Gastrointestinal Tract Infections
59
GENERAL CONSIDERATIONS
ANATOMY
We are all connected to the external environment
through our gastrointestinal (GI) tract (Figure 59-1).
What we swallow enters the GI tract and passes through
the esophagus into the stomach, through the small and
large intestines, and finally to the anus. During passage,
fluids and other components are added to this material
as secretory products of individual cells and as enzy-
matic secretions of glands and organs, and are removed
from this material by absorption through the gut
epithelium.
The major components of the tract are listed in
Box 59-1. The nature of the epithelial cells lining the GI
tract varies with each portion. The lining of the GI tract
is called the mucosa. Because of the differing nature of
the mucosal surfaces of various segments of the bowel,
specific infectious disease processes tend to occur in each
segment.
The wall of the small intestine has folds that have
millions of tiny, hairlike projections called villi. Each
villus contains an arteriole, venule, and lymph vessel
(Figure 59-2). The function of villi is to absorb fluids and
nutrients from the intestinal contents. Epithelial cells
lining the surface of villi have a surface that resembles a
fine brush and is referred to as a brush border. The
brush border is formed by nearly 2000 microvilli per
epithelial cell. Intestinal digestive enzymes are produced
in brush border cells toward the top of the villi. Villi and
microvilli help make the small intestine the primary site
of digestion and absorption by significantly increasing
the surface area; more than 90% of physiologic net fluid
absorption occurs here. Mucus-secreting goblet cells are
found in large numbers of villi and intestinal crypts.
Similar to the small intestine, the large intestine is
composed of several segments (see Box 59-1). The wall
of the large intestine consists of columnar epithelial
cells, many of which are mucus-producing goblet cells.
In contrast to the small intestine, there are no villous
projections into the lumen. The remaining excess fluid
within the GI tract is resorbed by the cells lining the
large intestine before waste is finally discharged through
the rectum.
In addition to the previously discussed compo-
nents of the GI tract, numerous other organs and struc-
C H A P T E R
tures are either located in the main digestive organs or
open into them. These accessory organs and structures
include the salivary glands, tongue, teeth, liver, gall-
bladder, and pancreas. Except for the teeth and salivary
glands, these organs are illustrated in Figure 59-1.
RESIDENT MICROBIAL FLORA
The GI tract contains vast, diverse normal flora.
Although the acidity of the stomach prevents any signi-
ficant colonization in a normal host under most cir-
cumstances, many species can survive passage through
the stomach to become resident within the lower
intestinal tract. Normally, the upper small intestine
contains only sparse flora (bacteria, primarily strep-
tococci; lactobacilli; and yeasts; 101 to 103/mL), but in
the distal ileum, counts are about 106 to 107/mL, with
Enterobacteriaceae and Bacteroides spp. predominately
present.
Infants usually are colonized by normal human
epithelial flora, such as staphylococci, Corynebacterium
spp., and other gram-positive organisms (bifidobacteria,
clostridia, lactobacilli, streptococci), within a few hours
of birth. Over time, the content of the intestinal flora
changes. The normal flora of the adult large bowel
(colon) is established relatively early in life and consists
predominantly of anaerobic species, including Bacte-
roides, Clostridium, Peptostreptococcus, Bifidobacterium, and
Eubacterium.
Aerobes, including Escherichia coli, other Entero-
bacteriaceae, enterococci, and streptococci are outnum-
bered by anaerobes 1000:1. The number of bacteria per
gram of stool within the bowel lumen increases steadily
as material approaches the sigmoid colon (the last seg-
ment). Eighty percent of the dry weight of feces from a
healthy human consists of bacteria, which can be pre-
sent in numbers as high as 1011 to 1012 colony-forming
units (CFU)/g of stool.
GASTROENTERITIS
SCOPE OF THE PROBLEM
Worldwide, diarrheal diseases are the second leading
cause of death; about 25 million enteric infections occur
each year. These infections cause significant morbidity
873
874 Part VII DIAGNOSIS BY ORGAN SYSTEM

Oral cavity
Pharynx
Tongue
Epiglottis

Trachea
Esophagus

Diaphragm

Stomach
Liver
Spleen
Gallbladder

Duodenum Pancreas
Transverse colon

Descending colon
Ascending colon
Small intestine

Cecum

Appendix
Sigmoid colon

Rectum

Anus

Figure 59-1 General anatomy of the gastrointestinal tract. (From Broadwell DC, Jackson BS:
Principles of ostomy care, 1982, St Louis, Mosby.)

and death, particularly in elderly people and children PATHOGENESIS

younger than 5 years of age. It has been estimated that 4
to 6 million children die each year of diarrheal diseases,
particularly in developing countries in Asia and Africa.
Even in developed countries, significant morbidity
occurs as a result of diarrheal illness. Although acute
diarrheal syndromes are usually self-limited, some per-
sons with infectious diarrhea require diagnostic studies
and treatment.
Similar to the pathogenesis of urinary tract infections,
the host and the invading microorganism possess key
features that determine whether an enteric pathogen is
able to cause microbial diarrhea.
Host Factors
The human host has numerous defenses that nor-
mally prevent or control disease produced by enteric

BOX 59-1 Components of the Gastrointestinal Tract
Mouth
Oropharynx
Esophagus
Stomach
• Fundus: enlarged portion of the stomach to the left and above
the opening of the esophagus into the stomach
• Body: central part of the stomach
• Pylorus: lower portion of the stomach
Small intestine
• Duodenum: uppermost division; attached to pyloric end of
the stomach
• Jejunum: midsection of the small intestine
• Ileum: lower portion of the small intestine
Large intestine
• Cecum
• Colon
Ascending colon: lies on the right side of the abdomen and ex-
tends up to the lower portion of the liver; the ileum joins the large
intestine at the junction of the cecum and the ascending colon
Transverse colon: passes horizontally across the abdomen
Descending colon: lies on the left side of the abdomen in a
vertical position
Sigmoid colon: extends downward, subsequently joining the
rectum
• Rectum
• Anal canal
pathogens. For example, the acidity of the stomach
effectively restricts the number and types of organisms
that enter the lower GI tract. Normal peristalsis helps
move organisms toward the rectum, interfering with
their ability to adhere to the mucosa. The mucous layer
coating the epithelium entraps microorganisms and
helps propel them through the gut. The normal flora
prevents colonization by potential pathogens.
Mucous membranes line the GI tract, as well as the
respiratory and urogenital tracts. Although technically
inside the body, some of these membranes can be con-
sidered as outside in the sense that they are exposed to
the external environment in the form of food, water,
and air. These membranes contain multiple cell types;
some are secreting or absorbing cells that perform
physiologic functions of the membrane while others
serve a protective function. For example, sets of spe-
cialized cells called follicles are part of the mucous
membrane lining the GI tract and serve a protective
function. Collections of follicles are called Peyer’s
patches. Follicles contain M cells, macrophages, and B
and T cells. As a result of the collective action of the
follicle components following uptake and processing of
the bacteria or antigens, secretory immunoglobulin
A (sIgA) is released. Phagocytic cells and sIgA within
the gut help destroy etiologic agents of disease, as do
eosinophils, which are particularly active against para-
sites. Follicles and Peyer’s patches are found in the small
and large intestines.
Chapter 59 Gastrointestinal Tract Infections 875
Other factors that come into play include the host’s
personal hygiene and age. An initial step in the patho-
genesis of enteric infections is ingestion of the pathogen.
The majority of enteric pathogens, including bacteria,
viruses, and parasites, is acquired by the fecal-oral
route. Enteric infections can be spread by conta-
mination of food products or drinking water and then
subsequent ingestion. The age of the host also plays a
role in whether disease is established. For example,
diarrheal infections caused by rotavirus or entero-
pathogenic Escherichia coli tend to affect young child-
ren and not adults.
Finally, the normal intestinal flora is an important
factor in the host response to the introduction of a
potentially harmful microorganism. Whenever a reduc-
tion in normal flora occurs because of antibiotic treat-
ment or some host factor, resistance to GI infection is
significantly reduced. The most common example of the
protective effect of normal flora is the development of
the syndrome pseudomembranous colitis (PMC). This
inflammatory disease of the large bowel is caused by the
toxins of the anaerobic organism Clostridium difficile and
occasionally other clostridia and perhaps even Staphylo-
coccus aureus, and seldom occurs except after antimicro-
bial or antimetabolite treatment has altered the normal
flora. Almost every antimicrobial agent and several can-
cer agents have been associated with the development
of PMC. C. difficile, usually acquired from the hospital
environment, is suppressed by normal flora. When nor-
mal flora are reduced, C. difficile is able to multiply and
produce its toxins. This syndrome is also known as
antibiotic-associated colitis. Other microorganisms that may
gain a foothold when released from selective pressure of
normal flora include Candida spp., staphylococci, Pseudo-
monas spp., and various Enterobacteriaceae.
Microbial Factors
The ability of an organism to cause GI infection depends
not only on the susceptibility of the human host to the
invading organism but also on the organism’s virulence
traits. To cause GI infection, a microorganism must
possess one or more factors that allow it to overcome
host defenses or it must enter the host at a time when
one or more of the innate defense systems is inactive.
For example, certain stool pathogens are able to survive
gastric acidity only if the acidity has been reduced by
bicarbonate, other buffers, or by medications for ulcers
(e.g., cimetidine, ranitidine, H2 blockers). Pathogens
ingested with milk have a better chance of survival,
because milk neutralizes stomach acidity. Organisms
such as Mycobacterium tuberculosis, Shigella, E. coli
O157:H7, and C. difficile (a spore-forming Clostridium
spp.) are able to withstand exposure to gastric acids and
thus require much smaller infectious inocula than do
acid-sensitive organisms such as Salmonella.
876 Part VII DIAGNOSIS BY ORGAN SYSTEM

