# 66 - 182 Infections Due to Mixed Anaerobic Organisms

### 182 Infections Due to Mixed Anaerobic Organisms

Neeraj K. Surana, Dennis L. Kasper

Infections Due to Mixed 
Anaerobic Organisms
Anaerobes constitute the predominant class of bacteria of the normal 
human microbiota that reside on mucous membranes and predomi­
nate in many infectious processes, particularly those arising from 
mucosal surfaces. These organisms generally cause disease subsequent 
to the breakdown of mucosal barriers and the leakage of the microbiota 
into normally sterile sites. Infections resulting from contamination by 
the microbiota are usually polymicrobial and involve both aerobic and 
anaerobic bacteria. However, the difficulties encountered in handling 
specimens in which anaerobes may be important and the technical 
challenges entailed in cultivating and identifying these organisms 
in clinical microbiology laboratories continue to leave the anaerobic 
etiology of an infectious process unproven in many cases. Therefore, 
an understanding of the types of infections in which anaerobes can 
play a role is crucial in selecting appropriate microbiologic tools to 
identify the organisms in clinical specimens and in choosing the most 
appropriate treatment, including antibiotics and surgical drainage or 
debridement of the infected site. This chapter focuses on infections 
caused by anaerobic bacteria other than Clostridium species, which are 
covered elsewhere (Chaps. 139 and 159).
■
■HISTORIC PERSPECTIVE
Anaerobic organisms were first identified by Antonie van Leeuwen­
hoek in 1680—nearly a century before oxygen itself was discovered. 
Leeuwenhoek set up culture medium (crushed pepper powder and 
clean rainwater) in two glass tubes—one open to ambient air and the 
other sealed closed—that he incubated for several days. Although he 
did not expect to observe anything in the sealed tube, he was surprised 
to find “animalcules” in both tubes. He noted that these bacteria in the 
sealed tube were “bigger than the biggest sort” in the tube left open to 
air. It was not until the mid- to late nineteenth century that Leeuwen­
hoek’s findings were confirmed by Pasteur and others. However, these 
principles described by Leeuwenhoek underlie the basic pathogenesis 
of anaerobic infections: development of an anaerobic environment in 
a closed space is due to consumption of oxygen by aerobic organisms 
and results in the outgrowth of anaerobic organisms.
■
■DIFFERENCES BETWEEN ANAEROBIC AND 
AEROBIC ORGANISMS
Anaerobic bacteria can be categorized as obligate anaerobes (killed in 
the presence of ≥0.5% oxygen), aerotolerant organisms (can tolerate the 
presence of oxygen but cannot use it for growth), and facultative anaer­
obes (can grow in the presence or absence of oxygen). Most clinically 
relevant anaerobes, such as Bacteroides fragilis, Prevotella melaninogenica, 
and Fusobacterium nucleatum, are relatively aerotolerant. These organ­
isms contrast with obligate aerobes, which require high concentrations 
of oxygen for growth, and microaerophilic organisms, which are dam­
aged by atmospheric concentrations of oxygen (~21%) but require 
low concentrations of oxygen (typically 2–10%) for growth. Given 
that molecular oxygen can reduce to superoxide (O2
−) and hydrogen 
peroxide (H2O2), which are damaging to cells, the ability to tolerate the 
presence of oxygen is due, in part, to the expression of superoxide dis­
mutase and catalase. The variation in anaerobic organisms tolerating 
anywhere from <0.5 to 8% O2 may reflect the amount of these enzymes 
that is produced.
Furthermore, aerobic and anaerobic organisms differ in their 
energy metabolism. Cellular respiration requires establishment of an 
electrochemical gradient across the membrane, resulting in an electric 
potential (often related to a proton gradient) across the membrane. In 
aerobic respiration, electrons are shuttled through an electron transport 
chain, with oxygen as the final electron acceptor. Anaerobic organisms 
can metabolize energy by either anaerobic respiration or fermentation. 

Given that the final electron acceptor in anaerobic respiration (e.g., 
sulfate, nitrate, carbon dioxide, or fumarate) is not as highly oxidizing 
as oxygen, this pathway is less efficient than aerobic respiration and 
produces less ATP per glucose molecule. In contrast, fermentation does 
not use an electrochemical gradient. Rather, it releases energy from an 
organic molecule (e.g., pyruvate and its derivatives) via substrate-level 
phosphorylation and is therefore a less efficient process than either 
aerobic or anaerobic respiration; for comparison, fermentation results 
in ~5% of the energy released by aerobic respiration. For these reasons, 
facultative anaerobes will preferentially utilize oxygen if it is available; 
in oxygen-limiting situations, organisms will use anaerobic respiration 
rather than fermentative processes, if possible.

■
■ANAEROBES OF THE HUMAN MICROBIOTA
Most human mucocutaneous surfaces harbor a rich indigenous nor­
mal microbiota composed of aerobic and anaerobic bacteria. These 
surfaces are dominated by anaerobic bacteria, which often account for 
99.0–99.9% of the cultivable microbiota and range in concentration 
from 103/mL in the nose to 1012/mL in gingival scrapings and the colon 
(Table 182-1). It is interesting that anaerobes inhabit many areas of 
the body that are exposed to air: skin, nose, mouth, and throat. Anaer­
obes are thought to reside in the portions of these sites that either are 
relatively well protected from oxygen (e.g., gingival crevices) or have a 
local anaerobic environment conferred by neighboring aerobic organ­
isms (e.g., tooth surfaces). The ability to cultivate these organisms is 
improving, and—with strict attention to anaerobic conditions—more 
than 80% of the microscopic counts in fecal samples can be cultured. 
However, culture-independent approaches (e.g., sequencing of the 16S 
rDNA gene) show that the overwhelmingly diverse low-abundance 
bacterial species present in the microbiota remain uncultivated. Several 
projects, including the Human Microbiome Project (funded by the U.S. 
National Institutes of Health) and MetaHIT (financed by the European 
Commission), have characterized the normal microbiota of healthy 
individuals and have demonstrated the presence of >10,000 different 
bacterial species in the collective human microbiota. The human gut 
CHAPTER 182
Infections Due to Mixed Anaerobic Organisms
TABLE 182-1  The Anaerobic Human Microbiota: An Overview
ANAEROBIC/
AEROBIC 
RATIO
POTENTIAL PATHOGEN(S)
TOTAL 
BACTERIAa
ANATOMIC SITE
Nose
103–104
2:1
Peptostreptococcus spp., 
Prevotella spp.
Oral cavity
Saliva
108–109
10:1
Fusobacterium nucleatum, 
Prevotella melaninogenica, 
Prevotella oralis group, 
Bacteroides ureolyticus 
group, Peptostreptococcus 
spp.
Tooth surface
1010–1011
1:1
Gingival crevices
1011–1012
103:1
Gastrointestinal tract
Stomach
100–103
1:10
Lactobacillus spp.
Duodenum
101–105
1:1
Lactobacillus spp., 
Streptococcus spp.
Jejunum
103–106
1:1
Streptococcus spp., 
Lactobacillus spp., 
Peptostreptococcus spp.
Ileum
104–109
10:1
Bacteroides spp., 
Streptococcus spp., 
Enterococcus spp.
Cecum and colon
1011–1012
103:1
Bacteroides spp. (principally 
members of the B. fragilis 
group), Prevotella spp., 
Clostridium spp.
Female genital tract
107–109
10:1
Peptostreptococcus spp., 
Bacteroides spp., Prevotella 
bivia
Skin
104–106
100:1
Cutibacterium acnes
aPer gram or milliliter.

alone harbors >1000 bacterial species, with 100–200 species present in 
any given individual.

