Clinical History
A 39 year old man presented to the Emergency Department with a two day history of severe, crampy
abdominal pain accompanied by nausea and vomiting. The patient was reportedly writhing, moaning, and
yelling due to the pain. There was no significant past medical history except for mild asthma, and there
was no past history of intra-abdominal surgery. Review of systems indicated intermittent abdominal pain
and diarrhea in the past, but no fever, chills, weight loss, or vomiting. Social history was significant
for extensive methamphetamine use. Physical examination revealed a soft, distended, diffusely tender
abdomen with decreased bowel sounds; there were no masses, rebound effect, or point tenderness.
Laboratory evaluation was significant for an elevated white blood cell count of 12,000 with increased
eosinophils. All other labs were unremarkable. Abdominal xrays showed multiple air-fluid levels, and CT
scan showed a small bowel obstruction. The patient was taken to surgery for exploratory laparotomy,
which revealed a segment of dilated and thickened jejunum associated with a small amount of ascitic fluid
and adhesions to adjacent loops of bowel. The abnormal segment of small intestine was resected, and an
H&E section from the partial small bowel resection is submitted for review.
Case 3 - Figure 1 - Low power view of a section of jejunum shows a dense eosinophilic infiltrate involving all layers of the bowel wall.
|
Case 3 - Figure 2 - The lamina propria contains a predominantly eosinophilic infiltrate, with focal infiltration of small bowel epithelium.
|
Case 3 - Figure 3 - The eosinophilic infiltrate percolates between muscle bundles in the muscularis propria.
|
Case 3 - Figure 4 - The serosa of the small bowel is involved by a striking eosinophilic infiltrate, and there was associated eosinophilic ascites.
|
Pathologic Findings
Grossly, the specimen consisted of a segment of jejunum that was markedly edematous. In the center of
the specimen there was a 4-5cm area that was dilated, with a thickened wall. Upon opening the specimen,
there was blunting of the mucosal folds, but no areas of ulceration. No masses or other focal lesions
were seen, and there were no areas of perforation. Subserosal fat was focally adherent to the surface of
the bowel.
At low power, sections showed marked edema and a striking transmural eosinophilic infiltrate. At
higher power, the villous architecture was intact, and the lamina propria contained numerous eosinophils
that focally infiltrated the epithelium. The infiltrate extended across the muscularis mucosa to
extensively involve the submucosa and muscularis propria, where muscle fibers within the wall were
splayed by the infiltrate. The serosa of the resected small bowel also contained an exudate composed
predominantly of eosinophils. The infiltrate was almost exclusively eosinophilic, with few other
inflammatory cells noted. There was no frank ulceration, nor any evidence of vasculitis. No parasites
were present.
Hospital Course
The patient did well postoperatively, with resolution of symptoms, and was discharged home on
postoperative day five. He subsequently did not keep any of his clinic appointments and was lost to
follow-up.
Diagnosis: Eosinophilic Gastroenteritis
Discussion
Eosinophilic gastroenteritis (EG), originally described in 1937, connotes a spectrum of diseases
characterized by eosinophilic infiltration of one or more segments of the gastrointestinal (GI) tract.
Many patients have peripheral eosinophilia and/or a history of atopy or asthma. The name is somewhat of
a misnomer, since the entire GI tract may be involved, including the esophagus, colon, biliary tree, and
pancreas.
Pathogenesis
The pathogenesis of EG remains unclear. Although familial
cases have been rarely reported, no specific genetic abnormalities or inheritance patterns have been
observed. Rare cases have been associated with rheumatologic diseases. Some patients have a history of
atopy; although many patients with EG have positive skin test responses to food antigens, they do no have
anaphylactic reactions, suggesting a delayed type hypersensitivity reaction. In addition, many have high
circulating levels of IgE. Many workers theorize that immunologic mechanisms involving interleukins 4
and 5 are responsible for EG, and that EG may result from food antigen exposures, stimulating eosinophil
degranulation and release of toxic major basic protein into gastrointestinal tissues. Mast cell
involvement has also been proposed, as mast cells may be increased in EG, and it is postulated that they
are involved in eosinophil chemotaxis and activation.
