Primary eosinophilic gastroenteritis is a disorder that selectively affects the gastrointestinal tract. First described in 1937 by Kaisjer, it is characterized by an eosinophilpredominant chronic inflammatory process of unknown etiology. The eosinophilia is present not only in the peripheral blood but also in the intestinal tissue.²⁷⁶ However, it has been reported that up to 23% of individuals do not present with peripheral eosinophilia.²²³ With approximately 75% of affected individuals found to have an allergic or atopic history, such an association has been suggested, but it has yet to be proven.^219,223 However, it is known that the expansion and tissue distribution of eosinophils are regulated by IL-5 and eotaxin 1.²⁶³ Furthermore, eosinophilic gastroenteritis has strong genetic and allergic components and presents immunopathogenic features similar to those of other allergic disorders, such as asthma.

Eosinophilic gastroenteritis can potentially affect any portion of the gastrointestinal tract. This can lead to eosinophilic esophagitis, eosinophilic gastritis, eosinophilic gastroenteritis, and eosinophilic colitis. The gastric antrum and the proximal small intestine are most frequently involved, whereas the colon is a rare location for the condition. Perianal disease has also been reported.¹⁷⁶ Eosinophilic colitis has a bimodal distribution, primarily affecting neonates and young adults.

The three major clinical patterns have been described based on the level of eosinophilic infiltration within the intestinal wall. The extent of disease corresponds to the clinical signs and symptoms of eosinophilic colitis. Primary mucosal disease is associated with enteric protein loss, diarrhea, and malabsorption. Predominant muscle layer disease
is characterized by obstructive symptoms as well as bowel thickening and may result in volvulus, intussusception, or perforation. Primary subserosal disease classically presents with eosinophilic ascites.^{35,171,223,235} The most common presenting symptoms are abdominal pain and change in bowel habits. Nausea, vomiting, weight loss, and rectal bleeding are also frequently noted.

Tissue eosinophilia is a common feature of eosinophilic gastroenteritis. Although eosinophiles are recognized manifestation in numerous gastrointestinal conditions, especially Crohn’s disease and ulcerative colitis, there is no comparison with the massive infiltration present in eosinophilic gastroenteritis and colitis.²¹⁹ When the condition occurs in the colon, it may radiologically mimic tuberculosis, amebiasis, or Crohn’s disease.²⁷⁶ Based on the few cases reported thus far, the proximal colon seems to have a greater predilection for involvement. Colonoscopic evaluation may reveal changes from erythema and friability, to granularity and narrowing.²³⁵ Tedesco and colleagues suggested that eosinophilic ileocolitis may be more common than originally thought, and postulated that biopsy and counting of high-power microscopic fields for eosinophils may be a useful means for distinguishing eosinophilic colitis from that of Crohn’s disease (Figure 33-1).³⁰² Eosinophilic colitis is largely a diagnosis of exclusion. In addition to inflammatory bowel disease (IBD), the differential diagnosis should include infectious colitides, such as helminthiasis and amebiasis (see later), systemic hypereosinophilia syndrome, milk protein colitis, vasculitis, and allergic gastroenteropathy.³⁵ Evaluation of IgE levels may help in determining the possibility of allergen involvement. Moreover, immunohistochemical analysis of intestinal eosinophil activation in patients with eosinophilic gastroenteritis suggests that the eosinophil cationic protein stored in eosinophil granules and secreted by activated eosinophils may be a tissue marker for eosinophilic gastroenteritis.²⁸ However, the combination of histologic and clinical patterns is the mainstay for the diagnosis of inflammatory conditions of the colon.⁴⁷ Although diagnostic criteria have been established for eosinophilic esophagitis, there has yet to be a consensus reached for eosinophilic colitis. Many investigators have used 20 eosinophils per high power field as being diagnostic. It is important to note, however, that the amount of eosinophils present varies depending on the location within the colon, ranging significantly above and below the aforementioned threshold.²²³