The major reservoir of anaerobic bacteria is the lower gastrointesti­
nal tract, but these organisms are also present in considerable numbers 
in the oral cavity, skin, and female genital tract (Table 182-1). In the 
oral cavity, the ratio of anaerobic to aerobic bacteria ranges from 1:1 
on the surface of a tooth to 1000:1 in the gingival crevices. Prevotella 
and Porphyromonas species make up much of the indigenous oral 
anaerobic microbiota. Fusobacterium and Bacteroides (non–B. fragilis 
group) species are present in lower numbers. Anaerobic bacteria are 
not found in appreciable numbers in the normal stomach and proximal 
small intestine. In the distal ileum, the microbiota begins to resemble 
that of the colon, where the ratio of anaerobes to aerobic species is high 
(~1000:1). The predominant anaerobes in the human intestine belong 
to the phyla Bacteroidetes and Firmicutes and include a number 
of Prevotella and Bacteroides species (e.g., members of the B. fragilis 
group, such as B. fragilis, B. thetaiotaomicron, B. ovatus, B. vulgatus, 
B. uniformis, and Parabacteroides distasonis) as well as various Clos­
tridium, Peptostreptococcus, Blautia, and Fusobacterium species. In the 
female genital tract, there are ~109 organisms/mL of secretions, with 
an anaerobe-to-aerobe ratio of ~10:1. The predominant anaerobes in 
the female genital tract are Prevotella, Bacteroides, Fusobacterium, Clos­
tridium, and the anaerobic Lactobacillus species. The skin microbiota 
contains anaerobes as well; Cutibacterium acnes (which was previously 
Propionibacterium acnes and will be considered as one of the Propioni­
bacterium species for the remainder of this chapter) is the predominant 
species, and other species of propionibacteria and peptostreptococci 
are present in lower numbers.
PART 5
Infectious Diseases
■
■ANAEROBES AND HUMAN HEALTH
Commensal anaerobes have been implicated as crucial mediators of 
physiologic, metabolic, and immunologic functions in the mammalian 
host. The intestinal microbiota is essential for fermenting dietary car­
bohydrates into forms that are more usable by the host, among which 
polysaccharides are the most abundant biologic source of energy. Of 
the organisms found within the intestines, Bacteroides species express 
the widest array of polysaccharide-degrading enzymes, providing 
important nutrients for both the host and other commensal organisms. 
For example, B. thetaiotaomicron expresses 172 glycosyl hydrolases. 
The anaerobic intestinal microbiota is also responsible for the produc­
tion of secreted products that promote human health (e.g., vitamin K 
and bile acids useful in fat absorption and cholesterol regulation).
One of the most important roles that anaerobes serve as components 
of the normal colonic microbiota is the promotion of resistance to 
colonization. The presence of commensal bacteria effectively interferes 
with colonization by potentially pathogenic bacterial species through 
the depletion of oxygen and nutrients, the production of enzymes and 
toxic end products, and the modulation of the host’s intestinal innate 
immune response. For example, the normal intestinal microbiota plays 
an important role in protection against enteric infections, including 
those due to Salmonella enterica serotype Typhimurium and Clostridium 
difficile.
The anaerobic intestinal microbiota also has immunomodulatory 
properties that help regulate the immune system. The first example 
of this role was demonstrated with B. fragilis, which can balance the 
effector functions of T cells in the peripheral immune system and 
induce colonic regulatory T cells via expression of polysaccharide A 
(PSA). Moreover, B. fragilis expresses a glycosphingolipid that regu­
lates the number of colonic invariant natural killer T cells. There are 
now numerous examples of commensal anaerobes that can modulate 
different aspects of the intestinal and extraintestinal immune system—
everything from specific effector T cells to dendritic cells to antimicro­
bial peptides.
Clearly, the gut microbiota confers many benefits, and its dysregula­
tion may play a role in the pathogenesis of diseases characterized by 
inflammation and aberrant immune responses, such as inflammatory 
bowel disease, rheumatoid arthritis, multiple sclerosis, asthma, and 
type 1 diabetes. Furthermore, the gut microbiota has been associated 
with obesity and metabolic syndrome. A more complete discussion of 

the intersection between the microbiota and human health is covered 
elsewhere (Chap. 484).
■
■ETIOLOGY
There are >10,000 species of bacteria—the overwhelming majority of 
which are anaerobes—in the human microbiota, with each individual 
colonized by hundreds of species. Anaerobic infections occur when the 
harmonious relationship between the host and the host’s microbiota is 
disrupted. Any site in the body is susceptible to infection with these 
indigenous organisms if they are introduced into otherwise sterile tis­
sue, either through disruption of mucosal surfaces (e.g., intestinal per­
foration, ischemia, surgery) or via direct inoculation of organisms into 
tissue (e.g., bite wounds, trauma). Because the sites that are colonized 
by anaerobes contain many species of bacteria, the resulting infections 
are often polymicrobial, involving multiple species of anaerobes in 
combination with synergistically acting facultative and/or microaero­
philic organisms.
Despite the complex array of bacteria in the normal microbiota, 
relatively few genera are isolated commonly from human infections 
(Fig. 182-1). While the specific organisms identified vary with the 
site and source of infection, the etiologic agents typically reflect the 
neighboring microbiota. For example, organisms normally found 
in the oro- and nasopharyngeal microbiota (e.g., P. melaninogenica, 
Fusobacterium necrophorum, F. nucleatum, Peptostreptococcus species, 
Porphyromonas gingivalis, Porphyromonas asaccharolytica, and Actino­
myces species) can cause disease in contiguous areas, including odon­
togenic infections, peripharyngeal space infections, chronic sinusitis, 
and pleuropulmonary infections. In female genital tract infections, 
organisms normally colonizing the vagina (e.g., Prevotella bivia and 
Prevotella disiens) are the most common isolates. Escherichia coli and B. 
fragilis, both of which are components of the intestinal microbiota, are 
the most commonly identified isolates from intraabdominal abscesses. 
Indeed, the B. fragilis group, which encompasses 25 species and 
includes B. thetaiotaomicron, B. vulgatus, B. uniformis, and B. ovatus, 
contains the anaerobic organisms among the most frequently isolated 
from clinical infections.
It is useful to think about anaerobic infectious etiologies with regard 
not only to their anatomic location but also to their microbiologic 
features. While many anaerobic gram-negative bacilli cause disease 
(e.g., Prevotella, Bacteroides, Fusobacterium, and Porphyromonas species), 
Veillonella species, which are part of the oral and intestinal microbiota, are 
among the few anaerobic gram-negative cocci that have been impli­
cated in human disease. Similarly, the peptostreptococci (e.g., P. micros, 
P. asaccharolyticus, and P. anaerobius) and Finegoldia magnus (which 
was previously Peptostreptococcus magnus and will be considered as 
part of the peptostreptococci for the remainder of this chapter) are the 
chief anaerobic gram-positive cocci that have pathogenic potential. 
Clostridium species are the primary anaerobic spore-forming grampositive rods that produce human disease (Chap. 159). Uncommonly, 
anaerobic gram-positive non-spore-forming bacilli cause infection; 

Gram-positive cocci
Clostridium spp.
Other Gram-positive rods
Bacteroides fragilis
Other Bacteroides spp.
Prevotella spp.
Fusobacterium spp.
Porphyromonas spp.
Other Gram-negative rods
Veillonella spp.
FIGURE 182-1  Distribution of anaerobic organisms isolated from clinical 
materials. (Data combined from Y Park et al: Clinical features and prognostic 
factors of anaerobic infections: A 7-year retrospective study. Korean J Intern 
Med 24:13, 2009; and Japanese Association for Anaerobic Infections Research: 
Anaerobic infections (general): Epidemiology of anaerobic infections. J Infect 
Chemother 17:4, 2011.)

