Oral Maxillofacial Surg Clin N Am 15 (2003) xi – xii
Preface
Current concepts in the management of
maxillofacial infections
Daniel M. Laskin, DDS, MS
Guest Editors
There have been many advances in the manage-
ment of head and neck infections since we last edited a
Clinics of North America issue on this subject more
than 10 years ago. New classes of antimicrobial
agents, noninvasive imaging techniques, improved
culturing methods, and a clearer understanding of
the normal and pathologic functioning of the immune
system are just some of the changes that have
occurred. These and other technologic advances have
enhanced dramatically our ability to diagnose and
treat these infections rapidly and accurately. As a
result, the incidence of serious morbidity and mortal-
ity from odontogenic infections has fallen signifi-
cantly over the years.
It should be noted, however, that vigilance regard-
ing research and technologic change must be main-
tained. Just as we have made many advances in the
management of these infections, an equal number of
new and serious issues have arisen to test our resolve.
Newly recognized bacterial and viral strains, the
effects of global antibiotic resistance to once univer-
sally effective agents, viral mutations, ‘‘flesh eating’’
bacteria causing necrotizing fasciitis, currently un-
treatable catastrophic infections, such as those caused
by the Ebola virus, and even biologic warfare and
bioterrorism have emerged as important issues over
the last 10 years, and articles about these issues flood
1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 8 0 - 8
Robert A. Strauss, DDS, MD
the public news media and professional journals
almost daily. In some cases we even have created
new sources of infection ourselves by developing
environments conducive to bacterial growth, such as
around implants.
In this issue of the Oral and Maxillofacial Surgery
Clinics of North America, we have attempted to
address some of the important and topical areas of
knowledge in the diagnosis and management of
infectious diseases that practicing oral and maxillofa-
cial surgeons face. Some articles examine new areas
of interest, such as peri-implantitis and infections
associated with facial skin resurfacing, whereas others
provide timely updates of our understanding of con-
tinually evolving topics, such as HIV/AIDS, chronic
sclerosing osteomyelitis, selection and use of anti-
biotics, and the changing microbiology of infections
of the head and neck.
The impressive ability of bacteria and viruses to
adapt, change, and mutate in response to our phar-
macologic bombardment is a testimony to the com-
plex, surreptitious, and unpredictable nature of these
small yet hardy microbes. For every new drug we
formulate, resistance develops to an older and often-
used one. For every organism that we eradicate,
another one suddenly emerges to take its place. To
those of us in the clinical trenches, it seems that we
xii D.M. Laskin, R.A. Strauss / Oral Maxillofacial Surg Clin N Am 15 (2003) xi–xii

are in a war with an ever-expanding number of Daniel M. Laskin, DDS, MS*

increasingly virulent and destructive bacteria, fungi, Robert A. Strauss, DDS, MD
and viruses. At times it seems that despite our Department of Oral and Maxillofacial Surgery
advances in technology and knowledge, the out- School of Dentistry
come of that war is not as clear as we might wish. Virginia Commonwealth University
It is our hope that the information provided by Richmond, VA 23298-0566, USA
the outstanding contributors to this issue will help E-mail address: dmlaskin@vcu.edu
to resolve some of the important issues we cur-
rently face. * Corresponding author.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 1 – 15
The changing microbiology of maxillofacial infections
Richard H. Haug, DDS
Division of Oral and Maxillofacial Surgery, University of Kentucky College of Dentistry, D-509 Chandler Medical Center,
Lexington, KY 40536-0084, USA
Dramatic headlines in the recent news media have
provided such alarming infectious disease stories as
‘‘Five year-old girl battles flesh eating bacteria,’’ ‘‘The
killers all around: new viruses and drug resistant
bacteria are reversing human victories over infectious
disease,’’ and ‘‘Losing the battle of the bugs’’ [1 – 3].
These reports conjure up images of an almost science
fiction type of horror story. Concerns among the lay
population regarding the perceived shifting of infec-
tious disease patterns and their subsequent manifes-
tations have become a topic of discussion in the news
media, often creating public alarm through misinter-
pretation and occasional exaggeration of events. For
example, Time magazine has suggested that antibiotics
are so overused that the human body has become
saturated and that the human immune system is so
depressed that it provides an environment for the
creation of ‘‘bacterial monsters’’ [2]. The reports of
‘‘flesh-eating’’ Streptococcus strains and ‘‘killer’’ bac-
teria that are resistant to most common antibiotics raise
public concerns that the normal bacterial flora is
mutating uncontrollably and that infections are much
more severe than they used to be [1 – 3]. Is this actually
the case, and if so, do the same principles hold true for
oral and maxillofacial infections? The purpose of this
article is to review the past and present oral and
maxillofacial literature, with a focus on identifying
Portions of this article have been reprinted from Storoe
W, Haug RH, Lillich TT. The changing face of odontogenic
infections. J Oral Maxillofac Surg 2001;59:739 – 48; and
Haug RH, Assael LA. Infection in the maxillofacial trauma
patient. In: Hupp J, Topazian RG, Goldberg MH, editors.
Management of infections of the oral and maxillofacial
regions. 5th edition. Philadelphia: WB Saunders Co.; 2002;
p. 267 – 92.
1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 7 6 - 6
any changes that may have occurred over the past
decades in infections of the maxillofacial region. It
attempts to answer whether there actually have been
any changes in the microbiology of maxillofacial
infections and, if so, what those changes, patterns,
and trends might be.
The availability of the ‘‘wonder drug’’ penicillin in
the years immediately after World War II ushered in an
era of complacency in infectious disease management.
Many traditional management techniques, such as
isolation, quarantine, and scrupulous application of
aseptic methods in the office and operatory, were
minimized or discarded altogether as no longer being
necessary because the new drugs were so effective in
treating common infections. Antibiotics were pre-
scribed when symptoms initially appeared, without
determining either the cause of the disease or antibiotic
susceptibility of the microbe. Many infectious disease
and public health professionals were ready to declare
ultimate victory over infection, being convinced that
antibiotics and vaccines would solve any existing and
future disease threats. Events of the last five decades
suggest that such enthusiasm was premature.
Although there have been many dramatic improve-
ments in the rates of morbidity and mortality associ-
ated with infectious diseases the past few decades,
microorganisms have proved to be adaptable and have
displayed the alarming ability to reemerge in continu-
ing cycles of disease in somewhat different forms [4].
In the journal Science, Harold Neu suggested that
‘‘bacteria are more clever than men’’ [5]. They have
adapted to every environmental niche on the planet
and are assimilating to surroundings saturated with
antibiotics [5]. This ability has been demonstrated over
the past decade in an array of infections not previously
recognized in humans. The reemergence of diseases in
different forms caused by common microorganisms,
2 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15
such as Staphylococci, Streptococci, Escherichia coli,
and Mycobacterium tuberculosis, further substantiates
this observation. The reemergence is caused partly by
acquisition of antibiotic resistance mechanisms, either
through mutation or by transfer of genetic information
from other bacteria; thus, there has been a rise in the
antibiotic resistance of important pathogenic genera
[4 – 6].
The most dramatic example in the development of
resistance over the past three decades and its effects
in terms of cost—human and economic—has been
the evolution of methicillin-resistant Staphylococcus.
It is estimated that methicillin resistance in certain
strains of Staphylococcus has risen from 2.4% in
1975 to 29% in 1999 and that the costs of hospi-
talization for these types of nosocomial infections
could approach $3 billion annually in the United
States alone [6 – 16]. One must remember, however,
that the ultimate human cost is death and that post-
surgical patients who develop nosocomial infections
from methicillin-resistant Staphylococcus are twice as
likely to die as patients who do not [13,15]. One must
ask the question: Do oral and maxillofacial infections
follow the same principles of reemergence and res-
istance as those described in other areas? If so, are the
patterns and trends similar? This article attempts to
answer those questions using the following format.
First, the microorganisms of the maxillofacial region
are identified in terms of indigenous flora. If this
process is not performed first, abnormalities and
outliers could not be identified. Next, oral and max-
illofacial infections are identified in terms of odonto-
genic infections, trauma-related infections, nasal and
paranasal sinus infections, and cutaneous infections.
Finally, changes or similarities in the microbiology of
maxillofacial infections over time are discussed in the
context of the previously named categories.
Normal oral and maxillofacial microflora
Factors that affect microbial flora
Most areas of the human body harbor an indigen-
ous microbial flora that is specific to the anatomic
area and the individual person. [17 – 21]. The specific
ecosystem helps to play a role in protecting the
individual from invasion by pathogenic organisms.
Numerous factors may alter the type, frequency, and
distribution of these microbes, including the anatomic
region, the individual host’s immune system, the
relative humidity, the presence or lack of oxygen,
local surface characteristics, available nutrition, and
any interaction between the microbes present
[17,19,20]. Although there is generally a normal
resident flora for any particular anatomic region, a
transient flora also may be found to colonize for
periods ranging from hours to weeks without being
retained permanently. This discussion is limited to the
indigenous flora (not transient), which, in the interest
of brevity, have been listed in Boxes 1 – 3.
Cutaneous microflora
Skin is relatively dry, has a mildly acidic pH, and
desquamates. This is an inhospitable environment for
microbes [17,21]. In the newborn, the cutaneous flora
is initially that of the mother’s birth canal. With time,
the skin develops its own unique flora (see Box 1)
and can include one animal form, five types of yeast,
and multiple species of bacteria [17,19,22 – 24]. This
flora is generally present on the hair follicles and
stratum corneum. The types and number of micro-
organisms are relatively constant, with minimal
change observed in the healthy individual [17,21].
Oral microflora
The mouth, unlike the skin, is warm, has a
relatively constant temperature, is wet and dark,
possesses environments with and without oxygen,
Box 1. Some of the most common
indigenous and colonizing maxillofacial
cutaneous flora [17,19,22 – 24]
Aerobic and facultative isolates
Acinetobacter species
Enterobacteriaceae
Corynebacterium species
Micrococcus species
Staphylococcus epidermidis
Staphylococcus species
Streptococcus viridans
Yeasts
Candida albicans
Malassezia furfur
Pityrosporum ovale
Pityrosporum orbiculare
Torulopsis glabrata
Animal
Demodex folliculorum
 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15 3

Penicillium species
Box 2. Some of the most common
Scopulariopsis species
indigenous and colonizing oral flora
Viruses
[17 – 19,22,24]
Herpes simplex virus
Aerobic and facultative isolates Cytomegalovirus
Protozoa
Diphtheroids
Lactobacilli Entamoeba gingivalis
Rothia dentocariosa Trichomonas tenax
Streptococcus faecalis
Streptococcus milleri
Streptococcus mitis and provides a constant source of nutrition [17,19 –
Streptococcus mutans 21]. The newborn’s mouth is not sterile but, like the
Streptococcus salivarius skin, contains a mixture of organisms representative
Streptococcus sanguis of the mother’s vagina, including Streptococci, Enter-
Anaerobic isolates ococci, and Microaerophilic species [17 – 21]. The
Actinobacillus newborn mouth is generally populated with aerobic
actinomycetemcomitans organisms because there are initially no crevices for
Actinomyces israelii the development of an anaerobic environment. That
Actinomyces naeslundii status changes with the eruption of teeth. In adult-
Actinomyces odontolyticus hood, the flora become more constant in composition
Actinomyces viscosus (see Box 2) and contain mostly streptococcal bac-
Bacteroides asaccharolyticus terial species (30% – 60%), other bacteria, yeasts,
Bacteroides melaninogenicus fungi, protozoa spirochetes, and viruses. Local dis-
Bacteroides oralis eases, such as caries and periodontitis, can alter the
Bacteroides ochraceus microflora [17 – 20,22,24]. Environmental agents,
Bacterionema matruchotii such as alcohol, tobacco smoke, medications, and
Branhamella catarrhalis radiation therapy, also change the microflora.
Capnocytophaga gingivalis
Capnocytophaga ochraceus
Nasal and paranasal sinus microflora
Capnocytophaga sputigena
Eikenella corrodens
Unlike the skin and mouth, the nasal passages and
Fusobacterium plauti
paranasal sinuses are sterile at birth. Within days,
Fusobacterium nucleatum
however, the infant acquires flora that are mostly
Neisseria sicca
associated with the mother, nursing staff, and hospital
Peptostreptococcus
[24 – 28]. As a human breathes, air and its contents
Veillonella alcalescens
pass through the nose. There the air is filtered, and
Veillonella parvula
most of the microorganisms are trapped by mucous
Spirochetes
secretions and swallowed. These secretions contain
Spirochaeta species
enzymes and immunoglobulins that further kill or
Christispira species
inhibit the growth of microorganisms. With time
Treponema denticola
and altered host resistance, the sinuses of the adult
Treponema orale
human acquire an indigenous microflora and occa-
Treponema macrodentium
sionally are subjected to colonization (see Box 3)
Treponema vincentii
[1,24,25,27,28].
Borrelia species
Leptospira species
Yeasts
Aspergillus species Odontogenic infections
Candida albicans
Other Candida species Although numerous articles in the surgical litera-
Geotrichum species ture discuss odontogenic infections, they do not
Hemispora species identify changes in the microbiology of these infec-
tions over time, nor do they make comparisons using
4 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15
case controlled cohorts [23,29 – 39]. A recent inves-
tigation has attempted to identify such changes and
make such comparisons [40]. Although not ideal in
design (possessing all of the imperfections inherent in
a retrospective chart review), it remains the only case-
controlled cohort comparison of odontogenic infec-
tions over time.
The retrospective record review performed by
Storoe et al [40] was conducted at the Cleveland
MetroHealth Medical Center, a large county teaching
hospital that serves an urban population of 1.9 million
and a rural population of 2 million in northeastern
Ohio. Hospital charts and radiographs were reviewed
from two patient cohorts admitted to that institution;
the first during the 81 months between March 1983
and November 1989 (1980s patients) [33] and the
other during the 81 months between July 1992 and
March 1999 (1990s patients). Admission criteria were
face or neck swelling, which suggested abscess or
cellulitis and one or more of the following: temper-
ature above 38
C, white blood cell count more than
10.8  103/mL, or concern about airway compromise.
Patient characteristics reviewed were age, gender,
race, admission temperature, admission white blood
count count, fascial space(s) involved, tooth of etiol-
ogy, duration of hospitalization, and bacteria isolated.
An exhaustive statistical analysis was undertaken to
analyze and compare the two populations and any
changes in their microbiology.
Box 3. Some of the most common
indigenous and colonizing nasal and
paranasal sinus flora [24 – 28]
Aerobic and facultative isolates
Corynebacterium diptheria
Corynebacterium species
Haemophilus influenzae
Haemophilus parainfluenza
Neisseria species
Staphylococcus species
Streptococcus pneumoniae
Moraxella species
Micrococcus species
Neisseria meningitidis
Staphylococcus aureus
Staphylococcus epidermidis
Streptococcus viridans
Streptococcus pneumonia
Anaerobic isolates
Propionibacterium acnes
No significant differences were found between the
two cohorts for age, gender, race, admission temper-
ature, admission white blood cell count, space
involvement, or length of stay. The manner in which
the bacterial isolates were reported was limited by the
ability to retrieve data in a retrospective review of
records. The reporting was performed in the follow-
ing manner. All isolates obtained from any one
patient, whether from a ‘‘sterile abscess,’’ multiple
sites, or multiple cultures, were recorded for that
patient. Most patients from the 1980s and 1990s
had multiple organisms isolated. The reporting mech-
anism was the number of isolates per patient per
admission. 180 bacterial isolates were identified from
79 patients from the 1980s compared to 115 from 71
patients from the 1990s. For the initial analysis,
isolates were grouped into four broad categories:
gram-positive cocci, other gram-positive bacteria,
gram-negative anaerobes, and other gram-negative
bacteria (Table 1). Gram-positive cocci and gram-
negative anaerobes were isolated significantly less
frequently from the 1990s patients than from the
1980s patients.
Table 2 reports the number of isolates per patient
per admission. For instance, a-hemolytic streptococci
were isolated in 47 patients from among 79 during
the 1980s and in 24 patients from among 71 during
the 1990s. When individual isolates from the cohorts
were compared, significant differences were found
for a-hemolytic streptococci, Bacteroides melanino-
genicus, coagulase-negative staphylococci, Eikenella
corrodens, Staphylococcus epidermidis, Neisseria
species, and b-lactamase – positive Bacteroides.
Culture and antibiotic sensitivity data were avail-
able from only ten 1990s patients (14%) (Table 3). The
1980s patient data were unavailable. Of the 32 bac-
terial isolates, 26 (81%) were resistant to one or more
antibiotic. 62% of the resistant bacteria were gram
positive and 38% were gram negative; Staphylococcus
aureus and coagulase-negative staphylococci were the
most common gram-positive antibiotic-resistant bac-
teria isolated. Klebsiella pneumoniae was the most
common gram-negative antibiotic-resistant isolate.
Until the mid 1970s, researchers believed that
odontogenic infections were caused by a single species
of aerobic or facultative bacteria [37]. Subsequently, it
has been well established that odontogenic infections
are polymicrobial [23,29,30,33 – 37,39,41 – 43].
Researchers also have observed that the bacteria iden-
tified from odontogenic infections have changed over
the decades [37]. Moenning et al [37] reported that the
bacterial flora of odontogenic infections is no longer
predominately facultative or microaerophilic, with
Staphylococcus and Streptococcus as the primary
 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15
Table 1
Bacteria isolated
Number isolated from the
1980s patients (percent of total)
Gram-positive cocci 101 (56)
Other gram-positive bacteria 9 (5)
Gram-negative anaerobes 41 (23)
Other gram-negative bacteria 29 (16)
Total isolates 180 (100)
From Storoe W, Haug RH, Lillich TT. The changing face of odontogenic infections. J Oral Maxillofac Surg 2001;59:739 – 48;
with permission.
genera, but is more often a mixed flora, with anaerobes
outnumbering aerobes 2:1. a-hemolytic streptococci
are the most frequently isolated bacteria [23,29,35],
although Bacteroides melaninogenicus has been
reported as the most common in some studies
[30,32]. In one study, 19 bacterial genera or species
were isolated in the 1980s patients, with a-hemolytic
streptococci predominating almost 2:1 over B. mela-
ninogenicus and b-hemolytic streptococci predominat-
ing 3:1 over S. epidermidis and almost 4:1 over S.
aureus and E. corrodens (see Table 2). 24 genera or
species were isolated in the 1990s patients, with
coagulase-negative staphylococci, a-hemolytic strep-
tococci, and b-hemolytic streptococci having nearly a
1:1:1 ratio and predominating equally over all other
bacteria by more than a 2:1 ratio (see Table 2).
No isolates of B. melaninogenicus were reported
in the 1990s patients; however, 13 other Bacteroides
isolates were identified. Significant differences
between the groups were found for a-hemolytic
streptococci, B. melaninogenicus, coagulase-negative
staphylococci, E. corrodens, S. epidermidis, Neisse-
ria species, and b-lactamase – positive Bacteroides,
which suggests that there has been a shift in the
microbiologic flora involved in odontogenic infec-
tions. a-hemolytic Streptococci were the most com-
mon isolates in both groups, which confirms previous
reports of the frequency with which these bacteria are
isolated from odontogenic infections [23,29,35]. It is
clear that these common gram-positive cocci remain a
threat to patients with odontogenic infections. The
significance of the differences in the other isolates is
less clear because of changes in bacterial nomencla-
ture and laboratory protocols in the past 10 years. For
example, there was a statistically significant differ-
ence between groups for B. melaninogenicus largely
because there were no isolates identified from the
1990s patients. This is because B. melaninogenicus
was reclassified in 1990; bacteria identified as
B. melaninogenicus in the 1980s patients would have
5
Number isolated from the Chi-square
1990s patients (percent of total) statistic P value
80 (70) 5.357 0.021
5 (4) 0.066 0.8
15 (13) 4.323 0.038
15 (13) 0.520 0.5
115 (100)
been reported as either Prevotella or Porphyromonas
in the 1990s patients [44]. Consequently, the differ-
ence between the two groups for this obligate anaer-
obe is a statistical anomaly.
The statistically significant difference between
groups for S. epidermidis and coagulase-negative
staphylococci can be explained by changes in labo-
ratory protocol. S. epidermidis is a coagulase-nega-
tive bacterium. During the 1980s, it was common to
speciate isolates, although this provided little clin-
ically useful information. Consequently, protocol
changes were made that resulted in all coagulase-
negative staphylococci being reported as such with-
out further identification. A more accurate picture
of the statistically significant differences between
groups emerges by analyzing S. epidermidis and
coagulase-negative staphylococci together. When this
is done, there is no significant difference between the
groups. The significant decreases in Neisseria species
and E. corrodens from the 1980s patients to the
1990s patients are most probably caused by labo-
ratory protocol changes and changes in therapy,
respectively. The complete absence of Neisseria iden-
tified in the 1990s patients suggests a decision by the
clinical laboratory not to seek this species in clinical
specimens from odontogenic infections. The decrease
in E. corrodens isolated from the 1990s patients may
be the result of improved treatment of anaerobic
infection with antibiotics that have been developed
in the past 10 years. Unfortunately, no data about
preadmission treatment were available.
The only other bacteria with a significant differ-
ence between groups were b-lactamase – positive Bac-
teroides, which were not reported at all in the 1980s
patients. The reason for this change is unclear, but
rather than reflecting some important shift in odonto-
genic infection etiology, it is probably caused by a
change in nomenclature or protocol. For example, it
is curious that no Prevotella or Porphyromonas
species were reported in the 1990s patients despite
6 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15

Table 2
Frequency of bacteria isolated, per patient, per admission
Bacteria 1980s patients 1990s patients Chi-square statistic P value
Gram-positive cocci
a-hemolytic streptococci 47 24 9.900 0.002
b-hemolytic streptococci 22 21 0.055 0.815
Staphylococcus aureus 12 8 0.498 0.480
Coagulase-negative staphylococci — 18 22.759 0.001
Staphylococcus epidermidis 15 — 14.979 0.001
g-hemolytic streptococci 2 5 0.846a 0.358
Peptostreptococcus species 3 3 0.000a 1.000
Enterococcus species — 1 0.003a 0.957
Other gram-positive bacteria
Diphtheroids 6 4 0.023a 0.878
Actinomyces species 2 — 0.406a 0.524
Lactobacillus species 1 — 0.000a 1.000
Corynebacterium species — 1 0.003a 0.957
Gram-negative anacrobes
Bacteroides malaninogenicus 27 — 29.59 0.001
Bacteroides (b-lactamase +) — 6 4.928a 0.03
Bacteroides (b-lactamase ) — 4 2.660a 0.103
Bacteroides (not fragilis) — 2 0.622a 0.430
Eusobacterium necrophorum — 2 0.622a 0.430
Bacteroides fragilis 1 1 0.000a 1.000
Other gram-negative bacteria
Eikenella corrodens 13 3 5.870 0.02
Haemophilus influenzae 8 2 2.144a 0.143
Neisseria species 9 — 6.704a 0.01
Klebsiella species 4 3 0.000a 1.000
Enterobacter species 3 1 0.159a 0.690
Escherichia coli 3 — 1.155a 0.283
Citrobacter species 1 — 0.000a 1.000
Haemophilus hemolyticus 1 — 0.000a 1.000
Proteus mirabilis — 1 0.003 0.957
Actinobacter — 1 0.003a 0.957
calcoaceticus (Iwoffi)
Sorralia marcescens — 1 0.003a 0.957
Pseudomonas aeruginosa — 1 0.003a 0.957
Pseudomonas species — 1 0.003a 0.957
Stenotrophomonas maltophilia — 1 0.003a 0.957
Total isolates 180 115
From Storoe W, Haug RH, Lillich TT. The changing face of odontogenic infections. J Oral Maxillofac Surg 2001;59:739 – 48;
with permission.
a
Continuity adjustment. Chi-square is reported because more than 25% of the cells had expected counts less than 6.

their importance in periodontal and other odontogenic
infections. It has been reported that more than 50% of
Prevotella species are b-lactamase producers [45]. It
is possible that the significant difference seen
between groups is another consequence of the
nomenclature change in B. melaninogenicus. It can
be concluded that contrary to previous suggestions,
there has been little change in the kinds of bacteria
isolated from odontogenic infections in the two
groups of patients 10 years apart; a-hemolytic strep-
tococci remain the most frequently isolated bacteria.
From the limited antibiotic sensitivity data avail-
able, it was not possible to determine if there have
been changes between groups in the kind, number,
and frequency of resistant isolates. There were no
data available for the 1980s patients, and only 10% of
the 1990s patients had a culture and antibiotic sen-
sitivity test performed at least once during their
hospitalization (Table 3). Despite their limitations,
the data confirm some disturbing trends. 81% of the
isolates were resistant to one or more antibiotics.
Despite the relatively low overall occurrence (see
 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15
Table 3
Antibiotic resistance for the 1990s patient’s isolates
Isolate
g- streptococci (not Enterococcus)
b-streptococci (non A/B)
Coagulase-negative staphylococci
Enterococcus secies
Staphylococcus aureus
Total gram positive
Acinetobacter calcoacelicus (Iwoffi)
Bacteroides species
Eikenella species
Enterobacter cloacae
Klebsiella pneumoniae
Pseudomonas aeruginosa
Pseudomonas species
Stenotrophomonas maltophilia
Total gram negative
Totals
From Storoe W, Haug RH, Lillich TT. The changing face of odontogenic infections. J Oral Maxillofac Surg 2001;59:739 – 48;
with permission.
Table 2), S. aureus was the most frequently identified
antibiotic resistant isolate. In addition to its innate
virulence, S. aureus is of increasing concern because
it, along with other gram-positive cocci, is becoming
resistant to most common antibiotics. For example,
methicillin-resistant S. aureus that also exhibits inter-
mediate resistance (minimum inhibitory concentra-
tion = 8 mg/mL) to vancomycin, the antimicrobial
regarded as the ‘‘antibiotic of last resort’’ for these
bacteria, recently has been reported in Japan and the
United States [46].
Another trend is the number of antibiotic-resistant,
coagulase-negative staphylococci isolated. These bac-
teria long have been regarded as ‘‘apathogenic’’
members of the normal flora but are increasingly
becoming recognized as important causes of infec-
tions, especially those acquired in hospitals [47].
Because most of the infections they cause are noso-
comial, it should not be surprising to see increasing
multiple antibiotic resistance. Against a background
of the empiricism with which many clinicians con-
tinue to treat these patients (eg, only 10% of the
patients in the 1990s had a culture and antibiotic
sensitivity test performed), one may predict that there
will be more initial treatment failures and a con-
sequent increase in patient morbidity and total cost
of care.
Moenning et al [37] suggested that wounds or
cavities could become contaminated with normal
skin microflora, including Staphylococcal species,
through external drainage. It is possible that such
contamination occurred during the harvest of culture
7
No. Resistant No. Sensitive Total
0 1 1
0 1 1
6 0 6
1 0 1
9 1 10
16 3 19
1 0 1
0 1 1
1 0 1
1 0 1
3 0 3
1 2 3
2 0 2
1 0 1
10 3 13
26 6 32
material in the cases included in this investigation,
but it is unlikely because proper surgical technique
and rigorous preoperative antimicrobial skin prepara-
tion were used. One would expect surgical technique
and skin preparation to improve with time, thereby
reducing the incidence of sample contamination
from cutaneous sources. This does not seem to be
the case in this comparison of odontogenic infections
from the 1980s and the 1990s, however, which
showed no change in the frequency of Staphylococ-
cus species isolated. Assuming that proper technique
was followed and that the sites were prepared
properly, it is improbable that there has been a shift
in the kinds of bacteria causing odontogenic infec-
tions in the last 10 years.
Trauma-related infections
Infection is a concern during the management of
patients who have sustained maxillofacial trauma
[48]. It may be a result of the injury itself or the
treatment of that injury. Infection may cause systemic
sepsis, fracture nonunion, failure in cutaneous wound
healing, and the need for a prolonged hospitalization
or additional surgery [49,50]. Infection in the trauma
patient also may produce irreversible damage, includ-
ing disfigurement, dysfunction, and death. From the
most empirical standpoint, infection subsequent to
trauma involves the introduction or inoculation of
microorganisms that routinely exist as normal flora
(Boxes 1 – 3, Table 1) beyond the body’s external
8 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15

protective media (the skin or mucosa) and the altera- the use of a rigorous protocol of tetanus globulin
tion of the homeostatic relationship between the and toxoid administration, the incidence of wartime
individual and his or her environment. Nothing more clostridial infection (in US casualties) was reduced
suddenly alters homeostasis than acute trauma, and from 5% in World War I, to 0.7% in World War II,
the major organ systems are mobilized to respond to and finally 0.08% during the Korean War [68].
increased demands. Because of the systemic response Civilian data are unavailable.
to injury, patients who have sustained major trauma
are far more susceptible to wound infection than
individuals who have suffered mere localized injuries
[48,51,52]. Box 4. Some of the most common
The skin and mucosa are essential barriers to the microorganisms transmitted by human
entry of microorganisms into the underlying connec- bites [53,54,56,58,59,63,64]
tive tissue and organs. The integument is normally
exposed to bacterial flora only to the depth of the Aerobic and facultative isolates
adnexa. Scalp and maxillofacial cutaneous lacerations Acinetobacter species
expose the underlying structures to the microorgan- Corynebacterium species
isms listed in Box 1. Intraoral lacerations and even Eikenella corrodens
simple fractures through the periodontal ligament Enterococcus species
expose the bone and viscera to the otherwise normal Haemophilus influenzae
flora (see Box 2). Finally, mid- and upper-facial Haemophilus parainfluenzae
fractures contaminate the victim with multiple micro- Klebsiella pneumonia
organisms from the skin, nose, and paranasal sinuses Neisseria species
(Box 3). Nocardia species
The introduction of microorganisms beyond the Proteus species
individual’s own indigenous flora also may occur Pseudomonas
with bite wounds (Boxes 4,5) or exposure to a Staphylococcus aureus
contaminated environment. Streptococcus viridans, Staphylococcus epidermidis
Bacteroides species, Peptostreptococcus species, Streptococcus
and E. corrodens are among the microorganisms most n-hemolytic
commonly isolated from humans who develop infec- p-hemolytic
tions after having been bitten by other humans, and S. Group A o-hemolytic
aureus, Pasteurella multocida, Streptococci species, Non – group A o-hemolytic
and Bacteroides species are pathogens that often have Anaerobic isolates
been isolated from mammalian bite wound infections Bacteroides fragilis group
[12,53 – 64]. The rabies virus is always a concern for Bacteroides oralis
raccoon, skunk, fox, or bat bites. Contamination also Bacteroides ureolyticus
may include debris or foreign material impacted into Bifidobacterium species
the wound, farm-related injuries that inoculate the Eubacterium species
victim with animal feces (gram-negative enteric Fusobacterium nucleatum
organisms, group D streptococci, and anaerobes) Fusobacterium necrophorus
[65], earth-borne contaminants, such as Clostridium Peptococcus niger
tetani and Actinomyces species, and contamination Peptostreptococcus
with free-standing water (E. coli). Fresh-water con- asaccharolyticus
tamination, particularly in brackish water, has caused Peptostreptococcus magnus
Aeromonas hydrophila infection [59]. This aerobic Prevotella melaninogenica
gram-negative rod can cause severe facial cellulitis Prevotella intermedia
and myonecrosis. Enteric organisms such as E. coli Veillonella species
also frequently contaminate wounds exposed to Other rare pathogens
coastal seawater. Hepatitis B virus
When burns, concussion, maceration, or avulsion Herpes simplex virus
injure muscle, myonecrosis may ensue, and if con- HIV
taminated, it frequently results in infection [66]. Mycobacterium tuberculosis
Anaerobic bacteria such C. perfringens and C. Treponema pallidum
tetani are the most frequent causes [67]. Through
 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15 9

Box 5. Some of the most common Rio Bravo virus

microorganisms transmitted by animal Simian herpes virus (macaque
bites [12,53 – 64] monkeys only)
Spirillum minus
Aerobic and facultative isolates Streptobacillus moniliformis
Yersinia pestis
Acinetobacter species
Aeromonas hydrophila
Bacillus subtilis
Bordetella species
Brucella canis Neither the indigenous maxillofacial flora nor the
Capnocytophagia canimorsus introduction of common contaminants has changed
Chromobacterium species in the management of maxillofacial trauma. Rather, the
Clostridium perfringens development of nosocomial infections and the human
Corynebacterium species body’s weakened systemic response to trauma are
EF-4 responsible. Central nervous system injury sharply
EF-7 increases the risk of infection in the trauma patient,
Eikenella corrodens with gram-negative pneumonia being among the great-
Enterococcus species est adverse outcomes [52]. Serratia and Pseudomonas
Flavobacterium species species are becoming more prominent offenders in the
Haemophilus aphrophilus head-injured patient because of their frequent presence
Klebsiella in intensive care units. The emergence of new and
Moraxella catarrhalis more resistant strains of staphylococci, especially
Moraxella weaveri methicillin-resistant S. aureus, has made the manage-
Neisseria species ment of nosocomial infection tenuous, with the under-
Proteus species standing that surgical patients who develop
Pasteurella multocida nosocomial methicillin-resistant S. aureus infection
Proteus mirabilis are twice as likely to die than those who do not
Pseudomonas [13,15]. The fear of the emergence of new and resistant
Serratia marcescens strains of bacteria has prompted the trauma surgeon to
Staphylococcus aureus withhold the administration of antibiotics until after
Staphylococcus epidermidis the trauma patient has become infected (if an infection
Staphylococcus saprophyticus does develop) [7,11]. This trend also has extended to
Streptococcus species the management of maxillofacial trauma victims.
Weeksella zoohelcum Many patients with clean-contaminated wounds who
Anaerobic isolates would previously have been considered candidates for
Arachnia propionica antibiotics are no longer considered such.
Bacteroides spp
Eubacterium spp
Fusobacterium spp
Leptotrichia Nasal and paranasal sinus infections
Peptococcus
Peptostreptococcus spp Acute maxillary sinusitis
Propionibacterium acnes
Veillonella spp Sinusitis is a disease that results from an infection
Other rare pathogens of one or more of the paranasal sinuses [69,70].
Bartonella henselae Acute sinusitis is generally categorized by its etiol-
Clostridium tetani ogy: nosocomial or community-acquired and viral,
Francisella tularensis bacterial, or fungal [69,70]. Combinations of causes
Hepatitis B virus are possible. Perhaps the most common cause for
Herpes simplex virus acute maxillary sinusitis is a rhinovirus infection. The
Leptospira species ability to culture viruses accurately is a relatively
Rabies virus recent development, and changes or trends in viral
pathogenicity have not been established.
10 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15

The origin of acute community-acquired bacterial

Box 6. Some common microorganisms
sinusitis (ACABS) has been well substantiated for the
responsible for acute maxillary and
past five to six decades (Box 6) [24,69 – 74].
ethmoid sinusitis [49,69,70,72,73]
S. pneumonia and H. influenzae account for more
than half of the occurrences of this form of sinusitis.
Aerobic and facultative isolates
Other Streptococcus species—S. aureus, M. catar-
Escherichia coli
rhalis, a-hemolytic streptococci, and anaerobic bac-
Haemophilus influenzae
teria—also have been isolated. Complicating
Moraxella catarrhalis
diagnosis and treatment even more has been the
Neisseria species
observation that viruses and bacteria may be iden-
Pseudomonas aeruginosa
tified as simultaneously causing ACABS. When
Staphylococcus aureus
cultures fail to yield bacteria for ACABS (approx-
Streptococcus pneumoniae
imately 40% of the time), a viral cause should be
Streptococcus pyogenes
suspected. Fungi also may be responsible for ACABS
Streptococcus
but are more associated with nosocomial sinusitis and
n-hemolytic
sinus disease that develops in the compromised host,
o-hemolytic
such as someone with diabetes. S. aureus, P. aerugi-
Anaerobic isolates
nosa, Serratia marcescens, K. pneumonia, Entero-
Bacteroides species
bacter species, Proteus mirabilis, and Legionella
Fusobacterium species
pneumonia are also associated with nosocomial
Peptostreptococcus species
sinusitis and the compromised host [69].
Viruses
The relative distribution of microorganisms re-
sponsible for ACABS has not changed in either Adenovirus
appearance or prevalence over the past five decades Influenza virus
(Box 6) [69,73,74]. What has changed is their Parainfluenza virus
antimicrobial susceptibility. Penicillin-resistant Rhinovirus
S. aureus has emerged as a difficult offender to
manage, along with b-lactam – resistant strains of
H. influenzae and M. catarrhalis. Worse yet has
been the emergence of multiple strains of resistant Infectious rhinitis
S. pneumoniae. The changes in the microbiology of
acute maxillary sinusitis have not been in the type or Infectious rhinitis may be classified as acute, self-
frequency of microorganisms but rather the emer- limiting, or chronic, and it is the major component of
gence of resistant strains. the ‘‘common cold’’ [75]. Viruses are the most fre-
quent cause of infectious rhinitis, and rhinoviruses are
Chronic maxillary sinusitis the most frequent etiologic agent among them (Box 8)
[26 – 28,75]. Bacteria are an infrequent cause of iso-
The specific cause and pathogenicity of chronic lated nasal infections but are common as a cause of
sinusitis disease remain unanswered, and its success- combination nasal/paranasal sinus infections [75].
ful treatment modalities continue to be elusive Granulomatous nasal infections, such as rhinoscle-
[49,72,73]. Although some cases of chronic sinusitis roma, tuberculosis, syphilis, aspergillosis, and mucor-
disease exist as a separate and distinct entity, it is mycosis, are rare and are caused by numerous
believed that most cases arise from ACABS treatment microorganisms, including fungi, protozoa, and myco-
failures. S. pneumoniae, H. influenzae, and other bacterium [26 – 28,75].
streptococcal species have been isolated from patients
with chronic sinusitis disease (Box 7) [24,69 – 74]. Frontal and sphenoid sinusitis
Recent reports have identified b-lactamase – pro-
ducing strains of S. pneumonia, H. influenzae, and The microbial flora of the infected frontal and
M. catarrhalis in ACABS aspirates [65,73]. It seems sphenoid sinuses are separate and distinct from those
probable that chronic sinusitis disease is merely the of the maxillary and ethmoid sinuses. Acute sphenoid
continued manifestation of ACABS caused by inef- and frontal sinusitis is uncommon and difficult to
fective treatment. The changes in the microbiology of diagnose and carries with it a significant degree of
chronic sinusitis disease have been in the emergence morbidity but accounts for less than 5% of paranasal
of resistant strains. sinus infections [76 – 78]. Headache is the most
 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15 11

consistent finding, and it may be accentuated with

Box 8. Some common microorganisms
activity. The degree of pain associated with it can be
responsible for rhinitis [26 – 28,75]
debilitating. Several organisms have been associated
with this disease (Box 9) [70,76 – 78]. S. aureus,
Aerobic and facultative isolates
S. pneumoniae, H. influenzae, and other streptococci
Klebsiella rhinoscleromatis
are the most frequent pathogens [76 – 78]. Although
Mycobacteria
changes in the microbiology of frontal and sphenoid
Mycobacterium tuberculosis
sinusitis have not been identified by case-controlled
Spirochetes
cohort investigations, changes in therapeutic trends
Treponema pallidum
point to the need for antibiotics that address b-lacta-
Fungi
mase – producing strains of organisms. Especially
problematic are b-lactamase – producing staphylo- Absidia species
cocci, streptococci, and H. influenzae. This finding Aspergillus flavus
suggests that the changes in the microbiology of acute Aspergillus fumigatus
frontal and sphenoid sinusitis are, as in other nasal and Bipolaris species
paranasal infections, not in the type or frequency of the Cladosporium species
microorganisms isolated but in the emergence of Curvularia species
resistant strains. Exophiala species
Exserohilum species
Histoplasma capsulatum
Mucor species
Cutaneous infections Rhinosporidium seeberi
Rhizopus species
A plethora of cutaneous maxillofacial infections Wangiel1la species
occur with relative infrequency in the population Protozoa
[79,80]. Many are rare and exotic, such as noma Leishmaniasis
(cancrum oris), the progressively deforming gangre- Viruses
nous stomatitis most often seen in debilitated and Adenovirus
malnourished Third World children. It is believed Enterovirus
to result from P. aeruginosa and the fusospirochetal Influenza virus
organisms, Borrelia vincentii and Fusobacterium Parainfluenza virus
nucleatum [79,80]. Other unusual cutaneous max- Respiratory syncytial virus
illofacial infections include such zoonotic diseases Rhinovirus
as cat scratch disease, caused by Bartonella hense-
lae; leptospirosis, caused by Leptospira species;
and the Bubonic Plague, caused by Y. pestis [81 –
83]. Zoonotic diseases occur from the inoculation the aforementioned diseases is characterized by
into the humans of microorganisms that exist non- raised, red, painful cutaneous lesions that are man-
pathogenically in the animal population. Each of aged with antibiotic therapy [81 – 83]. Although
exotic and unusual, changes in frequency of
appearance or virulence of these types of infections
Box 7. Some common microorganisms have not been observed.
responsible for chronic maxillary sinus- More common cutaneous infections that affect
itis [24,69 – 74] the maxillofacial region are erysipelas, caused by
group A streptococci; impetigo, the result of group A
Aerobic and facultative isolates streptococci or S. aureus; and furuncles, carbuncles,
Corynebacterium species and folliculitis, which may be caused by S. aureus,
Haemophilus influenzae C. albicans, or P. aeruginosa [79,80]. Although
Streptococcus pneumoniae associated with compromised hosts and institutional
Anaerobic isolates and regional outbreaks, there have been neither
Bacteroides species changes in the frequency of appearance of these
Peptostreptococcus species infections in the general population nor changes in
Veillonella species virulence of the organisms. Buccal cellulitis of
infancy, often confused with an odontogenic infec-
12 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15
Box 9. Some common microorganisms
responsible for sphenoid and frontal
sinusitis [44,70,76,78]
Aerobic and facultative isolates
o-hemolytic Streptococci
Citrobacter
Coliform bacilli
Enterococci
Escherichia coli
Haemophilus influenzae
Klebsiella pneumonia
Pneumococci
Proteus
Pseudomonas aeruginosa
Serratia
Staphylococcus aureus
Staphylococcus epidermidis
Streptococcus pneumoniae
Streptococcus viridans
Anaerobic isolates
Anaerobic Streptococci
Fungi
Aspergillus
tion from the deciduous teeth, is another unusual
cutaneous maxillofacial infection [80]. Its cause is
H. influenzae, and it is cared for with constitutional
support, antibiotics, and antipyretics. This condition
also has shown no apparent change in virulence or
frequency of distribution.
Perhaps the most frequent cutaneous infection of
the maxillofacial region is that caused by the herpes
simplex virus [9]. Although it may produce mucocu-
taneous lesions of the entire face, it is mostly limited
to the skin and mucosa of the oral cavity and perioral
region. Conjunctival and nasal/perinasal manifesta-
tions are also common. The virus is introduced into
the body through mucosal exposure or abraded skin.
Although the initial infection may be subclinical,
recurrence is common. More than 90% of adults have
antibodies to the herpes simplex virus. Prevalence
increases with decreasing socioeconomic status, and
it is more prevalent in African Americans and persons
living in Western Europe than in the United States.
Although relief and complacency developed in the
medical and lay populations after the discovery of
acyclovir, concern has arisen again because of the
increasing frequency of isolation of acyclovir-resist-
ant strains of herpes simplex virus.
Summary
What is considered the indigenous maxillofacial
flora begins during birth with the acquisition of the
flora of the birth canal and then changes from infancy
to adulthood as a result of the specific anatomic
region, environmental influences, and host factors.
The changes in the microbiology of infections of
odontogenic etiology rest in changes in nomenclature
and in our ability to isolate organisms. The changes in
the microbiology of infections associated with max-
illofacial trauma are not caused by the injuries but are
related to nosocomial systemic infection and the
emergence of resistant microorganisms. The changes
in the microbiology of nasal and paranasal sinus
infections are associated with the emergence or
reemergence of strains of bacteria that are resistant
to common antibiotics. Cutaneous infections appear
infrequently, but the most common etiologic agent,
herpes simplex virus, also has been associated with
the isolation of resistant strains at an alarming rate.
References
[1] Ohira R. Five-year-old girl battles flesh-eating bacteria.
Honolulu Star-Bulletin, June 28, 2002. p. 1.
[2] Lemonick MD. The killers all around: new viruses and
drug resistant bacteria are reversing human victories
over infectious disease. Time 1994;144:62.
[3] Spake A. Losing the battle of the bugs. US News and
World Report 1999;126:52 – 60.
[4] Satcher D. Emerging infections: getting ahead of the
curve. Emerg Infect Dis 1995;1:1 – 6.
[5] Neu HC. The crisis in antibiotic resistance. Science
1992;257:1064 – 73.
[6] Dacre J, Emmerson A, Jenner E. Gentamicin-methicil-
lin – resistant Staphylococcus aureus: epidemiology of
an outbreak. J Hosp Infect 1986;7:130 – 6.
[7] American Society of Health-System Pharmacists.
ASHP therapeutics guidelines on antimicrobial pro-
phylaxis in surgery. Am J Health Syst Pharm 1999;
56:1839 – 88.
[8] Barett F, McGehee R, Finland M. Methicillin-resistant
Staphylococcus aureus at Boston City Hospital. N
Engl J Med 1998;279:441 – 8.
[9] Corey L. Herpes simplex virus. In: Mandell GL, Ben-
nett JE, Dolin R, editors. Principles and practice of
infectious disease. 5th edition. Philadelphia: Church-
ill-Livingstone; 2000. p. 1564 – 80.
[10] Crossley K, Loesch D, Landesman B, et al. An out-
break of infections caused by strains of Staphylococcus
aureus resistant to methicillin and aminoglycosides. I:
clinical studies. J Infect Dis 1979;139:273 – 9.
[11] Fukatsu K, Saito H, Matsuda T, et al. Influences of
type and duration of antimicrobial prophylaxis on an
 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15
outbreak of methicillin-resistant Staphylococcus aur-
eus and on the incidence of wound infection. Arch
Surg 1997;132:1320 – 5.
[12] Goldstein E, Citron D, Feingold S. Dog bite wounds
and infection: a prospective clinical study. Ann Emerg
Med 1980;9:508 – 12.
[13] Kirkland KB, Briggs JP, Trivette SL, et al. The im-
pact of surgical-site infections in the 1990s: attribut-
able mortality, excess length of hospitalization, and
extra costs. Infect Control Hosp Epidemiol 1999;
20:725 – 30.
[14] Klimek J, Marsik F, Bartlett R, et al. Clinical, epide-
miologic and bacteriologic observations of an outbreak
of methicillin-resistant Staphylococcus aureus at a
large community hospital. Am J Med 1976;61:340 – 5.
[15] Panlilio AL, Culver DH, Gaynes RP, et al. Methicillin-
resistant Staphylococcus aureus in US hospitals,
1975 – 1991. Infect Control Hosp Epidemiol 1992;13:
582 – 6.
[16] Pavillard R, Harvey K, Douglas D, et al. Epidemic of
hospital-acquired infection due to methicillin-resistant
Staphylococcus aureus in Victorian hospitals. Med J
Aust 1982;1:451 – 4.
[17] Gallis HA. Normal flora and opportunistic infections.
In: Joklik WK, Willett HP, Amos DB, et al. Zinsser
microbiology. 19th edition. Norwalk (CT): Appleton
and Lange; 1988; p. 337 – 42.
[18] Koneman EW, Allen SD, Dowell VR, et al. Color atlas
and textbook of diagnostic microbiology. 3rd edition.
Philadelphia: Lippincott; 1988; p. 17 – 8.
[19] Schuster GS. Microbiology of the orofacial region.
In: Topazian RG, Goldberg MH, Hupp JR, editors.
Oral and maxillofacial infections. 4th edition. Phila-
delphia: WB Saunders Co.; 2002. p. 30 – 42.
[20] Shuster GS. Oral flora and pathogenic organisms. In-
fect Dis Clin N Am 1999;13:757 – 74.
[21] Tramont EC, Hoover DL. Innate (general or nonspe-
cific) host defense mechanisms. In: Mandell GL, Ben-
nett JE, Dolin R, editors. Principles and practice of
infectious disease. 5th edition. Philadelphia: Churchill-
Livingstone; 2000. p. 31 – 8.
[22] De Pauw BE, Donnelly JP. Infections in the compro-
mised host: general principles. In: Mandell GL, Ben-
nett JE, Dolin R, editors. Principles and practice of
infectious disease. 5th edition. Philadelphia: Churchill-
Livingstone; 2000. p. 3081 – 2.
[23] Labiola JD, Mascaro J, Alpert B. The microbiologic
flora of orofacial abscesses. J Oral Maxillofac Surg
1983;41:711 – 4.
[24] McCarter YS. Laboratory microbiological diagnostic
techniques. In: Topazian RG, Goldberg MH, Hupp
JR, editors. Oral and maxillofacial infections. 4th edi-
tion. Philadelphia: WB Saunders Co.; 2002. p. 43 – 61.
[25] Bamberger DM. Antimicrobial treatment of sinusitis.
Semin Respir Infect 1991;6:77 – 84.
[26] Larsen HS. Host-parasite interaction. In: Mahon CR,
Manuselis G, editors. Textbook of diagnostic microbi-
ology. 2nd edition. Philadelphia: WB Saunders Co.;
2000. p. 213 – 6.
13
[27] Weiser JN, Kim JO. The respiratory tract microflora
and disease. In: Tannock GW, editor. Medical impor-
tance of the normal microflora. Boston: Kluwer Aca-
demic; 1999. p. 47 – 73.
[28] Volk WA, Gebhart BM, Hammarskjold M, et al. Nor-
mal flora, infections, and bacterial invasiveness. In:
Volk WA, Gebhart BM, Hammarskjold M, et al,
editors. Essentials of medical microbiology. 5th edition.
Philadelphia: Lippincott-Raven; 1996. p. 315 – 32.
[29] Bartlett JG, O’Keefe P. The bacteriology of periman-
dibular space infections. J Oral Surg 1979;37:407 – 9.
[30] Chow AW, Roser SM, Brady FA. Orofacial odonto-
genic infections. Ann Intern Med 1978;88:392 – 402.
[31] Dodson TB, Perrott DH, Kaban LB. Pediatric maxillo-
facial infections: a retrospective study of 113 patients.
J Oral Maxillofac Surg 1989;47:327 – 30.
[32] Har-el G, Aroesty JH, Shaha A, et al. Changing trends
in deep neck space abscess: a retrospective study of
110 patients. Oral Surg Oral Med Oral Pathol 1994;77:
446 – 50.
[33] Haug RH, Hoffman MJ, Indresano AT. An epidemio-
logic and anatomic survey of odontogenic infections.
J Oral Maxillofac Surg 1981;49:976 – 80.
[34] Heimdahl A, Von Konow L, Satoh T, et al. Clinical
appearance of orofacial infections of odontogenic ori-
gin in relation to microbiological findings. J Clin Mi-
crobiol 1985;22:299 – 302.
[35] Hunt DE, King TJ, Fuller GE. Antibiotic susceptibility
of bacteria isolated from oral infections. J Oral Surg
1978;36:527 – 9.
[36] Kannangara DW, Thadepalli H, McQuirter JL. Bacteri-
ology and treatment of dental infections. Oral Surg
Oral Med Oral Pathol 1980;50:103 – 9.
[37] Moenning JE, Nelson CL, Kohler RB. The microbiol-
ogy and chemotherapy of odontogenic infections.
J Oral Maxillofac Surg 1989;47:976 – 85.
[38] Sakaguchi M, Sato S, Ishigama T, et al. Characteristics
and management of deep neck infections. Int J Oral
Maxillofac Surg 1997;26:131 – 4.
[39] Sethi DS, Stanley RE. Deep neck abscess-changing
trends. J Laryngol Otol 1994;108:138 – 42.
[40] Storoe W, Haug RH, Lillich TT. The changing face of
odontogenic infections. J Oral Maxillofac Surg 2001;
59:739 – 48.
[41] Gorbach SL, Gilmore WC, Jacobus NV, Doku C, Tally
FP. Microbiology and antibiotic resistance in odonto-
genic infections. Ann Otol Rhinol Laryngol 1991;
154:S40 – 2.
[42] Greenburg RN, James RB, Marier RL, et al. Micro-
biologic and antibiotic aspects of infections in the
oral and maxillofacial region. J Oral Surg 1979;
37:873 – 84.
[43] Sands T, Pynn BR, Katsikeris N. Part two odontogenic
infections: microbiology, antibiotics and management.
Oral Health 1995;85:11 – 4.
[44] Lewis MAD, MacFarlane TW, McGowan DA. Quan-
titative bacteriology of acute dento-alveolar abscesses.
J Med Microbiol 1986;21:101 – 4.
[45] Jousimies-Somer HR, Summanen PH, Finegold SM.
14 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15
Bacteroides, porphyromonas, prevotella, fusobac-
terium, and other anaerobic gram-negative rods and
cocci. In: Murray PR, Baron EJ, Pfaller MA, et al,
editors. Manual of clinical microbiology. 7th edition.
Washington, DC: ASM Press; 1999. p. 690 – 711.
[46] Lowy FD. Staphylococcus aureus infections. N Engl J
Med 1998;339:520 – 32.
[47] Huebner J, Goldmann DA. Coagulase-negative staph-
ylococci: role as pathogens. Annu Rev Med 1999;50:
223 – 36.
[48] Haug RH, Assael LA. Infection in the maxillofacial
trauma patient. In: Topazian RG, Goldberg MH,
Hupp JR, editors. Oral and maxillofacial infections.
4th edition. Philadelphia: WB Saunders Co.; 2002.
p. 267 – 92.
[49] Bikle DD. Effects of alcohol abuse on bone. Compr
Ther 1988;14:16 – 20.
[50] Gavin LA. Perioperative management of the diabetic
patient. Endocrinol Metab Clin North Am 1992;21:
457 – 75.
[51] Abubaker AO, Rollert MK. Postoperative antibiotic
prophylaxis in mandibular fractures: a preliminary
randomized, double blind, and placebo-controlled
clinical study. J Oral Maxillofac Surg 2001;59:
1415 – 9.
[52] Helling T, Evans L, Fowler DL, et al. Infectious com-
plication in patients with severe head injury. J Trauma
1989;28:1575 – 7.
[53] Brook I. Microbiology of human and animal bite
wounds in children. Pediatr Infect Dis J 1987;6:29 – 32.
[54] Callaham M. Prophylactic antibiotics in common dog
bite wounds: a controlled study. Ann Emerg Med
1980;9:410 – 4.
[55] Callaham M. Treatment of common dog bites: in-
fection risk factors. J Am Coll Emerg Phys 1978;1:
83 – 7.
[56] Doan-Wiggins L. Animal bites and rabies. In: Rosen P,
Barkin RM, editors. Emergency medicine concepts and
clinical practice. 3rd edition. Philadelphia: Mosby-Year
Book; 1992. p. 864 – 5.
[57] Feder HM, Shanley JD, Barbera VA. Review of 50
patients hospitalized with animal bites. Pediatr Infect
Dis J 1987;6:24 – 8.
[58] Goldstein EJ, Richwald GA. Human and animal bite
wounds. Am Fam Physician 1987;36:101 – 9.
[59] Johnson D, Kuzmik M, Chorazy C. Aeromonas infec-
tion following facial injury. J Oral Maxillofac Surg
1986;44:563 – 5.
[60] Klein JD. Animal bite infections. Del Med J 1989;
61:17 – 20.
[61] Morgan JP, Haug RH, Murphy MT. Management of
facial dog bite injuries. J Oral Maxillofac Surg 1995;
53:435 – 41.
[62] Peel M. Dog-associated bacterial infection in humans:
isolates submitted to an Australian reference labora-
tory, 1981 – 1992. Pathology 1993;25:379 – 84.
[63] Ruskin JD, Laney T, Wendt SV, et al. Treatment of
mammalian bite wounds of the maxillofacial region.
J Oral Maxillofac Surg 1993;51:174 – 6.
[64] Tandberg D, Rusnak R. Mammalian bites. In: Schwartz
GR, Safar P, Stone JH, editors. Principles and practice
of emergency medicine. 2nd edition. Philadelphia: WB
Saunders Co.; 1986. p. 1618 – 25.
[65] Brennen SR, Rhodes KH, Peterson HA. Infection after
farm related injuries in children and adolescents. Am J
Dis Child 1990;144:710 – 3.
[66] Schwarz M. Myositis. In: Mandel G, Douglas R,
Bennett J, editors. Principles and practice of infec-
tious diseases. 3rd edition. New York: Churchill-Liv-
ingstone; 1990. p. 813.
[67] Burgess J, Wambaugh G, Koczarski M. Review-
ing cephalic tetanus. J Am Dent Assoc 1992;123:
67 – 70.
[68] Fackler M. Wound ballistics misconceptions. JAMA
1990;259:2730 – 6.
[69] Gwaltney JM. Sinusitis. In: Mandell GL, Bennett JE,
Dolin R, editors. Principles and practice of infectious
disease. 5th edition. Philadelphia: Churchill-Living-
stone; 2000. p. 676 – 86.
[70] Sandler NA, Johns FR, Braun TW. Advances in the
management of acute and chronic sinusitis. J Oral
Maxillofac Surg 1996;54:1005 – 13.
[71] Hamory BH, Sande MA, Sydnor A, et al. Etiology and
antimicrobial therapy of acute maxillary sinusitis.
J Infect Dis 1979;139:197 – 202.
[72] Sydnor A, Gwaltney J, Cacchetto DM, et al. Compar-
ative evaluation of Cefuroxime Axetil and Cefaclor for
treatment of acute bacterial maxillary sinusitis. Arch
Otolaryngol Head Neck Surg 1989;115:1430 – 3.
[73] Wald ER, Reilly JS, Casselbrant M, et al. Treatment of
acute maxillary sinusitis in childhood: a comparative
study of amoxicillin and cefaclor. J Pediatr 1984;104:
297 – 302.
[74] Winther B, Vickery CL, Gross CW, et al. Microbiology
of the maxillary sinus in adults with chronic sinus dis-
ease. Am J Med Sci 1998;316:13 – 20.
[75] Kopke RD, Jackson RL. Rhinitis. In: Bailey BJ, editor.
Head and neck surgery: otolaryngology. Philadelphia:
JB Lippincott; 1993. p. 269 – 89.
[76] Kibblewhite DJ, Cleland J, Mintz DR. Acute sphenoid
sinusitis: management strategies. J Otolaryngol 1988;
17:159 – 63.
[77] Lew D, Southwick FS, Montgomery WW, et al. Sphe-
noid sinusitis: a review of 30 cases. N Engl J Med
1983;309:1149 – 54.
[78] Middleton WG, Briant TDR, Fenton RS. Frontal si-
nusitis: a 10 year experience. J Otolaryngol 1985;14:
197 – 200.
[79] Swartz MN. Cellulitis and subcutaneous tissue in-
fections. In: Mandell GL, Bennett JE, Dolin R, ed-
itors. Principles and practice of infectious disease.
5 th edition. Philadelphia: Churchill-Livingstone;
2000. p. 1037 – 57.
[80] Topazian RG. Uncommon inflammatory conditions
and infections of the orofacial region. In: Topazian
MH, Goldberg MH, Hupp JR, editors. Oral and max-
illofacial infections. 4th edition. Philadelphia: WB Sa-
unders Co; 2002. p. 423 – 38.
 R.H. Haug / Oral Maxillofacial Surg Clin N Am 15 (2003) 1–15 15

[81] Butler T. Yersinia species, including plague. In: Man-
dell GL, Bennett JE, Dolin R, editors. Principles and
practice of infectious disease. 5th edition. Philadelphia:
Churchill-Livingstone; 2000. p. 2406 – 13.
[82] Slater LN, Welch DF. Bartonella species, including cat-
scratch disease. In: Mandell GL, Bennett JE, Dolin R,
editors. Principles and practice of infectious disease.
5 th edition. Philadelphia: Churchill-Livingstone;
2000. p. 2444 – 57.
[83] Tappero JW, Ashford DA, Perkins BA. Leptospira spe-
cies (leptospirosis). In: Mandell GL, Bennett JE, Dolin
R, editors. Principles and practice of infectious disease.
5th edition. Philadelphia: Churchill-Livingstone; 2000.
p. 2495 – 501.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 17 – 38

Antibiotic selection in head and neck infections

Thomas R. Flynn, DMDa,b,*, Leslie R. Halpern, DDS, MD, MPH, PhDa,b
a
Department of Oral and Maxillofacial Surgery, Harvard School of Dental Medicine,
188 Longwood Avenue, Boston, MA 02115, USA
b
Department of Oral and Maxillofacial Surgery, Massachusetts General Hospital, 55 Fruit Street, Boston, MA 02114, USA

Oral and maxillofacial surgeons see patients with
infections as part of their everyday practice. It is
imperative to understand the mechanisms of antimi-
crobial resistance, its potential problems, and the
means of overcoming it. This situation raises several
important questions with respect to antimicrobial
therapy for odontogenic infections:
1. Is there a problem of antibiotic resistance?
2. How does antibiotic resistance arise?
3. Is antibiotic resistance the fault of the bacteria
or the host or the result of treatment (ie, the
medical and surgical community)?
4. What can be done to remedy the problem?
The purpose of this article is to examine the
problem of antimicrobial resistance in the oral cavity
and make recommendations for antibiotic selection in
the treatment of head and neck infections.
Molecular biology of antibiotic resistance
Generally speaking, bacteria acquire antibiotic
resistance in one of four ways:
1. Alteration of a drug’s target site
2. Inability of a drug to reach its target
3. Inactivation of an antimicrobial agent
4. Active elimination of an antibiotic from the cell
* Corresponding author. Department of Oral and
Maxillofacial Surgery, Harvard School of Dental Medicine,
188 Longwood Avenue, Boston, MA 02115.
E-mail address: thomas_flynn@hsdm.harvard.edu
(T.R. Flynn).
The acquisition of antibiotic resistance genes by
bacteria allows such mechanisms to be implemented.
There are four specific mechanisms by which bacteria
acquire resistance genes:
1. Spontaneous mutation. This is the original
source for all antibiotic resistance, because
bacteria have maintained genes that encode
for resistance of naturally occurring anti-
biotics of other species. For example, the
DNA encoding of b-lactamases and penicillin-
binding proteins have several homologous
sequences [1].
2. Gene transfer. Bacteria can undergo conjuga-
tion with a transfer of genes as plasmids, which
are a composition of cytoplasmic loops of
DNA that encode for antibiotic resistance, and
transposons, which are able to insert them-
selves into the genome of the recipient cell. An
example of a plasmid-mediated genetic event is
acquisition of the ability to produce b-lacta-
mase by some species.
3. Bacteriophages. Viruses infect bacteria and
can insert genetic material and take control
of the host’s genetic and metabolic machin-
ery, which may encode for antibiotic re-
sistance mechanisms.
4. Mosaic genes. Bacteria can absorb directly
the fragments of the virally altered genome of
dead members of related species to form a
‘‘mosaic genome’’ of genetic material from
varying sources. This type of gene derivation
is responsible for the non – b-lactamase pen-
icillin resistance in Streptococcus pneumoniae
and meningococci and ampicillin resistance in
Haemophilus influenzae and gonococci [1].