Prognosis is generally good, with clinical, hematologic, roentgenographic, and histologic improvement occurring spontaneously (usually in children) or with corticosteroid therapy.¹²⁵ Commonly, symptoms will resolve within 2 weeks on steroid treatment. However, recurrence is not infrequent. Novel therapeutic approaches to the management of eosinophilic gastroenteritis are currently being assessed, including treatment with newer antihistamines, leukotriene receptor antagonists, and monoclonal antibodies, such as anti-IL-5 (mepolizumab) and anti-IgE (omalizumab).^108,223 However, resection of the involved bowel may be required for unremitting symptoms or to exclude another diagnosis.²¹⁹

Microscopic colitis is an inflammatory condition that was initially described in 1976 by Lindström and colleagues.¹⁸³ It is characterized by chronic, watery diarrhea and grossly normal appearing colonic mucosa on endoscopic examination, with inflammation evident only on microscopic examination. Terminology for the disease has evolved as further investigation has allowed the disease to be more well defined. At present, it is widely accepted that microscopic colitis consists of two subtypes: lymphocytic colitis and collagenous colitis. Although the clinical manifestations are similar, the subtypes can be differentiated based on the presence of intraepithelial lymphocytosis and subepithelial collagen deposition, respectively.²⁷⁴

The actual incidence of microscopic colitis is unknown because it is a poorly reported condition. A study from Sweden by Olesen and colleagues²²⁴ noted an increasing incidence of microscopic colitis over the past decades from 1.8 per 100,000 to 6 per 100,000 inhabitants. The authors concluded that the incidence of microscopic colitis is higher than previously described and is as common as Crohn’s disease in Örebro, Sweden. Microscopic colitis was diagnosed in 10% of patients with nonbloody diarrhea and up to 20% of those individuals who were older than age 70 years. There is no association of microscopic colitis with an increased risk of colorectal cancer.¹⁸⁶ The incidence of microscopic colitis within the United States was recently investigated in a population-based study in Olmsted County, Minnesota. Over the duration of the study, investigators also reported an increasing incidence from 1.1 to 19.6 cases per 100,000 person-years. A positive association with older age was noted as well.²³⁴ In a Canadian population-based study conducted by Williams and colleagues, the authors found that individuals aged 65 and older were five times more likely to have the disease, and an increased risk was found in women for both subtypes of microscopic colitis.³²⁴

A definitive cause of microscopic colitis has yet to be determined. Reports in the literature have described a correlation between the disease and certain medications,
such as nonsteroidal anti-inflammatory drugs (NSAIDs). However, no causal relationship has been established.²⁹⁹ With respect to disease associations, links to autoimmunity are widely accepted, and an increased prevalence of microscopic colitis among patients with celiac disease has been described.^57,224 Moreover, Koskela and colleagues recently determined that HLA DR3-DQ2 (already linked to celiac disease) and TNF2 allele carriage (associated with autoimmunity) were also associated with both subtypes of microscopic colitis.¹⁷⁰

The clinical course of this disease is benign, with spontaneous remission of symptoms occurring within a few years of onset. Watery diarrhea may persist over weekly or monthly intervals in association with fecal incontinence. Anemia, increased erythrocyte sedimentation rate, hypokalemia, and hypoalbuminemia are common findings. Microscopic colitis is often misdiagnosed as irritable bowel syndrome (IBS). However, its incidence in older individuals as well as the absence of abdominal discomfort as a significant complaint are distinguishing features.²⁷⁴ Endoscopic examination and radiologic studies have demonstrated a normal appearing mucosa in the majority of individuals. However, edema, friability, pinpoint mucosal hemorrhages, and hyperemia have been occasionally seen.²⁴⁸ Definitive diagnosis is based on histologic evaluation of colonic biopsies. Although microscopic colitis is a diffuse disease, it is recommended that one avoid rectal biopsies because the subepithelial collagen layer tends to be thicker in the rectum.²⁷⁴ Both subtypes demonstrate damage to the epithelial cells, such as flattening and depletion along with a predominantly mononuclear inflammation of the lamina propria and preservation of crypt architecture. The inflammation is found to be remarkably uniform, indicative of a total colitis. As mentioned, additional histologic features distinguish lymphocytic colitis from collagenous colitis. Lymphocytic colitis is characterized by greater than 20 intraepithelial lymphocytes per 100 surface epithelial cells (Figure 33-2), whereas a thickened subepithelial collagen layer is typical in collagenous colitis (Figure 33-3).⁸⁹ The thickening of the subepithelial collagen layer may be a response to chronic inflammation or a local abnormality of collagen synthesis.²⁸⁹ It has been postulated that the disease may be attributable to reduced cell turnover, allowing fibrocytes to remain longer in the mature phase, hence producing more collagen and a thicker collagen plate.¹⁵⁸ Another subtype, although rare, has been described by Sandmeier and Bouzourene.²⁷¹ This type is characterized by the presence of subepithelial multinucleated giant cells, which the authors consider an important diagnostic criterion in the differential diagnosis of granulomatous infections and Crohn’s disease. Moreover, it seems that this atypical subtype of microscopic colitis has a more favorable prognosis and a better response to corticosteroid therapy.

