrosis of the intestinal muscularis propria. Visceral neuropathies encompass a variety of degenerative disorders of the myenteric and submucosal plexuses. Both sporadic and familial forms of visceral myopathies and neuropathies exist. Systemic disorders involving the smooth muscle, such as progressive systemic sclerosis and progressive muscular dystrophy, and neurologic diseases, such as Parkinson’s disease, can also be complicated by chronic intestinal pseudo-obstruction. In addition, viral infections, such as those associated with cytomegalovirus and Epstein-Barr virus, can cause intestinal pseudo-obstruction.

Clinical Presentation

The clinical presentation of ileus resembles that of small bowel obstruction. Inability to tolerate liquids and solids by mouth, nausea, and lack of flatus or bowel movements are the most common symptoms. Vomiting and abdominal distension may occur. Bowel sounds are characteristically diminished or absent, in contrast to the hyperactive bowel sounds that usually accompany mechanical small bowel obstruction. The clinical manifestations of chronic intestinal pseudo-obstruction include variable degrees of nausea and vomiting and abdominal pain and distention.

Diagnosis

Routine postoperative ileus should be expected and requires no diagnostic evaluation. If ileus persists beyond 3 to 5 days postoperatively or occurs in the absence of abdominal surgery, diagnostic evaluation to detect specific underlying factors capable of inciting ileus and to rule out the presence of mechanical obstruction is warranted.

Patient medication lists should be reviewed for the presence of drugs, especially opiates, known to be associated with impaired intestinal motility. Measurement of serum electrolytes may demonstrate electrolyte abnormalities commonly associated with ileus. Abdominal radiographs are often obtained, but the distinction between ileus and mechanical obstruction may be difficult based on this test alone. In the postoperative setting, CT scanning is the test of choice because it can demonstrate the presence of an intra-abdominal abscess or other evidence of peritoneal sepsis that may be causing ileus and can exclude the presence of complete mechanical obstruction. Distinction of postoperative ileus from early postoperative obstruction can be difficult but is helpful in developing the appropriate management plan.

The diagnosis of chronic pseudo-obstruction is suggested by clinical features and confirmed by radiographic and manometric studies. Diagnostic laparotomy or laparoscopy with full-thickness biopsy of the small intestine may be required to establish the specific underlying cause.

Therapy

The management of ileus consists of limiting oral intake and correcting the underlying inciting factor. If vomiting or abdominal distention is prominent, the stomach should be decompressed using a nasogastric tube. Fluid and electrolytes should be administered intravenously until ileus resolves. If the duration of ileus is prolonged, TPN may be required.

Given the frequency of postoperative ileus and its financial impact, a large number of investigations have been conducted to define strategies to reduce its duration. Although often recommended, the use of early ambulation and routine nasogastric intubation has not been demonstrated to be associated with earlier resolution of postoperative ileus. There is some evidence that early postoperative feeding protocols and are generally well tolerated, reduce postoperative ileus, and can result in a shorter hospital stay.³¹ The administration of nonsteroidal anti-inflammatory drugs such as ketorolac and concomitant reductions in opioid dosing have been shown to reduce the duration of ileus in most studies. Similarly, the use of perioperative thoracic epidural anesthesia/analgesia with regimens containing local anesthetics combined with limitation or elimination of systemically administered opioids has been shown to reduce duration of postoperative ileus, although they have not reduced the overall length of hospital stay.³² Interestingly, recent data have suggested that limiting intra- and postoperative fluid administration can also result in reduction of postoperative ileus and shortened hospital stay.³³ Table 28-6 summarizes some of the measures used to minimize postoperative ileus.

Most other pharmacologic agents, including prokinetic agents, are associated with efficacy-toxicity profiles that are too unfavorable to warrant routine use. Recently, administration of alvimopan, a novel peripherally active mu-opioid receptor antagonist with limited oral absorption, has been shown to reduce duration of postoperative ileus, hospital stay, and rate of readmissions in several prospective, randomized, placebo-controlled trials and a subsequent meta-analysis.³⁴ However, any cost savings associated with the use of this drug outside of a clinical trial have been debated.³⁵

Surgical interest in minimizing postoperative ileus has led to exploration of other potential avenues. Experiments in rodents have shown reduced postoperative ileus with enteral administration of lipid-rich nutrition, which is believed to work through a CCK-dependent vagovagal reflex.³⁶

Therapy of patients with chronic intestinal pseudo-obstruction focuses on palliation of symptoms as well as fluid, electrolyte, and nutritional management. Surgery should be avoided if at all possible. No standard therapies are curative or delay the natural history of any of the specific disorders causing intestinal pseudo-obstruction. Prokinetic agents, such as metoclopramide and erythromycin, are associated with poor efficacy. Cisapride has been associated with palliation of symptoms; however, because of cardiac toxicity and reported deaths, this agent is restricted to compassionate use in the United States.

Patients with refractory disease may require strict limitation of oral intake and long-term TPN administration. Despite these measures, some patients will continue to have severe abdominal pain or such copious intestinal secretions that vomiting and fluid and electrolyte losses remain substantial. These patients may require a decompressive gastrostomy or an extended small bowel resection to remove abnormal intestine. Small-intestinal transplantation has been applied in these patients with increasing frequency; the ultimate role of this modality remains to be defined.

Crohn’s disease is a chronic, idiopathic transmural inflammatory disease with a propensity to affect the distal ileum, although any part of the alimentary tract can be involved. Estimates of the incidence of Crohn’s disease in the United States have ranged from 3.6 to 8.8 per 100,000, with recent studies suggesting a prevalence of about 200 cases per 100,000.³⁷ A dramatic increase in incidence in the United States was observed to occur from the mid-1950s though the early 1970s. Incidence rates have been stable since the 1980s. Substantial regional variations in incidence have been observed, with the highest incidences reported to exist in northern latitudes. The incidence of Crohn’s disease varies among ethnic groups within the same geographic region. For example, members of Eastern European Ashkenazi Jewish population are at two- to four-fold higher risk of developing Crohn’s disease than members of other populations living in the same location. In countries such as China, the prevalence of Crohn’s disease is estimated at 1.38 cases per 100,000, substantially below that seen in the West, but the rates have been on a rapid increase recently.³⁸

Most studies suggest that Crohn’s disease is slightly more prevalent in females than in males. The mean age at which patients are diagnosed with Crohn’s disease falls in the third decade of life years, with a second smaller peak in the sixth decade of life, giving it a bimodal distribution. However, the age at diagnosis can range from early childhood through the entire lifespan.

Both genetic and environmental factors appear to influence the risk for developing Crohn’s disease. The relative risk among first-degree relatives of patients with Crohn’s disease is 14 to 15 times higher than that of the general population. Approximately one in five patients with Crohn’s disease will report having at least one affected relative. The concordance rate among monozygotic twins is as high as 67%; however, Crohn’s disease is not associated with simple Mendelian inheritance patterns. Although there is a tendency within families for either ulcerative colitis or Crohn’s disease to be present exclusively, mixed kindreds also occur, suggesting the presence of some shared genetic traits as a basis for both diseases.

Higher socioeconomic status is associated with an increased risk of Crohn’s disease. Most studies have found breastfeeding to be protective against the development of Crohn’s disease. Crohn’s disease is more prevalent among smokers. Furthermore, smoking is associated with the increased risk for both the need for surgery and the risk of relapse after surgery for Crohn’s disease.

Pathophysiology

Crohn’s disease is characterized by sustained inflammation. Whether this inflammation represents an appropriate response to a yet unrecognized pathogen or an inappropriate response to a normally innocuous stimulus is unknown. Various hypotheses on the roles of environmental and genetic factors in the pathogenesis of Crohn’s disease have been proposed. Many infectious agents have been suggested to be the causative organism of Crohn’s disease. Candidate organisms have included chlamydia, Listeria monocytogenes, Pseudomonas species, reovirus, Mycobacterium paratuberculosis, and many others. There is no conclusive evidence that any of these organisms is the causative agent. Studies using animal models suggest that in a genetically susceptible host, nonpathogenic, commensal enteric flora are sufficient to induce a chronic inflammatory response resembling that associated with Crohn’s disease. In these models, the sustained intestinal inflammation is the result of either abnormal epithelial barrier function or immune dysregulation. Poor barrier function is hypothesized to permit inappropriate exposure of lamina propria lymphocytes to antigenic stimuli derived from the intestinal lumen. In addition, a variety of defects in immune regulatory mechanisms (e.g., overresponsiveness of mucosal T cells to enteric flora-derived antigens) can lead to defective immune tolerance and sustained inflammation.

Specific genetic defects associated with Crohn’s disease in human patients are beginning to be defined. For example, the presence of a locus on chromosome 16 (the so-called IBD1 locus) has been linked to Crohn’s disease. The IBD1 locus has been identified as the NOD2 gene. Persons with allelic variants on both chromosomes have a 40-fold relative risk of Crohn’s disease compared to those without variant NOD2 genes. The relevance of this gene to the pathogenesis of Crohn’s disease is biologically plausible, as the protein product of the NOD2 gene mediates the innate immune response to microbial pathogens. Other putative IBD loci have been identified on other chromosomes (IBD2 on chromosome 12q and IBD3 on chromosome 6) and are under investigation.

Although the pathologic hallmark of Crohn’s disease is focal, transmural inflammation of the intestine, a spectrum of pathologic lesions can be present. The earliest lesion characteristic of Crohn’s disease is the aphthous ulcer. These superficial ulcers are up to 3 mm in diameter and are surrounded by a halo of erythema. In the small intestine, aphthous ulcers typically arise over lymphoid aggregates. Granulomas are highly characteristic of Crohn’s disease and are reported to be present in up to 70% of intestinal specimens obtained during surgical resection. These granulomas are noncaseating and can be found in both areas of active disease and apparently normal intestine, in any layer of the bowel wall, and in mesenteric lymph nodes.

As disease progresses, aphthae coalesce into larger, ­stellate-shaped ulcers. Linear or serpiginous ulcers may form when multiple ulcers fuse in a direction parallel to the longitudinal axis of the intestine. With transverse coalescence of ulcers, a cobblestoned appearance of the mucosa may arise.

With advanced disease, inflammation can be transmural. Serosal involvement results in adhesion of the inflamed bowel to other loops of bowel or other adjacent organs. Transmural inflammation can also result in fibrosis with stricture formation, intra-abdominal abscesses, fistulas, and, rarely, free perforation. Inflammation in Crohn’s disease can affect discontinuous portions of intestine—so-called “skip lesions” that are separated by intervening normal-appearing intestine.

A feature of Crohn’s disease that is grossly evident and helpful in identifying affected segments of intestine during surgery is the presence of fat wrapping, which represents encroachment of mesenteric fat onto the serosal surface of the bowel (Fig. 28-17). This finding is virtually pathognomonic of Crohn’s disease. The presence of fat wrapping correlates well with the presence of underlying acute and chronic inflammation.