C. acnes, a component of the skin microbiota and a cause of foreignbody infections, and Actinomyces species are relevant examples.
■
■PATHOGENESIS
First and foremost, anaerobic infections require an anaerobic environ­
ment with a lowered oxidation-reduction potential. In some circum­
stances, this environment can occur directly—e.g., in tissue ischemia, 
trauma, surgery, or a perforated viscus. In many other situations, the 
infection is polymicrobial, and the facultative organisms maintain a 
lowered oxidation-reduction potential in the local microenvironment 
that allows for the propagation of obligate anaerobes. Once the proper 
anaerobic environment is established, the organisms must still contend 
with the host’s immune defenses. Similar to aerobic organisms, anaer­
obes express an array of virulence factors that help evade host defenses, 
they can form abscesses as a protective measure, and they can act syn­
ergistically with other bacteria to better persist in the host.
Virulence factors associated with anaerobes typically confer the abil­
ity to evade host defenses, adhere to cell surfaces, produce toxins and/
or enzymes, or display surface structures such as capsular polysaccha­
rides and lipopolysaccharides that contribute to pathogenic potential. 
The ability of an organism to adhere to host tissues is often critical to 
the establishment of infection. Some oral species adhere to the epithe­
lium in the oral cavity. P. gingivalis, a common isolate in periodontal 
disease, has fimbriae that facilitate attachment. In supragingival plaque, 
many oral anaerobes are able to attach directly to aerobic bacteria (e.g., 
Streptococcus species) that are adherent to the tooth’s surface. F. nuclea­
tum is a notable example of these secondary colonizers: it expresses 
receptors to which almost all oral bacteria can bind and serves as an 
important bridge between the primary colonizers and subsequent lay­
ers of bacteria. B. fragilis synthesizes pili, fimbriae, and hemagglutinins 
that aid in attachment to host cell surfaces in the intestine.
Anaerobic bacteria produce several exoproteins that can enhance 
the organisms’ virulence. P. gingivalis produces a collagenase that 
enhances tissue destruction. Exotoxins produced by clostridial species, 
including botulinum toxins, tetanus toxin, C. difficile toxins A and B, 
and five toxins produced by Clostridium perfringens, are among the 
most virulent bacterial toxins in mouse lethality assays. Anaerobic 
gram-negative bacteria, such as B. fragilis, P. gingivalis, and Prevotella 
intermedia, possess lipid A molecules (endotoxins) that are 100–1000 
times less biologically potent than endotoxins associated with aero­
bic gram-negative bacteria; these differences relate to variations in 
acylation status, length of fatty acids, and number of phosphate groups. 
This relative biologic inactivity may account for the lower frequency 
of disseminated intravascular coagulation and purpura in anaerobic 
gram-negative bacteremia than in facultative and aerobic gramnegative bacillary bacteremia. An exception is the lipopolysaccharide 
from Fusobacterium, which may account for the severity of Lemierre’s 
syndrome (see “Complications of Anaerobic Head and Neck Infec­
tions,” below).
The most extensively studied virulence factor of the nonsporulating 
anaerobes is the capsular polysaccharide complex of B. fragilis. This 
organism is unique among anaerobes in its potential for virulence dur­
ing growth at normally sterile sites. Although it constitutes only 0.5–1% 
of the normal colonic microbiota, B. fragilis is the anaerobe most com­
monly isolated from intraabdominal infections and bacteremia. In an 
animal model of intraabdominal sepsis, the capsular polysaccharide 
was identified as the major virulence factor of B. fragilis; this polymer 
plays a specific, central role in the induction of abscesses. A series of 
detailed biologic and molecular studies of this virulence factor showed 
that B. fragilis produces at least eight distinct capsular polysaccharides, 
far more than the number reported for any other encapsulated bac­
terium. B. fragilis can exhibit distinct surface polysaccharides either 
alone or in combination by regulating the expression of these differ­
ent capsules in an on–off manner through a reversible inversion of 
DNA segments within the promoters for operons containing the genes 
required for polysaccharide synthesis. Structural analysis of two of 
these polysaccharides, PSA and polysaccharide B (PSB), revealed that 
each polymer consists of repeating units with positively charged free 
amino groups and negatively charged groups. This structural feature is 

rare among bacterial polysaccharides, and the ability of PSA—and, to 
a lesser extent, PSB—to induce abscesses in animals depends on this 
zwitterionic charge motif. Intraabdominal abscess induction is related 
to the capacity of PSA to stimulate macrophages to release cytokines 
and chemokines—in particular, interleukin (IL) 8, IL-17, and tumor 
necrosis factor α (TNF-α)—from resident peritoneal cells through a 
Toll-like receptor 2–dependent mechanism. The release of cytokines 
and chemokines results in the chemotaxis of polymorphonuclear neu­
trophils (PMNs) into the peritoneum, where they adhere to mesothelial 
cells induced by TNF-α to upregulate their expression of intercellular 
adhesion molecule 1 (ICAM-1). PMNs adherent to ICAM-1-expressing 
cells probably represent the nidus for an abscess. PSA also activates T cells 
to produce certain cytokines, including IL-17 and interferon γ, that are 
necessary for abscess formation.

These virulence factors not only promote persistence of the anaer­
obe that produces them but also aid in the survival of bystander organ­
isms and result in bacterial synergies. Clinically, these synergies are 
evidenced by the fact that anaerobic infections typically involve three 
to six different organisms. Examples of this synergistic pathogenesis 
include creation of a favorable environment for growth (e.g., establish­
ment and maintenance of an anaerobic environment by facultative 
organisms), inhibition of host defenses (e.g., production of short-chain 
fatty acids and succinic acid that inhibit the ability of phagocytes to 
clear facultative organisms), provision of necessary growth factors for 
other organisms (e.g., oral diphtheroids that produce vitamin K, which 
is needed by P. melaninogenica), and creation of tissue damage that 
promotes spread of the infection. In these ways, facultative and obligate 
anaerobes synergistically potentiate abscess formation.
CHAPTER 182
APPROACH TO THE PATIENT
Infections Due to Anaerobic Bacteria
The physician must consider several points when approaching the 
patient with a possible infection due to anaerobic bacteria.
Infections Due to Mixed Anaerobic Organisms
1.	 The organisms colonizing mucosal sites are commensals, very 
few of which typically cause disease. When these organisms do 
cause disease, it often occurs in proximity to the mucosal site 
they colonize.
2.	 For anaerobes to cause tissue infection, they must spread beyond 
the normal mucosal barriers.
3.	 Conditions favoring the propagation of anaerobic bacteria, par­
ticularly a lowered oxidation-reduction potential, are necessary. 
These conditions exist at sites of trauma, tissue destruction, 
compromised vascular supply, and necrosis.
4.	 Frequently, a complex array of infecting microbes can be found, 
occasionally with >10 different species isolated from a suppura­
tive site.
5.	 Anaerobic organisms tend to be found in abscess cavities or 
in necrotic tissue. The failure of an abscess to yield organisms 
on routine culture is a clue that the abscess is likely to contain 
anaerobic bacteria. Often smears of this “sterile pus” are found to 
be teeming with bacteria when Gram’s stain is applied. Although 
some facultative organisms (e.g., Staphylococcus aureus) are also 
capable of causing abscesses, abscesses in organs or deeper body 
tissues should call anaerobic infection to mind.
6.	 Gas is found in many anaerobic infections of deep tissues but is 
not diagnostic because it can be produced by aerobic bacteria 
as well.
7.	 Although a putrid-smelling infection site or discharge is con­
sidered diagnostic for anaerobic infection, this manifestation 
usually develops late in the course and is present in only 30–50% 
of cases.
8.	 Some species (the best example being the B. fragilis group) 
require specific therapy. However, many synergistic infections 
can be cured with antibiotics directed at some but not all of the 
organisms involved. Antibiotic therapy, combined with debride­
ment and drainage, disrupts the interdependent relationship

among the bacteria, and some species that are resistant to the 
antibiotic do not survive without the co-infecting organisms.
9.	 Manifestations of severe sepsis and disseminated intravascu­
lar coagulation are unusual in patients with purely anaerobic 
infection.
■
■EPIDEMIOLOGY
Difficulties in the performance of appropriate cultures, contamination 
of cultures by components of the normal microbiota, and the lack of 
readily available, reliable culture techniques have made it challenging 
to obtain accurate data on the incidence or prevalence of anaerobic 
infections. However, anaerobic infections are encountered frequently, 
with anaerobes constituting 7–8% and 13–15% of bacteria isolated 
from inpatients and outpatients, respectively. Bacteremia and soft 
tissue infections are the most common types of anaerobic infection 
(Fig. 182-2). Typically, anaerobic bacteria account for <1% of all cases 
of bacteremia.
■
■CLINICAL MANIFESTATIONS
Although anaerobes can cause infection anywhere in the body, cer­
tain clinical findings and characteristics are commonly found. These 
include abscess formation, putrid purulence (due to volatile fatty acid 
by-products), septic thrombophlebitis, tissue necrosis, and failure 
to respond clinically to broad-spectrum antibiotics that lack activity 
against anaerobes.
Anaerobic Infections of the Mouth, Head, and Neck 
Anaero­
bic bacteria are commonly involved in infections of the mouth, head, 
and neck (Chap. 37). The predominant isolates are components of the 
normal microbiota of the upper airways—mainly Prevotella species, 

P. asaccharolytica, Fusobacterium species, peptostreptococci, and 
microaerophilic streptococci.
PART 5
Infectious Diseases
OROFACIAL INFECTIONS  The most common oral infections are 
odontogenic and include dental caries and periodontal disease (gingi­
vitis and periodontitis). While dental caries usually manifest with pain, 
sensitivity, and discoloration of the tooth, periodontal disease involves 
inflammation of the gums and underlying tissue. In its more severe 
forms, periodontitis can result in difficulty chewing, loose teeth, and 
occasionally tooth loss. Severe orofacial infections typically develop as 
a consequence of dental infection, and the infection can spread from 
the tooth to different anatomic areas that provide the least resistance, 
resulting in periapical, periodontal, or pericoronal infections. If the 
dental surface is completely breached, an endodontic infection (pul­
pitis) can occur. In late stages of pulpitis, the tooth is generally very 
sensitive to heat, but cold stimuli may provide relief. Left untreated, 
pulpitis can progress to invade the alveolar bone and develop into a 
periapical abscess. The abscesses, particularly those involving the sec­
ond and third molars, can occasionally extend into the submandibular, 
Head and neck
Lung
Abdomen
Soft tissue and joints
Bacteremia
Catheter-related
Surgical site infection
FIGURE 182-2  Distribution of types of infection from which anaerobic organisms 
were cultured at a single hospital over a 7-year period. Head and neck infections 
included sinusitis, otitis media, and retropharyngeal abscess; abdominal infections 
included liver abscess, biliary tract infection, bowel obstruction, and intraabdominal 
abscess; catheter-related infections included those related to peritoneal dialysis 
catheters and ventriculoperitoneal shunts. (Data from Y Park et al: Clinical features 
and prognostic factors of anaerobic infections: A 7-year retrospective study. Korean 
J Intern Med 24:13, 2009.)