1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 8 2 - 1
18 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38
Antibiotic resistance mechanisms
Once the genetic machinery is in place, bacteria
exert antibiotic resistance by various pathways that
are broadly classified in four ways.
Drug inactivation or modification. The destruc-
tion or inactivation of the antimicrobial agent is
accomplished by the induction of specific drug-inac-
tivating enzymes, such as those that inhibit b-lactams
or aminoglycosides. Numerous gram-positive and
gram-negative bacteria, such as Staphylococcus au-
reus, Enterococcus faecium, Escherichia coli, Pseu-
domonas aeruginosa, H. influenzae, Bacteroides, and
many strains of Prevotella have this capability.
Another method used by bacteria to withstand anti-
microbial attack is the ability to synthesize neutral-
izing enzymes. The best examples are penicillinase
and the methylation of erythromycin and clindamy-
cin. Other antibiotics that are neutralized include
vancomycin, sulfonamides, aminoglycosides and
rifampin. Bacterial organisms with this capability
include S. pneumoniae, S. aureus, Clostridium per-
fringens, Bacteroides fragilis, Campylobacter spe-
cies, and Neisseria gonorrhoeae.
Alteration of microbial membrane permeability.
Alterations in membrane permeability can cause
decreased uptake or increased efflux of the antibiotic.
The types of antibiotics most often affected by this
mechanism are the b-lactams, quinolones, tetracy-
clines, erythromycin, and the aminoglycosides. The
gram-negative rods E. coli, P. aeruginosa, and Sal-
monella typhimurium also have this capability. Porins
within the transmembrane protein matrix are specific
for various antibiotics, and the loss of a specific porin
confers resistance. Lack of the D2 porin, for exam-
ple, confers imipenem resistance in P. aeruginosa.
Increased efflux of the antibiotic before lethal damage
occurs is seen in the Enterobacteriae with the mar,
norA, and tetA genes, which convey resistance by
pumping tetracycline out of the cells. E. coli and
Staphylococcus epidermidis also can resist tetracy-
clines, macrolides and quinolones by this mechanism
[1,2].
Alteration of target site. Enzymes responsible for
cell wall synthesis, the transpeptidases, can be altered
slightly to produce less affinity for penicillins. These
altered penicillin-binding proteins are most often seen
in S. aureus and S. pneumoniae [3].
Alteration in the concentration of drug target
receptors. Many of the gram-negative rods (ie, E.
coli and Proteus, Enterobacter, and Klebsiella spe-
cies) have the ability to alter the number of drug
receptors that bind antibiotics. The sulfonamide fam-
ily is affected by such a mechanism.
Strategies in the prevention of antibiotic resistance
Extending surgical prophylaxis beyond 48 hours
and inappropriately low dosing that encourages sub-
populations of organisms to survive in increasing
concentrations of antibiotics can select for resistant
bacteria [3]. Although culture and sensitivity studies
are crucial and should not preclude empiric therapy
when warranted, there is also the risk that the latter
can produce bacterial resistance. A case series to
examine the bacteriology of dentoalveolar abscesses
in patients who received empiric antibiotic therapy
suggested that the polymicrobial nature of the abscess
and the administration of empiric therapy with ampi-
cillin or cephalosporins often results in resistant
strains [4]. The predominant species were anaerobic
(ie, Prevotella and Peptostreptococcus species, both
resistant to the therapy initially given). Kuriyama et al
[5] examined the relationship between past adminis-
tration of b-lactamase antibiotics and an increase in
b-lactamase – producing bacteria in patients with
odontogenic infections. The algorithm of treatment
derived from their study is a course of b-lactamase
antibiotics for 1 to 2 days, but if the infection is
unresolved by 3 days or more, one should assume
the presence of b-lactamase – producing organisms,
and treatment should involve a penicillinase-stable
b-lactam or a non – b-lactam antibiotic. No definitive
studies with large sample sizes clearly define ways to
manage antibiotic resistance in odontogenic infec-
tions, however.
The question of whether antibiotic resistance in
patients with odontogenic infections who need hos-
pitalization is caused by the therapeutic modality
given, the characteristics of the patient population,
or the ability to isolate and characterize more
carefully the vector of disease is paramount because
of the possibility that the increased incidence of
antibiotic-resistant strains is an unavoidable direct
effect of therapy. Retrospective studies that com-
pared populations decades apart have shown that
although no clinically significant differences exist
between cohorts examined, there are differences
in types of microorganisms in terms of their no-
menclature [6,7]. Flynn et al [8] performed a
prospective study of 34 hospitalized patients with
odontogenic infections and found a 26% rate of
clinical failure with penicillin therapy and a 60%
rate of penicillin resistance.
This finding is exemplified by data on treatment
of upper respiratory tract infections. In a study of
children with pharyngitis, Brook [9] found a 9%
incidence of penicillin resistance in throat swab
cultures at the initiation of treatment. After 1 week
 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38
of penicillin therapy, 46% of the subjects and 45% of
the subjects’ parents and siblings harbored resistant
strains. The number declined to 27% in the subjects
over the ensuing 3 months. Several hospitals have
substituted the cephalosporins for penicillin/b-lacta-
mase inhibitor combinations, with or without an
aminoglycoside, which in some cases has resulted
in dramatic recovery of antibiotic susceptibility rates
among pathogens such as Enterobacter cloacae,
Klebsiella pneumoniae, P. aeruginosa, and Clos-
tridium difficile [1].
Issues in antibiotic selection
The selection of an appropriate antibiotic for a
given case can be complex, but usually it is a
straightforward process. The factors that must be
considered can be categorized into host-specific and
pharmacologic factors.
Host factors in antibiotic selection
Usual pathogens
The type of infection that presents can be char-
acterized by cause and location, and each has its
own characteristic flora. Odontogenic infections are
generally characterized by a combination of faculta-
tive streptococci and oral anaerobes. Within the
viridans group of facultative streptococci, the Strep-
tococcus milleri group, which consists of S. angino-
sus, S. intermedius, and S. constellatus, is most
frequently associated with orofacial cellulitis and
abscess. This is fortunate because only approxi-
mately 3% of the strains of these species are resistant
to the penicillins. On the other hand, other members
of the viridans streptococci, such as Streptococcus
mitis, Streptococcus sanguis, and Streptococcus sal-
ivarius, are more frequently found in endocarditis,
and they can be highly penicillin resistant—up to
58% in one study [10].
Among the anaerobes, anaerobic peptostreptococci
and members of the genera Prevotella and Porphyro-
monas predominate. Although the peptostreptococci
remain penicillin sensitive, approximately 25% of
strains of Prevotella and Porphyromonas are penicillin
resistant [8].
The penicillin-sensitive streptococci predominate
during the first 3 days of clinical symptoms, and the
more resistant gram-negative obligate anaerobes
appear in significant numbers thereafter. This fact
suggests the selection of the penicillins over other
19
antibiotics in early cases. Another factor is the
severity of the odontogenic infection. Flynn et al
[8] found a clinical failure rate of 26% for penicillin
in hospitalized cases. On the other hand, little or no
difference was found between the effectiveness of
penicillin and various other antibiotics in outpatient
odontogenic infections [11 – 14].
The clinician must keep in mind the occasional
pathogen that is resistant to the usual empiric anti-
biotic of choice. In odontogenic infections and dog
and cat bites, Eikenella corrodens is fairly resistant to
the penicillins and completely resistant to clindamy-
cin. The fluoroquinolones have become the antibiotic
of choice for this pathogen. E. corrodens should be
considered a possible pathogen in treatment failure of
odontogenic infections and routinely in animal bite
wounds [15]. The usual flora of various types of head
and neck infections are listed in Table 1.
Allergy or intolerance
A history of antibiotic allergy is usually readily
obtained from the conscious patient or, alternatively,
from the family. Penicillin allergy is common, and
macrolide (erythromycin family) intolerance and drug
interactions are frequent. The choice of clindamycin,
metronidazole, or newer antibiotics may be prudent
when anamnestic information is unavailable.
The penicillins are the antibiotics most frequently
prescribed for infections in the oral cavity. It is not
surprising that their use is associated with hypersen-
sitivity reactions. Between 1% and 10% of patients
who initially take penicillin develop an allergic reac-
tion, and persons who do not develop a reaction have
less than a 1% chance of developing an allergy with
reexposure [16]. It is judicious to clarify whether the
person has a true allergy to penicillin. Cross-sectional
studies of penicillin allergy indicate that in many
hospital chartings of penicillin allergy, subsequent
skin testing proved that more than 60% of pa-
tients were not allergic to either penicillin or other
b-lactams, which warrants more careful vigilance by
doctors who are recording medical histories and
allergies of their patients [17,18]. Fortunately, hyper-
sensitivity reaction to clindamycin, often substituted
in penicillin-allergic patients, is a rare event.
All clinicians should be aware of the potential for
cross-allergy between the penicillins and other mem-
bers of the b-lactam group. Approximately 10% to
15% of penicillin-allergic patients are also sensitive
to the cephalosporins. The cross-allergic group tends
to include persons who have had an anaphylactoid
reaction to the penicillins. The cephalosporins should
be avoided in these patients.
20 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38

Table 1
Major pathogens of head and neck infections
Type of infection Microorganisms
Odontogenic cellulitis/abscess Streptococcus milleri group
Peptostreptococci
Prevotella and Porphyromonas
Fusobacteria
Rhinosinusitis Acute Streptococcus pneumoniae
Haemophilus influenzae
Head and neck anaerobes (peptostreptococci, Prevotella,
Porphyromonas, fusobacteria)
Group A b-hemolytic streptococci
Staphylococcus aureus
Moraxella catarrhalis
Viruses
Chronic Head and neck anaerobes
Fungal Aspergillus
Rhizopus sp. (mucor)
Nosocomial Enterobacteriaceae (especially Pseudomonas,
(especially if intubated) Acinetobacter, Escherichia coli)
S. aureus
Yeasts (Candida species)
Osteomyelitis of the jaws Acute Odontogenic flora
S. aureus and skin flora in trauma
Salmonella in sickle cell disease
Chronic Actinomyces species
Necrotizing fasciitis Group A b-hemolytic streptococci
Regional flora (oral and sinus pathogens
in head and neck)
Fungal Mucosal or disseminated Candida species
Soft tissue Histoplasma species
Blastomyces species
Sinus Aspergillus
Rhizopus (mucor)

The newer b-lactam antibiotics, the monobactams
(aztreonam) and the carbapenems (imipenem and
meropenem), have much less frequent cross-sensitiv-
ity with the penicillin group. A history of adverse
reaction or intolerance of an antibiotic, such as
phototoxicity with the tetracyclines or antibiotic-
associated colitis with clindamycin, would preclude
its subsequent use unless strongly indicated.
Immune system compromise
Because the immunocompromised patient is less
able to kill invading pathogens by host resistance
mechanisms, a bactericidal rather than bacteriostatic
antibiotic should be selected whenever possible. This
stratagem should result in a more rapid clinical
response. The bactericidal antibiotics generally inter-
fere with either cell wall synthesis, which causes
lysis, or with nucleic acid synthesis, which arrests
vital processes. The bacteriostatic antibiotics interfere
with protein synthesis, arresting growth and mul-
tiplication. Some antibiotics, such as clindamycin,
seem to be bacteriostatic at lower doses and bacte-
ricidal at higher doses.
HIV-infected individuals seem to be able to han-
dle oral bacterial infections almost as well as non-
infected persons. This ability is probably caused by
the antibody-mediated immunity provided by the
B-lymphocytes, which is largely responsible for com-
bating the extracellular bacterial pathogens of most
head and neck infections. Resistance to these com-
mon infections remains fairly robust until the terminal
stages of AIDS, when all types of lymphocytes are
severely depleted. On the other hand, fungal and viral
infections, which are resisted by cell-mediated
immunity (T cells), are prevalent in poorly controlled
HIV-infected individuals.
 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38 21

Table 2 They are discussed in the section on adverse anti-

Bactericidal and bacteriostatic antibiotics biotic reactions.
Bactericidal Bacteriostatic
b-lactams Macrolides
penicillins erythromycin Pharmacologic factors in antibiotic selection
cephalosporins clarithromycin
carbapenems azithromycin Antimicrobial spectrum
monobactams Clindamycin
Aminoglycosides Tetracyclines The most important pharmacologic consideration
Vancomycin Sulfa antibiotics in antibiotic selection is whether it is effective against
Metronidazole the likely pathogens. Table 3 describes the general
Fluoroquinolones
spectrum of selected antibiotics. Table 4 lists the
bacteria and fungi most likely to be encountered
and the antibiotics of choice for those pathogens.
Table 2 lists common antibiotics by their ability to The antibiotics effective against the highly resistant
kill bacteria or merely suppress their growth. organisms are also included in Table 4. Table 5 lists
the antibiotics to which selected highly resistant
organisms have become resistant. These data, among
Previous antibiotic therapy
others, are used in constructing the recommendations
for empiric antibiotics of choice for various head and
All antibiotic therapy inherently selects for re-
neck infections, and Tables 4 and 5 especially can be
sistant organisms. Studies of patients who are cur-
used in selecting an appropriate antibiotic for organ-
rently taking or recently have taken antibiotics
isms identified by culture, for which sensitivity data
consistently yield a higher incidence and proportion
may not be available.
of organisms resistant to that antibiotic [10,19]. On
the other hand, these effects persist for a consid-
Tissue distribution of antibiotics
erable time after antibiotic therapy and may be
permanent [19,20].
Although abscess cavities are not vascular, some
The previous use of different antibiotics during
penetration of antibiotics into these spaces does
the course of an acute infection definitely clouds the
occur. The antibiotic that best penetrates an abscess
bacteriologic picture. In this situation, the clinician
is clindamycin; the abscess concentration of clinda-
has the choice of changing the current antibiotic or
mycin reaches 33% of the serum level [22]. This fact
increasing its dose, perhaps by using the parenteral
may partially explain the usefulness of clindamycin
route. With penicillins V (oral) and G (intravenous),
in odontogenic infections.
peak serum blood levels are 5.6 mg/mL and 20 mg/mL,
Bone penetration of antibiotics is an important
respectively. The dramatic increase in efficacy
consideration, especially in osteomyelitis. The
afforded by the parenteral route of administration
antibiotics that best penetrate or even accumulate
may be more advantageous than changing to another
in bone are the tetracyclines, clindamycin, and
antibiotic that is less effective than the penicillins.
the fluoroquinolones.
The penicillin resistance rate of the endocarditis-
Cerebrospinal fluid penetration, or the ability of
associated viridans streptococci (S. mitis, S. sanguis,
an antibiotic to cross the blood-brain barrier, is
and S. salivarius) is high—up to 58% [21] in persons
paramount in the treatment of infections that threaten
with a history of prior endocarditis. Clindamycin
the central nervous system, as in actual or impending
resistance of these bacteria in such patients remains
cavernous sinus thrombosis. The antibiotics that can
low. In patients with a history of endocarditis, it may
attain therapeutic levels in cerebrospinal fluid when
be advisable to use clindamycin rather than amox-
the meninges are inflamed are listed in Table 6. The
icillin for endocarditis prophylaxis before oral proce-
antibiotics that do not penetrate the cerebrospinal
dures. This approach, however, has not been tested in
fluid well are clindamycin, the macrolides (including
a clinical study.
clarithromycin and azithromycin), cefazolin, and
most other cephalosporins (except those listed in
Special conditions Table 6), aminoglycosides, amphotericin, itracona-
zole, ethambutol, and saquinavir.
Certain temporary host conditions may affect Penicillin G in high doses reaches 5% to 10% of
antibiotic selection, such as childhood and pregnancy. the serum concentration in the cerebrospinal fluid
22 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38

Table 3
Spectrum of selected antibiotics
Antibiotic category Antibiotic Susceptible organisms
Natural penicillins Penicillin G and V Viridans streptococci
Oral anaerobes
Actinomyces sp. (penicillin G only)
Pasteurella multocida
Semisynthetic penicillins Ampicillin As with natural penicillins, plus enterococci
Amoxicillin Actinomyces
b-lactam/b-lactamase inhibitors Amoxicillin/clavulanate As with amoxicillin, plus
Ampicillin/sulbactam S. aureus, not MRSA
S. epidermidis, not MRSE
H. influenzae
M. catarrhalis
Klebsiella species
E. coli
Bacteroides fragilis
Penicillinase-resistant penicillins Oxacillin S. aureus, not MRSA
Dicloxacillin S. epidermidis, not MRSE
Antipseudomonal penicillins Ticarcillin/clavulanate As with natural penicillins, plus
Piperacillin/tazobactam S. aureus, not MRSA
S. epidermidis, not MRSE
H. influenzae
M. catarrhalis
Klebsiella species
E. coli
Bacteroides fragilis
Enterobacteriaceae (most)
Pseudomonas aeruginosa
Carbapenems Imipenem As with antipseudomonal penicillins, plus
Meropenem Actinomyces (imipenem)
Ertapenem
Monobactam Aztreonam Enterobacteriaceae, except Salmonella
(no data) and Acinetobacter (resistant)
Cephalosporins First generation Streptococci
Cephalexin S. aureus, not MRSA
Cefazolin H. influenzae
Klebsiella
E. coli
Second generation As with first generation, plus
Cefaclor M. catarrhalis (cefuroxime)
Cefuroxime Oral anaerobes
Cefoxitin B. fragilis (cefoxitin)
Third generation As with first generation, plus
Cefotaxime M. catarrhalis
Ceftriaxone Oral anaerobes
Actinomyces (ceftriaxone)
Macrolides Erythromycin Streptococci
Clarithromycin Actinomyces
Azithromycin Peptostreptococci (azithromycin)
Clindamycin Clindamycin Streptococci
Oral anaerobes
Actinomyces
S. aureus, not MRSA
Metronidazole Obligate anaerobes
Fluoroquinolones Ciprofloxacin S. aureus, not MRSA
Enterobacteriaceae (most)

(continued on next page)

 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38
Table 3 (continued )
Antibiotic category Antibiotic
Moxifloxacin
Aminoglycosides Gentamicin
Tobramycin
Glycopeptides Vancomycin
Teicoplanin
Oxazolidinones Linezolid
Pristinamycins Quinupristin/dalfopristin
Ketolides Telithromycin
Abbreviations: MRSA, methicillin-resistant S. aureus; MRSE, methicillin-resistant S. epidermidis; VISA, vancomycin-inter-
mediate S. aureus; VRSE, vancomycin-resistant S. epidermidis.
when the meninges are inflamed. In odontogenic
infections that threaten the central nervous system,
the addition of metronidazole (30% – 100% penetra-
tion) to ampicillin (13% – 14% penetration) is more
efficacious than using penicillin G alone [15].
Pharmacokinetics
The effectiveness of some antibiotics, such as the
fluoroquinolones and aminoglycosides, is concentra-
tion dependent, whereas with other antibiotics, such
as the b-lactams and vancomycin, it is time depen-
dent. In concentration-dependent antibiotics, efficacy
is determined by the ratio of the serum concentration
of the antibiotic to the minimum inhibitory concen-
tration (MIC), which is the concentration of the
antibiotic required to kill a given percentage of the
strains of a particular species, usually 50% or 90%. In
time-dependent antibiotics, it is necessary to maintain
the serum concentration above the MIC for at least
40% of the dosage interval.
It is necessary with time-dependent antibiotics to
know the serum elimination half-life (t/2) of the
antibiotic to determine its proper dosage interval.
23
Susceptible organisms
Streptococci
Oral anaerobes
S. aureus, not MRSA
Actinomyces
B. fragilis
Enterobacteriaceae (most)
S. aureus, not MRSA
Enterococci (gentamicin synergistic with ampicillin)
Enterobacteriaceae (many)
Pseudomonas
Streptococci
S. aureus, including MRSA
S. epidermidis, including MRSE (vancomycin)
Streptococci
Staphylococci, including VISA, VRSE, MRSA, MRSE
Peptostreptococci
Enterococci, including VRE
Streptococci
Staphylococci, including VISA, VRSE, MRSA, MRSE
Legionella
Streptococci
S. aureus (not MRSA?)
H. influenzae
M. catarrhalis
Legionella
For example, the t/2 of penicillin G is 0.5 hours.
During each half hour, 50% of the remaining penicil-
lin is eliminated from the serum. By five half-lives, or
2.5 hours, only approximately 3% of the peak serum
level of penicillin remains. Because the MIC-90 of the
viridans streptococci (the concentration that kills 90%
of the strains) is 0.2 mg/mL and because the peak
serum level achieved with 2 million U of intravenous
penicillin G is 20 mg/mL, the serum concentration of
penicillin after 4 hours (eight half-lives) is approx-
imately 0.15 mg/mL. The serum level will have fallen
below the MIC-90 roughly for only the last 15% of the
dosage interval. Intravenous penicillin G, 2 million U
every 4 hours, should be highly effective against the
viridans group of streptococci.
Using the same analysis, the peak blood level
achieved with amoxicillin, 500 mg orally, is 7.5 mg/mL,
and its t/2 is 1.2 hours. The MIC-90 for the viridans
streptococci is 2 mg/mL for amoxicillin. Using an
8-hour dosage interval, the remaining serum concen-
tration of amoxicillin should have fallen below the
MIC-90 of the viridans streptococci at approximately
2.5 hours, which is only 31% of the dosage interval.
Oral amoxicillin therapy may not kill 90% of all the
24 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38

Table 4
Antibiotics of choice for head and neck pathogens
Pathogen Type First choice antibiotics Alternative antibiotics
Actinomyces +, R, A Penicillin G or ampicillin Doxycycline
Clindamycin
Erythromycin
Bacteroides fragilis
, R, AN Metronidazole Clindamycin
Cefoxitin, not cefotetan (DOT)
Ampicillin/sulbactam
Clostridium species +, R, AN Penicillin G F clindamycin Metronidazole
(except C. difficile) Doxycycline
Cephalosporin (1st)a
Clostridium difficile +, R, AN Metronidazole p.o. Vancomycin p.o.
Bacitracin p.o.
Eikenella corrodens
, R, A Penicillin G or V Fluoroquinolones
Amoxicillin TMP/SMX (avoid
Amoxicillin/clavulanate clindamycin)
Enterococcus faecalis +, C, F Ampicillin F gentamicin Vancomycin
(group D streptococcus) (for endocarditis or meningitis Ampicillin/sulbactam
Linezolid
Enterococcus faecium (group D +, C, F Linezolid + quinupristin/dalfopristin F Teicoplanin + aminoglycoside
streptococcus: b-lactamase +, choramphenicol F doxycycline (van B)
aminoglycoside and For some strains: no effective
vancomycin resistant) regimen (I.D. consultation)
Escherichia coli
, R, A Ticarcillin/clavulanate Meropenem for central nervous system
Cephalosporins Aztreonam
Imipenem TMP/SMX
Fluoroquinolones Tobramycin
Fusobacterium species
, R, AN Penicillin G or V Metronidazole
Clindamycin
Haemophilus influenzae
, R, F Amoxicillin/clavulanate Cefotaxime (if life threatening)
(b-lactamase positive) Cefaclor Ciprofloxacin
Azithro/clarithromycin TMP/SMX
Klebsiella pneumoniae
, R, A Cephalosporin (3rd)* Tobramycin
Fluoroquinolones Ticarcillin/clavulanate
Imipenem/cilastatin
Klebsiella pneumoniae
, R, A Imipenem/cilastatin Meropenem
(producing extended spectrum Fluoroquinolones
b-lactamases: ESBLs)
Pasteurella multocida
, R, A Penicillin G Doxycline
(eg, dog and cat bites) Amoxicillin/Clavulanate Cephalosporin (2nd)a
TMP/SMX
Peptostreptococcus +, C, AN Penicillin G or V Clindamycin
(and former Peptococcus) Doxycline
Vancomycin
Black pigmented oral
, R, AN Clindamycin PCN + metronidazole
anaerobes (Prevotella and Amoxicillin
Porphyromonas) Cefotetan
Proteus vulgaris (indole +)
, R, A Cephalosporin (3rd) Tobramycin
Fluoroquinolones Imipenem
Ticarcillin/clavulanate
Pseudomonas aeruginosa
, R, A Ciprofloxacin Aztreonam + ceftazidime
Tobramycin Piperacillin + tobramycin
Cefepime + tobramycin
Salmonella typhi
, R, A Fluoroquinolones Chloramphenicol
Ceftriaxone Amoxicillin
TMP/SMX
(continued on next page)
 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38 25

Table 4 (continued )
Pathogen Type First choice antibiotics Alternative antibiotics
Serratia marcescens
, R, A Cephalosporin (3rd) Gentamicin
Imipenem Aztreonam
Meropenem
Fluoroquinolones
Shigella
, R, A Fluoroquinolones TMP/SMX + ampicillin
Azithromycin
Staphylococcus aureus +, C, A Penicillinase-resistant Cephalosporin (1st)a
(methicillin sensitive) penicillin Vancomycin
Clindamycin
Staphylococcus aureus +, C, A Vancomycin Teicoplanin
(methicillin resistant) Quinupristin-dalfopristin
TMP/SMX (some strains)
Linezolid
Staphylococcus aureus +, C, A No effective regimen Quinupristin/dalfopristin
(methicillin and Try vancomycin F rifampin Linezolid
vanco mycin resistant)
Staphylococcus epidermidis +, C, A Vancomycin (+ rifampin Quinupristin/dalfopristin
(methicillin resistant) + gentamicin
for prosthetic valve endocarditis)
Staphylococcus epidermidis +, C, A Quinupristin/dalfopristin Vancomycin (high dose)
(methicillin and Linezolid New fluoroquinolones?b
glycopeptide resistant) (rapid resistance a problem)
Streptococcus pneumoniae +, C, A Penicillin G or V Cefuroxime, cefipime
(Pneumococcus) Ceftriaxone Imipenem
(penicillin sensitive) Amoxicillin New fluoroquinolonesb
Streptococcus pneumoniae +, C, A Vancomycin + Rifampin Clindamycin
(Pneumococcus) (multiantibiotic New fluoroquinolones (in vitro)
resistant, including high-level
penicillin, erythromycin,
tetracycline, chloramphenicol,
and TMP/SMX)
Streptococcus pyogenes +, C, A Penicillin G or V (+ gentamicin Cephalosporin (1st)a
(b-hemolytic streptococcus) if serious group B infection) Erythromycin
Streptococcus viridans +, C, A Penicillin G or V Cephalosporin (1st)a
(a-hemolytic streptococcus) Macrolides
Fungal organisms
Blastomyces Fungus Amphotericin B Itraconazole (if surface)
(for systemic cases) Fluconazole (if surface)
Candida Fungus Fluconazole Nystatin (if surface)
Amphotericin B Clotrimazole (if surface)
(for systemic cases) Ketoconazole (if surface)
Itraconazole (if surface)
Coccidioides immitis Fungus Itraconazole Fluconazole
Amphotericin B
Histoplasma Fungus Amphotericin B (for systemic Itraconazole (immunocompetent)
or immunocompromised cases) Itraconazole (immunocompromised)
Mucormyces Fungus Amphotericin B Control underlying systemic disease
Abbreviations: A, aerobe; AN, anaerobe; C, coccus; DOT, distasonis, ovatus, and thetaiotamicron group of B. fragilis species;
F, facultative; PCN, penicillin; R, rod; TMP-SMX, trimethoprim-sulfamethoxazole.
Data from Gilbert DN, Moellering RC Jr, Sande MA. The Sanford guide to antimicrobial therapy 2002. 32nd edition. New Hyde
Park (VT): Antimicrobial Therapy Inc.; 2002.
+ = gram positive.

= gram negative.
a
Number in parentheses after cephalosporins refers to generations within the cephalosporin family.
b
New fluoroquinoles are gati-, gemi-, lero-, moxi-, sparfloxacin.
26 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38

strains of the viridans streptococci. Fortunately for oral

and maxillofacial surgeons, the Streptococcus milleri
group associated with odontogenic infections is highly
sensitive to the penicillins, whereas the endocarditis-
Table 5 associated strains are less so.
Highly resistant organisms and the antibiotics to which they The pharmacokinetics of the clinically available
are resistant antibiotics have been determined during drug devel-
Organism Resistant to opment. It is incumbent on the clinician to prescribe
Acinetobacter baumanii Penicillins antibiotics within the accepted ranges for dose
Third generation and interval.
cephalosporins Once-daily dosing for the aminoglycosides as a
Antipseudomonal means of reducing their ototoxicity and nephrotox-
aminoglycosides icity recently has been evaluated in a systematic
Fluoroquinolones review [23]. The available well-designed studies
Imipenem indicate that this practice results in a modest increase
Enterococcus faecalis Glycopeptides
in therapeutic advantage and possibly a decrease in
b-lactamase negative Streptomycin
Gentamicin
toxicity. The cost saving of once-daily intravenous
Enterococcus faecalis All b-lactams dosing makes this approach appealing. Caution is
b-lactamase positive Glycopeptides advised in patients with limited volumes of fluid
Aminoglycosides distribution, however.
Enterococcus faecium Glycopeptides
b-lactamase negative Streptomycin Adverse reactions
Gentamicin
Enterococcus faecium all b-lactams The adverse reactions and toxicities of the anti-
b-lactamase positive Glycopeptides biotics commonly used in head and neck infections
Aminoglycosides
are generally mild and uncommon. Table 7 lists the
Klebsiella pneumoniae Penicillins
ESBL positive Third generation
major serious adverse reactions of the commonly
cephalosporins used antibiotics. The clinician especially should note
Aztreonam allergic reactions to the penicillins and cephalospo-
Pseudomonas Penicillins rins, gastrointestinal intolerance of the erythromycins,
aeruginosa Cephalosporins nephrotoxicity and ototoxicity of the amino-
Carbapenems glycosides, and antibiotic-associated colitis with the
Staphylococcus Methicillin b-lactam/b-lactamase inhibitor combinations (eg,
aureus MRSA Augmentin, Unasyn), antipseudomonal penicillins
S. aureus VISA or GISA Methicillin (eg, ticarcillin, piperacillin), cephalosporins, and clin-
Vancomycin only
damycin, among others.
Vancomycin and
teicoplanin (both
available glycopeptides) Special conditions
Staphylococcus Methicillin
epidermidis MRSE Antibiotics that should be avoided in children
S. epidermidis VRMRSE Methicillin include the tetracyclines (under the age of 8), because
Glycopeptides of permanent intrinsic dental staining, and the fluo-
Streptococcus pneumoniae Penicillin G roquinolones, because of chondrotoxicity in growing
penicillin intermediate cartilage. Among the carbapenems, imipenem is not
or resistant recommended because of the risk of seizures. Mer-
S. pneumoniae Penicillins
openem is an acceptable alternative.
multi-antibiotic resistant Cephalosporins
Aztreonam
The use of antibiotics in pregnancy almost always
involves an evaluation of risk versus benefit. The
Abbreviations: ESBL, extended-spectrum b-lactamase;
antibiotics that must be avoided in pregnancy include
GISA, glycopeptide-intermediate S. aureus; MRSE, methi-
cillin-resistant S. epidermidis; VRMRSE, vancomycin- the antimycobacterial agent, thalidomide, and the
resistant methicillin-resistant S. epidermidis. antiparasitic agent, quinine, for which the risk clearly
Data from Gilbert DN, Moellering RC Jr, Sande MA. The outweighs the benefit.
Sanford guide to antimicrobial therapy. 32nd edition. Hyde Table 8 lists the pregnancy risk categories of
Park (VT): Antimicrobial Therapy, Inc.; 2002. selected antibiotics.
 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38 27

Table 6 Antibiotic drug interactions

Selected antibiotics and the blood-brain barrier
Cerebrospinal fluid Antibiotic
Two important categories of antibiotic drug inter-
action are interference with the effectiveness of oral
Therapeutic Penicillins contraceptives and interference with the metabolism
levels achieved ampicillin
of drugs, which involves the cytochrome P450 sys-
nafcillin
penicillin G, high dose
tem. These and other selected antibiotic drug inter-
ticarcillina actions are listed in Table 9.
piperacillina Antibiotic interference with the effectiveness of
Cephalosporins oral contraceptive pills remains a controversial topic.
ceftazidime The only antibiotic that has been shown conclusively
cefuroxime to interfere with oral contraception is rifampin. The
ceftriaxone evidence that implicates ampicillin, amoxicillin, dap-
Carbapenem sone, trimethoprim/sulfamethoxazole, and the antivi-
meropenemb ral protease inhibitors is less strong. It is important to
Fluoroquinolones
note that antibiotics do not interfere with injectable or
levofloxacin
ciprofloxacinc
implantable contraceptives. Only oral contraceptives
Other antibiotics are affected [24].
metronidazole A possible mechanism for this interaction stems
trimethoprim/ from efforts to decrease the adverse effects, such as
sulfamethoxazoled thromboembolism and activation of uterine and
vancomycine breast carcinomas associated with older contracep-
Antifungal drugs tive formulations that contained higher estrogen
fluconazole doses. Currently, oral contraceptive preparations
flucytosine have minimally effective estrogen doses, and the
Antiviral drugs
serum level of the estrogen is supported by enter-
acyclovir
foscarnet
ohepatic recirculation. In this process, the liver
ganciclovir conjugates absorbed estrogen with glucuronide, and
zidovudine the estrogen-glucuronide complex is excreted in the
Therapeutic levels Cephalosporins bile. In turn, the gut flora breaks the estrogen-
not achieved cefazolin glucuronide bond, which allows the pure estrogen
cephalexin molecule to be reabsorbed by the gut, thus support-
Aminoglycosides ing the serum estrogen level. If an antibiotic kills
Macrolides enough of the gut flora, then the conjugated estrogen
erythromycin is not broken down, and the estrogen-glucuronide
clarithromycin
complex stays in the intestine until it is excreted.
azithromycin
Clindamycin
The serum estrogen level falls, which results in
Antifungal drugs breakthrough menstrual bleeding or ovulation and
amphotericin unwanted pregnancy.
itraconazole The cytochrome P450 system is a complex set
Antiviral drugs of drug-metabolizing enzymes that is responsible for
saquinavir the breakdown of many classes of drugs. Enzymes
zidovudine within this system include CYP3A4, CYP2C19, and
Data from Gilbert DN, Moellering RC Jr, Sande MA. The CYP2D6. Drugs that share this metabolic pathway
Sanford guide to antimicrobial therapy. 32nd edition. Hyde may interact. The metabolism of one or the other
Park (VT): Antimicrobial Therapy, Inc.; 2002. may be either increased or decreased as a result.
a
Levels effective for P. aeruginosa and coliforms may The adverse affect is usually caused by an increased
not be reached.
b effect of the drug whose metabolism is inhibited,
Imipenem is avoided in meningitis because of seizure
but in some of the most serious cases, life-threat-
potential. Meropenem is preferred.
c
Does not reach adequate cerebrospinal fluid levels ening or fatal cardiac dysrhythmias, such as ven-
for streptococci. tricular fibrillation and torsade des pointes, have
d
Not adequately effective against Neisseria species occurred. The most significant interactions invol-
and coliforms. ving the cytochrome P450 system are included in
e
High doses are needed for resistant streptococci. Table 9.
 28
Table 7

T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38
Major adverse reactions of selected antibiotics
Ampicillin,
Penicillin G amoxicillin F Ticarcillin F Gentamicin, Cephalexin,
Adverse reactions and V clavulanate clavulanate Impenem Meropenem tobramycin cefazolin Cefuroxime Cefoxitin Cefotaxime Cefaclor
Local, phlebitis +
Hypersensitivity
Rash + + + + +
Photosensitivity
Anaphylaxis +
Serum sickness +
Anemia +
Nausea, vomiting
Diarrhea +
Antibiotic-associated colitis (AAC) +
Renal: z BUN, creatinine +
Headache
Seizures +
Hypotension
Ototoxicity +
Vestibular dysfunction +
Alcohol interaction
‘‘Red man’’ flushing
Drug interactions +
Pregnancy risk C or D +
Data from Gilbert DN, Moellering RC Jr, Sande MA. The Sanford guide to antimicrobial therapy. 32nd edition. Hyde Park (VT): Antimicrobial Therapy, Inc.; 2002.
 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38
Table 7 (continued )
Clarithromycin, Tetracycline,
Erythromycin azithromycin Clindamycin Metronidazole Ciprofloxacin Moxifloxacin Vancomycin doxycycline Linezolid Telithromycin
Local, phlebitis +
Hypersensitivity +
Rash
Photosensitivity +
Anaphylaxis
Serum sickness
Anemia +
Nausea, vomiting + + + +
Diarrhea + +
AAC + +
Renal: z BUN, creatinine
Headache +
Seizures
Hypotension +
Ototoxicity
Vestibular dysfunction
Alcohol interaction +
‘‘Red man’’ flushing +
Drug interactions + + + + + +
Pregnancy risk C or D + + + + +

29
30 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38

Table 8
Pregnancy risk categories of selected antibiotics
Antibiotic Pregnancy risk category Pregnancy risk
Penicillins
penicillin G and V B
ampicillin B
amoxicillin B
amoxicillin/clavulanate B
ticarcillin/clavulanate B
Cephalosporins
cephalexin B
cefazolin B
cefaclor B
cefuroxime B
cefoxitin B
cefotaxime B
Carbapenems
imipenem C Spontaneous abortions in monkeys
meropenem B
Macrolides
erythromycin B
clarithromycin C Fetal defects in mice and monkeys
azithromycin B
Antianaerobic
clindamycin B
metronidazole B
Fluoroquinolones
ciprofloxacin C Spontaneous abortions in rabbits
moxifloxacin C Fetal toxicity in rodents and monkeys
Aminoglycosides
gentamicin D Ototoxicity in human fetuses
tobramycin D Ototoxicity in human fetuses
Other
vancomycin C Potential ototoxicity in human fetuses
tetracyclines D Intrinsic dental staining
doxycycline D Intrinsic dental staining
linezolid C Fetal toxicity in rodents
telithromycin B
Data from Gilbert DN, Moellering RC Jr, Sande MA. The Sanford guide to antimicrobial therapy. 32nd edition. Hyde Park (VT):
Antimicrobial Therapy, Inc.; 2002.
A = Studies in pregnancy; no risk.
B = Animal studies no risk, but human studies inadequate or animal toxicity, but human studies no risk.
C = Animal studies show toxicity, and human studies inadequate, but benefit of use may outweigh risk.
D = Evidence of human risk, but benefits may outweigh risk.
X = Risk outweighs benefit.

Cost cially by the intravenous route. Table 10 compares the

Although clinical effectiveness and reduction of
the morbidity of infection and treatment are of para-
mount concern in the management of head and neck
infections, cost is a factor that should be considered
when other factors do not predominate. The costs of
oral antibiotic therapy can be compared based on the
cost for a standard prescription for the antibiotics of
interest, because there is no additional cost of admin-
istration, as there is with parenteral antibiotics, espe-
retail cost of a 1-week prescription of the antibiotics
listed. The penicillin V cost ratio is calculated by
dividing the retail cost of the standard 1-week pre-
scription for the given antibiotic by that of penicillin V.
Table 11 compares the cost of intravenous anti-
biotics. The cost of administration assumes great
importance. Each dose requires sterile intravenous ad-
ministration supplies, professional labor, and hospital
sterile processing and drug error prevention systems.In
Table 11, these costs are conservatively estimated at
 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38 31

Table 9
Selected antibiotic interactions with other drugsa,b
Antibiotic Second drug Adverse effects Mechanism
Erythromycin, clarithromycin, Theophylline Seizures, dysrhythmias Antibiotic inhibits
ketoconazole, itraconazole cytochrome P450
metabolism of second
drug; ketoconazole
not implicated
Erythromycin, clarithromycin, Cisapride Dysrythmias (torsades) Antibiotic inhibits
ketoconazole, itraconazole cytochrome P450
metabolism of
second drug
Erythromycin, clarithromycin, Alfentanil z Respiratory depression Antibiotic inhibits
ketoconazole, itraconazole cytochrome P450
metabolism of second
drug; ketoconazole
not implicated
Erythromycin, clarithromycin, Bromocriptine z CNS effects, hypotension Antibiotic inhibits
ketoconazole, itraconazole cytochrome P450
metabolism of
second drug
Erythromycin, clarithromycin, Carbamazepine Ataxia, vertigo, drowsiness Antibiotic inhibits
ketoconazole, itraconazole cytochrome P450
metabolism of
second drug
Erythromycin, clarithromycin, Cyclosporine z Immunosuppression Antibiotic inhibits
ketoconazole, itraconazole and nephrotoxicity cytochrome P450
metabolism of
second drug
Erythromycin, clarithromycin, Felodipine, possibly other Hypotension, tachycardia, Antibiotic inhibits
ketoconazole, itraconazole calcium channel blockers edema cytochrome P450
metabolism of
second drug
Erythromycin, clarithromycin, Methylprednisolone, z Immunosuppression Antibiotic inhibits
ketoconazole, itraconazole prednisone cytochrome P450
metabolism of
second drug
Erythromycin, clarithromycin, Lovastatin, possibly Muscle pain, rhabdomyolysis Antibiotic inhibits
ketoconazole, itraconazole other -statins cytochrome P450
metabolism of
second drug
Erythromycin, clarithromycin, Triazolam, oral midazolam z Sedative depth and duration Antibiotic inhibits
ketoconazole, itraconazole cytochrome P450
metabolism of
second drug
Erythromycin, clarithromycin, Disopyramide Dysrhythmias Antibiotic inhibits
ketoconazole, itraconazole cytochrome P450
metabolism of
second drug
Erythromycin Clindamycin # Antibiotic effect Mutual antagonism
Erythromycin, tetracyclines Digoxin Digitalis toxicity, Antibiotic kills
dysrhythmias, visual Eubacterium lentum,
disturbances, which metabolizes
hypersalivation digoxin in the gut
Erythromycin, clarithromycin, Warfarin z Anticoagulation Antibiotic interferes
metronidazole Anisindione with metabolism of
the second drug

(continued on next page)

32 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38
Table 9 (continued )
Antibiotic
Tetracycline, cefamandole,
cefotetan, cefoperazone,
sulfonamides, aminoglycosides
Metronidazole, cephalosporins
Metronidazole
Metronidazole, tetracyclines
Tetracyclines, fluoroquinolones
Clindamycin, aminoglycosides,
tetracyclines, bacitracin
Clindamycin
Penicillins, cephalosporins,
metronidazole, erythromycin,
clarithromycin, tetracyclines,
rifampin
Ampicillin, amoxicillin
Cephalosporins
Trimethoprim/sulfamethoxazole
Vancomycin
Second drug Adverse effects Mechanism
Warfarin, anisindione z Anticoagulation Antibiotic kills gut
flora that synthesize
vitamin K, which
antagonizes the second
drug; poor vitamin K
intake a factor
Alcohol, ritonavir Flushing, headache, Antibiotic inhibits
palpitations, nausea acetaldehyde
dehydrogenase,
causing accumulation
of acetaldehyde;
ritonavir preparations
contain alcohol
Disulfiram Acute toxic psychosis
Lithium Lithium toxicity: confusion, Antibiotic inhibits
ataxia, kidney damage lithium excretion by
kidney; tetracycline
interaction not well
established
Divalent and trivalent cations # Absorption of antibiotic Second drug interferes
(dairy, antacids, vitamins) with absorption of
didanosine antibiotic; didanosine
is formulated with
calcium carbonate and
magnesium hydroxide
buffers
Neuromuscular blocking z Depth and duration Additive effect caused
agents of paralysis by inherent minor
neuromuscular
blocking effect of the
antibiotic; seen with
clindamycin in the
presence of low
pseudocholinesterase
levels and abnormal
liver function tests
Erythromycin # Antibiotic effect Mutual antagonism
Estrogen- and progestin- Contraceptive failure Interference with
containing oral contraceptives enterohepatic
recirculation of
estrogen caused by
killing of gut flora;
rifampin is the only
antibiotic in which
this has been
clinically proven
Allopurinol Rash Unknown, possibly
caused by hyperuricemia
in patients taking
allopurinol
Aminoglycosides z Nephrotoxicity Additive or
potentiating effect
Thiazide diuretics Purpura, bleeding in Thrombocytopenia
elderly patients
Aminoglycosides z Renal toxicity Additive effect
(continued on next page)
 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38 33

Table 9 (continued )
Antibiotic Second drug Adverse effects Mechanism
Fluoroquinolones, sulfonamides, Oral hypoglycemic agents Hypoglycemia Antibiotic displaces
chloramphenicol, fluconazole, second drug from
itraconazole plasma proteins
Ciprofloxacin, sulfonamides, Phenytoin z Serum level of phenytoin, Interference with
chloramphenicol, fluconazole, confusion, delirium phenytoin metabolism
ketoconazole, itraconazole
Sulfonamides Methotrexate z Methotrexate concentration Antibiotic displaces
methotrexate
from plasma proteins
Protease inhibitors (ritonavir, Hydrocodone, fentanyl, z Levels of second drug, Serious interaction:
amprenavir, saquinavir, alfentanil, amiodarone, with possible toxic effects avoid using the drugs
nelfinavir, indinavir, lidocaine, anticonvulsants, in bold print
and others) loratidine, Ritonavir has high
Benzodiazepines affinity for various
b-blockers isoenzymes in the
Calcium channel blockers cytochrome P450 system
Cisapride and has the most frequent
Corticosteroids and severe drug
-statin type interactions among the
antihyperlipidemics protease inhibitors
Warfarin Warfarin reaction is
only with ritonavir
Protease inhibitors Codeine, morphine, # Levels of second drug Antibiotic enhances
contraceptives cytochrome P450
metabolism of
second drug
Delavirdine (Rescriptor) Cisapride, clarithromycin, z Levels of second drug, Antibiotic inhibits
protease inhibitors, warfarin with possible toxic effects cytochrome P450
metabolism of
second drug
Didanosine (ddl, Videx) Metronidazole z Risk of peripheral Additive effect
neuropathy
Foscarnet (Foscavir) Ciprofloxacin z Risk of seizures Additive effect
From Flynn TR. Update on the antibiotic therapy of oral and maxillofacial infections. In: Piecuch JF, editor. Oral and
maxillofacial surgery knowledge update 2001. Rosemont (IL): American Association of Oral and Maxillofacial Surgeons, 2001;
with permission.
a
Interactions among the various anti-HIV antibiotics are frequent and complex. The reader is referred to appropriate sources
on the subject.
b
This list of antibiotic-drug interactions is only partial and selected according to the interests of oral and maxillofacial
surgeons. Drug prescribers remain responsible to ascertain the complete drug interactions of any medications they may prescribe.