Features that allow for differentiation between Crohn’s disease of the colon and ulcerative colitis include the layers of the bowel wall affected (inflammation in ulcerative colitis is limited to the mucosa and submucosa but may involve the full thickness of the bowel wall in Crohn’s disease) and the longitudinal extent of inflammation (inflammation is continuous and characteristically affects the rectum in ulcerative colitis but may be discontinuous and spare the rectum in Crohn’s disease). In the absence of full expression of features of advanced disease, Crohn’s colitis can sometimes be difficult to distinguish from ulcerative colitis. It is also important to remember that although ulcerative colitis is a disease of the colon, it can be associated with inflammatory changes in the distal ileum (backwash ileitis).

Clinical Presentation

The most common symptoms of Crohn’s disease are abdominal pain, diarrhea, and weight loss. However, the clinical features are highly variable among individual patients and depend on which segment(s) of the GI tract is (are) predominantly affected, the intensity of inflammation, and the presence or absence of specific complications. Patients with Crohn’s disease can be classified by their predominant clinical manifestation as having primarily (a) fibrostenotic disease, (b) fistulizing disease, or (c) aggressive inflammatory disease. There is substantial overlap among these disease patterns in individual patients, ­however. The onset of symptoms is insidious, and once present, their severity follows a waxing and waning course. Constitutional symptoms, particularly weight loss and fever, or growth retardation in children, may also be prominent and are occasionally the sole presenting features of Crohn’s disease.

The disease affects the small bowel in 80% of cases and the colon alone in 20%. In those with small bowel disease, the majority have ileocecal disease. The small bowel alone is affected in 15% to 30% of patients. Isolated perineal and anorectal disease occurs in 5% to 10% of affected patients. Uncommon sites of involvement include the esophagus, stomach, and duodenum.

An estimated one fourth of all patients with Crohn’s disease will have an extraintestinal manifestation of their disease. One fourth of those affected will have more than one manifestation. Many of these complications are common to both Crohn’s disease and ulcerative colitis, although as a whole, they are more prevalent among patients with Crohn’s disease than those with ulcerative colitis. The most common extraintestinal manifestations are listed in Table 28-7. The clinical severity of some of these manifestations, such as erythema nodosum and peripheral arthritis, are correlated with the severity of intestinal inflammation. The severity of other manifestations, such as pyoderma gangrenosum and ankylosing spondylitis, bear no apparent relationship to the severity of intestinal inflammation.

Diagnosis

The diagnosis is usually established with endoscopic findings in a patient with a compatible clinical history. The diagnosis should be considered in those presenting with acute or chronic abdominal pain, especially when localized to the right lower quadrant, chronic diarrhea, evidence of intestinal inflammation on radiography or endoscopy, the discovery of a bowel stricture or fistula arising from the bowel, and evidence of inflammation or granulomas on intestinal histology. Disorders associated with clinical presentations that resemble those of Crohn’s disease include ulcerative colitis, functional bowel disorders such as irritable bowel syndrome, mesenteric ischemia, collagen vascular diseases, carcinoma and lymphoma, diverticular disease, and infectious enteritides. These infectious enteritides are most frequently diagnosed in immunocompromised patients but can also occur in patients with normal immune function. Acute ileitis caused by Campylobacter and Yersinia species can be difficult to distinguish from that caused by an acute presentation of Crohn’s disease. Typhoid enteritis caused by Salmonella typhosa can lead to overt intestinal bleeding and perforation, most often affecting the terminal ileum. The distal ileum and cecum are the most common sites of intestinal involvement by infection due to Mycobacterium tuberculosis. This condition can result in intestinal inflammation, strictures, and fistula formation, similar to those seen in Crohn’s disease. Cytomegalovirus (CMV) can cause intestinal ulcers, bleeding, and perforation.

No single symptom, sign, or diagnostic test establishes the diagnosis of Crohn’s disease. Instead, the diagnosis is based on a complete assessment of the clinical presentation with confirmatory findings derived from radiographic, endoscopic, and in most cases pathologic tests. Colonoscopy with intubation of terminal ileum is the main diagnostic tool and can reveal focal ulcerations adjacent to areas of normal-appearing mucosa along with polypoid mucosal changes that give a “cobblestone appearance.” Skip areas of involvement are typical, with segments of normal-appearing bowel interrupted by large areas of obvious disease; this pattern is different from the continuous involvement in ulcerative colitis. Pseudopolyps, as seen in ulcerative colitis, are also often present. Contrast examinations of the small bowel and colon may reveal strictures or networks of ulcers and fissures. CT scanning may reveal intra-abdominal abscesses and is useful in acute presentations to rule out the presence of other intra-abdominal disorders. Esophagogastroduodenoscopy (EGD) is done for disease of the proximal alimentary tract. Because Crohn’s disease often affects the small bowel, which is difficult to image, capsule endoscopy has been increasingly used to make this diagnosis (Fig. 28-18).³⁹

Several antibodies have also been identified in patients with inflammatory bowel disease, which may have diagnostic value. The most commonly tested antibodies are antineutrophil cytoplasmic antibody (pANCA) and anti-Saccharomyces cerevisiae antibody (ASCA). ASCA +/pANCA– is associated with a diagnosis of Crohn’s disease, while ASCA–/pANCA+ correlates with ulcerative colitis. Although these antibody tests are commercially available, their widespread use has been hampered by low test sensitivities.

Because of the insidious, and often nonspecific, presentation, the diagnosis of Crohn’s disease is typically made only after symptoms have been present for several years. However, in acute presentations, the diagnosis is sometimes made intraoperatively or during surgical evaluation. The initial manifestation of Crohn’s disease can consist of right lower quadrant abdominal mimicking the presentation of acute appendicitis. In patients with this presentation, Crohn’s disease can be discovered for the first time during laparotomy or laparoscopy performed for presumed appendicitis. In some patients, the initial manifestation of Crohn’s disease is an acute abdomen related to small bowel obstruction, intra-abdominal abscess, or free intestinal perforation. In other patients, perianal abscesses and fistulas requiring surgical therapy may be the first manifestation of Crohn’s disease.

Therapy

Because no curative therapies are available for Crohn’s disease, the goal of treatment is to palliate symptoms rather than to achieve cure. Medical therapy is used to induce and maintain disease remission. Surgery is reserved for specific indications described later. In addition nutritional support in the form of aggressive enteral regimens or, if necessary, parenteral nutrition is used to manage the malnutrition that is common in patients with Crohn’s disease.

Medical Therapy

Pharmacologic agents used to treat Crohn’s disease include antibiotics, aminosalicylates, corticosteroids, and immunomodulators. Antibiotics have an adjunctive role in the treatment of infectious complications associated with Crohn’s disease. They are also used to treat patients with ­perianal disease, enterocutaneous fistulas, and active colonic disease.

Most studies have shown oral 5-aminosalicylic acid (5-ASA) drugs (e.g., mesalamine) to be superior to placebo in inducing disease remission. Their efficacy in the maintenance of remission is less clear. Aminosalicylates are associated with minimal toxicity and are available in a variety of formulations that allow for their delivery to specific regions of the alimentary tract. In Crohn’s disease, sulfasalazine, the parent compound of 5-ASA and widely used in ulcerative colitis, has been shown to be less effective than 5-ASA.

Orally administered glucocorticoids are used to treat patients with mildly to moderately severe disease that does not respond to aminosalicylates. Patients with severe active disease usually require intravenous administration of glucocorticoids. Although glucocorticoids are effective in inducing remission, they are ineffective in preventing relapse, and their adverse effect profile makes long-term use hazardous. Therefore, they should be tapered once remission is achieved. Some patients are unable to undergo glucocorticoid tapering without suffering recurrence of symptoms. Such patients are said to have steroid dependence. In these patients, along with those who do not respond to steroids at all (steroid resistant), use of immune modulators should be considered.

The thiopurine antimetabolites azathioprine and its active metabolite, 6-mercaptopurine, have demonstrated efficacy in inducing remission, in maintaining remission, and in allowing for glucocorticoid tapering in glucocorticoid-dependent patients. A response to these medications is usually observed in 3 to 6 months. There is also some evidence that they decrease the risk of relapse after intestinal resection for Crohn’s disease. These agents are relatively safe but can induce bone marrow suppression and promote infectious complications. For patients who do not respond to the thiopurines, methotrexate is an alternative that is usually given intramuscularly before switching to an oral form after achieving symptomatic control. There is little role for cyclosporine in Crohn’s disease; its efficacy/toxicity profile in this disease is poor.

The successful introduction of infliximab (Remicade), an anti–tumor necrosis factor-α (TNF-α) antibody, heralded the era of biologic therapies for inflammatory bowel disease. Infliximab is a chimeric monoclonal anti-TNFα antibody that has been shown to have efficacy in inducing remission and in promoting closure of enterocutaneous fistulae. It is generally used for patients resistant to standard therapy, in order to help taper steroid dosage. Infliximab is generally well tolerated but should not be used in patients with ongoing septic processes, such as undrained intra-abdominal abscesses. Whereas infliximab is a mouse-human chimeric antibody, the newer drugs in this group include adalimumab (Humira), which is a fully human antibody. Antibodies against other targets in this inflammatory pathway have also been developed and are in various stages of clinical evaluation. Recent studies have shown that these biologic therapies can be used safely after surgery for Crohn’s disease. In a randomized study of 24 patients, those receiving infliximab starting 4 weeks after ileal resection had improved endoscopic and histologic scores at 1-year follow-up compared to those receiving placebo.⁴⁰

For patients with perianal disease, antibiotic therapy with metronidazole or ciprofloxacin is the primary step. Two to 4 weeks of therapy is needed before improvements are seen, and often long-term therapy is required to prevent relapse. In cases of relapse, azathioprine can be considered. In patients with fistulas, infliximab and azathioprine are drugs of choice.

Surgical Therapy

Fifty to seventy percent of patients with Crohn’s disease will ultimately require at least one surgical intervention for their disease.^41,42 Surgery is generally reserved for patients whose disease is unresponsive to aggressive medical therapy or who develop complications of their disease (Table 28-8). Failure of medical management may be the indication for surgery if symptoms persist despite aggressive therapy for several months or if symptoms recur whenever aggressive therapy is tapered. Surgery should be considered if medication-induced complications arise, specifically corticosteroid-related complications, such as cushingoid features, cataracts, glaucoma, systemic hypertension, compression fractures, or aseptic necrosis of the femoral head. Growth retardation constitutes an indication for surgery in 30% of children with Crohn’s disease.

One of the most common indications for surgical intervention is intestinal obstruction. Abscesses and fistulas are frequently encountered during operations performed for intestinal obstruction in these patients, but are rarely the only indication for surgery. Most abscesses are amenable to percutaneous drainage, and fistulas, unless associated with symptoms or metabolic derangements, do not require surgical intervention. Less common complications that require surgical intervention are acute GI hemorrhage, perforations, and development of cancer.

Although most commonly, surgery for Crohn’s disease is planned, an uncommon, but not rare, scenario is the intraoperative discovery of inflammation limited to the terminal ileum during operations performed for presumed appendicitis. This scenario can result from an acute presentation of Crohn’s disease or from acute ileitis caused by bacteria such as Yersinia or Campylobacter. Both conditions should be treated medically; ileal resection is not generally indicated. However, the appendix, even if normal appearing, should be removed (unless the cecum is inflamed, increasing the potential morbidity of this procedure) in order to eliminate appendicitis from the differential diagnosis of abdominal pain in these patients, particularly those with Crohn’s disease who may be destined to have recurring symptoms.