sublingual, and submental spaces (Ludwig’s angina). This infection 
results in marked local swelling of tissues, with pain, trismus, and supe­
rior and posterior displacement of the tongue. Submandibular swelling 
of the neck and obstruction by the tongue can impair swallowing and 
cause respiratory obstruction. In some cases, tracheotomy is lifesaving.
Microbiologically, dental caries begin with the binding of Strep­
tococcus mutans and Streptococcus sanguis to the tooth surface, with 
subsequent further colonization by anaerobes. In contrast, periodon­
titis is typically associated with P. gingivalis, Tannerella forsythia, 
Aggregatibacter actinomycetemcomitans, and Treponema denticola. 
Fusobacterium, Actinomyces, Peptostreptococcus, and Bacteroides spe­
cies (other than B. fragilis) are the organisms most commonly isolated 
from periapical abscesses.
ACUTE NECROTIZING ULCERATIVE GINGIVITIS  Gingivitis may 
become a necrotizing infection (trench mouth, Vincent’s stomatitis). 
The onset of disease is usually sudden and is associated with painful 
bleeding gums, foul breath, and a bad taste. The gingival mucosa, 
especially the papillae between the teeth, becomes ulcerated and may 
be covered by a yellowish-white or gray “pseudomembrane,” which is 
removable with gentle pressure. Patients may become systemically ill, 
developing fever, malaise, cervical lymphadenopathy, and leukocytosis. 
The infection can sometimes extend into the pharynx, resulting in an 
extremely sore throat, foul breath, and tonsillar pillars that are swollen, 
red, ulcerated, and covered by a pseudomembrane. Prevotella, Treponema, 
and Fusobacterium species have been implicated.
In some cases, acute necrotizing gingivitis can rapidly progress to 
noma (cancrum oris), a gangrenous infection that destroys the soft and 
hard tissues related to the oral cavity. Noma occurs most frequently 
in young children (1–4 years of age) who have immune dysfunction 
related to malnutrition and endemic infections (particularly measles). 
This infection occurs worldwide but is most common in sub-Saharan 
Africa, where the incidence is 1–7 cases per 1000 children. Although 
the pathogenesis is not fully understood, infections with F. necrophorum 
and P. intermedia are thought to be key drivers of this disease. Without 
treatment, the mortality rate is 70–90%.
PERIPHARYNGEAL SPACE INFECTIONS  These infections arise from 
the spread of organisms from the upper airways to potential spaces 
formed by the fascial planes of the head and neck. The etiology is 
typically polymicrobial and represents the normal microbiota of the 
mucosa of the originating site.
Peritonsillar abscess (quinsy) is the most common peripharyngeal 
infection and occurs as a complication of acute tonsillitis. Consistent 
with its association with tonsillitis, adolescents are most commonly 
affected. Patients present with a sore throat, dysphagia, peritonsillar 
swelling, muffled voice, and uvular deviation to the contralateral side. 
The abscess material typically grows group A Streptococcus in conjunc­
tion with obligate anaerobes (e.g., Bacteroides, Prevotella, and Pepto­
streptococcus species) (Chap. 37). Retropharyngeal abscesses typically 
occur in children 2–4 years of age, although they can occur at any age. 
Although a suppurative infection of the retropharyngeal lymph nodes 
is the usual precursor to these abscesses in children, foreign-body 
ingestion and/or local trauma is more commonly the inciting factor in 
adults. The clinical presentation shares many features with peritonsil­
lar abscesses, but difficulty extending the neck and torticollis are more 
common with retropharyngeal abscesses. The etiologic agents are the 
same as in peritonsillar abscesses, with additional aerobic organisms 
(e.g., S. aureus, viridans streptococci) also playing a role.
SINUSITIS AND OTITIS  Anaerobic bacteria have been implicated 
in chronic sinusitis but play little role in acute sinusitis. Numerous 
studies related to the microbiology of chronic sinusitis have been con­
ducted; on average, anaerobic bacteria have been found in two-thirds 
of patients, with many studies demonstrating their presence in >90% 
of patients. Anaerobic bacteria represent ~40% of all bacteria cultured, 
with Peptostreptococcus, Prevotella, and Porphyromonas species the 
most commonly isolated anaerobes. S. aureus and Enterobacteriaceae 
are the aerobes most commonly recovered in chronic sinusitis. Anaero­
bic bacteria have been isolated in ~60% of cases of chronic suppurative 
otitis media in children, but they are not involved in acute otitis media.

COMPLICATIONS OF ANAEROBIC HEAD AND NECK INFECTIONS  Direct 
extension of these infections into contiguous areas can result in addi­
tional disease manifestations. Cranial spread of these infections can 
result in osteomyelitis of the skull or mandible or in intracranial infec­
tions, such as brain abscess and subdural empyema. Caudal spread can 
produce mediastinitis or pleuropulmonary infection. Hematogenous 
complications can also result from anaerobic infections of the head 
and neck. Bacteremia, which occasionally is polymicrobial, can lead 
to endocarditis or other distant infections. Lemierre’s syndrome 
(Chap. 37), which is usually due to F. necrophorum, is an acute oro­
pharyngeal infection with secondary septic thrombophlebitis of the 
internal jugular vein and frequent septic emboli, most commonly to 
the lung. This infection typically begins with pharyngitis, which is fol­
lowed by local invasion in the lateral pharyngeal space, with resultant 
internal jugular vein thrombophlebitis.
Central Nervous System (CNS) Infections 
CNS infections 
associated with anaerobic bacteria are brain abscess, epidural abscess, 
and subdural empyema, in which anaerobes are recovered in up to 30, 
20, and 10% of cases, respectively. The frequency with which anaerobes 
are recovered depends in large part on the underlying reason for the 
infection. For example, brain abscesses are typically due to hematog­
enous seeding, contiguous spread, penetrating head trauma, or recent 
surgical intervention. Anaerobic bacteria are most commonly associ­
ated with brain abscesses resulting from contiguous spread (related 
to otogenic, odontogenic, and sinus infections), and the pathogens 
recovered are the same as in these antecedent infections. Facultative or 
microaerophilic streptococci and coliforms are often part of a mixed 
infecting flora in brain abscesses. The location of the abscess may 
also provide insight into the pathogens. Abscesses in the frontal lobe 
(often associated with sinusitis) are due to anaerobes, streptococci, and 
staphylococci; temporal lobe and cerebellar abscesses are often related 
to the oral microbiota and middle-ear pathogens.
Obligate anaerobes rarely cause meningitis. Only one obligate 
anaerobe was identified in a seminal study of 188 bacterial meningitis 
isolates, and a U.S. national surveillance study of 18,642 such isolates 
collected between 1977 and 1981 found only five obligate anaerobes. 
This low incidence may be due, in part, to the fact that many clinical 
microbiology laboratories do not routinely culture cerebrospinal fluid 
(CSF) for anaerobes.
Pleuropulmonary Infections 
The lungs are constantly seeded 
with organisms from the oral microbiota via subclinical microaspira­
tion that normally occurs in all people. Even though the lung is the site 
of oxygen exchange and is therefore an overwhelmingly aerobic envi­
ronment, the organisms most abundant in the lower respiratory tract 
(as assessed by culture-independent methods) include anaerobes such 
as Prevotella and Veillonella species, with oral microaerophilic strep­
tococcal species (e.g., the Streptococcus milleri group) also present in 
significant abundances. In patients who have impaired bacterial clear­
ance (due to decreased cough, dysfunctional mucociliary transport, or 
alcohol intoxication) and/or increased rates of aspiration (due to neu­
rologic disorders, impaired consciousness, or swallowing dysfunction), 
these anaerobic bacteria can establish an infection and result in aspira­
tion pneumonia, lung abscess, or empyema. These anaerobic infections 
have an indolent course that may serve as a clinical clue differentiating 
them from conditions with other etiologies (e.g., chemical pneumoni­
tis, pneumococcal pneumonia) that often present more acutely.
ASPIRATION PNEUMONIA  Bacterial aspiration pneumonia must be 
distinguished from two other clinical syndromes associated with 
aspiration that are not of bacterial etiology. One syndrome results 
from aspiration of food or, rarely, other foreign bodies. Obstruction 
of major airways typically results in difficulty breathing, atelectasis, 
and moderate nonspecific inflammation. Therapy consists of removal 
of the foreign body. The second aspiration syndrome relates to 
chemical pneumonitis caused by inhalation or aspiration of alveolar 
irritants. Perhaps the most common cause of chemical pneumonitis 
is Mendelson syndrome, which results from regurgitation and aspira­
tion of acidic gastric juices. Pulmonary inflammation—including the 