Segment of jejunum

Serosa
Lumen
Mesentery
Plica (fold)

Interstitial
crypts
Muscularis
Submucosa Submucosa
Plica (fold)
Mucosa
Lymph node

Muscle
Serosa

Epithelium
of villus Three-dimensional magnification
of jejunal wall

Artery
Three cells of the
villus epithelium
showing brush
Vein border (microvilli)

Single villus
Figure 59-2 Wall of the small intestine. Villi cover the folds of the mucosal layer; in turn, each villus is covered with
epithelial cells.

Primary Pathogenic Mechanisms. Because the
normal adult GI tract receives up to 8 L of ingested fluid
daily, plus the secretions of the various glands that
contribute to digestion (salivary glands, pancreas, gall-
bladder, stomach), of which all but a small amount must
be resorbed, any disruption of the normal flow or
resorption of fluid will profoundly affect the host.
Depending on how they interact with the human host,
enteric pathogens may cause disease in one or more of
the following three ways:
● By causing cell destruction and/or a marked inflam-
matory response following invasion of host cells and
possible cytotoxin production, usually in the colon
● By penetrating the intestinal mucosa with subse-
quent spread to and multiplication in lymphatic or
reticuloendothelial cells outside of the bowel; these
infections are considered systemic infections
Examples of microorganisms for each of these
pathogenic mechanisms are listed in Table 59-1.

● By changing the delicate balance of water and Toxins

electrolytes in the small bowel, resulting in massive ENTEROTOXINS. Enterotoxins alter the metabolic
fluid secretion. In many cases, this process is activity of intestinal epithelial cells, resulting in an out-
mediated by enterotoxin production. This is a non- pouring of electrolytes and fluid into the lumen. They
inflammatory process act primarily in the jejunum and upper ileum, where
 Chapter 59 Gastrointestinal Tract Infections 877

Table 59-1 Examples of Microorganisms That Cause GI Infection inhabits sea and stagnant water and is spread in con-
for Each Primary Pathogenic Mechanism taminated water. The organsims have been isolated
from coastal waters of several states, and sporadic cases
Mechanism Examples of Microorganisms of cholera occur in the United States. Additional infor-
mation about V. cholerae is provided in Chapter 28.
Toxin Production
Other organisms also produce a cholera-like
Enterotoxin Vibrio cholerae enterotoxin. A group of vibrios similar to V. cholerae but
Noncholera vibrios
serologically different, known as the noncholera
Shigella dysenteriae type 1
Enterotoxigenic Escherichia coli vibrios, produce disease clinically identical to cholera,
Salmonella spp. effected by a very similar toxin. The heat-labile toxin
Clostridium difficile (toxin A) (LT) elaborated by certain strains of E. coli, called ente-
Aeromonas rotoxigenic E. coli (ETEC), is similar to cholera toxin,
Campylobacter jejuni
sharing cross-reactive antigenic determinants. The ente-
Cytotoxin Shigella spp. rotoxins of some Salmonella spp. (including S. arizonae),
Clostridium difficile (toxin B) Vibrio parahaemolyticus, the Campylobacter jejuni group,
Enterohemorrhagic Escherichia coli
Clostridium perfringens, Clostridium difficile, Bacillus cereus,
Neurotoxin Clostridium botulinum Aeromonas, Shigella dysenteriae, and many other Entero-
Staphylococcus aureus
bacteriaceae also cause positive reactions in at least one of
Bacillus cereus
the tests for enterotoxin (discussed below). The exact
Attachment Within or Enteropathogenic Escherichia coli contribution of these enterotoxins to the pathogenicity
Close to Mucosal Enterohemorrhagic Escherichia coli
Cells/Adherence Cryptosporidium parvum of most stool pathogens remains to be elucidated.
Isospora belli Certain strains of E. coli, in addition to producing a
Rotavirus heat-labile toxin similar to cholera toxin (LT), also
Hepatitis A, B, C produce a heat-stable toxin (ST) with other properties.
Norwalk virus Although ST also promotes fluid secretion into the
Invasion Shigella spp. intestinal lumen, its effect is mediated by activation of
Enteroinvasive Escherichia coli guanylate cyclase, resulting in increased levels of cyclic
Entamoeba histolytica
guanylate monophosphate (GMP), which yields the
Balantidium coli
Campylobacter jejuni same net effect as increased cAMP. Tests for ST include
Plesiomonas shigelloides enzyme-linked immunosorbent assay (ELISA), immu-
Yersinia enterocolitica nodiffusion, cell culture, and the classic suckling mouse
Edwardsiella tarda assay, in which culture filtrate is placed into the stomach
of a suckling mouse, with the intestinal contents later
measured for fluid volume increase. Molecular tech-
niques, including the use of DNA probes as well as
most fluid transport takes place. The stool of patients several amplification assays, have been used to identify
with enterotoxic diarrheal disease involving the small ETEC directly in clinical samples or isolated bacterial
bowel is profuse and watery, and blood or polymor- colonies.
phonuclear neutrophils are not prominent features. Although several tests are available for the detec-
The classic example of an enterotoxin is that of tion of enterotoxin, they are not performed routinely in
Vibrio cholerae (Figure 59-3). This toxin consists of two diagnostic microbiology laboratories. These tests include
subunits, A and B.21 The A subunit is composed of one the ligated rabbit ileal loop test, the Chinese hamster
molecule of A1, the toxic moiety, and one molecule of ovary cell assay, and the Y-1 adrenal cell assay. Because
A2, which binds an A1 subunit to five B subunits. The B many enterotoxins are antigenic, homologous antibo-
subunits bind the toxin to a receptor (a ganglioside, an dies can be used to identify them specifically. Immuno-
acidic glycolipid) on the intestinal cell membrane. Once diffusion, ELISA, and latex agglutination tests are all
bound, the toxin acts on adenylate cyclase enzyme, available to identify specific toxins. Molecular probes
which catalyzes the transformation of adenosine tri- and amplification assays for toxin detection are also
phosphate (ATP) to cyclic adenosine monophosphate available primarily for research use.
(cAMP). Increased levels of cAMP stimulate the cell to
actively secrete ions into the intestinal lumen. To CYTOTOXINS. The second category of toxins, cyto-
maintain osmotic stabilization, the cells then secrete toxins, acts to disrupt the structure of individual
fluid into the lumen. The fluid is drawn from the intra- intestinal epithelial cells. When destroyed, these cells
vascular fluid store of the body. Patients therefore can slough from the surface of the mucosa, leaving it raw
become dehydrated and hypotensive rapidly. V. cholerae and unprotected. The secretory or absorptive functions
878 Part VII DIAGNOSIS BY ORGAN SYSTEM

Figure 59-3 Diagrammatic representation

of the structure and action of cholera toxin.
B subunits

A2 subunit
s
s Gml ganglioside
receptor
ss
Cell membrane

A1 subunit Adenylate
cyclase
1. 2.
B unit remains
membrane bound.
1. Cholera toxin binds to cell
membrane via B subunits.