When the diagnosis of Crohn’s disease is known and surgery planned, thorough examination of the entire intestine should be performed. The presence of active disease is suggested by thickening of the bowel wall, narrowing of the lumen, serosal inflammation and coverage by creeping fat, and thickening of the mesentery. Skip lesions are present in approximately 20% of cases and should be sought. The length of uninvolved small intestine should be noted.

Segmental intestinal resection of grossly evident disease followed by primary anastomosis is the usual procedure of choice. Microscopic evidence of Crohn’s disease at the resection margins does not compromise a safe anastomosis, and frozen-section analysis of resection margins is unnecessary. In a randomized prospective trial, the effects of achieving 2-cm resection margins beyond grossly evident disease were compared with achieving 12-cm resection margins.⁴³ There were no evident differences with respect to clinical recurrence rates or anastomotic recurrences. Recurrence rates were similar whether margins were histologically free of or involved with Crohn’s disease. An area of controversy in surgical management of Crohn’s disease has been the ideal anastomotic technique for the bowel after intestinal resection. This issue was addressed in a randomized study of 139 patients undergoing an ileocolic resection for Crohn’s disease, with a mean follow-up of 11.9 months. There were no differences in endoscopic or symptomatic disease recurrence between the groups reconstructed using end-to-end sutured (2-0 PDS) anastomosis vs. those with side-to-side staples anastomosis.⁴⁴

An alternative to segmental resection for obstructing lesions is stricturoplasty (Fig. 28-19). This technique allows for preservation of intestinal surface area and is especially well suited to patients with extensive disease and fibrotic strictures who may have undergone previous resection and are at risk for developing short bowel syndrome. In this technique, the bowel is opened longitudinally to expose the lumen. Any intraluminal ulcerations should be biopsied to rule out the presence of neoplasia. Depending on the length of the stricture, the reconstruction can be fashioned in a manner similar to the Heinecke-Mickulicz pyloroplasty (for strictures <12 cm in length) or the Finney pyloroplasty (for longer strictures as much as 25 cm in length). For longer strictures, variations on the standard stricturoplasty, namely the side-to-side isoperistaltic enteroenterostomy, have been advocated and used for strictures with mean lengths of 50 cm.⁴⁵ Stricturoplasty sites should be marked with metallic clips to facilitate their identification on radiographs and during subsequent operations. Stricturoplasty is associated with recurrence rates that are no different from those associated with segmental resection. Because the affected bowel is left in situ rather than resected, there is the potential for cancer developing at the stricturoplasty site. However, as data on this complication are limited to anecdotes, this risk remains a theoretical one. Stricturoplasty is contraindicated in patients with intra-abdominal abscesses or intestinal fistulas. The presence of a solitary stricture relatively close to a segment for which resection is planned is a relative contraindication. In general, stricturoplasty is performed in cases where single or multiple strictures are identified in diffusely involved segments of bowel or where previous resections have been performed and maintenance of intestinal length is of great importance.

Intestinal bypass procedures are sometimes required in the presence of intramesenteric abscesses or if the diseased bowel is coalesced in the form of a dense inflammatory mass, making its mobilization unsafe. Bypass procedures (gastrojejunostomy) are also used in the presence of duodenal strictures, for which stricturoplasty and segmental resection can be technically difficult.

Since the 1990s, laparoscopic surgical techniques have been applied to patients with Crohn’s disease. The inflammatory changes associated with Crohn’s disease such as thickened and foreshortened mesentery, obliterated tissue planes, and friable tissues with engorged vasculature can make the laparoscopic approach challenging. Randomized studies and a meta-analysis have confirmed that laparoscopic surgery for Crohn’s disease is associated with less postoperative pain, shorter duration of ileus, and a shorter hospital stay. The rates of disease recurrence were similar between the two groups.⁴⁶

Outcomes

Overall complication rates following surgery for Crohn’s disease range from 15% to 30%. Wound infections, postoperative intra-abdominal abscesses, and anastomotic leaks account for most of these complications.

Most patients whose disease is resected eventually develop recurrence. If recurrence is defined endoscopically, 70% recur within 1 year of a bowel resection and 85% by 3 years.⁴⁷ Clinical recurrence, defined as the return of symptoms confirmed as being due to Crohn’s disease, affects 60% of patients by 5 years and 94% by 15 years after intestinal resection. Reoperation becomes necessary in approximately one third of patients by 5 years after the initial operation, with a median time to reoperation of 7 to 10 years.⁴⁸

A fistula is defined as an abnormal communication between two epithelialized surfaces. The communication occurs between two parts of the GI tract or adjacent organs in an internal fistula (e.g., enterocolonic fistula or colovesicular fistula). An external fistula (e.g., enterocutaneous fistula or rectovaginal fistula) involves the skin or another external surface epithelium. Enterocutaneous fistulas that drain less than 200 mL of fluid per day are known as low-output fistulas, whereas those that drain more than 500 mL of fluid per day are known as high-output fistulas.

Over 80% of enterocutaneous fistulas represent iatrogenic complications that occur as the result of enterotomies or intestinal anastomotic dehiscences. Fistulas that arise spontaneously without antecedent iatrogenic injury are usually manifestations of progression of underlying Crohn’s disease or cancer.

Pathophysiology

The manifestations of fistulas depend on which structures are involved. Low-resistance enteroenteric fistulas, which allow luminal contents to bypass a significant proportion of the small intestine, may result in clinically significant malabsorption. Enterovesicular fistulas often cause recurrent urinary tract infections. The drainage emanating from enterocutaneous fistulas are irritating to the skin and cause excoriation. The loss of enteric luminal contents, particularly from high-output fistulas originating from the proximal small intestine, results in dehydration, electrolyte abnormalities, and malnutrition.

Fistulas have the potential to close spontaneously. Factors inhibiting spontaneous closure, however, include malnutrition, sepsis, inflammatory bowel disease, cancer, radiation, obstruction of the intestine distal to the origin of the fistula, foreign bodies, high output, short fistulous tract (<2 cm) and epithelialization of the fistula tract (Table 28-9).

Clinical Presentation

Iatrogenic enterocutaneous fistulas usually become clinically evident between the fifth and tenth postoperative days. Fever, leukocytosis, prolonged ileus, abdominal tenderness, and wound infection are the initial signs. The diagnosis becomes obvious when drainage of enteric material through the abdominal wound or through existing drains occurs. These fistulas are often associated with intra-abdominal abscesses.

Diagnosis

CT scanning following the administration of enteral contrast is the most useful initial test. Leakage of contrast material from the intestinal lumen can be observed. Intra-abdominal abscesses should be sought and drained percutaneously. If the anatomy of the fistula is not clear on CT scanning, a small bowel series or enteroclysis examination can be obtained to demonstrate the fistula’s site of origin in the bowel. This study is also useful to rule out the presence of intestinal obstruction distal to the site of origin. Occasionally, contrast administered into the intestine does not demonstrate the fistula tract. A fistulogram, in which contrast is injected under pressure through a catheter placed percutaneously into the fistula tract, may offer greater sensitivity in localizing the fistula origin.

Therapy

The treatment of enterocutaneous fistulas should proceed through an orderly sequence of steps.⁴⁹

1. Stabilization. Fluid and electrolyte resuscitation is begun. Nutrition is provided, usually through the parenteral route initially. Sepsis is controlled with antibiotics and drainage of abscesses. The skin is protected from the fistula effluent with ostomy appliances or fistula drains.

2. Investigation. The anatomy of the fistula is defined using the studies described earlier.

3. Decision. The available treatment options are considered, and a timeline for conservative measures is determined.

4. Definitive management. This entails the surgical procedure and requires appropriate preoperative planning and surgical experience.

5. Rehabilitation.

The overall objective is to increase the probability of spontaneous closure. Nutrition and time are the key components of this approach. Most patients will require TPN; however, a trial of oral or enteral nutrition should be attempted in patients with low-output fistulas originating from the distal intestine. The somatostatin analogue octreotide is a useful adjunct, particularly in patients with high-output fistulas; its administration reduces the volume of fistula output, thereby facilitating fluid and electrolyte management. Further, octreotide may accelerate the rate at which fistulas close; however, its administration has not clearly been demonstrated to increase the probability of spontaneous closure.

Timing of Surgical Intervention

Most surgeons would pursue 2 to 3 months of conservative therapy before considering surgical intervention. This approach is based on evidence that 90% of fistulas that are going to close do so within 5 weeks and also that surgical intervention after this time period is associated with better outcomes and lower morbidity.⁵⁰

If the fistula fails to resolve during this period, surgery may be required during which the fistula tract, together with the segment of intestine from which it originates, should be resected. Simple closure of the opening in the intestine from which the fistula originates is associated with high recurrence rates. Patients with intestinal fistulas typically have extensive and dense intra-abdominal adhesions. As a result, operations performed for nonhealing fistulas can present formidable challenges. Successful applications of alternative therapies to close intestinal fistulas such as the use of biologic sealants have been reported. The indications for their use remain to be defined.

Outcomes

Over 50% of intestinal fistulas close spontaneously. A useful mnemonic designates factors that inhibit spontaneous closure of intestinal fistulas: “FRIEND” (Foreign body within the fistula tract, Radiation enteritis, Infection/Inflammation at the fistula origin, Epithelialization of the fistula tract, Neoplasm at the fistula origin, Distal obstruction of the intestine) (Table 28-10).

In a recent 23-year old retrospective review of 153 cases of enterocutaneous fistulas that were treated surgically, the majority of fistulas were found to originate from the small bowel and be iatrogenic in nature, with patients having undergone five or more previous abdominal surgeries. Operative repair was associated with a 30-day mortality of approximately 4% and a 1-year mortality of 15%. Morbidity was over 80%. First attempt at surgical repair was successful in 70% of cases, with an overall closure rate of 84% and some patients requiring up to three attempts at surgical repair. The authors identified closure of the abdominal fascia as an important factor in reducing rates of refistulization and postoperative mortality.⁵¹ In another similar study, fistula recurrence rates of 30% were again documented and were independently associated with high-output fistulas and the type of surgical treatment: operations not involving resection of the fistula had a much higher rate of recurrence.⁵²

Adenomas are the most common benign neoplasm of the small intestine. Other benign tumors include fibromas, lipomas, hemangiomas, lymphangiomas, and neurofibromas. The prevalence of small bowel tumors identified at autopsy is 0.2% to 0.3%, which is significantly higher than the rate of operation for small bowel tumors. This suggests that the majority of small bowel tumors are asymptomatic. These lesions are most frequently encountered in the duodenum as incidental findings during esophagogastroduodenoscopic (EGD) examinations (Fig. 28-20). The reported prevalence of duodenal polyps, as detected during EGD performed for other reasons, ranges from 0.3% to 4.6%.⁵³

Benign neoplasms account for 30% to 50% of small bowel tumors and include adenomas, lipomas, hamartomas, and hemangiomas. Primary small bowel cancers are rare, with an estimated incidence of 5300 cases per year in the United States.⁵⁴ Among small bowel cancers, adenocarcinomas comprise 35% to 50% of all cases, carcinoid tumors comprise 20% to 40%, and lymphomas comprise approximately 10% to 15%. In a retrospective review of a large U.S. database (Surveillance, Epidemiology, and End Results) between 1992 and 2006, of a total of 10,945 small intestine cancers, 4315 were neuroendocrine in origin, 3412 were carcinomas, 2023 were lymphomas, and 1084 were sarcomas.⁵⁵ GI stromal tumors (GISTs) are the most common mesenchymal tumors arising in the small intestine and comprise the vast majority of tumors that were formerly classified as leiomyomas, leiomyosarcomas, and smooth muscle tumors of the intestine. The small intestine is frequently affected by metastases from or local invasion by cancers originating at other sites. Melanoma, in particular, is associated with a propensity for metastasis to the small intestine.