Recommended treatments have included nonspecific antidiarrheal agents, such as loperamide, bismuth subsalicylate, sulfasalazine, mesalamine, bile acid-binding agents (e.g., cholestyramine), budesonide, prednisone, and in refractory cases, immunosuppressive drugs.^118,274 In the Swedish experience, prednisolone was most effective, with a response rate of 82%. However, the required dose was often quite high, and the effect was not maintained following withdrawal of the medication. The response rates for antibiotics, cholestyramine, and loperamide were 63%, 59%, and 71%, respectively. Conversely, spontaneous resolution may occur.³² Reported success rates with sulfasalazine and mesalamine are lower (40% to 50%).²⁷⁴ A publication by the Cochrane database concerned a review of five randomized trials. It identified three treatment options for collagenous colitis. Budesonide was found to be an effective agent for the management of this disease; however, there was weak evidence to support the value of bismuth subsalicylate and prednisolone.⁵⁸ More recently, several randomized, double-blinded, placebo-controlled trials have demonstrated the efficacy of oral budesonide in both subtypes of microscopic colitis. Two trials involved the administration of budesonide, 6 mg daily, as maintenance treatment over 6 months in patients with collagenous colitis. The response rates in the treatment groups as compared with the placebo groups were 74.0% and 76.5% compared to 35% and 12%, respectively.^33,201 A third trial reported a response rate of 86% versus 48% placebo in patients with lymphocytic colitis who were given budesonide, 9 mg daily, for a 6-week duration.
Clinical relapses were noted but were treatable with repeated administration of budesonide.²⁰²

Rare refractory cases may benefit from operative management, but the role of surgery in this condition is extremely limited.¹²⁶ Fecal diversion, subtotal colectomy,¹⁹⁰ and restorative proctocolectomy with ileo-pouch anal anastomosis have all been suggested as therapeutic options.³¹³

Neutropenic enterocolitis also known as ileocecal syndrome or typhlitis is a syndrome associated with bowel wall necrosis that was initially recognized in pediatric leukemic patients.⁶⁹ With the development of effective chemotherapeutic agents, this clinical entity has become increasingly prevalent in the adult population. It can occur during treatment of hematologic malignancies, especially leukemia, lymphoma, and aplastic anemia.¹⁷² Those chemotherapeutic agents associated with neutropenic enterocolitis include paclitaxel, gemcitabine, cytosine, arabinoside, vincristine, doxorubicin, cyclophosphamide, 5-fluorouracil, leucovorin, and daunorubicin.⁶⁹ Neutropenic enterocolitis has also been reported as a complication of other therapeutic modalities, such as antithyroid therapy,⁶⁰ and chemotherapy for lung, breast, ovarian, and colon cancer,^69,101 or as the presenting complication of acute lymphoblastic leukemia.²⁴⁴ It has also been described as a complication of cyclic neutropenia, a rare benign hematologic disorder.²²² Profound neutropenia secondary to chemotherapy has been considered the hallmark of the disease and the major etiologic factor in its development.²¹³ This is a condition characterized by regular oscillations in blood neutrophil counts, in which sporadic disappearance of these cells from the circulation occurs. Involvement of the process is most commonly seen in the cecum, possibly due to its distensibility and limited blood supply.⁸⁵ Other frequent sites include the terminal ileum and right colon, possibly because of the higher concentration of lymphatic tissue in these areas.