destruction of the alveolar lining, with transudation of fluid into the 
alveolar space—occurs with remarkable rapidity. This syndrome typi­
cally develops within 4–6 h, often following anesthesia when the gag 
reflex is depressed. The patient becomes tachypneic, tachycardic, and 
hypoxic, often in the absence of fever. The leukocyte count may rise, 
and the chest x-ray may evolve from normal to a complete bilateral 
“whiteout” within 8–24 h. Sputum production is minimal. The pulmo­
nary signs and symptoms often resolve quickly with symptom-based 
therapy, but this condition can culminate in respiratory failure due, in 
part, to pulmonary edema. Antibiotic therapy is not indicated unless 
bacterial superinfection occurs.

In contrast to these syndromes, bacterial aspiration pneumonia 
develops over a period of several days or weeks rather than hours. 
The pathogenesis includes some combination of an increased bacterial 
burden, increased virulence of the organisms aspirated, and potential 
airway damage related to aspiration of gastric fluid. Patients gener­
ally report fever, malaise, and sputum production. In some patients, 
weight loss and anemia reflect a more chronic process. Usually the 
history reveals factors predisposing to aspiration, such as significant 
alcohol consumption or neurologic impairment due to a previ­
ous stroke. Severe dental disease is often associated with aspiration 
pneumonia, but it is not clear whether this association relates to an 
increased number of oral microbes and/or the presence of organisms 
with increased virulence. Sputum characteristically is not malodorous 
unless the process has been ongoing for at least a week. Chest x-rays 
show consolidation in dependent pulmonary segments: in the basilar 
segments of the lower lobes if the patient has aspirated while upright 
and in either the posterior segment of the upper lobe (usually on the 
right side, given that the right mainstem bronchus has a more vertical 
orientation) or the superior segment of the lower lobe if the patient has 
aspirated while supine.
CHAPTER 182
A mixed bacterial population with many PMNs is evident on Gram’s 
staining of sputum. Expectorated sputum is unreliable for anaerobic 
cultures because of inevitable contamination by the normal oral micro­
biota. Reliable specimens for culture can be obtained by transtracheal 
or transthoracic aspiration—techniques that are rarely used at present. 
Although the culture of protected-brush specimens or bronchoalveolar 
lavage fluid obtained by bronchoscopy is controversial, more recent 
data suggest that these approaches can be used without oropharyngeal 
contamination and can recover anaerobic organisms from the lower 
respiratory tract in a site-directed manner. Further research is needed 
to determine how these approaches compare with the previous gold 
standards.
Infections Due to Mixed Anaerobic Organisms
ANAEROBIC LUNG ABSCESSES  (See also Chap. 132) These abscesses 
result from subacute anaerobic pulmonary infection. The clinical 
presentation typically involves a history of constitutional signs and 
symptoms (including malaise, weight loss, fever, night sweats, and 
foul-smelling sputum) that have typically persisted for 1–3 weeks 
prior to hospitalization. Patients who develop lung abscesses often, but 
not always, have an antecedent dental infection. Abscess cavities may 
be single or multiple and generally occur in dependent pulmonary 
segments (Fig. 182-3). The differential diagnosis for lung abscesses 
includes pneumonia (including necrotizing pneumonia), a purulent 
pleural effusion with a bronchopleural fistula, and a pneumatocele. Of 
note, infection with some aerobic organisms, particularly S. aureus, 
can develop into a lung abscess without an anaerobic component. 
Approximately 90% of cases have an anaerobe identified—usually 
three to six isolates per sample—if careful attention is paid to handling 
and processing of the abscess sample. The most common isolates 
include peptostreptococci, Prevotella and Porphyromonas species, 
and F. nucleatum. An important finding is that ~90% of cultures also 
demonstrate the presence of aerobic organisms, such as S. aureus, 
Streptococcus pneumoniae, and Klebsiella pneumoniae. Consistent with 
the notion that anaerobes are contributing to disease, patients often 
do not improve clinically until they receive an antibiotic regimen that 
includes anaerobic coverage.
EMPYEMA  Empyema is a manifestation of long-standing anaerobic 
pulmonary infection and manifests with thick, purulent material in

FIGURE 182-3  Chest radiograph (left) and CT image (right) of a lung abscess. The patient aspirated while supine and developed an abscess in the posterior segment of 
the right upper lobe. Cultures were pretreated and grew only Klebsiella pneumoniae. (Images provided by Gita N. Mody, MD, MPH, Division of Cardiothoracic Surgery, 
Department of Surgery, The University of North Carolina at Chapel Hill.)
PART 5
Infectious Diseases
the pleural space, often in association with a bronchopleural fistula. 
Alternatively, a subdiaphragmatic infection may extend into the pleural 
space and similarly result in an empyema. The clinical presentation 
resembles that of other anaerobic pulmonary infections and may 
include foul-smelling sputum, pleuritic chest pain, and marked chestwall tenderness. This disease process must be differentiated from a 
parapneumonic effusion resulting from more routine causes of pneu­
monia (e.g., S. pneumoniae). In the latter instance, the fluid is a thin exu­
date that has a mean white blood cell (WBC) count of ~5000 cells/mL, 

a lactate dehydrogenase level of >200 IU/L, and a pH of ~7.4. In 
contrast, empyema is characterized by foul-smelling thick pus with a 
mean WBC count of ~55,000 cells/mL, a lactate dehydrogenase level 
of >1000 IU/L, and a pH of <7.2 as well as loculations and a thick 
pleural peel on imaging. Drainage and occasionally decortication of the 
visceral and parietal pleura are required. Defervescence, a return to a 
feeling of well-being, and resolution of the process may require several 
months, particularly in the absence of surgical intervention.
Intraabdominal Infections 
Breach of the gut mucosal surface 
(e.g., due to trauma, intestinal perforation, or malignancy) allows 
members of the microbiota to enter the normally sterile perito­
neum. Accordingly, the offending organisms reflect the microbiota 
in the affected intestinal region. For example, recovery of Candida 
species from intraabdominal infections should prompt evaluation 
of the stomach and proximal small bowel for potential perforation. 
Furthermore, a study of patients with perforated and gangrenous 
appendicitis demonstrated that virtually all samples yielded E. coli 

and members of the B. fragilis group; peptostreptococci and 
Bilophila wadsworthia—additional components of the appendiceal 
and colonic microbiota—also were recovered from >50% of samples. 
Notably, some studies have identified an average of 10 different 
bacterial species, with an anaerobe-to-aerobe ratio of ~3:1. Given 
that >1000 bacterial species are present in the colonic microbiota, 
the dominance of such a limited repertoire of bacterial genera and 
species recovered in intraabdominal infections reflects a combina­
tion of two factors: the increased propensity of these organisms to 
result in intraabdominal abscesses and the difficulty faced by clinical 
microbiology laboratories in culturing the diverse organisms pres­
ent in these samples. See Chap. 137 for a complete discussion of 
intraabdominal infections.