2. B subunits change conformation

to allow A subunits to enter
membrane.

3. A subunits dissociate and A1

activates adenylate cyclase.

3. ATP cAMP

of the cells are no longer performed. The damaged tissue
evokes a strong inflammatory response from the host,
further inflicting tissue damage. Numerous polymor-
phonuclear neutrophils and blood are often seen in the
stool, and pain, cramps, and tenesmus (painful straining
during a bowel movement) are common symptoms. The
term dysentery refers to this destructive disease of the
mucosa, almost exclusively occurring in the colon. Cyto-
toxin has not yet been shown to be the sole virulence
factor for any etiologic agent of GI disease, because most
agents produce a cytotoxin in conjunction with another
factor.
E. coli strains seem to possess virulence mechan-
isms of many types.15,20 Some strains produce a cyto-
toxin that destroys epithelial cells and blood cells.
Certain strains produce a cytotoxin that affects Vero cells
(African green monkey kidney cells) and resemble the
cytotoxin produced by Shigella dysenteriae (Shiga toxin);
such strains of E. coli are associated with hemorrhagic
colitis and the sequelae following infection of hemolytic-
uremic syndrome (HUS) and thrombotic thrombocyto-
penia purpura (TTP).8,9,15 These strains of E. coli are
referred to as enterohemorrhagic E. coli (EHEC).
Table 59-2 summarizes the key pathogenic features of
the primary groups of diarrheagenic E. coli.
C. difficile produces a cytotoxin, the presence of
which is a most useful marker for diagnosis of PMC.
S. dysenteriae, Staphylococcus aureus, C. perfringens, and
V. parahaemolyticus produce cytotoxins that probably
contribute to the pathogenesis of diarrhea, although they
may not be essential for initiation of disease. Other vibrios,
Aeromonas hydrophila (a relatively newly described agent
of GI disease), and Campylobacter jejuni, the most com-
mon cause of GI disease in many areas of the United
States, have been shown to produce cytotoxins. The role
that these toxins play in the pathogenesis of the disease
syndromes is not yet completely delineated.
NEUROTOXINS. Food poisoning, or intoxication,
may occur as a result of ingesting toxins produced by
microorganisms. The microorganisms usually produce
their toxins in foodstuffs before they are ingested; thus
the patient ingests preformed toxin. Strictly speaking,
these syndromes are not GI infections but rather into-
xications; because they are acquired by ingestion of
microorganisms or their products, they are considered in
this chapter. Particularly in staphylococcal food poison-
ing and botulism, the causative organisms may not be
present in the patient’s bowel at all.
Bacterial agents of food poisoning that produce
neurotoxins include Staphylococcus aureus and Bacillus
cereus. Toxins produced by these organisms cause vomit-
ing, independent of other actions on the gut mucosa.
Staphylococcal food poisoning is one of the most fre-
quently reported categories of food-borne disease. The
organisms grow in warm food, primarily meat or dairy
 Chapter 59 Gastrointestinal Tract Infections 879

Table 59-2 Overview of the Primary Groups of E. coli That Cause Diarrhea in Humans

Type Primary Mode of Pathogenesis Other Comments

Enterotoxigenic (ETEC) Produces heat-labile (LT) and/or heat stable (ST) enterotoxins; Common cause of traveler’s diarrhea; infects all ages
genes of both toxins reside on a plasmid. LTs are closely
related in structure and function to cholera toxin. STs result in
net intestinal fluid secretion by stimulating guanylate cyclase.
Enteroaggregative (EAEC) Binds to small intestine cells via fimbriae encoded by a large Infects primarily young children
molecular weight plasmid, forming small clumps of bacteria on
the cell surface. Other plasmid-borne virulence factors include
structured pilin, a heat-stable enterotoxin, novel anti-aggregative
protein, and a heat-labile enterotoxin, all believed to be the
cause of the associated diarrhea.
Enteroinvasive (EIEC) Pathogenesis has yet to be totally elucidated. Studies suggest that Very difficult to distinguish from Shigella spp. and
mechanisms by which diarrhea results are virtually identical to other E. coli strains
those of Shigella spp.
Enteropathogenic (EPEC) Initially attaches in the colon and small intestine and then becomes Diarrhea in infants, particularly in large urban
intimately adhered to intestinal epithelial cells, subsequently hospitals
causing the loss of enterocyte microvilli (effacement). Genes for
attachment/effacement reside in a cluster on the bacterial
chromosome (i.e. pathogenicity island).
Enterohemorrhagic (EHEC) Attaches to and effaces gut epithelial cells in a similar manner as Although many outbreaks are caused by E. coli
EPEC. In addition, EHEC elaborates Shiga toxins. O157:H7, other serotypes have been implicated in
outbreaks and sporadic cases