$4.00 per dose. Even this small additional cost can
make an infrequently administered but more expensive
antibiotic more economical than a cheaper, more fre-
quently dosed antibiotic. An example of this effect can
be found by comparing the cost ratio of penicillin G
(analogous to the penicillin V cost ratio) with cefazolin.
Table 11 also illustrates the markedly increased
cost of combined antibiotic therapy as compared to
monotherapy. For example, 1 week’s intravenous
therapy of penicillin G plus metronidazole costs
$690, whereas 1 week’s treatment with clindamycin
costs only $375, a reduction of 46%. On the other
hand, the combination approach may be advantageous
in an infection that threatens the brain, for example,
because clindamycin does not cross the blood-brain
barrier and penicillin does so only to a limited extent.
Metronidazole crosses the blood-brain barrier well.
New antibiotics of interest to oral and
maxillofacial surgeons
New fluoroquinolones
Moxifloxacin (Avelox) and gemifloxacin are
two new fluoroquinolones whose spectrum includes
34 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38

Table 10
Oral antibiotic costs
Usual dose Usual interval Pharmacy Cost for Retail cost for Penicillin
Antibiotic (mg) (h) Cost ’01 * * 24 hours 1 weekd cost ratioc
Penicillins
Penicillin V 500 6 $0.14 $0.56 $9.99 1.00
Amoxicillin 500 8 $0.31 $0.93 $13.89 1.39
Augmentina 500 8 $3.65 $10.95 $104.99 10.51
Augmentin 875 12 $4.76 $9.52 $97.59 9.77
Dicloxacillin 500 6 $0.66 $2.64 $26.69 2.67
Cephalosporins (generation)
Cephalexin caps (1st) 500 6 $1.07 $4.28 $24.89 2.49
Keftabs (1st)b 500 6 $3.11 $12.44 $104.99 10.51
Cephradine (1st) 500 6 $0.52 $2.08 $70.59 7.07
Cefuroxime (2nd) 500 8 $7.43 $22.29 $199.99 20.02
Cefaclor (2nd) 500 8 $4.00 $12.00 $77.59 7.77
Erythromycins
Erythromycin base 500 6 $0.36 $1.44 $13.89 1.39
Erythromycin stearate 333 6 $0.36 $1.44 $16.29 1.63
Erythromycin estolate 250 6 $0.31 $1.24 $13.49 1.35
Dirythromycin (Dynabec) 500 24 $12.39 $12.39 $63.99 6.41
Clarithromycin (Biaxin) 500 12 $3.57 $7.14 $71.99 7.21
Azithromycin (Zithromax) 250 24 $6.75 $6.75 $60.59 6.07
Anti-anaerobic
Clindamycin (generic) 150 6 $0.98 $3.92 $31.29 3.13
Clindamycin (2 T generic) 300 6 $1.96 $7.84 $54.86 5.49
Clindamycin (Cleocin) 300 6 $4.22 $16.88 $118.27 11.84
Metronidazole (250 mg = $0.08) 500 6 $0.72 $2.88 $10.02 1.00
Other
Trimethoprim/sulfamethoprim 160/800 12 $0.15 $0.30 $11.69 1.17
Ciprofloxacin 500 12 $4.15 $8.30 $80.59 8.07
Doxycycline 100 12 $0.08 $0.16 $9.99 1.00
Vancomycin 125 6 $5.38 $21.52 $187.99 18.82
Usual doses and intervals are for moderate infections, and are not to be considered prescriptive.
From Gilbert DN, Moellering RC Jr, Sande MA. The Sanford guide to antimicrobial therapy 2001. 31st edition. Hyde Park (VT):
Antimicrobial Therapy, Inc; 2001.
a
Augmentin = amoxicillin plus clavulanic acid.
b
Keftab = cephalexin hydrocloride in tablet form (Dista).
c
Penicillin cost ratio = retail cost of antibiotic for 1 week retail cost of penicillin V for 1 week.
d
Retail cost/1 week = retail price charged for a 1-week prescription at a large pharmacy chain in the Boston region. Courtesy
of Chris Gonzalez, RPh.

the viridans streptococci, oral anaerobes, and acti-
nomyces. They are also effective against sinus
pathogens, staphylococci, Enterobacteriaceae, and
B. fragilis. Their broad spectrum is a relative dis-
advantage when the target is a fairly small range of
bacteria. These new fluoroquinolones probably
should be reserved for situations in which a narrower
spectrum alternative antibiotic is not available.
Oxazolidinones
Linezolid (Zyvox) is the prototype of this new
class of antibiotics. It is effective against virtually all
gram-positive pathogens but not against the gram-
negative oral anaerobes. Its effectiveness against
methicillin- and vancomycin-resistant staphylococci
and enterococci indicates that it should be reserved
for these highly resistant organisms [25].
Ketolides
Telithromycin (Ketek) is the first representative of
this new class, which is related to the macrolides. Its
spectrum includes the pathogens against which the
macrolides have been historically effective, including
S. pneumoniae, mycoplasma, H. influenzae, Chlamy-
 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38 35

Table 11
Intravenous antibiotic costs
Usual Usual interval Pharmacy Pharmacy Total cost Total cost Penicillin G
Antibiotic doseb (hour)b cost ’00 cost ’01 24 hours for 7 days cost ratioa
Penicillins
Penicillin G 2 mu 6 $1.33 $1.32 $21.28 $148.96 1.00
Ampicillin 1g 6 $1.64 $1.31 $21.24 $148.68 1.00
Unasyn 2g 6 $10.18 $14.45 $73.80 $516.60 3.47
Oxacillin 1g 6 $2.68 $5.14 $36.56 $255.92 1.72
Ticarcillin 3g 4 $12.92 $13.43 $104.58 $732.06 4.91
Timentin 3.1 g 4 $15.40 $15.20 $115.20 $806.40 5.41
Cephalosporins (generation)
Cefazolin (1st) 1g 8 $1.74 $1.90 $17.70 $123.90 0.83
Cefotetan (2nd) 1g 12 $11.58 $11.60 $31.20 $218.40 1.47
Cefuroxime (2nd) 1.5 g 8 $13.93 $13.80 $53.40 $373.80 2.51
Cefotaxime (3rd) 2g 8 $21.16 $26.38 $91.14 $637.98 4.28
Ceftazidime (3rd) 2g 8 $28.45 $28.45 $97.35 $681.45 4.57
Ceftriaxone (3rd) 1g 24 $42.00 $40.18 $44.18 $309.26 2.08
Monobactam
Aztreonam 1g 8 $16.97 $16.97 $62.91 $440.37 2.96
Carbapenem
Imipenem-cilastatin 0.5 g 6 $30.32 $30.32 $137.28 $960.96 6.45
Penicillin allergy
Erythromycinc 1g 6 $22.16 $23.00 $108.00 $756.00 5.08
Azithromycin 0.5 g 24 $23.70 $24.44 $28.44 $199.08 1.34
Vancomycin 0.5 g 6 $7.80 $8.28 $49.12 $343.84 2.31
Vancomycin 1.0 g 12 $15.60 $16.56 $41.12 $287.84 1.93
Anti-anaerobic
Clindamycin 0.9 g 8 $13.88 $13.88 $53.64 $375.48 2.52
Metronidazole 0.5 g 6 $19.03 $15.34 $77.36 $541.52 3.64
Other
Doxycycline 0.1 g 12 $21.07 $4.16 $16.32 $114.24 0.77
Trimethoprim-sulfa 800 mg 6 $16.42 $16.42 $81.68 $571.76 3.84
Ciprofloxacind 400 mg 12 $30.00 $30.00 $68.00 $476.00 3.20
Total cost of therapy includes $1.00 for infusion materials and $3.00 labor cost, per dose.
From Gilbert DN, Moellering RC Jr, Sande MA. The Sanford guide to antimicrobial therapy 2001. 31st edition. Hyde Park (VT):
Antimicrobial Therapy, Inc; 2001.
a
Penicillin cost ratio = 24-hour cost of antibiotic/24-hour cost of penicillin G.
b
Usual doses and intervals are for moderate infections and are not to be considered prescriptive.
c
Only the brand name price is listed in the reference. Price is selected from the lowest available average whole-
sale price.
d
Cipro IV is for NPO patients only because of excellent oral absorption.

dia pneumoniae, and Legionella pneumophila. Its
most frequent use probably is in respiratory tract
infections, especially pneumonia [26,27].
Pristinamycins
Quinupristin/dalfopristin (Synercid), a combina-
tion of two pristinamycin antibiotics, is especially
effective against vancomycin-resistant staphylococci.
Its use generally has been reserved for infections
caused by these organisms.
Empiric antibiotics of choice for head and
neck infections
Odontogenic infections
Empiric antibiotics are administered before cul-
ture and sensitivity test results are available; specific
antibiotic therapy is selected based on culture and
sensitivity results. Table 12 lists the empiric anti-
biotics of choice for selected types of head and neck
infections, including odontogenic infections.
36 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38
In a prospective case series of 34 cases of odonto-
genic infection, Flynn et al reported therapeutic
failure of penicillin in 26% of cases using the
following criteria for failure: allergic or toxic reaction
(no cases); failure of swelling, temperature, and white
blood cell count to decline after at least 48 hours of
intravenous penicillin; and a postoperative CT scan
that demonstrated adequate surgical drainage. If in-
adequate drainage was found on the postoperative CT
scan, surgery was repeated. All of the patients with
therapeutic penicillin failure (8 of 31 cases initially
treated with penicillin) subsequently yielded at least
one penicillin-resistant strain when culture and sen-
sitivity test results became available. This finding
suggests a correlation between infection severity
and penicillin resistance and is the basis for the
recommendation of clindamycin as the empiric anti-
biotic of choice in odontogenic infections serious
enough to require hospitalization [8].
On the other hand, penicillin resistance has not yet
been shown to be a significant problem in outpatient
odontogenic infections [11 – 14]. Penicillin V remains
the empiric antibiotic of choice for outpatient odon-
togenic infections. Because of their ineffectiveness
against the oral anaerobes, the macrolides are no
longer considered among the empiric antibiotics
of choice for odontogenic infections. Because the
oral anaerobic gram-negative rods are fairly resist-
ant to most cephalosporins, especially those in the
first generation, the cephalosporins remain second-
line choices.
Sinus infections
Acute rhinosinusitis of odontogenic origin is char-
acterized by the same flora as other odontogenic
infections, except that not all of the species found
in the periapical infection survive in the sinus loca-
tion [28]. Non-odontogenic acute rhinosinusitis is
frequently caused by S. pneumoniae, H. influenzae,
Moraxella catarrhalis, and streptococci. S. aureus is
found in only approximately 4% of cases of acute
rhinosinusitis [15].
Antibiotic treatment should be reserved for
patients who already have been treated for 7 days
with only decongestants and analgesics and who have
maxillary or facial pain or purulent nasal discharge.
Patients with severe pain or fever may need antibiotic
therapy sooner, and hospitalization may be required
in these cases. If antibiotics have been used in
the previous month or if the local incidence of
penicillin-resistant S. pneumoniae is more than
30%, amoxicillin and clavulanic acid or a second-
or third-generation cephalosporin is prescribed for
two weeks [15]. On the other hand, a recent system-
atic literature review indicates that penicillin or
amoxicillin alone is as effective as the other broader
spectrum and more expensive antibiotics [29].
In chronic rhinosinusitis, the flora becomes more
anaerobic, including B. fragilis and the peptostrep-
tococci, such as Fusobacterium, Prevotella, and
Porphyromonas. Antibiotics alone are not usually
effective in these cases, and corrective surgery, usually
with otorhinolaryngology consultation, is indicated.
Fungal infection of the sinuses should be sus-
pected and treated urgently with antibiotics and
surgery in patients with acute rhinosinusitis who have
diabetes mellitus with acute ketoacidosis, neutro-
penia, or previous treatment with deferoxamine.
Amphotericin B and surgery are indicated, along with
discontinuation of deferoxamine, if applicable. Defer-
oxamine (Desferal) is an iron-chelating agent used in
Alzheimer’s disease. Mucormycosis has been found
in patients who are undergoing simultaneous defero-
xamine treatment and hemodialysis.
Osteomyelitis of the jaw
The microbiology of osteomyelitis of the jaws has
not been reported specifically in a large case series. It
is increasingly apparent from case reports, however,
that the usual odontogenic pathogens are the most
frequent cause. One also may suspect skin and soil
pathogens in traumatic osteomyelitis and salmonella
in sickle-cell osteomyelitis. Actinomyces are another
prominent pathogen in chronic osteomyelitis, and
culture and microscopic examination may be required
to identify this organism. Molecular methods ulti-
mately may become the most rapid and reliable
method for identifying Actinomyces [30]. Long
courses of the antibiotics effective against the Actino-
myces are required (see Table 4). Oral penicillins plus
probenecid can be used for long-term outpatient
therapy. Probenecid inhibits the renal excretion of
penicillin and increases the blood level obtained by
the oral route.
Fungal infections
Various fungi cause a wide spectrum of infectious
manifestations in the head and neck. An excellent
review of the topic can be found in a recent chapter by
Bergman [30]. The major fungal infections of concern
to oral and maxillofacial surgeons are histoplasmosis
and blastomycosis, which may cause granulomatous
oral lesions; aspergillosis and mucormycosis, which
tend to cause sinusitis; and candidiasis, which causes
surface lesions in non-immunocompromised patients
 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38 37

Table 12
Empiric antibiotics of choice for head and neck infections
Type of infection Empiric antibiotic of choice
Odontogenic infections
Outpatient Penicillin
Clindamycin
Cephalexin (or other first-generation cephalosporin)
Penicillin allergy Clindamycin
Cephalexin (only if nonanaphylactoid penicillin reaction)
Inpatient Clindamycin
Ampicillin + metronidazole
Ampicillin + sulbactam
Penicillin allergy Clindamycin
Moxifloxacin
Cefotaxime (only if nonanaphylactoid penicillin reaction)
Rhinosinusitis
Acute Amoxicillin
Amoxicillin/clavulanate
Cefuroxime
Moxifloxacin (over 18 years of age)
Penicillin allergy Clarithromycin or azithromycin
Telithromycin
Moxifloxacin (over 18 years of age)
Chronic Antibiotics not effective:
otolaryngologic consultation
Intubated Imipenem or meropenem
Ticarcillin or piperacillin
Ceftazidime + vancomycin
Cefepime
Fungal Amphotericin B
Osteomyelitis of Clindamycin
the jaw
Ampicillin + metronidazole
Ampicillin + sulbactam
Penicillin allergy Clindamycin
Moxifloxacin
Histoplasmosis and Itraconazole
blastomycosis
Fluconazole
Amphotericin B
(systemic or disseminated)
Candidiasis
Oral, non-AIDS Fluconazole or itraconazole
Nystatin or clotrimazole
Oral, AIDS Fluconazole or itraconazole
Amphotericin B
Data from Gilbert DN, Moellering RC Jr, Sande MA. The Sanford guide to antimicrobial therapy. 32nd edition. Hyde Park (VT):
Antimicrobial Therapy Inc.;2002.

and may cause disseminated and invasive disease in
immunocompromised persons. Histoplasmosis, blas-
tomycosis, and mucormycosis are diagnosed by sur-
gical sampling for culture, histologic examination
with special stains, and use of molecular methods,
such as polymerase chain reaction. In general, fungal
infections are treated with the azole-type antifungal
agents for less severe cases and amphotericin B for
disseminated and severe disease. In surface candidia-
sis in a patient with a healthy immune system,
clotrimazole is a better-tasting yet economical alter-
native to nystatin.
38 T.R. Flynn, L.R. Halpern / Oral Maxillofacial Surg Clin N Am 15 (2003) 17–38
Summary
Antibiotic selection remains as much of an art as it
is a science. It requires the integration of many factors
that are host specific, pharmacologic, and even geo-
graphic. Much more research is necessary in this field
to solve the current problems with the need for more
timely culture and sensitivity results, increasing anti-
biotic resistance, and best practices in antibiotic usage.
References
[1] Flynn TR. Update on the antibiotic therapy of oral and
maxillofacial infections. In: Piecuch JF, editor. Oral
and maxillofacial surgery knowledge update 2001.
Rosemont (IL): American Association of Oral and
Maxillofacial Surgeons; 2001. p. 23 – 50.
[2] Molinari JA. Antibiotic resistance and maxillofacial
pathogens: emerging treatment issues. J California
Dental Assoc 1999;27:386 – 92.
[3] Neu HC. Emerging trends in antimicrobial resistance
in surgical infections: a review. Eur J Surg Suppl 1994;
573:7 – 18.
[4] Kulekei G, Inane D, Kocak H, et al. Bacteriology of
dentoalveolar abscesses in patients who have received
empirical antibiotic therapy. Clin Infect Dis Suppl
1996;1:S51 – 3.
[5] Kuriyama T, Nakagawa K, Karasawa T, et al. Past
administration of b-lactam antibiotics and increase in
the emergence of b-lactamase-producing bacteria in
patients with orofacial odontogenic infections. Oral
Surg Oral Med Oral Pathol Oral Radiol Endod
2000;89:186 – 92.
[6] Haug RH, Hoffman MJ, Indresano AT. An epidemio-
logic and anatomic survey of odontogenic infections.
J Oral Maxillofac Surg 1991;49:976 – 80.
[7] Storoe W, Haug RH, Lillich TT. The changing face of
odontogenic infections. J Oral Maxillofac Surg 2001;
59:739 – 48.
[8] Flynn TR, Wiltz M, Adamo AK, et al. Predicting
length of hospital stay and penicillin failure in severe
odontogenic infections. Int J Oral Maxillofac Surg
1999;28(Suppl 1):48.
[9] Brook I. Microbiology of common infections in the
upper respiratory tract. Prim Care 1998;25:637 – 47.
[10] Doern GV, Ferraro MJ, Brueggemann AB, Ruoff
KL. Emergence of high rates of antimicrobial resist-
ance among viridans group streptococci in the
United States. Antimicrob Agents Chemother 1996;
40:891 – 4.
[11] Fazakerley MW, McGowan P, Hardy P, Martin MV.
A comparative study of cephradine, amoxycillin
and phenoxymethylpenicillin in the treatment of acute
dentoalveolar infection. Br Dent J 1993;174:359 – 63.
[12] Gilmore WC, Jacobus NV, Gorbach SL, et al. A pro-
spective double-blind evaluation of penicillin versus
clindamycin in the treatment of odontogenic infections.
J Oral Maxillofac Surg 1988;46:1065 – 70.
[13] Lewis MA, Carmichael F, MacFarlane TW, Milligan
SG. A randomised trial of co-amoxiclav (Augmentin)
versus penicillin V in the treatment of acute dentoal-
veolar abscess. Br Dent J 1993;175:169 – 74.
[14] Paterson SA, Curzon ME. The effect of amoxycillin
versus penicillin V in the treatment of acutely ab-
scessed primary teeth. Br Dent J 1993;174:443 – 9.
[15] Gilbert DN, Moellering Jr RC, Sande MA. The San-
ford guide to antimicrobial therapy, 32nd edition. Hyde
Park (VT): Antimicrobial Therapy Inc.; 2000.
[16] Craig TJ, Mende C. Common allergic and allergic-like
reactions to mediations: when the cure becomes the
curse. Postgrad Med 1999;105:173 – 81.
[17] Bowrey DJ, Morris-Stiff GJ. Drug allergy: fact or fic-
tion? Int J Clin Pract 1998;52:20 – 1.
[18] Warrington RJ, Lee KR, McPhillips S. The value of
testing for penicillin allergy in an inpatient population:
analysis of the subsequent patient management. Al-
lergy Asthma Proc 2000;21:297 – 9.
[19] DeFonseca MA. Adverse reaction to amoxicillin:
a case report. Pediatr Dent 2000;22:401 – 4.
[20] Ebersole JL, Cappelli D. Acute-phase reactants in in-
fections and inflammatory diseases. Periodontology
2000;23:19 – 49.
[21] Bancescu G, Skaug N, Dumitriu S, et al. Antimicrobial
susceptibility of some streptococci strains of anginosus
group isolated from oral and maxillofacial infections.
Roum Arch Microbiol Immunol 1999;58:57 – 63.
[22] Kasten MJ. Clindamycin, metronidazole, and chloram-
phenicol. Mayo Clin Proc 1999;74:825 – 33.
[23] Fisman DN, Kaye KM. Once-daily dosing of amino-
glycoside antibiotics. Infect Dis Clin North Am
2001;14:475 – 87.
[24] Hersh EV. Adverse drug interactions in dental practice:
interactions involving antibiotics. Part II. J Am Dent
Assoc 1999;130:236 – 51.
[25] Marchese A, Schito GC. The oxazolidinones as a new
family of antimicrobial agent. Clin Microbiol Infect
2001;7(Suppl 4):66 – 74.
[26] Felmingham D. Microbiological profile of telithromy-
cin, the first ketolide antimicrobial. Clin Microbiol In-
fect 2001;7(Suppl 3):2 – 10.
[27] Linden PK. Treatment options for vancomycin-resist-
ant enterococcal infections. Drugs 2002;62:425 – 41.
[28] Brook I, Frazier EH, Gher Jr ME. Microbiology of
periapical abscesses and associated maxillary sinusitis.
J Periodontol 1996;67:608 – 10.
[29] Williams Jr JW, Aguilar C, Makela M, et al. Antibi-
otics for acute maxillary sinusitis. Cochrane Library
2000;3:1 – 51.
[30] Bergman SA. Fungal, viral, and protozoal infections
of the maxillofacial region. In: Topazian RG, Goldberg
MH, Hupp JR, editors. Oral and maxillofacial infec-
tions. 4th edition. Philadelphia: W.B. Saunders Co.;
2002. p. 243 – 78.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 39 – 49

Diagnostic imaging of maxillofacial infections

Thomas E. Underhill, DDS, MDa,*, Fred J. Laine, MDa,b, John George, BSc
a
Department of Radiology, Division of Neuroradiology, VCU Health Systems/MCV Hospitals and Physicians,
1250 East Marshall Street, PO 980615, Richmond, VA 23298, USA
b
Department of Otolaryngology, VCU Health Systems/MCV Hospitals and Physicians, 1250 East Marshall Street,
PO 980615, Richmond, VA 23298, USA
c
Medical Student, 8377 NW 57th Drive, Coral Springs, FL 33067, USA

Diagnosis and management of head and neck
infections are common clinical challenges for the oral
and maxillofacial surgeon. Symptoms, signs, and
laboratory data are often suggestive of an infectious
or inflammatory process. Given the right clinical
conditions, however, several noninfectious conditions
can mimic these processes. Based on clinical exam-
ination and occasionally laboratory data, the exam-
ining surgeon must determine the need for advanced
imaging studies.
Opinions still vary as to whether computed tomog-
raphy (CT) or magnetic resonance imaging (MRI) is
the best imaging modality for acute neck infections
[1 – 4]. When imaging is needed, it is the authors’
opinion that the least invasive and least expensive
examination that adequately evaluates the patient
should be used. When an odontogenic origin (necrotic
pulp) is suspected as the source of a localized infec-
tion, intraoral films or panoramic radiographs are
usually adequate. Ultrasound also has been used in
the evaluation of superficial neck infections, espe-
cially to determine fluid accumulation, but is not
recommended for deep neck infections because it
cannot penetrate bone or the airway space. Conven-
tional films still can be used for a preliminary survey,
especially of the retropharyngeal space and evaluation
of the airway.
Although most maxillofacial infections remain
localized and are managed successfully by antibiotic
therapy or surgical intervention, serious and poten-
tially life-threatening complications, such as airway
* Corresponding author.
E-mail address: tunderhill@attbi.com (T.E. Underhill).
compromise [5], cavernous sinus thrombosis [4,6],
cerebral empyema [4,7], internal jugular vein throm-
bophlebitis [8], Ludwig’s angina [9], or necrotizing
fasciitis [10 – 12], can occur.
This article discusses mainly CT and MRI as they
relate to localization of infectious processes and the
evaluation of abscess cavities.
Techniques
Computed tomography
Traditional or single slice acquisition CT uses a
gantry that houses an x-ray tube and a row of
detectors. Images are produced by data collected from
the detectors after a 360° rotation. After each tomo-
graphic image the patient table is moved and another
image obtained. A time delay of 10 to 15 seconds
between each slice is necessary. Spiral CT involves
the simultaneous movement of the patient table and
the x-ray tube, which results in a volume acquisition
of data from which individual tomographic images
can be reconstructed. Because a volume data set is
acquired, excellent multiplanar reformations are pos-
sible when using thin image slices (3 mm or less).
Picture archival communication systems are becom-
ing more common in hospitals. Some of these systems
allow viewing multiplanar reformation in any plane
desired, not just the standard sagittal and coronal
planes. In the past, CT reformation programs, such
as DentaScan, were recommended for true cross-
sectional images of the jaws, not only for implant
planning but also for evaluation of tumors, osteomye-
litis, or other pathologic conditions [13,14]. Picture

1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 7 5 - 4
40 T.E. Underhill et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 39–49

archival communication systems with multiplanar
reformation capability obviate the need for such pro-
grams in the evaluation of pathologic conditions.
Other advantages of spiral CT in applications to the
head and neck include one breath hold, which mini-
mizes artifacts because of swallowing, and improved
vascular opacification and lesion enhancement using a
smaller contrast bolus [15]. Multidetector CT is yet
another improvement over spiral CT. Whereas spiral
CT uses a single row of detectors, multidetector CT
uses a matrix of detectors that allows the acquisition of
multiple tomographic images per revolution, which
greatly increases the speed of imaging.
Accuracy of computed tomography
Deep neck infections in adult patients
Miller et al [16] performed a prospective study that
compared the efficacy of contrast-enhanced CT
(CECT) to clinical examination in detecting the pres-
ence of a drainable fluid collection in suspected deep
neck infections. The accuracy (frequency of a test to
diagnosis correctly the presence or absence of disease)
of clinical examination alone in identifying a drain-
able collection was 63%, the sensitivity (ability of a
test to identify correctly a disease when it is truly
present) was 55%, and the specificity (ability of a test
to identify correctly the absence of disease when it is
truly absent) was 73%. The accuracy of CECT alone
was 77%, the sensitivity was 95%, and the specificity
53%. When CECT and clinical examination were
combined, the accuracy in identifying a drainable
collection was 89%, the sensitivity was 95%, and
the specificity was 80% [16].
Deep neck infections in pediatric patients
Nagy et al [3] compared the sensitivity of lateral
neck films and CECT in evaluating children with a
high index of suspicion for a deep neck infection,
based on clinical presentation. More than 25% of
lateral neck radiographs were unable to determine
the presence of a deep neck infection, whereas CECT
had a sensitivity of 100%. They concluded that CECT
is the study of choice in the evaluation of children
strongly suspected of having a deep neck infection
based on clinical presentation [3].

Fig. 1. (A) Chronic dentoalveolar granuloma. Axial T1-weighted MRI without contrast shows the normal bright signal from fatty
marrow in the maxillary alveolar ridge. There is a focal area with low signal surrounding the maxillary canine root (arrow),
which indicates replacement of normal fatty marrow by either a neoplastic or inflammatory process. (B) Parasagittal T1-weighted
MRI with contrast and fat saturation demonstrates enhancement of tissue surrounding the root of the right maxillary canine. The
uniform enhancement indicates a granuloma rather than an abscess, which would exhibit only peripheral enhancement.
 T.E. Underhill et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 39–49 41

Wetmore et al [17] evaluated the use of CT in coils distort the uniform field, which causes only those
children with deep neck infections. CECT demonstra- hydrogen protons in a certain plane to resonate when
ted an accuracy of 92% for detection of an abscess excited by a specific radiofrequency. These excited
confirmed by surgery. In children with superficial neck protons emit a signal that, when analyzed by a two-
infections, only 24% of those who were treated sur- dimensional Fourier, is transformed into an image.
gically showed definite CT evidence of an abscess.
The decision to perform surgery was based on clinical Accuracy of magnetic resonance imaging
findings, such as palpable fluctuance and skin changes
[17]. As with adults, the combination of clinical Munoz et al [2] prospectively evaluated 47 patients
examination and radiographic data yielded the with neck infections. Each patient underwent CT and
best results. MRI with contrast of the area of interest using similar
slice thickness. MRI was superior to CT in regard to
NewTom lesion conspicuity, number of anatomic spaces
involved, extension, and source. CT was superior to
A new development in maxillofacial imaging is the
NewTom (Schick NIM S.r.l., Verona, Italy). This
machine performs axial panoramic and multiplanar
imaging of the maxillofacial region and produces
images similar to those acquired by the DentaScan
(General Electric Medical Systems, Milwaukee, WI)
as viewed in bone windows. Image acquisition by the
NewTom is fundamentally different from conventional
CT. The images are derived from a three-dimensional
data set acquired from multiple planar images (ie,
multiple digital projections), which renders excellent
high tissue contrast resolution (bony detail). Because
of this ability, the NewTom is excellent for detection of
lytic bony lesions; however, it lacks the ability to
demonstrate the low tissue contrast resolution needed
to detect subtle soft tissue changes as seen with
cellulitis, soft tissue abscess, cerebral empyema, or
retropharyngeal fluid/edema. The lower neck cannot
be imaged. For these reasons, conventional CECT is
recommended for the evaluation of head and neck
inflammatory processes.

Magnetic resonance imaging

A disadvantage of MRI is the lack of availability.
Most MRI units are ‘‘tightly booked’’ with long
waiting periods for appointments. Even a 1-day wait
is not acceptable for a patient with a potentially life-
threatening head and neck infection. Over the years,
MRI scanners have become more numerous and are
more available for emergent scans. Some imaging
centers have time reserved each day for emergency
scanning, and some large medical centers have MR
scanners in the emergency department. Even with
emergency time reserved, however, maxillofacial
and neck infections would have a lower priority for
the MRI scanner than examinations to evaluate for
acute spinal cord compression or acute stroke.
An MRI is obtained by placing the patient in a
strong and uniform magnetic field. Smaller gradient
Fig. 2. Retropharyngeal abscess. Axial image of CT scan
with contrast obtained at the level of the maxilla. A 1.5-cm
retropharyngeal abscess is present (large arrow). This is seen
as a fluid density collection surround by a rim of enhancing
tissue. Notice that the density or attenuation of this fluid
collection is the same as the cerebrospinal fluid that
surrounds the cord. A second area of fluid density, which
is not as well defined (smaller arrows), is seen in the lateral
neck. This represents a multiloculated abscess. Notice that
normal fat on the soft tissue windows is dark. There is an area
of edema or infiltration in the fat (stars). This appearance is
referred to as dirty fat. Compared to the right sternocleido-
mastoid muscle (SCM), the left sternocleidomastoid muscle
cannot be distinguished from the surrounding inflammatory
process. The left jugular vein is compressed and is not
visualized. The right jugular vein and the right carotid artery
are in their normal location (J,C). The left carotid artery (c) is
displaced anteriorly by the inflammatory process.
42 T.E. Underhill et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 39–49

MRI in the detection of intralesional gas and calcium
and showed fewer motion artifacts. On this basis, MRI
was considered superior to CT in the initial evaluation
of neck infections. These findings suggest that MRI
should be used as the first and perhaps only modality
to evaluate patients initially with neck infections [2].
whereas muscle decreases in attenuation on noncon-
trasted CT [18]. Increased blood flow occurs in
inflamed tissue. After administration of iodinated
contrast medium, areas of increased blood flow dem-
onstrate enhancement. Intravenous iodinated contrast
is indicated for CT evaluation of a patient with
suspected cellulitis or abscess.

Interpretation Magnetic resonance imaging

Computed tomography The MRI parameters can be manipulated to pro-

Computed tomography uses the differences in
attenuation of the x-ray beam by different tissues to
form an image. The lowest attenuation occurs in air,
and the highest attenuation occurs in bone, dentin,
enamel, or metal. Fat has a lower attenuation than
water, which in turn has lower attenuation than
muscle. When edema occurs, there is an increase in
water content. Edematous fat increases in attenuation,
duce several different appearances. As an oversim-
plification, T1-weighted images produce images in
which fat is bright and fluid such as cerebrospinal
fluid is dark. The presence or absence of the bright
signal from fat can help detect pathologic processes.
Normally fatty marrow is bright. Any pathologic
process, such as metastatic disease or infection, that
has replaced the normal fatty marrow appears abnor-
mally dark (Fig. 1A). Gadolinium contrast medium is

Fig. 3. (A) Subperiosteal abscess. Axial CT image with contrast at the level of the inferior border of the mandible in a 43-year-old
man with painful swelling of the neck. There is a fluid mass density surrounding the angle of the mandible and extending along
the inferior border (arrows). This abnormal fluid collection extends on both sides of the mandible. The geniohyoid muscle
(asterisks) is displaced medially. (B) Bone window demonstrates a small periapical radiolucency around the distal root of the
mandibular right second molar. The next inferior image demonstrates a small break in the cortex. The source of the large
subperiosteal abscess was from the necrotic second molar.
 T.E. Underhill et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 39–49 43

used only with T1-weighted images. When concen-
trated in tissues, gadolinium contrast medium dem-
onstrates an increase in signal intensity. Unlike
imaging the brain, where (usually) no fat is located,
in maxillofacial, orbital, or neck imaging, fat satura-
tion is critical when using gadolinium contrast
medium. Fat saturation makes the normal fat dark,
and normal and abnormal enhancement is much more
conspicuous (Fig. 1B). When evaluating a contrasted
T1-weighted image of the head and neck, there
should be an avid enhancement of the nasal mucosa.
If it does not enhance, there is a problem with the
contrast bolus and pathologic processes also would
fail to enhance. This occurrence could lead to a false-
negative examination result. The ocular muscles
normally enhance, but the larger muscles, such as
paraspinal muscles or muscles of mastication, do
not enhance.
With T2-weighted images, the ‘‘juicier or squish-
ier’’ [19] the tissue, the brighter it is. Stationary fluid,
such as cerebrospinal fluid, appears bright. Edematous
or inflamed tissues demonstrate increased signal
intensity, especially when fat saturation is used.
Pathology
Odontogenic infections
In a retrospective study of 210 patients with neck
infections, the most common cause was dental infec-
tion (43%). Dental infection was the cause of 76% of
Ludwig’s angina [20]. In patients with deep neck
space infections, airway compromise was more fre-
quent and severe in odontogenic than in nonodonto-
genic deep neck space infections. The parapharyngeal,
submandibular, and masticator spaces are more vul-
nerable in odontogenic deep neck space infections
than in nonodontogenic infections. The predilection
for certain spaces of the neck to be involved in

Fig. 4. (A) Septic thrombophlebitis. Axial CT examination with contrast performed at the level of the inferior border of the
mandible in a 31-year-old woman with a painful swollen left neck. There is an area of low attenuation surrounded by avid
enhancement (arrows). Compared to Fig. 1, this has the appearance of an abscess, but the hypodensity extends in a craniocaudal
dimension from the skull base to the level of the hyoid bone. This represents a low-density thrombus surrounded by enhancing
vasovasorum. Note the normally enhancing jugular vein on the right (JV ). (B) Axial enhanced CT scan at the level of the thyroid
gland (T ). Note the normally enhancing jugular vein at this level on the left and the normal carotid arteries (C ).
44 T.E. Underhill et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 39–49
 T.E. Underhill et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 39–49 45

odontogenic deep neck space infection originates

from the intimate relationship of the mandibular
molars to the adjacent deep neck spaces [5] (Fig. 2).

Retropharyngeal abscess versus cellulitis

On CT, cellulitis appears as soft tissue swelling,

increased density of surrounding fat, enhancement of
involved muscles, and obliteration of fat planes [3,5].
An abscess is considered to be present if there is a low
(fluid) density area with a peripheral rim of enhance-
ment (Fig. 3) [3,5]. Symptoms include fever, neck
swelling, sore throat, dysphasia, and cervical rigidity.
Sometimes small children present with nonspecific
symptoms [21]. Children younger than 5 years who
present with poor oral intake, high fever, drooling, and
trismus should be suspected of having a peritonsillar
abscess. A CT scan of the neck is usually required to
confirm a suspected diagnosis [22].
Recurrent deep neck infections
Most deep neck infections are the result of sup-
purative adenitis. The location of the primary focus is
usually from the mucosa of the upper aerodigestive
tract or from an odontogenic source. Recurrence in
these situations is unusual. Less commonly, congen-
ital lesions can present as deep neck infections, and
recurrences are common. In patients with recurrence
of deep neck infections, the possibility of an under-
lying congenital lesion, such as branchial cleft cyst,
infected lymphangioma, or thyroglossal duct cyst,
should be considered [23].
Nontuberculous and tuberculous
mycobacteria infection
Nontuberculous mycobacteria infection most
often presents in children as an asymmetric adenop-
athy with contiguous low-density ring-enhancing
masses, minimal or absent inflammatory stranding
of the subcutaneous fat, and cutaneous extension
Fig. 6. Periorbital abscess. Coronal CT scan with contrast in
a 14-year-old boy who experienced orbital swelling and
pain. Notice the subperiosteal fluid collection adjacent to the
orbital roof on the left (arrows). There is opacification of
the maxillary sinuses and left ethmoid sinuses. Also notice is
the periorbital soft tissue swelling. The left medial rectus
(asterisk) is displaced laterally from the lamina propria. The
cause for this orbital abscess was spread of infection from
ethmoid sinusitis.
[24]. Cervical tuberculous lymphadenitis is also an
important cause of a neck mass in many countries.
Multilocular low densities with peripheral enhance-
ment and a large confluent low density with a lesser
degree of fat plane obliteration than a pyogenic
abscess are suggestive features of advanced cervical
tuberculous lymphadenitis.
Septic thrombophlebitis
Septic thrombophlebitis of the jugular vein
(Lemierre’s syndrome) is an uncommon cause of
painful neck swelling [8]. On CT, a thrombus within
the jugular vein appears as an area of low attenuation
with peripheral enhancement (Fig. 4), which easily
could be mistaken for abscess. Normal enhancing
vascular structures of the neck must be identified

Fig. 5. (A) Sialadenitis with abscess formation. The patient presented with painful swelling in the left floor of the mouth.
Clinically, an odontogenic source of infection was suspected. Axial CT scan performed at the level of the submandibular gland
showed multiple small locules of gas present as low attenuation or dark areas (arrows). Notice that gas has extended into the
submandibular space, buccal space, and parapharyngeal space. Compare to the normal submandibular gland (SMG) on the right.
(B) Axial CT image performed superior to the submandibular gland at the level of the floor of the mouth. Note the presence of
gas in the fascial plane of the floor of the mouth, with some crossing the midline. The gas is the product of gas-producing
organisms. The appearance is concerning for early development of Ludwig’s angina. (C) Axial CT scan performed at the level of
the maxillary sinuses. Gas in the parapharyngeal space (black arrow) has tracked up toward the skull base. There is also a large
amount of gas in the buccal space (white arrow), anterior to the left masticator space.
46 T.E. Underhill et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 39–49

for abscess/cellulitis is recommended. Frequently,

Fig. 7. Pott’s puffy tumor. A CT image with contrast at the
level of the frontal sinus of a 14-year-old boy with painful
frontal swelling. A subperiosteal abscess is present (asterisk).
There is erosion of the inner wall of the frontal sinus and
abnormal dural enhancement (arrow). A bony sequestrum
is present.
and traced in the craniocaudal direction. Tracing of
other neck structures also is important. Necrotic
lymph nodes also can mimic cervical abscesses on
CT scans and result in negative surgical findings [22].
Sialadenitis
In general, CT has supplanted sialography for the
evaluation of inflammatory processes in the salivary
glands (Fig. 5). If salivary calculi are suspected as the
cause of an inflammatory process, a noncontrasted CT
to detect calculi followed by contrasted CT to evaluate
inflammatory processes in the salivary glands present
as a mass and clinically are difficult to distinguish
from malignancy. In such cases, contrasted CT or MRI
can be used to determine if the mass is intrinsic or
extrinsic to the gland [25].
Acute suppurative bacterial sialadenitis is associ-
ated with painful, swollen glands. The parotid gland is
most commonly involved [1]. Salivary calculi are best
seen on CT and appear as a high-density mass along
the course of the duct. CT also demonstrates gas in the
tissue better than MRI [5].
Necrotizing fasciitis
Craniocervical necrotizing fasciitis is a rapidly
progressive, severe bacterial infection of the super-
ficial fascial planes of the head and neck. Group A
beta-hemolytic Streptococcus, Staphylococcus au-
reus, and obligate anaerobic bacteria are commonly
involved pathogens. The disease usually results
from a dental source or facial trauma. Extensive
fascial necrosis and severe systemic toxicity are
common manifestations of craniocervical necro-
tizing fasciitis. Most patients have an underlying
medical problem that created an immunocompro-
mised state, usually diabetes mellitus or chronic
alcoholism [26]. Constant CT features of necro-
tizing fasciitis are diffuse cellulitis, diffuse enhance-
ment or thickening of the superficial and deep

Fig. 8. (A) Chronic fungal sinusitis. Axial CT image without contrast at the level of the orbits. Notice the high-density material in
the ethmoid sinuses (asterisks). This is a chronic process and has caused expansion of the ethmoid sinuses, which resulted in
hypertelorism. The fungal sinusitis also caused erosion of the clivus (arrows). The high-density material in the ethmoid sinus is
characteristic of fungal sinusitis. The surrounding lower density material represents mucosa. (B) Axial T2-weighted MRI at the
same level as (A). Notice the signal void (asterisk), which corresponds to the high-density material seen on the CT examination.
This signal void could be mistaken for normal, aerated ethmoid sinus. The lack of signal is highly characteristic of fungal sinusitis.
 T.E. Underhill et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 39–49 47

cervical fasciae (fasciitis), enhancement and thick-
ening of the neck muscles (myositis), and fluid
collections in multiple neck compartments.
Sinusitis
The paranasal sinuses are a common source of
maxillofacial infection. Occasionally, paranasal
sinusitis can have devastating sequelae from involve-
ment of the orbits (Fig. 6), cavernous sinuses, or brain
parenchyma [4]. Orbital spread is most commonly
secondary to bacterial or fungal sinus infections.
Optic neuritis that arises from such infections requires
prompt recognition and aggressive treatment if vision
is to be preserved [27]. Intracranial empyema sec-
ondary to sinusitis is generally a disease of young
adult men without an antecedent history of sinus
disease [7]. The dominant clinical features are fever,
intense headache, and facial swelling. The patient
should receive an immediate noncontrasted and con-
trasted head CT and MRI. If CT is used, additional
thin cuts (3 mm or less) in the axial and coronal
planes of the paranasal sinuses should be obtained.
Because of the late development of the frontal
sinuses, frontal sinus infection in children is rare.
When present, it can lead to osteomyelitis associated
with forehead swelling (Pott’s puffy tumor) (Fig. 7).
Early diagnosis and active treatment prevent progres-
sion to life-threatening intracranial spread [28].
In fungal disease of the paranasal sinuses, expan-
sion of the bony walls occurs by increased mucus
secretion and fungal growth (Fig. 8). The fungus is
confined to the lumen and does not invade the
tissues. Within the lumen of the sinus, fungus
appears as areas of high attenuation on the CT scan
because of the presence of calcium and traces of
metallic elements. On MRI, the fungus-filled lumen
appears as an area of low signal intensity because of

Fig. 9. (A) Invasive fungal sinusitis. Coronal nonenhanced T1-weighted MRI in a 33-year-old immunocompromised man. There
is an oroantral fistula in the region where maxillary teeth were removed because of loss of all bony support. On removal of teeth,
there was no normal tissue with which to close the sinus opening. Abnormal tissue along the left infraorbital rim has eroded the
bone and replaced the normal bright orbital fat (asterisk). (B) Sagittal postcontrast T1-weighted MRI with fat saturation shows
the oroantral fistula (large arrows). Abnormal enhancing tissue extends back to the orbital apex along the expected route of the
infraorbital nerve (small arrows). Not imaged on this slice is the abnormal enhancing tissue that extends into the cavernous sinus.
This is an example of perineural spread of a fungus infection.
48 T.E. Underhill et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 39–49
the presence of ferromagnetic elements (Fig. 8) [29].
Invasive fungal sinusitis involves destruction of the
bony walls and spread into surrounding structures
(Fig. 9) [29].
Summary
Computed tomography and MRI are quick and
accurate methods for the evaluation of complex head
and neck infections. Because of new spiral CT
scanners and picture archival communication sys-
tems, multiplanar imaging is common with CT. CT
has supplanted sialography for the evaluation of
inflammatory processes in the salivary glands. In
the search for a drainable abscess collection, MRI
is superior to CT in regard to lesion conspicuity and
determining the number of anatomic spaces involved
and the degree of extension and the source. MRI is
also superior to CT for detection of intracranial
extension of infection. CT is superior to MRI in the
detection of intralesional gas, calcifications, and
cortical destruction.
References
[1] Ginsberg LE. Inflammatory and infectious lesions
of the neck. Semin Ultrasound CT MR 1997;18:
205 – 219.
[2] Munoz A, Castillo M, Melchor MA, Gutierrez RJ.
Acute neck infections: prospective comparison be-
tween CT and MRI in 47 patients. Comput Assist To-
mogr 2001;25:733 – 41.
[3] Nagy M, Backstrom J. Comparison of the sensitivity of
lateral neck radiographs and computed tomography
scanning in pediatric deep-neck infections. Laryngo-
scope 1999;109:775 – 9.
[4] Sahjpaul RL, Lee DH. Infratentorial subdural empye-
ma, pituitary abscess, and septic cavernous sinus
thrombophlebitis secondary to paranasal sinusitis: case
report. Neurosurgery 1999;44:864 – 6.
[5] Kim HJ, Park ED, Kim JH, Hwang EG, Chung SHJ.
Odontogenic versus nonodontogenic deep neck space
infections: CT manifestations. Comput Assist Tomogr
1997;21:202 – 8.
[6] Feldman DP, Picerno NA, Porubsky ES. Cavernous
sinus thrombosis complicating odontogenic paraphar-
yngeal space neck abscess: a case report and dis-
cussion. Otolaryngol Head Neck Surg 2000;123:
744 – 5.
[7] Fenton JE, Smyth DA, Viani LG, Walsh MA. Sino-
genic brain abscess. Am J Rhinol 1999;13:299 – 302.
[8] Nakamura S, Sadoshima S, Doi Y, Yoshioka M, Yama-
shita S, Gotoh H, et al. Internal jugular vein thrombo-
sis, Lemierre’s syndrome, oropharyngeal infection with
antibiotic and anticoagulation therapy: a case report.
Angiology 2000;51:173 – 7.
[9] Nguyen VD, Potter JL, Hersh-Schick MR. Lud-
wig angina: an uncommon and potentially lethal
neck infection. AJNR Am J Neuroradiol 1992;13:
215 – 9.
[10] Haraden BM, Zwemer Jr FL. Descending necroti-
zing mediastinitis: complication of a simple den-
tal infection. Ann Emerg Med 1997;20:683 – 6.
[11] Raboso E, Llavero MT, Rosell A, Martinez-Vidal A.
Craniofacial necrotizing fasciitis secondary to sinusitis.
J Laryngol Otol 1998;112:371 – 2.
[12] Tsunoda R, Suda S, Fukaya T, Saito K. Descending
necrotizing mediastinitis caused by an odontogenic in-
fection: a case report. J Oral Maxillofac Surg 2000;58:
240 – 2.
[13] Abrahams JJ, Glassberg RM. Dental disease: a fre-
quently unrecognized cause of maxillary sinus ab-
normalities? AJR Am J Roentgenol 1996;166:
1219 – 23.
[14] King JM, Caldarelli DD, Petasnick JP. DentaScan: a
new diagnostic method for evaluating mandibular
and maxillary pathology. Laryngoscope 1992;102:
379 – 87.
[15] Heiken J, Brink J, Vannier M. Spiral (helical) CT. Ra-
diology 1993;189:647 – 56.
[16] Miller WD, Furst IM, Sandor GK, Keller MA. A pro-
spective, blinded comparison of clinical examination
and computed tomography in deep neck infections.
Laryngoscope 1999;109:1873 – 9.
[17] Wetmore RF, Mahboubi S, Soyupak SK. Computed
tomography in the evaluation of pediatric neck in-
fections. Otolaryngol Head Neck Surg 1998;119:
624 – 7.
[18] Ariti E, Moriguchi S, Kuroki T, et al. Computed to-
mography of maxillofacial infections. Dentomaxillofac
Radiol 1991;20:147 – 51.
[19] Elster AD, Burdette JH. Introduction to nuclear mag-
netic resonance. In: Questions and answers in magnetic
resonance imaging. St Louis: Mosby; 2001. p. 37.
[20] Parhiscar A, Har-El G. Deep neck abscess: a retrospec-
tive review of 210 cases. Ann Otol Rhinol Laryngol
2001;110:1051 – 4.
[21] Chan WL, Fernandes VB, Carolan MG. Retropharyng-
eal abscess on a Ga-67 scan: a case report. Clin Nucl
Med 1999;24:942 – 4.
[22] Friedman NR, Mitchell RB, Pereira KD, Younis RT,
Lazar RH. Peritonsillar abscess in early childhood:
presentation and management. Arch Otolaryngol Head
Neck Surg 1997;123:630 – 2.
[23] Nusbaum AO, Som PM, Rothschild MA, Shugar JM.
Recurrence of a deep neck infection: a clinical indica-
tion of an underlying congenital lesion. Arch Otolar-
yngol Head Neck Surg 1999;125:1379 – 82.
[24] Robson CD, Hazra R, Barnes PD, Robertson RL, Jones
D, Husson RN. Nontuberculous mycobacterial infec-
tion of the head and neck in immunocompetent chil-
dren: CT and MR findings. AJNR Am J Neuroradiol
1999;20:1829 – 35.
 T.E. Underhill et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 39–49 49

[25] Conway W, Laine F, Blinder R. Newer diagnostic
imaging techniques. Oral Maxillofacial Surg Clin N
Am 1991;3:259 – 72.
[26] Henrich DE, Smith TL, Shockley WW. Fatal cranio-
cervical necrotizing fasciitis in an immunocompetent
patient: a case report and literature review. Head Neck
1995;17:351 – 7.
[27] Osguthorpe JD, Hochman M. Inflammatory sinus dis-
eases affecting the orbit. Otolaryngol Clin North Am
1993;26:657 – 71.
[28] Hore I, Mitchell RB, Radcliffe G, De Casso, Moxo C.
Pott’s puffy tumour: a rare cause of forehead swelling
in a child. Int J Clin Pract 2000;54:267 – 8.
[29] Michaels L, Lloyd G, Phelps P. Origin and spread of
allergic fungal disease of the nose and paranasal si-
nuses. Clin Otolaryngol 2000;25:518 – 25.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 51 – 58

Diagnosis and treatment of cervicofacial actinomycosis

Morton H. Goldberg, DMD, MDa,b,*
a
Department of Oral and Maxillofacial Surgery, University of Connecticut, USA
b
Department of Dentistry, Hartford Hospital, Seymour Street, Hartford, CT 06107, USA

Actinomycosis is an uncommon and frequently and their reproduction by typical bacterial fission
misdiagnosed infection of the cervicofacial region rather than by budding or by spores. Actinomyces
that may be acute or indolent. The term ‘‘actino- are anaerobic (whereas fungi are aerobic) and lack the
mycosis’’ derives from the Greek aktino (ray) and nuclear membrane of fungi and yeasts. Actinomyces
mykos (fungus). Historically, Langenback may have are destroyed by antibacterial agents such as penicil-
been the first to describe the disease in humans lin and erythromycin but are not affected by antifun-
(1848), although it had been considered to be a gal medications, such as amphotericin B (see box 1)
sarcoma in cattle 20 years earlier by Leblanc. Bol- [2,3].
linger (1870) coined the term ‘‘lumpy jaw’’ in its
bovine form, whereas Harz (1877) described the
appearance of a ray-like microorganism, which he Box 1. Taxonomy of the order
named Actinomycoses bovis. In 1878, Israel and Actinomycetales [2,38]
Ponfick observed and described ‘‘sulfur granules’’
in human disease, and in 1891, Israel and Wolf Class: Schizomycetes
isolated the anaerobic filamentous organism from Order: Actinomycetales
humans. In 1898, the organism found in humans Family: Actinomycetes
was named Actinomycoses israelii, and by the Genus: Actinomyces
1940s further research had confirmed that whereas * Species:
A. bovis was responsible for ‘‘lumpy jaw’’ in cattle, A. israelii
A. israelii was the etiologic agent of human disease A. neerlundi
[1 – 3]. A. viscosus
Originally considered to be a fungus (mykos) A. odontolyticus
because of its slow growth and filamentous appear- A. meyerii
ance, it is currently generally accepted that Actino- A. bovis
myces are bacteria and are taxonomically classified Family: Mycobacteriaceae
accordingly as members of the order Actinomycetales Genus: Mycobacterium
and the family Actinomycetacae. Other orders clas- Family: Nocardiaceae
sified as Actinomycetales include Mycobacteriaceae Genus: Nocardia
and Nocardiaceae. * A. israelii is the etiologic organism of
Evidence that Actinomyces are bacteria include most human actinomycosis,
their filaments, which are narrower than the hyphae but all species except A. bovis can
of fungi, their morphology, including bacillary forms, be commensals in the normal oro-
pharyngeal flora. Human disease is
occasionally reported with species
* Department of Dentistry, Hartford Hospital, 928 other than A. israeii [4].
Farmington Avenue, West Hartford, CT 06107.