Most patients with small-intestinal cancers are in their fifth or sixth decade of life. Reported risk factors for developing small-intestinal cancers include consumption of red meat, ingestion of smoked or cured foods, Crohn’s disease, celiac sprue, hereditary nonpolyposis colorectal cancer (HNPCC), familial adenomatous polyposis (FAP), and Peutz-Jeghers syndrome.

Pathophysiology

The small intestine contains over 90% of the mucosal surface area of the GI tract, but only 1.1% to 2.4% of all GI malignancies. Proposed explanations for the low frequency of small-intestinal neoplasms include (a) dilution of environmental carcinogens in the liquid chyme present in the small-intestinal lumen, (b) rapid transit of chyme, limiting the contact time between carcinogens and the intestinal mucosa, (c) a relatively low concentration of bacteria in small-intestinal chyme and, therefore, a relatively low concentration of carcinogenic products of bacterial metabolism, (d) mucosal protection by secretory IgA and hydrolases such as benzpyrene hydroxylase that may render carcinogens less active, and (e) efficient epithelial cellular apoptotic mechanisms that serve to eliminate clones harboring genetic mutations.

Recent advances have begun to clarify the molecular pathogenesis of small-intestinal adenocarcinomas and GISTs; there has been less progress with respect to the pathogenesis of the other small-intestinal malignancies (see Table 28-10). Small-intestinal adenocarcinomas are believed to arise from pre-existing adenomas through a sequential accumulation of genetic abnormalities in a model similar to that described for the pathogenesis of colorectal cancer. Adenomas are histologically classified as tubular, villous, and tubulovillous. Tubular adenomas have the least aggressive features. Villous adenomas have the most aggressive features and tend to be large, sessile, and located in the second portion of the duodenum. Malignant degeneration has been reported to be present in up to 45% of villous adenomas by the time of diagnosis. Patients with FAP have a nearly 100% cumulative lifetime risk of developing duodenal adenomas that have the potential to undergo malignant transformation. The risk of duodenal cancer in these patients is over 100-fold that in the general population. Indeed, duodenal cancer is the leading cause of cancer-related death among patients with FAP who have undergone colectomy. Patients with Peutz-Jeghers syndrome develop hamartomatous polyps; however, these polyps can contain adenomatous foci that can undergo malignant transformation (Fig. 28-21).

A defining feature of GISTs is their gain-of-function mutation of proto-oncogene KIT, a receptor tyrosine kinase. Pathologic KIT signal transduction is believed to be a central event in GIST pathogenesis. The majority of GISTs have activating mutations in the c-kit protooncogene, which cause KIT to become constitutively activated, presumably leading to persistence of cellular growth or survival signals. Because the interstitial cells of Cajal normally express KIT, these cells have been implicated as the cell of origin for GISTs. KIT expression is assessed by staining the tissues for CD117 antigen, which is part of the KIT receptor, and present in 95% of GISTs.

Clinical Presentation

Most small-intestinal neoplasms are asymptomatic until they become large. Partial small bowel obstruction, with associated symptoms of crampy abdominal pain and distention, nausea, and vomiting, is the most common mode of presentation. Obstruction can be the result of either luminal narrowing by the tumor itself or intussusception, with the tumor serving as the lead point. Hemorrhage, usually indolent, is the second most common mode of presentation.

Physical examination may be unrevealing. Up to 25% of patients with small-intestinal malignancies are reported to have a palpable abdominal mass. Findings of intestinal obstruction are reported to be present in 25% of patients. Fecal occult blood test may be positive. Jaundice secondary to biliary obstruction or hepatic metastasis may be present. Cachexia, hepatomegaly, and ascites may be present with advanced disease.

Although the clinical presentation is usually not specific for tumor type, some general comments are appropriate. Adenocarcinomas, as well as adenomas (from which most are believed to arise), are most commonly found in the duodenum, except in patients with Crohn’s disease, in whom most are found in the ileum. Lesions in the periampullary location can cause obstructive jaundice or pancreatitis. Adenocarcinomas located in the duodenum tend to be diagnosed earlier in their progression than those located in the jejunum or ileum, which are rarely diagnosed prior to the onset of locally advanced or metastatic disease.

Carcinoid tumors of the small intestine are also usually diagnosed after the development of metastatic disease. These tumors are associated with a more aggressive behavior than the more common appendiceal carcinoid tumors. Approximately 25% to 50% of patients with carcinoid tumor–derived liver metastases will develop manifestations of the carcinoid syndrome. These manifestations include diarrhea, flushing, hypotension, tachycardia, and fibrosis of the endocardium and valves of the right heart. Candidate tumor-derived mediators of the carcinoid syndrome such as serotonin, bradykinin, and substance P undergo nearly complete metabolism during first passage through the liver. As a result, symptoms of carcinoid syndrome are rare in the absence of liver metastases.

Lymphoma may involve the small intestine primarily or as a manifestation of disseminated systemic disease. Primary small-intestinal lymphomas are most commonly located in the ileum, which contains the highest concentration of lymphoid tissue in the intestine. Although partial small bowel obstruction is the most common mode of presentation, 10% of patients with small-intestinal lymphoma present with bowel perforation.

Sixty to seventy percent of GISTs are located in the stomach. The small intestine is the second most common site, containing 25% to 35% of GISTs. There appears to be no regional variation in the prevalence of GISTs within the small intestine. GISTs have a greater propensity to be associated with overt hemorrhage than the other small-intestinal malignancies (Fig. 28-22).

Metastatic tumors involving the small intestine can induce intestinal obstruction and bleeding.

Diagnosis

Because of the absent or nonspecific symptoms associated with most small-intestinal neoplasms, these lesions are rarely diagnosed preoperatively. Laboratory tests are nonspecific, with the exception of elevated serum 5-hydroxyindole acetic acid (5-HIAA) levels in patients with the carcinoid syndrome. Elevated carcinoembryonic antigen (CEA) levels are associated with small-intestinal adenocarcinomas, but only in the presence of liver metastases.

Contrast radiography of the small intestine may demonstrate benign and malignant lesions. Enteroclysis is reported to have a sensitivity of over 90% in the detection of small bowel tumors and is the test of choice, particularly for tumors located in the distal small bowel. Upper GI with small bowel follow-through examinations have reported sensitivities ranging from only 30% to 44%.⁵⁵ CT scanning has low sensitivity for detecting mucosal or intramural lesions but can demonstrate large tumors and is useful in the staging of intestinal malignancies. Tumors associated with significant bleeding can be localized with angiography or radioisotope-tagged red blood cell (RBC) scans.

Tumors located in the duodenum can be visualized and biopsied on EGD. In addition, endoscopic ultrasonography (EUS) can offer additional information such as the layers of the intestinal wall involved by the lesion. Occasionally, the distal ileum can be successfully visualized during colonoscopy. Intraoperative enteroscopy can be used to directly visualize small-intestinal tumors beyond the reach of standard endoscopic techniques. Recently, capsule endoscopy and double-balloon endoscopy have been used to evaluate small bowel.

Therapy

Benign neoplasms of the small intestine that are symptomatic should be surgically resected or removed endoscopically, if feasible. Tumors located in the duodenum, including asymptomatic lesions incidentally found during EGD, can pose the greatest therapeutic challenges. These lesions should be biopsied; symptomatic tumors and adenomas, because of their malignant potential, should be removed. In general, duodenal tumors less than 1 cm in diameter are amenable to endoscopic polypectomy. Lesions greater than 2 cm in diameter are technically difficult to remove endoscopically and may need to be removed surgically. Surgical options include transduodenal polypectomy and segmental duodenal resection. Tumors located in the second portion of the duodenum near the ampulla of Vater may require pancreaticoduodenectomy. EUS may offer utility for duodenal tumors ranging in size between 1 and 2 cm in diameter, with those limited to the mucosa being amenable to endoscopic ­polypectomy. Recent studies have shown that endoscopic resection of biopsy-proven benign duodenal periampullary adenomas leads to equivalent efficacy to surgery but with lower morbidity. Adenomas can recur; therefore, surveillance endoscopy is required after these procedures.⁵⁶

Duodenal adenomas occurring in the setting of FAP require an especially aggressive approach to management. Patients with FAP should undergo screening EGD starting sometime during their second or third decade of life. Adenomas detected should be removed endoscopically, if possible, followed by surveillance endoscopy in 6 months and yearly thereafter, in the absence of recurrence. If surgery is required, pancreaticoduodenectomy is generally necessary because adenomas in patients with FAP tend to be multiple and sessile, with a predilection for the periampullary region. Further, localized resections are complicated by high recurrence rates. Given the potential for recurrences in the duodenal remnant following pylorus-preserving pancreaticoduodenectomy, there is rationale for recommending the application of standard pancreaticoduodenectomy in these patients. However, recurrences have been reported even following this procedure; therefore, continuing surveillance is necessary. For most adenocarcinomas of the duodenum, except those in the third or fourth portion of the duodenum where a local resection could be considered, pancreaticoduodenectomy is required.

The surgical therapy of jejunal and ileal malignancies usually consists of wide local resection of the intestine harboring the lesion. For adenocarcinomas, a wide excision of corresponding mesentery is done to achieve regional lymphadenectomy, as is done for adenocarcinomas of the colon. In the presence of locally advanced or metastatic disease, palliative intestinal resection or bypass is performed. Chemotherapy has no proven efficacy in the adjuvant or palliative treatment of small-intestinal adenocarcinomas.

The goal of surgical therapy for carcinoids is resection of all visible disease. Localized small-intestinal carcinoid tumors should be treated with segmental intestinal resection and regional lymphadenectomy. Nodal metastases are unusual with tumors less than 1 cm in diameter but are present with 75% to 90% of tumors larger than 3 cm in diameter. In approximately 30% of cases, multiple small-intestinal carcinoid tumors are present. Therefore, the entire small intestine should be examined before planning extent of resection. In the presence of metastatic disease, tumor debulking should be conducted as it can be associated with long-term survival and amelioration of symptoms of the carcinoid syndrome. Response rates of 30% to50% have been reported with chemotherapy regimens based on agents such as doxorubicin, 5-fluorouracil, and streptozocin. However, none of these regimens is associated with a clearly demonstrable impact on the natural history of disease. Octreotide is the most effective pharmacologic agent for management of symptoms of carcinoid syndrome.