Patients exhibit symptoms of diarrhea, abdominal pain, sepsis, and findings typical of acute appendicitis. Even when the pain is localized to the right lower quadrant, the rapidity of the appearance of toxicity, with tachycardia, fever, and delirium, can be extremely dramatic. Koea and Shaw reported three individuals with leukemia who presented with fever, abdominal pain, and signs of peritonitis localized to the right iliac fossa.¹⁶⁶ Each patient was found to have a nonviable cecum.

The pathogenesis of the disease involves disruption of the intestinal mucosa in a neutropenic state compounded by damage secondary to chemotherapy. The disruption of gut mucosal integrity allows for the invasion of pathogenic bacteria leading to the release of endotoxins and to the development of necrosis and hemorrhage (Figure 33-4). A wide spectrum of bacteria have been identified in surgical specimens, such as clostridial species, gram-negative rods, gram-positive cocci, and Candida species.⁶⁹ The diagnosis is based on clinical presentation and a low neutrophil count. The presence of radiologic abnormalities, however, is essential with CT and ultrasound imaging serving as the studies of choice (Figure 33-5).⁸⁵ Kirkpatrick and Greenberg described pneumatosis intestinalis as the most common computed tomographic observation suggestive of neutropenic enterocolitis (21% of patients).¹⁶⁰ Bowel wall thickening, mesenteric stranding, bowel dilatation, mucosal enhancing, and ascites are also frequently reported.¹⁶⁰ In patients who are hemodynamically unstable, ultrasound should be considered as an alternative diagnostic modality given its ease of use and ability to quickly detect bowel wall thickening as well as rapidly eliminating other possible diagnoses, such as appendicitis and pancreatitis.⁶⁹ One prospective study evaluating the use of ultrasound in leukemic patients postchemotherapy found that bowel wall thickening greater than
4 mm was present only in those who symptomatically satisfied the criteria for neutropenic enterocolitis. As such, some advocate that the diagnostic criteria for neutropenic enterocolitis should include the presence of bowel wall thickening identified on CT scan or ultrasound (greater than 4 mm on transverse scan over greater than 30 mm on longitudinal scan), in addition to the clinical presentation of fever and abdominal pain.⁸⁵

Given high mortality rates, the management of neutropenic enterocolitis has been traditionally aggressive, with a low threshold for surgical intervention. However, because an increasing number of cases successfully treated with medical therapy alone have been reported, there has been a shift in the treatment paradigm. Intensive support with close monitoring, fluid resuscitation, bowel rest, and broad-spectrum antibiotics are key components to conservative management. Antibiotic treatment may be initiated with single drug regimens, such as piperacillin-tazobactam, or with dual agents, such as a beta-lactam antipseudomonal drug paired with an aminoglycoside. Coverage should be targeted at grampositive and gram-negative bacteria, as well as pseudomonas, clostridial, and fungal species. Moreover, the use of recombinant granulocyte colony-stimulating factor (G-CSF) has been administered to help normalize white blood cell levels and consequently to lead to clinical improvement. Its efficacy, however, is not well established and is currently recommended for only the sickest of patients.⁶⁹

Surgical intervention should be considered in those individuals with persistent gastrointestinal bleeding, free intraabdominal perforation, clinical worsening, or to exclude other acute abdominal disease processes. Resection is necessary with perforated or necrotic bowel; however, in the setting of neutropenia, a primary bowel anastomosis is not recommended. Rather, resection with diversion is the preferred alternative.⁸⁵

Diverting the fecal stream and defunctionalizing the bowel can produce a noninfectious colitis, an observation made by Glotzer and colleagues in 1981.¹¹³ Generally, patients are asymptomatic, although mucus discharge and bleeding may be noted. Fenton and Siegel have shown that 70% of their ostomy patients had gross or microscopic findings of diversion colitis despite the absence of symptoms.¹⁰⁰ Roe and colleagues reported that all of their 12 patients who had undergone a Hartmann’s procedure demonstrated histologic abnormalities consistent with diversion colitis by 3 months.²⁵³ An association of microcarcinoids with diversion colitis has also been reported.¹²² This occurrence has been attributed to neurogenic hyperplasia, which represents proliferation of a separate, neuron-associated, extraglandular population of endocrine cells.¹²⁰