1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 7 4 - 2
52 M.H. Goldberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 51–58

Microbiology and natural history

Box 3. Reported diagnostic sites of head
and neck infection by Actinomycoses
Actinomyces are endogenous to the human oral
[2,11,16,26 – 32,36,37,39 – 42]
cavity and cervicofacial region and are not restricted
to soil or soil contaminants as previously hypothe-
Dentigerous cyst
sized. The organism is a human commensal, but its
Dental caries
presence results only in a low incidence of clinical
Dental plaque
infection, even when tissue undergoes trauma, such
Dental pulp
as tooth extraction. This low incidence of infection,
Ear (external canal and middle ear)
despite the high incidence of oral trauma, argues for
Eye (lids, conjunctiva, lacrimal gland)
the organism’s minimal potential for invasion and its
Fascial spaces (submandibular, sub-
lack of virulence.
mental, temporal, masseteric,
Actinomyces require a polymicrobial ecosystem
buccal)
and tissue trauma to proliferate and create clinically
Gingiva
evident infection. These microflora work synergisti-
Larynx
cally to destroy vascularized aerobic tissue and create
Lip
poorly oxygenated granulation tissue, which becomes
Masseter muscle, mandible, maxilla
an environment that supports the growth and mul-
(exodontia, fracture)
tiplication of Actinomyces. The severity and chro-
Osteotomy site
nicity of tissue trauma are factors that can enhance
Palate
the development of anaerobic growth by these organ-
Parotid gland
isms, which normally exhibit low virulence. Other
Scalp
oral bacterial species involved in this ecosystem
Sinuses
include Actinobacillus, Actinomycetencomitans (fre-
Temporal bone
quent), Bacteroides, Eikenella, and Fusobacterium
Temporomandibular joint
(see box 2). Actinomyces have been isolated from
Thyroid
teeth, extraction sites, periodontal infections, jaw
Tongue
fractures, postoperative surgical sites, salivary glands,
Tonsils
the tongue, tonsils, and other areas of the orofacial
Tooth (extraction site)
and neck region (see box 3).
Trachea

Histopathology
nibacteria. The thin filaments of Actinomyces may be
Morphologically, Actinomyces may be filamen- observed to have V or Y branching and may appear in
tous or diptheroidal. If diptheroidal, they may bacillary or coccoid forms.
resemble nonpathogenic diptheroids such as Propio- Sulfur granules are macroscopic yellowish grains
of firm consistency that are visible to the eye when
on a gauze sponge. They vary in size from 100 to
Box 2. Bacterial ecosystem associated 1000 u in diameter and are visible by light micro-
with Actinomycoses infection scopy. When stained with methylene blue, Gram’s
[1,11,33 – 35] stain, periodic acid-Schiff, or silver methenamine,
gram-positive branching mycelia and filaments are
Actinobacillus actinomycetemcomitas visible (Figs. 1,2). The filaments may be mistaken
Bacteroides species (pigmented) for streptococci.
Capnocytophaga species Granules may not always be present or may be
Eikenella corrodens few in number, which necessitates multiple micro-
Fusobacterium species scopic sections of sinus tracts or abscess walls for
Haemophilus species their identification. One study found only one to three
Peptostreptococcus granules in 56% of specimens examined and none at
Porphymonas all in seven cases, which were confirmed by culture
Streptococcus [5]. Although granule formation is common in A.
Staphylococcus gingivalis israelii infection, it is not observed with A. odonto-
lyticus, a noninvader of tissue. The presence of
 M.H. Goldberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 51–58 53

Fig. 1. Low magnification of sulfur granule. Note amorphous center and peripheral rosette of filaments. (Courtesy of Dr. Ellen
Eisenberg, University of Connecticut.)

granules in exudate from an infection in the cervico-
facial region, although highly suggestive of actino-
mycosis, is not, of itself, pathognomonic, because
granules may be produced by Actinobacillus ligniersi
and some pathogenic mycoses, such as Sporotrichum
and Phialophore [6], and are found in nocardial
infections. Although fluorescent antibody staining
may contribute to an accurate diagnosis, confirmation
by culture is necessary.
Microscopic examination of an excised or biop-
sied mass of tissue can reveal acute or chronic
inflammation, with loculation of neutrophils, lym-
phocytes, plasma cells, macrophages, and dense fib-
rous tissue. The number of plasma cells increases
with the chronicity of the lesion. Fibrosis may be
mild in early infections, whereas chronicity is asso-
ciated with intense avascular fibrosis. This tissue is
believed to maintain the anaerobic environment that

Fig. 2. Higher magnification of edge of sulfur granule, which demonstrates individual and branching filaments. (Courtesy of Dr.
Mark Fletcher, Hartford Hospital.)
54 M.H. Goldberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 51–58
stimulates growth of the actinomyces organisms and
inhibits antibiotic penetration [7,8].
Clinical manifestations
Actinomycosis is an uncommon disease. It is
possible for an oral maxillofacial surgeon to have a
lengthy career and never encounter the disease. It
may occur at any age, however, most commonly
between age 30 and 60. Cases among men outnumber
cases that involve women 4:1. In the pediatric popu-
lation, actinomycosis is rare, and a low level of
suspicion may cause diagnostic confusion and delay
in treatment [9].
The disease may present as either an acute or
chronic form. The less common acute form manifests
as a fluctuant swelling that resembles or mimics an
acute odontogenic infection. It may be painful and
associated with temperature elevation, and it can
spread rapidly through tissue [10]. More common,
however, is the chronic lesion of actinomycosis,
which presents as a slowly enlarging, progressive,
inflammatory mass that may or may not be painful
and usually is associated with an afebrile course or
minimal elevation of temperature. The chronic or
chronic-recurring form may develop or progress or
recur over weeks, months, or even years, during
which time lethargy may be present, but patients do
not typically complain of feeling ill. Sedimentation
rates tend to remain low or normal, whereas leuko-
cyte counts may be normal or minimally elevated.
Reddish-brown discoloration of the facial skin over-
lying the jaw is common, and recurrent episodes of
suppuration result in firm or fluctuant irregular skin
masses; hence the historic description of the infection
as ‘‘lumpy jaw.’’ Spread of the infection is indepen-
dent of lymphatic channels or fascial planes. Ulti-
mately, sinus tracts develop deep in the lesion,
burrowing through the dense fibrosis and discharging
onto the surface of the skin (or mucosa). These sinus
tracts may persist or close, or new tracts may form. A
pattern of remission and exacerbation may character-
ize the disease, especially if the diagnosis remains
obscure and short-term episodic antibiotic therapy is
performed [2,11,12].
A primary actinomycotic bone infection is uncom-
mon but may occur in up to 12% of cases, because
osseous invasion and destruction may result from
spread of the infection from adjacent soft tissues,
extraction sites, or fractures sites. Such infections
occur more commonly in the mandible than the
maxilla (4:1) [13 – 16] (see box 4). Periostitis with
evidence of periosteal elevation may be the earliest
radiographic finding. True actinomycotic osteomyel-
itis may demonstrate imaging evidence of cortical
erosion or localized lytic destruction in the midst of
increased bone density. Sinus tracts may develop
from the bone to the surface. Radiographically, acti-
nomycotic osteomyelitis may be diagnostically inter-
preted as suppurative osteomyelitis or neoplasm.
Box 4. Differential diagnosis of osteolytic
lesions of the jaws associated with
‘‘lumpy jaw’’ or sinus tracts
Infection:
 Pyogenic osteomyelitis
 Periapical infection
 Tuberculosis
 Nocardiosis or other fungal
infection
 Infected hardware
Tumor:
 Benign bone tumors
 Primary or metastatic malignant
bone tumors
 Gingival carcinoma invading bone
The recovery of Actinomyces and Eikenella from
cultures of chronic diffuse sclerosing osteomyelitis
by Marx et al suggests that a pathogenic mutalism
exists between these organisms and perhaps with
Arachnia. This research must be repeated and
expanded [17,18].
Although cervicofacial infection may account for
60% of all cases of actinomycosis, other anatomic
sites or organ systems may be infected, including the
thoracic cavity (15%) and the abdominal cavity
(20%), where it may surprise the unsuspecting physi-
cian or surgeon. Superficial cutaneous infection and
genitourinary tract infection also may occur. Central
nervous system infection is uncommon (2%) and can
originate by direct extension of cervicofacial infec-
tion via the sinuses, orbits, or auditory canals and by
perineural pathways [19]. It can gain access to the
trigeminal ganglion via the foramen ovale. Meningeal
spread may result in diffuse meningitis or brain
abscess, and it may be initially misdiagnosed as a
tumor. Central nervous system involvement may be
the result of hematogenous spread from the oral,
pulmonary, abdominal, or pelvic regions. Dissemi-
nated hematogenous ‘‘sepsis’’ is rare but may be
fatal. Focal neurologic signs and elevated intracranial
pressure should alert the examiner to this possibility.
A delay in diagnosis of 2 months or more caries a
 M.H. Goldberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 51–58
mortality rate of 28%, even when adequate antibiotic
therapy is ultimately instituted.
The most common diseases occur most frequent-
ly; therefore the diagnostician is more likely to be
correct when looking for ‘‘horses’’ rather than
uncommon or unlikely ‘‘zebras.’’ Some diseases
masquerade or mimic other diseases—‘‘horses wear-
ing striped pajamas’’ [20]. Actinomycosis that
mimics tumors of the maxillary sinus or neural neo-
plasms is an example of this phenomenon [21,22].
Diagnosis
Although history and physical examination are
always basic to diagnosis of any disease, the indolent
but progressive natural process of actinomycotic infec-
tion requires a high index of suspicion of what is often
a diagnostic dilemma. With a history of recurrent or
late infection after tooth extraction and the presence of
one or more sinus tracts not associated with high fever
or elevated neutrophil count, actinomycosis should be
high on the differential diagnosis list.
Radiographs of bone may be helpful in recognizing
the extent of the osseous process but are nonspecific
for actinomycosis even in the presence of soft tissue
swelling and sinus tract. CT and gallium scintigraphy
may be useful in differentiating between inflammatory
and neoplastic changes. CT may reveal loculations and
dense fibrosis, whereas scintigraphy may be useful as a
method to determine the efficacy of therapy; however,
no single imaging modality is diagnostic.
Absolute confirmation of the diagnosis depends
on careful culturing of these fastidiously oxygen-
sensitive anaerobes, preferably when the patient has
received no antibiotics for 7 to 10 days. If culturing is
delayed, improperly performed, or negated by recent
or concurrent antibiotic therapy, the diagnosis may
remain obscure or uncertain.
Culturing may be performed directly from sinus
tract drainage, an open biopsy of tissue, or after
needle aspiration of the lesion [23]. The clinician
must be fastidious in technique to avoid contamina-
tion by skin or oral flora and prevent excessive
exposure of the specimen to room air. Even with
these precautions, recovery of the pathogen is com-
monly unsuccessful, often in more than 50% of
specimens. All cultures should be transported imme-
diately to the microbiology laboratory in a nonaerobic
medium with a carbon dioxide atmosphere. If the
laboratory has been alerted that Actinomyces is sus-
pected as the pathogen, appropriate culture media in
an anaerobic environment can be available, without
unnecessary delay, which contributes to a higher
55
percentage of positive identification. Immediate
Gram stain of the specimens may yield a presumptive
diagnosis of Actinomyces, especially if sulfur gran-
ules have been included [2].
Culture plates must be incubated at least 14 days,
under completely anaerobic conditions, in a 5%
carbon dioxide atmosphere at 37
C. Blood agar,
brain-heart infusion agar, and thioglycolate liquid
medium are commonly used. During the incubation
period, direct observation of the cultures may reveal
sulfur granules, which can be Gram stained, a tech-
nique that adds evidence to the final diagnosis [2,11].
The fine-needle aspiration technique permits eas-
ier anaerobic transport and rapid inoculation into
anaerobic culture media. Gram staining of aspirated
material may reveal the filamentous appearance of the
Actinomyces. Aspiration is most successful when
easily palpable tissue masses can be aspirated accu-
rately rather than when a blind attempt is made to
obtain material from deep tissue.
Classic microscopic examination of excised tissue
(biopsy) can be useful diagnostically but has limita-
tions, especially if intense inflammation is present in
the tissue in the early stages of the infection. Histo-
logically, it is difficult to differentiate Actinomyces
from Nocardia species. Small colonies or isolated
filaments of Actinomyces may be overlooked in
routine hematoxylin and eosin – stained tissue sec-
tions. Other diagnostic tests include fluorescent typ-
ing and gel-diffusion techniques, but these are not
routinely available in many laboratories.
It is important for clinicians, pathologists, and
microbiologists to consider that the presence of
A. israelii in tissue specimens or even in cultures does
not unequivocally confirm its role as the etiologic
agent in an infection; it may represent contamination
or commensal growth. Its presence should correlate
with clinical findings.
Therapy
Before the availability of antibiotics, treatment of
cervicofacial actinomycosis included surgery, irra-
diation, vaccination, and the use of iodides and other
ineffective nostrums. The successful use of sulfona-
mides in 1938 simplified the therapy. Penicillin was
introduced into the therapeutic regimen for actino-
mycosis in 1948, and by 1960, prolonged high-level
penicillin therapy was being used, as was the radical
excision of recalcitrant fibrotic lesions and persistent
sinus tracts that resisted antibiotic penetration [24].
Later, Lerner recommended the use of intravenous
antibiotics alone for early lesions, whereas resection
56 M.H. Goldberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 51–58
was advised for more mature and densely fibrotic
lesions [6,11]. Perhaps the efficacy of penicillin
therapy is related not only to its direct effect on A.
israelii but also to its antimicrobial suppression of the
other microorganisms that participate in the metabolic
ecosystem and enable Actinomyces to flourish.
In actinomycotic osteomyelitis, sequestrectomy or
saucerization of nonvital bone is indicated. Débride-
ment of the dead bone, an excellent anaerobic growth
medium that lacks any vascularity and is unpenetra-
ble by antibiotics, contributes to the healing process.
Initial therapy for actinomycosis is high-dose in-
travenous infusion of penicillin. Three to 12 million U
daily is recommended and can be administered at
home by indwelling heparin-lock intravenous cath-
eter. This treatment should be followed by orally
administered penicillin, 2 to 4 g daily, for an addi-
tional 3 to 12 months, depending on the infection’s
response [16,25]. This therapeutic regimen may be
extended if central nervous system infection is pre-
sent. If actinomycotic osteomyelitis has developed,
intravenous penicillin doses may reach 20 million U
daily, given in divided doses. If potassium penicillin is
used in high doses for prolonged periods, potassium
levels should be monitored and one should observe for
hyperkalemia, especially in patients who are simulta-
neously taking potassium-sparing diuretics or who
have impaired renal function.
Other antibiotics that are effective include eryth-
romycin, cephaloridine, minocycline, clindamycin,
chloramphenicol, and imipenem. Flagyl (metronida-
zole) and aminoglycosides are ineffective against
A. israelii (Table 1).
If the clinical response to monoantibiotic therapy
is less than optimal, subsequent options include sur-
gical exploration and excision of undrained obscure
abscesses or the use of antibiotics in combination.
Polyantibiotic therapy may be successful, perhaps
Table 1
Antibiotic therapy for actinomycosis [2,11]
Antibiotic Mode MIC mg/mL Daily dose
Penicillin G IV 0.03 – 0.5 3 – 12
(soft tissue infection) million U
Penicillin G (osseous IV 0.03 – 0.5 12 – 20
and central million U
nervous system)
Penicillin V PO 0.03 – 0.5 2–4 g
Erythromycin PO 0.12 2g
Clindamycin PO 2.0 – 8.0 1200 mg
Tetracycline PO — 2g
Minocycline PO — 2g
Abbreviations: IV, intravenous; PO, orally.
because penicillin-resistant organisms develop in the
polymicrobial ecosystem that supports the growth of
actinomycoses. After successful therapy, pharma-
ceutical or surgical, secondary repair or reconstruction
may be indicated. Bone loss, absence of teeth, and
aesthetically unacceptable scarring from chronic si-
nus tracts can be treated by bone grafting, soft
tissue procedures, or implants, but only when the sur-
geon is confident that the infection has been com-
pletely eliminated.
Medicolegal considerations
Because A. israelii is an oral commensal, actino-
mycotic infection cannot be predicted or prevented. It
would seem that malpractice claims secondary to
infection are specious and without scientific validity.
Nonetheless, the author has reviewed such cases,
which are based on the claims that practitioners have
failed to diagnose or treat appropriately.
The decision whether to prescribe antibiotics at
the time of tooth extraction should not be based on
the statistically highly unlikely possibility that the
patient might eventually develop actinomycosis. Nei-
ther routine culturing nor routine antibiotic use after
exodontia can be justified medically or economically.
The most common causes of early postextraction
infection are food trapping (inadequate hygiene) in
the operative site or the presence of a small fragment
of nonviable bone. Such infections are usually treated
with irrigation, débridement, and antibiotics if, in the
surgeon’s judgment, they are indicated. Late or mul-
tiple infections should alert the surgeon to the pos-
sibility of actinomycotic infection, however. Failure
to culture late or recurrent infections and the use of
multiple short courses of empirical antibiotic therapy
before the eventual establishment of the correct
diagnosis are the common filaments in the fabric of
actinomycosis malpractice actions. Without a high
index of suspicion, clinicians may be lulled into a
false sense of security because the infections may be
initially subtle, nonpyogenic, and indolent and
because actinomycosis may become evident months
after the surgery or after years of quiescence.
Summary [5,7,8]
Actinomyces israelii is an anaerobic commensal
microorganism of the human oral cavity. It infre-
quently causes an infection that is polymicrobial and
mutalistic in nature. The infection, actinomycosis,
has a variable clinical course, which may be acute
 M.H. Goldberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 51–58
or indolently chronic. The differential diagnosis
includes pyogenic and fungal infections, tuber-
culosis, and neoplasm. A high diagnostic index of
suspicion for actinomycosis when late or recurrent
postextraction infection occurs benefits the patient
and the practitioner. A prolonged course of penicillin
therapy remains the contemporary treatment of
choice. Excisional or secondary restorative surgery
may be necessary.
References
[1] Holm P. Studies on the etiology of human actinomy-
cosis. I: The ‘‘other microbes’’ of actinomycosis and
their importance. Acta Pathol Microbiol Scand 1950;
27:736 – 42.
[2] Miller M, Haddad AJ. Cervicofacial actinomycosis.
Oral Surg Oral Med Oral Pathol 1998;85:496 – 508.
[3] Richtsmeier WJ, Johns ME. Actinomycosis of the
head and neck. CRC Crit Rev Clin Lab Sci 1979;11:
175 – 202.
[4] Pordy RC. Lumpy jaw due to Actinomycoces meyerii:
report of the first case and review of the literature. Mt
Sinai J Med 1988;55:190 – 3.
[5] Brown JR. Human actinomycosis: a study of 181 sub-
jects. Hum Pathol 1973;4:319 – 30.
[6] Lerner PI. Actinomyces and arachnia. In: Wonsiewicz
MJ, editor. Infectious Diseases. Philadelphia: WB Sa-
unders Co.; 1992. p. 1626 – 32.
[7] Bennhoff DF. Actinomycosis: diagnostic and therapeu-
tic considerations and a review of 32 cases. Laryngo-
scope 1984;94:1198 – 217.
[8] Dusek JJ, Howe AG, Carr RF, et al. Case 37, part II:
cervico-facial actinomycosis [clinicopathological con-
ference]. J Oral Maxillofac Surg 1982;40:113 – 6.
[9] Foster SV, Demmier GJ, Hawkins EP, et al. Pediatric
cervico-facial actinomycosis. South Med J 1993;86:
1147 – 50.
[10] Nielsen PM, Novak A. Acute cervicofacial actinomy-
cosis. Int J Oral Maxillofac Surg 1987;16:440 – 4.
[11] Lerner PI. The lumpy jaw: cervicofacial actinomycosis.
Infect Dis Clin N Am 1988;2:203 – 19.
[12] Schuster GS. Bacterial and protozoal infections with
oral manifestations. In: Topazian R, Goldberg M, edi-
tors. Oral and maxillofacial infections. 3rd edition. Phi-
ladlphia: WB Saunders Co.; 1994. p. 557 – 77.
[13] Gupta DS, Gupta MK, Naidu NG. Mandibular osteo-
myelitis caused by Actinomyces israelii. J Oral Max-
illofac Surg 1986;14:291.
[14] Lewis RP, Sutter VL, Finegold SM. Bone infections
involving anaerobic bacteria. Medicine (Baltimore)
1978;57:279 – 305.
[15] Schaal KP, Beaman BL. Clinical significance of actino-
mycetes. In: Goodfellow M, Mordarski M, Williams
ST, editors. The biology of the Actinomycetes. New
York: Academic Press; 1983. p. 389 – 424.
57
[16] Topazian R. Osteomyelitis of the jaws. In: Topazian R,
Goldberg M, Hupp J, editors. Oral and maxillofacial
infections. 4th edition. Philadelphia: WB Saunders Co.;
2002. p. 231 – 4.
[17] Jacobsson S. Isolation of Actinomyces species and Ei-
chenella corrodens from patients with chronic diffuse
sclerosing osteomyelitis [discussion]. J Oral Maxillo-
fac Surg 1994;52:33 – 4.
[18] Marx RE, Carlson ER, Smith BR, et al. Isolation of
Actinomyces species and Eichenella corrodens from
patients with chronic diffuse sclerosing osteomyelitis.
J Oral Maxillofac Surg 1994;52:26 – 33.
[19] Lad SD, Chandy MJ. Craniofacial actinomycosis. Br J
Neurosurg 1991;5:361 – 70.
[20] Goldberg MH. Zebra rounds and horses wearing
striped pajamas. Pharos 2001;64:53.
[21] Har-El G, Prager DA, DeSoto La Paix F, et al. Actino-
mycotic granuloma masquerading as infraorbital nerve
neoplasm. Head Neck 1990;12:261 – 2.
[22] Pradhan S, Datta NR, Prasad KN, et al. Actinomycosis
mimicking carcinoma of the maxillary sinus. Indian J
Cancer 1993;30:1 – 4.
[23] Das DK. Actinomycosis in fine-needle aspiration cy-
tology. Cytopathology 1994;5:243 – 50.
[24] Peabody JW, Seabury JH. Actinomycosis and nocar-
diosis. Am J Med 1960;28:99 – 115.
[25] Goldberg MH. Antibiotics: old friends and new
acquaintances. Oral Maxillofacial Surg Clin N Am
2001;13:15 – 30.
[26] Barnard NA, Magennis JPM. Intra-masseteric actino-
mycosis: report of a case. Br J Oral Maxillofac Surg
1992;30:190 – 1.
[27] Bradley P. Actinomycosis of the temporomandibular
joint. Br J Oral Surg 1971;9:54 – 6.
[28] Carrau RL, Greenwall K, Canaan RE, et al. Actino-
mycosis of the infratemporal fossa. Am J Otolaryngol
1993;14:1 – 4.
[29] Chuong R, Goldberg M. Case 60: preauricular mass
(clinicopathological conference). J Oral Maxillofac
Surg 1986;44:214 – 7.
[30] Deloach-Banta LJ, Barber FA. Nonhealing masses of
the right cheek and submandibular areas. Arch Derma-
tol 1991;127:1831 – 4.
[31] Fenton R, Rotenberg D. Actinomycosis of the parotid.
Is it usually a masticator space infection? J Otolaryngol
1977;6:233 – 8.
[32] Ficarra G, DiLollo S, Pierleoni F, et al. Actinomyces of
the tongue: a diagnostic challenge. Head Neck 1993;
15:53 – 5.
[33] Jordan HV, Kelly DM. Persistence of associated gram-
negative bacteria in experimental actinomycotic lesions
in mice. Infect Immunol 1983;40:847 – 9.
[34] Jordan HV, Kelly DM, Heeley JD. Enhancement of
experimental actinomycosis in mice by Eikenella cor-
rodens. Infect Immunol 1984;46:367 – 71.
[35] Lerner PI. Susceptibility of pathogenic Actinomycetes
to antimicrobial compounds. Antimicrob Agents Che-
mother 1974;5:302 – 9.
[36] Nakamatsu K, Shinohara M, Takenoshita Y, et al. Two
58 M.H. Goldberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 51–58
cases of actinomycosis of the submandibular region.
Jpn J Oral Maxillofac Surg 1986;32:68 – 74.
[37] Ozaki W, Abubaker AQ, Sotereanos GC, et al. Cervi-
cofacial actinomycosis following sagittal split ramus
osteotomy: a case report. J Oral Maxillofac Surg
1992;50:649 – 52.
[38] Rippon JW. Medical mycology. In: The pathogenic
fungi and the pathogenic actinomycetes. 3rd edition.
Philadlphia: WB Saunders Co.; 1988. p. 30 – 52.
[39] Roth M, Montone KT. Actinomycosis of the paranasal
sinuses: a case report and review. Otolaryngol Head
Neck Surg 1996;114:818 – 21.
[40] Shaheen S, Ellis F. Actinomycosis of the larynx. J R
Soc Med 1983;76:226 – 8.
[41] Weir J, Buck W. Periapical actinomycosis. Oral Surg
1982;54:336 – 40.
[42] Zajac I, Oribovac Z, Bagatin M. Temporal actinomy-
cosis: report of a case. J Oral Maxillofac Surg 1999;57:
1370 – 2.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 59 – 67
Diagnosis and treatment of necrotizing fasciitis in the head
and neck region
Mark McGurk, MD, FRCS, DLO, FDSRCS
Department of Oral and Maxillofacial Surgery, Guy’s, King’s, and St. Thomas’ Dental Institute, Floor 23 Guy’s Tower,
Guy’s Hospital, London Bridge, London SE1 9RT, UK
A rare event encountered in current medical prac-
tice is necrotizing fasciitis. Its presence is occasionally
brought to public attention by the press through head-
lines such as ‘‘Killer or Flesh-Eating Bug’’ [1], when
the impression is conveyed of the advent of a newly
discovered disease. Nothing could be further from the
truth because the condition was well known to the
military surgeons of past centuries. In his book ‘‘Lec-
tures on Inflammation,’’ J. Thomson, the Regius
Professor of Military Surgery at Edinburgh [2], attri-
buted the first description of the disease to L. Gillespie,
who was a surgeon in the Royal Navy [3]. The
condition was known to Amboise Pare and can be
found in the writings of most army surgeons who kept
careful records of their experiences. Knowledge of the
disease can be traced back to Hippocrates, who gave a
classic description of the disease process:
‘‘Sometimes a very small wound broke out and if
such an accident was neglected great inflammation
took place. In most of them the abscess ended in
suppurations and there was great falling off of the
flesh, tendons and bones; and the defluxion which
seated in the parts was not like pus, but a sort of
putrefaction and the running was large and of various
characters. About the head these things were
accompanied by falling off of the hairs of the head
and chin, the bones were laid bare and separated and
there were excessive runnings; and these symptoms
happened in fevers and without fevers.’’ [4]
In modern practice, necrotizing fasciitis occurs
sporadically, which makes a true estimate of its
prevalence difficult to ascertain. Invasive streptococ-
E-mail address: mark.mcgurk@kcl.ac.uk
1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 8 3 - 3
cal infections are monitored nationally, and in the
5-year period from1989 to1994, 160 cases of necro-
tizing fasciitis were reported in England and Wales.
Most were attributed to group A streptococci [1].
Because many cases are caused by polymicrobial
infection, however, this was almost certainly an
underestimate. In the United States, an estimated
10,000 to 15,000 cases of invasive group A strep-
tococcal infections occur annually, of which 5% to
10% are necrotizing fasciitis, with a case fatality of
28% [5]. A prospective population-based study of
group A streptococcal necrotizing fasciitis conducted
in Ontario between November 1991 and May 1995
showed that the incidence increased from 0.085 per
100,000 population in the first 12 months to 0.4 per
100,000 population in the last year of the study [6].
This pattern mirrors a world trend increase in group A
streptococcal infections since 1980 [7].
Typically, necrotizing fasciitis occurs on the abdo-
men/perineum or lower limbs after trauma or surgery.
In a few cases (1% – 10%), however, it occurs in the
head and neck region, particularly when the patient’s
health is already compromised. Delayed diagnosis is
a common event because the condition can arise
unexpectedly out of a seemingly trivial infection or
injury. The defining characteristic is rapid, progress-
ive tissue destruction that is disproportionate to the
initial clinical signs and symptoms.
Despite modern advances in medicine, this dis-
ease presents similar problems currently just as it did
200 years ago. Necrotizing fasciitis is a clinical
syndrome rather than a pathologic entity. Mortality
rates have been reported in the region of 40%.
Success depends on prompt diagnosis and treatment
without delay for microbiologic confirmation.
60 M. McGurk / Oral Maxillofacial Surg Clin N Am 15 (2003) 59–67
An historical perspective
The identification of disease entities from histor-
ical texts can be difficult because of changing nomen-
clature and vague clinical descriptions that lack
diagnostic detail; however, necrotizing fasciitis was
described clearly in the late eighteenth century by
Claude Pouteau, chief surgeon to the Hotel Dieu in
Lyon in 1783 [8]. At that time the disease complex
was given many names, such as ‘‘malignant ulcer,’’
‘‘gangrenous ulcer,’’ ‘‘putrid ulcer,’’ ‘‘phagedenis
ulcer,’’ ‘‘phagedena gangraenosa,’’ and ‘‘hospital
gangrene.’’ In the late eighteenth century, a series
of outbreaks affected the British Home Fleet. In
confined quarters the disease could spread quickly.
On HMS San Josef the surgeon observed ‘‘. . .an ulcer
that had devoured the one side of a sailor’s face,
which had followed a blow on the ear, that was
attended by a very slight wound’’ [9]. In the early
nineteenth century, the disease was reported from
military hospitals by the name of ‘‘hospital gan-
grene’’ or ‘‘phagedena gangraenosa.’’ In one case,
’’half the cranium was denuded, the bones having
become as black as charcoal; in another the neck was
denuded to expose the trachea.’’ The characteristic
features of the disease were as follows:
 extreme rapidity with which the disease pro-
gressed (measured in hours), which distin-
guishes it from standard gangrene;
 a tendency to turn soft parts into a putrid, pulpy
substance;
 severe pain together with a smell, which was
peculiar and extremely offensive;
 starting at the site of a wound or following a
trivial scratch and attacking young and healthy
persons and debilitated soldiers.
The disease was recorded in the Gendarmerie
Hospital at Brussels after Waterloo [10], and Miss
Nightingale noted 80 cases in 1 month at Scutari
[11]. The disease was well known to the surgeons
in the American Civil War (Fig. 1), and Joseph
Jones (Confederate Army surgeon) is credited with
the first clear investigation and characterization of
hospital gangrene [12]. It is not possible to discrim-
inate between the different types, but 2642 cases
were reported, of which 1142 were fatal (Table 1).
A serious outbreak occurred in September 1862 in
the hospitals at Fredrick and West Philadelphia after
the battles of South Mountain and Antietam. In
January 1863, because of the poor prisoner of war
sanitation in Richmond, three outbreaks occurred
as the sick were transferred to Annapolis. The dis-
Fig. 1. Private Milton Warren, aged 41 years, of the 1st
Kentucky Cavalry was captured in August 1863 at the
Cumberland River. While in captivity at Richmond he was
shot in the right elbow. An amputation was required on June
1, 1864. On June 20, the stump looked inflamed, and by
June 24, the whole stump and bone were exposed. He was
treated with charcoal and yeast poultices, a generous diet,
and ale. By August the sloughing process had stopped and
he survived to claim his pension in 1873.
ease was also present at the Douglas hospital in
Fredricksburg [13].
In civilian life it was much less common and
occurred sporadically in clusters, being much less
contagious. The hospital surgeon of London knew a
form of it as a genital disease that was said to be
confined to prostitutes and the destitute, and a few
cases were admitted to St. Bartholomew’s Hospital
London [14], where the disease was aggressive and if
unchecked ‘‘. . .involves in its ravages the vagina,
perineum and anus and sometimes even the bladder
and uterus.’’ Fournier’s classical description of the
condition was of phagedena of the penis and scrotum
[15], but case histories demonstrated that it was the
same disease process as reported by the military
surgeons [16].
Sporadic cases continued to be reported into the
early twentieth century. An American surgeon
reported a hospital outbreak in Peking, where it was
more common than in the West [17]. Melaney iso-
lated a hemolytic Streptococcus from the wounds,
and his name subsequently was associated with the
disease (Melaney’s gangrene). In 1952, Wilson
coined the name ‘‘necrotizing fasciitis,’’ which
described the main feature of the disease and empha-
sized the polymicrobial nature of some of the infec-
tions. Currently, it occurs as unexpected isolated
attacks so that few oral and maxillofacial surgeons
have any experience with it. The effects remain as
devastating as ever if not checked, however. This fact
is illustrated by the aftermath of the Nevado del Ruiz
volcano eruption in Colombia, where 38 patients
 M. McGurk / Oral Maxillofacial Surg Clin N Am 15 (2003) 59–67
Table 1
Cases of gangrene by site and mortality recorded in the War
of Rebellion
No. of cases Fatal cases of
Site of gangrene (n) gangrene (n)
Head and neck 60 23 (38%)
Trunk 216 129 (60%)
Upper extremity 844 295 (35%)
Lower extremity 1522 695 (46%)
From Barnes J. Gangrene. In: Medical and surgical history of
the War of Rebellion. 1867; Part III, Vol II Surgical History.
Government Printing Office. Washington, DC. p. 823 – 51.
developed necrotizing fasciitis, and the disease proved
fatal in 47.7% of those people.
Microbiology
Melaney was the first to associate group A strep-
tococcal infections with severe necrotizing condi-
tions, and this organism should be the principal
suspect in any rapidly progressive necrotizing infec-
tion. Because culture techniques have improved,
however, it has become clear that most necrotizing
wounds sustain a mixture of bacteria working syn-
ergistically. Various bacterial strains may dominate
different wounds, but essentially necrotizing fasciitis
may be categorized into three types according to
the causative organism. (1) In cool and temperate
climates it tends to be associated with group A
b-hemolytic streptococci (Streptococcus pyogenes)
[6] alone or with Staphylococcus aureus. They are
the only bacteria that seem to be able to generate
solely this clinical picture. Serotypes M1 and M3 are
the most common S. pyogenes serotypes associated
with invasive disease [7], but multilocus sequence
typing has confirmed that a genetically diverse range
of strains is associated with these infections [18]. (2)
In many cases (up to 60%) the necrotizing fasciitis
may be polymicrobial, including one or more obligate
anaerobes [19,20]. Brook and Frazier [21] reviewed
87 cases of necrotizing fasciitis over a 17-year period.
Of these cases, only 4 were monoinfections with
S. pyogenes. In the remaining cases, anaerobic bac-
teria were predominant, with Peptostreptococcus,
Prevotella, Porphyromonas, Bacteroides, and Clos-
tridium the commonest genera isolated. Facultative
anaerobic bacteria, such as Enterobacteriacae, are
also important. Up to 11 bacterial species have been
cultured with various streptococci (groups B and F) in
attendance, not just group A [21]. (3) In tropical
climates, the condition can be caused by members
of the family Vibrionacae, which are of seawater
origin [22]. In Colombia, the dominant genus was a
61
mycosis that was particularly virulent and proved
lethal in 70% of cases [23].
Streptococcus pyogenes produce several virulence
factors that are likely to be involved in necrotizing
fasciitis, including the extracellular pyrogenic exo-
toxins A, B, and C together with other exotoxins
and superantigens [24]. Given this powerful viru-
lence armory, it is perhaps surprising that invasive
S. pyogenes infections are relatively rare. One reason
for this might be the demonstration that mutations in
the two-component CsrS/CsrR 2-component regula-
tory system led to increased virulence in a mouse
model [25]. It could be hypothesized that exotoxin
production by S. pyogenes is normally tightly regu-
lated but that when that control is lost through
mutation in the regulatory gene, a hypervirulent
phenotype results. The pathogenesis of the polymi-
crobial form of the infection is unclear, although it is
well known that consortia of bacteria work together
to evade the host defenses and cause tissue damage.
Host factors may predispose to the rapid spread of
some infections.
Classification, pathogenesis, and clinical features
of necrotizing fasciitis
The history of necrotizing fasciitis has been domi-
nated by bacteriology. Pruitt [26] and Gorbach et al
[27] produced a bacterial classification, each of which
comprised five entities, and Simmons [28] produced
a third classification, which incorporated seven. Even
a simplified clinical classification based on necro-
tizing cellulitis, necrotizing fasciitis, and myonecrosis
[23] is difficult to adopt because these entities are rare
and clinical experience is unavailable to distinguish
between them. In practice this is not important
because the initial treatment for all rapidly progres-
sive necrotizing infections is the same: wide surgical
débridement. Attempts to subclassify the disorder are
unnecessary [29] and may be a disadvantage if it
leads to delay in surgery (even 24 – 48 hours) to
obtain culture results. This principle greatly simpli-
fies the approach to clinical management. Ultimately,
success depends on rapid diagnosis of the early lesion
and prompt treatment.
Pathogenesis
The overarching feature of necrotizing fasciitis is
a rapid, progressive liquefaction of the subcutaneous
fat and connective tissue below a relatively normal
looking skin surface. The fascial planes disintegrate,
and with the ensuing necrosis come edema and the
62 M. McGurk / Oral Maxillofacial Surg Clin N Am 15 (2003) 59–67

Fig. 2. (A,B) Necrotizing fasciitis arising from an infected tooth. The infection spread relentlessly, first into the neck and then the
chest wall. (C) The patient became systemically unwell. (D,E) After wide débridement the patient made a full recovery and is
married, with her own family. (Courtesy of Mr. P. McAndrew.)

release of tissue fluid. Early in the development of
the disease the veins that traverse the liquefying
subdermal fat become inflamed and start to throm-
bose, which gives the skin first a red and then a
mottled color. Later the arterial supply is also jeop-
ardized and the skin becomes pale, which leads to
necrosis and wet (coliquative) gangrene. The bacteria
initiate an acute local inflammatory response within
the dermis that is characterized by an intense poly-
morphonuclear infiltrate, focal necrosis, and micro-
 M. McGurk / Oral Maxillofacial Surg Clin N Am 15 (2003) 59–67
Fig. 2 (continued ).
abscess formation. The histologic picture is one of
arteriolar and venous thrombosis of the subcutaneous
fat, whereas the adjoining muscle shows compara-
tively little inflammation.
Clinical features
The rate of necrosis is disproportionate to the
signs and symptoms of infection. A small wound
can be painful. Early systemic symptoms may be
subtle and amount to little more than a feeling of
malaise or tachycardia. If there are systemic symp-
toms in the presence of an apparently innocuous
wound, however, necrotizing fasciitis should be con-
sidered early. The incidence of this disease increases
with age (median age, 57 years) and most adult cases
(70%) occur in patients with at least one underlying
chronic illness (immunosuppression, diabetes, alco-
hol/drug abuse, malignancy, or chronic systemic dis-
ease). Children by contrast tend not to have chronic
illness, but necrotizing fasciitis may complicate
chickenpox. Occasionally the disease afflicts appa-
rently healthy individuals. It has been suggested
(unconvincingly) that antiinflammatory medication
might predispose to these spreading infections by
interfering with granulocyte function [30], but Kauls
et al [6] could find no such association. In two thirds
of cases, the necrosis followed either a skin lesion or
trauma. In an otherwise normal Western population,
more than 50% of episodes occurred in the limbs,
63
whereas a few cases involved the trunk and perineum.
The head and neck were involved in 1% to 10% of
cases. The condition even has been reported after
routine dental surgery or dental sepsis (Fig. 2).
The variable clinical picture means that delay in
diagnosis is common, because the prodromal period
in which the synergistic consortia of bacteria are
evolving may be only 3 or 4 days before the phase
of rapid acceleration. Diagnosis depends on being
alert to the possibility of the disease and recognizing
the pattern of clinical events, the main feature of
which is a rapidly progressive necrotizing infection.
The area is acutely painful, and the surrounding
tissues are red (the signs depend on the specific
mix of bacteria), but on close inspection a central
portion of skin is pale and toxic (Fig. 3). The skin
subsequently develops a slightly mottled appearance
as it becomes congested through venous stasis. As the
perfusion is further reduced through arterial failure,
the skin starts to blister. Sensory perception is lost as
nerves are destroyed and the wound weeps fluid from
the underlying liquefaction. Gross edema is a feature
of the disease, and gas may be present in up to 40%
of cases [21]. The presence of gas is neither a reliable
nor discriminatory sign for clostridial infections
because it can be absent in gas gangrene and present
in various nonclostridial infections [20]. Gas simply
denotes the presence of anaerobic bacteria [29]. A
marked leukocytosis (median 16,000 leucocytes/mm)
is common, but 20% of patients have a normal white
cell count, and in some cases the count even may be
low. Fever is not always present, especially in the
early stages of the disease.
With advancing disease the patient becomes pro-
gressively unwell, with a general malaise and tachy-
cardia. More than 50% of patients develop significant
hypotension. In 10% to 30% of cases the disease is
complicated by one or more of the following con-
ditions: acute renal failure, coagulopathy, abnormal
liver function, acute respiratory distress syndrome, or
hemolytic anemia. The rapid progression of the
disease is a distinguishing feature.
Diagnosis
If the clinical features are suggestive of necro-
tizing fasciitis, the diagnosis should not wait for the
results of bacterial culture. Clinical inspection of the
wound demonstrates that the subcutaneous fat has no
structural integrity and offers little resistance to the
exploring finger. The skin is widely undermined by
the progressing infection. Histologic criteria have
been described for the early diagnosis of necrotizing
fasciitis by frozen section [31], and the typical pattern
64 M. McGurk / Oral Maxillofacial Surg Clin N Am 15 (2003) 59–67

Fig. 3. An emaciated patient presented with a large oral carcinoma (A) and a history of alcoholism with associated liver failure,
pancreatitis, and diabetes. Necrosis of the skin flaps occurred abruptly 72 hours after surgery (mixed flora) despite antibiotic
prophylaxis (B). Surgical débridement was followed by a regimen of dressing changes every 4 hours (C). Healthy granulating
tissue developed eventually (D) and the skin flaps reattached. The patient remains disease free 2 years after surgery (E).

of a dense polymorphonuclear infiltrate in the dermal Management

layers of the skin clinches the diagnosis. Samples of
necrotic tissue are uninformative; a biopsy should be
taken from normal-looking adjacent tissue. A Gram’s
stain may be helpful in selecting first-line antibiotics,
and blood cultures typically produce positive results.
Shock and multiorgan failure are relatively com-
mon, so resuscitation and general supportive mea-
sures are vital in the established case. Two treatments
are recommended: (1) surgery and (2) antibiotics, to
 M. McGurk / Oral Maxillofacial Surg Clin N Am 15 (2003) 59–67 65

which hyperbaric oxygen might be added. Of these
treatments, the single most important modality is
surgery. There is no controversy regarding the initial
management of spreading necrosis, the extent of
débridement being determined clinically [29]. Under-
lying muscle can be preserved, but all necrotic tissue
and overlying skin must be removed. Resected tissue
(skin, muscle, connective tissue) should be sent for
culture and antibiotic sensitivity (aerobic and anaer-
obic), and Gram’s stain results should be obtained.
Time is of the essence at this stage, because mortality
is associated with delayed intervention. Even if
recognized promptly, significant necrosis usually
has taken place before resection is undertaken. More
than one débridement may be necessary and it is
considered prudent to make a second operative
inspection of the wound after 24 to 36 hours.
Management is similar to that of an extensive
burn. The wounds should be washed (hydrogen
peroxide is useful for débridement) and packed
regularly (every 4 hours) (Fig. 4), a procedure best
done personally by the attending surgeon. Slowly the
slough clears and shiny granulation tissue emerges
from beneath the yellow slime. The undermined skin
at the edge of the wound reattaches to the underlying
granulation tissue and the packing can be withdrawn
slowly day by day. Regular dressing still should be
maintained at 8-hour intervals, which demands a
heavy nursing commitment.
Ultimately, antibiotic therapy is dictated by the
cultures, but intravenous penicillin is the initial drug
of choice. If the Gram’s stain shows a mixed flora, a
broad-spectrum antibiotic also should be used (gen-
tamycin), and it can be supplemented as appropriate
information is obtained. Finally, hyperbaric oxygen
has been suggested as a supportive measure, but
there is no definitive evidence of efficacy and there
are obvious logistic problems if this technology
is contemplated.
Results
Despite proper management of necrotizing fascii-
tis, mortality remains high. A collective review [32]
[Janevicious, Han & Batt 1982] of 146 cases reported

Fig. 4. During the preparation of this article a 27-year-old man presented with a recent history of third molar infection (A). He
was otherwise healthy. The tooth was removed manually by the patient himself on a Sunday, he was admitted with low-grade
cervical infection the following Thursday, and submental necrosis developed overnight (B). Note the ring of ischemia around the
ulcer and the congestion in the surrounding tissue. Fluid can be seen leaking from the ulcer. The infection (mixed flora) settled
quickly after débridement (C) and standard antibiotic therapy. The wound was allowed to granulate before repair with a split
skin graft (D).
66 M. McGurk / Oral Maxillofacial Surg Clin N Am 15 (2003) 59–67
a 38% mortality rate. In a prospective study, Kaul [6]
record ed a case fatality rate of 34%. Pessa and
Howard [33] applied severity of illness scoring sys-
tems to predict outcome. They found that the scores
continued to rise after débridement in patients who
were to die, and accurate prediction of survival could
be made as early as the third postoperative day.
Mortality rates increase with age ( > 50 years of
age), concomitant illness (diabetes), delay in diag-
nosis or treatment [20], inadequate débridement,
lesions of the abdomen, and mucormycosis [23,32].
In a multivariate analysis, age, hypotension, and
bacteremia were independent variables that predicted
mortality. Patients can die from systemic problems
some days or weeks after the infection, and in
historical texts there are reports of fatal arterial bleeds
occurring approximately 10 days after surgery, just as
the infective process is settling.
Summary
Necrotizing fasciitis is a dramatic but rare disease.
In the head and neck it often strikes unexpectedly.
Early diagnosis and radical treatment are important to
maximize the chances of a good outcome. The
outcome depends on the clinician having a high
threshold of suspicion when rapidly progressive local
and systemic symptoms appear in the presence of
what was initially an apparently innocuous wound.
Diagnosis is achieved principally by inspection and
manual examination during explorative surgery but
can be supported by frozen section examination.
Treatment is mainly surgical, helped by general
measures of support to combat the systemic effects
of circulating toxins. The key to success is captured
in a line by Shakespeare in Macbeth: ‘‘Be bloody,
bold and resolute.’’
References
[1] Deans M. Flesh-eating bugs scare. Lancet 1994;
343:1418.
[2] Thomson J. Hospital gangrene or malignant ulcers. In:
Lectures on inflammation. Edin: James Ballantyne and
Co.; 1813. p. 456 – 500.
[3] Gillespie L. Observations on the putrid ulcer. London
Medical Journal 1785;6:373 – 400.
[4] Adams F. The genuine works of Hippocrates. London:
Sydenham Society; 1771. p. 400 – 1.
[5] Anonymous. Invasive group A streptococcal infec-
tions. JAMA 1994;272:16.
[6] Kauls R, McGeer A, Low DE, Green K, Schwartz
AE, Simor AE. Population based surveillance for
group A streptococcal necrotizing fasciitis: clinical
features, prognostic indications and microbiological
analysis of seventy seven cases. Am J Med 1977;
103:18 – 24.
[7] Schlievert PM, Aris P, Assimacopoulos Cleary PP. Se-
vere invasive group A streptococcal disease: clinical
description and mechanism of pathogenesis. J Lab Clin
Med 1996;127:13 – 22.
[8] Blackadder HH. Observations on phagedena gangrae-
nosa. Edinburgh: David Brown; 1818.
[9] Trotter T. Medicina nautical. London: Longman, Hurst,
Rees and Orme; 1797. p. 1 – 111.
[10] Henne J. Principles of military surgery. Edinburgh:
Constable; 1820.
[11] Nightingale F. Notes on hospitals, 3rd edition. London:
Longman, Green; 1863.
[12] Jones J. Investigation upon the nature, causes and treat-
ment of hospital gangrene as it prevailed in the Con-
federate armies 1861 – 1865. In: Hamilton FH, editor.
United States Sanitary Commission memoirs: surgical
II. New York: Riverside Press; 1871. p. 146 – 70.
[13] Barnes J. Gangrene. In: Barnes J, editor. Medical
and surgical history of the War of Rebellion. 1867.
p. 823 – 51.
[14] Welbank R. On sloughing phagedaena. London: Long-
mans Brown and Green; 1844.
[15] Fournier JA. Gangrene fourdroyant de la verge. Sem-
aine Medicale 1883;3:345 – 7.
[16] Travers T. Two cases of slough ulceration. London
Medical and Physical Journal 1824;122 – 34.
[17] Melaney F. Hemolytic Streptococcus gangrene. Arch
Surg 1924;9:317 – 64.
[18] Enright MC, Spratt BG, Kalia A, Cross JH, Bessen
DE. Multilocus sequence typing of Streptococcus pyo-
genes and the relationships between emm type and
clone. Infect Immunol 2001;69:2416 – 27.
[19] Giuliano A, Lewis F, Hadley K, Blaisdell FW. Bacteri-
ology of necrotizing fasciitis. Am J Surg 1977;
134:52 – 7.
[20] Freischlag JA, Ajalat G, Busuttil RW. Treatment of
necrotizing soft tissue infections. Am J Surg 1985;
149:751 – 5.
[21] Brook I, Frazier EH. Clinical and microbiological fea-
tures of necrotizing fasciitis. J Clin Microbiol 1995;
33:2382 – 7.
[22] Joynt GM, Gommersall CD, Lyon DJ. Severe necrot-
izing fasciitis of the extremities caused by Vibrionacae:
experience of a Hong Kong Territory referral hospital.
Hong Kong Med J 1999;5:63 – 8.
[23] Patino J, Castro D, Valencia A, Morales P. Necrotizing
soft tissue lesions after a volcanic cataclysm. World J
Surg 1991;15:240 – 7.
[24] Cunningham MW. Pathogenesis of group A streptococ-
cal infections. Clin Microbiol Rev 2000;13:470 – 511.
[25] Engleberg NC, Heath A, Miller A, Rivera C, Di Rita
VJ. Spontaneous mutations in the CsRS two compo-
nent regulatory system of Streptococcus pyogenes re-
sulting in enhanced virulence in a murine model of
 M. McGurk / Oral Maxillofacial Surg Clin N Am 15 (2003) 59–67
skin and soft tissue infection. J Infect Dis 2001;183:
1043 – 54.
[26] Pruitt BA. Burns and soft tissues. In: Polk Jr HC,
editor. Infection and the surgical patient: clinical sur-
gery international. London: Churchill-Livingstone;
1982. p. 113 – 31.
[27] Gorbach SL, Bartlett JG, Nichols RL. Manual of sur-
gical infections: skin and soft tissue infections. Boston:
Little Brown and Co.; 1984.
[28] Simmons RL, Ahrenholz DH. Infections of the skin
and soft tissues. In: Howard RJ, Simmons RL, editors.
Surgical infections diseases. 2nd edition. Norwalk: Ap-
pleton and Lange; 1988. p. 404 – 8.
[29] Dellinger EP. Severe necrotizing soft-tissue infections:
67
multiple disease entities requiring a common approach.
JAMA 1981;246:1717 – 21.
[30] Brun-Buisson CJL, Saada M, Trunet P, Rapin M, Rou-
jeau J, Revuz J. Haemolytic streptococcal gangrene
and non-steroidal anti-inflammatory drugs. BMJ
1985;290:1786.
[31] Stamenkovic I, Lew PD. Early recognition of poten-
tially fatal necrotizing fasciitis: the use of frozen sec-
tion biopsy. N Engl J Med 1984;310:1689.
[32] Janevicius RV, Hann SE, Batt MD. Necrotizing fascii-
tis. Surg Gynecol Obstet 1982;154:97 – 102.
[33] Pessa ME, Howard RJ. Necrotizing fasciitis. Surg Gy-
necol Obstet 1985;161:357.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 69 – 78

Diagnosis and treatment of diffuse sclerosing

osteomyelitis of the jaws
Marjut Montonen, MD, DDSa,*, Christian Lindqvist, MD, DDS, PhDa,b
a
Department of Oral and Maxillofacial Diseases, Helsinki University Central Hospital, Surgical Hospital,
P.O. Box 263, Fin-00029 HUS, Finland
b
Department of Oral and Maxillofacial Surgery, Helsinki University Central Hospital, Institute of Dentistry,
P.O. Box 263, Fin - 00029 HUS, Helsinki, Finland

Diffuse sclerosing osteomyelitis of the jaws is an
inflammatory condition that is a consequence of
vascular changes and nonspecific bacterial infection
in which bone deposition rather than bone resorption
occurs [1 – 3]. Diagnosis is based on clinical and
radiologic findings and histopathologic findings relat-
ing to bone specimens. Histopathologic findings in
the maxilla and in the mandible are similar, but the
symptoms—which are characterized by intermittent
pain resistant to almost all treatment—and specific
radiological changes relate almost exclusively to the
mandible [3 – 10].
Etiology and pathogenesis
In osteomyelitis of the jaws following pulpal or
periodontal infections, there is inflammation of both
the cortical and cancellous bone. In the acute form,
there is pain, swelling, and purulent discharge. The
condition is generally curable by antibiotics, drain-
age, and removal of the offending teeth; however, if
the condition is untreated or inadequately treated, the
infection becomes chronic and eradication may
require extensive sequestrectomy as well as antibiotic
therapy. Diffuse sclerosing osteomyelitis differs from
the acute and chronic suppurative forms of osteo-
* Corresponding author.
E-mail address: marjut.montonen@hus.fi
(M. Montonen).
myelitis by its more insidious onset, which is char-
acterized by acute exacerbations of pain and swelling
but a general lack of purulent drainage.
Most investigators believe that diffuse sclerosing
osteomyelitis is caused by bacteria that are not
particularly virulent; however, no specific causative
micro-organisms have yet been identified. Various
bacteria have been reported in some studies, but
contamination of the culture specimen by organisms
on the skin or oral mucosa could not be excluded
[11 – 13]. It has been suggested that a hyperactive
immunologic reaction to the bacterial toxins may be
responsible for the chronic inflammatory response.
There is generally no identifiable bacterial port of
entry, but chronic periodontal disease has been impli-
cated as a possible source. Diffuse sclerosing osteo-
myelitis is difficult to eradicate and may persist for
years, with intermittent symptoms of pain and swell-
ing of the face [5,11,14].
Differential diagnosis
Diffuse sclerosing osteomyelitis should be distin-
guished from secondarily infected florid osseous
dysplasia. In the latter, apart from inflammation and
reactive changes, histological studies reveal a fibro-
blastic stroma with bone and cementum-like struc-
tures that have been formed as a result of metaplasia
[15]. Diffuse sclerosing osteomyelitis may even
indicate the mandibular location of SAPHO (syn-
ovitis, acne, pustulosis, hyperostosis, osteitis) syn-
drome [16 – 19].

1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 7 3 - 0
70 M. Montonen, C. Lindqvist / Oral Maxillofacial Surg Clin N Am 15 (2003) 69–78

Clinical findings
The most important criteria for diagnosis of dif-
fuse sclerosing osteomyelitis of the jaws are history
of intermittent pain, swelling, trismus, pressure and
paresthesia. These symptoms can appear at any age
and often persist for years. Exacerbations are charac-
terized by a marked inflammatory reaction in the
bone marrow leading to sensitization and stimulation
of primary afferent nociceptive nerves, which results
in intense, sometimes unbearable pain, particularly in
the lower jaw [8,9,20].
Because pain is the most important symptom of
diffuse sclerosing osteomyelitis, standardization of
measurement of pain and employment of a question-
naire detailing the history of the pain are useful in
clinical examination. Using such means, the nature
and intensity of pain can be classified. Classification
is necessary from a differential diagnostic point of
view, and in deciding treatment and evaluating its
effects [20]. Pain can be measured using a visual
analogue scale, the McGill Pain Questionnaire,a pain-
relief scale, or a pain diary [21,22]. The visual
analogue scale involves the patient drawing a mark
on a 100-mm horizontal line on a piece of paper. The
left edge of the line indicates no pain, and the right
edge indicates the worst possible pain [22]. The
McGill Pain Questionnaire involves the patient
choosing words from various categories that describe
the nature of the pain as well as its intensity [21].
Radiologic findings
Radiographic and scintigraphic investigations are
used to increase diagnostic accuracy and improve
prognostic and therapeutic judgements [7,23 – 26].
Whereas a typical feature of suppurative osteomye-
litis of infectious origin is a radiolucent lesion spread-
ing through the cancellous bone, with cortical bone
perforation and a lamellated periosteal reaction, in
chronic sclerosing osteomyelitis, intermingled scle-
rotic and osteolytic lesions with a solid periosteal
reaction or external bone resorption are common
findings [24]. However, with the passage of time,
sclerosis becomes increasingly marked, and normal-
ization of bone structure is rare [7].
Panoramic and intraoral radiographs and 99mTc-
scintigraphy are most often used for diagnosis, deter-
mination of disease activity, and follow-up [7,9]. CT

Fig. 1. (A) Vertical and horizontal osteotomies over the osteomyelitic lesion. (B) Removal of the lateral cortical bone to the
exposed cancellous bone. (C) Bur holes have been drilled through lingual cortex.
 M. Montonen, C. Lindqvist / Oral Maxillofacial Surg Clin N Am 15 (2003) 69–78 71

scans and MRI techniques can also be used to deter-

mine extent of disease, especially when planning
primary surgical therapy [2,24,27,28]. However,
MRI techniques may be less useful during follow-up
after surgery because of disturbing artifacts related to Table 1
the use of metallic instruments. Yoshimura et al [28] Unpublished data on patients with diffuse sclerosing
osteomyelitis treated in the Department of Oral and
showed by CT scan a close interaction between cortical
Maxillofacial Surgery at Helsinki University Central Hospital
plate disruption and muscle inflammation. Compared
with plain film, the extent of the soft tissue involve- Characteristics Years Number
ment could be better appreciated with CT scans, Sex
especially in mixed pattern cases [28]. Along with Female 59
this, detecting and localization of active sequestra will Male 43
be improved by the use of MRI, CT scans, or single Age at onset of symptoms Mean 27.5
Range 5 – 77
photon emission computerized tomography (SPECT)
Dental status
with 99mTc hydroxymethylene diphosphonate (HDP)
Edentulous 10
when planning surgical debridement [2,29]. Dentate 92
Location of symptoms
Maxilla 2
Histologic findings Mandible 100
Dentate area 102
Histologic examination and various enzyme Other regions 98
immunohistochemical investigations of bone speci- Symptoms
mens are necessary to establish diagnosis. These can Pain 102
Swelling 100
be obtained by using an accurate biopsy technique.
Trismus 100
Use of a rotating trepan bur is preferred in attempting
Duration of disease Mean 13.5
to obtain representative bone specimens [30]. Scler- Range 0.5 – 42
otic changes have been described mainly in the Diagnosis
subperiosteal parts of biopsy specimens from the Clinical 102
lower jaw using this technique. Coarse trabeculae Radiologic 102
and necrotic foci, partly calcified, thin trabeculae Histologic 102
reminiscent of fibrous dysplasia, and granulation Conservative treatment
tissue with chronic inflammatory cell and foreign- Outpatients 102
body giant cells have been found in cancellous bone Inpatients 77
Inpatients in pain clinic 36
[8,9,30 – 32]. The histopathology of diffuse sclerosing
Systemic antibiotics 102
osteomyelitis in the maxilla and mandible is probably
Nonsteroidal 102
the same; however, diffuse sclerosing osteomyelitis antiinflammatory drugs
of the maxilla has been mentioned in only a few Systemic steroids 42
reports [1,33,34]. Disodium clodronate 36
Weak or strong opioids 102
Anticonvulsants 10
Laboratory findings Antidepressants 8
Physiotherapy 67
In general, there are no specific laboratory tests Hyperbaric 9
oxygen treatment
for determining the stage of diffuse sclerosing osteo-
Surgery
myelitis except for the mean erythrocyte sedimenta-
Dentoalveolar surgery 28
tion rate, which can be slightly elevated during Revision, removal 15
exacerbations. In attempts to discover a suitable tool of sequestrum
for evaluation of the results of treatment, the serum Decortication 47
levels of alpha-1-antitrypsin, orosomucoid, and hap- Partial resection 2
toglobin have been measured during different stages; Resection and 1
however, caution must be exercised when general- reconstruction with
izing from results of single measurements. Serum microvascular bone craft
levels of C-reactive proteins seem to reflect the Combination of conservative 94
therapy and surgery
clinical activity of the disease, but a certain ‘‘mass
72 M. Montonen, C. Lindqvist / Oral Maxillofacial Surg Clin N Am 15 (2003) 69–78

of inflammation’’ is necessary before elevated values
of these proteins can be detected [35]. Leukocyte and
platelet counts, as well as serum levels of osteocalcin,
parathyroid hormone, creatinine, calcium, phosphate,
aspartate aminotransferase, alanine aminotransfer-
ase, albumine, and alkaline phosphatase, remained
within normal limits [20].
Because the synthesis of immunoglobulins only
reflects inflammatory activity indirectly, immunoglo-
bulines are not regarded as appropriate markers of
inflammation. However, their concentration in serum
can vary with clinical activity [35]. Immunological
findings suggest that most patients with diffuse scle-
rosing osteomyelitis have normal humoral and cel-
lular responses. The inflammatory events and
chronicity of the disease cannot be explained by
immunologic findings [11,36].
Bacteriologic findings
Various attempts have been made to determine the
causes of diffuse sclerosing osteomyelitis. Bacterio-
logic investigations of specimens from diseased bone
have been undertaken, paying particular attention to
anaerobic culture techniques. Few studies involving
healthy controls have been performed. Some bac-
teria—including Propionibacterium acnes, Pepto-
streptococcus intermedius, Eikenella corrodens,
Actinomyces species, Streptococcus sanguis, Strep-

Fig. 2. (A) Panoramic radiographs of a 17-year-old female with a 9-month history of pain and swelling on the right side of the
mandible preceded by prolonged endodontic treatment of the second molar. (B) Wide lytic areas can be seen in connection with
the teeth and in the body of the mandible. (C) Uptake of 99mTc-diphosphonate corresponding to the radiographic changes.
 M. Montonen, C. Lindqvist / Oral Maxillofacial Surg Clin N Am 15 (2003) 69–78 73

tococcus mitis, and Fusobacterium nucletum—have
been found, but in very small numbers [11 – 13].
However, contamination with skin or oral flora cannot
be avoided despite improvements of sampling techni-
ques. The results of bacteriological and serological
investigations indicate that Propionibacterium acnes
and Peptostreptococcus intermedius may be of sig-
nificance in relation to the disease but cannot explain
its chronicity [11]. One bacteriologic study comparing
patients with diffuse sclerosing osteomyelitis of the
mandible with healthy controls has been done, and this
study found no difference in the quantity or quality of
bacteria in the bone samples [13]. No particular group
of bacteria was unequivocally associated with disease.
Thus, the findings probably reflected contamination.
Therapy
Nonsurgical treatment
There is no cure for diffuse sclerosing osteomye-
litis [37 – 39]. The main reason is that knowledge of
the disease, its causes, and its natural history is not
known. Long-term antibiotic therapy can have a
beneficial effect on the course of the disease in its
early stages, while corticosteroid therapy and some-
times decortication can be more effective once the
condition has become chronic [37]. The type of
antibiotics and the exact duration of treatment are
difficult to establish [40]. The clinical efficacy of
long-term roxithromycin treatment (300 mg/day
orally for 68 days to 66 months) in patients with
diffuse sclerosing osteomyelitis of the mandible has
been studied [41]. In seven of nine cases (77.8%)
the symptoms disappeared 1 month to 1 year after
the start of therapy. Radiography showed that
osteolysis had decreased but that osteosclerosis
had persisted or increased by the end of therapy.
The optimum duration of treatment should be
determined based on the amelioration of symptoms
and the disappearance of osteolytic findings on
radiographs. The mechanism of action of roxithro-
mycin is not yet fully understood, but long-term
roxithromycin treatment may be useful before sur-
gical treatment is considered [41].