Neutropenic enterocolitis (typhlitis) involves marked thickening 
of the bowel wall (typically >4 mm) in the setting of neutropenia, 
abdominal pain, and fever. This condition most commonly affects 
the cecum and may extend to the neighboring terminal ileum and/or 
proximal colon, but any intestinal region may be involved. Typhlitis 
generally occurs after 1–2 weeks of chemotherapy-induced neutrope­
nia associated with treatment of hematologic or, less commonly, solid 
tumor malignancies, but it can occur regardless of the cause of neutro­
penia. At least 5% of adults hospitalized for malignancy are thought to 
develop typhlitis, but this is likely an underestimate. Although the right 
lower quadrant is the most common location of abdominal pain and 
tenderness, these symptoms are absent in nearly half of cases; more­
over, some patients, particularly those taking glucocorticoids, may 
not experience abdominal pain at all. Given the weakened integrity of 
the bowel wall and the associated neutropenia, patients often develop 
bacteremia due to one or more organisms related to the microbiota of 
the affected intestinal segment. Patients who develop bacteremia due 
to Clostridium septicum often have relatively severe disease, and identi­
fication of this organism is highly associated with the presence of 
malignancy—notably, colon cancer. Medical management including 
bowel rest, intestinal decompression, and broad-spectrum antibiotic 
administration is generally successful, although surgical intervention may 
be required in cases of persistent intestinal bleeding, necrotic bowel, or 
clinical deterioration suggestive of an ongoing intestinal process.
Pelvic Infections 
Anaerobes are frequently encountered in pel­
vic inflammatory disease, pelvic abscess, endometritis, tubo-ovarian 
abscess, septic abortion, and postoperative or postpartum infections. 
These infections are often of mixed etiology, involving both anaerobes 
and coliforms; pure anaerobic infections without coliform or other 
facultative bacterial species occur more often in pelvic than in intraab­
dominal sites. The major anaerobic pathogens in pelvic abscesses are 
P. bivia, P. disiens, and the B. fragilis group, but many other anaerobes 
also have been implicated. See Chap. 141 for a complete discussion of 
pelvic inflammatory disease.
Anaerobic bacteria have been thought to be contributing factors in 
bacterial vaginosis. This syndrome of unknown etiology is character­
ized by a profuse malodorous discharge and a change in bacterial 
ecology that results in replacement of the Lactobacillus-dominated 
normal microbiota with an overgrowth of anaerobic bacterial species.

Culture-based and culture-independent approaches have identified 
numerous organisms, including Gardnerella vaginalis, peptostrepto­
cocci, genital mycoplasmas, and species within the genera Prevotella, 
Mobiluncus, Atopobium, Leptotrichia, Megasphaera, and Eggerthella. 
This wide array of implicated bacteria may reflect differences in the 
overall disease spectrum of bacterial vaginosis and/or a shared physi­
ologic response to these different organisms.
Skin and Soft Tissue Infections 
Similar to other anatomic sites, 
skin or soft tissue injured by trauma, ischemia, or surgery creates a 
suitable environment for anaerobic infections. The infecting bacteria 
either are introduced directly (e.g., wounds associated with intestinal 
surgery, decubitus ulcers, or human bites) or originate in contiguous 
areas (e.g., cutaneous abscesses, rectal abscesses, and axillary sweat 
gland infections [hidradenitis suppurativa]). Anaerobes also are often 
cultured from foot ulcers of diabetic patients. The most common loca­
tions for anaerobic cellulitis include the neck, trunk, groin (including 
the genitalia), and legs. The deep soft tissue infections associated with 
anaerobic bacteria are gas gangrene, synergistic cellulitis (both progres­
sive and necrotizing), necrotizing fasciitis, and myositis (Chaps. 134 
and 159).
Gas gangrene (crepitus cellulitis) is most often due to C. perfringens, 
although other clostridial species have been implicated as well. This 
infection involves extensive gas formation in the tissue leading to 
crepitus and a thin, dark, occasionally malodorous discharge. True gas 
gangrene typically presents with fever and tenderness around the lesion 
and can rapidly spread; in contrast, there are somewhat more indolent 
forms of anaerobic cellulitis that may involve some gas formation but 
often present without fever or extensive local pain and can spread over 
the course of days rather than minutes.
Progressive bacterial synergistic gangrene (Meleney gangrene) is 
characterized by an area of exquisite pain, redness, and swelling fol­
lowed by ulceration. As the ulcer enlarges, it is surrounded by a viola­
ceous ring that fades into a pink edematous border. If it is not promptly 
treated, the ulcer continues to enlarge, and new, distant ulcers may 
emerge. Symptoms are limited to pain; fever is not typical. Peptostrep­
tococci and microaerophilic streptococci are commonly found in the 
leading edge of the lesions, and S. aureus and Proteus species can be 
isolated from the ulcerated lesion. Treatment includes surgical removal 
of necrotic tissue and antimicrobial administration. In contrast, syn­
ergistic necrotizing cellulitis involves the deep fascia and occurs near 
the point of bacterial entry. Pain, fever, and systemic symptoms are 
common. If this form of cellulitis involves the scrotum, perineum, and 
anterior abdominal wall, it is referred to as Fournier gangrene. S. aureus, 
the B. fragilis group, Peptostreptococcus species, Clostridium species, 
Fusobacterium species, and members of the family Enterobacteriaceae 
are the predominant organisms identified.
Necrotizing fasciitis, a rapidly spreading destructive disease of the 
fascia, is usually attributed to group A streptococci (Chap. 153) but can 
also be a mixed infection involving anaerobes and aerobes. Polymicro­
bial necrotizing fasciitis differs from stereotypical group A streptococ­
cal necrotizing fasciitis in that the initial erythematous, swollen, tender 
lesions progress over 3–5 days (as opposed to 1–3 days), with conse­
quent skin breakdown and cutaneous gangrene. Fever, subcutaneous 
gas, development of anesthesia (often before skin necrosis), and a foulsmelling discharge are common. The particular clinical findings some­
times suggest the causative agent: regional lymphadenopathy suggests 
the B. fragilis group; necrosis and gangrene suggest Clostridium species, 
peptostreptococci, the B. fragilis group, and Enterobacteriaceae; bul­
lous lesions suggest Enterobacteriaceae; a foul-smelling odor suggests 
Bacteroides and Clostridium species; and subcutaneous gas suggests 
peptostreptococci, Clostridium species, and the B. fragilis group. More­
over, diabetic infections are often associated with Bacteroides species, 
Enterobacteriaceae, and S. aureus, and infections related to trauma are 
associated with Clostridium species.
Although S. aureus is the typical cause of myositis, anaerobes—
particularly C. perfringens—are often recovered from patients with 
pyogenic myositis. In anaerobic streptococcal myonecrosis, pepto­
streptococci are often identified along with group A streptococci or 

S. aureus. Patients typically present with fever, muscle pain, fatigue, and 
an elevated creatinine kinase level suggestive of muscle inflammation.

Bone and Joint Infections 
A comprehensive review of the world 
literature on anaerobic bone infections through 1975 included >650 
cases. Of these, ~400 cases were caused by Actinomyces species; anaero­
bic cocci and Bacteroides, Fusobacterium, and Clostridium species were 
most commonly identified in the remaining cases. Actinomycotic 
involvement of the jaw was the most common bone infection, with 
the mandible involved four times as frequently as the maxilla. Patients 
with cervicofacial actinomycosis (Chap. 180) are often described as 
having a “lumpy jaw” because of the prominent soft tissue swelling 
that is sometimes mistaken for malignancy or granulomatous disease. 
These infections can be chronic in nature, can include the development 
of sinus tracts, can progress across normal tissue boundaries, and can 
require prolonged antibiotic treatment to prevent relapse. The verte­
brae are the second most common location for Actinomyces infection; 
involvement of the thorax, abdomen, or pelvis is much less frequent.
Osteomyelitis involving anaerobes other than Actinomyces species 
most commonly develops by extension of an adjacent infection (e.g., 
soft tissue, paranasal sinus, or middle-ear infection). For example, 
diabetic foot ulcers and decubitus ulcers may be complicated by mixed 
aerobic–anaerobic osteomyelitis (Chap. 136). Hematogenous seeding 
of bone by anaerobes is uncommon and is thought to occur in fewer 
than 10% of cases. The most common sites of anaerobic osteomyelitis 
are the head (skull and jaw) and the extremities. Fusobacteria have 
been isolated in pure culture from infections of the mastoid, mandible, 
and maxilla. Clostridium species have been reported as anaerobic 
pathogens in cases of osteomyelitis of the long bones following trauma. 
Anaerobic and microaerophilic cocci are most frequently isolated from 
infections involving the skull or mastoid; usually, these organisms are 
present in mixed cultures.
CHAPTER 182
In contrast to anaerobic osteomyelitis, anaerobic arthritis (Chap. 135) 
is uncommon, typically involving a single isolate, and most cases are 
secondary to hematogenous spread. Although Fusobacterium spe­
cies accounted for nearly one-third of cases in the preantibiotic era, 
C. acnes, peptostreptococci, and B. fragilis are now among the more 
frequent causes of anaerobic septic arthritis. Peptostreptococci and 