products, and produce the toxin. Onset of disease is
usually within 2 to 6 hours of ingestion. B. cereus pro-
duces two toxins, one of which is preformed, called the
emetic toxin, because it produces vomiting. The second
type, probably involving several enterotoxins, causes
diarrhea. Often acquired from eating rice, B. cereus has
also been associated with cooked meat, poultry, vege-
tables, and desserts.
Perhaps the most common cause of food poisoning
is from type A Clostridium perfringens, which produces
toxin in the host after ingestion. As a result, a relatively
mild, self-limited (usually 24-hour) gastroenteritis
occurs often in outbreaks in hospitals. Meats and gravies
are typical offending foods.
One of the most potent neurotoxins known is pro-
duced by the anaerobic organism Clostridium botulinum.
This toxin prevents the release of the neurotransmitter
acetylcholine at the cholinergic nerve junctions, causing
flaccid paralysis. The toxin acts primarily on the peri-
pheral nerves but also on the autonomic nervous
system. Patients exhibit descending symmetric paralysis
and ultimately die of respiratory paralysis unless they
are mechanically ventilated. In most cases, adult
patients who develop botulism have ingested the pre-
formed toxin in food (home-canned tomato products and
canned, cream-based foods are often implicated), and the
disease is considered to be an intoxication, although
C. botulinum has been recovered from the stools of many
adult patients. A relatively recently recognized syn-
drome, infant botulism, is a true GI infection. In adults,
the normal flora probably prevents colonization by
C. botulinum, whereas the organism is able to multiply
and produce toxin in the infant bowel. Infant botulism is
not an infrequent condition; babies acquire the organism
by ingestion, although the source of the bacterium is not
always clear. Because an association has been found
with honey and corn syrup, infants younger than 9
months of age should not be fed honey. The effect of the
toxin is the same, whether ingested in food or produced
by growing organisms within the bowel.
Attachment. An organism’s ability to cause disease
can also depend on its ability to colonize and adhere to a
relevant region of the bowel. To illustrate, ETEC must be
able to adhere to and colonize the small intestine, as
well as produce an enterotoxin. These organisms pro-
duce an adherence antigen, called colonization factor anti-
gen (CFA), that gives the organism this adherence
capacity. Certain strains of E. coli referred to as the ente-
ropathogenic E. coli (EPEC), attach and then adhere to
the intestinal brush border. This localized adherence is
mediated by the production of pili. Subsequent to attach-
ing, EPEC disrupts normal cell function by effacing the
brush epithelium, thereby causing diarrheal disease. This
complete process is referred to as attachment and
effacement. Genes responsible for the initial adherence
of ETEC, EHEC, and EPEC to intestinal epithelial cells
reside on a transmissible plasmid.15,20 Of note, EHEC
880 Part VII DIAGNOSIS BY ORGAN SYSTEM
Origins of EHEC
SLT
O157:H7
Intimin +
O55:H7
or
Intimin +
Non-O157:H7
Enteropathogenic E. coli Intimin +
that cause infant diarrhea
Enterohemorrhagic or
Shiga toxin-producing
E. coli that cause
bloody diarrhea
Figure 59-4 It appears that the presence of EHEC strain
O157:H7 has actually increased in recent years and was not
simply overlooked before 1982. E. coli O157:H7 strains are closely
related to a Shiga toxin–negative EPEC strain O55:H7. It is
proposed that this EPEC strain O55:H7 became infected by a
bacteriophage that encoded Shiga toxin (SLT); it is now
recognized that more than 100 different E. coli serotypes can
express Shiga toxin.9,20
have the same ability to attach to intestinal epithelial
cells and cause effacement. In addition, EHEC produces
a Shiga toxin that spreads to the bloodstream, causing
systemic damage to vascular endothelial cells of various
organs, including kidney, colon, small intestine, and
lung. EHEC is believed to have arisen as a result of an
EPEC strain having become infected with a bacterio-
phage that carried the Shiga toxin gene (Figure 59-4).
Giardia lamblia, a parasite, has increasingly become
more common as an etiologic agent of GI disease in the
United States. Excreted into fresh water by natural
animal hosts such as the beaver, the organism can be
acquired by drinking stream water or even city water in
some localities, particularly in the Rocky Mountain
states, as well as throughout the world. The organism, a
flagellated protozoan, adheres to the intestinal mucosa
of the small bowel, possibly by means of a ventral
sucker, destroying the mucosal cells’ ability to participate
in normal secretion and absorption. No evidence indi-
cates invasion or toxin production.
Cryptosporidium and Isospora spp., parasitic etiologic
agents of diarrhea in animals and poultry and more
recently recognized as causing human disease, probably
also act by adhering to intestinal mucosa and disrupting
function. Cryptosporidia are often seen in the diarrhea
of patients with acquired immunodeficiency syndrome
(AIDS), as well as in travelers’ diarrhea, day care epide-
mics, and diarrhea in people with animal exposure;
cryptosporidia and Isospora spp. may cause severe, pro-
tracted diarrhea in AIDS patients. Other coccidian
parasites, such as microsporidia, produce diarrhea by
destroying intestinal cell function.
Invasion. Following initial and essential adhe-
rence to GI mucosal cells, some enteric pathogens are
able to gain access to the intracellular environment.
Invasion allows the organism to reach deeper tissues,
access nutrients for growth, and possibly avoid the host
immune system.
In the case of diarrhea caused by Shigella, the
primary mechanism of disease production consists of (1)
the triggering and directing by Shigella of its own entry
into colonic epithelial cells by genes located on a
plasmid, and once internalized, (2) the rapid multiplica-
tion of Shigella in the submucosa and lamina propria and
its intracellular and extracellular spread to other adja-
cent colonic epithelial cells.19 Once in the host cell
cytoplasm, Shigella spp. cause apoptosis and release of
the cytokines interleukin (IL)-1 and IL-8.20 The inflam-
matory response to these cytokines damages the colonic
mucosa and exacerbates (aggravates) the infection. Of
note, the genes for invasiveness are located on a large
invasion plasmid. These activities lead to extensive
superficial tissue destruction. If these two steps do not
occur, one does not get the clinical presentation of
classic dysentery (Table 59-3). The entry process is
illustrated in Figure 59-5.
Salmonellae interact with the apical (top) micro-
villi of colonic epithelial cells, disrupting the brush bor-
der. Similar to Shigella, Salmonella spp. also stimulate the
host cell to internalize them through rearrangements of
host actin filaments and other cytoskeleton proteins.3,4,6
Once the whole bacteria are internalized within endo-
cytic vesicles of the host epithelial cell, organisms begin
to multiply within the vacuoles. In contrast to Shigella
spp. that use the colonic mucosal epithelium as a site of
multiplication, certain serotypes of Salmonella, such as
S. typhi and S. choleraesuis, use the colonic epithelium as a
route to gain access to the submucosal layers, mesen-
teric lymph nodes, and subsequently the bloodstream.
The entry of Salmonella is a complex process that
involves several essential genes, as well as particular
environmental conditions of the host cell; this process is
still being delineated. Many virulence factors for inva-
sion of salmonellae into nonphagocytic cells as well as
their ability to cause systemic infections by surviving in
phagocytic cells and replicating within the Salmonella-
containing vesicle in a variety of eukaryotic cells are
determined by chromosomal genes, many of which are
located within pathogenicity islands. Invasiveness is also
thought to contribute to the pathogenesis of disease
associated with species of vibrios, campylobacters,
Yersinia enterocolitica, Plesiomonas shigelloides, and Edward-
siella tarda.
Certain parasites, particularly Entamoeba histolytica
and Balantidium coli, invade the intestinal epithelium of
the colon as a primary site of infection. The ensuing
amebic dysentery is characterized by blood and nume-
rous white blood cells, and the patient experiences
cramping and tenesmus. Other parasites that are