Fig. 3. (A) Two years after onset of the disease, an exacerbation occurred with swelling of the mandible and severe pain. (B)
Predominantly sclerotic areas are seen in the right body and angle and new lytic areas as present in the ramus of the mandible.
(C) An intensive uptake of 99mTc-diphosphonate corresponds to the swollen and painful area of the mandible.
74 M. Montonen, C. Lindqvist / Oral Maxillofacial Surg Clin N Am 15 (2003) 69–78

The inflammatory nature of the disease has been Surgery

demonstrated by alleviation of ostomyelitic pain
through administration of glucocorticosteroids [37].
Hyperbaric oxygen has often been recommended as
an adjunct to treatment of diffuse sclerosing osteo-
myelitis of the jaws with antibiotics and surgery
[37,39]. The results of hyperbaric oxygen treatment
of chronic osteomyelitis of the jaws have been
described by van Merkesteyn et al [39].
Disodium clodronate, a bisphosphonate, is a
potent anti-osteolytic agent which has been used to
treat diseases of bone and calcium metabolism [42 –
44]. It has also been found to be useful for treatment
of recurrent pain of diffuse sclerosing osteomyelitis
unresponsive to conservative or surgical therapy
[13,45]. In a randomized, placebo-controlled, dou-
ble-blind study, patients received disodium clodro-
nate (300 – 900 mg) or placebo intravenously.
Disodium clodronate administration resulted in no
better immediate pain relief than administration of
placebo. However, 6 months after treatment there was
a statistically significant difference in pain intensities
between the groups, with the disodium clodronate
group experiencing significantly less pain [13].
Surgical treatment of diffuse sclerosing osteomye-
litis involves decortication and removal of any foci of
infection and sequestra [37,40,46]. The results of
decortication have been described by Hjørting-Hansen
[47], Jacobsson and Hollender [37], and Montonen et
al [38]. They reported a success rate in relation to to
mandibular decortication of about 50%. Symptoms
generally recurred within a year of surgery. Advanced
age at the time of decortication and the presence of
carious or poorly endodontically treated teeth in the
decorticated area correlated with recurrence [14].
Decortication can be performed intraorally or
extraorally and should be preceded by intravenous
antibiotics and hyperbaric oxygen treatment for opti-
mial results. After horizontal and vertical burr cuts
have been made in the lateral cortex covering the
diseased area, pieces of cortical and cancellous bone
can be removed using a chisel. Perforations can then be
drilled into the bleeding marrow cavity before closure
of the wound to improve bone nutrition (Fig. 1) [38].
In cases in which decortication fails, resection of
affected areas has been recommended [48 – 50]; how-

Fig. 4. (A) The patient was symptom-free 6 months after combined conservative treatment consisting of removal of the third
molar, intravenous antibiotics, hyperbaric oxygen, and disodium clodronate. (B) There was remodeling of the mandible bone
along with increasing sclerosis of the angle and the body of the mandible.
 M. Montonen, C. Lindqvist / Oral Maxillofacial Surg Clin N Am 15 (2003) 69–78 75

ever, resection should be limited to severe, therapy-
resistant cases because of potential disadvantages
such as loss of function of the inferior alveolar nerve
and problems related to mandibular reconstruction
[14]. It would therefore seem preferable to choose
saucerization combined with particulate cancellous
bone and marrow grafting as a relatively conservative
surgical treatment for diffuse sclerosing osteomyelitis
of the mandible [51]. The topical application of a
broad spectrum antibiotic to the surgical bed after
removal of the affected jaw bone has been reported in
only two case reports [46,52]. The outcome of this
therapy was not better than decortication alone.
Personal experience
Records relating to 102 patients (59 women,
43 men) with diffuse sclerosing osteomyelitis treated
between 1962 and 2002 in the Department of Oral
Maxillofacial Surgery at Helsinki University Central
Hospital were retrospectively analyzed (unpublished
data, M. Montonen, 2002). Patients with acute and
subacute suppurative osteomyelitis and osteoradio-
necrosis were excluded. Thirty-three patients are still
being followed and are receiving therapy as neces-
sary. All patients had been clinically and radiograph-
ically examined, and the diagnoses have been verified
by histological investigation (Table 1). All patients
had typical symptoms with intermittent pain, swell-
ing, trismus, and radiological changes.
A typical example is a young woman with a
6-year history of the disease whose radiographs are
shown in Fig. 2. Her symptoms first occured in the
right second molar area at the age of 17. The location
of symptoms varied subsequently between the ramus
and symphysis of the mandible. Erupting third molars
were considered to be the infectious foci and were
extracted. Hyperbaric oxygen treatment was under-
taken before and after the operation, and antibiotics
were administered intravenously (Fig. 3). Bone –
tissue specimens were taken from the affected man-
dible by means of minor decortication. They exhibited
typical histologic changes, namely subcortical scle-
rosis and moderate chronic medullar inflammation
with mononuclear cell infiltration. Microbial findings

Fig. 5. (A) After a symptom-free period of 13 months, increased pain and swelling occurred in the right ramus, body, and
symphysis of the mandible. (B) Panoramic radiograph showing sclerotic changes in the body, angle, and ramus of the mandible
and lytic lesions in the coronoid process (arrow) and adjacent to the right first premolar and central incisors, which had a positive
response to electrical stimulation. (C) Intense uptake of 99mTc-diphosphonate corresponds in the dentate area of the mandible.
76 M. Montonen, C. Lindqvist / Oral Maxillofacial Surg Clin N Am 15 (2003) 69–78

Fig. 6. (A) Panoramic radiograph of a 22-year-old woman with a 3-year history of diffuse sclerosing osteomyelitis of the
mandible taken 2 weeks before surgery. Mixed lytic radiographic lesions are present in the mandibular body, ramus, and condylar
process. The teeth with periapical lesions had a positive response to electrical stimulation. (B) Panoramic radiograph taken
6 years after surgery. The mandible appears radiologically normal and the patient is symptom-free.

with spare oral flora were equal from the diseased and
the contralateral side of the mandible bone, resem-
bling the results from the previous study with mis-
taken contamination [13]. Seven months after surgery
the patient was symptom-free, but sclerosis in the
body and the ramus of the mandible was increasing
(Fig. 4). Despite operative treatment; long-term ad-
ministration of antibiotics, anti-inflammatory drugs,
and disodium clodronate; and hyperbaric oxygen
treatment, symptoms recurred after 1.5 years in the
right premolar and incisor regions. There were lytic
radiological changes and increased uptake in the
99m
Tc-scintigram (Fig. 5). The teeth in the affected
area were vital and were not endodontically treated. If
symptoms persist despite continued medical treat-
ment, either the new endodontic therapy of the right
second molar should be performed or the tooth
should be extracted, because it may act as the focus
of infection despite an apparently satisfactory
root filling.
Only two cases of maxillary diffuse sclerosing
osteomyelitis have been reported; the disease mostly
affects the mandible. The pain in the maxilla differed
from that associated with diffuse sclerosing osteo-
myelitis of the mandible. These patients had neuro-
pathic pain with a burning sensation or a feeling of
dull, deep pressure. In both cases there was a dental
problem before symptoms occurred. Radiologically,
the disease was characterized by mixed sclerotic and
osteolytic changes of the mandibular bone, but there
were few radiological findings in the upper jaw.
Only 1 of 100 patients with mandibular diffuse
sclerosing osteomyelitis experienced complete clin-
ical and radiological recovery from the disease. This
patient had suffered typical intermittent symptoms
with typical findings for 9 years but has now been
free from symptoms for over 10 years (Fig. 6) [38].
Summary
Early clinical and radiological diagnosis of diffuse
sclerosing osteomyelitis and effective, combined con-
servative and surgical treatment, including meticu-
lous dental care by a specialist, may improve the
chance of avoiding a poor outcome. There should be
regular follow-up by a specialist in a facial pain clinic
where there are possibilities of oral and maxillofacial
consultation. Such an approach should facilitate rapid
and efficient evaluation of pain and treatment of
exacerbations. To minimize the likelihood of exacer-
bations, even regular dental checkups and treatments
should be performed.
References
[1] Bell WH. Sclerosing osteomyelitis of the mandible and
maxilla. Oral Surg Oral Med Oral Pathol 1959;12:
391 – 402.
[2] Schuknecht BF, Carls FR, Valavanis A, Sailer HF.
Mandibular osteomyelitis: evaluation and staging in
18 patients, using magnetic resonance imaging, com-
puted tomography and conventional radiograps. J Cra-
nio-Maxillofacial Surg 1997;25:24 – 33.
[3] Thomas DW, Shepherd JP. Paget’s disease of bone:
current consepts in pathogenesis and treatment. J Oral
Pathol Med 1994;23:12 – 6.
[4] Adekeye EO, Cornah J. Osteomyelitis of the jaws:
a review of 141 cases. Br J Oral Maxillofac Surg
1985;23:24 – 35.
[5] El-Mofty S. Chronic diffuse sclerosing osteomyelitis.
Oral Surg Oral Med Oral Pathol 1973;36:898 – 904.
[6] Hudson JW. Osteomyelitis of the jaws. J Oral Maxil-
lofac Surg 1993;51:1294 – 301.
[7] Jacobsson S, Hollender L. Chronic sclerosing osteo-
myelitis of the mandible. Scintigraphic and radio-
 M. Montonen, C. Lindqvist / Oral Maxillofacial Surg Clin N Am 15 (2003) 69–78
graphic findings. Oral Surg Oral Med Oral Pathol
1978;45:167 – 74.
[8] Jacobsson S. Diffuse sclerosing osteomyelitis of the
mandible. Int J Oral Surg 1984;13:363 – 85.
[9] van Merkesteyn JP, Groot RH, Bras J, Bakker DJ. Dif-
fuse sclerosing osteomyelitis of the mandible: Clinical,
radiographic and histologic findings in twenty-seven
patients. J Oral Maxillofac Surg 1988;46:825 – 9.
[10] Wannfors K, Hammarström L. A proliferative inflam-
mation in the mandible caused by implantation of an
infected dental root: a possible experimental model for
chronic osteomyelitis. Int J Oral Maxillofac Surg
1989;18:179 – 83.
[11] Jacobsson S, Dahlén G, Möller Å. Bacteriologic and
serologic investigation in diffuse sclerosing osteomye-
litis (DSO) of the mandible. Oral Surg Oral Med Oral
Pathol 1982;54:506 – 12.
[12] Marx R, Carlson E, Smith B, Toraya N. Isolation of
Actinomyces species and Eikenella corrodens from
patients with chronic diffuse sclerosing osteomyelitis.
J Oral Maxillofac Surg 1994;52:26 – 33.
[13] Montonen M, Jousimies-Somer H, Lindqvist C.
Quantitative microbial findings in diffuse sclerosing
osteomyelitis (DSO) of the mandible. Presented at
the 15th Congress of European Association for Cra-
nio-Maxillo-Facial Surgery. Edinburgh, Scotland, Sep-
tember 5 – 9, 2000.
[14] Panders AK, Hadders HN. Chronic sclerosing inflam-
mations of the jaw. Oral Surg Oral Med Oral Pathol
1970;30:396 – 412.
[15] Shakenovsky BN, Ripamonti U, Lownie JF. Chronic
osteomyelitis of the jaws. Int J Oral Maxillofac Surg
1986;15:352 – 6.
[16] Eyrich GK, Harder C, Sailer HF, Langenegger T,
Bruder E, Michel BA. Primary chronic osteomyelitis
associated with synovitis, acne, pustulosis, hyperosto-
sis and osteitis (SAPHO syndrome). J Oral Pathol Med
1999;28:456 – 64.
[17] Kahn MF, Hayem F, Hayem G, Grossin M. Is diffuse
sclerosing osteomyelitis of the mandible part of the
synovitis, acne, pustulosis, hyperostosis, osteitis (SA-
PHO) syndrome? Analysis of seven cases. Oral Surg
Oral Med Oral Pathol 1994;78:594 – 8.
[18] Suei Y, Taguchi A, Tanimoto K. Diffuse sclerosing
osteomyelitis of the mandible: its characteristics and
possible relationship to synovitis, acne, pustulosis, hy-
perostosis, osteitis (SAPHO) syndrome. J Oral Maxil-
lofac Surg 1996;54:1194 – 200.
[19] Valleala H, Montonen M, Lindqvist C, Kahn M-F,
Konttinen YT. Tuntematon SAPHO [The unknown
SAPHO]. Duodecim 2000;116:1985 – 9.
[20] Montonen M, Kalso E, Pylkkänen L, Lindström BM,
Lindqvist C. Disodium clodronate in the treatment of
diffuse sclerosing osteomyelitis (DSO) of the mandi-
ble. Int J Oral Maxillofac Surg 2001;30:313 – 7.
[21] Melzack R. The McGill Pain Questionnaire: major
properties and scoring methods. Pain 1975;1:277 – 99.
[22] Onhaus EE, Adler R. Methodological problems in the
measurement of pain: a comparison between the verbal
77
rating scale and the visual analogue scale. Pain 1975;1:
379 – 84.
[23] Rohlin M. Diagnostic value of bone scintigraphy in
osteomyelitis of the mandible. Oral Surg Oral Med
Oral Pathol 1993;75:650 – 7.
[24] Suei Y, Taguchi A, Tanimoto K. Radiographic evalua-
tion of possible etiology of diffuse sclerosing osteo-
myelitis of the mandible. Oral Surg Oral Med Oral
Pathol Oral Radiol Endod 1997;84:571 – 7.
[25] von Wowern N, Hjørting-Hansen E, Edeling CJ. Bone
scintigraphy of benign jaw lesions. Int J Oral Surg
1978;7:528 – 33.
[26] Tsuchimochi M, Higashino N, Okano A, Kato J. Study
of combined technetium 99m methylene diphoshonate
and gallium 67 citrate scintigraphy in diffuse scleros-
ing osteomyelitis of the mandible. J Oral Maxillofac
Surg 1991;49:887 – 97.
[27] Orpe EC, Lee L, Pharoah MJ. A radiological analysis
of chronic sclerosing osteomyelitis of the mandible.
Dentomaxillofac Radiol 1996;25:125 – 9.
[28] Yoshiura K, Hijiya T, Ariji E, Sa`do B, Nakayama E,
Higuchi Y, et al. Radiographic patterns of osteomyelitis
in the mandible. Plain film/CT correlation. Oral Surg
Oral Med Oral Pathol 1994;78:116 – 24.
[29] Robinson CB, Higginbotham-Ford EA. Determination
of sequestrum activity by SPECT with CT correlation
in chronic osteomyelitis of the head and neck. J Oto-
laryngol 1986;15:279 – 81.
[30] Jacobsson S, Heyden G. Chronic sclerosing osteomye-
litis of the mandible. Histologic and histochemical
findings. Oral Surg Oral Med Oral Pathol 1977;43:
357 – 64.
[31] Shafer WG. Chronic sclerosing osteomyelitis. J Oral
Surg 1957;15:138 – 42.
[32] Wannfors K, Hammarström L. Infectious foci in
chronic osteomyelitis of the jaws. Int J Oral Surg
1985;14:493 – 503.
[33] Deane CE. Chronic diffuse sclerosing osteomyelitis of
the maxilla and mandible. Oral Surg Oral Med Oral
Pathol 1978;46:872 – 3.
[34] Tooley M, Towns BS. Chronic sclerosing osteomyelitis
of maxilla and mandible: review of the literature and
report of case. J Oral Surg 1972;30:903 – 5.
[35] Wannfors K, Hansson LO. Plasma protein changes in
chronic osteomyelitis of the jaws. J Oral Pathol Med
1991;20:81 – 5.
[36] Malmström M, Fyhrquist F, Kosunen TU, Tasanen A.
Immunological features of patients with chronic scle-
rosing osteomyelitis of the mandible. Int J Oral Surg
1983;12:6 – 13.
[37] Jacobsson S, Hollender L. Treatment and prognosis of
diffuse sclerosing osteomyelitis (DSO) of the mandi-
ble. Oral Surg Oral Med Oral Pathol 1980;49:4 – 14.
[38] Montonen M, Iizuka T, Hallikainen D, Lindqvist C.
Decortication in the treatment of diffuse sclerosing os-
teomyelitis of the mandible. Oral Surg Oral Med Oral
Pathol 1993;75:5 – 11.
[39] van Merkesteyn JP, Bakker DJ, van der Waal I, Kusen
GJ, Egyedi P, van den Akker HP, et al. Hyperbaric
78 M. Montonen, C. Lindqvist / Oral Maxillofacial Surg Clin N Am 15 (2003) 69–78
oxygen treatment of chronic osteomyelitis of the jaws.
Int J Oral Surg 1984;13:386 – 95.
[40] Ord RA, El Attar A. Osteomyelitis of the mandible in
the children: clinical presentation and review of man-
agement. Br J Oral Maxillofac Surg 1987;25:204 – 17.
[41] Yoshii T, Nishimura H, Yoshikawa T, Furudoi S, Yosh-
ioka A, Takenono I, et al. Therapeutic possibilities of
long-term roxithromycin treatment for chronic diffuse
sclerosing osteomyelitis of the mandible. J Antimicrob
Chemother 2001;47:631 – 7.
[42] Ascaari E, Attardo-Parrinello G, Merlini G. Treatment
of painful bone lesions and hypercalcemia. Eur J Hae-
matol 1989;51(Suppl):135 – 9.
[43] Ernst DS, McDonald RN, Paterson AH, Jensen J,
Brasher P, Bruera E. A double-blind, crossover trial
of intravenous clodronate in metastatic bone pain.
J Pain Symptom Manage 1992;7:4 – 11.
[44] Fleich H. Bisphosphonates in bone disease. From the
laboratory to the patient. 2nd edition. New York: The
Parthenon Publishing Group; 1995.
[45] Jokinen J, Lindqvist C, Elomaa I, Tasanen A. Treat-
ment of recurrent diffuse sclerosing osteomyelitis of
the mandible with clodronate. Presented at the Euro-
pean Association for Cranio- Maxillo-Facial Surgery,
9th Congress. Athens, Greece, September 5 – 9, 1988.
[46] Grime PD, Bowerman JE, Weller PJ. Gentamicin im-
pregnated polymethylmethacrylate (PMMA) beads in
the treatment of primary chronic osteomyelitis of the
mandible. Br J Oral Maxillofac Surg 1990;28:367 – 74.
[47] Hjørting-Hansen E. Decortication in treatment of os-
teomyelitis of the mandible. Oral Surg Oral Med Oral
Pathol 1970;29:641 – 55.
[48] Glahn M. The surgical treatment of chronic osteo-
myelitis of the mandible. J Maxillofac Surg 1974;2:
238 – 41.
[49] Head MD, Sanger JR, Matloub HS, Yousif NJ, Rooney
GE. Bilateral microvascular free iliac grafts for man-
dibular reconstruction in intractable osteomyelitis: re-
port of case. J Oral Maxillofac Surg 1986;44:724 – 7.
[50] Obwegeser HL, Sailer HF. Experiences with intraoral
partial resection and simultaneous reconstruction in
cases of mandibular osteomyelitis. J Maxillofac Surg
1978;6:34 – 40.
[51] Ogawa A, Miyate H, Nakamura Y, Shimada M, Seki S,
Kudo K. Treating chronic diffuse sclerosing osteomye-
litis of the mandible with saucerization and autogenous
bone grafting. Oral Surg Oral Med Oral Pathol Oral
Radiol Endod 2001;9:390 – 4.
[52] Chisholm BB, Lew D, Sadasivan K. The use of tobra-
mycin-impregnated polymethylacrylate beads in the
treatment of osteomyelitis of the mandible. Report of
three cases. J Oral Maxillofac Surg 1993;51:444 – 9.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 79 – 89

Diagnosis and treatment of viral infections

Sol Silverman, Jr, MA, DDSa,*, Craig S. Miller, DMD, MSb
a
Department of Stomatology, University of California School of Dentistry, Box 0422 S-612, 513 Parnassus Avenue,
San Francisco, CA 94143, USA
b
Department of Oral Medicine, Microbiology, Immunology, and Molecular Genetics,
University of Kentucky College of Dentistry, College of Medicine, MN-118,
Lexington, KY 40536-0297, USA

Viral infections are of concern to dental profes-
sionals because of ease of transmission, the oral,
latent, recurrent, and systemic diseases they can
produce, their association with opportunistic infec-
tions and malignant transformation, and their influ-
ence on infection control. In this article, information
regarding the following viruses is provided: (1) HIV,
(2) human herpesviruses, (3) human papillomavi-
ruses, (4) enteroviruses, and (5) hepatitis C virus.
HIV
We are currently in the third decade of an RNA
virus pandemic. There are approximately 40 million
people throughout the world infected with HIV. It is
estimated that more than 16,000 new infections occur
each day. HIV is spread predominantly by sexual con-
tact, blood or blood products, or perinatal exposure.
Infection also results from high-risk activities, such as
sharing needles with infected drug users, having
unprotected sexual activity with one or more infected
partners, receiving infected blood or blood products, or
being accidentally exposed to infected materials.
In the United States, almost 500,000 persons are
reported to be living with HIV and AIDS. More than
2 million persons are believed to be infected, how-
ever. Approximately 98% of infections occur in
adults and adolescents, and approximately one third
of new cases occur in women. New infections are
* Corresponding author.
E-mail address: ssjr@itsa.ucsf.edu (S. Silverman).
occurring at a disproportionately higher rate in Afri-
can Americans. In reversal since the late 1990s, the
number of new cases and deaths is increasing once
again because of viral resistance to multiple drug
therapy, apathy toward barrier techniques, the
increasingly large number of individuals living with
HIV who serve as a reservoir for transmission, and
widespread drug abuse and prostitution.
The HIV epidemic is an important concern to the
dental profession for many reasons. First, infection
control measures are required in the dental office.
Second, many oral and systemic manifestations occur
in immunocompromised individuals who have falling
numbers of physiologically incompetent lymphocytes
and rising viral loads. Third, recognition of the signs
and symptoms of HIV infection should lead to referral
of the patient to a physician for diagnostic testing.
Acute HIV infection in most cases produces flu-
like symptoms that develop 2 to 6 weeks after the
initial infection. Soon thereafter, persistent general-
ized lymphadenopathy occurs, which is followed by a
latent phase. Initially the latent phase is asympto-
matic. Later as the viral loads rise and the CD4+ cell
count drops, lymphadenopathy, weight loss, fever,
diarrhea, fatigue, skin anergy, neurologic decline,
parotid enlargement, and opportunistic infections
develop. Many of the numerous oral lesions that
develop are caused by increased individual suscep-
tibility to viral transmission and proliferation. These
infections, which involve the herpes family viruses
and human papillomaviruses, are more prevalent in
HIV-seropositive patients and are usually concurrent
and recurrent. An increasing number of HIV patients
are coinfected with the hepatitis C virus.

1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 7 2 - 9
80 S. Silverman, C.S. Miller / Oral Maxillofacial Surg Clin N Am 15 (2003) 79–89
The final stage of the disease (AIDS) is often
marked by fatal respiratory infections, lymphoma, or
cancer. Treatment involves the use of highly active
antiretroviral agents, such as nucleoside, nucleotide,
and nonnucleoside reverse transcriptase inhibitors,
and protease inhibitors, which are used in combina-
tion to block virus replication and maturation.
Human herpesviruses
The human herpesvirus (HHV) family includes
herpes simplex viruses (HSV-1 and -2), varicella-
zoster virus (HVV, HHV-3), Epstein-Barr virus
(EBV, HHV-4), cytomegalovirus (CMV, HHV-5),
lymphotrophic viruses (HHV-6 and -7), and Kaposi’s
sarcoma virus (HHV-8). These large DNA viruses
have the hallmark of establishing latent infection. The
latent infection serves as a reservoir for the periodic
activation of virus. Although the molecular factors
that regulate activation of HHVs are still undefined,
aging, immunosuppression, stress, and tissue damage
predispose HHVs to reactivation. Reactivation occurs
despite cell-mediated and humoral HHV immunity.
Clinical manifestations are diverse and are more
severe during immunosuppression.
Herpes simplex viruses
HSV-1, the most common of the HHV oral infec-
tions, must contact mucosa or abraded skin to initiate
infection. By puberty, most individuals have been
Fig. 1. Primary adult-onset herpetic gingivostomatitis in a 28-year-old man. It was manifested by sudden onset, pain, fever,
lymphadenopathy, and gingivitis. There had been no similar previous attacks, and the signs and symptoms resolved completely
in 2 weeks.
exposed to HSV and have developed circulating
antibodies. The virus incubates and replicates for 2
to 12 days within epithelium and then penetrates local
nerve endings [1]. After the primary infection, the
virus travels to regional ganglia, where it remains
latent indefinitely. Asymptomatic shedding and reac-
tivation are common [2]. Although population studies
are variable, clinical recurrences are estimated to
occur in up to 40% of cases [3].
More than 67% of initial exposures are asympto-
matic subclinical infections [4]. The remaining indi-
viduals who acquire the primary infection experience
marked signs and symptoms that last up to 2 weeks
(Fig. 1). Features include gingivostomatitis, lip ves-
icles and coalescing ulcerations, fever, lymphade-
nopathy, and oropharyngeal pain. Approximately
10% of adults are not exposed to or do not obtain
adequate level of antibodies. These persons are at risk
for developing adult-onset acute herpetic gingivosto-
matitis. Although the signs and symptoms in adults
are usually more severe, the attack is usually com-
plete in 2.5 weeks. The initial HSV infection incurs
permanent immunity from a similar future attack;
however, it does not prevent reactivation and recur-
rent mucosal and labial flares.
Recurrent mucosal (intraoral herpes) and labial
(cold sores) infections create a lifetime problem in
persons who are susceptible to reactivation of latent
HSV (Fig. 2). These lesions recur near the point of
entry into the body and are usually caused by HSV-1,
although HSV-2 (usually associated with genital
herpes) occasionally can be identified. The signs,
 S. Silverman, C.S. Miller / Oral Maxillofacial Surg Clin N Am 15 (2003) 79–89 81

Fig. 2. Typical recurrent oral herpes manifested by irregular, shallow gingival ulcerations that tend to coalesce and usually
resolve within 7 to 10 days. The lesions are often mistaken for trauma.

symptoms, and treatment are similar for both sero-
types. Lesions are often preceded by prodromal
symptoms of burning, tingling, itching, or pain.
Recurrent intraoral HSV is often mistaken for
some form of traumatic injury, because lesions
appear on periosteal bound mucosa as small, irregu-
lar, erosive areas or ulcers that usually disappear
within 1 week. An important implication is trans-
mission, because vesicular and ulcerative lesions
shed virus during the first 2 to 3 days, during which
time HSV can be transmitted. Common sites of
spread are the eye (herpetic keratitis) and fingers
(herpetic whitlow) (Fig. 3). ‘‘Cold sores’’ are a
problem because of aesthetics, pain, and source of
transmission. The nature of HSV recurrences varies
and often is preceded by stress, irritation, exposure
to sunlight, cold, fever, trauma, and immunosuppres-
sion [5 – 7]. There are reports that HSV infection can
precede, or be subclinically involved in, an attack of
erythema multiforme, which indicates a possible
antigenic role [8].
Diagnosis of HSV infections is usually based on
the history and clinical findings. Cultures, cytologic
smears that show multinucleation, and special immu-
nofluorescent processing can be helpful when clinical
recognition is uncertain. Serologic diagnosis is of
value only to determine past exposure.
Definitive treatment includes the use of systemic or
topical antiviral drugs [9 – 12] (Table 1). Precursor
antiviral agents, such as valacyclovir and famciclovir,
have better oral bioavailability than acyclovir and pen-
ciclovir. If not used early in the infection (first 3 to

Fig. 3. Herpetic whitlow of the finger contracted by contact with a cold sore.
82 S. Silverman, C.S. Miller / Oral Maxillofacial Surg Clin N Am 15 (2003) 79–89

Table 1 Epstein-Barr viruses

Antiviral drugs
Generic (trade name) Suggested dosage Epstein-Barr virus is associated with infectious
(days)a mononucleosis, hairy leukoplakia, nasopharyngeal
Systemic: acyclovir (Zovirax) 400 mg 3  carcinomas, and lymphomas [13]. The primary infec-
daily (7) tion, referred to as infectious mononucleosis, usually
famciclovir (Famvir) 125 mg 1  causes a sore throat, fever, cervical lymphadenopathy,
daily (5) malaise and pain, and occasional hepatosplenome-
valacyclovir (Valtrex) 500 mg 2  galy. It occurs chiefly in adolescents and young
daily (5) adults. The disease is of low contagiousness, and
Topical: acyclovir (Zovirax) 5% ointment transmission is through exchange of EBV-contami-
penciclovir (Denavir) 1% cream
nated saliva. Affected patients often demonstrate
docosanol (Abreva) 10% cream
multiple petechiae located on the soft palate or lips.
(over-the-counter)
The lymphadenopathy is often bilateral and affects
Apply topical medications to oral lesions at least four
the posterior cervical nodes. Blood studies reveal
times daily.
a atypical lymphocytes, heterophile antibodies, and
Dosage levels are adjusted according to clinical
severity and response. For severe infection: acyclovir 5 – mildly elevated transaminase levels. Treatment in
10 mg/kg IV q8h for 7 – 10 days, famciclovir 500 mg t.i.d., most cases is palliative and supportive. Recovery
or valacyclovir 1000 – 2000 mg b.i.d. For acyclovir-resistant usually occurs within 1 to 2 months; however, the
cases: foscarnet (Foscavir) 40 – 60 mg/kg IV q8h for 7 – 10 virus enters latency in lymphocytes.
days, or cidofovir (Vistide) 5 mg/kg. Epstein-Barr virus is associated with hairy leuko-
plakia, which is a benign manifestation of epithelial
hyperplasia and hyperkeratosis that primarily occurs
4 days), antiviral drugs are usually ineffective. Anti- on the lateral border(s) of the tongue. It appears as a
septic, analgesic, and antiinflammatory medications corrugated white lesion that does not rub off (Fig. 4).
can be beneficial in reducing pain and transmission of Hairy leukoplakia is predominately seen in individuals
the disease. Systemic antiviral agents are used in com- infected with HIV; however, it may be seen in non-
plicated primary infections, HSV infections in the HIV immunosuppressed patients. Hairy leukoplakia is
immunocompromised, prophylaxis for seropositive almost always asymptomatic, with the principal sig-
patients who undergo chemotherapy or transplanta- nificance being a sign of immunosuppression.
tion, HSV-associated central nervous system disease, The diagnosis of hairy leukoplakia is clinically
and recurrent erythema multiforme. Systemic antiviral suggestive and can be confirmed by biopsy. Cyto-
agents are used often as daily prophylaxis. Nephrotox- logic scrapings are somewhat characteristic by fea-
icity, although rare, is a concern with the use of high- turing nucleoprotein condensations in nuclei.
dose systemic acyclovir, valacyclovir, and famciclovir, Treatment is elective and includes high-dose antiviral
particularly if patients have renal insufficiency. drugs, topical podophyllin in 25% tincture of ben-

Fig. 4. Hairy leukoplakia in an HIV-positive patient. The lesion was asymptomatic and chronic.
 S. Silverman, C.S. Miller / Oral Maxillofacial Surg Clin N Am 15 (2003) 79–89
Fig. 5. A lytic bone lesion caused by lymphoma.
zoin, or laser vaporization. Recurrence is common.
Although the term ‘‘leukoplakia’’ has been used to
describe the condition, there is no known associated
precancerous risk.
Epstein-Barr virus is well known for its ability to
cause lymphocyte immortalization and malignant
transformation [13,14]. Such transformation occurs
in a small percentage of infected persons. The trans-
formation process is complex and involves the upre-
gulation of several viral and host gene products and
immunosuppression [13,15]. Critical is the upregula-
Fig. 6. A painful ulceration of the palate of 1 month duration in an HIV-positive patient. Biopsy showed this to be a CMV-
induced lesion that responded to high-dose antiviral medication.
83
tion of EBV latency membrane protein 1 and the
NF-kB pathway that alters apoptotic and growth
pathways [16]. EBV-associated lymphomas in the
head and neck region present as nontender swellings
in Waldeyer’s ring, cervical lymph nodes, salivary
glands, oral mucosa, and lytic bone lesions (Fig. 5).
Persistent fever of unknown cause, weight loss,
malaise, sweating, and abdominal or chest pain often
accompany the condition. Radiation and chemother-
apy are used for treatment. Of interest is the recent
detection of EBV DNA in the serum of patients who
have nasopharyngeal carcinoma, certain lymphomas,
and gastric carcinoma.
Cytomegalovirus
Cytomegalovirus rarely causes mouth lesions but
can be associated with persistent mucosal ulcerations
(Fig. 6) in transplant patients and immunosuppressed
persons. Transmission is person-to-person mainly
through sexual and blood contacts. The primary
infection in most persons goes unrecognized. In utero
and perinatal transmission can lead to deafness,
learning disabilities, and mental retardation, however.
CMV enters latency in peripheral blood mononuclear
cells and reactivates during immunosuppression.
CMV infection recurs commonly in HIV-infected
individuals and presents as pneumonitis, retinitis, or
nervous system disease [17]. CMV can infect and
replicate in major salivary glands, particularly in
immunocompromised patients, which can lead to
swelling, pain, and xerostomia, secondary to lympho-
cytic salivary gland infiltrates.
The diagnosis of CMV infection is established by
special stains and immunoprocessing (polymerase
84 S. Silverman, C.S. Miller / Oral Maxillofacial Surg Clin N Am 15 (2003) 79–89
chain reaction) of biopsy specimens and specific
serologic antibody tests. Successful treatment is
based on establishing the diagnosis and the use of
antiviral agents, such as acyclovir or ganciclovir.
Herpes varicella virus or varicella-zoster virus
Varicella-zoster virus is well known for causing
chickenpox. It is a highly contagious virus that is
generally spread during the late winter and spring
months to young children who lack antibodies against
varicella-zoster virus. After exposure and a 2- to
3-week incubation period, mild prodromal features
appear. The first recognizable signs are fever, mal-
aise, and a distinctive red, itchy truncal rash. The rash
spreads quickly to the neck, face, and extremities and
is followed shortly by the eruption of papules that
form vesicles and pustules. Occasionally, oral inflam-
matory vesicular-ulcerative lesions develop that may
be seen on the posterior palate or buccal mucosa.
Anorexia, chills, fever, nasopharyngitis, and muscu-
loskeletal aches may accompany the disease. Lesions
heal within 7 to 10 days as patient antibody titers rise
and control the infection. Complications such as
pneumonitis and encephalitis are infrequent.
Varicella-zoster virus resides latently in the sen-
sory ganglia of the host after the initial infection. The
virus reactivates in approximately 0.2% of adults,
more often in elderly and immunosuppressed patients
and has an increasing incidence with age. The recur-
rent attack is known as ‘‘herpes zoster’’ or ‘‘shingles.’’
The disease is preceded by hypersensitive skin over-
lying the area of attack. Within a few days, the classic
manifestations are painful vesicles that occur unilat-
erally along nerve dermatomes. Most commonly,
Fig. 7. Varicella (herpes) zoster or ‘‘shingles’’ in a 65-year-old patient. Note the unilateral distribution. The vesicles/ulcers
formed scabs and healed after 3 weeks. Treatment was supportive and included high-dose acyclovir.
lesions occur on the trunk between vertebrae T3 and
L2 and on the face along the ophthalmic division of
the trigeminal nerve and extend up to the midline
(Fig. 7). In immunocompetent patients, vesicles break
down, scab, and resolve within 2 to 4 weeks.
The diagnosis of varicella-zoster virus infection is
made by history, clinical findings, and serology.
Polymerase chain reaction is used for diagnosis in
severe cases that may involve the central nervous
system [18]. Treatment involves antiviral drugs in
high dosages (acyclovir, 4000 mg/d in divided doses;
famciclovir, 1500 mg/d in three divided doses; vala-
cyclovir, 3000 mg/d in two divided doses) given
within the first week [19]. Supportive treatment for
pain and pruritus is also in order. The most painful
and discouraging complication is postherpetic neu-
ropathy, which can be debilitating. Early use of
corticosteroids along with antiviral treatment may
help minimize, or even prevent, the neuropathy.
Longer standing cases benefit from the use of ami-
triptyline, nortriptyline, topical lidocaine patches, and
gabapentin, which should be considered early in the
course of treatment [20,21]. A live-attenuated vac-
cine (Varivax) virus is currently available for the
prevention of chickenpox.
Human herpesvirus-6 and -7
Human herpesvirus-6 and -7 are T-cell lympho-
tropic herpesviruses that have significance in den-
tistry. Like most human herpesviruses, they are
ubiquitous and capable of establishing a lifelong,
latent infection in humans. HHV-6 is particularly
efficient at infecting infants and young children and
produces exanthem subitum (roseola) and febrile
 S. Silverman, C.S. Miller / Oral Maxillofacial Surg Clin N Am 15 (2003) 79–89
seizures. HHV-6 is present in saliva [22] and has been
detected in salivary glands and tonsillar tissue [23].
Primary infection in adults can cause a mononuc-
leosis-like illness. The virus is harbored during
latency in peripheral blood mononuclear cells, sali-
vary glands, mucosa, and tonsils [23]. Reactivation
occurs in the immunocompromised host and is asso-
ciated with fever, leukopenia, encephalitis, interstitial
pneumonitis, skin rash, and bone marrow suppres-
sion. A role for HHV-6 in malignant transformation
remains to be defined, although HHV-6 can trans-
activate other viruses, such as human papillomavirus
(HPV), associated with malignant disease [24].
Human herpesvirus-7 is associated with exanthem
subitum and febrile seizures and is detected in the
saliva of healthy adults [25]. HHV-7 infection gen-
erally occurs later in childhood after infection with
HHV-6. Case reports suggest that HHV-7 may reac-
tivate after immunosuppression or the flu and can
cause encephalitis, encephalopathy, and febrile con-
vulsions [26].
Human herpesvirus -8 (Kaposi’s sarcoma herpesvirus)
Human herpesvirus-8 is a sexually transmitted
herpesvirus that infects endothelial cells. A secondary
source of infection is organ transplantation [27]. Up to
15% of the US adult population is infected [28,29].
Kaposi’s sarcoma herpesvirus contains several viral
oncogenes, and Kaposi’s sarcoma is the primary dis-
ease associated with its infection [30]. Primary effu-
sion lymphoma and multicentric Castleman’s disease
also have a strong association with this virus [31].
Kaposi’s sarcoma is a pseudomalignancy that has
a particularly high prevalence in immunocompro-
mised patients, homosexual men infected with HIV,
Fig. 8. Nodular Kaposi sarcoma of the palate in a patient with AIDS whose CD4 count was less than 200/mm3.
85
and elderly persons. It attacks many different organs
and is usually multifocal. Kaposi’s sarcoma does not
metastasize but can be the cause of death. It occurs
most commonly in the skin, with the mouth being the
second most common site. Within the mouth, the
palate and gingiva are common sites. Kaposi’s sar-
coma is highly vascular and typically appears purple-
red. The initial appearance of Kaposi’s sarcoma is a
purplish macule that enlarges to become a violaceous
papule or nodule. Lesions are single or multifocal,
can cause discomfort and bleeding, and can impact
appearance (Fig. 8).
Since the institution of ‘‘high activity anti-retro-
viral therapy’’, the occurrence of Kaposi’s sarcoma
has been much less frequent. Treatment involves low-
dose radiation, chemotherapy (systemic or intrale-
sional), and surgery.
Human papillomavirus
Human papillomaviruses are small, double-
stranded, non-enveloped DNA viruses That have a
propensity for infecting epithelium. Approximately
5.5 million new HPV infections occur every year in
the United States, and 10% to 33% of sexually active
individuals are infected with the virus. Current esti-
mates indicate that 10% to 18% of adults carry HPV,
with the highest rates of infection found in 19- to
26-year-old individuals.
There are more than 100 types of HPV, and at
least 45 types are known to infect genital and oral
epithelium. HPVs can be harbored latently within
epithelium for the life of the host, undergo a lytic
infection and alter epithelial cell growth and replica-
tion, or dysregulate the cell cycle, which results in
86 S. Silverman, C.S. Miller / Oral Maxillofacial Surg Clin N Am 15 (2003) 79–89
premalignant changes. Infection outcomes depend on
the infecting HPV genotype, anatomic site, and
immune response. Benign genotypes (HPV 6, 11,
44, 55) replicate in the lower epidermis and differ-
entiated cells (keratinocytes) and produce increased
numbers of epithelial cells and koilocytosis. High-
risk genotypes [16,18,31,33,35] are associated with
premalignant and malignant epithelial disease.
Low-risk HPV types are associated with benign
proliferations, such as the squamous papilloma, ver-
ruca vulgaris, condyloma acuminatum (venereal
warts), and focal epithelial hyperplasia (Heck’s dis-
ease). The squamous papilloma is a well-defined,
pink-white, exophytic and pedunculated or stalk-like
mass. This small, usually asymptomatic growth may
occur on any mucosal surface.
Condyloma acuminata intraorally appear as small
verrucous lesions of varying sizes or they can mimic
small fibromas. Condyloma acuminata occur on any
mucosal surface and may be single or multiple, clus-
tered or coalesced, and widespread (Fig. 9). Condylo-
mas are more commonly seen in immunocompromised
patients and are transmitted by oral sex. A higher
prevalence of oral condyloma in HIV-positive patients
on antiretroviral drugs compared to HIV-positive
patients not on antiretroviral drugs has been reported.
Reports have shown a decreased occurrence of most
other HIV-associated oral lesions, however, since
high-activity antiretroviral therapy was initiated.
The common skin wart (verruca vulgaris) is a rare
intraoral finding. It is most often seen on the com-
missures of the lips in children and adolescents and
is caused by autoinoculation. Occasionally it occurs
on the tongue, labial mucosa, or gingiva. Warts have
Fig. 9. Condyloma accuminatum associated with the human papillomavirus in a sexually active homosexual man. Treatment was
by laser surgical removal.
a rough, pebbly, clefted surface and a well-demar-
cated border.
Focal epithelial hyperplasia is a multipapular con-
dition associated with HPV-13 and -32 infection that
was originally reported in Native American Indians
and Eskimos. The infection is transmitted by kissing.
The HPV-induced, small, flat papules occur on labial
and buccal mucosa and tongue. Lesions coalesce and
develop a cobblestone-like surface.
Benign HPV lesions can enlarge and spread or
resolve spontaneously; however, most do not regress
without treatment. Ablative and chemotherapeutic
approaches are useful. Surgery, laser treatment, and
cryotherapy have good success in removing HPV
proliferations when the basal and adjacent epithelium
is removed. High-speed evacuation is recommended
during laser surgery to prevent aspiration of viral
DNA that might be contained within the plume. A
useful topical pharmacologic approach involves
podofilox 0.5% (Condylox, Oclassen), an agent that
causes necrosis by arresting cells in mitosis. It is
applied twice daily for 3 days followed by no
treatment for the next 4 days; then the cycle
is repeated up to four times. Immunomodulatory
approaches include intralesional interferon, imiqui-
mod (Aldara) 5% cream at bedtime three times per
week for up to 16 weeks (alters cell cytokines and
stimulates interferon production), and cimetidine (a
histamine receptor antagonist), 30 mg/kg body
weight (bw), given daily in divided doses usually
over 8 weeks. Recurrence is seen in approximately
10% to 25% of patients generally within 3 months.
Sexual partners should be examined and treated to
minimize the risk of transmission and recurrences.
 S. Silverman, C.S. Miller / Oral Maxillofacial Surg Clin N Am 15 (2003) 79–89
Infection with high-risk HPV types has an asso-
ciation with precancerous oral leukoplakia and
squamous cell carcinoma. In one form of leukoplakia,
proliferative verrucous leukoplakia, HPV-16 has been
identified frequently. This finding helps explain the
high rate of malignant transformation associated with
proliferative verrucous leukoplakia. It can be charac-
terized clinically as a progressive leukoplakic lesion
that can vary from somewhat corrugated and flat to
one that is exophytic and verrucous. Proliferative
verrucous leukoplakia may have a red component
and a variable microscopic presentation that ranges
from hyperkeratosis and epithelial dysplasia to car-
cinoma. It may be asymptomatic, cause slight discom-
fort, or be painful. The lesion usually involves more
than one mucosal surface. It occurs four times more
often in women than men and occurs less commonly
in smokers than nonsmokers. In one long-term study
[32], more than half of proliferative verrucous leuko-
plakias were reported to have transformed into car-
cinoma less than 8 years after diagnosis.
Human papillomavirus can be detected in approx-
imately 30% of oral squamous cell carcinoma cases,
which suggests that the virus may play a causative
role, similar to its role in anogenital carcinoma [33].
The role of HPV in oral cancer seems different from
the carcinogenic effects of tobacco and alcohol. For
example, HPV-positive oropharyngeal cancers occur
less often among moderate to heavy alcohol drinkers
and tobacco smokers. Patients in this category also
have improved survival from cancer when compared
with HPV-negative head and neck squamous cell
carcinoma [34]. The role of HPV in carcinogenesis
may involve binding to p53 suppressor protein,
which in turn enhances epithelial cell proliferation
and neoplasia.
Fig. 10. Oropharyngeal ulcers associated with coxsackievirus infection.
87
Diagnosis of HPV-infected mucosa is based on the
history, clinical findings, and cellular characteristics
seen in biopsy specimens (koilocytes). Confirmation
can be made by immunohistochemistry, in situ hybrid-
ization, or polymerase chain reaction processing.
Currently, specific identification and typing are com-
plex and expensive. There is no effective vaccine, and
antiviral medications do not eliminate latent infection.
The most effective treatment is by surgical approaches
and behavior modification (eg, barrier techniques). In
the future, HPV analyses may prove important for
directing treatment, such as by gene therapy.
Enteroviruses
Enteroviruses are single-stranded, small RNA
positive-sense, nonenveloped viruses that cause a
spectrum of human disease. The family includes
polioviruses, coxsackieviruses (23 serotypes), echo-
viruses (32 serotypes), human enteroviruses 68 to 71,
hepatitis virus A, and several nonhuman enteric
viruses that are well known for causing foot-and-
mouth disease in livestock. Most nonpolio enterovi-
ruses infect nasopharyngeal cells and inhabit the
alimentary (enteric) tract. They are transmitted by
ingestion (fecal-oral route) and contaminated saliva
of close contacts in poor sanitary environments and
warm climates. Many enterovirus infections are
endemic in Southeast Asia [35].
The bulk of enterovirus infections are benign and
self-limiting, manifested mostly by fever alone [36].
The primary enterovirus infections that involve the
oropharyngeal complex are hand-foot-and-mouth dis-
ease and herpangina. These infections occur primarily
in children. Hand-foot-and-mouth disease is caused
88 S. Silverman, C.S. Miller / Oral Maxillofacial Surg Clin N Am 15 (2003) 79–89
Fig. 11. Erosive lichen planus of the buccal mucosa developed in a patient who previously had been diagnosed as hepatitis C
virus positive. A cause-and-effect relationship has not been proved conclusively.
by various members of the Coxsackie group and
enterovirus 71. They produce oral ulcerations and
extraoral ‘‘rashes’’ of the hands and feet (Fig. 10).
Herpangina, caused by group A and sometimes
group B coxsackieviruses, has the main feature of
erythema and multiple vesicles of the oropharynx and
posterior tongue. Both conditions are accompanied
by pain, fever, headache, lymphadenopathy, and
malaise. Infection by nonpolio enteroviruses can
produce lymphonodular pharyngitis, which is man-
ifested mainly by nonulcerative oropharyngeal nod-
ules, abdominal pain, vomiting, conjunctivitis, and
croup. Serious infections can result in meningitis,
encephalitis, paralysis, myocarditis, hepatitis, and
death. Although there is no viral latency, patients
can become reinfected by one of the many serotypes.
The diagnosis of enteroviral infection can be made
by viral culture, serology, and nucleic acid amplifica-
tion [36]. Treatment for self-limiting oropharyngeal
enterovirus infections in immunocompetent individ-
uals is supportive and palliative. Immunoglobulin,
interferon-a, and the antiviral agent pleconaril are
available for treatment of persons with meningitis and
an immunocompromised status [37].
Hepatitis C virus
Hepatitis C virus (HCV) is a small, single-stranded
RNA virus of the Flaviviridae family that was pre-
viously known as one of the non-A/non-B hepatitis
viruses. Infection by HCV is a widespread global
disease. In the United States, it is estimated that
4 million Americans are infected and potential carriers
of HCV, with 30,000 new infections and 10,000
deaths each year [38]. Transmission is highest among
drug users, who share HCV-contaminated needles,
and persons who have large or repeated percutaneous
exposures. Of individuals infected, approximately
80% develop hepatitis, and 10% to 20% of those
persons develop cirrhosis. Treatment involves com-
bination chemotherapy and liver transplantation.
Dental implications relate to the need for
bodily fluid borne – pathogen control in the dental
office, demand for proper sterilization and disinfec-
tion protocols, and use of barrier techniques to protect
against contaminated blood and saliva. HCV has been
detected in saliva [39]; however, it is less infectious
than hepatitis B virus. HCV carriers often are not
aware of their seropositive status.
Many questionable reports [40] indicate a relation-
ship between HCV infection and oral lichen planus.
Although this occurrence has not been documented
conclusively, there seems to be an unexplainable risk
that may not be coincidental. Because oral lichen
planus can be a source of oral pain and interfere with
oral functions, its diagnosis and treatment are impor-
tant considerations. The diagnosis of oral lichen planus
is based its on clinical characteristics (Wickham striae,
erosions, ulcers) and biopsy (Fig. 11). Effective treat-
ment requires the elimination of drugs that can cause
similar appearing lichenoid eruptions, followed by the
use of topical or systemic corticosteroids.
References
[1] Whitley RJ. Herpes simplex virus infections. Lancet
2001;357:1513 – 8.
[2] Kameyama T, Yamamoto S, Hwang CB, Shillitoe EJ.
 S. Silverman, C.S. Miller / Oral Maxillofacial Surg Clin N Am 15 (2003) 79–89
Shedding of herpes simplex virus type 1 into saliva.
J Oral Pathol 1988;17:478 – 81.
[3] Higgins CR, Tatnall FM, Leigh IM. Natural history,
management and complications of herpes labialis.
J Med Virol 1993;1:22 – 6.
[4] Langenberg AG, Ashley RL, Leong WP, Straus SE.
A prospective study of new infections with herpes sim-
plex virus type 1 and type 2. N Engl J Med 1999;3412:
1432 – 8.
[5] Eisen D. The clinical characteristics of intraoral herpes
simplex virus infection in 52 immunocompetent pa-
tients. Oral Surg Oral Med Oral Pathol Oral Radiol
Endod 1998;86:432 – 7.
[6] Logan HL, Lutgendorf S, Hartwig A, et al. Immune
stress, and mood markers related to recurrent oral her-
pes outbreaks. Oral Surg Oral Med Oral Pathol Oral
Radiol Endod 1998;86:48 – 54.
[7] Miller CS. Diagnosis and management of orofacial
herpes simplex virus infections. Dent Clin North Am
1992;36:879 – 95.
[8] Darragh TM, Berger TG, Yen TS. Identification of
herpes simplex virus DNA in lesions of erythema mul-
tiforme by the polymerase chain reaction. J Am Acad
Dermatol 1991;24:23 – 6.
[9] Abramowicz M. Drugs for non-HIV viral infections.
Medical Letter 1999;41:113 – 20.
[10] Balfour Jr HH. Antiviral drugs. N Engl J Med 1999;
16:1255 – 68.
[11] Horowitz E, Pisanty S, Czerninski R, et al. A clinical
evaluation of a novel liposomal carrier for acyclovir in
the topical treatment of recurrent herpes labialis. Oral
Surg Oral Med Oral Pathol Oral Radiol Endod 1999;
87:700 – 5.
[12] Raborn GW, Martel AY, Grace MGA, et al. Oral acy-
clovir in prevention of herpes labialis: a randomized,
double-blind, multi-centered trial. Oral Surg Oral Med
Oral Pathol Oral Radiol Endod 1998;85:55 – 9.
[13] Kieff E. Epstein-Barr virus: new insights. J Infect Dis
1995;171:1323 – 4.
[14] Cruchley AT, Niedobitek G. Epstein-Barr virus: biology
and disease. Oral Diseases 1997;S156 – 63.
[15] Knecht H, Al-Homsi AS, McQuain C, Brousset P. Ep-
stein-Barr virus oncogenesis. Crit Rev Oncol Hematol
1997;26:117 – 35.
[16] Hatzivassiliou E. Cellular signaling pathways engaged
by the Epstein-Barr virus transforming protein LMP1.
Front Biosci 2002;7:319 – 29.
[17] Cunha B. Central nervous system infections in the
compromised host: a diagnostic approach. Infect Dis
Clin North Am 2001;15:567 – 90.
[18] Markoulatos P, Siafakas N, Plakokefalos E, et al. Lab-
oratory diagnosis of common herpesvirus infections of
the central nervous system by a multiplex PCR assay.
J Clin Microbiol 2001;39:426 – 32.
[19] Kost RG, Straus SE. Postherpetic neuralgia: pathogen-
esis, treatment, and prevention. N Engl J Med 1999;
335:32 – 42.
[20] Kanazi GE, Dworkin RH. Treatment of postherpetic
neuralgia: an update. Drugs 2000;59:1113 – 26.
89
[21] Watson CPN. A new treatment for postherpetic neu-
ralgia. N Engl J Med 2000;343:1563 – 5.
[22] Levy J. Three new human herpesviruses (HHV6,7, and
8). Lancet 1997;349:558 – 63.
[23] Roush KS, Margraf LR, Krisher K, et al. Prevalence
and cellular reservoir of latent human herpesvirus 6 in
tonsillar lymphoid tissue. Am J Clin Pathol 2001;116:
648 – 54.
[24] Dockrell DH, Paya CV. Human herpesvirus 6. Mayo
Clin Proc 1999;74:163 – 70.
[25] Wyatt LS. Human herpesvirus 7 is a constitutive inhab-
itant of adult human saliva. J Virol 1992;66:3206 – 9.
[26] Sugaya N, Miura M, Ishizuka T, et al. Influenza ence-
phalopathy associated with infection with human her-
pesvirus 6 and/or human herpesvirus 7. Clin Infect Dis
2002;34:461 – 6.
[27] Regamey N, Wernli M, Witschi A, et al. Transmission
of human herpesvirus 8 infection from renal-transplant
donors to recipients. N Engl J Med 1998;339:1358 – 63.
[28] Baillargeon J, Hettler E, Harrison C, et al. Seropreva-
lence of Kaposi’s sarcoma-associated herpesvirus infec-
tion among blood donors from Texas. Ann Epidemiol
2001;11:512 – 8.
[29] Schulz T. Kaposi’s sarcoma-associated herpesvirus
(human herpesvirus-8). J Gen Virol 1998;79:1573 – 91.
[30] Jaffe HW, Pellett PE. Human herpes virus 8 and Ka-
posi’s sarcoma: some answers, more questions. N Engl
J Med 1999;340:1912 – 3.
[31] Chadburn A, Nador RG, Liu YF, et al. Kaposi’s sarco-
ma-associated herpesvirus sequences in benign lym-
phoid proliferations not associated with human
immunodeficiency virus. Cancer 1997;80:788 – 97.
[32] Silverman S, Gorsky M. Proliferative verrucous leuko-
plakia: a follow-up of 54 cases. Oral Surg Oral Med
Oral Pathol Oral Radiol Endod 1997;84:154 – 7.
[33] Miller CS. Human papillomavirus as a risk factor for
oral squamous cell carcinoma: a meta-analysis, 1982 –
1997. Oral Surg Oral Med Oral Pathol Oral Radiol
Endod 2001;91:622 – 35.
[34] Gillison ML, Capone RB, Spafford M, et al. Evidence
for a causal association between human papillomavirus
and a subset of head and neck cancers. J Natl Cancer
Inst 2000;92:709 – 20.
[35] Lum LC, McMinn PC, Goh AY, et al. Echovirus 7 asso-
ciated encephalomyelitis. J Clin Virol 2002;23:153 – 60.
[36] Sawyer M. Enterovirus infections: diagnosis and treat-
ment. Curr Opin Pediatr 2001;13:65 – 9.
[37] Nigrovic L. What’s new with enteroviral infections?
Curr Opin Pediatr 2001;13:89 – 94.
[38] Lauer GM, Walker BD. Hepatitis C virus infection.
N Engl J Med 2001;345:41 – 52.
[39] Hermida MFM, Barral S, Laredo R, et al. Detection of
HCV RNA in saliva of patients with hepatitis C virus
infection by using a highly sensitive test. J Virol Meth-
ods 2002;101:29 – 35.
[40] Bagan JV, Ramon C, Gonzales L, et al. Preliminary
investigation of the association of oral lichen planus
and hepatitis C. Oral Surg Oral Med Oral Pathol Oral
Radiol Endod 1998;85:532 – 6.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 91 – 102