Localized small-intestinal lymphoma should be treated with segmental resection of the involved intestine and adjacent mesentery. If the small intestine is diffusely affected by lymphoma, chemotherapy rather than surgical resection should be the primary therapy. The value to adjuvant chemotherapy after resection of localized lymphoma is controversial.

Small-intestinal GISTs should be treated with segmental intestinal resection. If the diagnosis is known prior to resection, wide lymphadenectomy can be avoided as GISTs are rarely associated with lymph node metastases. GISTs are resistant to conventional chemotherapy agents. Imatinib (Gleevec) is a tyrosine kinase inhibitor with potent activity against the tyrosine kinase KIT and is used in those with metastatic disease. Clinical trials have shown that 80% of patients with unresectable or metastatic GISTs derive clinical benefit from the administration of imatinib, with 50% to 60% having objective evidence of reduction in tumor volume.⁵⁷ Imatinib has shown great promise as a neoadjuvant and adjuvant therapy of GISTs. Studies have emphasized the potential for development of tumor resistance to this agent. In this setting, an alternative tyrosine kinas inhibitor, sunitinib, has been used with good results.⁵⁷

Metastatic cancers affecting the small intestine that are symptomatic should be treated with palliative resection or bypass except in the most advanced cases. Systemic therapy may be offered if effective chemotherapy exists for the primary cancer.

Outcomes

Complete resection of duodenal adenocarcinomas is associated with postoperative 5-year survival rates ranging from 50% to 60%. Complete resection of adenocarcinomas located in the jejunum or ileum is associated with 5-year survival rates of 20% to 30%.⁵⁸ Five-year survival rates of 75% to 95% following resection of localized small-intestinal carcinoid tumors have been reported. In the presence of carcinoid tumor–derived liver metastases, 5-year survival rates of 19% to 54% have been reported. The overall 5-year survival rate for patients diagnosed with intestinal lymphoma ranges from 20% to 40%. For patients with localized lymphoma amenable to surgical resection, the 5-year survival rate is 60%.

The recurrence rate after resection of GISTs averages 35%. The 5-year survival rate after surgical resection has been reported to range from 35% to 60%. Both tumor size and mitotic index are independently correlated with prognosis. Low-grade tumors (mitotic index <10 per high-power field) measuring less than 5 cm in diameter are associated with excellent prognosis.

Radiation therapy is a component of multimodality therapy for many intra-abdominal and pelvic cancers such as those of the cervix, endometrium, ovary, bladder, prostate, and rectum. An undesired side effect of radiation therapy is radiation-induced injury to the small intestine, which can present clinically as two distinct syndromes: acute and chronic radiation enteritis. Acute radiation enteritis is a transient condition that occurs in approximately 75% of patients undergoing radiation therapy for abdominal and pelvic cancers. Chronic radiation is enteritis is inexorable and develops in approximately 5% to15% of these patients.

Pathophysiology

Radiation induces cellular injury directly and through the generation of free radicals. The principal mechanism of radiation-induced cell death is believed to be apoptosis resulting from free radical–induced breaks in double-stranded DNA. Because radiation has its greatest impact on rapidly proliferating cells, the small-intestinal epithelium is acutely susceptible to radiation-induced injury. Pathologic correlates of this acute injury include villus blunting and a dense infiltrate of leukocytes and plasma cells within the crypts. With severe cases, mucosal sloughing, ulceration, and hemorrhage are observed. The intensity of injury is related to the dose of radiation administered, with most cases occurring in patients who have received at least 4500 cGy. Risk factors for acute radiation enteritis include conditions that may limit splanchnic perfusion such as hypertension, diabetes mellitus, coronary artery disease, and restricted mobility of the small intestine due to adhesions. Injury is potentiated by concomitant administration of chemotherapeutic agents such as doxorubicin, 5-fluorouracil, dactinomycin, and methotrexate that act as radiation sensitizers. Because of the intestinal epithelium’s capacity for regeneration, the mucosal injury that is characteristic of acute radiation enteritis resolves after the cessation of radiation therapy.

In contrast, chronic radiation enteritis is characterized by a progressive occlusive vasculitis that leads to chronic ischemia and fibrosis that affects all layers of the intestinal wall, rather than the mucosa alone. These changes can lead to strictures, abscesses, and fistulas, which are responsible for the clinical manifestations of chronic radiation enteritis.

Clinical Presentation

The most common manifestations of acute radiation enteritis are nausea, vomiting, diarrhea, and crampy abdominal pain. Symptoms are generally transient and subside after the discontinuation of radiation therapy. Because the diagnosis is usually obvious, given the clinical context, no specific diagnostic tests are required. However, if patients develop signs suggestive of peritonitis, CT scanning should be performed to rule out the presence of other conditions capable of causing acute abdominal syndromes.

The clinical manifestations of chronic radiation enteritis usually become evident within 2 years of radiation administration, although they can begin as early as several months or as late as decades afterward. The most common clinical presentation is one of partial small bowel obstruction, with nausea, vomiting, intermittent abdominal distention, crampy abdominal pain, and weight loss being the most common symptoms. The terminal ileum is the most frequently affected segment. Other manifestations of chronic radiation enteritis include complete bowel obstruction, acute or chronic intestinal hemorrhage, and abscess or fistula formation.

Diagnosis

Evaluation of patients suspected of having chronic radiation enteritis should include review of the records of their radiation treatments for information on total radiation dose administered, fractionation, and volume of treatment. Areas that received high doses should be noted, as lesions subsequently found in imaging studies usually localize to areas that had received high radiation doses. Enteroclysis is the most accurate imaging test for diagnosing chronic radiation enteritis, with reported sensitivities and specificities of over 90% (Fig. 28-23). CT scan findings are neither very sensitive nor specific for chronic radiation enteritis. However, CT scanning should be obtained to rule out the presence of recurrent cancer since its clinical manifestations may overlap with those of chronic radiation enteritis.

Therapy

Most cases of acute radiation enteritis are self-limited. Supportive therapy, including the administration of antiemetics, is usually sufficient. Patients with diarrhea-induced dehydration may require hospital admission and parenteral fluid administration. Rarely are symptoms severe enough to necessitate reduction in or cessation of radiation therapy.

In contrast, the treatment of chronic radiation enteritis represents a formidable challenge. Surgery for this condition is difficult, is associated with high morbidity rates, and should be avoided in the absence of specific indications such as high-grade obstruction, perforation, hemorrhage, intra-abdominal abscesses, and fistulas. The goal of surgery is limited resection of diseased intestine with primary anastomosis between healthy bowel segments. However, the characteristically diffuse nature of fibrosis and dense adhesions among bowel segments can make limited resection difficult to achieve. Further, it is difficult to distinguish between normal and irradiated intestine intraoperatively by either gross inspection or even frozen section analysis. This distinction is important as anastomoses between irradiated segments of intestine have been associated with leak rates as high as 50%.⁵⁹ If limited resection is not achievable, an intestinal bypass procedure may be an option, except in cases for which hemorrhage is the surgical indication. There remain cases in which resections extensive enough to cause short bowel syndrome are unavoidable. This condition is discussed in detail later in the Short Bowel Syndrome section.

Outcomes

Acute radiation injury to the intestine is self-limited; its severity is not correlated with the probability of chronic radiation enteritis developing. Surgery for chronic radiation enteritis is associated with high morbidity rates and reported mortality rates averaging 10%.

Prevention

In view of significant morbidity associated with radiation enteritis, groups have studied possible measures to reduce or prevent such side effects. Keeping radiation exposure to below 5000 cGy is associated with minimal long-term side effects and recommended where clinically possible.

Multi-beam radiation techniques to minimize the area of maximal radiation exposure and tilt tables to move the bowel out of the pelvis during radiation are increasingly used. Few small studies have suggested that oral sulfasalazine may help reduce the incidence of acute radiation-induced enteritis.⁶⁰

In patients undergoing pelvic surgery who are likely to require postoperative radiation therapy, surgical techniques that keep the small bowel out of the pelvis have been recommended. These measures include use of absorbable mesh sling to separate the pelvis from the true abdominal cavity and prevent the small bowel from being exposed to pelvic radiation.⁶¹

Meckel’s diverticulum is the most prevalent congenital anomaly of the GI tract, affecting approximately 2% of the general population. Meckel’s diverticula are designated true diverticula because their walls contain all of the layers found in normal small intestine. Their location varies among individual patients, but they are usually found in the ileum within 100 cm of the ileocecal valve (Fig. 28-24). Approximately 60% of Meckel’s diverticula contain heterotopic mucosa, of which over 60% consist of gastric mucosa. Pancreatic acini are the next most common; others include Brunner’s glands, pancreatic islets, colonic mucosa, endometriosis, and hepatobiliary tissues. A useful, although crude, mnemonic describing Meckel’s diverticula is the “rule of two’s”: 2% prevalence, 2:1 male predominance, location 2 feet proximal to the ileocecal valve in adults, and half of those who are symptomatic are under 2 years of age.

Pathophysiology

During the eighth week of gestation, the omphalomesenteric (vitelline) duct normally undergoes obliteration. Failure or incomplete vitelline duct obliteration results in a spectrum of abnormalities, the most common of which is Meckel’s diverticulum. Other abnormalities include omphalomesenteric fistula, enterocyst, and a fibrous band connecting the intestine to the umbilicus. A remnant of the left vitelline artery can persist to form a mesodiverticular band tethering a Meckel’s diverticulum to the ileal mesentery.

Bleeding associated with Meckel’s diverticulum is usually the result of ileal mucosal ulceration that occurs adjacent to acid-producing, heterotopic gastric mucosa located within the diverticulum. Intestinal obstruction associated with Meckel’s diverticulum can result from several mechanisms:

1. Volvulus of the intestine around the fibrous band attaching the diverticulum to the umbilicus

2. Entrapment of intestine by a mesodiverticular band (Fig. 28-25)

3. Intussusception with the diverticulum acting as a lead point

4. Stricture secondary to chronic diverticulitis

Meckel’s diverticula can be found in inguinal or femoral hernia sacs (known as Littre’s hernia). These hernias, when incarcerated, can cause intestinal obstruction.

Clinical Presentation

Meckel’s diverticula are asymptomatic unless associated complications arise. The lifetime incidence rate of complications arising in patients with Meckel’s diverticula has been estimated to be approximately 4% to 6%.^62,63 Although initial data had suggested that the risk of developing a complication related to Meckel’s diverticulum decreases with age, this has been now been questioned. In a population-based review at Olmsted County, MN, Cullen and colleagues showed that the risk of developing Meckel’s-related complications does not change with age.⁶³

The most common presentations associated with symptomatic Meckel’s diverticula are bleeding, intestinal obstruction, and diverticulitis. Bleeding is the most common presentation in children with Meckel’s diverticula, representing over 50% of Meckel’s diverticulum–related complications among patients less than 18 years of age. Bleeding associated with Meckel’s diverticula is rare among patients older than 30 years of age.