C. acnes are often found in association with prosthetic joints, Fusobac­
terium species have a predilection for the sternoclavicular and sacro­
iliac joints, and clostridial arthritis is especially common after trauma. 
As a frequent cause of bacteremia, B. fragilis is a common cause of 
anaerobic septic arthritis; however, arthritis occurs in fewer than 5% of 
patients with B. fragilis bacteremia.
Infections Due to Mixed Anaerobic Organisms
Bacteremia 
B. fragilis is the anaerobe most commonly isolated from 
blood cultures. Although the frequency of positive cultures appeared to 
be decreasing in the 1980s, more recent evidence suggests that the rate 
is now increasing and that the increase may be related to changing 
demographics, with more patients who are elderly, immunocompro­
mised, and/or receiving medications that may disrupt the mucosal 
barrier (e.g., chemotherapy). The source of bacteremia is most often an 
abscess in the abdomen, female genital tract, or soft tissue. At a large 
tertiary-care U.S. hospital, 0.8% of all positive blood cultures yielded 
an anaerobic gram-negative bacillus, with 0.5% yielding B. fragilis. A 
similar study in France revealed that 0.6% of all positive blood cultures 
yielded an anaerobic organism; 60% of these isolates were Bacteroides 
species, and 22% were Clostridium species. Peptostreptococcus and 
Fusobacterium species are also recovered with significant frequency.
Once the organism in the blood has been identified, both the portal 
of bloodstream entry and the underlying problem that probably led to 
seeding of the bloodstream can often be deduced from an understand­
ing of the organism’s normal site of residence. For example, mixed 
anaerobic bacteremia including B. fragilis usually implies a colonic 
pathology, with mucosal disruption from neoplasia, diverticulitis, 
or some other inflammatory lesion. The initial manifestations are 
determined by the portal of entry and reflect the localized condition. 
Although the clinical manifestations of B. fragilis bacteremia (e.g., 
rigors, hectic fevers) are similar to those of aerobic gram-negative

bacillary bacteremia, the incidence of septic shock is lower with B. fra­
gilis. This difference may be due to differences in the immunostimula­
tory effects of the different endotoxin structures.

In virtually all cases, isolation of a member of the B. fragilis group 
from blood indicates underlying infection that is associated with a 
mortality rate of 60% if untreated. It has been suggested that the mor­
tality rate depends in part on the species recovered (B. thetaiotaomicron 
> P. distasonis > B. fragilis), but it is unclear whether differences in 
mortality rates relate to intrinsic differences in the virulence of these 
organisms, in their antimicrobial susceptibility profiles, and/or in the 
host’s immune response. Case–fatality rates appear to increase with 
the increasing age of the patient (with reported rates of >66% among 
patients >60 years old), with the isolation of multiple species from 
the bloodstream, and with the failure to surgically remove a focus of 
infection.
Endocarditis 
(See also Chap. 133) Although gram-negative anaer­
obic bacteria only rarely cause endocarditis, their involvement is asso­
ciated with significant mortality rates (21–43%). Members of the B. 
fragilis group are the most commonly identified gram-negative anaer­
obes in endocarditis. Anaerobic streptococci, which are often classified 
incorrectly, are likely responsible for this disease more frequently than 
is generally appreciated. Compared with aerobic bacterial endocarditis, 
endocarditis due to Bacteroides species is less likely to be associated 
with a history of cardiovascular disease and more likely to involve 
thromboembolic complications.
■
■DIAGNOSIS
There are three critical steps in the diagnosis of anaerobic infection: (1) 
proper collection of specimens; (2) rapid transport of the specimens to 
the microbiology laboratory, preferably in anaerobic transport media; 
and (3) proper handling of the specimens by the laboratory. Specimens 
must be collected by meticulous sampling of infected sites, with avoid­
ance of contamination by the normal microbiota. Samples from sites 
known to harbor numerous anaerobes (e.g., the mouth, nose, vagina, 
feces) are not acceptable for anaerobic culture as the presence of the 
normal microbiota will complicate interpretation of the results in a 
clinically meaningful manner. In contrast, samples from normally 
sterile locations (e.g., blood, pleural fluid, peritoneal fluid, CSF, and 
aspirates or biopsy samples from normally sterile sites) are appropriate 
for anaerobic culture in clinical microbiology laboratories. As a general 
rule, liquid or tissue specimens are preferred; if swab specimens must 
be used, special anaerobic swab systems should be used to help main­
tain persistence of anaerobes. Liquid samples should be collected in an 
air-free syringe that is then capped, injected into anaerobic transport 
bottles, or quickly transported to the clinical microbiology laboratory 
for immediate culture.
PART 5
Infectious Diseases
Because of the time and difficulty involved in the isolation of anaer­
obic bacteria, the diagnosis of anaerobic infections must frequently be 
based on presumptive evidence. As mentioned previously, anaerobic 
infections are sometimes suggested by specific clinical factors, such as 
origins from a site with an anaerobic-rich microbiota (e.g., the intesti­
nal tract, oropharynx), the presence of an abscess, involvement of sites 
with lowered oxidation-reduction potential (e.g., avascular necrotic 
tissues), a foul odor, and the presence of gas in tissues. None of these 
features is necessarily pathognomonic or required for the diagnosis 
of an anaerobic infection, but these are helpful clues to keep in mind 
when constructing a differential diagnosis.
When cultures of obviously infected sites or purulent material 
yield no growth, streptococci only, or a single aerobic species (such 
as E. coli) and Gram’s staining reveals a mixed bacterial population, 
the involvement of anaerobes should be suspected; the implication 
is that the anaerobic microorganisms have failed to grow because of 
inadequate transport and/or culture techniques. It is also important 
to remember that prior antibiotic therapy reduces the cultivability of 
these bacteria. Failure of an infection to respond to antibiotics that 
are not active against anaerobes (e.g., aminoglycosides and—in some 
circumstances—penicillin, cephalosporins, or tetracyclines) suggests 
an anaerobic etiology.

TREATMENT
Anaerobic Infections
Similar to successful therapy for other types of infection, treatment 
for anaerobic infections requires the administration of appropri­
ate antibiotics, surgical debridement of devitalized tissues, and 
drainage of any large abscess. Any mucosal breach must be closed 
promptly to prevent ongoing infection. 
ANTIBIOTIC THERAPY AND RESISTANCE
The antibiotics used to treat anaerobic infections should be active 
against both aerobic and anaerobic organisms because many of 
these infections are of mixed etiology. Antibiotic regimens can usu­
ally be selected empirically on the basis of the location of infection 
(which provides insight into the likely species involved), the sever­
ity of infection, and knowledge of local antimicrobial resistance pat­
terns. Other factors influencing the selection of antibiotics include 
need for penetration into certain organs (such as the brain) and 
associated toxicity (Chap. 149). As with all infections, the general 
maxim is to use the the narrowest-spectrum agent(s) possible so as 
to minimize the impact on the normal microbiota and the develop­
ment of resistance.
Because of the slow growth rate of many anaerobes, the lack of 
standardized testing methods and of clinically relevant standards for 
resistance, and the generally good results obtained with empirical 
therapy, the role of antibiotic susceptibility testing of these organ­
isms has been limited in most clinical microbiology laboratories. 
Instead, isolates are sent to reference laboratories for susceptibility 
testing when an infection is serious (e.g., brain abscess, meningitis, 
joint infection), is refractory, or requires prolonged therapy (e.g., 
osteomyelitis, prosthetic joint infection, endocarditis). Such testing 
should also be considered when a patient is not responding to anti­
microbial therapy as expected; multidrug-resistant anaerobes have 
been reported. Antimicrobial susceptibility testing is also helpful in 
monitoring the activity of new drugs and recording current resis­
tance patterns among anaerobic pathogens.
The need for susceptibility testing of anaerobic organisms is 
highlighted by increasing rates of antimicrobial resistance, geo­
graphic and institutional differences in susceptibility profiles, spe­
cies-specific antibiograms, and the potential for worse clinical 
outcomes when ineffective antibiotics are used. These differences 
preclude making any sweeping generalizations regarding antibiotic 
therapy for anaerobic infections. For example, rates of resistance to 
piperacillin-tazobactam have remained low (≤1%) for all Bacteroi­
des species in the United States, but B. thetaiotaomicron isolates in 
Korea have a notably higher resistance rate (17%). Clindamycin was 
historically effective against members of the B. fragilis group, but 
rates of resistance have increased to 30–43% in the United States 
and are >80% in some parts of the world. Furthermore, metronida­
zole is effective against many different anaerobic organisms and is 
considered a first-line agent for many anaerobic infections world­
wide, but, in a population of Colombian patients with refractory 
periodontitis, 45% of Fusobacterium isolates and 25% of Prevotella 
and Porphyromonas strains were resistant to metronidazole; this 
finding underscores the importance of understanding the local 
antibiogram and of assessing susceptibility profiles in refractory 
disease. 
Empirical Therapy  Not every anaerobe isolated must be specifically 
targeted by the antibiotic regimen. Given that infections involving 
anaerobes are typically polymicrobial, that the cultivation and iden­
tification of anaerobes are challenging (i.e., not all organisms may 
be recovered), and that organisms often depend on one another for 
persistence, clinical resolution of the infection is often achieved with 
empirical antibiotics targeting the bulk of the organisms recovered. 
Antibiotics that demonstrate no useful activity against anaerobes 
include aminoglycosides, monobactams, and trimethoprim-sulfa­
methoxazole. With the caveat that susceptibility profiles may change 
with time and geography, the antibiotics that are commonly used