Diagnosis and treatment of oropharyngeal candidiasis

Joel B. Epstein, DMD, MSD, FRCD(C)a,b
a
Department of Oral Medicine and Diagnostic Sciences, College of Dentistry, Chicago Cancer Center,
801 South Paulina, Chicago, IL 60612, USA
b
Interdisciplinary Program on Oral Cancer, College of Medicine, Chicago Cancer Center,
801 South Paulina, Chicago, IL 60612, USA

Candida, a yeast-like fungus, is present in the oral
cavity of 40% to 60% of the population [1,2]. Candida
albicans is the most commonly isolated species [1,2]
and is the one most likely to cause disease in humans
[3]. Other Candida species include Candida tropica-
lis, Candida krusei, (both of which are prevalent in
immunosuppressed and cancer patients), Candida
guilliermondii, and Candida parapsilosis (of limited
pathogenicity but associated with infection of indwel-
ling vascular access devices). The species of coloniz-
ing organism is important, because resistance to
antifungal agents is more common in some species,
including C. tropicalis and C. krusei. Reports of
resistance to systemic agents are increasing.
Candida is common in the oral and gastrointestinal
flora, and presence in the oral cavity is increased in
patients with dentures, individuals who smoke, per-
sons with xerostomia, and patients who use broad-
spectrum antibiotics and steroids. The use of immu-
nosuppressive purine analogus (eg, fludarabine, cla-
dribine) and antithymocyte globulin increases risk of
invasion. The risk of invasive candidiasis is increased
in persons with neutropenia and mucosal injury. The
use of growth factors to promote recovery of white cell
count in neutropenic patients may reduce the risk of
candidiasis and require further study. Age and prior
splenectomy increase the risk of infection. Many
sedatives, tranquilizers, and some antihypertensive
medications may cause dry mouth, which increases
the risk of oropharyngeal infection. The use of den-
tures and tobacco increases the frequency of coloniza-
tion of the oropharynx and increases the risk of clinical
E-mail address: jepstein@uic.edu
infection. Cases of superficial and invasive candidiasis
are occurring more frequently because of increased
use of antibiotics and immunosuppressive agents and
advances in medical management, such as chemother-
apy, solid organ transplantation and hematopoietic cell
transplantation (HCT), parenteral nutrition, and inva-
sive surgical procedures [4]. Candidiasis of the oro-
pharynx and esophagus is associated with HIV
infection and is a clinical predictor of disease progres-
sion in HIV-infected patients [5,6].
The common portals of entry of Candida are the
oropharynx and gut, sites that are commonly colo-
nized by the organisms [7]. Invasive fungal infections
are one of the major complications in transplant
medicine, potentially resulting in mortality [8]. Heart
transplant recipients develop fungal infections in
approximately 20% of cases [8]. In renal transplant
patients, approximately 5% of infections are caused by
Candida species that present as urinary tract infection
and fungemia [8]. Diagnosis of systemic infection is
difficult, resulting in limited ability to make an early
diagnosis, and invasive infection caused by Candida
and Aspergillus is potentially fatal in neutropenic
patients [7]. Higher rates of infection and mortality
are seen in patients with hematologic cancer as com-
pared to patients with solid tumors. Candida accounts
for up to 70% of invasive fungal infections in cancer
patients, and a doubling of the rate of fungal infections
in all hospitalized patients was reported in the 1980s
[7]. Candida is the fourth most common pathogen in
the blood [7].
Fungal infections, particularly candidiasis and
aspergillosis, in HCT are common and result in
morbidity and mortality in these high-risk patients
[9]. Increasing infections with non-albicans Candida

1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 7 1 - 7
92 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102
have been identified and include species such as C.
tropicalis, C. glabrata, C. parapsilosis, and C. krusei,
which show increased resistance to a commonly used
antifungal agent, fluconazole. C. glabrata and C.
parapsilosis are more common in patients with solid
tumors, and C. tropicalis and C. krusei are more
common in HCT recipients [7]. Intact mucosa serves
as a barrier to systemic infection, but cancer therapies
may lead to a violation of the barrier, which promotes
invasive, systemic infection. Fever caused by fungi
typically occurs 7 to 14 days after neutropenia and is
the most common cause of fever of unknown origin
not responsive to antibiotics [7]. Aspergillus tends to
occur later, 20 days after HCT, and most patients have
nasal or respiratory symptoms, in contrast to invasive
candidiasis, which is typical 7 to 10 days after onset of
neutropenia [7]. In transplant medicine, the risk for
candidiasis also depends on the occurrence of graft-
versus-host disease and the use of immunosuppressive
therapy for its prevention and management [10].
Gradual restoration of immune function occurs after
3 months of HCT, depending on the discontinuation of
immunosuppressive therapy for graft-versus-host dis-
ease. The time to engraftment has decreased because
of changes in transplantation and use of hematopoietic
growth factors, reducing the risk of invasive fungal
infection, which occur more commonly during the
postengraftment period [10].
‘‘Fungal hypersensitivity syndrome,’’ a condition
promoted by the public media, has been associated
with multiple and nonspecific complaints, including
fatigue, impaired concentration and memory, respi-
ratory symptoms and asthma, gastrointestinal com-
plaints, muscle and joint pain, and skin and urogenital
problems. Recent support for this possible condition
was reported in a study in which nystatin capsules to
116 patients were provided in a 4-week randomized
controlled trial. The symptoms were assessed by
questionnaire, and the results showed nystatin to be
superior to placebo in reducing localized and systemic
symptoms [11].
Despite the various topical and systemic agents
available to treat patients with oropharyngeal and
systemic candidiasis, optimal management can be
elusive. Topical therapy is generally effective in
uncompromised patients with oropharyngeal candi-
diasis; however, the optimal treatment for immuno-
compromised patients and patients with chronic or
recurrent infection is not well documented. A general
overview of the etiology, pathogenesis, and treatment
of candidiasis has been presented [4].
The purpose of this article is to review the local
and systemic risk factors for oropharyngeal candi-
diasis, review the clinical manifestations and diag-
nosis of infection, and discuss current approaches
to management.
Risk factors
Local and systemic risk factors and characteristics
of the organisms increase an individual’s susceptibil-
ity to candidiasis. Host factors include age (ie, neo-
nates and elderly people), diabetes, oral prostheses
(particularly acrylic dentures), use of broad-spectrum
antibiotics, steroids and other immunosuppressive
drugs, hyposalivation, disruption of oral mucosa,
dietary factors (eg, high carbohydrate diet, iron defi-
ciency anemia), tobacco use, cancer and cancer thera-
py, and HIV infection (see Box 1). Patients may have
multiple factors that predispose them to candidiasis.
Box 1. Risk factors for development of
oropharyngeal candidiasis
Local factors
Xerostomia (eg, because of radiotherapy,
chemotherapy, Sjögren’s syndrome,
diabetes, medications)
Use of broad-spectrum antibiotics or
steroids
High carbohydrate diet
Leukoplakia, oral cancer
Dentures (eg, poor fit, trauma, unclean-
liness)
Cigarette use
Systemic factors
Neonate, advanced age
Diabetes
Nutritional deficiencies (eg, iron, folate or
vitamin B12 deficiency)
Malignancies (eg, leukemia, agranulocy-
tosis)
Immunosuppression (eg, AIDS, steroid
use)
Microbial factors
Factors related to the organism that have been
implicated include cell-wall components that enhance
adhesion to epithelial cells, hydrophobicity of the
organism, germ tube formation, presence of mycelia,
ability to persist in epithelial cells, production of
enzymes and toxins, induction of tumor necrosis
factor, and phenotypic switching [12]. Candida is a
 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102
dimorphic yeast that demonstrates phenotypic switch-
ing from the yeast form to a filamentous form (pseu-
dohyphal), and although the pseudohyphal form has
been believed to be more pathogenic, both forms may
be associated with clinical infection.
Systemic risk factors
Neonates are susceptible to oropharyngeal candi-
diasis because of their immature immune system and
exposure to the organism during or shortly after birth
[2]. Exposure to C. albicans from infected bottle
nipples or the hands of nurses or the mother also
may lead to oropharyngeal candidiasis [2]. The infec-
tion usually can be treated effectively with topical
agents [13]. Elderly patients may present with a
number of risk factors related to reduced saliva
production, denture use, and immune function, which
result in increased risk, although age alone does not
seem to be a risk factor.
Patients with diabetes are at increased risk of
oropharyngeal candidiasis [2]. They are colonized
more commonly and with higher than normal num-
bers of C. albicans [5,12]. Elevated glucose levels,
reduced chemotactic factor in saliva, altered neutro-
phil function, and reduced saliva volume may play a
role in the pathogenesis of clinical infection [2].
Patients with cancer are at high risk for developing
oropharyngeal candidiasis [3,6,14 – 24]. Oropharyng-
eal candidiasis is reported in up to one third of patients
who receive head and neck radiation therapy, and risk
is increased in patients who are colonized before
radiation therapy [20]. The presence of clinical oro-
pharyngeal candidiasis may complicate oral mucosi-
tis; in patients treated with radiotherapy, risk factors
include xerostomia, mucosal damage, the presence of
oral prostheses, and continuing tobacco and alcohol
use during treatment [20].
In addition to causing mucosal damage, chemo-
therapy may result in transient xerostomia, which may
result in acute oropharyngeal infections. Use of anti-
biotics may result in overgrowth of fungi [22]. Oro-
pharyngeal candidiasis is reported in up to one third of
patients with leukemia [1,3,4,18,20,21]. The intensive
radiation and chemotherapy used to treat these patients
can disrupt the oral mucosa and cause a shift in the oral
flora, thus favoring overgrowth of Candida [22].
Frequent treatment with broad-spectrum antibiotics
in neutropenic patients also alters the normal oral flora
and predisposes to oropharyngeal candidiasis [14].
Candida infections, although generally superficial,
can become invasive or systemic in these patients
and may result in mortality [16,18]. It has been
93
reported that systemic candidiasis develops more
frequently in patients with leukemia who have oro-
pharyngeal candidiasis than in patients who are colon-
ized before treatment, and it is associated with
prolonged neutropenia [16,24]. Candida is present in
approximately 90% of patients with acute leukemia
who undergo chemotherapy [24]. Candidiasis in
patients with leukemia and patients who undergo
HCT causes morbidity (eg, altered taste, oral sensitiv-
ity, dysphagia caused by esophagitis, fever) and results
in considerable risk of mortality during neutropenia
[18,19,21,23]. Systemic candidiasis was shown to be
increased in patients with leukemia with ulcerative
oral mucositis [18]. Oropharyngeal candidiasis also is
associated with long-lasting fever and decreased bone
marrow function in patients with leukemia [23].
Patients with depletion of CD4+ lymphocytes
because of HIV infection are at high risk for oro-
pharyngeal candidiasis. Oropharyngeal candidiasis is
the most prevalent opportunistic oral infection in
AIDS, and it occurs in as many as 95% of patients
[5,6,17,24 – 27]. The development of candidiasis in
HIV-infected patients suggests progression of immu-
nosuppression [5,6], and in patients in whom candi-
diasis has been controlled with therapy, progression or
recurrence of the candidiasis is commonly seen in
advancing disease.
The clinical manifestations of oropharyngeal can-
didiasis in AIDS patients include the pseudomembra-
nous and erythematous forms and angular cheilitis
[5,26]. Pain and change in taste that often accompany
this infection can result in poor appetite, which leads
to weight loss [17]. Oropharyngeal candidiasis can
extend into the esophagus and cause gastrointestinal
bleeding and severe regional and systemic infection
[17,25,28,29]. Therapy should consider an altered
host response to help combat the infection, frequent
relapses, persistent infection that requires prolonged
treatment, potential antifungal drug resistance, and
adverse drug effects [6].
Broad-spectrum antibiotics create a favorable
environment for the proliferation of Candida species
by altering the oral flora [30]. Use of corticosteroids
also places patients at increased risk [30]. Xerostomia-
inducing medications may place the patient at risk of
infection by affecting saliva.
Local and oral factors
Impaired salivary gland function increases the risk
of oropharyngeal candidiasis [1,12,30] Because of the
decreased saliva secretion and low pH. Saliva protects
against candidiasis by diluting and moving organisms
94 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102
from mucosal surfaces. Antimicrobial proteins in the
saliva, including defensins, lactoferrin, sialoperoxi-
dase, lysozyme, histidine-rich polypeptides, and spe-
cific anti-Candida antibodies, interact with the
organism [4]. Lactoferrin, a nonspecific defense factor
in saliva and mucosal secretions, seems to have
fungicidal activity against Candida species.
Drugs that cause hyposalivation predispose
patients to oropharyngeal candidiasis. Patients with
head and neck cancer treated with radiation to the
head and neck develop disturbances of the salivary
glands that may be permanent [12]. HCT recipients
and patients who receive chemotherapy also can
experience salivary gland dysfunction [7].
Oropharyngeal candidiasis is reported in up to
65% of older patients who wear full upper dentures
[3]. Increased susceptibility to oropharyngeal candi-
diasis in denture wearers may be caused by enhanced
adherence of Candida to acrylic, reduced saliva flow
under the denture, poorly fitted dentures, or poor oral
hygiene [1,2].
The risk for oropharyngeal candidiasis is increased
when there is disruption of the oral mucosa [1,12].
The oral epithelium is a physical barrier that prevents
invasion of microorganisms, and epithelial turnover
helps clear adherent organisms from the mouth.
Patients who receive chemotherapy have an altered
rate of mucosal regeneration, which may result in
increased vulnerability to infection [12]. Patients with
mucositis, especially individuals with an immuno-
compromised system, are at high risk for invasive
candidiasis [9,10,17 – 19].
Local use of antimicrobial products and antiin-
flammatory agents (steroids) increases the risk for
colonization and infection [31]. Other studies have
shown that chronic use of oral topical steroids and
steroid inhalers also increases the risk of oropharyng-
eal candidiasis, possibly by suppressing cellular
immunity and inhibiting phagocytosis [2,30,31].
Tobacco use represents a local factor that is asso-
ciated with increased risk of colonization and clinical
infection [20,30].
Clinical presentation
Symptoms that may be associated with Candida
infection include oral and pharyngeal burning, sensi-
tivity, altered taste, and change in the sense of smell. If
involvement extends to the oropharynx, dysphagia and
odynophagia may occur. Oropharyngeal candidiasis
can be classified clinically in several ways (see Box 2)
[3,30,32]. The use of ‘‘acute’’ and ‘‘chronic’’ descrip-
tors of candidiasis should be avoided, particularly in
individuals with chronic immunosuppression, because
oropharyngeal manifestations can persist for extended
periods regardless of the clinical findings.
Box 2. Classification of oropharyngeal
candidiasis
Pseudomembranous (thrush)
Erythematous
Denture stomatitis (atrophic)
Angular cheilitis
Leukoplakia caused by hyperplastic can-
didiasis
Candidal ulceration
The classic form of candidiasis is the pseudomem-
branous form (thrush), which is characterized by soft,
yellowish-white plaques on the oral mucosa that can
be removed with vigorous rubbing and may leave red
or bleeding sites after removal [3,32]. The erythema-
tous form of candidiasis frequently develops in
patients who take antibiotics or use steroid inhalers
and in patients with HIV [3,30]. This form of candi-
diasis is characterized by sensitive and painful ery-
thematous mucosa with few, if any, white plaques.
The dorsal aspect of the tongue and the palate is
generally involved, and the patient presents with red
mucosa with loss of papillae on the tongue and patchy
red changes in the palate, although any portion of the
oral mucosa can be affected [30].
In denture stomatitis, or denture sore mouth (atro-
phic candidiasis), the palatal mucosa in contact with
the denture is affected and is chronically erythematous
and edematous. A hyperplastic response may be seen,
although patients generally experience no symptoms
[3]. Treatment includes correction of denture faults,
careful cleaning of dentures, and antifungal therapy
[33]. Angular cheilitis, an inflammatory reaction at
one or both corners of the mouth, is characterized by
painful red fissures. C. albicans is a common patho-
gen, and infection is frequently accompanied by
Staphylococcus aureus infection [3]. Denture wearers
and patients with HIV are predisposed to this type of
presentation. Although denture stomatitis and angular
cheilitis usually do not indicate serious disease, severe
infections may occur in immunocompromised indi-
viduals [4].
Leukoplakia caused by hyperplastic candidiasis
may represent a precancerous condition that presents
as unilateral or bilateral, elevated, white mucosal
lesions on the buccal mucosa, tongue, lips, and floor
of the mouth [3,29]. Candida species are present in
 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102
the tissue, can be identified on biopsy, and may be
cultured from the mouth. They may not be the cause
of the lesions, however, and may be merely secondary
invaders [3]. Clinical signs and symptoms range from
painless plaque-like white patches to nodular eryth-
roplakic lesions that cause the patient discomfort.
Smoking seems to be a risk factor for development
of hyperplastic candidiasis.
Systemic infection
Systemic Candida infection is usually caused by
C. albicans and less frequently by C. krusei and C.
tropicalis in immunosuppressed patients. Patients
with neutropenia caused by leukemia or from treat-
ment of their malignancy, patients with indwelling
intravascular lines, and patients who receive antibi-
otics or parenteral nutrition are at risk for Candida
septicemia. Candida endocarditis is related to intra-
vascular trauma (eg, cardiac catheterization, surgery)
and is more common on prosthetic valves, where it
can become a life-threatening infection. Candidiasis
that involves the esophagus, trachea, bronchi, or lungs
is an AIDS-defining condition [28].
All forms of systemic disease are serious and cause
morbidity and risk of mortality. Intravenous ampho-
tericin B is the treatment of choice, although ketoco-
nazole or fluconazole is preferred for chronic
mucocutaneous candidiasis [34].
Diagnosis
The diagnosis of candidiasis is based on clinical
signs and symptoms, laboratory testing, and the
response to antifungal treatment. Laboratory tests
include smears from the lesions using Gram’s stain
or a potassium hydroxide preparation or cultures
from the skin, mouth, vagina, urine, sputum, or stool
[5,35]. A culture is obtained if identification of the
species or strain is desired. One must remember,
however, that the diagnosis of oropharyngeal candi-
diasis cannot be based solely on the presence of
Candida species because the organism is a common
commensal in the oral flora [4]. Studies have shown
that high colony counts of Candida species in saliva
collections correlate with the presence of clinical
infection, and quantitative counts may be used in
diagnosis [36]. Occasionally, histologic evidence
is necessary, specifically in cases of hyperplastic
candidiasis (candidal leukoplakia), mucocutaneous
candidiasis, and invasive candida ulceration in immu-
nosuppressed patients.
95
Treatment
Susceptibility testing: identification of species and
resistant strains
Susceptibility testing of fungi to antifungal agents
is not as standardized as that of bacteria but is
evolving, and guidelines are being developed by the
National Committee for Clinical Laboratory Standards
[37]. The current applicability of in vitro antifungal
susceptibility tests is limited by poor standardization
and insufficient correlation of in vitro test results with
clinical outcome [38 – 41]. Antifungal susceptibility
studies may help clarify which strains are likely to
develop resistance during long-term therapy [42].
History of treatment
Gentian violet, an aniline dye, was used for topical
treatment until the first topical polyene antibiotic,
nystatin, became available in 1951 [43]. The first
systemic antifungal agent, amphotericin B, was
reported in 1956 and is the standard against which
newer systemic therapies are compared.
The azoles include the imidazoles and the tria-
zoles. The first azole, benzimidazole, was discovered
in 1944 [43]. Miconazole and clotrimazole (both
identified in 1969) and ketoconazole (1977) are imi-
dazoles. The antifungal activity of the triazoles,
including fluconazole and itraconazole, was reported
in the mid-1980s.
Treatment approach
Underlying risk factors for colonization and infec-
tion should be managed whenever possible. For
example, in patients with hyposalivation, treatment
with sialagogues should be considered. Oral pros-
theses should be disinfected or topical antifungal
agents should be applied to impact colonization of
the acrylic surface, and instructions in denture use
should be provided. If tobacco use is identified as a
risk factor, tobacco cessation should be encouraged.
In patients with diabetes, improved glucose control
may reduce oral candidiasis. In patients with neutro-
penia, use of hematopoietic growth factors may speed
recovery of blood counts, which reduces risk of local
and systemic Candida infection. Changes in use of
medications that increase the risk of Candida col-
onization and infection should be considered (eg, it
may be possible to replace topical steroid use with the
use of other topical immunosuppressive medications).
Various systemic and topical agents are available
to treat oropharyngeal candidiasis (Table 1) [4,44].
96 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102

Table 1 taste and texture may not be tolerated if they are

Routes of action of antifungal agents for the treatment of applied without an occlusive appliance. Creams are
oropharyngeal candidiasis easily applied to the corners of the mouth. A single-
Route of action dose trial of an antifungal agent provides guidance on
Class and agent Topical Systemic product acceptability by the patient.
Gentian violet was commonly used to treat oro-
Polyenes
pharyngeal candidiasis until the advent of polyene
Nystatin X
Amphotericin B X X antifungal medications [30] and was replaced quickly
Imidazoles by the polyenes because of emerging resistance and
Clotrimazole X local side effects, such as staining and irritation of the
Miconazole X oral mucosa [43,47]. Three polyenes—nystatin,
Ketoconazole X amphotericin B, and natamycin (used only for ocular
Triazoles infection)—are available. These agents act by binding
Fluconazole X to ergosterol in the cell membrane of fungi, altering
Itraconazole X cell-membrane permeability, and causing pores and
Other
leakage of cellular components, which leads to micro-
Chlorhexidine gluconate 0.2% X
bial death [43,48].
(antiseptic for denture disinfection)
Denture cleansers X Nystatin is the most widely used agent for the
initial treatment of patients with oropharyngeal candi-
diasis [1,2]. It is available in oral rinse, topical cream,
Topical agents have been the mainstay of therapy, oral pastille, and vaginal tablet and powder [1,2,5].
particularly in uncomplicated cases. If possible, top- Nystatin is not absorbed systemically and lacks serious
ical preparations should be used before systemic toxicity. Adverse effects include nausea, vomiting, and
antifungal drugs [45] because they are not absorbed diarrhea [43], which can be problematic for cancer
systemically and lack adverse systemic effects and the patients who are already nauseated from chemother-
possible drug interactions of systemic agents. Sali- apy. The oral rinse is heavily sweetened with sucrose,
vary action may dilute and rapidly eliminate topical which increases the risk of caries formation. If chosen,
drugs from the oral cavity, however, which reduces nystatin suspension should be used with precautions to
their effectiveness [46], and compliance with required reduce caries risk. The oral rinse may be useful for
repeated topical application impacts efficacy. Topical edentulous patients, but the tablets may not dissolve
agents are available in various formulations, includ- well in the mouth and may be irritating to the mucosa,
ing oral rinses, troches, powders, and vaginal tablets particularly if it is friable or damaged [1]. Although
and creams. Systemic agents are used when topical nystatin is commonly used as prophylaxis for treat-
therapy has been ineffective or not tolerated, pri- ment of oropharyngeal candidiasis in AIDS and cancer
marily in immunocompromised patients with cancer patients, several reports have cited disappointing
or HIV infection. Topical and systemic medications results in these patients, with frequent treatment failure
are used to attempt prophylaxis in patients who re- and early relapse [6,15,16,19].
ceive HCT. Antiseptic agents have been assessed for antifun-
gal effect. An in vitro study of antifungal effects of
Topical therapy ListerineR and chlorhexidine (0.12% and 0.2%, re-
spectively) [49] showed potential value in suppression
The medication chosen should be appropriate for of Candida species. Chlorhexidine rinse has been
the conditions in the oral cavity, and cost should be assessed in patients with leukemia, and suppression
considered. The taste and texture should be consid- of Candida colonization has been demonstrated [31].
ered when selecting a medication and formulation. In patients with symptomatic oral mucositis, alcohol
Unfortunately, many of the oral products include high content, the presence of phenol, and intense flavoring
concentrations of sucrose and have a high cariogenic agents may limit use of commercial antiseptic prod-
potential, which is a particularly important considera- ucts because of mucosal irritation.
tion in dentate patients with dry mouth. In patients Care should be taken when using more than one
with dry mouth, oral rinses should be used because topical agent simultaneously. Two studies that exam-
tablets not only may dissolve poorly but also may ined the compatibility of nystatin with chlorhexidine
become rough with use, which results in physical digluconate in the treatment of patients with oropha-
irritation of the oral tissue. Antifungal creams may ryngeal candidiasis found that both drugs become
be readily applied to the surface of dentures, but their ineffective when combined [50,51]. Several studies
 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102
have compared various topical and systemic agents
(eg, nystatin and amphotericin B with fluconazole or
ketoconazole) to determine what is best tolerated,
most effective, and least expensive in different pa-
tients [19,52 – 55].
The azoles are fungistatic and interfere with syn-
thesis of ergosterol in the fungal organism, which
causes an increased permeability of the cell membrane
[1]. Clotrimazole, an imidazole, was the first broad-
spectrum antifungal agent of its class [1]. Clotrima-
zole troches are more palatable than nystatin oral
suspension and may be used in patients who cannot
tolerate the taste of nystatin [12,56]. Clotrimazaole
has been reported to be effective for prophylaxis and
treatment of oropharyngeal candidiasis in cancer
patients, in whom it may prevent the development
of esophagitis [57], but it seems to be less effective
than fluconazole in treating HIV-infected patients with
oropharyngeal candidiasis [5,58]. Like nystatin oral
troches, however, clotrimazole troches may be poorly
tolerated by AIDS patients or patients with xerostomia
caused by cancer therapy [12]. Miconazole also has
been shown to be effective in patients with oropha-
ryngeal and esophageal candidiasis [59].
When topical agents cannot effectively control
oropharyngeal candidiasis, combining topical therapy
with a systemic agent may eradicate the infection
while allowing a lower dose and shorter course of
the systemic agent [1].
Systemic therapy
Systemic therapy may be necessary in patients
with oropharyngeal candidiasis that is refractory to
Table 2
Strategy for managing fungal infections in high-risk neutropenic patients
Risk factor
Oral colonization
Oral candidiasis
Probable or proven aspergillosis
Possible fungal infection (fever of unknown origin)
Infection by Candida
Infection by Aspergillus
Oral route for prophylaxis and maintenance, parenteral for treatment; dose range 200 – 400 mg/d for fluconazole, with higher
doses in North America.
Modified from Wingard JR. Approach to invasive fungal infection after blood or marrow transplantation. Transplant Proc
2000;32:1543 – 4.
97
topical treatment, in patients who cannot tolerate
topical agents, and in patients at high risk for systemic
infection. It may be chosen for convenience of use and
ease of patient compliance (Table 2).
Amphotericin B
Intravenous amphotericin B has been shown to be
effective in patients with severe oropharyngeal candi-
diasis and in patients with infection refractory to other
agents [2,6]. Low-dose, prophylactic use has been
shown to be effective in neutropenic HCT recipients
[60]. Amphotericin B also has been used prophylacti-
cally in solid organ transplant recipients [61]. Ampho-
tericin B is used primarily in patients at risk for
progressive and potentially fatal fungal infections
because of the potential toxicity of the medication.
Potential toxicities include general toxicities (eg,
fever, shaking chills, malaise, weight loss), renal
toxicity, gastrointestinal effects (eg, nausea, vomiting,
diarrhea, cramping, altered liver function), neurologic
symptoms (eg, headache, hearing loss, dizziness,
visual changes, peripheral neuropathy), muscle/joint
pain, dermatologic reactions (eg, rash, itching), hem-
atologic effects (eg, anemia), and cardiovascular and
pulmonary toxicity. Despite its toxicity, amphotericin
B remains the ‘‘gold standard’’ of antifungal medica-
tions because of its broad spectrum of action, clinical
efficacy, and limited evidence of fungal resistance.
Azoles
The azoles (eg, miconazole, clotrimazole, ketoco-
nazole, fluconazole, and itraconazole) have been
Oral Parenteral
Topical antiseptic/polyene
Fluconazole 400 mg/d 200 mg/d if no po
Topical and oral/parenteral
Fluconazole 400 mg/d 200 mg/d if no po
Itraconazole 400 mg/d Ambisome 200 mg/d
Itraconazole 200 mg/d if no po
Itraconazole 400 mg/d Ambisome 200 mg/d
Itraconazole 200 mg/d if no po
Fluconazole 400 mg/d Fluconazole 400 mg/d for treatment
for maintenance
Intraconaole 400 mg/d Ambisome 200 mg/d
for maintenance
98 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102
widely used to treat patients with superficial and
systemic fungal infections. Ketoconazole was the first
imidazole found to have systemic activity, and
although it is effective in the treatment of oropha-
ryngeal candidiasis in patients with cancer and HIV,
several studies have shown that ketoconazole is less
effective than fluconazole [5,14,30,62,63]. Ketocona-
zole requires gastric acidity for absorption, which may
reduce its effectiveness in patients with AIDS or other
gastrointestinal disorders and persons with limited
food intake [1]. The most frequent adverse effects
described for ketoconazole include nausea, vomiting,
abdominal pain, and itching [1,62]. The adverse event
of greatest concern is hepatotoxicity, which may result
in hepatitis and, rarely, hepatic failure [1,40,43,45,62].
Ketoconazole is less costly than the newer azoles. A
study by Donnelly et al [54] found ketoconazole to be
no more effective than amphotericin B in preventing
yeast infection in neutropenic patients, and use was
not recommended [64].
The triazoles, fluconazole and itraconazole, are in
common use. Both medications have been shown to be
effective in the treatment of oropharyngeal candidiasis
in patients with cancer and HIV infection, and they are
widely used in these patients [15,30,55,57,63,65].
Itraconazole and fluconazole have fewer side effects
than the azoles, and are available in oral formulations.
Fluconazole offers no protection against Aspergillus
species and has been associated with fungal resistance
[27], although this may be overcome in many cases by
increasing the dose. Fluconazole is useful in patients
who require prolonged antifungal therapy, because it is
well tolerated and is taken only once a day [25].
Fluconazole prophylaxis against candidiasis in HCT
patients and patients with leukemia has been con-
ducted, and it is used in some centers [18,66].
Although studies show reduction in Candida coloniza-
tion, not all studies have shown an effect on systemic
infection [18]. Unfortunately, C. krusei is resistant to
fluconazole, and increased infection by C. krusei in
patients with leukemia indicates the need for caution in
the use of fluconazole in these patients [18]. Further
studies are needed to establish optimal doses in
patients with HIV infection and different types of
oropharyngeal candidiasis [25]; studies are also
needed regarding the usefulness of fluconazole in
and HCT patients and patients with leukemia [18,66].
Resistance to antifungal agents
Resistance of Candida to polyenes is virtually
unknown despite years of common clinical use [67].
Reports of resistance to the azoles, however, particu-
larly among AIDS patients, are increasing [26,65,68 –
76]. Therapeutic failure caused by fluconazole-resist-
ant strains in AIDS patients is increasingly reported
[26,68,69,73 – 76], although one study found flucona-
zole effective when other agents were not [77]. Risk of
resistance seems to develop in patients with advanced
HIV disease or after repeated or long-term fluconazole
therapy [26,68,69,73]. Resistant species, specifically
C. krusei, have been increasingly identified in HCT
patients and patients with leukemia because of effec-
tive killing of C. albicans and selection for C. krusei
[18] and have been described with ketoconazole [68].
A study by Fan-Havard et al [78] did not find large-
scale resistance in the Candida populations isolated
from patients who receive long-term azole therapy.
There are reports of fluconazole selecting more re-
sistant strains of Candida species and of cross-resist-
ance between the azoles [30,47,62,72,79]. The
optimal management of fungal infections in children
with HIV remains to be determined [80].
The azoles, particularly ketoconazole, can interact
with many other agents, including antacids, omepra-
zole, histamine-2 antagonists, rifampin, phenytoin,
oral anticoagulants, insulin, cyclosporine, and cortico-
steroids [5,43,62]. Such interactions may result in
either increased or decreased blood levels of these
agents, thus altering their potential efficacy or toxicity.
New therapies
Unlike with the azoles, resistance to polyenes,
including amphotericin B, rarely develops during
therapy [41]. The drug’s principal limitation is its
toxicity when used systemically [41,82]. The use of
amphotericin B oral suspension as a topical antifungal
agent was first studied nearly 40 years ago and has
been available commercially in Europe for the treat-
ment of patients with oropharyngeal candidiasis
[5,47,48]. In vitro studies have shown that amphoter-
icin B is more active than nystatin against C. albicans
[83,84]. Virtually no amphotericin B oral suspension
is absorbed systemically, and the toxicity concerns
associated with intravenous amphotericin B use are
eliminated [47,85 – 88]. Unfortunately, the formula-
tion is poorly tolerated and the rinse has been with-
drawn from the US market [47,62,63,65]. An
amphotericin B lozenge/chlorhexidine combination
has been used in Scandinavia to treat patients with
denture stomatitis [81]. Amphotericin B lozenges are
effective in patients susceptible to Candida infection,
because long-lasting concentrations of the drug in the
saliva can be achieved [81,82].
New formulations of systemic amphotericin B
have been developed with the goal of reducing tox-
icity. Systemic amphotericin B has been studied in
 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102
lipid complex, colloidal dispersion, and liposomal
forms. Although there are limited data on the thera-
peutic advantage of new formulations, toxicity seems
to be reduced [8].
New triazoles in late stage development (vorico-
nazole, posaconazole, ravuconazole) have increased
the spectrum of antifungal activity, including against
Aspergillus and additional Candida species, and
safety seems excellent. New classes of antifungal
agents, including echinocandin agents that affect the
fungal cell membrane (versus cell wall), are in devel-
opment [7]. These agents, with different mechanisms
of action, offer the future potential to manage cases
with new drug combinations.
Combination therapy antifungal agents
Combined therapies have been evaluated to seek
synergy of different agents for use in potentially fatal
fungal infections, including Cryptococcus, C. albi-
cans, and Aspergillus, in immunocompromised and
neutropenic patients [89]. In vitro study of possible
synergy of fluconazole and granulocyte-colony stimu-
lating factor (G-CSF), granulocyte-macrophage-col-
ony stimulating factor (GM-CSF), human serum
treated neutrophils, showed that the combination
caused increased killing by the antifungal agents
[89]. Monoclonal antibodies also were shown to
facilitate fungal cell death in combination with anti-
fungals, as have combinations with proinflammatory
cytokines, interleukin-1, interferon, and tumor ne-
crosis factor-alpha [89]. Use of polyene and azole
antifungal agents has interfered with fungal killing in
vitro; however, animal studies and clinical trials
suggest that dosing schedules of the drugs may be
key, and conflicting results may be related to the
dosing schedule and the animal model used. Although
combining azoles and polyenes may lead to enhanced
effect, there are no clear guidelines for their combined
use [89].
Discussion
Oropharyngeal candidiasis is increasingly com-
mon because of changes in medical care, HIV
infection, and advances in medicine that allow
patients with cancer and other debilitating diseases
to live longer, although in an immunocompromised
state. Use of xerostomia-inducing medications,
broad-spectrum antibiotics, and topical and systemic
steroids has led to increased oropharyngeal coloniza-
tion and subsequent infection. Patients with xerosto-
mia, persons who smoke tobacco, and denture
99
wearers are also at increased risk for candidiasis.
The clinical presentation of this infection may vary
and may be asymptomatic or cause severe, persistent
symptoms. Common oral findings may include white
plaques that can be wiped off, erythema, and, less
commonly, leukoplakia.
Treatment of uncomplicated oropharyngeal can-
didiasis in the non-immunocompromised patient is
usually straightforward. Selecting the form of a
topical medication requires consideration of the oral
condition, the length of contact time with the medi-
cation, and the taste, texture, and cost of the product.
In patients with severe oropharyngeal candidiasis,
particularly individuals with a compromised immune
system, or in patients with refractory oropharyngeal
candidiasis, treatment is often more difficult. Relap-
ses are common, and unless the underlying condi-
tions that predispose to infection are managed
effectively, long-term or frequent intermittent therapy
is often required.
Topical agents are generally effective in uncom-
plicated cases but may be inadequate in some cases,
including complicated infections in patients with
AIDS or neutropenic cancer patients. The most effec-
tive regimen for fungal prophylaxis in patients with
protracted neutropenia has yet to be determined.
Systemic agents, such as amphotericin B, ketocona-
zole, fluconazole, and itraconazole, have been used
extensively in these patients. A major concern with
these agents, particularly with fluconazole, is the
increase in fungal resistance and in selection of
non-albicans species, particularly in patients who
require long-term or recurrent therapy. Strategies
must be developed for preventing and managing the
increasing prevalence of more resistant strains of C.
albicans and selecting resistant species. Intravenous
amphotericin B is the gold standard for patients with
systemic mycoses and, occasionally, persons with
refractory oropharyngeal candidiasis; however, its
routine use for oropharyngeal candidiasis is limited
by its toxicity.
References
[1] Epstein JB. Antifungal therapy in oropharyngeal my-
cotic infections. Oral Surg Oral Med Oral Pathol
1990;69:32 – 41.
[2] Guida RA. Candidiasis of the oropharynx and esoph-
agus. Ear Nose Throat J 1988;67:832 – 40.
[3] Dreizen S. Oral candidiasis. Am J Med 1984;30:28 – 33.
[4] Budtz-Jörgensen E. Etiology, pathogenesis therapy and
prophylaxis of oral yeast infections. Acta Odontol
Scand 1990;48:61 – 9.
100 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102
[5] Greenspan D. Treatment of oropharyngeal candidiasis
in HIV-positive patients. J Am Acad Dermatol 1994;
31(Suppl):51 – 5.
[6] Meyer RD, Holmberg K. Fungal infection in HIV-
infected patients. In: Homberg K, Meyer R, editors.
Diagnosis and therapy of systemic fungal infections.
New York: Raven Press; 1989. p. 79 – 100.
[7] Wingard JR, Leather HL. Empiric antifungal therapy
for the neutropenic patient. Oncology 2001;15:351 – 69.
[8] de Pauw BE. Advances in the management of invasive
fungal infections in organ transplant recipients: step by
step. Transpl Infect Dis 2000;2:48 – 50.
[9] Donnelly JP. A strategy for managing fungal infections
in haematopoietic stem cell transplantation. Transplant
Infect Dis 2000;3:88 – 95.
[10] Wingard JR. Approach to invasive fungal infection
after blood or marrow transplantation. Transplant Proc
2000;32:1543 – 4.
[11] Santelmann H, Laerum E, Roennevig J, Fagertun HE.
Effectiveness of nystatin in polysymptomatic patients:
a randomized, double-blind trial with nystatin versus
placebo in general practice. Fam Pract 2001;18:258 – 65.
[12] Peterson DE. Oral candidiasis. Clin Geriatr Med 1992;
8:513 – 27.
[13] Dhondt F, Ninane J, De Benle K, et al. Oral candidosis:
treatment with absorbable and non-absorbable antifun-
gal agents in children. Mycoses 1992;35:1 – 8.
[14] Francis P, Walsh TJ. Current approaches to the man-
agement of fungal infections in cancer patients. Oncol-
ogy 1992;6:81 – 92.
[15] Prentice AG. Oral and gastrointestinal candidosis: pro-
phylaxis during immunosuppressive therapy. Mycoses
1989;32(Suppl 2):42 – 6.
[16] DeGregorio MW, Lee WMF, Ries CA. Candida infec-
tions in patients with acute leukemia: ineffectiveness of
nystatin prophylaxis and relationship between oropha-
ryngeal and systemic candidiasis. Cancer 1982;50:
2780 – 4.
[17] Stevens DA, Green SI, Lang OS. Thrush can be pre-
vented in patients with acquired immunodeficiency
syndrome and the acquired immunodeficiency syn-
drome-related complex. Arch Intern Med 1991;151:
2458 – 64.
[18] Epstein JB, Ransier A, Lunn R, Chin E, Jacobson J, Le
N, et al. Prophylaxis of candidiasis in patients with
leukemia and bone marrow transplants. Oral Surg Oral
Med Oral Pathol Oral Radiol Endod 1996;81:291 – 6.
[19] Epstein JB, Vickars L, Spinelli J, Reece D. Efficacy of
chlorhexidine and nystatin rinses in prevention of oral
complications in leukemia and bone marrow trans-
plantation. Oral Surg Oral Med Oral Pathol 1991;
73:682 – 9.
[20] Epstein JB, Freilich MM, Le ND. Risk factors for
oropharyngeal candidiasis in patients who receive ra-
diation therapy for malignant conditions of the head
and neck. Oral Surg Oral Med Oral Pathol 1993;76:
169 – 74.
[21] Bergmann OJ. Alterations in oral microflora and
pathogenesis of acute oral infections during remis-
sion-induction therapy in patients with acute myelo-
leukemia. Scand J Infect Dis 1991;2:355 – 66.
[22] Bergmann OJ, Andersen PL. Acute oral candidiasis
during febrile episodes in immunocompromised pa-
tients with haematologic malignancies. Scand J Infect
Dis 1990;22:353 – 8.
[23] Rodu B, Carpenter JT, Jones MR. The pathogenesis and
clinical significance of cytologically detectable oral
Candida in acute leukemia. Cancer 1988;62:2042 – 6.
[24] Dupont B, Graybiull JR, Armstrong D, et al. Fungal
infections in AIDS patients. Journal of Medical and
Veterinary Mycology 1992;30(Suppl 1):19 – 28.
[25] Koks CHW, Schepens MHJ, Burger DM, et al. Drug
development report (9): fluconazole in the treatment
and prophylaxis of oral candidosis in HIV-infected pa-
tients. J Drug Dev Clin Pract 1993;5:235 – 49.
[26] Heinic GS, Stevens DA, Greenspan D. Fluconazole-
resistant Candida in AIDS patients. Oral Surg Oral
Med Oral Pathol 1993;76:711 – 5.
[27] Pollock JJ, Santarpia III RP , Heller HM, et al. De-
termination of salivary anticandidal activities in
healthy adults and patients with AIDS: a pilot study.
J Acquir Immune Defic Syndr Hum Retrovirol 1992;
5:610 – 8.
[28] Conant MA. The AIDS epidemic. J Am Acad Derma-
tol 1994;31:847 – 50.
[29] Jewell ME, Sweet DE. Oral and dermatologic mani-
festations of HIV infection. Postgrad Med 1994;96:
105 – 16.
[30] Edwards JE. Candida species. In: Mandell GL, Bennett
JE, Dolin R, editors. Principles and practices of infec-
tious diseases. New York: Churchill-Livingstone;
1995. p. 2289 – 301.
[31] Epstein JB, Komiyama K, Duncan D. Oral topical ste-
roids and secondary oral candidiasis. J Oral Med 1986;
41:223 – 7.
[32] Crislip MA, Edwards JE. Candidiasis. Infect Dis Clin
North Am 1989;3:103 – 33.
[33] Lewis MAO, Samaranyake LP, Lamey PJ. Diagnosis
and treatment of oral candidiasis. J Oral Maxillofac
Surg 1991;49:996 – 1002.
[34] Bissell V, Felix DH, Wray D. Comparative trial of
fluconazole and amphotericin in the treatment of den-
ture stomatitis. Oral Surg Oral Med Oral Pathol 1993;
76:35 – 9.
[35] Berkow R. The Merck manual of diagnosis and ther-
apy. 16th edition. Rahway (NJ): Merck & Co.; 1992.
[36] Epstein JB, Truelove EL, Izutzu KT. Oral candidiasis:
pathogenesis and host defense. Rev Infect Dis 1984;
6:96 – 106.
[37] Epstein JB, Pearshall NN, Truelove EL. Quantitative
relationships between Candida albicans in saliva and
the clinical status of human subjects. J Clin Microbiol
1980;12:475 – 6.
[38] Rex JH, Cooper CR, Merz WG, et al. Detection of
amphotericin B-resistant Candida isolates in a broth-
based system. Antimicrob Agents Chemother 1995;39:
906 – 9.
[39] Barchiesi F, Del Poeta M, Morbiducci V, et al. Turbi-
 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102
dimetric and visual criteria for determining the in vitro
activity of six antifungal against Candida spp and
Cryptococcus neoformans. Mycopathologia 1993;
124:19 – 25.
[40] Woods GL, Washington JA. The clinical and the mi-
crobiology laboratory. In: Mandell GL, Bennett JE,
Dolin R, editors. Principles and practices of infectious
diseases. New York: Churchill-Livingstone; 1995.
p. 169 – 99.
[41] Galgiani JN. Antifungal susceptibility tests. Antimi-
crob Agents Chemother 1987;31:1867 – 70.
[42] Galls HA, Drew RH, Pickard WW. Amphotericin B:
30 years of clinical experience. Rev Infect Dis 1990;
12:308 – 29.
[43] Korting HC, Ollert M, Georgii A, Fröschl M. In vitro
susceptibilities and biotypes of Candida albicans iso-
lates from the oral cavities of patients infected with
human immunodeficiency virus. J Clin Microbiol
1988;26:2626 – 31.
[44] Gupta AK, Sauder DN, Shear NH. Antifungal agents:
an overview. Part I. J Am Acad Dermatol 1994;30:
677 – 98.
[45] Judd G. Oral candidiasis: treatment choices. Continu-
ing Education 1989;16:21 – 2.
[46] Datry A. Candidose digestive et infection vih: actu-
alites cliniques et therapeutiaques. Mycology Medicine
1992;2(Suppl I):5 – 14.
[47] Samaranayake LP, Ferguson MM. Delivery of antifun-
gal agents to the oral cavity. Adv Drug Deliv Rev
1994;13:161 – 79.
[48] Dewsnup DH, Stevens DA. Efficacy of oral amphoter-
icin B in AIDS patients with thrush clinically resistant
to fluconazole. Journal of Medical and Veterinary My-
cology 1994;32:389 – 93.
[49] Meiller TF, Kelley JI, Jabra-Rizk MA, DePaola LG,
Baqui A, Falkler WA. In vitro studies of the efficacy of
antimicrobials against fungi. Oral Surg Oral Med Oral
Pathol Oral Radiol Endod 2001;91:663 – 70.
[50] Bennett JE. Antimicrobial agents, antifungal agents.
In: Gilman AG, Rall TM, Nies AS, editors. The phar-
macological basis of therapeutics. 8th edition. New
York: Pergamon Press; 1990. p. 1165 – 81.
[51] Barkvoll P, Atramadal A. Effect on nystatin and chlo-
rhexidine digluconate on Candida albicans. Oral Surg
Oral Med Oral Pathol 1989;67:279 – 81.
[52] Barkvoll P, Hurlen B. Conventional treatment or oral
candidiasis: new aspects. Nor Tannlaegeforen Tid
1989;99:116 – 9.
[53] Philpott-Howard JN, Wade JJ, Mufti GJ, et al.
Randomized comparison of oral fluconazole versus
oral polyenes for the prevention of fungal infection
in patients at risk of neutropenia. J Antimicrob Che-
mother 1993;31:973 – 84.
[54] Egger T, Gratwohl A, Tichelli A, et al. Comparison of
fluconazole with oral polyenes in the prevention of
fungal infections in neutropenic patients: a prospective,
randomized, single-center study. Support Care Cancer
1995;3:139 – 46.
[55] Donnelly JP, Starke ID, Galton DAG, et al. Oral keto-
101
conazole and amphotericin B for the prevention of
yeast colonization in patients with acute leukemia.
J Hosp Infect 1984;5:83 – 91.
[56] Menichetti F, Del Favero A, Martino P, et al. Prevent-
ing fungal infection in neutropenic patients with acute
leukemia: fluconazole compared with oral amphoteri-
cin B. Ann Intern Med 1994;120:913 – 8.
[57] Owens NJ, Nightingale CH, Schweizer RT, et al. Pro-
phylaxis of oral candidiasis with clotrimazole troches.
Arch Intern Med 1984;144:290 – 3.
[58] Shechtman LB, Funaro L, Robin T, et al. Clotrimazole
treatment of oral candidiasis in patients with neoplastic
disease. Am J Med 1984;76:91 – 4.
[59] Goldstein SM. Advances in the treatment of superficial
Candida infections. Semin Dermatol 1993;12:315 – 30.
[60] Uchida K, Yamaguchi H. Susceptibility to miconazole
(base) of isolates from the oral cavity and esophagus of
patients with mycosis. Jpn J Antibiot 1991;44:109 – 16.
[61] Perfect JR, Klotman ME, Gilbert CC, et al. Prophylac-
tic intravenous amphotericin B in neutropenic autolo-
gous bone marrow transplant recipients. J Infect Dis
1992;165:891 – 7.
[62] Gupta KL, Ghosh AK, Kochhar R, et al. Esophageal
candidiasis after renal transplantation: comparative
study in patients on different immunosuppressive pro-
tocols. Am J Gastroenterol 1994;89:1062 – 5.
[63] van der Bijl P, Arendorf TM. Itraconazole and fluco-
nazole in oropharyngeal candidiasis. Ann Dent 1993;
52:12 – 6.
[64] Meunier F. Fluconazole treatment of fungal infections
in the immunocompromised host. Semin Oncol 1990;
17(Suppl 6):19 – 23.
[65] Hansen RM, Reinerio N, Sohnle PG, et al. Ketocona-
zole in the prevention of candidiasis in patients with
cancer: a prospective, randomized, controlled, double-
blind study. Arch Intern Med 1987;147:710 – 2.
[66] Goodman JL, Winston DJ, Greenfield RA, et al.
A controlled trial of fluconazole to prevent fungal in-
fections in patients undergoing bone marrow trans-
plantation. N Engl J Med 1992;326:845 – 51.
[67] Hay RJ. Overview of studies of fluconazole in oropha-
ryngeal candidiasis. Rev Infect Dis 1990;12(Suppl 3):
334 – 7.
[68] Conly A, Renin R, Johnson J, et al. Disseminated can-
didosis due to amphoteric B resistant Candida albi-
cans. J Infect Dis 1992;165:761 – 4.
[69] Rex JH, Rinaldi MG, Pfaller MA. Resistance of Can-
dida species to fluconazole. Antimicrob Agents Che-
mother 1995;39:1 – 8.
[70] White A, Goetz MB. Azole-resistant Candida albi-
cans: report of two cases of resistance to fluconazole
and review. Clin Infect Dis 1994;19:687 – 92.
[71] Ng TTC, Denning DW. Fluconazole resistance in Can-
dida in patients with AIDS: a therapeutic approach.
J Infect 1993;26:117 – 25.
[72] Newman SL, Flanigan TP, Fisher A, et al. Clinically
significant mucosal candidiasis resistant to fluconazole
treatment in patients with AIDS. Clin Infect Dis 1994;
19:684 – 6.
102 J.B. Epstein / Oral Maxillofacial Surg Clin N Am 15 (2003) 91–102
[73] Chavanet P, Lopez J, Grappin M, et al. Cross-sectional
study of the susceptibility of Candida isolates to anti-
fungal drugs and in vitro-in vivo correlation in HIV-
infected patients. AIDS 1994;8:945 – 50.
[74] Samaranayake LP, Holmstrup P. Oral candidiasis and
human immunodeficiency virus infection. J Oral Path-
ol Med 1989;18:554 – 64.
[75] Manso E, Montillo M, Discepoli G, Leoni P. Flucona-
zole resistance of Candida krusei. Boli Ist Sieroter
Milan 1991;70:527 – 9.
[76] Redding S, Smith J, Farinacci G. Resistance of Candida
albicans to fluconazole during treatment of orophar-
yngeal candidiasis in a patient with AIDS: documenta-
tion by in vitro susceptibility testing and DNA subtype
analysis. Clin Infect Dis 1994;18:240 – 2.
[77] Laguna F, Rodriguez-Tudela JL, Enriquez A. Funge-
mia due to fluconazole-resistant Candida albicans in a
patient with AIDS. Clin Infect Dis 1994;19:542 – 3.
[78] Lucatorto FM, Franker C, Hardy WD, Chafey S. Treat-
ment of refractory oral candidiasis with fluconazole:
a case report. Oral Surg Oral Med Oral Pathol 1991;
71:42 – 4.
[79] Fan-Havard P, Capano D, Smith SM, et al. Develop-
ment of resistance in Candida isolates from patients
receiving prolonged antifungal therapy. Antimicrob
Agents Chemother 1991;35:2302 – 5.
[80] Akova M, Akalin HE, Uzun O, Gur D. Emergence of
Candida krusei infections after therapy of oropharyng-
eal candidiasis with fluconazole. Eur J Clin Microbiol
Infect Dis 1991;10:598 – 9.
[81] de Vries-Hospers HG, Mulder NH, Sleijfer DT, van
Saene HK. The effect of amphotericin B lozenges on
the presence and number of Candida cells in the oro-
pharynx of neutropenic leukemia patients. Infection
1982;10:71 – 5.
[82] de Vries-Hospers HG, van der Waaij D. Salivary con-
centrations of amphotericin B following its use as an
oral lozenge. Infection 1980;8:63 – 5.
[83] Terrell CL, Hermans PE. Antifungal agents used for
deep-seated mycotic infections. Mayo Clin Proc
1987;62:1116 – 28.
[84] Ghosh A, Ghosh JJ. Effect of nystatin and amphoter-
icin B on the growth of Candida albicans. Ann Bio-
chem Exp Med 1963;23:29 – 44.
[85] Lechevalier H. Comparison of the in vitro activity of
four polyenic antifungal antibiotics. Antibiotic Annual
1959;614 – 8.
[86] Alban J, Groel JT. Amphotericin B oral suspension in the
treatment of thrush. Curr Ther Res 1970;12:479 – 84.
[87] Brandell R, Chase SL, Cohn JR. Treatment of oral
candidiasis with amphotericin B solution. Clin Pharm
1988;7:70 – 2.
[88] Stough AR, Groel JT, Kroeger WH. Amphotericin B, a
new antifungal agent for the prophylaxis of antibiotic-
induced moniliasis. Antibiot Med Clin Ther 1959;6:
653 – 61.
[89] Stevens DA, Kullberg BJ, Brummer E, Casadevall A,
Netea MG, Sugar AM. Combined treatment: antifungal
drugs with antibodies, cytokines or drugs. Medical
Mycology 2000;3:305 – 15.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 103 – 110