Intestinal obstruction is the most common presentation in adults with Meckel’s diverticula. Diverticulitis, present in 20% of patients with symptomatic Meckel’s diverticula, is associated with a clinical syndrome that is indistinguishable from acute appendicitis. Neoplasms, most commonly carcinoid tumors, are present in 0.5% to 3.2% of symptomatic Meckel’s diverticula that are resected.⁵⁶

Diagnosis

Most Meckel’s diverticula are discovered incidentally on radiographic imaging, during endoscopy, or at the time of surgery. In the absence of bleeding, Meckel’s diverticula rarely are diagnosed prior to the time of surgical intervention. For those presenting with symptoms suggestive of a Meckel’s diverticulum, confirmatory imaging can be challenging. The sensitivity of CT scanning for the detection of Meckel’s diverticula is too low to be clinically useful. Enteroclysis is associated with an accuracy of 75% but is usually not applicable during acute presentations of complications related to Meckel’s diverticula. Radionuclide scans (^99mTc-pertechnetate) can be helpful in the diagnosis of Meckel’s diverticulum; however, this test is positive only when the diverticulum contains associated ectopic gastric mucosa that is capable of uptake of the tracer (Fig. 28-26). The accuracy of radionuclide scanning is reported to be 90% in pediatric patients but less than 50% in adults. Angiography can localize the site of bleeding during acute hemorrhage related to Meckel’s ­diverticula.

Therapy

The surgical treatment of symptomatic Meckel’s diverticula should consist of diverticulectomy with removal of associated bands connecting the diverticulum to the abdominal wall or intestinal mesentery. If the indication for diverticulectomy is bleeding, segmental resection of ileum that includes both the diverticulum and the adjacent ileal peptic ulcer should be performed. Segmental ileal resection may also be necessary if the diverticulum contains a tumor or if the base of the diverticulum is inflamed or perforated.

The management of incidentally found (asymptomatic) Meckel’s diverticula is controversial. Until recently, most authors recommended against prophylactic removal of asymptomatic Meckel’s diverticula, given the low lifetime incidence of complications. More recently, greater enthusiasm for prophylactic diverticulectomy has appeared in the literature.⁶³ Proponents of this approach cite the minimal morbidity associated with removing Meckel’s diverticula and the possibility that previous estimates of the lifetime incidence of complications related to Meckel’s diverticula may be erroneously low. Other authors have advocated a selective approach, with a recommendation to remove diverticula attached by bands and those with narrow bases, on the assumption that these diverticula are more likely to develop complications. No controlled data supporting or refuting these recommendations exist.

Acquired diverticula are designated false diverticula because their walls consist of mucosa and submucosa but lack a complete muscularis. Acquired diverticula are more common in the duodenum and tend to be located near the ampulla; such diverticula are known as periampullary, juxtapapillary, or perivaterian ­diverticula. Approximately 75% of juxtapapillary diverticula arise on the medial wall of the duodenum. Acquired diverticula in the jejunum or ileum are known as jejunoileal diverticula. Eighty percent of jejunoileal diverticula are localized to the jejunum, 15% to the ileum, and 5% to both jejunum and ileum. Diverticula in the jejunum tend to be large and accompanied by multiple other diverticula, whereas those in the ileum tend to be small and solitary.

The prevalence of duodenal diverticula, as detected on upper GI examinations (Fig. 28-27), has been reported to range from 0.16% to 6%.⁶⁴ Their prevalence, as detected during endoscopic retrograde cholangiography (ERCP) examinations, has been reported to range from 5% to 27%. A 23% prevalence rate has been reported in an autopsy series. The prevalence of duodenal diverticula increases with age; they are rare in patients under the age of 40 years. The mean age of diagnosis ranges from 56 to 76 years.

The prevalence of jejunoileal diverticula (Fig. 28-28) has been estimated to range from 1% to 5%.⁶⁵ Their prevalence increases with age; most patients diagnosed with these diverticula are in the sixth and seventh decades of life.

Pathophysiology

The pathogenesis of acquired diverticula is hypothesized to be related to acquired abnormalities of intestinal smooth muscle or dysregulated motility, leading to herniation of mucosa and submucosa through weakened areas of muscularis.

Acquired diverticula can be associated with bacterial overgrowth, leading to vitamin B₁₂ deficiency, megaloblastic anemia, malabsorption, and steatorrhea. Periampullary duodenal diverticula have been described to become distended with intraluminal debris and to compress the common bile duct or pancreatic duct, thus causing obstructive jaundice or pancreatitis, respectively. Jejunoileal diverticula can also cause intestinal obstruction through intussusception or compression of adjacent bowel.

Clinical Presentation

Acquired diverticula are asymptomatic unless associated complications arise. Such complications are estimated to occur in 6% to 10% of patients with acquired diverticula and include intestinal obstruction, diverticulitis, hemorrhage, perforation, and malabsorption. Periampullary duodenal diverticula may be associated with choledocholithiasis, cholangitis, recurrent pancreatitis, and sphincter of Oddi dysfunction. However, a clear link between the presence of the diverticula and the development of these ­conditions has not been demonstrated. Symptoms such as intermittent abdominal pain, flatulence, diarrhea, and constipation are reported to be present in 10% to 30% of patients with jejunoileal diverticula. The relationship between these symptoms and the presence of diverticula is similarly unclear.

Diagnosis

Most acquired diverticula are discovered incidentally on radiographic imaging, during endoscopy, or at the time of surgery. On ultrasound and CT scanning, duodenal diverticula may be mistaken for pancreatic pseudocysts and fluid collections, biliary cysts, and periampullary neoplasms. These lesions can be missed on endoscopy, particularly with forward-viewing endoscopes, and are best diagnosed on upper GI radiographs. Enteroclysis is the most sensitive test for detecting jejunoileal diverticula.

Therapy

Asymptomatic acquired diverticula should be left alone. Bacterial overgrowth associated with acquired diverticula is treated with antibiotics. Other complications, such as bleeding and diverticulitis, are treated with segmental intestinal resection for diverticula located in the jejunum or ileum.

Bleeding and obstruction related to lateral duodenal diverticula are generally treated with diverticulectomy alone. These procedures can be technically difficult for medial duodenal diverticula that penetrate into the substance of the pancreas. Complications related to these medial duodenal diverticula should be managed nonoperatively, if possible, using endoscopy. In emergent situations, bleeding related to medial duodenal diverticula can be controlled using a lateral duodenotomy and oversewing of the bleeding vessel. Similarly, perforation can be managed with wide drainage rather than complex surgery. Whether diverticulectomy should be done in patients with biliary or pancreatic symptoms is controversial and is not routinely recommended.⁵⁸

Mesenteric ischemia can present as one of two distinct clinical syndromes: acute mesenteric ischemia and chronic mesenteric ischemia.

Four distinct pathophysiologic mechanisms can lead to acute mesenteric ischemia:

1. Arterial embolus

2. Arterial thrombosis

3. Vasospasm (also known as nonocclusive mesenteric ischemia [NOMI])

4. Venous thrombosis

Embolus is the most common cause of acute mesenteric ischemia and is responsible for over 50% of cases. The embolic source is usually in the heart, most often the left atrial or ventricular thrombi or valvular lesions. Indeed, up to 95% of patients with acute mesenteric ischemia due to emboli will have a documented history of cardiac disease. Embolism to the superior mesenteric artery accounts for 50% of cases; most of these emboli become wedged and cause occlusion at branch points in the mid to distal superior mesenteric artery, usually distal to the origin of the middle colic artery. In contrast, acute occlusions due to thrombosis tend to occur in the proximal mesenteric arteries, near their origins. Acute thrombosis is usually superimposed on pre-existing atherosclerotic lesions at these sites. NOMI is the result of vasospasm and is usually diagnosed in critically ill patients receiving vasopressor agents.

Mesenteric venous thrombosis accounts for 5% to 15% of cases of acute mesenteric ischemia and involves the superior mesenteric vein in 95% of cases.⁶⁶ The inferior mesenteric vein is only rarely involved. Mesenteric venous thrombosis is classified as primary if no etiologic factor is identifiable or as secondary if an etiologic factor, such as heritable or acquired coagulation disorders, is identified.

Regardless of the pathophysiologic mechanism, acute mesenteric ischemia can lead to intestinal mucosal sloughing within 3 hours of onset and full-thickness intestinal infarction by 6 hours.

In contrast, chronic mesenteric ischemia develops insidiously, allowing for development of collateral circulation, and, therefore, rarely leads to intestinal infarction. Chronic mesenteric arterial ischemia results from atherosclerotic lesions in the main splanchnic arteries (celiac, superior mesenteric, and inferior mesenteric arteries). In most patients with symptoms attributable to chronic mesenteric ischemia, at least two of these arteries are either occluded or severely stenosed. A chronic form of mesenteric venous thrombosis can involve the portal or splenic veins and may lead to portal hypertension, with resulting esophagogastric varices, splenomegaly, and hypersplenism.

Severe abdominal pain, out of proportion to the degree of tenderness on examination, is the hallmark of acute mesenteric ischemia, regardless of the pathophysiologic mechanism. The pain is typically perceived to be colicky and most severe in the mid-abdomen. Associated symptoms can include nausea, vomiting, and diarrhea. Physical findings are characteristically absent early in the course of ischemia. With the onset of bowel infarction, abdominal distension, peritonitis, and passage of bloody stools occur.

Chronic mesenteric ischemia presents insidiously. Postprandial abdominal pain is the most prevalent symptom, producing a characteristic aversion to food (“food fear”) and weight loss. These patients are often thought to have a malignancy and suffer a prolonged period of symptoms before the correct diagnosis is made.

Most patients with chronic mesenteric venous thrombosis are asymptomatic because of the presence of extensive collateral venous drainage routes; this condition is usually discovered as an incidental finding on imaging studies. However, some patients with chronic mesenteric venous thrombosis present with bleeding from esophagogastric varices.

The diagnosis and management of these disorders, which are of primary vascular origin, are discussed in the section on mesenteric artery occlusive disease in the chapter on arterial disease.

Obscure Gastrointestinal Bleeding

Up to 90% of lesions responsible for GI bleeding are within the reach EGD and colonoscopy. Obscure GI bleeding refers to GI bleeding for which no source has been identified by routine endoscopic studies (EGD and colonoscopy). Overt GI bleeding refers to the presence of hematemesis, melena, or hematochezia. In contrast, occult GI bleeding occurs in the absence of overt bleeding and is identified on laboratory tests (e.g., iron-­deficiency anemia) or examination of the stool (e.g., positive guaiac test). Obscure GI bleeding is occult in 20% of cases.⁶⁷

Obscure bleeding can be frustrating for both the patient and the clinician, and this is particularly true for obscure-overt bleeding, which cannot be localized despite aggressive diagnostic measures. One study from a tertiary referral center reported that the typical patient with obscure-overt bleeding had suffered intermittent episodes of hemorrhage for 26 months, had undergone up to 20 diagnostic tests, and had received an average of 20 units of blood before reaching a diagnosis.⁶⁸ Most of the small bowel is beyond the reach of these examinations and, hence, contains most lesions responsible for obscure GI bleeding. Small-intestinal angiodysplasias account for approximately 75% of cases in adults, and neoplasms account for approximately 10%. Meckel’s diverticulum is the most common etiology of obscure GI bleeding in children. Other etiologies of obscure GI bleeding include Crohn’s disease, infectious enteritides, nonsteroidal anti-inflammatory drug (NSAID)-induced ulcers and erosions, vasculitis, ischemia, varices, diverticula, and intussusception.