Proctosigmoidoscopic findings are essentially that of a mild inflammatory bowel disease suggestive of ulcerative colitis. Microscopic alterations, however, tend to be rather focal and include crypt abscesses, epithelial cell degeneration, acute and chronic inflammation in the lamina propria, and regenerative changes in the crypts (Figure 33-6).¹¹³ The crypt cell production rate has been determined to be less than half that of controls. Additionally, crypt length and width are lower.¹⁰ However, the histologic picture with the condition is quite variable. Pathologic evaluation of more severely diseased, resected specimens may demonstrate diffuse nodularity caused by lymphoid hyperplasia and an inflammatory process confined to the mucosa and submucosa.²¹⁴

The mucosa of the intestinal tract is unique in that it draws nutrients not only from the vasculature, but also from the bowel lumen.²⁵⁴ It is not difficult to appreciate how this entity can develop when examining the radiologic picture shown in Figure 33-7. Nutrition of the colonic epithelial cells is mainly from short-chain fatty acids produced by bacterial fermentation in the colonic lumen.²⁵⁴ Deficiency in these substances initially leads to mucosal hypoplasia and subsequently to a more typical picture of nonspecific inflammatory bowel
disease. Harig and colleagues have shown that negligible amounts of short-chain fatty acids can be identified in a segment of excluded rectosigmoid.¹²⁹ They further demonstrated that installation of a solution containing short-chain fatty acids twice daily resulted in the disappearance of symptoms and in the inflammatory changes observed at endoscopy within a period of 6 weeks. Short-chain fatty acids have also been used for the management of diversion colitis in children.¹⁵⁶ Others have shown that the condition may be successfully treated with the use of 5-aminosalicylic acid enemas.³⁰⁷ However, the only curative approach is stoma reversal or rectal excision.

The affected bowel rapidly returns to a normal appearance following reestablishment of intestinal continuity. Orsay prospectively evaluated 34 patients who were scheduled to undergo colostomy closure and who were demonstrated to have diversion colitis.²²⁵ There was no increased infection rate or any other complications following colostomy closure.

Chemical colitis is a rare clinical entity that can be induced by iatrogenic means with contaminated endoscopes or with intentional or accidental administration of corrosive enemas. Iatrogenic chemical colitis was first described in the 1960s and has been reported in the literature in numerous case reports and series.²⁸¹ Although disinfectant solutions are known to be caustic to the intestinal mucosa, controversy exists as to the exact causative agent. The main disinfectants implicated are glutaraldehyde, peracetic acid, and hydrogen peroxide (a breakdown product of peracetic acid when it is combined with water). The residual presence of such chemicals on endoscopic equipment may be the result of malfunction of automated disinfecting devices or lack of staff education in proper cleansing techniques.

Injury to the colon typically becomes manifest at the time of endoscopy with the classic “snow white sign,” multiple areas of elevated whitish yellow plaques of variable size and confluence. Histologically, such lesions are characterized by empty vacuoles within the submucosa; hence, the term pseudolipomatosis (Figure 33-8).^145,210

Signs and symptoms range from no clinical sequelae to fever, abdominal pain, and bloody diarrhea within 48 hours of endoscopic examination. Radiologic findings on CT scan after glutaraldehyde-induced colitis may demonstrate a “target sign” of circumferential colonic wall thickening. Resolution is usually seen with bowel rest and conservative treatment. Empiric antibiotics and steroids may be administered, but no evidence of their efficacy has been reported.²⁸¹

Oral administration of a host of corrosives is a well-recognized entity for which surgical intervention is often required. Less well recognized, however, is the installation of various toxic or corrosive materials by means of an enema. Cappell and Simon reviewed the literature on the reported intra-rectally administered toxic agents and found that the most common was hydrogen peroxide.⁴⁵ Other agents included detergent enemas, herbal medicines, acetic acid, ethyl alcohol, sodium hydroxide, and hydrofluoric acid. Pikarsky and colleagues presented a case of severe formalin-induced colitis 5 days after rectal instillation of formalin for radiation proctitis.²³⁸ Drugs such as ergotamine can also be associated with inflammatory reaction and toxicity. Toxic exposure may result from conventional medical therapy, unconventional medical therapy, radiographic examination, colonoscopic examination (see previous discussion), deliberate self-mutilation, or accidental self-administration.⁴⁵

Rectal bleeding, diarrhea, and abdominal pain are frequent symptoms. Endoscopic examination usually reveals nonspecific changes consistent with inflammatory bowel disease, including erythema, ulceration, granularity, friability, and purulent exudate. A history of medication use or exposure to toxic agents is obviously helpful.