Management of head and neck infections in the

immunocompromised patient
Newton C. Gordon, DDS, MS*, Stephen Connelly, DDS
Department of Dentistry/Oral and Maxillofacial Surgery, San Francisco General Hospital, 1001 Potrero Avenue,
NH-1N1, San Francisco, CA 94110, USA

The term immunocompromised has traditionally Normal immune system

been used to describe patients with a serious impair-
ment of one or more aspects of their immune defense
mechanisms. As a result, these patients are more
susceptible than immunocompetent hosts to the estab-
lishment of bacterial and nonbacterial infections [1].
The number of patients classified as immunocompro-
mised has increased in the last 15 years. Among the
many factors contributing to the increased numbers of
such patients is the fact that people are outliving once
fatal diseases because of advances in modern treat-
ment strategies [2].
The most common conditions leading to an
immunocompromised state can be divided into four
general categories: systemic conditions, congenital
defects or primary immunodeficiencies, iatrogenic
causes, and social factors (Box 1). The purpose of
this article is to review three common causes of
immunosuppression in the oral surgery patient: dia-
betes mellitus, alcoholism/substance abuse, and HIV/
AIDS, with emphasis on their impact on head and
neck infections. Each of these conditions leads to a
host environment that is more susceptible to severe
pathogenic invasion, causing infections such as
osteomyelitis, pan-facial abscesses, and necrotizing
fascitis. Management of these infections requires
accurate diagnosis, aggressive incision and drainage,
proper antimicrobial therapy, and improved nutri-
tional status to achieve resolution.
* Corresponding author.
E-mail address: newtong@itsa.ucsf.edu (N.C. Gordon).
To understand the defects in the immune system
produced by these conditions, it is first necessary to
discuss the immune system in the noncompromised
host. The normal functioning immune system
involves a complex network of specialized cells and
defensive barriers designed to protect an individual
from potential pathogens. Its development begins in
the first month of gestation with the hematopoietic
stem cells located in the yolk sac [3]. In the third
month of gestation, hematopoiesis occurs mainly in
the liver and continues up to the point when the
skeletal elements are formed and the bone marrow
becomes the major site of blood cell formation. A
variety of cells differentiate from the hematopoietic
stem cell, including granulocytes, monocytes, lym-
phocytes, megakaryocytes, and erythrocytes. Two
months into gestation, lymphocytic cells destined to
become T cells emigrate from the bone marrow into
the developing thymus for maturation. The matura-
tion of B cells occurs under the influence of stromal
reticular cells. As the fetus continues to develop, so
do the peripheral components of the immune system,
including the blood, thymus, lymphatic system,
spleen, skin, and mucosa [4].
The immune system is often divided into the
innate and adaptive, or acquired immune systems.
Innate immunity consists of antigen-nonspecific
defense mechanisms activated immediately on
encounter with an antigen [5,6]. These mechanisms
include physical barriers like epithelium, fatty acids,
mucus, and cilia. Soluble factors such as proteins of
the complement cascade, chemokines (proteins that

1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 7 9 - 1
104 N.C. Gordon, S. Connelly / Oral Maxillofacial Surg Clin N Am 15 (2003) 103–110

tion because their specificity, selected during thymic

Box 1. Etiology of immunodeficiency
Systemic conditions
AIDS
Diabetes mellitus
End-stage renal disease
Leukemia/lymphoma
Systemic lupus erythematous
Advanced age
Primary immunodeficiencies
X-linked severe combined immunode-
ficiency
Wiskott-Aldrich syndrome
Chediak-Higashi syndrome
DiGeorge syndrome
Iatrogenic causes
Immunosuppressive drugs
Broad-spectrum antibiotics
Chemotherapy
Radiation therapy
Bone marrow transplantation
Social factors
Alcoholism
Illicit drug use
Obesity
induce leukocyte migration), cytokines (proteins that
modulate leukocyte function), and leukocytes other
than T cells and B cells are all part of the innate
system that becomes activated in response to the
chemical properties of the insulting agent.
Adaptive immunity is a system that is antigen-
specific. The main effector cells of this antigen-
specific system are the T cells and B cells [7,8]. This
system requires antigen processing, which involves
the recognition of antigenic epitopes by T cells and
B cells, clonal expansion, and differentiation of these
antigen-specific lymphocytes into effector cells [9].
In addition, random rearrangement of genes that
encode T-cell receptors (TCRs) on T cells and immu-
noglobulins (Ig) on B cells results in a vast repertoire
of antigen-specific receptors. This enables the host to
react to thousands of foreign substances [10,11].
Ig genes have the ability to increase specificity and
affinity of antibodies by undergoing further somatic
mutation. TCR genes do not undergo somatic muta-
ontogeny, is retained for self – nonself discrimination.
The adaptive immune system can form memory of
antigenic exposure, decreasing the time for the sub-
sequent immune response to re-exposure to an anti-
gen. The antigen-driven clonal expansion of T cells
and B cells is central to adaptive immunity and is the
basis for immunologic memory.
The innate and adaptive immune systems are
complementary. For example, the chemokines and
cytokines produced by macrophages in an inflamma-
tory reaction serve to attract and modulate T cell and
B cell activation and function. In addition, the
complement cascade can be activated via the classic
pathway by Ig stimulation [12].
The central cellular elements of the immune
system are the leukocytes, which consist of granulo-
cytes, specialized antigen-presenting cells, and lym-
phocytes. Granulocytes include the neutrophils,
eosinophils, basophils, and mast cells. The neutro-
phils, monocytes, and macrophages are responsible
for the phagocytic destruction of antigens. Their
function is crucial in the acute inflammatory re-
sponse. Eosinophils play a role in the late phase of
allergic inflammation and are responsible for the
destruction of parasitic infections. Basophils and mast
cells both express IgE and participate in the imme-
diate hypersensitivity immune response. Monocytes,
macrophages, Langerhans cells, Kupffer cells, and
dendritic cells comprise the antigen processing and
presenting cells [13].
There are three major types of lymphocytes:
T cells, B cells, and natural killer (NK) cells. T cells
are phenotypically defined by the expression of the
TCR heterodimeric receptor on their cell surface,
which binds antigen displayed by antigen-presenting
cells [14]. B cells express transmembrane Ig on their
surface, which binds unprocessed antigen indepen-
dent of antigen-presenting cells. NK cells are mor-
phologically large granular lymphocytes. They are
phenotypically defined by the absence of either trans-
membrane cell surface expression of TCR or Ig and
by the presence of the cell surface molecules (protein
markers), CD16 and CD56 [15]. T cells and B cells
are responsible for clonally specific immune re-
sponses, while NK cells provide innate cytotoxic
immune responses directed against virus-infected
cells and tumor cells. There is also cooperation
between the NK cells and the adaptive immune
response through Fc-bound IgG and the production
of cytokines [4,5].
All nucleated cells of the body display the trans-
membrane class I human leukocyte antigen (HLA)
molecule on their surface, which is encoded by the
 N.C. Gordon, S. Connelly / Oral Maxillofacial Surg Clin N Am 15 (2003) 103–110
major histocompatability complex. The class I HLA
molecule is restricted to presenting linear peptides of
8 to 25 amino acids long to CD8+ T cells [16]. They
present endogenous antigens, proteins synthesized
within the antigen-presenting cells. These proteins
are derived from the processing of genetic informa-
tion from either viral, intracellular bacterial, or tumor
sources. This is consistent with the specificity of
CD8+ T cells. Only macrophages, monocytes, den-
dritic cells, and B lymphocytes present the class II
HLA molecule. However, any nucleated cell stimu-
lated with interferon-g may also express the class II
HLA molecule [17]. The specific antigen-presenting
cells present only the class II HLA molecule to CD4+
cells. The antigens are exogenous peptides, derived
from phagocytized bacteria, parasites and virus par-
ticles. Class I and II HLA molecules are cell surface
heterodimeric structures with a groove for presenta-
tion of linear peptides only. However, there is also a
mechanism for the presentation of polysaccharides.
CD4 and CD8 T-cells are responsive to lipoglycan
antigens presented by CD1 molecules, which are
expressed on most antigen presenting cells [18].
The TCR is responsible for the specificity and
sensitivity of T cell recognition of antigens presented
by the HLA molecule [19]. T cells go through a
process of positive and negative selection that occurs
in the thymus during T cell development. Positive
selection occurs in the cortical region of the thymus.
T cells whose antigen receptor fails to bind to self-
HLA molecules are programmed to die by apoptosis,
whereas T cells whose antigen receptor binds to self-
HLA molecules survive and migrate to the medulla.
Negative selection occurs next when T cells whose
antigen receptor binds to self-HLA molecules with
too high an affinity display auto-antigens and
undergo apoptosis. This process creates T cells that
can recognize HLA molecules correctly but yet not
bind to those HLA molecules presenting self antigens
[20]. T cells can be further defined by their cytokine
production and can be either T helper1 (Th1) cells
(which generate cell-mediated immune responses) or
Th2 cells (which generate humoral allergic immune
responses). Both CD4+ and CD8+ T cells can exhibit
either cytokine profile.
The hallmark of the B cell is the production of Ig.
Maturation of B cells depends on bone marrow
stromal cells and stromal cell-produced interleukin-
7 (IL-7) [21]. B cells can express IgM, IgD, IgG1-4,
IgE, and IgA1-2 isotypes. IgM and IgD isotypes
characterize a naı̈ve mature B cell and switching to
the other isotypes is T cell – dependent. B cell mem-
brane Ig and secreted Ig are alternative products of
the differentially spliced Ig heavy chain gene. They
105
are produced when a specific antigen binds to the
membrane Ig receptor, which transduces an intra-
cellular signal stimulating clonal expansion to pro-
duce more cells and secretes Ig specific for that
antigen. This production is a T cell – dependent pro-
cess, in which these cells direct immunoglobulin
isotype switching through a series of cell surface
molecular interactions with B cells, resulting in
reciprocal intracellular signaling and T cell elabora-
tion of cytokines [22]. Repetitive antigen stimulation
is associated with somatic hypermutation within the
Ig gene segments, encoding the heavy and light chain
variable regions, resulting in Ig with greater antigen
specificity. The daughter clones are preferentially
expanded and produce antibody with higher affinity,
all occurring within the germinal centers of the
lymphoid organs.
The complement system facilitates antibody-medi-
ated immunity. The complement system consists of
plasma proteins activated along an enzymatic cascade,
resulting in a spectrum of bioactive molecules that
facilitate opsonization, osmotic lysis of targeted cells,
and recruitment of phagocytic cells [23]. Through the
classic complement pathway, antigen – antibody com-
plexes efficiently activate the complement cascade.
The alternative pathway involves independent activa-
tion of cascade protein C3.
Cell adhesion molecules are also very important to
the proper functioning of the immune system. Selec-
tins, integrins, and Ig superfamily adhesion molecules
are the three families of adhesion molecules that
allow leukocytes to attach to extracellular matrices
and to adhere to each other. Selectins are found on all
leukocytes and function as lectins, which bind to
carbohydrate moieties expressed by endothelial cells
or other leukocytes. Selectin L, E, and P participate in
leukocyte rolling along vascular endothelium. Integ-
rins and Ig superfamily adhesion molecules bind
through protein – protein interactions, which is
important for stopping leukocyte rolling and medi-
ating leukocyte aggregation and transendothelial
migration [31]. The adhesion molecules are often
dysfunctional in patients with immunocompromising
condition, such as diabetes.
Immune defects in specific diseases
Diabetes mellitus
The negative effects of diabetes mellitus on the
immune system have been extensively investigated.
These effects impact greatly on the host’s ability to
prevent the establishment of, and bring resolution to,
106 N.C. Gordon, S. Connelly / Oral Maxillofacial Surg Clin N Am 15 (2003) 103–110
a variety of head and neck infections [24,25]. The
main etiologic factor in diabetes mellitus that leads to
dysfunction in the immune system is hyperglycemia
[26,27]. All the major cell types involved in the
immune defense are affected. Cellular elements of
the innate immune system, including neutrophils and
monocytes/macrophages, have altered function. In
the neutrophils, functions such as adherence, chemo-
taxis, and phagocytosis may be down- regulated.
This results in a less effective defense against a
microbial challenge [28 – 30]. The neutrophils from
diabetic patients also produce less free oxygen rad-
icals, which reduce their ability to make toxic metab-
olites for release against microbes [8]. Monocytes
and macrophages may have up-regulated catabolism
of pro-inflammatory cytokines as well as increased
production of matrix metalloproteases, such as colla-
genase [31 – 33]. This creates an imbalance that is
detrimental to the containment of head and neck
infections. The hyperglycemic state may also lead
to a decrease in fibroblast proliferation and synthesis
of collagen, impairing tissue turnover and wound
repair [34,35].
It has been proposed that the formation of
advanced glycation end-products (AGEPs), which
form as a result of glucose irreversibly binding to
proteins and lipids in the face of prolonged hyper-
glycemia, is a key event in the generation of the
defects seen in diabetes. [29,33]. Glycation end-
products can bind to receptors on various cells, such
as leukocytes, and affect their function. The up-
regulation of tissue destructive cytokines produced
by the monocytes and macrophages may be a result
of AGEP binding. AGEPs also alter the solubility of
collagen and may play a role in the changes seen in
small and large blood vessels. This collagen inter-
action may result in the accumulation of AGEPs on
the basement membrane, affecting the exchange of
nutrition, neutrophil migration, and the diffusion of
antibodies and oxygen. As a group, these effects
should have a detrimental effect on the wound heal-
ing apparatus.
There is a significant body of evidence that
indicates that cystolic Ca++ is elevated in many cell
types in patients with both type I and type II diabetes
mellitus [36,37]. The high levels of cystolic Ca++,
ultimately induced by hyperglycemia, could lead to a
reduced ATP content and decreased phagocytic abil-
ity in neutrophils [38]. Additionally, increased intra-
cellular Ca++ may affect the cellular components of
the acquired immune system. B cells may have an
impaired proliferative response to mitogen, not unlike
that seen in the altered ionic environment of chronic
renal failure [39].
Alcoholism
It has been demonstrated that ingestion of large
amounts of ethanol leads to a relatively broad impair-
ment of host defense mechanisms [40]. Ethanol
impairs phagocytic cells, including neutrophils,
monocytes, and macrophages [41]. Also, significant
decreases in T cell subpopulations, including CD4,
CD8, and CD3, are seen in patients with alcoholic
hepatitis [42]. These abnormalities were significantly
correlated with protein malnutrition or kwashiorkor-
like changes, but not with primary caloric malnutri-
tion or marasmus-like changes. In addition, in this
same population, it was noted that a large percentage
of patients displayed lowered CD4 cell count num-
bers (250 – 300 cells/mm3), which were similar to
counts seen in HIV-infected individuals vulnerable
to the onset of opportunistic infections. Ethanol
seems to blunt the activation of normal human
circulating CD3 T cells in terms of their ability to
produce IL-2 and thus decreases their proliferative
potential in response to mitogenic stimulation. The
final stages of CD4 T cell maturation may be
impaired, affecting the generation of Th1 and Th2
patterns of cytokine response [43]. It is noteworthy
that the most profound impairment in T cell prolif-
eration occurred during the period after cessation of
ethanol ingestion when withdrawal symptoms are
evident. In this period, there are high levels of
corticosteroids in the circulation, suggesting that this
may be the primary mode for immunosuppression in
alcoholics. An additional subset of T cells, the NK
cells, have been shown to be dysfunctional in alco-
holic patients, decreasing the hosts ability to elim-
inate virus-infected cells and tumor cells [44].
B cell activation is dependent on the proper
functioning of the T cell and its ability to produce a
Th2 cytokine response. If this mechanism is dysfunc-
tional, as has been demonstrated in alcoholic patients,
then the production of antigen specific immunoglo-
bulins is greatly reduced [45].
HIV-infected patients
The characteristic immune defect in HIV infection
is the destruction of CD4+ T cells [46]. This loss of
CD4+ T cells affects a variety of immune cells and
their respective functions. CD4+ T cells exhibit
decreased lymphokine secretion (IL-2 and inter-
feron-g) and a decreased response to soluble antigens,
increasing susceptibility to opportunistic infections
and neoplasms. CD8+ T cells have a decreased
cytotoxic response, decreasing a host’s ability to fend
off intracellular organisms. NK cells have a decreased
 N.C. Gordon, S. Connelly / Oral Maxillofacial Surg Clin N Am 15 (2003) 103–110
ability to kill tumor cells. Macrophages exhibit
diminished cytotoxic ability, decreased chemotaxis,
reduced IL-1 secretion, poor antigen presentation,
and decreased class II HLA antigen expression,
contributing to a depressed antigen response and
defective wound healing. B cells show depressed Ig
production in response to new antigens, and they are
refractory to normal signals for B cell activation [47].
There is also polyclonal activation of B cells resulting
in hypergammaglobulinemia and circulating immune
complexes [48].
A significant number of patients with AIDS will
develop neutropenia as a result of direct retroviral
infection, use of antiretroviral drugs and other drug
therapy, systemic infections, and autoimmune mech-
anisms [49]. In addition, the neutrophils of patients
with AIDS have defective bactericidal function and
chemotatic defects [50,51]. Consequently, it has been
proposed that the impairment in neutrophil function
along with defective immunoglobulin synthesis are
important causes of the increased risk of bacterial
infections in patients with advanced HIV disease [52].
Also, there has been a strong correlation between the
degree of neutropenia and risk of developing serious
bacterial infections in cancer patients [53].
Management of infections
The management of head and neck infections in
the immunocompromised patient should follow the
same basic steps established for immunocompetent
individuals. These steps are outlined in Box 2. The
Box 2. Management of head and neck
infections in the immunocompromised
patient
1. Airway monitoring and possible sur-
gical airway establishment
2. Comprehensive history and physical
examination
3. Obtaining appropriate laboratory and
imaging studies
4. Empiric antimicrobial therapy
5. Surgical debridement and irrigation,
as needed
6. Culture and antibiotic sensitivity test-
ing of infectious organisms to appro-
priately adjust antibiotic therapy
7. Close follow-up to monitor for reso-
lution and recurrence
107
first consideration is assuming that there is no acute
airway compromise. If the airway is compromised,
the first course of action is either oral or nasal
endotracheal intubation. In instances where swelling
of the oropharyngeal airway is severe, it may be
necessary to perform a tracheostomy to establish a
competent airway.
Once the airway is secured, the next step should
be to obtain a thorough history and physical exam-
ination. This provides information about concomitant
diseases, social habits, or other processes that may be
indications of an immunocompromised status. Such
indicators include a positive history of HIV, a pre-
vious diagnosis of diabetes, or signs and symptoms of
the disease, alcohol or illicit drug use, renal dialysis,
and a recent history of recurrent infections. When
performing the physical examination, one should
keep in mind that immunocompromised individuals
may have an attenuated immune response resulting in
decreased signs and symptoms of inflammation.
Thus, a serious infection may not necessarily have a
dramatic clinical presentation.
The next step is to obtain the necessary laboratory
and imaging studies to establish the diagnosis and
determine the extent of the infection. Routine labor-
atory studies such as a white blood cell count,
hemoglobin and hematocrit determination, a platelet
count, and measurement of electrolytes, blood urea
nitrogen, creatinine, and glucose should be per-
formed. It is also helpful to obtain a differential white
cell count. A high percentage of immature neutro-
phils would indicate that the immune system is
struggling to produce cells to fight the infection.
Also, a decreased lymphocyte count may be indic-
ative of an HIV infection. Imaging studies may
include plain films, CT scans with or without con-
trast, MRI, and radionucleotide bone scanning (skel-
etal scintigraphy; see elsewhere in this issue).
Empiric antibiotic treatment should be instituted
as soon as possible to rapidly obtain minimum
inhibitory concentrations in the plasma. A recom-
mended antibiotic regimen is outlined in Box 3.
The initial selection is based on the duration of the
infection and the level of immunocompetence of the
patient. If the polymicrobial infection has been pre-
sent for less than 3 days, the amount of cross-
colonization or synergy that has developed is limited.
After 3 days, enough time has passed so that there is
increased virulence of some microorganisms due to
the changes in the environment. For example, aerobic
bacteria may produce a more favorable environment
for anaerobic bacteria, allowing them to multiply
more readily. If the infection responds well to the
empiric use of an antibiotic, the regimen should be
108 N.C. Gordon, S. Connelly / Oral Maxillofacial Surg Clin N Am 15 (2003) 103–110
Box 3. Empiric antibiotic treatment
Early infection (first 3 days of symptoms
or mildly immunocompromised)
Penicillin
Clindamycin
Cephalexin (or other first-generation
cephalosporin)
Late infection (after 3 days of symptoms
or moderately to severely immunocom-
promised)
Clindamycin (maximum dose)
Penicillin and metronidazole
Ampicillin and sulbactam
Cephalosporin (first or second
generation)
Mild, moderate, and severe comprom-
ise is based on CD4/viral loads, glycemic
control, and the degree of alcoholic-
related disease.
Modified from Flynn TR. The swol-
len face. Severe odontogenic infec-
tions. Emerg Med Clin N Am 2000;18:
481 – 519.
continued even if the culture and antibiotic sensitivity
test indicates a change may be appropriate. However,
in the absence of clinical improvement, the culture
and antibiotic sensitivity test results should form the
basis for continued antimicrobial therapy.
If a patient is mildly immunocompromised, there
should be an adequate immune response to assist the
antimicrobial drug. However, in the moderate to
severely immunocompromised patient there is more
dependence on the antimicrobial drug to control the
infection. Along with the use of antibiotics, surgical
debridement or incision and drainage rank as the most
important interventions in the management of head
and neck infections. This includes removal of all nidi
of infection (teeth, necrotic tissue, and nonvital bone),
exploration and irrigation of all involved fascial
spaces, and proper wound care. The lack of vascular-
ity can result in failure of immune effectors and
antibiotics to reach the infected sites. Proper place-
ment of drains and periodic irrigation through them is
essential for the continuous removal of necrotic
debris and enhancement of vascularization. Speci-
mens for culture should be taken at the time of
surgical debridement or incision and drainage. Both
aerobic and anaerobic cultures should be done and
antibiotic sensitivity testing should be performed to
provide guidance in selecting the correct antimicro-
bial treatment. Finally, close postoperative monitor-
ing, including repeated imaging studies, additional
surgical interventions, and long-term follow-up are
mandatory for resolution of the infection.
Management of head and neck infections involves
the contribution of a triad of factors: the host, the
antibiotic, and the surgical intervention. In the normal
host, serious polymicrobial infections can be pro-
duced by organisms that may not be infectious in
pure culture but that become infectious through
microbial synergism. Consequently, antimicrobial
therapy aimed at one major organism may be enough
to break the chain and change the environment
sufficiently to allow the immune system to take over
and bring about resolution of the infection. In the
normal host, surgical intervention, as well as anti-
biotic therapy, significantly alter the microbial envir-
onment, allowing the patient’s immune system to
phagocytize the remaining bacteria. In an immuno-
compromised patient, the host component of the triad
is weakened, leaving the surgeon to rely almost
entirely on surgical intervention and antimicrobial
therapy to resolve the infection. This is the rationale
for recommending aggressive surgical incision and
drainage, frequent irrigation of the drains, and use of
high-dose, broad-spectrum antimicrobial therapy.
Again, it must be emphasized that antimicrobial
therapy must be supported by aerobic and anaerobic
culturing and antibiotic sensitivity testing.
Summary
The immunocompromised host has a potential
increased risk for severe head and neck infections
that usually require aggressive antimicrobial therapy
and prolonged hospitalization. The causes of the
immunocompromised status of patients who seek care
from the oral and maxillofacial surgeon are multi-
factorial and include diabetes, malnutrition, obesity,
alcohol abuse, tobacco abuse, intravenous drug abuse,
cocaine abuse, HIV infection, and AIDS. Patients with
other diseases, such as organ transplant recipients and
those receiving cancer therapy or therapy for various
autoimmune diseases, are not frequently encountered
with severe infections of the head and neck.
The number of patients with multifaceted causes of
immunocompromise will clearly increase in the future
as the population ages and medical treatments for
 N.C. Gordon, S. Connelly / Oral Maxillofacial Surg Clin N Am 15 (2003) 103–110
previously morbid or lethal conditions are improved.
Recognizing the conditions associated with decreased
immune function is critical for the proper management
of these patients. Concurrent with recognition of
immunocompromising diseases, it is important to
have a basic understanding of the normal immune
system and associated defects, because advances in
therapy will undoubtedly increase in complexity. For
instance, new antimicrobials, as well as pharmaceut-
icals that alter cytokine function and affect the gen-
eration of progenitor cells in the bone marrow stroma,
are on the horizon. Future treatment strategies will not
only include aggressive use of traditional management
methods but also these new approaches. Ultimately,
this should provide a shorter course of treatment and
improved outcomes for immunocompromised patients
with head and neck infections.
References
[1] Klastersky J. Infections in immunocompromised pa-
tients I. Pathogenesis, etiology, and diagnosis. Clin
Ther 1985;8:90 – 9.
[2] Kumate J. Infectious diseases in the 21st century. Arch
Med Res 1997;28:155 – 61.
[3] Lewis D, Harriman GR. Cells and tissues of the im-
mune system. In: Rich RR, Fleisher TA, Schwartz
BD, et al, editors. Clinical immunology principles and
practice. St. Louis: Mosby-Year Book Inc.; 1996.
p. 15 – 38.
[4] Huston DP. The biology of the immune system. JAMA
1997;278:1804 – 14.
[5] Fearon DT, Locksley RM. The instructive role of in-
nate immunity in the acquired immune response. Sci-
ence 1996;272:50 – 3.
[6] Fleisher TA. Immune function. Pediatr Rev 1997;18:
251 – 6.
[7] Kishimoto TK, Anderson DC. The role of integrins in
inflammation. In: Gallin JI, Goldstein IM, Snyderman
R, editors. Infammation: basic principles and clinical
correlates. New York: Raven Press; 1992. p. 353 – 406.
[8] Weissmann IL, Cooper MD. How the immune system
develops. Sci Am 1993;269:64 – 71.
[9] Brodsky FM, Guagliardi L. The cell biology of antigen
processing and presentation. Annu Rev Immunol
1991;9:707 – 44.
[10] Davis MM. T cell receptor gene diversity and selection.
Annu Rev Biochem 1990;59:475 – 96.
[11] Schatz DG, Oettinger MA, Schlissel MS. V(D)J re-
combination. Annu Rev Immunol 1992;10:359 – 83.
[12] Roitt IM. Immunology. 4th edition. Barcelona: Mosby;
1998.
[13] Rodgers JR, Rich RR. Antigens and antigen presenta-
tion. In: Rich RR, Fleisher TA, Schwartz BD, et al,
editors. Clinical immunology principles and practice.
St. Louis: Mosby-Year Book Inc.; 1996. p. 114 – 31.
109
[14] Germain RN, Margulies DH. The biochemistry and
cell biology of antigen processing and presentation.
Annu Rev Immunol 1993;11:403 – 50.
[15] Trinchieri G. Biology of natural killer cells. Adv Im-
munol 1989;47:187 – 376.
[16] Salter RD, Benjamin RJ, Wesley PK, et al. A binding
site for the T-cell-co-receptor CD8 on the alpha3 do-
main of HLA-A2. Nature 1990;345:41 – 6.
[17] Skoskiewcz MJ, Colvin RB, Scheenberger EE, et al.
Widespread and selective induction of major histocom-
patibility complex-determined antigens by gamma in-
terferon. J Exp Med 1985;162:1645 – 64.
[18] Sieling PA, Chatterjee D, Porceli SA, et al. CD1-re-
stricted T cell recognition of microbial lipoglycan anti-
gens. Science 1995;269:227 – 30.
[19] Bjorkman PJ, Davis MM. Model for the interaction of
T-cell receptors with peptide/MHC complexes. Cold
Spring Harb Symp Quant Biol 1989;54:365 – 73.
[20] Haynes B, Denning SM, Le PT, et al. Human intra-
thymic T cell differentiation. Semin Immunol 1990;2:
67 – 77.
[21] Goodwin RG, Lupton S, Schmierer A, et al. Human
interleukin 7. Proc Natl Acad Sci U S A 1989;86:
302 – 6.
[22] Noelle RJ, Ledbetter JA, Aruffa A. CD40 and its li-
gand, an essential ligand-receptor pair for thymus de-
pendent B cell activation. Immunol Today 1992;13:
431 – 3.
[23] Frank MM, Fries LF. The role of complement in in-
flammation and phagocytosis. Immunol Today 1991;
12:322 – 6.
[24] Harrison GA, Schultz TZ, Schaberg SJ. Deep neck
infection complicated by diabetes mellitus. Oral Surg
Oral Med Oral Pathol 1983;55:133 – 7.
[25] Sugata T, Fujita Y, Myoken Y, Fujioka Y. Cervical
cellulitis with mediastinitis from an odontogenic infec-
tion complicated by diabetes mellitus: report of a case.
J Oral Maxillofac Surg 1997;55:864 – 9.
[26] Leibovici L, Yehezkelli Y, Porter A. Influence of dia-
betes mellitus and glycemic control on the character-
istics and outcome of common infections. Diabet Med
1996;13:457 – 63.
[27] Vernillo AT. Diabetes mellitus: relevance to dental
treatment. Oral Surg Oral Med Oral Pathol Oral Radiol
Endod 2001;91:263 – 70.
[28] Cianciola LJ, Park BH, Bruck E, et al. Prevalence of
periodontal disease in insulin-dependent diabetes mel-
litus. J Am Dent Assoc 1982;104:653 – 60.
[29] Mattson J, Cerutis D. Diabetes mellitus: a review of the
literature and dental implications. Compendium
2001;22:757 – 72.
[30] Naghibi M, Smith R, Baltch A, et al. The effect of
diabetes mellitus on chemotatic and bactericidal activ-
ity of human polymorphonuclear leukocytes. Diabetic
Research and Clinical Practice 1987;4:27 – 35.
[31] Campbell MJ. A light and electron microscope study
of blood vessels from the gingival tissues of non-
diabetic and diabetic patients. Aust Dent J 1971;16:
235 – 9.
110 N.C. Gordon, S. Connelly / Oral Maxillofacial Surg Clin N Am 15 (2003) 103–110
[32] Grossi SG, Genco RJ. Periodontal disease and diabetes
mellitus: a two-way relationship. Ann Periodontol
1998;3:51 – 61.
[33] Lalla E, Lamster IB, Schmidt AM. Enhanced interac-
tion of advanced glycation end-products with their
cellular receptor for RAGE: implications for the patho-
genesis of accelerated periodontal disease in diabetes.
Ann Periodontol 1998;3:13 – 9.
[34] Golub LM, Nicoll GA, Iacono VJ, et al. In vivo
crevicular leukocyte response to a chemotatic chal-
lenge: inhibition by experimental diabetes. Infect
Imm 1982;37:1013 – 20.
[35] Ryan ME, Ramamurthy NS, Sorsa T, et al. MMP-medi-
ated events in diabetes. Ann N Y Acad Sci 1994;878:
311 – 34.
[36] Draznin B, Lewis D, Houlder N, et al. Mechanism of
insulin resistance induced by sustained levels of cyto-
solic free Ca in rat adipocytes. Endocrinology 1989;
125:2341 – 9.
[37] Rapoport Y, Himelfarb MZ, Zikk D, et al. Cervical
necrotizing fasciitis of odontogenic origin. Oral Surg
Oral Med Oral Pathol 1991;72:15 – 8.
[38] Alexiewicz J, Kumnar D, Smogorzewski M, et al.
Polymorphonuclear leukocytes in non-insulin depend-
ent diabetes mellitus: abnormalities in metabolism
and function. Ann Intern Med 1995;123:919 – 24.
[39] Gaciong Z, Alexiwcz JM, Linker-Israeli M, et al. In-
hibition of immunoglobulin prodution by parathyroid
hormone: implications in chronic renal failure. J Am
Soc Nephrol 1990;1:236 – 44.
[40] MacGregor RR. Alcohol and immune defense. JAMA
1986;256:1474 – 9.
[41] Brayton RG, Stokes PE, Schwartz MS, et al. Effect of
alcohol and various diseases on leukocyte mobiliza-
tion, phagocytosis and intracellular bacterial killing.
N Engl J Med 1970;282:123 – 8.
[42] Roselle GA, Medenhal CL, Chedid A, et al. Alco-
hol modulation of immune function: clinical and
experimental data. Alcohol Clin Exp Res 1995;19:
551 – 4.
[43] Chiappelli F, Kung M, Lee P, et al. Alcohol modulation
of human normal T-cell activation, maturation, and
migration. Alcohol Clin Exp Res 1995;19:539 – 44.
[44] Saxena QB, Mezey E, Adler WH. Regulation of natu-
ral killer activity in vivo. II: The effect of alcohol
consumption on human peripheral blood natural killer
activity. Int J Cancer 1980;26:413 – 7.
[45] Smith KA. Interleukin-2: inception, impact, and impli-
cations. Science 1988;242:1169 – 76.
[46] Fichtenbaum CJ, Dunagan WC, Powderly WG. Bac-
teremia in hospitalized patients infected with human
immunodeficiency virus: a case-control study of risk
factors and outcome. J Acquir Immune Defic Syndr
Hum Retrovirol 1995;8:51 – 7.
[47] Janeway CA, Travers P, Hunt S, Walport M. Immu-
nobiology. 3rd edition. New York: Garland Publish-
ing; 1997. p. 153 – 75.
[48] Lane HC, Masur H, Edgar LC, et al. Abnormalities of
B-cell activation and immunoregulation in patients
with the acquired immunodeficiency syndrome.
N Engl J of Med 1983;309:453 – 8.
[49] Murphy MF, Metcalfe P, Waters AH, et al. Incidence
and mechanism of neutropenia and thrombocytopenia
in patients with HIV. Br J Haematol 1987;66:337 – 40.
[50] Murphy PM, Clifford LH, Fauci AS, et al. Impairment
of neutrophil bactericidal capacity in patients with
AIDS. J Infect Dis 1988;158:627 – 30.
[51] Nielsen H, Kharazmi A, Faber V. Blood monocyte and
neutrophil functions in acquired immune deficiency
syndrome. Scand J Immunol 1986;24:291 – 6.
[52] Moore RD, Keruly JC, Chaisson RE. Neutropenia and
bacterial infection in acquired immunodeficiency syn-
drome. Arch Intern Med 1995;155:1965 – 70.
[53] Chanock S. Evolving risk factors for infectious com-
plications of cancer therapy. Hematol Oncol Clin North
Am 1993;7:771 – 93.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 111 – 122
Diagnosis and treatment of recurrent aphthous stomatitis
Ellen Eisenberg, DMD
Division of Oral and Maxillofacial Pathology, University of Connecticut School of Dental Medicine,
University of Connecticut Health Center, 263 Farmington Avenue, Farmington, CT 06030, USA
What is recurrent aphthous stomatitis (RAS)? Is it
appropriate to regard it as an infectious disease? Are
there effective measures for managing aphthous
ulcers? These are just a few of the questions that
are continually asked about this benign but highly
symptomatic oral problem.
Although it is one of the most common recurrent
oral ulcerative conditions of adults and children
recognized throughout the world, RAS is also one
of the least understood oral diseases and is among the
most vexing problems faced by affected patients and
clinicians alike [1]. The triggering factors that pre-
cipitate recurrent episodes in RAS patients seem to be
as diverse and unique as the affected individuals
themselves [2], which has posed a challenge for
scientists in their attempts to identify a specific
causation for this disease. Although the origin of
RAS remains obscure, there is growing lucidity with
regard to its pathogenesis, and that enlightenment has
influenced contemporary approaches to its manage-
ment significantly [3].
Etiology and pathogenesis
Historically, conjecture about the origin of RAS
focused on a wide spectrum of potential local and
systemic factors that encompassed microbial agents,
hematologic and hormonal disturbances, physical
injury, emotional stress, and other influences [4]. To
date, however, despite years of collective effort on the
part of many researchers, the precise cause of RAS
continues to elude disclosure. Also confounding the
search for a singular cause is the observation that
E-mail address: eeisenberg@nso2.uchc.edu
1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 7 0 - 5
aphthous-like oral ulcers often occur in conjunction
with diverse conditions of a systemic nature. Included
among those conditions are cyclic neutropenia [5],
selected anemias [6], inflammatory bowel diseases [7],
Behçet’s disease (see Box 1) [8,9], gluten-sensitive
enteropathy (celiac sprue) [10,11], relapsing poly-
chondritis syndromes (including the so-called
‘‘MAGIC’’ syndrome, which consists of mouth and
genital ulcers with inflamed cartilage) [12], HIV
infection [13], the purported symptom complex of
recurring fevers, aphthous stomatitis, pharyngitis, and
lymphadenopathy (FAPA syndrome) [14], and others
[2,4 – 6,15]. Given the various presentations of RAS
as either an exclusively local oral phenomenon or as a
systemically related oral condition, it is not surprising
that the search for its origin has proved frustrating.
In recent years a body of evidence has emerged to
suggest a genetic and an immunologic basis for RAS.
These revelations largely have eclipsed speculation
that RAS is caused by an infectious microorganism
or one of the other previously suspected etiologic
factors. They also have led to more rational and ef-
fective contemporary approaches to the management
of RAS [16].
Genetic factors
There is an apparent familial predisposition for
aphthous stomatitis. As compared to the general
population, the prevalence of RAS is higher when
there is a positive family history, especially when both
parents are affected [17]. There is also increased
disease correlation observed in identical twins as
compared to fraternal twins [18]. In familial cases of
RAS, the onset of disease is earlier and attacks tend to
occur more frequently than in nonfamilial cases [19].
112 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122
Box 1. Behçet’s disease
A recurrent, multisystemic, immuno-
logically mediated vasculitic syndrome
Essential diagnostic component:
Oral aphthous ulcers (three or more
recurrences in one 12-month period)
accompanied by any two of the following:
Skin
 Pustular or nodular cutaneous lesions
 Positive cutaneous pathergy (ie, pus-
tule formation at injection site 24 to
48 hours after injection of sterile
solution of inert substance)
 Superficial thrombophlebitis, ery-
thema, ulcerations, acneiform lesions
Ocular
 Uveitis
 Retinal vasculitis
Genital
 Recurring ulcers of penis, scrotum,
or vulva
Less common manifestations:
Inflammatory bowel disease-related
symptoms
Arthritis
Neurologic disturbances
 Psychiatric disorders
 Meningoencephalitis
 Brain stem abnormalities
Data from International Study Group
for Behçet’s Disease. Criteria for diagnosis
of Behçet’s disease. Lancet 335:1078,
1990.
The likelihood that RAS is a genetically grounded
disease is further supported by the recognized,
although not entirely consistent, identification of
certain histocompatibility antigen (HLA) types (eg,
HLA B12, B51, Cw7), among some groups of
aphthous patients [2]. The latter type includes the
largest group of individuals whose aphthous ulcers
(‘‘canker sores’’) are isolated to the oral cavity
exclusively, without an attendant underlying systemic
disease, individuals with the mucocutaneous form of
Behcß et’s disease [8], persons with celiac disease (in
which there is immune-induced gluten sensitivity and
resultant intestinal malabsorption), and patients with
inflammatory bowel diseases (specifically Crohn’s
disease and ulcerative colitis) and oral aphthae [7,10].
Immunologic factors
Although findings have not been entirely consist-
ent and conflicting theories persist, mounting scien-
tific testimony supports immune dysregulation as a
key mechanism underlying the pathogenesis of RAS.
It is believed that the altered immune reactivity arises
perhaps in response to, or in concert with, a state of
presumably heightened antigenic stimulation [16]
exacted on a diminished mucosal barrier [2].
A constellation of cell-mediated immunologic
phenomena seems to be a consistent factor in the
disease. Serologic studies that compared RAS
patients and unaffected controls revealed diminished
ratios of circulating CD4+ helper cells to CD8+
suppressor cells in the former group [20,21]. It has
been proposed that in RAS some unspecified anti-
genic influence [2,22] is at the epicenter of an
antibody-dependent, T cell-mediated immune re-
sponse that involves a shift in local lymphocytic
subpopulations that eventuates in tissue damage
[23 – 25]. From the observations of several investiga-
tors it has been hypothesized that the entire process of
aphthous ulceration, from initiation through progres-
sion, is instigated by the expression on oral epithelial
cells of not only normally found HLA class I antigens
but also HLA class II antigens [26]. Presumably, this
renders the cells antigenically ‘‘foreign’’ and conse-
quently they become the targets of a cell-mediated
immune reaction perpetrated by lymphocytes and
Langerhans cells [26]. It has been shown that in
patients with aphthous stomatitis there is a heightened
lymphocytotoxic effect directed against oral epithelial
cells when compared to unaffected controls [25 – 27].
Evidence for this pathogenetic mechanism is also
inferred from observations that tissue biopsies of
newly erupted aphthous ulcerations demonstrate
agglomeration of activated T lymphocytes at the
periphery of the lesions, whereas in well-established
aphthae the initially predominant CD4+ helper/sup-
pressor cell population is subsumed and succeeded by
cytotoxic CD8+ lymphocytes [24,25].
Additional indirect support for primary immune
dysregulation is reflected in the long-recognized cor-
relation between stress and outbreaks of aphthous
ulcers that is reported by many RAS patients, in con-
trast to the notable decrease in frequency of episodes
during periods of reduced stress [28]. This observed
 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122
association is not entirely surprising, because stress is
known to affect immunologic function [2]. It also has
been observed that HIV-infected patients experience
oral aphthous-like ulcerations with relatively high
frequency. With advancing immune depletion, their
aphthous outbreaks are often dominated by larger
ulcers that run a more protracted course [13,29,30].
Attempts to explore the possibility that RAS is
fundamentally an antibody-driven disorder have dis-
closed findings that are at best inconsistent and
largely unsupportive. It seems that any previously
held conjecture that aphthous ulcers stem from a
centrally generated humoral immune mechanism
rather than from local cellular immune responses to
an antigenically modified oral mucous membrane
was predicated on assumptions that have since been
discredited [3,18,31 – 33].
Microbial factors
To date, investigations have yielded little, if any,
consistent evidence to support the hypothesis that
RAS represents an infectious disease. In particular,
from studies to determine whether there might be a
connection between previously suspect L-forms of
streptococci and RAS, or the adenoviruses, herpes
simplex virus (HSV), varicella-zoster virus, or cyto-
megalovirus and RAS, the available evidence sug-
gests that none of these microorganisms seems to be
directly culpable for RAS despite continued specu-
lation about their possible role [2,6]. One should note
that an antiviral agent, acyclovir, offers no beneficial
effect in preventing or attenuating episodic flares of
the condition [34], which serves to weaken arguments
in favor of a possible viral causation for RAS [16].
From occasional anecdotal cases in which patients
report an apparent consistent temporal relationship
between their aphthous outbreaks and an immediately
antecedent reactivated (recurrent) HSV infection, it is
tempting to postulate that in a narrow subset of
individuals who get RAS, the herpes virus may serve
as an antigenic ‘‘trigger’’ that initiates the cascade of
immunologic events that result in ulceration [2]. In a
limited subset of RAS patients, it is possible that this
is actually the case. Presumably, such patients would
benefit from appropriate therapeutic and prophylactic
antiviral therapy, coupled with treatments specifically
aimed at lessening the severity and frequency of the
RAS episodes by modulating their supposedly height-
ened immune responses to the viral ‘‘trigger’’ [6].
Such therapeutic strategies probably would be best
carried out in consultation with an infectious disease
specialist [6]. (Analogous speculation and therapeutic
113
consideration also could be applied to the other human
herpes viruses, varicella-zoster virus and cytomegalo-
virus, which are known to affect such subclinical
phenomena as asymptomatic viral shedding and ele-
vated viral titers, similar to HSV [35,36].) It must be
emphasized, however, that regarding most aphthous
patients, any suggestion of a causative nexus between
RAS and HSV seems to represent unsubstantiated
conjecture rather than proven fact [16].
Nutritional factors
Other issues explored in the quest to determine a
cause for RAS include the possible relationship of
attacks to excess or deficiency of various nutritional
factors, such as serum iron, folate, and vitamin B12,
and speculation that aphthous ulcers represent the
manifestation of an allergic reaction to certain foods
or other ingested or contacted substances. Apart from
variably favorable responses to the avoidance of
gluten products in aphthous patients with docu-
mented intestinal malabsorption disease (compared
to controls [37]) and in some aphthous stomatitis
patients with normal intestinal function [38], evi-
dence that RAS primarily represents an allergic
response or is etiologically linked to diminished
serum iron, vitamin B12, or folate levels is lacking
or, at best, equivocal [6,16,24,28,39].
For any RAS patient who exhibits physical signs
and symptoms that suggest the possibility of an
underlying malabsorption or nutritional deficiency
state or a blood dyscrasia, it is prudent to obtain a
complete blood count and assays for serum folate,
vitamin B12, and ferritin. Should any of these tests
yield findings that suggest an underlying systemic
abnormality, referral to an internist or a hematologist
is indicated [6].
Clinical features
The classic clinical features of RAS are generally
well known and usually suffice for diagnostic pur-
poses [6]. The lesions affect the oral cavity exclu-
sively and, with the occasional exception of the
dorsal aspect of the tongue, overwhelmingly favor
nonkeratinizing, freely moveable mucosal surfaces.
From the latter characteristic, one might infer that the
abridged mucosal barrier of these attenuated epithe-
lial surfaces allows for intensified antigenic stimu-
lation, which in turn launches the autoimmune
reaction that results in mucosal disruption. Involve-
ment of keratinized epithelial surfaces is distinctly
114 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122

uncommon and almost always represents extension of
ulceration from its origin on an adjacent nonkeratin-
ized epithelial surface [2]. When a cluster of ulcers is
confined to a single keratinized mucosal location (eg,
hard palate, attached gingiva), the possibility that
they represent ruptured vesicles of a recurrent HSV
infection (Table 1) or are traumatic ulcers must be
considered in the differential diagnosis as more likely
than aphthous ulcers [6,40,41].
Aphthous ulcers’ preference for buccal and labial
mucosa and the lateral-ventral tongue surfaces also
lends oblique support to the condition’s immune-
related nature, because these sites are especially
susceptible to friction and prone to trauma. It is not
uncommon for some patients to report that their
aphthous ulcers typically arise on such locations
shortly after a preceding minor traumatic event, such
as an accidental self-inflicted bite, frictional irritation
or injury from tooth clenching or bruxing, or after an
inadvertent nonpenetrating ‘‘stab’’ from a utensil
during a meal [42]. These events are examples of
the so-called ‘‘Koebner phenomenon,’’ in which
lesions that are characteristic of a specific immuno-
logically mediated mucocutaneous disorder (eg,
psoriatic plaques in psoriasis) develop on sites of
recently traumatized but previously lesion-free skin
or mucosa [43]. ‘‘Koebnerization’’ also occurs in
individuals with diseases such as lichen planus [44],
various types of lupus erythematosus, and others
[45]. It seems that a subset of individuals with RAS
tends to experience a similar phenomenon with
relation to the location of at least some of their
aphthous lesions in the context of some, if not all,
of their outbreaks.
Unlike the ulcerations seen in primary or recurrent
oral HSV infections, aphthous ulcers are not preceded
by vesicles [6]. Although individual aphthae are often
categorized into one of three possible subtypes (minor,
major, and herpetiform) based primarily on their
respective size differences, this classification is purely
descriptive rather than substantive. Regardless of an
individual lesion’s dimensions, all of the subtypes
seem to represent mere superficial variations of a
single disease entity [2]. Any individual recurrence
may feature as few as one or two isolated ulcers or
upwards of virtually dozens of aphthous lesions. Any
one of the subtypes or various combinations of each of
the three subtypes can occur in the course of a single
RAS outbreak [16]. Notwithstanding size, all aphthous
ulcers are painful, and in many cases, patients also
relate histories of prodromal discomfort or other symp-
toms that routinely augur impending recurrences [40].
The frequency and severity of recurrences vary
widely among affected individuals. Some patients
experience recurrences at fairly regular intervals,
whereas for others the episodes occur with less
predictability. In many cases the attacks are precipi-
tated by a particular identifiable factor (eg, intense
stress [15] or, as reported by some women, in tem-
poral relation to their menstrual cycles [46]), whereas