Clinical Features
EG most commonly presents in children and adults
younger than 50 years of age. Signs and symptoms are related to the site of GI tract involvement, as
well as whether or not involvement is predominantly mucosal, mural, or serosal (see below). Presenting
clinical findings in EG include abdominal pain (often severe and recurrent), hemorrhage, nausea,
vomiting, diarrhea, protein losing enteropathy and other forms of malabsorption, anemia, obstruction, and
ascites. Infants and small children may present with failure to thrive. More unusual presentations
include solitary nonhealing gastric ulcers and intestinal perforation. The majority of patients have an
elevated white blood cell count with substantial peripheral blood eosinophilia, particularly during
attacks of abdominal pain, although up to 25% of patients with EG lack peripheral eosinophilia.
Erythrocyte sedimentation rate may be normal or slightly elevated. Rarely, EG involves the biliary tree,
liver, and pancreas, causing pancreatitis, cholangitis, or biliary obstruction.
Pathologic Features
The stomach is the most common site of involvement, followed by duodenum, esophagus, jejunum, ileum,
and colon. Small bowel and gastric involvement often coexist. In the colon, the cecum and ascending
colon are most commonly involved. The colon may be solely involved, or in combination with other areas
of the GI tract.
Historically, EG is often categorized into three types:
1. Mucosal predominant EG. This is the most common form of EG.
Patients with mucosal EG usually present with diarrhea, bleeding, and/or malabsorption. Grossly, the
mucosa may appear granular, friable, or nodular. Microscopic features include mucosal eosinophils with
variably present crypt abscesses, erosions, ulceration, and small bowel villous blunting.
2. Mural EG. Mural involvement usually presents as abdominal
pain, obstruction, nausea, and vomiting. Grossly, the bowel is thickened and edematous, with lumenal
narrowing. The stomach may contain prominent, thickened rugal folds. The eosinophilic infiltrate
involves the submucosa and muscular wall, often splaying the muscle fibers. The mucosa may appear
entirely normal in both mural and serosal EG.
3. Serosal EG. Serosal EG is the least common. This form is
usually associated with eosinophilic ascites, along with abdominal pain and obstruction. The peritoneal
surface of the bowel often is covered with a fibrinous exudate or thick plaques that upon microscopic
sectioning consist of eosinophils.
Some cases may show features of all three, as in the case presented here. The eosinophilic infiltrate
may be focal and patchy, so biopsies should be obtained from multiple sites. Typically there is a
remarkably dense infiltrate of eosinophils that may be so dense as to cause thickened gastric rugal
folds, intestinal pseudopolyps, and villous blunting in the small bowel. Eosinophils may be exclusively
present, or mast cells and IgE plasma cells may be admixed. Neutrophils are typically not prominent. In
mucosal predominant disease, there are increased numbers of both intact and degranulating eosinophils
within the lamina propria, and infiltrating the epithelium to varying degrees. Eosinophilic
micro-abscesses, crypt abscesses, erosions, and frank ulceration may be present. Eosinophils may
accumulate in the muscularis mucosa and submucosa, often associated with edema, and may produce splaying
of muscle bundles in mural EG. Serosal EG features marked serosal eosinophilic infiltrates associated
with ascites.
There are no exact numerical criteria for the diagnosis of EG. Numbers of eosinophils present
normally within the gut vary greatly from site to site within the GI tract, as well as in different
regions of the country. Asymptomatic patients may have high tissue eosinophil counts, particularly
within the colon. Infiltration of the epithelium, clustering of eosinophils, and crypt abscesses are
more important clues to the presence of real disease in a mucosal biopsy than counting numbers of
eosinophils. Knowledge of the peripheral eosinophil count, symptoms, and radiography can be very helpful
when evaluating biopsies for the possibility of EG.
Differential diagnosis.
Many diseases can produce increased numbers of eosinophils in the GI tract, including Crohn's disease,
vasculitides, food and drug allergies, idiopathic hypereosinophilic syndrome, parasitic infections, and
celiac disease. Parasitic infection can be excluded both on tissue sections and stool examination for
ova and parasites. Arteritis with damage to vessel walls and associated ischemic tissue damage are not
features of EG. Patients with idiopathic hypereosinophilic syndrome may have gut involvement, but
involvement of other organ systems (such as heart and lung) should be present as well. Crohn's disease
can be difficult to distinguish from EG on a biopsy, but granulomas, pseudopyloric metaplasia,
architectural distortion, and fibrosis are typically not features of EG. In addition, the inflammatory
infiltrate of Crohn's disease is usually mixed and not virtually exclusively eosinophilic. The features
of EG on resection are usually easy to distinguish from Crohn's disease. Celiac disease may feature
increased eosinophils, and EG may feature villous blunting so severe as to mimic celiac disease.