Table 59-3 Types of Enteric Infections
Pathogenic Mechanism
Upsetting of fluid and electrolyte balance/noninflammatory
Invasion and possible cytotoxin production/ inflammatory
(dysentery)
Penetration with subsequent access to the bloodstream
(enteric fever)
acquired by ingestion, such as Trichinella, may cause tran-
sient bloody diarrhea and pain during migration through
the intestinal mucosa to their preferred sites within the
host.
Other organisms selectively destroy absorptive cells
(e.g., villus tip cells) in the mucosa, disrupting their
normal cell function and thereby causing diarrhea. Rota-
viruses and Norwalk-like viruses are both visualized by
electron microscopy within the absorptive cells at the
ends of the intestinal villi, where they multiply and
destroy cellular function. As a result, the villi become
shortened, and inflammatory cells infiltrate the mucosa,
further contributing to the pathologic condition. In addi-
tion to these viral agents, hepatitis A, B, and C and occa-
sionally enteric adenoviruses have been associated with
diarrheal symptoms in infected patients.
Miscellaneous Virulence Factors. Other virulence
traits appear to be involved in the development of GI
infections and include characteristics such as motility,
chemotaxis, and mucinase production. Also, the pos-
session of certain antigens, such as the Vi antigen of
Salmonella typhi and certain cell wall components, are
also associated with virulence.
CLINICAL MANIFESTATIONS
The clinical symptoms experienced by a patient are
largely dependent on how the enteric pathogen causes
disease. To illustrate, patients infected with an enteric
pathogen that upsets fluid and electrolyte balance have
no fecal leukocytes present in the stool and complain of
watery diarrhea; fever is usually absent or mild.
Although nausea, vomiting, and abdominal pain may
also be present, the dominant feature is intestinal fluid
loss. In contrast, patients infected with an enteric patho-
gen that causes significant cell destruction and inflam-
Chapter 59 Gastrointestinal Tract Infections 881
Major Symptoms Examples of Etiologic Agents
Watery diarrhea Vibrio cholerae
No fecal leukocytes Rotavirus
No fever Norwalk virus
Enterotoxigenic Escherichia coli
Giardia lamblia
Bacillus cereus
Dysenteric-like diarrhea (mucus, Shigella spp.
blood, white cells) Enteroinvasive E. coli
Fever Salmonella enteritidis
Fecal leukocytes Entamoeba histolytica
Signs of systemic infection Salmonella typhi
(headache, malaise, sore throat) Yersinia enterocolitica
Fever
mation have fecal leukocytes present in the stool (Figure
59-6). Their diarrhea is often characterized by the pre-
sence of mucus and possibly blood; in many of these
patients, fever is a prominent component of their disease,
as well as abdominal pain, cramps, and tenesmus.
Finally, patients who become infected with a pathogen
that is able to penetrate the intestinal mucosa of the
small intestine without producing enterocolitis and then
subsequently spread and multiply at other sites will
present with signs and symptoms of a systemic illness
such as headache, sore throat, malaise, and fever;
diarrhea in these patients is not a prominent feature and
is absent or mild in many cases. Features of these three
types of enteric infections are summarized in Table 59-3.
EPIDEMIOLOGY
Gastrointestinal infections occur in numerous epide-
miologic settings. Awareness of these different settings is
important because knowledge of a particular epidemio-
logic setting can help provide a basis for the diagnosis
and clues to possible etiologies. When this knowledge is
combined with clinical findings, the etiology of the infec-
tion can often be narrowed to three or four organisms.
Institutional Settings
Diarrheal illness can be a major problem in institutional
settings such as day care centers, hospitals, and nursing
homes. Because individual hygiene is often difficult to
maintain in these settings, coupled with several organ-
isms with relatively low infecting doses such as Shigella
and Giardia lamblia, numerous outbreaks of diarrheal
illness caused by various organisms have been reported.
Organisms such as Shigella, Campylobacter jejuni, Giardia
lamblia, Cryptosporidium, and rotaviruses have been
reported to cause outbreaks in day care centers. Of
significance, these infections can be spread to family
882 Part VII DIAGNOSIS BY ORGAN SYSTEM
Shigellosis Salmonellosis
Penetration
Cell death
Mucosal abscesses Mesenteric lymph nodes
Bloodstream
Figure 59-5 The invasion of Shigella and Salmonella into
intestinal epithelial cells. (Modified from Sansonetti PJ: Genetic
and molecular basis of epithelial invasion by Shigella species, Rev
Infect Dis 13[suppl 4]:S282, 1991, University of Chicago Press.)
members. Similarly, outbreaks caused by these organ-
isms, as well as hemorrhagic E. coli O157:H7, have been
reported in nursing homes and other extended care
facilities.
Nosocomial diarrheal illness is also a problem for
hospital patients and personnel. Of importance, rota-
viruses, adenoviruses, and coxsackie viruses are also
nosocomially transmitted. In addition to these organ-
isms, Clostridium difficile is a major nosocomial enteric
pathogen in hospitals and other settings, including
nursing homes and extended-care facilities. This organ-
ism is a hardy pathogen that readily survives on fomites
(inanimate objects) such as floors, bed rails, call buttons,
and doorknobs, and on the hands of hospital personnel
caring for the patient. Of great concern has been the
Figure 59-6 Wright’s stain of stool from a patient with
shigellosis showing moderate numbers of polymorphonuclear
cells.
recent emergence of a new strain of C. difficile with
increased virulence and fluoroquinolone resistance.13 By
virtue of partial deletions in a toxin regulatory gene,
tcdC, these isolates are able to produce 16- to 23-fold
more toxin A and B. In addition, a separate binary toxin
has been described that is encoded by cdtA and cdtB
genes; cdtB mediates cell surface binding and cellular
translocation, whereas cdtA disrupts the assembly of the
actin filament causing cell death.11 These strains have
emerged as a cause of geographically dispersed out-
breaks of C. difficile–associated disease. Of significance,
many of the reported cases caused by these strains were
in otherwise healthy patients with minimal or no
exposure to a health care setting.
Traveler’s Diarrhea
Individuals who travel into developing geographic areas
that have poor sanitation are at particularly high risk for
developing diarrhea if they do not pay attention to their
eating and drinking habits. In areas with poor sanita-
tion, enteric pathogens heavily contaminate the water
and food. Although many types of enteric pathogens
can cause diarrhea in travelers, enterotoxigenic E. coli is
a leading cause in Asia, Africa, and Latin America,
accounting for about 50% of cases. Salmonellae, shigel-
lae, Campylobacter spp., vibrios, rotavirus, and Norwalk
virus can also cause diarrhea in travelers, depending on
the area or country they visit.
Food- and Water-Borne Outbreaks
The Centers for Disease Control and Prevention report
that more than 12,000 cases of food-borne illness occur
in the United States each year. Because most of these
illnesses are not reported, some estimate that millions of
cases occur annually. Eating raw or undercooked fish,
shellfish, or meats, and drinking unpasteurized milk

increases the risks of certain bacterial, parasitic, and viral
infections. Many food-borne outbreaks can be traced to
poor hygienic practices of food handlers such as not
washing hands after using the toilet; hepatitis A, Nor-
walk virus, and Salmonella are a few examples of
organisms that have contaminated food during prepara-
tion by a food handler and causing diarrheal disease.
Since 1968, the number of cases of salmonellosis has
gradually increased, with many of these infections asso-
ciated with eating raw or undercooked eggs. Also, the
potential for widespread dissemination of food-borne
pathogens has increased because of factors such as
the tendency to eat outside the home, the export and
import of food sources worldwide, and travel.
In addition to food-borne outbreaks of GI tract
infections, water-borne outbreaks of diarrheal disease
caused by Giardia lamblia and Cryptosporidium have been
traced to inadequately filtered surface water. Recrea-
tional waters, including swimming pools, can also
become contaminated with enteric pathogens such as
Shigella and G. lamblia because of poor toilet facilities or
practices.
Immunocompromised Hosts
GI tract infections in individuals infected with human
immunodeficiency virus (HIV) and other patients who
are immunosuppressed, such as organ transplant reci-
pients or individuals receiving chemotherapy, are a
diagnostic challenge for the clinician and microbiologist.
For example, cytotoxic chemotherapy and/or antibiotic
therapy may predispose patients to develop C. difficile
colitis.
Diarrhea is a common clinical manifestation of
infection with HIV, developing in about 30% to 80% of
cases. Numerous pathogens and opportunistic patho-
gens have been identified and are believed to cause
recurrent or chronic diarrhea. Commonly reported etio-
logic agents are:
● Species of Salmonella, Shigella, and Campylobacter
● Cytomegalovirus
● Cryptosporidia, Isospora belli
● Microsporidia
● Entamoeba histolytica
● Mycobacterial spp.
● Giardia lamblia
ETIOLOGIC AGENTS
Many microorganisms are able to cause enteric infec-
tions. A discussion of each organism is beyond the scope
of this chapter. Rather, these organisms are addressed in
Parts III through VI of the textbook. Table 59-4 sum-
marizes the general characteristics of the more common
agents of enteric infections.
Chapter 59 Gastrointestinal Tract Infections 883
OTHER INFECTIONS OF THE GASTROINTESTINAL
TRACT
Besides causing disease in the small and large intestine,
microorganisms can also infect other sites of the GI tract,
as well as the GI tract’s accessory organs.
ESOPHAGITIS
Infections of the mucosa of the esophagus (esopha-
gitis) can cause painful or difficult swallowing, and/or
the sensation that something is lodged in the throat
while swallowing. Individuals who have esophagitis
usually have local or systemic underlying illnesses such
as hematologic malignancies or HIV infection, or are
receiving immunosuppressive therapy. The most com-
mon etiologic agents are Candida spp. (primarily C. albi-
cans), herpes simplex virus, and cytomegalovirus.
GASTRITIS
Gastritis refers to inflammation of the gastric mucosa.
This illness is associated with nausea and upper abdo-
minal pain; vomiting, burping, and fever may also be
present. A curved organism called Helicobacter pylori is
seen on the surface of gastric epithelial cells of patients
with gastritis. The organism is recovered from gastric
biopsy material obtained endoscopically but not from
stool. Following acute infection, H. pylori can persist for
years in most individuals, with most remaining asymp-
tomatic. H. pylori is also the causative agent of peptic
ulcer disease and a significant risk factor for stomach
cancer.
PROCTITIS
Proctitis is the inflammation of the rectum (distal por-
tion of the large intestine). Common symptoms asso-
ciated with proctitis are itching and a mucous discharge
from the rectum; if the infection progresses, ulcers and
abscesses may form in the rectum. The majority of
infections are sexually transmitted through anal inter-
course. Chlamydia trachomatis, herpes simplex, syphilis,
and gonorrhea are the most common etiologic agents.
MISCELLANEOUS
Unusual agents and those that have not been cultured,
such as the mycobacteria that may be associated with
Crohn’s disease and the bacterium associated with
Whipple’s disease, identified by molecular methods as a
new agent, Trophyrema whipplei, are also candidates as
etiologic agents of GI disease. Occasionally, stool cultures
from patients with diarrheal disease yield heavy growth
of organisms such as enterococci, Pseudomonas spp., or
884 Part VII DIAGNOSIS BY ORGAN SYSTEM