TABLE 182-2  Antimicrobial Therapy That Is Typically Active against 
Commonly Encountered Anaerobes
ANTIBIOTIC(S)
CAVEATS
Metronidazole
This drug is clinically unreliable against 

gram-positive non-spore-forming anaerobes 
(e.g., Actinomyces spp., Propionibacterium spp., 
Peptostreptococcus spp.).
Rates of resistance are increasing in some 

gram-negative anaerobes. The newer 
cephalosporin/β-lactamase combinations have 
limited anaerobic activity.
β-Lactam/β-lactamase 
inhibitor combinations 
(ampicillin-sulbactam, 
ticarcillin–clavulanic acid, 
piperacillin-tazobactam)
Clindamycin
Rates of resistance are increasing in Bacteroides 
spp.
Carbapenems 
(meropenem, imipenem, 
ertapenem, doripenem)
Rates of resistance are currently very low (<5%), 
although some carbapenemase-producing strains 
have been identified.
Chloramphenicol
Some clinical failures have been noted, even 

when the isolate is found to be susceptible by in 
vitro testing.
as empirical therapy against anaerobic bacteria include metroni­
dazole, β-lactam/β-lactamase inhibitor combinations, clindamycin, 
carbapenems, and chloramphenicol (Table 182-2).
Metronidazole is active against gram-negative anaerobes, includ­
ing nearly all isolates of Bacteroides species, and gram-positive 
spore-forming organisms, such as C. difficile (Chap. 139) and 
other Clostridium species. Given intrinsically reduced susceptibility, 
metronidazole is clinically unreliable against gram-positive non- 
spore-forming organisms, such as Actinomyces, Propionibacterium, 
Lactobacillus, Bifidobacterium, Eubacterium, and Peptostreptococ­
cus. Of note, a few metronidazole-resistant Bacteroides isolates have 
been identified in the United States, and rates of such resistance 
have been increasing in Europe. Moreover, the rate of resistance 
to metronidazole has probably been greatly underestimated in 
some countries (e.g., the United Kingdom) that use metronida­
zole susceptibility to discriminate between obligate and facultative 
anaerobes (with obligate anaerobes defined by their susceptibility). 
Although the majority of metronidazole-resistant isolates have been 
identified in patients who have been exposed to the drug, resistant 
organisms have also been found in metronidazole-naïve patients.
More than 90% of clinical isolates from the B. fragilis group 
produce β-lactamases that are predominantly active against cepha­
losporins and that are highly active, cell associated, and produced 
constitutively. Thus, members of the B. fragilis group are presumed 
to be resistant to penicillin and ampicillin but may remain suscep­
tible to extended-spectrum penicillins, particularly in combination 
with a β-lactamase inhibitor (e.g., ampicillin-sulbactam, piperacillin-

tazobactam). Rates of resistance to ampicillin-sulbactam are increas­
ing, particularly in P. distasonis, which has a reported resistance rate 
of 21% in the United States. Because β-lactamase production is 
not common in Clostridium species, these combination agents are 
usually effective. Of note, the newer cephalosporin/β-lactamase 
inhibitors (e.g., ceftolozane-tazobactam, ceftazidime-avibactam) 
have limited anaerobic activity.
Clindamycin is active against many anaerobes. However, rates of 
resistance to clindamycin among Bacteroides species increased in 
the United States from 7% in 1981 to 33% in 2010–2012. Resistance 
to clindamycin among non-Bacteroides gram-negative anaerobes is 
much less common (<10%). Some Clostridium species are resistant 
to clindamycin, although C. perfringens typically is not.
Carbapenems (ertapenem, doripenem, meropenem, and imipe­
nem) are active against anaerobes, with fewer than 3% of Bacteroides 
isolates resistant. There is little difference among resistance rates for 
specific species, and, of the carbapenems, imipenem typically has 
the lowest resistance rate. Although the β-lactamase produced by 
most Bacteroides species is unable to inactivate carbapenems, rare 
B. fragilis strains have been reported to produce a carbapenemase.

Resistance to chloramphenicol is rare in Bacteroides species. 
Nationwide surveys in the United States have identified no resistant 
organisms, but some isolates with elevated minimal inhibitory con­
centrations (MICs)—i.e., 16 μg/mL—have been noted. Although 
chloramphenicol has excellent in vitro activity against all clinically 
relevant anaerobes, some clinical failures have been documented. 
Therefore, this drug may be less preferable if other active agents 
are available.

Other antibiotics with more variable activity against anaerobes 
include the fluoroquinolones and tigecycline. Although many 
fluoroquinolones (e.g., ciprofloxacin, levofloxacin, ofloxacin) display 
reasonable activity against anaerobic organisms other than Bacteroides 
species, these agents exhibit poor activity against the B. fragilis group. 
Rates of resistance to moxifloxacin are relatively high (39–83%) 
among Bacteroides isolates obtained in the United States but are much 
lower among B. fragilis and B. thetaiotaomicron isolates collected in 
Korea (8 and 2%, respectively) or Taiwan (8 and 15%, respectively). 
Tigecycline is active against most anaerobic bacteria, although MICs 
are somewhat higher for Clostridium species. Tigecycline’s efficacy for 
treatment of complicated intraabdominal infections is comparable to 
that of imipenem, and it is therefore recommended as single-agent 
therapy for these infections. 
Infections at Specific Sites  In clinical situations, specific antibi­
otic regimens and durations must be tailored to the initial site of 
infection; the reader is referred to specific chapters on infections at 
specific sites for recommendations. In general, anaerobic infections 
are often broadly categorized as originating above or below the dia­
phragm. This distinction is clinically useful in that the predominant 
pathogens—and therefore the empirical antibiotic regimens—differ 
between these two categories of infection.
CHAPTER 182
Infections above the diaphragm usually reflect the orodental 
microbiota, which includes Prevotella, Porphyromonas, Fusobacte­
rium, and Bacteroides species other than the B. fragilis group along 
with streptococci (both aerobic and microaerophilic). Accordingly, 
antibiotic regimens should cover both aerobic and anaerobic bacteria. 
Given that >70% of these infections include a β-lactamase-producing 
organism, β-lactam drugs (penicillins and cephalosporins) are poor 
options as monotherapy. The recommended regimens include 
clindamycin, a β-lactam/β-lactamase inhibitor combination, or met­
ronidazole in combination with a drug active against microaerophilic 
and aerobic streptococci (e.g., penicillin).
Infections Due to Mixed Anaerobic Organisms
Anaerobic infections arising below the diaphragm (e.g., colonic 
and intraabdominal infections) must be treated specifically with 
agents active against Bacteroides species, including B. fragilis. 
Single agents suitable for this purpose include cefoxitin, moxi­
floxacin, a β-lactam/β-lactamase inhibitor combination, or a 
carbapenem. A two-drug regimen is an alternative, with one drug 
active against anaerobes and the other against coliforms (e.g., 
metronidazole with either a cephalosporin or a fluoroquinolone). 
In addition, if the clinician suspects that gram-positive facultative 
organisms such as enterococci are involved, therapeutic regimens 
should include ampicillin or vancomycin. Although clindamycin 
and cefotetan were previously considered acceptable options for 
intraabdominal infections involving anaerobes, these drugs are 
no longer recommended because of escalating rates of resistance 
in the B. fragilis group. Ampicillin-sulbactam is not recommended 
because of high rates of resistance among community-acquired 
strains of E. coli rather than because of resistance in anaerobic 
bacteria.
CNS infections involving anaerobic organisms may be treated 
with metronidazole, a carbapenem, chloramphenicol, or—if only 
gram-positive anaerobes are involved—penicillin. Clindamycin and 
cefoxitin have poor penetration into the CSF and should not be 
used. Cases of osteomyelitis in which a polymicrobial infection is 
identified from a bone biopsy specimen should be treated with a 
regimen that covers both aerobes and anaerobes, as some organisms 
that are often regarded as a contaminant (e.g., C. acnes) may have 
a pathogenic role. When an anaerobic organism is recognized as a