However, the intra-epithelial lymphocytosis, crypt hyperplasia, and serologic findings of celiac disease
are absent in EG. Enterocolitis secondary to food or drug allergies can mimic EG, and gastrointestinal
food allergies may actually be a trigger for EG (see above). Knowledge of ingested food or drugs prior
to onset of symptoms can help address this.
When the biliary tree is involved by EG, radiographic and histologic findings may closely mimic
primary sclerosing cholangitis. The number of eosinophils present histologically, the presence of
peripheral eosinophilia, and the knowledge of EG in other parts of the GI tract can be helpful in the
differentiation from PSC.
Diagnosis. The clinicopathologic criteria for the diagnosis of EG are as
follows:
1. The presence of gastrointestinal symptoms
2. Histologic evidence of an eosinophilic infiltrate in one or more
areas of the GI tract
3. Exclusion of other causes of gastrointestinal eosinophil infiltrates,
such as parasites, idiopathic hypereosinophilic syndrome, etc.
Patients with gastrointestinal symptoms and peripheral eosinophilia, but no biopsy evidence of an
eosinophilic infiltrate, do not meet the criteria for EG.
Histologic diagnosis of EG may be obtained from either biopsy or surgical specimens, but diagnosis may
be missed on biopsy if the mucosa is not involved. Surgical resection of an obstructed segment of small
bowel is frequently performed prior to diagnosis, as making a definitive tissue diagnosis on mucosal
biopsy specimens can be very challenging. Cytologic evaluation of ascitic fluid for an eosinophilic
infiltrate can be helpful, particularly in the serosal form of EG.
Radiographic studies may demonstrate thickening of the gut wall and/or evidence of obstruction.
Endoscopic evaluation may show thickened gastric rugal folds or intestinal pseudopolyps, but as mentioned
above the mucosa may be normal if the disease is predominantly mural or serosal.
Treatment
Steroids are the mainstay of therapy. Sodium chromoglycerate
has been tried with variable success, and although trial elimination diets have been attempted, these do
not appear to be successful and relapse is common. Surgery may be required to relieve obstruction or
perforation.
References
- Talley NJ, Shorter RG, Phillips SF, Zinsmeister AR. Eosinophilic gastroenteritis: a clinicopathologic study of patients with disease of the mucosa, muscle layer, and subserosal tissues. Gut 31:54-8, 1990.
- Klein NC, Hargrove RL, Sleisenger MH, et al. Eosinophilic gastroenteritis. Medicine 49:299-319, 1970.
- Le CD, Nguyen H. Eosinophilic gastroenteritis, ascites, and pancreatitis: a case report and review of the literature. Southern Med J 97:906-5, 2004.
- Kelly KJ. Eosinophlic gastroenteritis. J Pediatr Gastroenterol Nutr 30:S28-S35, 2000.
- Kristopaitis T, Neghme C, Yong SL, et al. Giant antral ulcer: a rare presentation of eosinophilic gastroenteritis-case report and review of the literature. Am J Gastroenterol 92:1205-1208, 1997.
- McNabb PC, Fleming CR, Higgins JA, Davis GL. Transmural eosinophilic gastroenteritis with ascites. Mayo Clin Proc 54:119-22, 1979.
- Naylor AR, Pollet JE. Eosinophilic colitis. Dis Colon Rectum 28:615-18, 1985.
- Blackshaw AJ, Levison DA. Eosinophilic infiltrates of the gastrointestinal tract. J Clin Pathol 39:1-7, 1986.
- Schoonbroodt D, Horsmans Y, Laka A, et al. Eosinophilic gastroenteritis presenting with colitis and cholangitis. Dig Dis Sci 40:308-14, 1995.
- Rothenberg ME. Eosinophilic gastrointestinal disorders. J Allergy Clin Immunol 113:11-28, 2004.
- Wershil BK, Walker WA. The mucosal barrier, IgE-mediated gastrointestinal events, and eosinophilic gastroenteritis. Gastroenterol Clin North Amer 21:387-404, 1992.