Table 59-4 General Characteristics of the Common Agents of Enteric Infections

Common Sources
or Predisposing Predominant
Organism Condition Distribution Clinical Presentation Pathogenic Mechanism Fecal Leukocytes
Bacillus cereus Meats, vegetables, rice Worldwide Intoxication: vomiting or Ingestion of preformed –
watery diarrhea toxin (food poisoning)
Clostridium botulinum Improperly preserved Worldwide Neuromuscular paralysis Ingestion of preformed –
vegetables, meat, fish toxin (food poisoning)
Staphylococcus Meats, salads, dairy Worldwide Intoxication: vomiting Ingestion of preformed –
aureus products toxin (food poisoning)
Clostridium Meats, poultry Worldwide Watery diarrhea Ingestion of organism –
perfringens followed by toxin
production
Aeromonas Water Worldwide Watery diarrhea or ? Enterotoxin –
dysentery ? Cytotoxin
Campylobacter spp. Water, poultry, milk Worldwide Dysentery ? Invasion +
? Cytotoxins
Clostridium difficile Antimicrobial therapy Worldwide Dysentery Enterotoxin and +/–
cytotoxin

Diarrheagenic
Escherichia coli:
Enteropathogenic ? Worldwide Watery diarrhea Adherence/? invasion –
without multiplication
Enterotoxigenic Food, water Worldwide—more Watery diarrhea Enterotoxin –
prevalent in
developing
countries
Enteroinvasive Food Worldwide Dysentery Invasion, enterotoxin +
Enterohemorrhagic Meats Worldwide Watery, often bloody Cytotoxin –/+
diarrhea
Plesiomonas Fresh water, shellfish Worldwide ? Dysentery Unknown +/–
shigelloides ? Enterotoxin
Salmonella spp. Food, water Worldwide Dysentery Invasion +
(nontyphoidal)
Salmonella typhi Food, water Tropical, developing Enteric fever Penetration + (monocytes, not PMNs)
countries
Shigella spp. Food, water Worldwide Dysentery Invasion +
Shigella dysenteriae Water Tropical, developing Dysentery Invasion, cytotoxin +
countries
Vibrio cholerae Water, shellfish Asia, Africa, Middle Watery diarrhea ? Enterotoxin –/+
East, South and Cytotoxin
North American
(along coastal
areas)
Yersinia enterocolitica Milk, pork, water Worldwide Watery diarrhea and/or ? Invasion –
enteric fever ? Penetration
Giardia lamblia Food, water Worldwide Watery diarrhea Unknown-impaired –
absorption
Cryptospordium Animals, water Worldwide Watery diarrhea ? Adherence –
parvum
Entamoeba histolytica Food, water Worldwide (more Dysentery Invasion, cytotoxin –/+ (amoeba destroy the
common in white cells)
developing
countries)

Table 59-4 General Characteristics of the Common Agents of Enteric Infections (cont’d)
Common Sources
or Predisposing
Organism Condition Distribution
Rotavirus ? Worldwide
Norwalk viruses Shellfish, salads Worldwide
?, Questionable or uncertain; +, positive; –, negative; +/–, more frequently positive;
+/–, more frequently negative.
Klebsiella pneumoniae, not usually found in such num-
bers as normal flora. Only anecdotal evidence suggests
that these organisms actually contribute to the patho-
genesis of the diarrhea. Agents of sexually transmitted
disease may cause GI symptoms when they are intro-
duced into the colon via sexual intercourse. Mycobac-
terium avium-intracellulare complex may be transmitted
in this way, going on to cause systemic disease in
patients with AIDS. The pathogenesis of infections
resulting from Blastocystis hominis (a possible coccidian
etiologic agent of human diarrheal disease) is not well
documented, although these organisms are associated
with GI symptoms.
LABORATORY DIAGNOSIS OF GASTROINTESTINAL
TRACT INFECTIONS
SPECIMEN COLLECTION AND TRANSPORT
If enteric pathogens are to be detected by the laboratory,
adherence to appropriate guidelines for specimen
collection and transport is imperative (see Table 5-1 for a
quick guide to specimen collection, transport, and pro-
cessing). If an etiologic agent is not isolated with the first
culture or visual examination, two additional specimens
should be submitted to the laboratory over the next few
days. Because organisms may be shed intermittently,
collection of specimens at different times over several
days enhances recovery. Certain infectious agents, such
as Giardia, may be difficult to detect, requiring the pro-
cessing of multiple specimens over weeks, duodenal
aspirates (in the case of Giardia), or additional alterna-
tive methods.
General Comments
Specimens that can be delivered to the laboratory within
30 minutes may be collected in a clean, waxed card-
Chapter 59 Gastrointestinal Tract Infections 885
Predominant
Clinical Presentation Pathogenic Mechanism Fecal Leukocytes
Watery diarrhea Mucosal damage –
leading to impaired
absorption in small
intestine
Watery diarrhea Mucosal damage –
leading to impaired
absorption in small
intestine
board or plastic container. Stool for direct wet-mount
examination, Clostridium difficile toxin assay, immuno-
electron microscopy for detection of viruses, and ELISA
or latex agglutination test for rotavirus must be sent to
the laboratory without any added preservatives or
liquids. Volume of a liquid stool at least equal to 1
teaspoon (5 mL) or a pea-sized piece of formed stool is
necessary for most procedures.
Stool Specimens for Bacterial Culture
If a delay longer than 2 hours is anticipated for stools for
bacterial culture, the specimen should be placed in
transport medium. Cary-Blair transport medium best
preserves the viability of intestinal bacterial pathogens,
including Campylobacter and Vibrio spp. However, the
media produced by different manufacturers can vary.14
Most workers recommend reducing the agar content of
Cary-Blair medium from 0.5% to 0.16% (modified) for
maintenance of Campylobacter spp. Buffered glycerol
transport medium does not maintain these bacteria.
Several manufacturers produce a small vial of Cary-Blair
with a self-contained plastic scoop suitable for collecting
samples.
Because Shigella spp. are delicate, a transport
medium of equal parts of glycerol and 0.033 M phos-
phate buffer (pH 7.0) supports viability of Shigella better
than does Cary-Blair. For this purpose, maintaining the
glycerol transport medium at refrigerator or freezer
temperatures yields better results.
If stool is unavailable, a rectal swab may be sub-
stituted as a specimen for bacterial or viral culture, but
it is not as good, particularly for diagnosis in adults.
For suspected intestinal infection with Campylobacter,
the swab must be placed in Cary-Blair transport
medium immediately to avoid drying. Swabs are not
acceptable for the detection of parasites, toxins, or
viral antigens.
886 Part VII DIAGNOSIS BY ORGAN SYSTEM
Stool Specimens for Ova and Parasites
For detection of ova and parasites, specimen preserva-
tion with a fixative is recommended for visual exami-
nation (see Chapter 49).
Stool Specimens for Viruses
Stools for virus culture must be refrigerated if they are
not inoculated onto media into cell cultures within 2
hours. A rectal swab, transported in modified Stuart’s
transport medium or another viral transport medium, is
adequate for recovery of most viruses from feces. See
Chapter 51 for more information regarding collection
and transport of specimens for viral culture.
Miscellaneous Specimen Types
Other specimens that may be obtained for diagnosis of
GI tract infection include duodenal aspirates (usually for
detection of Giardia or Strongyloides), which should be
examined immediately by direct microscopy for the
presence of motile protozoan trophozoites, cultured for
bacteria, and placed into polyvinyl alcohol (PVA) fixative
for subsequent parasitic examination. The laboratory
should be informed in advance that such a specimen is
going to be collected so that the specimen can be pro-
cessed and examined efficiently.
The string test has proved useful for diagnosing
duodenal parasites, such as Giardia, and for isolating
Salmonella typhi from carriers and patients with acute
typhoid fever. The patient swallows a weighted gelatin
capsule containing a tightly wound length of string,
which is left protruding from the mouth and taped to
the cheek. After a predetermined period, during which
the capsule reaches the duodenum and dissolves, the
string, now covered with duodenal contents, is retracted
and delivered immediately to the laboratory. There the
technologist, using sterile-gloved fingers, strips the
mucus and secretions attached to the string and deposits
some material on slides for direct examination and some
material into fixative for preparation of permanent
stained mounts. The technologist also inoculates some
material to appropriate media for isolation of bacteria.
DIRECT DETECTION OF AGENTS OF GASTROENTERITIS
IN FECES
Wet Mounts
A direct wet mount of fecal material, particularly with
liquid or unformed stool, is the fastest method for detec-
tion of motile trophozoites of Dientamoeba fragilis, Enta-
moeba, Giardia, and other intestinal parasites that may
not contribute to disease but may alert the micro-
biologist to the possibility of finding other parasites, such
as Entamoeba coli, Endolimax nana, Chilomastix mesnili, and
Trichomonas hominis. Occasionally the larvae or adult
worms of other parasites may be visualized. Experienced
observers can also see the refractile forms of cryptos-
poridia and many types of cysts on the direct wet
mount, including Cyclospora cayetanensis, a parasite that is
associated with the consumption of contaminated food
such as raspberries.7 If present in sufficient numbers, the
ova of intestinal parasites can be seen.
Examination of a direct wet mount of fecal mate-
rial taken from an area with blood or mucus, with the
addition of an equal portion of Loeffler’s methylene
blue, is helpful for detection of leukocytes, which
occasionally aids in differentiating among the
various types of diarrheal syndromes. Under phase-
contrast and dark-field microscopy, the darting
motility and curved forms of Campylobacter may be
observed in a warm sample. Water, which will immo-
bilize Campylobacter, should not be used. However, for
practical reasons most laboratories do not use a wet
mount. Trained observers working in endemic areas
can recognize the characteristic appearance and moti-
lity of Vibrio cholerae.
Stains
Feces may be Gram stained for detection of certain
etiologic agents. For example, many thin, comma-
shaped, gram-negative bacilli may indicate Campylobacter
infection (if vibrios have been ruled out). In addition,
polymorphonuclear cells may also be detected. Of impor-
tance, an acid-fast stain can be used to detect Cryptospori-
dium spp., mycobacteria, and Isospora spp. Examination
of fixed fecal material for parasites by trichrome or other
stains is covered in Chapter 49. A permanent stained
preparation should be made from all stool specimens
received for detection of parasites.
Antigen Detection
An accurate, sensitive, indirect fluorescent antibody
stain for giardiasis and cryptosporidiosis is commercially
available. These organisms can be visualized easily and
unequivocally with a monoclonal antibody fluorescent
stain (Meridian Diagnostics, Cincinnati, Ohio). Park and
colleagues described a simple and rapid screening pro-
cedure using a direct fluorescent antibody stain for E. coli
O157:H7.18
Enzyme immunoassays (EIAs) or latex agglutina-
tion can detect numerous microorganisms that cause GI
tract infections. For example, EIAs are commercially
available to detect E. coli O157:H7, the presence of the
Shiga toxins produced by EHEC, or the presence of
C. difficile toxins A or A and B.5,10,12,15 In addition,
rotavirus is detected using a solid-phase EIA procedure
or a latex agglutination test. EIA methods are also
available for detection of antigens of Cryptosporidium and
Giardia lamblia as well as E. histolytica. EIA methods
have also been evaluated for detection of certain
bacterial pathogens.17