The diagnostic evaluation of patients with obscure GI bleeding should be tailored to the severity of bleeding and to the availability of technology and expertise. Enteroscopy is playing an increasingly important role. Several endoscopic techniques for visualizing the small intestine are available: push enteroscopy, Sonde enteroscopy, intraoperative enteroscopy, double-balloon endoscopy, and wireless capsule enteroscopy.

Push enteroscopy entails advancing a long endoscope (such as a pediatric or adult colonoscope or a specialized instrument) beyond the ligament of Treitz into the proximal jejunum. This procedure can allow for visualization of approximately 60 cm of the proximal jejunum. Diagnostic yield in patients with obscure GI bleeding ranges from 3% to 65%. In addition to diagnosis, push enteroscopy allows for cauterization of ­bleeding sites.

In Sonde enteroscopy, a long, thin fiberoptic instrument is propelled through the intestine by peristalsis following inflation of a balloon at the instrument’s tip. Visualization is done during instrument withdrawal; approximately 50% to 75% of the small-intestinal mucosa can be examined. However, this instrument lacks biopsy or therapeutic capability. Further, it lacks tip deflection capability, limiting complete mucosal visualization, and has therefore been abandoned in favor of capsule endoscopy.

Wireless capsule enteroscopy relies on a radiotelemetry capsule enteroscope that is small enough to swallow and has no external wires, fiberoptic bundles, or cables. While the capsule is being propelled through the intestine by peristalsis, video images are transmitted using radiotelemetry to an array of detectors attached to the patient’s body. These detectors capture the images and permit continuous triangulation of the capsule location in the abdomen, facilitating the localization of lesions detected. The entire system is portable, allowing the patient to be ambulatory during the entire examination. Capsule endoscopy is an excellent tool in the patient who is hemodynamically stable but continues to bleed. This technique has reported success rates as high as 90% in identifying a small bowel pathology. Although many studies have shown an improved diagnostic yield for small bowel lesions with capsule endoscopy, the impact of this new imaging modality on actual patient outcomes was not studied until recently. In a randomized study of patients with obscure GI bleeding, evaluation with capsule endoscopy vs. small bowel contrast study had a much higher diagnostic yield (30% vs. 7%, respectively); however, this did not translate into an improvement in outcomes. The rates of rebleeding, hospitalization, need for blood transfusion, and therapeutic interventions were similar between the two arms.

The inability to perform biopsies or carry out any therapeutic interventions of capsule endoscope likely prevents the improved diagnostic yield of the test from translating into improved patient outcomes and highlights the continuing challenge with evaluation of the small bowel.⁶⁹

For patients in whom bleeding from an obscure GI source has apparently stopped, push enteroscopy or capsule enteroscopy is a reasonable initial study. If these examinations do not reveal a potential source of bleeding, then enteroclysis should be performed. Standard small bowel follow-through examinations are associated with a low diagnostic yield in this setting and should be avoided. If still no diagnosis has been made, a “watch-and-wait” approach is reasonable, although angiography should be considered if the prior episode of bleeding was overt. Angiography can reveal angiodysplasia and vascular tumors in the small intestine even in the absence of ongoing bleeding.

For persistent mild bleeding from an obscure GI source, push and capsule enteroscopy can be used. If these examinations are nondiagnostic, then ^99m Tc-labeled RBC scanning should be performed and, if positive, followed by angiography to localize the source of bleeding. ^99m Tc-pertechnetate scintigraphy to diagnose Meckel’s diverticulum should be considered, although its yield in patients older than 40 years of age is extremely low. Patients who remain undiagnosed but continue to bleed and those with recurrent episodic bleeding significant enough to require blood transfusions should then undergo exploration with intraoperative enteroscopy.

Patients with persistent severe bleeding from an obscure source should undergo angiography to help localize the bleeding source. Therapy can be tailored based on the source. Push enteroscopy can also be attempted, but capsule enteroscopy is too slow to be applicable in this setting. If these examinations fail to localize the source of bleeding, exploratory laparoscopy or laparotomy with intraoperative enteroscopy is indicated. Intraoperative enteroscopy can be done during either laparotomy or laparoscopy. An endoscope (usually a colonoscope) is inserted into the small bowel through peroral intubation or through an enterotomy made in the small bowel or cecum. The endoscope is advanced by successively telescoping short segments of intestine onto the end to the instrument. In addition to the endoscopic image, the transilluminated bowel should be examined externally with the operating room lights dimmed, as this maneuver may facilitate the identification of angiodysplasias. Identified lesions should be marked with a suture placed on the serosal surface of the bowel; these lesions can be resected after completion of endoscopy. Examination should be performed during instrument insertion rather than withdrawal because instrument-induced mucosal trauma can be confused with angiodysplasias.

Fig. 28-29 provides a diagnostic and management algorithm for patients with obscure GI bleeding.

Small Bowel Perforation

Prior to the 1980s, duodenal perforation due to peptic ulcer disease was the most common form of small bowel perforation. Today, iatrogenic injury incurred during GI endoscopy is the most common cause of small bowel perforation. Other etiologies of small bowel perforation include infections (especially tuberculosis, typhoid, and CMV), Crohn’s disease, ischemia, drugs (e.g., potassium- and NSAID-induced ulcers), radiation-induced injury, Meckel’s and acquired diverticula, neoplasms (especially lymphoma, adenocarcinoma, and melanoma), and foreign bodies.

Among iatrogenic injuries, duodenal perforation during ERCP with endoscopic sphincterotomy (ES) is the most common. This complication occurs in 0.3% to 2.1% of cases. Patients who have undergone Billroth II gastrectomy are at increased risk of duodenal perforations as well as free jejunal perforations during ERCP. Although ERCP-related duodenal perforations can result in a free perforation, most are retroperitoneal. Manifestations of such contained duodenal perforations following ERCP can resemble those of ERCP-induced pancreatitis, including hyperamylasemia.

CT scanning is the most sensitive test for diagnosing duodenal perforations; positive findings include pneumoperitoneum for free perforations, but more commonly retroperitoneal air, contrast extravasation, and paraduodenal fluid collections. If all patients undergoing a therapeutic ERCP are imaged with a CT scan following the procedure, up to 30% will have evidence of air in the retroperitoneum, but the majority are asymptomatic. These patients do not require any specific therapy.⁷⁰

True cases of retroperitoneal perforations of the duodenum can be managed nonoperatively, in the absence of progression and sepsis. However, intraperitoneal duodenal perforations require surgical repair with pyloric exclusion and gastrojejunostomy or tube duodenostomy. Iatrogenic small bowel perforation incurred during endoscopy, if immediately recognized, can sometimes be repaired using endoscopic techniques.

Perforation of the jejunum and ileum occurs into the peritoneal cavity and usually causes overt symptoms and signs, such as abdominal pain, tenderness, and distention accompanied by fever and tachycardia. Plain abdominal radiographs may reveal free intraperitoneal air if intraperitoneal perforation has occurred. If perforation is suspected but not clinically obvious, CT scanning should be performed. Jejunal and ileal perforations require surgical repair or segmental resection.

Chylous Ascites

Chylous ascites refers to the accumulation of triglyceride-rich peritoneal fluid with a milky or creamy appearance, caused by the presence of intestinal lymph in the peritoneal cavity. ­Chylomicrons, produced by the intestine and secreted into lymph during the absorption of long-chain fatty acids, account for the characteristic appearance and triglyceride content of chyle.

The most common etiologies of chylous ascites in Western countries are abdominal malignancies and cirrhosis. In Eastern and developing countries, infectious etiologies, such as tuberculosis and filariasis, account for most cases. Chylous ascites can also develop as a complication of abdominal and thoracic operations and trauma. Operations particularly associated with this complication include abdominal aortic aneurysm repair, retroperitoneal lymph node dissection, inferior vena cava resection, and liver transplantation. Other etiologies of chylous ascites include congenital lymphatic abnormalities (e.g., primary lymphatic hypoplasia), radiation, pancreatitis, and right-sided heart failure.

Three mechanisms have been postulated to cause chylous ascites: (a) exudation of chyle from dilated lymphatics on the wall of the bowel and in the mesentery caused by obstruction of lymphatic vessels at the base of the mesentery or the cisterna chili (e.g., by malignancies); (b) direct leakage of chyle through a lymphoperitoneal fistula (e.g., those that develop as a result of trauma or surgery); and (c) exudation of chyle through the wall of dilated retroperitoneal lymphatic vessels (e.g., in congenital lymphangiectasia or thoracic duct obstruction).

Patients with chylous ascites develop abdominal distention over a period of weeks to months. Postoperative chylous ascites can present acutely during the first postoperative week. Delayed presentations following surgery can occur if the mechanism of ascites formation is adhesion-induced lymphatic obstruction rather than lymphatic vessel disruption. Dyspnea may result if abdominal distention is severe enough.

Paracentesis is the most important diagnostic test. Chyle typically has a cloudy and turbid appearance; however, it may be clear in fasting patients (such as those in the immediate postoperative period). Fluid triglyceride concentrations above 110 mg/dL are diagnostic. CT scanning may be useful in identifying pathologic intra-abdominal lymph nodes and masses and in identifying extent and localization of fluid. Lymphangiography and lymphoscintigraphy may help localize lymph leaks and obstruction; this information is particularly useful for surgical planning.

There are few data on optimal management of patients with chylous ascites. The general approach is to focus on evaluating and treating the underlying causes, especially for patients with infectious, inflammatory, or hemodynamic etiologies for this condition.

Most patients respond to administration of a high-protein and low-fat diet supplemented with medium-chain triglycerides. This regimen is designed to minimize chyle production and flow. Medium-chain triglycerides are absorbed by the intestinal epithelium and are transported to liver through the portal vein; they do not contribute to chylomicron formation.

Patients who do not respond to this approach should be fasted and placed on TPN. Octreotide can further decrease lymph flow. Paracentesis is indicated for respiratory difficulties related to abdominal distention. Overall, two thirds of patients will respond to conservative therapy. However, one third of patients will require surgical therapy for chylous ascites. In general, postoperative and trauma-related cases that fail to respond to initial nonoperative therapy are best managed by surgical repair. Lymphatic leaks are localized and repaired with fine nonabsorbable sutures. If extravasation of chyle is localized to the periphery of the small bowel mesentery, then a limited small bowel resection can be performed instead. For patients who are poor surgical candidates and do not respond to prolonged conservative therapy, peritoneovenous shunting may be an option. However, these shunts are associated with high rates of complications, including sepsis and disseminated intravascular coagulation. Because of the viscosity of chyle, these shunts are associated with a high occlusion rate.