Treatment is generally supportive care, although emergency laparotomy and bowel resection may be required for fulminant acute colitis, stricture, or perforation (see also Chapter 17).

Adverse effects of NSAIDs on the upper gastrointestinal tract and small intestine are well recognized.⁸⁴ Since the initial description of NSAID-induced colonic injury by Debenham,⁸⁶ a growing number of reports have associated the use of NSAIDs and salicylates with colonic complications. However, the true incidence of NSAID colopathy is unknown. A study from Jersey (United Kingdom) estimated that approximately 1:1,200 NSAID users may develop colitis.¹¹² Recognized colonic complications secondary to NSAID use include the development of primary macroscopic colitis, worsening of inflammatory bowel disease, collagenous colitis, gastrointestinal bleeding, and increased complications in association with diverticular disease.¹⁶

The clinical presentation of NSAID-induced colitis is based on the development of an acute, nonspecific inflammatory process involving the colonic mucosa.⁹⁴ With long-term NSAID use, it may progress to chronic disease characterized by the presence of fibrosis and strictures.² Symptoms include bloody diarrhea, sporadic melanotic stool leading to irondeficiency anemia, weight loss, fatigue, anorexia, and even death. The pathogenesis of NSAID-induced colitis may be related to the inhibition of prostaglandin synthesis.⁸⁴
In addition to the oral route, NSAID-induced suppository toxicity has been recognized.

Colonoscopy may demonstrate diffuse inflammation, with erosions or ulcerations. The presence of diaphragmatic-like strictures or webs, most commonly in the right colon, has been described as a pathognomonic sign of NSAID-induced colitis.¹⁴⁹ However, the most common finding is a solitary or limited number of cecal ulcerations, or they may be scattered throughout the colon and rectum.²⁴¹ The condition may be difficult to differentiate from that of ulcerative colitis in its early presentation.

NSAID ingestion should certainly be considered in the differential diagnosis of colitis. The use of NSAIDs in the form of suppositories may induce rectal bleeding and proctitis due to the high concentration of the drug in direct contact with the rectal mucosa.^111,180 Moreover, the severity of rectal symptoms may be dose dependent. It has been estimated that up to 30% of patients using NSAID-containing suppositories may develop rectal symptoms such as pain, inflammation, bleeding, ulcerations, and stricture.⁸⁴

One would certainly expect an ameliorative response of the bowel to the discontinuation of the medication. Additionally, the use of sulfasalazine and metronidazole has proven to be somewhat beneficial.⁹⁵ Surgical treatment is reserved for cases of life-threatening complications or refractory colitis.

Toxic epidermal necrolysis is a rare and severe reaction to certain drugs that results in full-thickness epidermal skin necrosis.⁴⁸ Other mucosal surfaces have been implicated, including the large bowel. The most common drugs associated with this condition include sulfonamides, penicillin, NSAIDs (see previous discussion), phenytoin anticonvulsants, and barbiturates.⁴⁸

Symptoms include abdominal pain and bloody diarrhea, usually concomitant with the development of the skin lesions. Several cases of colonic necrosis complicating this condition have been reported.⁴⁸

The main nonantibiotic agent for prophylaxis against traveler’s diarrhea is bismuth subsalicylate, which has been shown to decrease incidence rates from 40% to 14%. The Centers for Disease Control and Prevention recommend avoidance of prophylactic antibiotics for the average traveler due to concerns for possible adverse drug reactions. For those who are immunosuppressed or at high risk, prophylaxis with fluoroquinolones is the current recommendation. Although the majority of diarrheal illnesses represent little more than a self-limited nuisance, oral rehydration along with antibiotic treatment is recommended for severe cases. Fluoroquinolones such as levofloxacin are considered first-line therapy with azithromycin as an alternative in cases of drug resistance.⁷⁵

Comprehensive microbiologic testing and thorough gastrointestinal studies for every individual with the complaint of diarrhea are both impractical and costly.^130,329 As enteric infections frequently present with vague, nonspecific signs and symptoms, it is imperative that the clinician obtain a thorough history considering recent travel and other possible sources of exposure to infectious agents. Generally, for the younger patient, no investigations, except perhaps a rigid or flexible sigmoidoscopy and biopsy, are necessary if weight is stable and the stool is free of occult blood.¹³⁰ Nonetheless, a few individuals will require a more extensive evaluation, a decision that must rest with the physician’s clinical judgment.