Molecular Biological Techniques
The development of amplification techniques has led to
numerous publications for the direct detection of
many enteric pathogens, including all major organ-
ism groups—bacteria, viruses, and parasites. Probes
for Salmonella, Shigella, EHEC, and Yersinia are being
evaluated. A disadvantage with probe technology is
that the organism itself is not available for suscepti-
bility testing, which is important for certain bacterial
pathogens (e.g., Shigella) for which susceptibility
patterns vary.
CULTURE OF FECAL MATERIAL FOR ISOLATION OF
ETIOLOGIC AGENTS
Bacteria
Fecal specimens for culture should be inoculated to
several media for maximal yield, including solid agar
and broth. The choice of media is arbitrary and based on
the particular requirements of the clinician and the
laboratory. Recommendations for selection of media are
given in this section.
Organisms for Routine Culture. Stools received for
routine culture in most clinical laboratories in the
United States should be examined for the presence of
Campylobacter, Salmonella, and Shigella spp. under all
circumstances. Detection of Aeromonas and Plesiomonas
spp. should be incorporated into routine stool culture
procedures. The cost of doing a stool examination on
every patient for all potential enteric pathogens is pro-
hibitive. The decision as to what other bacteria are
routinely cultured should take into account the inci-
dence of GI tract infections caused by particular etiologic
agents in the area served by the laboratory. For example,
if the incidence of Yersinia enterocolitica gastroenteritis is
high enough in the area served by the laboratory, then
this agent should also be sought routinely. Similarly,
because of the increasing prevalence of disease caused
by Vibrio spp. in individuals living in high-risk areas of
the United States (seacoast), laboratories in these local-
ities may routinely look for these organisms. Conversely,
unless a patient has a significant travel history, a labora-
tory located in the Midwestern part of the United States
should not routinely look for these organisms except by
special request. Protocols for culture of enterohemorr-
hagic E. coli (e.g., E. coli O157:H7) vary greatly1,15,16;
based on incidence of disease, some laboratories rou-
tinely culture for this organism while others carry out
culture on request only. Other laboratories set up cul-
tures routinely on bloody stool specimens from children
only. Selective or screening media to detect E. coli
O157:H7 also vary greatly such as using a 1% sorbitol-
containing medium (most O157:H7 E. coli are sorbitol-
negative), a specific trypticase blood agar (Unipath
Chapter 59 Gastrointestinal Tract Infections 887
GmbH, Wesel, Germany) or Rainbow Agar O157 (Bio-
log, Inc., Hayward, Calif.).10,15,16
Routine Culture Methods. An in-depth discussion
regarding culture of all enteric pathogens is beyond the
scope of this chapter. Because U.S. laboratories should
routinely examine stools for the presence of Salmonella,
Shigella, and Campylobacter spp., culture of these organ-
isms is addressed. Culture conditions for all other
pathogens, including viruses, are covered in Parts III, IV,
and VI. Specimens received for detection of the most
frequently isolated Enterobacteriaceae and Salmonella and
Shigella spp. should be plated to a supportive medium, a
slightly selective and differential medium, and a mode-
rately selective medium. A highly selective medium
does not seem to be cost effective for most microbiology
laboratories.
Blood agar (tryptic soy agar with 5% sheep blood)
is an excellent general supportive medium. Blood agar
medium allows growth of yeast species, staphylococci,
and enterococci, in addition to gram-negative bacilli. Of
importance, the absence of normal gram-negative fecal
flora and/or the presence of significant quantities of
organisms such as Staphylococcus aureus, yeasts, and Pseu-
domonas aeruginosa can be evaluated. Another benefit of
blood agar is that it allows oxidase testing of colonies.
Several colonies that do not resemble Pseudomonas from
the third or fourth quadrant should be routinely
screened for production of cytochrome oxidase. If many
are present, Aeromonas, Vibrio, or Plesiomonas spp. should
be suspected.
The moderately selective agar should support
growth of most Enterobacteriaceae, vibrios, and other pos-
sible pathogens; MacConkey agar works well. Some
laboratories use eosin-methylene blue (EMB), which is
slightly more inhibitory. All lactose-negative colonies
should be tested further, ensuring adequate detection of
most vibrios and most pathogenic Enterobacteriaceae.
Lactose-positive vibrios (V. vulnificus), pathogenic E. coli,
some Aeromonas spp., and Plesiomonas spp. may not be
distinctive on MacConkey agar.
Salmonella/Shigella. The specimen should also be
inoculated to a moderately selective agar such as Hek-
toen enteric (HE) or xylose-lysine desoxycholate (XLD)
media. These media inhibit growth of most Entero-
bacteriaceae, allowing Salmonella and Shigella spp. to be
detected. Colony morphologies of lactose-negative,
lactose-positive, and H2S-producing organisms are illus-
trated in Figure 59-7. Other highly selective enteric
media, such as salmonella-shigella, bismuth sulfite, de-
oxycholate, or brilliant green, may inhibit some strains
of Salmonella or Shigella. All these media are incubated at
35° to 37° C in air and examined at 24 and 48 hours for
suspicious colonies.
888 Part VII DIAGNOSIS BY ORGAN SYSTEM