Intussusception

Intussusception refers to a condition where one segment of the intestine becomes drawn in to the lumen of the proximal segment of the bowel. It is usually seen in the pediatric population, where the cecum intussuscepts into the ileum (ileocolic intussusception). In children, it is often an idiopathic condition and treated nonsurgically by radiologic reduction.

Adult intussusceptions are far less common and usually have a distinct pathologic lead point, which can be malignant in up to one half of cases.⁷¹ They commonly present with a history of intermittent abdominal pain and signs and symptoms of bowel obstruction. CT scan is the investigation of choice, where a “target sign” may be seen (Fig. 28-30). Treatment is surgical resection of the involved segment and the lead point, which needs to undergo pathologic evaluation to rule out an underlying malignancy.

With increasing use of CT imaging, target signs are sometimes seen on CT scans of patients who do not have a clinical presentation indicative of bowel obstruction. In such cases, the finding is of little clinical significance and is probably related to normal peristalsis.

In patients who have undergone a Roux-en-Y gastric bypass surgery, an atypical form of intussusception has been increasingly described. In these cases, the proximal bowel is drawn in to the lumen of the distal bowel (retrograde intussusception). These intussusceptions are usually not associated with a lead point and may represent a motility disorder of the bowel following the Roux-en-Y reconstruction. Surgical reductions without resection have been successfully reported in these patients.⁷²

Pneumatosis Intestinalis

Pneumatosis intestinalis indicates the presence of gas within the bowel wall. It may affect any region of the GI tract, but is most commonly seen in the jejunum. Pneumatosis intestinalis is not a disease but merely a sign that can be idiopathic or associated with many intestinal or nonintestinal disorders such as obstructive pulmonary disease and asthma. Most cases of pneumatosis intestinalis are secondary to an identifiable cause, and 15% are idiopathic. The pathogenesis of pneumatosis intestinalis is not fully understood.

The surgical interest in this finding is the association of it with bowel ischemia and infarction, both of which necessitate emergent surgical intervention (see Fig. 28-15). Thus, patients with this radiologic finding need to be fully evaluated and monitored closely to rule out such intra-abdominal catastrophes.

Intestinal resection is performed for many of the diseases discussed in this chapter and generally is associated with minimal morbidity. However, when extent of resection is great enough, a devastating condition known as short bowel syndrome may result. Short bowel syndrome has been arbitrarily defined as the presence of less than 200 cm of residual small bowel in adult patients.⁷³ A functional definition, in which insufficient intestinal absorptive capacity results in the clinical manifestations of diarrhea, dehydration, and malnutrition, is more broadly applicable.

In adults, the most common etiologies of short bowel syndrome are acute mesenteric ischemia, malignancy, and Crohn’s disease. Seventy-five percent of cases result from resection of a large amount of small bowel at a single operation. Twenty-five percent of cases result from the cumulative effects of multiple operations during which small intestine is resected. This latter pattern is typical of patients with Crohn’s disease who develop short bowel syndrome; the former is typical of patients with acute mesenteric ischemia who develop intestinal infarction. In pediatric patients, intestinal atresias, volvulus, and necrotizing enterocolitis are the most common etiologies of short bowel syndrome.

The prevalence of short bowel syndrome is hard to estimate. The most recent available data on chronic home total TPN administration were obtained in 1992. At that time, approximately 40,000 patients were receiving TPN chronically at home. The most prevalent indication for TPN among these patients was short bowel syndrome; however, this estimate does not include patients with short bowel syndrome not receiving TPN at home. It also fails to include those who have been weaned off of TPN.⁷⁴

Pathophysiology

Resection of less than 50% of the small intestine is generally well tolerated. However, clinically significant malabsorption occurs when greater than 50% to 80% of the small intestine has been resected. Among adult patients who lack a functional colon, lifelong TPN dependence is likely to persist if there is less than 100 cm of residual small intestine. Among adult patients who have an intact and functional colon, lifelong TPN dependence is likely to persist if there is less than 60 cm of residual small intestine. Among infants with short bowel syndrome, weaning from TPN dependence has been achieved with as little as 10 cm of residual small intestine.

Residual bowel length is not the only factor predictive of achieving independence from TPN (enteral autonomy), however. Other determinants of the severity of malabsorption include the presence or absence of an intact colon, as indicated earlier. The colon has the capacity to absorb large fluid and electrolyte loads. In addition, the colon can play an important, albeit small, role in nutrient assimilation by absorbing short-chain fatty acids. Second, an intact ileocecal valve is believed to be associated with decreased malabsorption. The ileocecal valve delays transit of chyme from the small intestine into the colon, thereby prolonging the contact time between nutrients and the small-intestinal absorptive mucosa. Third, healthy, rather than diseased, residual small intestine is associated with decreased severity of malabsorption. Fourth, resection of jejunum is better tolerated than resection of ileum, as the capacity for bile salt and vitamin B₁₂ absorption is specific to the ileum (Table 28-11).

During the first 1 to 2 years following massive small bowel resection, the remaining intestine undergoes compensatory adaptation, as discussed earlier. Clinically, the period of adaptation is associated with reductions in volume and frequency of bowel movements, increases in the capacity for enteral nutrient assimilation, and reductions in TPN requirements. As this process completes, some patients are successfully weaned off TPN. Understanding the mechanisms mediating intestinal adaptation may suggest strategies for enhancing adaptation in patients with short bowel syndrome who are unable to achieve independence from TPN. To date, the phenomenon of intestinal adaptation in patients remains poorly understood.¹³

Malabsorption in patients who have undergone massive small bowel resection is exacerbated by a characteristic hypergastrinemia-associated gastric acid hypersecretion that persists for 1 to 2 years postoperatively. The increased acid load delivered to the duodenum inhibits absorption by a variety of mechanisms, including the inhibition of digestive enzymes, most of which function optimally under alkaline conditions.

Therapy

Medical Therapy

For patients having undergone massive small bowel resection, the initial treatment priorities include management of the primary condition precipitating the intestinal resection and the repletion of fluid and electrolytes lost in the severe diarrhea that characteristically occurs. Most patients will require TPN, at least initially. Enteral nutrition should be gradually introduced, once ileus has resolved. High-dose ­histamine-2 receptor antagonists or proton pump inhibitors should be administered to reduce gastric acid secretion. Antimotility agents, such as loperamide hydrochloride or diphenoxylate, may be administered to delay small-intestinal transit. Octreotide can be administered to reduce the volume of GI secretions, although in animal models, its use is associated with an inhibition of intestinal adaptation.

During the period of adaptation, generally lasting 1 to 2 years postoperatively, TPN and enteral nutrition are titrated in an attempt to allow for independence from TPN. Patients who remain dependent on TPN face substantial TPN-associated morbidities including catheter sepsis, venous thrombosis, liver and kidney failure, and osteoporosis. Liver failure is a significant source of morbidity and often leads to liver transplantation (always in combination with small bowel transplantation). Due to these complications, patients with short bowel syndrome on TPN have a reduced life expectancy, with 5-year survival rates of 50% to 75%. Costs associated with chronic TPN administration are estimated to be as high as $150,000 per patient annually. Because of these problems, alternative therapies for short bowel syndrome are under investigation.

Nontransplant Surgical Therapy

Among patients with stomas, restoration of intestinal continuity should be performed whenever possible to capitalize on the absorptive capacity of all residual intestine. Other forms of nontransplant surgery designed to improve intestinal absorption are associated with unclear efficacy and/or substantial morbidities and therefore should not be applied routinely.

The goal of these operations is to increase nutrient and fluid absorption by either slowing intestinal transit or increasing intestinal length. Operations designed to slow intestinal transit include segmental reversal of the small bowel, interposition of a segment of colon between segments of small bowel, construction of small-intestinal valves, and electrical pacing of the small intestine. Reported experience with these procedures is limited to case reports or series of a few cases. Objective evidence of increased absorption is lacking; further, these procedures are frequently associated with intestinal obstruction.

The intestinal lengthening operation that has the longest history is the longitudinal intestinal lengthening and tailoring (LILT) procedure, first described by Bianchi in 1980.⁷⁵ The procedure entails separation of the dual vasculature of the small intestine, followed by longitudinal division of the bowel with subsequent isoperistaltic end-to-end anastomosis. This procedure has the potential to double the length of small intestine to which it is applied. This procedure has generally been used for pediatric patients with dilated residual small bowel.

An alternative surgical approach to lengthening the small bowel is the serial transverse enteroplasty procedure (STEP). This procedure is designed to accomplish lengthening of dilated small intestine without the need for separating its dual vasculature (Fig. 28-31). A recent report from an international registry of 111 patients showed that 47% of patients achieved enteral autonomy at a median follow-up of 21 months.⁷⁶

Intestinal Transplantation

This complex procedure is being increasingly performed to treat patients with short bowel syndrome. Approximately 1850 patients worldwide had undergone this procedure by 2009.⁷⁷ The currently accepted indication for intestinal transplantation is the presence of life-threatening complications attributable to intestinal failure and/or long-term TPN therapy. Specific complications for which intestinal transplantation is indicated include (a) impending or overt liver failure, (b) thrombosis of major central veins, (c) frequent episodes of catheter-related sepsis, and (d) frequent episodes of severe dehydration.

Of the transplants involving the intestine, 37% were intestine-alone transplants, 30% included intestine + liver + pancreas, and 24% were intestine + liver.⁷⁷

Isolated intestinal transplantation is used for patients with intestinal failure who have no significant liver disease or failure of other organs. Combined intestine/liver transplantation is used for patients with both intestinal and liver failure. Multivisceral transplantation has been used for patients with giant desmoid tumors involving the vascular supply of the liver and pancreas as well as that of the intestine, for diffuse GI motility disturbances, and for diffuse splanchnic thrombosis.

Nearly 80% of survivors have full intestinal graft function with no need for TPN. However, morbidities associated with intestinal transplantation are substantial and include acute and chronic rejection, CMV infection, and posttransplant lymphoproliferative disease.

Alternative Therapies

Pharmacologic and biologic therapies designed to expand intestinal mucosal surface area or to enhance the efficiency of intestinal absorption are beginning to undergo clinical evaluation. Promising regimens include GLP-2 and the combination of glutamine and growth hormone with a modified, high-carbohydrate diet.

Outcomes

Approximately 50% to 70% of patients with short bowel syndrome who initially require TPN are ultimately able to achieve independence from TPN.⁷³ Prognosis for achieving enteral autonomy is better among pediatric patients than among adults.

Information on survival among patients with short bowel syndrome is limited. In a recently reported study of 124 adults with short bowel syndrome due to nonmalignant etiologies, the survival rates at 2 and 5 years of follow-up were 86% and 45%, respectively.⁷⁸ Patients with end-enterostomies and those having less than 50 cm of residual small intestine had significantly worse survivals than those without these features.

No randomized trials comparing intestinal transplantation to chronic TPN administration among patients with short bowel syndrome have been reported. One-, 5-, and 10-year graft survival rates of intestine-alone recipients were 80%, 44%, and 26%; whereas the rates for intestine + liver and intestine + liver + pancreas were 62%, 45%, and 36% and 69%, 48%, and 33%, respectively.⁷⁷