In this section, the various enteric infections of potential interest to the surgeon are discussed with respect to epidemiology, clinical presentation, diagnosis, therapy, and prevention.

The first reported case of pseudomembranous colitis actually antedated the antibiotic era and is generally attributed to Finney of Johns Hopkins.¹⁰² His patient developed bloody diarrhea, perhaps as a consequence of enema feedings following a gastric procedure. The discussants of his paper presented what were than a “who’s who” of American medicine prior to the turn of the 20th century—Halsted, Osler, Thayer, and Flexner.

Campylobacter jejuni and Campylobacter coli are the leading causes of infectious enterocolitis worldwide. In fact, Campylobacter has become one of the major causes of infectious diarrhea in the United States, with a reported 13 laboratory-confirmed cases per 100,000 annually.⁵³ Within the United States, it is noted to occur most commonly in children younger than the age of 4. Internationally, infection is typically sporadic in poor developing countries as well as in military personnel stationed abroad.¹⁵⁹ The organism is a curved or “gull wing” microaerophilic, grampositive rod. Transmission occurs by way of the fecal-oral route through contaminated fruit and water or by direct contact with infected animals or persons.³⁰ Epidemiologic studies have demonstrated an association with the handling and consumption of poultry and beef. The risk of contamination is higher with raw meats, especially at barbecues. This is believed to be due to the ready transfer of the bacteria to other foods and then to the mouth.⁴² This bacterium is also found in shellfish and dairy products and has also been diagnosed in hospitalized patients. Campylobacter is vulnerable to temperature extremes as well as to oxygen in atmospheric concentrations. In contrast to that of Salmonella, Campylobacter does not survive in pelleted meals, egg powder, and spices.

Symptoms and findings may be difficult to differentiate from those of other diseases affecting the intestinal tract, particularly nonspecific inflammatory bowel disease.²⁴⁶ In fact, Campylobacter enteritis must be considered in the differential diagnosis of any patient who presents with rectal bleeding and diarrhea. Abdominal pain, fever, nausea, and vomiting may also be associated complaints. Toxic megacolon, necessitating total colectomy, has been reported.⁹

Proctosigmoidoscopic examination usually reveals an edematous, inflamed mucosa. Histologic examination of biopsy specimens is nonspecific. A double-contrast barium enema may demonstrate aphthoid ulcers and a stippled appearance.³⁰⁴

A high index of suspicion must be maintained if one is to establish the diagnosis and to initiate therapy promptly. Examination of the fecal specimen within 2 hours of passage by dark-field or phase-contrast microscopy may identify the organism.³⁰ The presence of polymorphonuclear leukocytes in the fecal stream is not uncommon but is not pathognomonic for the condition.

Infection with C. jejuni may lead to various postinfectious sequelae, including reactive arthritis, Reiter’s syndrome, Guillain-Barré syndrome, IBS, and IBD.²³ The incidence of IBS- and IBD-associated cases ranges from 4% to 32%. Research has demonstrated a consistent association between length of clinical illness (greater than 1 week) and the development of postinfectious IBS. Debate still remains, however, over the mechanism of disease as well as the question of whether infection with Campylobacter leads to IBS and IBD or whether patients with such intestinal diseases are simply predisposed to the bacterial infection.¹⁵⁹

In the majority of patients, symptomatic infection is usually self-limited and does not require antibiotic therapy, but relapses are frequent. In severe cases, those associated with fever, bloody diarrhea, and cramping, hospitalization and fluid and electrolyte replacement may be necessary. Although the mainstay of treatment has been fluoroquinolones, antibiotic resistance has increased rapidly with a reported prevalence of 26% in 2007. Consequently, single dose administration of azithromycin has been recommended as an effective alternative.¹⁵⁹ Appropriate stool precautions (particularly for hospitalized patients) are indicated, with proper disposal of contaminated linens and washing of hands. With respect to prevention, avoiding raw meats, particularly poultry, is recommended.