A B

C D

E F

Figure 59-7 Colonies of a lactose-positive organism growing on xylose-lysine deoxycholate (XLD) agar (A) and Hektoen
enteric (HE) agar (B). Colonies of Salmonella enteritidis (lactose-negative) growing on XLD (C) and HE agar (D). (Note how
both agars detect H2S production.) Colonies of Shigella (lactose-negative) growing on XLD (E) and HE agar (F).

Campylobacter. Cultures for isolation of Campylo-
bacter jejuni and Campylobacter coli should be inoculated
to a selective agar containing antimicrobial agents that
suppress the growth of normal flora but not of Campy-
lobacter spp. The introduction of a blood-free, charcoal-
containing medium that has more selective antibiotic
components has resulted in better recovery of most
enteropathogenic Campylobacter spp. compared with
earlier media.2 Brucella broth base has yielded less
satisfactory recovery of Campylobacter spp. Commercially
produced agar plates for isolation of campylobacters are
available from several manufacturers. These plates are

incubated in a microaerophilic atmosphere at 42° C and
examined at 24 and 48 hours for suspicious colonies.
Culture methods for other campylobacters that are
associated with GI disease, such as C. hyointestinalis and
C. fetus subsp. fetus, are provided in Chapter 36.
Enrichment Broths. Enrichment broths are some-
times used for enhanced recovery of Salmonella, Shigella,
Campylobacter, and Y. enterocolitica although Shigella
usually does not survive enrichment. Gram-negative
broth (Hajna GN) or selenite F broth yields good reco-
very. Enrichment broths for Enterobacteriaceae should be
incubated in air at 35° C for 6 to 8 hours and then
several drops should be subcultured to at least two
selective medium plates. A commercial system that
allows such broth to be tested for antigen of Salmonella
or Shigella directly has been described; however, the
reported sensitivity is lower than desired. Stool would
be inoculated to broth only initially; those broths that
tested negative could be discarded without subculturing.
Campy-thioglycollate enrichment broth increases the
yields of positive cultures for Campylobacter spp.,
although it is not necessary for routine use. Enrichment
broth for Campylobacter is refrigerated overnight or for a
minimum of 8 hours before a few drops are plated to
Campylobacter agar and incubated at 42° C in a micro-
aerophilic atmosphere.
LABORATORY DIAGNOSIS OF CLOSTRIDIUM
DIFFICILE–ASSOCIATED DIARRHEA
The definitive diagnosis of C. difficile–associated diarrhea
is based on clinical criteria combined with laboratory
testing. Visualization of a characteristic pseudomem-
brane or plaque on endoscopy is diagnostic for pseudo-
membranous colitis and, with the appropriate history of
prior antibiotic use, meets the criteria for diagnosis of
antibiotic-associated pseudomembranous colitis. No
single laboratory test will establish the diagnosis unequi-
vocally. Three tests are currently available for routine
use: culture, detection of cytotoxin by tissue culture, and
antigen detection assays (e.g., enzyme immunoassay,
latex agglutination) for C. difficile toxin. The laboratory
diagnosis of C. difficile–associated diarrhea is reviewed in
Chapter 44.
C a s e S t ud y
A 30-year-old man developed diarrhea with severe
abdominal cramping 3 days after eating in a local
restaurant. He became febrile and weak and went to
his physician. A stool specimen was collected and
immediately hand-carried to the laboratory. Nume-
rous WBCs and bacteria were seen in a wet mount,
Chapter 59 Gastrointestinal Tract Infections 889
but the bacteria were noteworthy in that they were
nonmotile. The patient was treated with ciprofloxacin.
A culture was performed that grew a non–lactose-
fermenting gram-negative rod that was identified as
Shigella sonnei.
QUESTIONS
1. What agent of diarrhea becomes nonviable in a stool
that has not been cultured within 30 minutes of
collection?
2. If the culture is going to be delayed in transit, what
preservative and storage temperature is best for
preservation of the specimen?
3. What method is used to identify Shigella to the species
level, and what is the importance of such identifi-
cations?
R E F E R E N C E S
1. Abbott SL: Laboratory aspects of non-O157 toxigenic E. coli, Clin
Microbiol Newsl 19:105, 1997.
2. Endtz HP, Ruijs GJ, Zwinderman AH, et al: Comparison of six
media, including a semisolid agar, for the isolation of various
Campylobacter species from stool specimens, J Clin Microbiol
29:1007, 1991.
3. Finlay BB, Falkow S: Salmonella interactions with polarized human
intestinal Caco-2 epithelial cells, J Infect Dis 162:1096, 1990.
4. Galán JE, Ginocchio C, Costeas P: Molecular and functional
characterization of the Salmonella typhimurium invasion gene
invA: homology of invA to members of a new protein family,
J Bacteriol 17:4338, 1992.
5. Gavin PJ, Thomson RB: Diagnosis of enterohemorrhagic Esche-
richia coli infection by detection of Shiga toxins, Clin Microbiol Newsl
26:49, 2004.
6. Ginocchio C, Pace J, Galan JE: Identification and molecular cha-
racterization of a Salmonella typhimurium gene involved in
triggering the internalization of Salmonellae into cultured epithelial
cells, Proc Natl Acad Sci U S A 89:5976, 1992.
7. Herwaldt BL, Beach MJ: The return of Cyclospora in 1997: another
outbreak of cyclosporiasis in North America associated with
imported raspberries, Ann Intern Med 130:210, 1999.
8. Kaplan BS: Commentary on the relationships between HUS and
TTP. In Kaplan BS, et al, editors: Hemolytic-uremic syndrome and
thrombotic thrombocytopenic purpura, New York, 1992, Marcel Dekker.
9. Kaye SA, Obrig TG: Pathogenesis of E. coli hemolytic-uremic
syndrome, Clin Microbiol Newsl 18:49, 1996.
10. Kehl SC: Role of the laboratory in the diagnosis of entero-
hemorrhagic Escherichia coli infections, J Clin Microbiol 40:2711,
2002.
11. Loo VG, Poirier L, Miller MA, et al: A predominantly clonal multi-
institutional outbreak of Clostridium difficile-associated diarrhea
with high morbidity and mortality, N Engl J Med 353:2442, 2005.
12. MacKenzie AM, Orrbine E, Hyde L, et al: Performance of the
ImmunoCard STAT! E. coli O157:H7 test for detection of Esche-
richia coli O157:H7 in stools, J Clin Microbiol 38:1866, 2000.
13. McDonald LC, Killgore GE, Thompson A, et al: An epidemic, toxin
gene-variant of Clostridium difficile, N Engl J Med 353: 2433, 2005.
14. Mundy LS, Shanholtzer CJ, Willard KE, et al: An evaluation of
three commercial fecal transport systems for the recovery of
enteric pathogens, Am J Clin Pathol 96:364, 1991.