Currently, research is focusing on the identification and eradication of bacterial reservoirs. Additionally, molecular studies, leading to a better understanding of Campylobacter-associated diseases, are being conducted in order to improve prevention techniques and to offer alternative therapeutic approaches.¹⁸² The development of vaccines against C. jejuni has posed a significant challenge due to its microbiologic characteristics. Presently, phase I/II human trials are being conducted with preclinical experiments demonstrating promising data on a capsular conjugate vaccine.¹⁵⁹

Yersinia enterocolitica is a relatively recently recognized cause of enteric infection. Yersinia enterocolitis is of particular interest to the surgeon because of its prevalence and its occasional confusion with regional enteritis. A former name for the causative organism was Pasteurella pseudotuberculosis, again implying confusion with another bacterial infection that tends to involve the ileocecal region. Epidemics due to contamination of food, water, and milk have been reported.²⁹ Y. enterocolitica most commonly affects infants, young children, and young adults.

The disease is caused by a facultative anaerobic, gramnegative coccoid bacillus resembling nonlactose-fermenting E. coli.^78,312 It grows optimally in cold temperatures. The diagnosis is established by isolation of the bacteria from stool, blood, fluids, or tissue. However, Y. enterocolitica culture is not standard in most clinical laboratories and must be specifically requested.³¹² Biotyping and serotyping according to O antigens have been the most helpful of the epidemiologic techniques.⁷⁸

After a median incubation period of 4 days, the organism may produce signs and symptoms of an acute enterocolitis. This may be accompanied by pharyngitis as a consequence of invasion of epithelial cells and the penetration of the intestinal mucosa, as well as the presence of the lymphoidabundant tonsils.³⁰⁶ Bloody diarrhea is frequently observed in addition to abdominal pain.²⁶⁸ Joint pain may also be a manifestation of this disease. Drainage of the bacteria into regional lymph nodes accounts for the systemic complications. Mesenteric lymphadenitis may develop from colonization of Peyer’s patches communicating directly with mesenteric lymph nodes. A syndrome simulating appendicitis is seen in 40% of patients—that is, fever, leukocytosis, right lower quadrant abdominal tenderness, and pain.^78,268 However, a small number of patients with Y. enterocolitica will actually develop true appendicitis as a manifestation of the disease.²⁸³ The condition may produce generalized septicemia and “metastatic” abscesses in other organs. It can also present as a colonic abscess or as toxic megacolon.^272,296 Sometimes the disease may pursue a chronic course for many weeks, particularly if not treated with appropriate antibiotics. Postinfection manifestations include erythema nodosum and reactive arthritis.⁷⁸ Predisposing factors to the development of the infection follow:

Results of radiologic examination were evaluated in a review by Vantrappen and colleagues.³¹² A coarse, irregular, nodular mucosal pattern was seen in the terminal ileum; ulcerations were also noted. In contrast to Crohn’s disease, infection of the terminal ileum is usually confined to the mucosa and submucosa, and the characteristic “string sign” is absent. Endoscopic examination demonstrates signs of inflammatory disease in approximately one-half of the patients. Findings are nonspecific and variable, including aphthoid lesions within the cecum, as well as small round elevations in the terminal ileum with or without exudate.¹⁵⁵

Recommended treatment includes antipseudomonal aminoglycosides, trimethoprim/sulfamethoxazole, ceftizoxime, or ceftriaxone, but antimicrobial therapy has not been proven essential or necessarily efficacious in the uncomplicated situation.⁷⁸ However, when systemic illness supervenes or when the patient is immunocompromised, doxycycline or trimethoprim/sulfamethoxazole is advisable. Y. enterocolitica is generally resistant to penicillin and first generation cephalosporins. Sporadic resistance to macrolides and fluoroquinolones has been reported as well.¹⁵⁵

Prevention of yersiniosis includes adequate treatment and handling of raw poultry, beef, and pork as potential sources of infection. Consumption of raw milk should be avoided and hand washing after using the toilet or diaper changes as well as after handling pets and animals is mandatory.²²⁰