Table 1
Distinctions: Recurrent aphthous ulcers and recurrent (‘‘reactivated’’) intraoral HSV lesions
Factor Recurrent aphthous ulcers Recurrent HSV
Etiology Focal immune dysregulation HSV-I (HSV-II rarely)
Antecedents Commonly stress, trauma; other factors diverse Local trauma, stress, systemic
and unique as the affected individuals immune alteration
Prodromal awareness Sporadic Typical
Location of lesions Non-keratinizing ‘‘moveable’’ mucosa and Keratinizing (‘‘bound down’’)
tongue dorsum mucosa only
Distribution Unifocal or multifocal throughout oral cavity Unifocal only
(in immunocompetent individuals)
Lesions Ulcers of varying size and number possible Unifocal cluster of vesicles
per episode that ulcerate
Symptoms Pain (regardless of lesion size) Pain
Diagnostic Confirmation History, clinical features; negative cytology for History, clinical features; positive viral
virally-modified epithelial cells; biopsy usually cytopathic changes on smear; biopsy
not necessary usually not necessary
Clinical course 7 to 14 days, no scarring (minor and herpetiform types); 7 to 14 days
greater than 14 days, scarring possible (major types) (in immunocompetent individuals)
Management Palliative agents, corticosteroids, other immune Antiviral agents
modulating agents
Modified from Regezi JA, Sciubba JJ. Vesiculo-bullous diseases: Herpes simplex infections. In: Regezi JA, Sciubba JJ, editors.
Oral pathology: Clinical-pathologic correlations. 2nd edition. Philadelphia: W.B. Saunders; 1993; p. 9; with permission.
 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122
Fig. 1. Minor aphthous ulcers, lower labial mucosa.
some patients’ outbreaks arise without any identifi-
able triggering factor [16]. Episodes may occur rarely
or with some frequency, accompanied by a range of
symptomatic intensity. Most troubling for a small
subset of individuals is the experience of overlapping
recurrences for periods of time that can involve weeks
or months or, rarely, a year or more of almost
perpetual mucosal disruption from aphthous ulcer-
ations. During such periods of seemingly relentless
outbreaks, a patient’s ability to function comfortably
and even their emotional stability may be compro-
mised drastically because of protracted oral mucosal
pain [2]. Under such circumstances, patients are
occasionally suspected of having Behcß et’s disease
[8] or some other major systemic disorder. Accord-
ingly, they are often referred to appropriate medical or
dental specialists for a diagnostic consultation and
evaluation. One should note that in most cases even
the most exacting and thorough systemic evaluation
fails to reveal evidence that supports a diagnosis either
of Behcßet’s or any other identifiable systemic disease.
Instead, such evaluations only serve to confirm that
the condition is confined to the oral cavity and is
simply an extreme expression of RAS alone [2,6].
Commonly, RAS patients report sensations of
burning, tingling, or other mucosal discomfort just
before an outbreak of aphthous ulcers, similar to the
prodromal symptoms that frequently precede recur-
rent HSV infections [40]. During the prodromal phase
of RAS it may be possible to observe a transient
erythematous macule or papule at the site of mucosal
discomfort. Such lesions typically emerge onto move-
able oral mucosal surfaces or the dorsal aspect of the
tongue and can antecede the actual aphthous ulcer-
ations by at least several hours or longer. These
preulcerative lesions are succeeded by characteris-
tically painful ulcers with round, symmetrical, yel-
low-white or gray fibrinonecrotic centers surrounded
115
by slightly raised, erythematous, halo-like borders.
Several ulcers in close apposition to one another may
become confluent and result in a larger area of
mucosal disruption [2].
So-called ‘‘minor’’ (‘‘typical’’) aphthous ulcers are
smaller than 1 cm in diameter and tend to heal within
10 to 14 days (Fig. 1). Their larger counterparts, so-
called ‘‘major’’ aphthous ulcers (Mickulicz’s aphthae,
periadenitis mucosa necrotica recurrens, Sutton’s aph-
thae), are not as common. They are 1 cm in diameter or
larger, with edematous borders, and classically involve
a greater depth of submucosal tissue destruction. For
this reason, these types of aphthous ulcers tend to run a
more extended clinical course; they can take signifi-
cantly longer than 2 weeks to resolve and sometimes
heal with scar formation (Figs. 2,3). Herpetiform
aphthous ulcers, the least common type, tend to be
distributed in tight clusters of small ulcers, 1 to 3 mm
in diameter, that bear superficial resemblance to the
ruptured vesicles of oral HSV infections [2,4,6,36].
Frequently, close crops of these ulcers coalesce to
involve a broad surface area (Figs. 4,5). Although
their small size allows for resolution of herpetiform
aphthae within a 7- to 10-day period, some patients
experience recurrences that flare at relatively close
intervals. Although herpetiform aphthae also exhibit a
strong predilection for nonkeratinized epithelium,
unlike their minor and major counterparts they
coalesce readily and can occur on any oral mucosal
surfaces, including those that are keratinized [2].
Histopathology
Once an aphthous ulcer has emerged onto the oral
mucosa, biopsy is usually not required for diagnostic
Fig. 2. Major aphthous ulcer of longer than 8 weeks’
duration. Located on anterior buccal-labial mucosa, the
lesion had been secondarily traumatized.
116 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122
Fig. 3. Hyperplastic scar tissue from healed major
aphthous ulcer, lower right vestibular mucosa adjacent to
mandibular canine and first premolar tooth. Scar resembles
epulis fissuratum.
confirmation because there are no defined pathogno-
monic microscopic features of this disorder. At the
tissue level, the clearly nonvesicular nature of the
ulcerative lesion and the absence of virally modified
epithelial cells may serve to distinguish an early
aphthous ulcer from an herpetic lesion in cases in
which there is clinical diagnostic ambiguity [4,40].
Relatively early in their course, aphthous lesions
demonstrate an ulcer base covered by a fibrinopuru-
lent pseudomembrane. The marginal epithelium may
appear spongiotic with attendant lymphocytosis,
especially in the basal-most strata. At the marginal
epithelial-stromal interface, ‘‘tagging’’ of lympho-
cytes is prominent, and throughout the superficial
stroma polymorphonuclear leukocytes and chronic
inflammatory infiltrates dominated by lymphocytes
and histiocytes prevail. Within the deeper submucosa,
perivascular infiltrates of lymphocytes and histiocytes
Fig. 4. Minor, major, and herpetiform aphthae occurring
synchronously on right soft palatal mucosa.
Fig. 5. Herpetiform aphthous ulcers on the soft palate and
uvula. These clusters of small ulcerations have coalesced to
form extensive lesions that resemble major aphthae.
may be seen [2]. As an aphthous ulcer ‘‘ages,’’ the
histopathologic findings become increasingly non-
specific [16]. Biopsy of any long-standing, nonheal-
ing oral ulcer may be indicated, however, to determine
whether the lesion is benign or malignant or repres-
entative of some unsuspected unusual diagnostic
entity, such as a granulomatous infection or inflam-
matory disease.
Management of recurrent aphthous stomatitis
General considerations
Conventional wisdom and clinical experience
strongly suggest that most cases of RAS are isolated
to the oral cavity and are neither attributable to nor
associated with an underlying systemic disease or
some other generalized pathologic condition. If the
medical history is positive for or suggestive of any of
the systemic conditions known to be linked to oral
aphthae, however, and the patient experiences severe
bouts or unusually frequent recurrences of aphthous
stomatitis, it may be prudent to rule out the possibility
that the oral flare up reflects an exacerbation of the
systemic disease. Should that suspicion be confirmed
through an appropriately focused evaluation in con-
sultation with the patient’s primary care provider or
an appropriate medical specialist, the apparent incit-
ing condition must be addressed and treated accord-
ingly. The expectation is that once the underlying
systemic problem is brought under control, at least
temporary remission or regression of the RAS can be
achieved [2,6,16].
Whether it is of value to include complete hem-
atologic screening of RAS patients as part of the
 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122
standard diagnostic and treatment planning evalu-
ation is debatable. Although low serum iron and
ferritin levels have been documented in some
patients by some investigators [47,48], others have
found that RAS patients generally have normal
erythrocyte counts and hemoglobin levels, similar
to those of controls [49,50]. Some researchers recom-
mend that children with RAS be screened hemato-
logically because in approximately 20% of children
who get aphthous ulcers, dormant serum iron defi-
ciencies unaccompanied by anemia have been dis-
closed [51]. Regardless of the patient’s age, if
hematologic testing reveals a deficiency state or
other abnormality, management of the blood dyscra-
sia also may constitute a key strategy for managing
the oral lesions.
Objectives, principles, and caveats
Treatment is targeted at reducing the duration
and expediting the resolution of ulcers, alleviating
pain, lessening the frequency of attacks, and fore-
stalling recurrences. Management decisions should
be dictated by each patient’s perceptions of the
severity of the oral lesions, the frequency of recur-
rences, and the degree of debilitation that attends the
outbreaks. All treatment efforts must be applied in
balance with the safest, most judicious use of
therapeutic agents available.
In patients with strictly oral manifestations of
RAS and no concomitant systemic abnormality, a
broad scope of possible treatment options is avail-
able. At the extremes, choices range from doing
essentially nothing other than to confirm the diag-
nosis to prescribing major therapeutic agents aimed at
prevention and control of the immunologic mecha-
nisms that produce aphthous ulcers. In between these
extremes several relatively conservative strategies are
available for alleviating symptoms exclusively, cur-
tailing the duration and pain of a single episode,
attenuating an active outbreak that may be part of a
series of overlapping or frequent outbreaks of RAS,
and preventing future episodes (see Box 2). It is
vitally important for the patient and the clinician to
bear in mind that any and all interventions for RAS,
regardless of severity, are neither curative nor intend-
ed for persistent use [16]. Rather, they are prescribed
to effect temporary respite from a disease with a
proclivity for recurrences [1 – 4,6,16,28,40].
Because aphthous stomatitis is an immunologi-
cally mediated disease, therapies centered in
immunomodulation are most appropriate. They can
be applied successfully for managing individual epis-
117
Box 2. Recurrent aphthous stomatitis:
treatment options
No treatment
Palliative approach
 Topical agents: home remedies, over-
the-counter medications, prescribed
analgesics, cauterizing agents
Antiinflammatory and antimicrobial
agents
Immunomodulation*
 Topical: corticosteroid creams, oint-
ments, gels, rinses; ‘‘intralesional’’
(perilesional) corticosteroid injections
 Systemic: corticosteroids, nonsteroi-
dal immunosuppressive agents, anti-
inflammatory agents
Combined therapy: topical and sys-
temic agents, systemic agents in com-
bination
* Most efficacious approaches.
odic flares and can be used on an intermittent basis as
a preemptive strategy for attenuating an incipient
attack or they can be used to abort an imminent
aphthous outbreak in its prodromal stage. The unre-
mitting use of immunomodulating agents for this
essentially benign disease, whether topical, systemic,
or combined, is contraindicated, given these agents’
recognized potential for pernicious side effects
[2,3,6].
First-line therapeutic choices
When a recurrent episode consists of a relatively
limited number of aphthous lesions that are either
small or large, closely apposed to one another, and
distributed on readily accessible oral surfaces such
as the labial or vestibular mucosa or the anterior
portion of the tongue, first-line therapeutic manage-
ment should involve regimens based on conserva-
tive topical therapy [2 – 4,6,16]. Among the options
available are any of the over-the-counter nonsteroi-
dal occlusive preparations, such as Orabase, with or
without topical analgesic (usually benzocaine) that
may be applied primarily for symptomatic relief or
nonsteroidal antiinflammatory preparations, such
as amlexanox 5% paste (Aphthasol oral paste).
Applied directly to active lesions, amlexanox pro-
motes pain reduction by inhibiting release of his-
118 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122
tamine and leukotrienes. It has been shown that
aphthous ulcers treated with amlexanox resolve in
less time (by 1 day) than their untreated counter-
parts [2 – 4,6].
Topical corticosteroid gels, creams, or ointments
that are selectively prescribed and judiciously applied
to active lesions constitute an effective conservative
modality for managing RAS. The cream or ointment
preparations are preferred, because some patients
report stinging or burning with the use of cortico-
steroid gels [3,40].
Some patients who wish to use topical cortico-
steroid gels, creams, or ointments are reluctant to do
so or fail to comply because of the fear that attends
reading the warning on prescription packaging inserts
or receiving pharmacists’ admonishments that topical
corticosteroid preparations are for external use only.
Clinicians should reassure their patients that the use
of these topical agents strictly as prescribed for
limited periods of time is not likely to produce
untoward effects and is safe [3,40,52].
Over the course of several days, starting as early
as possible from the onset of the outbreak, direct
application of a thin film of the corticosteroid agent
three to five times daily and at bedtime after gentle
drying of the affected area alleviates discomfort and
reduces the duration of ulcers so that they heal within
several days, rather than linger for a week or more, as
is typical of untreated lesions [6]. Among the recog-
nized vagaries inherent in using creams or ointments
on oral mucous membranes is the likelihood that they
rapidly wash away from the site of application and
thereby diminish the therapeutic efficacy of the
medication. This problem can be addressed either
by prescribing a topical corticosteroid compounded
with tissue adhesive (eg, 0.1% or 0.5% triamcinolone
acetonide [Aristocort A, Kenalog] in Orabase) or,
as an alternative, by prescribing more potent topical
corticosteroids, such as 0.05% betamethasone dipro-
pionate cream or ointment (Diprolene; Lotrisone),
0.05% fluocinonide cream or ointment (Lidex),
or 0.05% clobetasol propionate ointment or cream
(Temovate) alone, without tissue adhesive [2 – 4,6].
The latter three topical agents seem to provide greater
therapeutic efficacy than those compounded in an
occlusive or adherent base, despite their limited time
in contact with active lesions [3]. There seems to be
no appreciable therapeutic advantage to applying
corticosteroid in a base preparation (ie, compounded
in Orabase or some other inert mucous membrane
dressing) as compared to applying a base preparation
alone (ie, without corticosteroid) [53]. Regardless of
whether a base preparation or corticosteroid alone or
a compound of both is applied, avoidance of food and
drink for 30 minutes after application is advised to
promote adherence of the preparation to the ulcer-
ations. To prevent or address corticosteroid-induced
candidiasis, an antifungal medication can be com-
pounded into the corticosteroid cream or ointment,
with or without a tissue adhesive base when indi-
cated [40].
For aphthous ulcers that are difficult to reach with
a finger—predominantly ulcers that involve the most
posterior oral regions—or are so numerous and widely
distributed as to render the direct application of
creams or ointments impractical, corticosteroid rinses
are an excellent topical therapeutic alternative [2 – 4,
6,16,28,40]. Rinsing over a period of several days
with betamethasone (Celestone) syrup 0.6 mg/5 mL,
dexamethasone (Decadron) elixir 0.5 mg/5 mL, or a
specially compounded aqueous suspension of 0.1% or
0.2% triamcinolone acetonide accelerates resolution
of the ulcers and alleviates discomfort. Three to four
times daily and before retiring for the night, 1 tea-
spoonful of the liquid is held and swished in the
mouth for 2 to 3 minutes and then expectorated. The
rinse regimen is followed by avoidance of food and
drink for 30 to 60 minutes afterward. This is not only
an effective method for addressing active RAS out-
breaks but also may be useful for aborting an immi-
nent attack in its prodromal phase [16]. The author
also has found that ‘‘customized’’ prophylactic corti-
costeroid rinse regimens can be adapted to an indi-
vidual patient’s needs, particularly in cases in which it
is necessary to intercept and prevent episodes known
to occur with some regularity or frequency. Triamcin-
olone acetonide suspension can be compounded with
nystatin (Nystatin oral suspension USP) for individu-
als predisposed to oral candidiasis or with 2% lido-
caine (Xylocaine 2% viscous solution) if there is a
need for expedient pain relief (Karen A. Baker, MS
Pharm, personal communication, 2001).
Another method for applying topical corticoste-
roid medication to active lesions (particularly major
aphthous ulcers or other types of aphthae located
primarily in intertriginous-like regions of the oral
mucosa, such as the upper or lower buccal or labial
vestibules or the sublingual vestibule) involves
using a gauze sponge either soaked in corticosteroid
rinse preparation or on which a small amount of
corticosteroid cream or ointment has been applied
for delivery. By laying the gauze sponge directly
onto the lesion or lesions and holding it in place for
15 to 20 minutes two or three times daily during
waking hours, the concentrated contact method for
delivery of the medication requires fewer applica-
tions and promotes more expeditious healing of
ulcers [3,4].
 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122
Perilesional injection
Perilesional injection of corticosteroid medication
is another therapeutic alternative. Such injections can
be highly efficacious for treating major aphthous
ulcers that have been resistant to other more conserv-
ative topical approaches or are unresponsive to sys-
temic treatment [2]. The author has found that this
route of corticosteroid administration can reduce
significantly the diameter of either a newly erupted
or a long-standing major aphthous lesion within the
24- to 48-hour postinjection period and dramatically
shorten healing time.
The procedure the author uses is based on a
modification of the adjuvant intralesional cortico-
steroid injection regimen described for use in treat-
ing recalcitrant pemphigus lesions [54]. After
administration of local anesthesia (with vasocon-
strictor), 10 to 40 mg of sterile triamcinolone
acetonide injectable suspension, USP (Kenalog-40)
40 mg/mL diluted to 10 mg/mL strength, is injected
into the perilesional tissue immediately adjacent to
the ulcer border. The lesion is clinically evaluated
48 hours after injection to determine whether the
treatment was adequate. If there seems to be neither
improvement in symptoms nor evidence of progres-
sion toward healing, it may be necessary to repeat
the procedure. Usually a single administration of
corticosteroid is sufficiently therapeutic, however,
so that even large, painful, or stubbornly recalcitrant
major aphthous lesions that were present for many
weeks before the injection resolve within 5 to
7 days.
Indications for second-line therapy
Prednisolone (Prelone) or betamethasone (Celes-
tone) syrup preparations can be used exclusively as
rinses or as several times daily rinse-and-swallow
regimens for cases that are recalcitrant to the topical
approach alone. The latter combined route of admin-
istration can be helpful for treating major aphthae and
is especially effective in cases in which the aphthous
ulcers are concentrated in the posterior oral cavity, soft
palate, and tonsillar fauces and are attended by con-
siderable pain. Used in this manner, the corticosteroid
agent provides topical and systemic immunomodula-
tory benefits and is considered to be a second-line type
of therapy [2].
For the occasional patient plagued by frequent
or overlapping RAS attacks over a protracted
period of time and whose quality of life has been
eroded as a result, the goal of treatment is two-
119
fold. The first goal is to remedy the current attack.
The second goal is to attempt to break the apparent
cycle of recurrence to achieve at least a temporary
period of remission or an interim in which recur-
rences are less frequent, ulcerations are fewer, and
symptoms are less intense. Achieving this goal may
require an approach that briefly couples systemic
corticosteroid ‘‘burst therapy’’ with topical cortico-
steroid therapy initially, followed by a ‘‘prophy-
lactic maintenance’’ regimen with continuance of
the topical corticosteroid alone. The author has
found that individual case-centered modifications
of the therapeutic approach reported by Vincent
and Lilly [16] can be effective in most cases for
accomplishing these goals. They recommend using
a ‘‘burst’’ therapeutic regimen of prednisone (40
mg taken 1 hour after rising in the morning for
5 days, followed by 20 mg every other day for an
additional week) along with oral rinses with com-
pounded 0.1% triamcinolone acetonide aqueous
suspension (5 mL of liquid swished four times
daily for 2 to 3 minutes and expectorated ; NPO
for 1 hour afterward). If the triamcinolone suspen-
sion is not available, dexamethasone elixir used
similarly three to four times daily at outset also
can be effective.
In the author’s experience, ‘‘burst therapy’’ also
can be used safely over a 3-week period if necessary
(eg, 40 mg prednisone every morning for 7 days
followed by 20 mg every morning for 7 additional
days followed by 20 mg every other morning for
another 7 days) and does not require any further
tapered doses [16]. Coupled with the initial four-
times-daily corticosteroid rinse regimen, burst ther-
apy is more effective than 5 to 7 days of abbreviated
systemic corticosteroid (dosepak) therapy alone for
obtaining control of an active outbreak and intercept-
ing and preventing subsequent recurrences [16].
Once the systemic corticosteroid regimen is com-
pleted, a prophylactic rinse regimen (eg, either one to
three times daily or one to three times on alternate
days as needed; NPO for 30 minutes to 1 hour after
rinsing) can provide long-term preventive mainte-
nance. Although systemic complications from using
attenuated corticosteroid rinse regimens over ex-
tended periods of time do not seem to be a significant
problem [16], oral candidiasis can be a potential
problem for some individuals on this therapy, par-
ticularly individuals whose cell-mediated immune
responses are compromised by an underlying medical
condition, such as diabetes. In such cases, inter-
vention with ketoconazole (Fluconazole, Nizoral)
tablets or clotrimazole (Mycelex) troches may be
indicated [2,3].
120 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122
Other medications
Various other medications and methods too numer-
ous to mention in this article and of highly variable
efficacy have been used for managing aphthous sto-
matitis. These medications include inert and more
active prescribed preparations (antimicrobial agents,
cauterizing solutions, monoamine oxidase inhib-
itors, tissue films, and others) and myriad home
remedies [2,4]. When cases are resistant to cortico-
steroid-based immunomodulatory regimens exclu-
sively, other nonsteroidal immunosuppressive agents,
such as cyclosporine (Sandimmune, Neoral), dapsone,
or azathioprine (Imuran) given systemically in com-
bination with adrenal-sparing doses of prednisone and
other agents with antiinflammatory, immunomodula-
tory properties alone (eg, pentoxyphylline [Trental] or
colchicine) may or may not prove useful as alternative
second- or third-line therapies [2 – 4,6,16,28,40]. Their
potential for producing serious side effects does limit
case selection for their use.
Recently, thalidomide (Thalomid), a potent immu-
nosuppressive medication, was found to be effective
for treating HIV-infected patients with severe RAS
when prescribed at daily doses of 100 to 200 mg [55]
and for managing oral aphthae in immunocompetent
individuals with Crohn’s disease [56] and Behcßet’s
disease [8,57]. Because of thalidomide’s recognized
potential for profound side effects, however, its use
must be reserved only for patients whose aphthous
ulcers have been so unrelenting and sufficiently
symptomatic as to significantly compromise quality
of life and who, after exhaustive efforts, also have
failed to respond to other, more conservative thera-
peutic measures. Use of this systemic agent requires
strict adherence to guidelines for patient selection.
Patients who take thalidomide must be monitored
frequently to intercept and prevent its potential toxic
effects [2,6].
Summary
Currently, RAS is recognized as an immunologi-
cally mediated, inflammatory oral condition rather
than an infectious disease. Contemporary approaches
to its management are focused on modulating the
aberrant immune responses that underlie its pathogen-
esis. Immunomodulation has been applied with great
success for managing existing episodes and prevent-
ing recurrences of RAS. Safe, effective, and reliable
therapies for actually curing this disease remain elu-
sive at this time. It is reasonable to suggest that for the
future, anticipated new insights into the condition’s
origin and pathogenesis, combined with ongoing
research directed toward the development of safer,
more effective immunomodulating agents, render the
prospect of a cure for RAS increasingly plausible.
Acknowledgment
The author wishes to express sincere gratitude to
Marilyn R. Holt, MS, for her technical expertise and
invaluable assistance in preparing the manuscript.
References
[1] Ship JA. Recurrent aphthous stomatitis: an update. Oral
Surg Oral Med Oral Pathol Oral Radiol Endod 1996;81:
141 – 7.
[2] Neville BW, Damm DD, Allen CM, Bouquot JE. Al-
lergies and immunologic diseases: recurrent aphthous
stomatitis. In: Oral and maxillofacial pathology.
2 nd edition. Philadelphia: W.B. Saunders; 2002.
p. 285 – 90.
[3] Eisen D, Lynch DP. Selecting topical and systemic
agents for recurrent aphthous stomatitis. Cutis 2001;
68:201 – 6.
[4] Reich RF, Kerpel SM, Freedman PD. Differential diag-
nosis and treatment of ulcerative, erosive, and vesicu-
lobullous lesions of the oral mucosa. Oral Maxillofac
Surg Clin N Am 1998;10:95 – 129.
[5] Wright DG, Dale DC, Fauci AC. Human cyclic neu-
tropenia: clinical review and long-term follow-up of
patients. Medicine (Baltimore) 1981;60:1 – 13.
[6] Ship JA, Chavez EM, Doerr PA, et al. Recurrent aph-
thous stomatitis. Quintessence International 2000;31:
95 – 112.
[7] Glickman RM. Inflammatory bowel disease: ulcerative
colitis and Crohn’s disease. In: Fauci AS, Braunwald
E, Isselbacher KJ, et al, editors. Harrison’s principles
of internal medicine. 14th edition. New York: McGraw-
Hill; 1998. p. 1633 – 45.
[8] Ghate JV, Jorizzo JL. Behcßet’s disease and complex
aphthosis. J Am Acad Dermatol 1999;40:1 – 18.
[9] International Study Group for Behcßet’s Disease. Criteria
for diagnosis of Behcß et’s disease. Lancet 1990;335:
1078 – 80.
[10] Magro C, Crowson AN, Mihm M. Cutaneous manifes-
tations of nutritional deficiency states and gastrointes-
tinal disease: Behcßet’s disease, inflammatory bowel
diseases (Crohn’s disease, ulcerative colitis, celiac dis-
ease). In: Elder D, Elenitsas R, Jaworsky C, Johnson B,
editors. Lever’s histopathology of the skin. 8th edition.
Philadelphia: Lippincott-Raven; 1997. p. 359 – 63.
[11] Rizzi R, Bruno S, Dammacco R. Behcß et’s disease: an
immune-mediated vasculitis involving vessels of all
sizes [review]. Int J Clin Lab Res 1997;27:225 – 32.
[12] Orme RL, Nordlund JJ, Barich L, et al. The MAGIC
 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122
syndrome (mouth and genital ulcers with inflamed car-
tilage). Arch Dermatol 1990;126:940 – 4.
[13] Ficarra G. Oral ulcers in HIV-infected patients: an up-
date on epidemiology and diagnosis [review]. Oral
Diseases 1997;3(Suppl 1):S183 – 9.
[14] Marshall GS, Edwards KM, Butler J, et al. Syndrome
of periodic fever, pharyngitis and aphthous stomatitis.
J Pediatr 1987;110:43 – 6.
[15] Rogers III RS. Recurrent aphthous stomatitis: clinical
characteristics and associated systemic disorders.
Semin Cutan Med Surg 1997;16:278 – 83.
[16] Vincent SD, Lilly GE. Clinical, historic, and therapeu-
tic features of aphthous stomatitis: literature review and
open clinical trial employing steroids. Oral Surg Oral
Med Oral Pathol 1992;74:79 – 86.
[17] Ship JA. Inheritance of aphthous ulcers of the mouth.
J Dent Res 1965;44:837 – 44.
[18] Miller MF, Garfunkel AA, Ram CA, et al. Inheritance
patterns in recurrent aphthous ulcers: twin and pedi-
gree data. Oral Surg Oral Med Oral Pathol 1977;43:
886 – 91.
[19] Miller MF, Garfunkel AA, Ram CA, et al. The inher-
itance of recurrent aphthous stomatitis. Oral Surg Oral
Med Oral Pathol 1980;49:409 – 12.
[20] Landesberg R, Fallon M, Insel R. Alterations of
T-helper/inducer and T-suppressor/inducer cells in pa-
tients with aphthous ulcers. Oral Surg Oral Med Oral
Pathol 1990;69:205 – 8.
[21] Savage NW, Mananonda R, Seymour GJ, et al. The
proportion of suppressor-inducer T-lymphocytes is re-
duced in recurrent aphthous stomatitis. J Oral Pathol
Med 1988;17:293 – 7.
[22] Schroeder HE, Muller-Glauser W, Sallay K. Stereolog-
ical analysis of leukocyte infiltration in oral ulcers of
developing Mickulicz aphthae. Oral Surg Oral Med
Oral Pathol 1983;56:629 – 40.
[23] Greenspan JS, Gadol N, Olson JA, et al. Antibody-
dependent cellular cytotoxicity in recurrent aphthous
ulceration. Clin Exp Immunol 1981;44:603 – 10.
[24] Pedersen A, Hougen P, Kenrad B. T-lymphocyte sub-
sets in oral mucosa of patients with recurrent aphthous
ulceration. J Oral Pathol Med 1992;21:176 – 80.
[25] Savage NW, Seymour GJ, Kruger BJ. T-lymphocyte
subset changes in recurrent aphthous stomatitis. Oral
Surg Oral Med Oral Pathol 1985;60:175 – 80.
[26] Savage NW, Seymour GJ, Kruger BJ. Expression of
class I and class II major histocompatibility complex
antigens on epithelial cells in recurrent aphthous sto-
matitis. J Oral Pathol Med 1986;15:191 – 5.
[27] Rogers RS, Sams WM, Shorter RG. Lymphocytotox-
icity in recurrent aphthous stomatitis. Arch Dermatol
1974;109:361 – 3.
[28] Woo S-B, Sonis ST. Recurrent aphthous ulcers: a re-
view of diagnosis and treatment. J Am Dent Assoc
1996;127:1202 – 13.
[29] MacPhail LA, Greenspan JS. Oral ulceration in HIV-
infection: investigation and pathogenesis. Oral Dis-
eases 1997;3(Suppl 1):S190 – 3.
[30] Phelan JA, Eisig S, Freedman PD, et al. Major aph-
121
thous-like ulcers in patients with AIDS. Oral Surg Oral
Med Oral Pathol 1991;71:68 – 72.
[31] Cohen L. Etiology, pathogenesis and classification of
aphthous stomatitis and Behcßet’s syndrome. J Oral Path-
ol 1978;7:347 – 52.
[32] Lehner T. Immunologic aspects of recurrent oral ul-
cers. Oral Surg Oral Med Oral Pathol 1972;33:80 – 5.
[33] Ben Aryeh H, Malberger E, Gutman D, et al. Salivary
IgA and serum IgG and IgA in recurrent aphthous
stomatitis. Oral Surg Oral Med Oral Pathol 1976;42:
746 – 52.
[34] Wormser GP, Mack L, Lenox T, et al. Lack of effect of
oral acyclovir on prevention of aphthous stomatitis.
Otolaryngol Head Neck Surg 1988;98:14 – 7.
[35] Kameyama T, Sujaku C, Yamamoto S, et al. Shedding
of herpes simplex virus type I into saliva. J Oral Pathol
1988;17:478 – 81.
[36] Neville BW, Damm DD, Allen CM, Bouquot JE. Viral
infections: HSV, VZV, CMV. In: Oral and maxillofa-
cial pathology. 2nd edition. Philadelphia: W.B. Saun-
ders; 2002. p. 213 – 24.
[37] Veloso FT, Saleiro JV. Small-bowel changes in recur-
rent ulceration of the mouth. Hepatogastroenterology
1987;34:36 – 7.
[38] Wray D. Gluten-sensitive recurrent aphthous stomati-
tis. Dig Dis Sci 1981;26:737 – 40.
[39] Olson JA, Feinberg I, Silverman Jr S, et al. Serum
vitamin B12, folate and iron levels in recurrent aph-
thous stomatitis. Oral Surg Oral Med Oral Pathol
1982;54:517 – 20.
[40] Regezi JA, Sciubba JJ. Ulcerative conditions: condi-
tions associated with immunologic dysfunction. Aph-
thous ulcers. In: Regezi JA, Sciubba JJ, editors. Oral
pathology: clinical-pathological correlations. 2nd edi-
tion. Philadelphia: W.B. Saunders; 1993. p. 52 – 60.
[41] Rodu B, Mattingly G. Differential diagnosis of oral
mucosal ulcerations. J Am Dent Assoc 1992;123:
83 – 6.
[42] Wray D, Graykowski EA, Notkins AL. Role of mu-
cosal injury in initiating recurrent aphthous stomatitis.
BMJ 1981;283:1569 – 70.
[43] Murphy GF, Mihm Jr MC. The skin: chronic inflamma-
tory dermatoses. In: Cotran RS, Kumar V, Robbins SL,
editors. Robbins pathologic basis of disease. 5th edition.
Philadelphia: W.B. Saunders; 1994. p. 1198.
[44] Boyd AS, Neldner KH. The isomorphic response to
Koebner. Int J Dermatol 1990;29:401 – 10.
[45] Domonokos AN. Cutaneous symptoms, signs and di-
agnosis: Koebner’s phenomenon. In: Domonokos AN,
editor. Andrews’ diseases of the skin: clinical derma-
tology. 6th edition. Philadelphia: W.B. Saunders; 1971.
p. 26.
[46] Ferguson MM, Carter J, Boyle P. An epidemiological
study of factors associated with recurrent aphthae in
women. J Oral Med 1984;39:212 – 7.
[47] Porter SR, Scully C, Flint S. Hematologic status in
recurrent aphthous stomatitis as compared with other
oral disease. J Oral Pathol 1978;7:418 – 23.
[48] Rogers III RS . Screening for hematinic deficiencies in
122 E. Eisenberg / Oral Maxillofacial Surg Clin N Am 15 (2003) 111–122
patients with recurrent aphthous stomatitis. Australas J
Dermatol 1986;27:98 – 103.
[49] Challacombe SJ, Barkhan P, Lehner T. Hematological
features and differentiation of recurrent oral ulceration.
Br J Oral Surg 1977;15:37 – 48.
[50] Porter SR, Kingsmill V, Scully C. Audit of diagnosis
and investigations in patients with recurrent aphthous
stomatitis. Oral Surg Oral Med Oral Pathol 1993;76:
446 – 52.
[51] Field EA, Rotter E, Speechley JA, et al. Clinical and
hematological assessment of children with recurrent
aphthous ulceration. Br Dent J 1987;163:19 – 22.
[52] MacPhail L. Topical and systemic therapy for recurrent
aphthous stomatitis. Semin Cutan Med Surg 1997;16:
301 – 7.
[53] Voute AB, Schulten EA, Langendijk PN, et al. Fluocin-
onide in an adhesive base for treatment of oral lichen
planus: a double-blind, placebo-controlled clinical
study. Oral Surg Oral Med Oral Pathol 1993;75:
181 – 5.
[54] Bystryn J-C, Steinman NM. The adjuvant therapy
of pemphigus: an update. Arch Dermatol 1996;132:
203 – 12.
[55] Jacobson JM, Greenspan JS, Spritzler J, et al, for the
National Institute of Allergy and Infectious Diseases
AIDS Clinical Trials Group. Thalidomide for the treat-
ment of aphthous ulcers in patients with human immu-
nodeficient virus infection. N Engl J Med 1997;336:
1487 – 93.
[56] Weinstein TA, Sciubba JJ, Levine J. Thalidomide for
the treatment of oral aphthous ulcers in Crohn’s dis-
ease. J Pediatr Gastroenterol Nutr 1999;28:214 – 6.
[57] Eisenbud L, Horowitz I, Kay B. Recurrent aphthous
stomatitis of the Behcßet’s type: successful treatment
with thalidomide. Oral Surg Oral Med Oral Pathol
1987;64:289 – 92.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 123 – 128
The surgical treatment of periodontal infections
Norman Trieger, DMD, MD*
Department of Dentistry, Oral and Maxillofacial Surgery, Albert Einstein College of Medicine, Montefiore Medical Center,
111 E. 210th Street, Bronx, NY 10467, USA
Identification of the specific pathogenic micro-
organisms responsible for periodontal destruction has
led to newer treatment methods. The basis for this
approach has been the application of well-established
surgical principles of infection management: the
identification of the responsible pathogens; mechan-
ical débridement of the infected sites to decrease the
bacterial load as well as products of the inflammatory
immune response; and the introduction of systemic
antibiotics known to be effective against the offend-
ing bacteria.
Experience has shown that eradication of the
infection is possible and regeneration of attachment
and bone regrowth in vertical defects is accomplished
when several important guidelines are followed. The
oral cavity is regarded as one ecosystem, with micro-
organisms readily transported to multiple sites by
saliva, food, and the toothbrush. Treatment by debrid-
ing only one quadrant at a time allows reinfection of
the site within the ecosystem from other untreated
quadrants. The usual pattern of multiple oral hygiene
visits before definitive full mouth debridement meant
to ‘‘prepare’’ the mouth for surgery does not elim-
inate the infection that is harbored in biofilms in deep
pockets. Eliminating the deep infection should take
precedence over the repeated superficial scalings.
Brushing, flossing and various mouth rinses do not
reach pockets deeper than 4 to 5 mm despite the
diligence of enhanced home care. The periodontal
attachment continues to secrete extracrevicular fluid
that inactivates and rapidly washes away any sub-
stances squirted into the pocket. It should be noted
that topical antimicrobials (eg, chlorhexidine) as
* 55 Lakeside Drive, Larchmont, NY 10538.
E-mail address: n.triege@verizon.net
1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 6 9 - 9
well as topical antibiotics are cytotoxic [1]. They
provoke the death of healthy cells and elicit their own
inflammatory reactions.
The field of periodontal disease has undergone
rapid expansion of basic information with respect to
microbiology and the origin and characterization of
endotoxins produced by gram-negative anaerobes,
which are responsible for various kinds of periodon-
titis. The endotoxins (lipopolysaccharides) from the
outer cell wall of the specific anaerobes are now
known to be water soluble, and their localization on
the cementum of the root is superficial and not deeply
imbedded [2 – 5]. Smart et al [6] and others have
reported that endotoxin can be washed away or
brushed away without resorting to root planing,
which destroys the cementum. Recently, the treat-
ment concept has changed from aggressive root
planing to debridement. It also has been shown that
connective tissue regrowth in the periodontal crevice
is positively influenced by the presence of adjacent
cementum. In the absence of cementum, reattachment
to dentin does not readily occur. Following peri-
odontal connective tissue reattachment, bone fills
the vertical defects, and the mobility of teeth
decreases when the infection is resolved.
It is difficult to identify the origin of the practice
of root planing. There is no scientific evidence to
show that root planing is any more efficacious than
debridement, without destroying cementum. The
shibboleth of ‘‘scaling and root planing’’ with which
we were taught to begin every treatment plan
deserves to be seriously questioned. In addition to
being destructive, the removal of cementum also
leads to thermal hypersensitivity and more postoper-
ative pain. It is likely that the original intent to
remove scale, formed from salivary calculus, was
extended down onto the exposed root surface and
124 N. Trieger / Oral Maxillofacial Surg Clin N Am 15 (2003) 123–128
continued even in the absence of calculus deposits.
Calculus deposits come from saliva and are most
pronounced opposite the openings of Wharton’s and
Stensen’s ducts. Calculus does not cause periodontitis
but may constitute a mechanical irritant and be
aesthetically objectionable. Calculus, when iden-
tified, is best removed using an ultrasonic cleaner
or by discrete curetting, which clears away just the
deposit and does not strip the cementum.
Diagnosis of periodontitis
A major contribution was made by Loesche [7],
who introduced the ‘‘specific plaque hypothesis.’’
Other investigators have identified several periopath-
ogens associated with patterns of bone loss, primarily
by using the techniques of deep-pocket sampling and
rapid anaerobic subculturing. Periodontitis was thus
identified not as a single entity but as a series of
infections based on microbiologic sources, host
immunocompetency, and the interaction between
these factors. Various syndromes of periodontal infec-
tions have been identified.
Rapidly progressive periodontitis
This is seen in children, young adults, and adults
who are immunocompromised, especially by virtue
of impaired cell-mediated immunity (eg, neutropenia,
cancer chemotherapy, acquired immunodeficiency
syndrome, etc.).
Juvenile periodontitis
Formerly known as periodontosis, this is a disease
of late childhood and early adolescence caused by
Actinobacillus actinomycetemcomitans, a gram-nega-
tive facultative organism that produces a potent
leukotoxin. Early in childhood the organism is found
in the tonsillar crypts and may cause repeated bouts
of tonsillitis. Later on, in addition to dental pathology
it may be found on damaged heart valves, causing
infective endocarditis. This organism is susceptible to
the tetracyclines and not consistently to the penicillin
derivatives or clindamycin. Another important aspect
of this syndrome is that there is usually a qualitative
inherited defect in the behavior of the patient’s
neutrophilic leukocytes.
Acute necrotizing ulcerative gingivitis
This uncommon infection, once called ‘‘trench
mouth,’’ is most often noted in young smokers.
Borrelia vincenti and Prevotella intermedia are two
major pathogens readily identified in this condition
which is often associated with susceptible individuals
who have been subjected to major psychologic stress,
such as a divorce, familial death, or dismissal from
college. Under these circumstances, direct invasion
by the periopathogens has been demonstrated, lead-
ing to necrosis, loss of interdental papillae, lympha-
denitis, gingival bleeding, and fetid odor. Treatment
should be directed first to using a systemic antibiotic,
such as penicillin or amoxicillin, to control the acute
phase before launching into painful manipulations,
such as curettage.
Chronic adult periodontitis
Several specific organisms have been associated
with the localized areas of periodontal bone destruc-
tion (Table 1). These organisms include Porphyromo-
nas gingivalis, Bacteroides forsythus, Prevotella
intermedia, Capnocytophaga species, Campylobacter
rectus, Eikenella corrodens, and Fusobacterium
nucleatum. Other periopathogens will undoubtedly
be identified as the field expands. Many of these
organisms produce an endotoxin (ie, lipopolysac-
charide) that provokes the release of host-tissue fac-
tors, contributing to further breakdown. Cytokines
such as interleukin-1b and tumor necrosis factor-a
from host lymphocytes and other mononuclear phago-
cytes are released. These and other substances can
destroy collagen and bone, causing advancing de-
struction of the periodontium. Bacteriologic culturing
Table 1
Species identified in refractory or recurrent periodontitis
(n = 196)
Organism Percentage of sites
Bacteroides forsythus 84
Spirochetes 83
Motile rods 76
Fusobacterium species 68
Porphyromonas gingivalis 63
Campylobacter rectus 47
Capnocytophaga species 38
Prevotella intermedia 23
Peptostreptococcus micros 18
Actinobacillus actinomycetemcomitans 16
Candida species 14
Enteric rods 9
Resistance to penicillin, tetracycline, and metronidazole
was high.
Adapted from Listgarten MA, Lai C-H, Young V: Micro-
biota and antibiotic resistance (abstract). J Dent Res
1993:72:819. [8]
 N. Trieger / Oral Maxillofacial Surg Clin N Am 15 (2003) 123–128
and the use of specific DNA probes (eg, University
of Pennsylvania Microbiological Testing Laboratory,
Philadelphia, PA) have served to identify putative
organisms. Other tests have been proposed based on
the enzymatic characteristics of the particular patho-
gens. Loesche et al [9] have reported on the use of the
benzoyl-DL-arginine naphthylamide (BANA) test at
chairside. BANA identifies three known pathogens:
P. gingivalis, B. forsythus, and Treponema denticola.
The ultimate aim is to develop a chairside test that will
identify the predominant organisms and guide treat-
ment and follow-up monitoring.
Periodontitis is readily transmissable between
spouses and other cohabitants [10,11]. In the author’s
experience, 80% of patients requiring treatment have
mates with similar organisms. To prevent ‘‘ping-
ponging’’ the disease, both partners should be placed
on antibiotics while the primary patient undergoes
surgical débridement and proceeds into a mainte-
nance phase.
Radiographs reflect the bone level (ie, bone loss)
but do not indicate the current status of the infection.
Bone loss does not necessarily translate into mobility
or condemn a tooth to extraction. Satisfactory resolu-
tion of the infection leads to reduction in bleeding,
pocket depth, and mobility, and new bone often forms
and fills vertical defects [12 – 14].
Surgical techniques
Current management of periodontitis has moved
away from some of the prior practices (see Box 1).
The gingivectomy is passé, as is the practice of bony
recontouring. The author prefers early, one-stage,
full-mouth débridement guided by the severity of
attachment and bone loss. Pockets deeper than 4 or
5 mm merit use of a conservative access flap reflected
just to the bone margin (Fig. 1).
This provides visibility for surgical curettage of
the granulation tissue and avoids the postoperative
edema and pain caused by excessive exposure of
labial bone. Pockets that measure less than 4 or 5 mm
are debrided and curetted at the same appointment
without flap reflection.
With the patient under intravenous sedation and
local anesthesia, the crevicular incisions are made in
areas where pocket depths exceed 4 to 5 mm,
preserving the interdental papillae and marginal gin-
givae. Bony crypts are debrided with a large Prichard
curette followed by use of Gracey curettes to reach
into and under bony defects. Copious saline irrigation
is used during debridement instead of cytotoxic
agents such as chlorhexidine and topical antibiotic
125
Box 1. A summary of guiding surgical
treatment of periodontal infections
1. Periodontitis is an infectious dis-
ease attributable to a relatively
small number of gram-negative
oral pathogenic bacteria that pro-
duce endotoxins.
2. Piecemeal mechanical treatment
(by quadrant or sextant) via scal-
ing and root planing above is
incomplete and invites relapse.
3. The mouth is one ecosystem that
must be treated as a comprehen-
sive unit to eradicate the infection,
preferably under intravenous seda-
tion and local anesthesia.
4. Periodontitis may be considered a
sexually transmitted disease.
5. Soft tissue débridement of perio-
dontal pockets without destruction
of root cementum (planing) is the
goal, in conjunction with systemic
antibiotics effective against the
specific bacterial pathogens.
6. Topical irrigants, including antibi-
otic and antimicrobial agents, are
cytotoxic and jeopardize healing
and regeneration.
7. Regrowth of bone follows reattach-
ment when the infection is resolved
and the cementum is intact. It is a
curable disease.
8. Reinfection is possible, and con-
tinued vigilance and follow-up care
are important.
9. Individual immune status affects
the incidence and long-term re-
sponse to therapy. Examples of
immunocompromised status in-
clude diabetes, HIV, and qualita-
tive and quantitative white blood
cell abnormalities.
10. Calculus has a minimal role in
the initiation and progression of
periodontitis.
The periodontal literature supports each
of these principles and should lead to a
modification of practice.
126 N. Trieger / Oral Maxillofacial Surg Clin N Am 15 (2003) 123–128
Fig. 1. (A) Crevicular incision reflected just to bony margin for débridement. (B) New bone regrowth in absence of infection.
solutions, which damage healthy cells [15,16]. It is
important to remove the granulation tissue that
adheres to the underside of the mucoperiosteal flap
either by curettage or with a soft-tissue rongeur
(Nipro Medical, Miami, FL) or nipper. With the flap
reflected, deposits on the cemental surface can be
discretely removed with a scaler or Cavitron.
Another important component of the recommend-
ed technique is to avoid root planing. The peri-
odontal literature is finally recognizing that root
planing is destructive and compromises reattachment.
Cementum is necessary to induce and enhance fibro-
blastic proliferation and reattachment [4,5]. The
apparent indication to remove diseased cementum is
predicated on the presence of endotoxin [2,3]. Endo-
toxin, a product of gram-negative anaerobic orga-
nisms, is water soluble and not deeply imbedded in
the cementum. It is limited to 40 – 50 mm of the sur-
face cementum and can be washed or even brushed
away. Cementum has specific growth-enhancing fac-
tors that stimulate the fibroblasts necessary for repair
and reattachment of the periodontal membrane.
Therefore, planing the cementum is a waste of time
and effort. Furthermore, root planing jeopardizes soft-
tissue regeneration and bone support and often leads
to thermal hypersensitivity.
The surgical objectives should be to clear the
infection by mechanical debridement and to use anti-
biotics effectively. Minimal periodontitis can be read-
ily controlled by frequent periodic debridement
probably because it disrupts the colonies that re-form
in the pocket. When bone loss has advanced beyond
4 to 5 mm, however, it does not make good sense to rely
only on conservative treatment and wait until the
infection recurs and is finally classified as refractory.
Treatment should be comprehensive and aim to debride
with the least possible damage to cementum, bone, and
mucoperiosteum. This should be accompanied by the
administration of systemic antibiotics known to be
effective against gram-negative anaerobes. Clindamy-
cin, metronidazole, or amoxicillin/clavulanic acid have
been used for 7 to 10 days with rewarding results. The
tetracyclines usually are prescribed for 2 to 3 weeks to
treat juvenile periodontitis [12 – 14,17,18].
Flaps are closed with interrupted sutures, avoiding
apical repositioning (Fig. 1). Healing and bony
regrowth are favored by adequate mucoperiosteal
covering. The surgeon should not try to eliminate
the pocket. The cervical defect fills initially with a
blood clot, which then differentiates into connective
tissue and subsequently into new bone. Periodontal
packs are never used; they retard healing and are
cumbersome, difficult to clean, and inappropriate
when treating an anaerobic infection (Fig. 2).
Even in the advanced periodontitis cases, the oper-
ating time for this procedure is usually 1 to 2.5 hours.
Very few patients undergo repeated sessions with the
hygienist before the main surgical debridement. It is
 N. Trieger / Oral Maxillofacial Surg Clin N Am 15 (2003) 123–128
Fig. 2. (Left) Preoperative first molar bone loss. (Right) One year postoperative bone regeneration.
the primary intent to reduce the bacterial load in deep
pockets that are beyond the reach of the patient’s home
care or the hygienist’s scaling.
Postoperatively, the patient is given a prescription
for a nonsteroidal antiinflammatory agent and
instructed to continue the prescribed antibiotic for
7 to 10 days. Brushing is avoided in the sutured areas,
and either saline irrigation or a chlorhexidine rinse is
recommended for home hygiene. A new brush is
provided on the 1-week return visit and a simple Bass
brushing technique is demonstrated.
Patients are maintained on a soft diet and rarely
report pain or swelling after the first day. The pockets
are reprobed at 1 month to evaluate healing and pocket
reduction. At this time, the patient can be referred
back to his or her dentist for definitive restorative care.
Individual persistent deep pockets may now be ad-
dressed if pocket elimination is requested by the
restorative dentist. Successive 3-month follow-up
visits are scheduled during the first year to evaluate
the long-term results. Reprobing characteristically
finds pockets significantly reduced, with no bleeding
on probing or brushing. If this is not the case, ad-
ditional antimicrobial therapy is pursued after cultur-
ing or taking a sample for DNA probe analysis [14].
Periodic (every 3 months) disruption of reforming
colonies by deep scaling has been shown to provide
adequate control if the case has been refractory to
previous treatment. Topical application of fluoride gel
127
or use of a toothpaste with 1.1% sodium fluoride on a
daily basis at home is effective in preventing reinfec-
tion from a supragingival source. The use of a
chlorhexidine mouth rinse is also helpful in control-
ling supragingival organisms.
Concern has been expressed about the devel-
opment of enterocolitis caused by Clostridium diffi-
cile, a gram-positive organism that produces exotoxin
and damages the cells lining the intestinal tract. This
complication is seen with a number of antibiotics,
such as clindamycin, the cephalosporins, and amox-
icillin. Patients are routinely alerted to this possibility
when postoperative instructions are given. They are
advised that approximately 10% of patients may
develop diarrhea after taking the antibiotic for at least
4 to 5 days. If this occurs, they are instructed to
discontinue the drug and call the office. The diarrhea
usually abates within 24 to 48 hours. If it persists,
oral metronidazole is recommended as the drug of
first choice (250 mg four times per day).
The success rate of combined mechanical and
antibiotic therapy has been cited by others [11 – 14,
17,18] to be superior to that of periodontal therapy
alone. When treating a destructive infectious disease
in the year 2002 it does not make sense to avoid using
effective antibiotics. This technique also has distinct
advantages in the salvage of failing implants. The
principles of debridement and irrigation and the ef-
fective use of systemic antibiotics can enhance the
128 N. Trieger / Oral Maxillofacial Surg Clin N Am 15 (2003) 123–128
repair of endosseous implants that are losing bone
support. No one knowingly root planes an implant. A
soft-tissue debridement is indicated instead [12].
The recognition that specific cytokines, prosta-
glandins, growth factors, and interleukins, are active
participants responding to the elaboration of bac-
terial lipopolysaccharide sheds new light on methods
that may influence the clinical outcome. Golub and
coworkers [19] have shown that the tetracyclines, in
low doses, also inhibit matrix metalloproteinases
such as collagenase and prevents damage to the
attachment apparatus.
Perhaps most intriguing is the realization that
immune-mediated injury is not only an integral part
of the host response to periodontitis, but it has also
been shown to be a significant contributor to myo-
carditis [20]. Cytokines released from activated
T-lymphocytes have been identified as releasing
tumor necrosis factor- a and other injurious mole-
cules that also cause myocardial and vascular injury.
Subsequent myocardial fibrosis correlates with the
presence of T-lymphocytes and macrophages, which
release fibrogenic cytokines such as fibroblast growth
factor and tissue growth factor-b Our knowledge base
is growing rapidly, and the future will bring fasci-
nating new therapies for the management of peri-
odontal infections.
References
[1] Shahan M. The effect of chlorhexidine irrigation on
tensile wound strength [abstract]. J Periodontol 1992;
63:1012.
[2] Hughes FJ, Auger DW, Smales FC. Investigation of
the distribution of cementum-associated lipopolysac-
charides in periodontal disease by scanning electron
microscope immunohistochemistry. J Periodontol Res
1988;23:100.
[3] Nakib NM, Bissada NF, Simmelink JW, Goldstine SN.
Endotoxin penetration into root cementum of perio-
dontally healthy and diseased human teeth. J Periodon-
tol 1982;368 – 78.
[4] Fukazawa E, Nishimura K. Superficial cemental curet-
tage: its efficacy in promoting improved attachment on
human root surfaces previously damaged by periodon-
titis. J Periodontol 1994;65:2, 168.
[5] Cobb CM. Non-surgical pocket therapy. Mechanical
Annals of Periodontology 1996;1:472.
[6] Smart JJ, Wilson M, Davies EH, et al. The assessment
of ultrasonic root débridement by determination of re-
sidual endotoxin levels. J Clin Periodontol 1990;
17:174.
[7] Loesche WJ. The bacterial etiology of periodontal dis-
ease: the specific plaque hypothesis. In: Clark JW, ed-
itor. Clinical dentistry. Philadelphia: Harper & Row;
1987. p. 11 – 20.
[8] Listgarten MA, Lai C-H, Young V. Microbiota and
antibiotic resistance [abstract]. J Dent Res 1993;
72:819.
[9] Loesche WJ, Giordano J, Hujael PP. The utility of the
BANA test for monitoring anaerobic infections due to
spirochetes (Treponema denticola) in periodontal dis-
ease. J Dent Res 1990;69:1696.
[10] Asikainen S, Chen C, Aloluusua S, et al. Can one
acquire periodontal bacteria and periodontitis from a
family member? J Am Dent Assoc 1997;128:1263.
[11] Offenbacher S, Olsvik B, Tonder A. The similarity of
periodontal microorganisms between husband and wife
cohabitants. J Periodontol 1985;56:317 – 23.
[12] Trieger N, Mark L, McKittrick J, et al. Clindamycin v.
tetracycline in the surgical treatment of advanced pe-
riodontitis: a double blind study with applicability for
implant salvage. Int J Oral Maxillofac Implants
1991;7:31.
[13] Trieger N. Surgical treatment of periodontal infections.
Atlas of Oral and Maxillofacial Surgery Clinics of
North America 2000;8:127 – 34.
[14] Loesche WJ, Giordano JR, Soehren S, Kaciroti N. The
nonsurgical treatment of patients with periodontal dis-
ease-results after five years. J Am Dent Assoc
2002;133:311 – 20.
[15] Pucher JJ, Daniel JC. The effects of chlorhexidine di-
gluconate on human fibroblasts in vitro. J Periodontol
1992;63:526.
[16] Dahlen G, Wennstrome JL, Grondahl K, Heijl L. Mi-
crobiological observations at periodic subgingival anti-
microbial irrigation of periodontal pockets. J Dent Res
1989;68:1714 – 5.
[17] Loesche WJ, Schmidt E, Smith BA, Morrison EC,
Caffesse R, Hujael PP. Effects of metromidazole on
periodontal treatment needs. J Periodontol 1991;
62:247 – 57.
[18] Trieger N, Chomenko A. New concepts in the treat-
ment of periodontitis. J Oral Maxillofac Surg 1982;
40:701 – 8.
[19] Golub LM, Wolff M, Roberts S, et al. Treating perio-
dontal diseases by blocking tissue destructive enzymes.
J Am Dent Assoc 1994;125:163.
[20] Beck J, Garcia R, Heiss G, et al. Periodontal disease
and cardiovascular disease. Periodontol 1996;27:1123.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 129 – 138

Management of peri-implantitis
R. Gilbert Triplett, DDS, PhDa,*, J. Adam Andrews, DDS, MDa,
William W. Hallmon, DDSb
a
Department of Oral and Maxillofacial Surgery and Pharmacology, Baylor College of Dentistry,
TAMUSHSC and Baylor University Medical Center, 3302 Gaston Avenue, Dallas, TX 75246, USA
b
Department of Periodontics, Baylor College of Dentistry, TAMUSHSC, 3302 Gaston Avenue, Dallas, TX 75246, USA

Failure of osseointegrated dental implants is a
frustrating problem for the patient and dentist. Peri-
implantitis and occlusal overload are the most com-
mon causes of implant failure after osseointegration,
and they often require removal of the involved
implant. A single failed implant can result in com-
plete prosthetic failure when load-sharing mechanics
of the prosthesis depend on the health and integrity of
each individual implant. Peri-implantitis, which is an
inflammatory process around an osseointegrated den-
tal implant in function with resulting bone loss,
affects approximately 5% to 10% of osseointegrated
implants [1]. Peri-implant mucositis refers to revers-
ible inflammation of the peri-implant soft tissues
without bone loss [1,2]. Accurate diagnosis and
appropriate intervention are essential if implant sal-
vage techniques are to be successful in preventing
implant failure. This article focuses on the methods
available for diagnosis and treatment of peri-implant-
itis that involve various implant systems.
The role of bacteria in the development
of peri-implantitis
The experimental gingivitis model of Löe and
Silness elegantly displays the interaction between
bacterial plaque and gingivitis in a human model
[3]. This landmark study provided the foundation for
* Corresponding author.
E-mail address: gtriplett@tambcd.edu (R.G. Triplett).
the indisputable evidence that linked bacteria to
periodontal disease and was later repeated in several
experimental mucositis/implantitis models [1]. The
evidence compiled from these well-designed studies
identified bacteria as the primary culprit in the
development of peri-implantitis. Large numbers of
gram-negative anaerobic bacteria (A. actinomycetem-
comitans, P. gingivalis, P. intermedia) tend to be
found around implants with objective signs of peri-
implantitis, whereas healthy implants are most often
colonized with flora dominated by gram-positive
cocci [1]. Endotoxins produced by gram-negative
bacteria have the capability to adhere to the implant
surface and produce inflammation and resulting bone
loss around implants in a similar fashion to peri-
odontitis [1]. Several studies have demonstrated the
ability of periodontal pathogens to infect peri-implant
tissues in partially edentulous patients (Fig. 1). This
observation may account for the higher implant
success rates reported in several studies when
implants were placed in edentulous mouths versus
partially edentulous mouths.
The convincing evidence that suggests a bacterial
cause of peri-implantitis provokes several interesting
clinical scenarios regarding the optimal restorative
treatment of the partially edentulous patient. Extra-
polation of research data, which suggest that implant
failures may be significantly higher in partially
edentulous patients with a history of periodontitis,
may lead to the possible recommendation of extrac-
tion of questionable teeth before implant placement.
A possible alternative is treatment of the periodon-
tally diseased tissues before implant placement in
persons with a history of severe periodontal prob-

1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 7 8 - X
130 R.G. Triplett et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 129–138
Fig. 1. (A) Peri-implantitis in the partially edentulous patient. (B,C) Peri-implant mucositis demonstrated by soft tissue
inflammation and increased probing depths without bone loss.
lems. Edentulous mouths tend to host a more benign
flora because of the lack of deep pockets and
crevices that commonly harbor gram-negative an-
aerobes and spirochetes in patients with periodontal
disease. Restoration of a more benign oral microbiota
seems to be a logical goal before implant placement
if long-term implant and prosthetic success is to be
achieved. Several protocols for local and systemic
decontamination of periodontal tissues have been
published; however, the overall long-term benefit
depends on many variables and is often unpredict-
able [4].
Titanium implant surface characteristics and the
relationship with implant success and failure
The implant market has been inundated with
various systems that use different materials, surface
coatings, and manufacturing processes. Implants can
be categorized by differences in macrostructure (eg,
cylindrical, threaded, screw design) and microstruc-
ture (rough versus smooth, commercially pure tita-
nium versus titanium alloy). It seems evident, based
on sound clinical and experimental data, that the
same surface implant characteristics that enhance
osseointegration may possibly increase the risk of
peri-implantitis when placed under certain unfavor-
able conditions.
Patient selection is as important as the type of
implant used if favorable long-term success rates are
to be expected. Implant selection should be based on
a combination of patient-dependent clinical and bio-
logic considerations. Several recent studies have
concluded that rough surface implants demonstrate
a significantly higher percentage of bone-to-implant
contact and faster and stronger osseointegration when
compared to machined surface titanium implants
[5 – 7]. A study by Trisi et al [5] evaluated differences
in the rate of osseointegration between smooth and
rough surface implants in low-density human jaw
bone. At 12 months, the implant-to-bone contact rates
for smooth and rough surface titanium implants were
noted to be 6.7% and 76.75%, respectively. The study
 R.G. Triplett et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 129–138
concluded that although a rough surface may enhance
the rate of osseointegration, it is not able to improve
bone density [5].
The surface oxide layer on commercially pure
titanium and titanium alloy (Ti-6Al-4V) determines
the biocompatibility of a particular implant because it
is the only portion of the implant in contact with host
tissues [6]. The same surface oxide layer has been
shown to exist on rough titanium surfaces, which
may contribute to enhanced osseointegration when
compared to machined surfaces of the identical
material [6]. Several other studies have reported
significantly better bone anchorage with titanium-
oxide-blasted (TiO2-blasted) screw implants than
with machined implants of similar composition and
higher torque removal for implants with increased
surface roughness [6]. Schwartz et al demonstrated
that as the surface roughness of titanium implants
increases, the cytokine and growth factor production
by host osteoblast-like cells also increases and could
account for improved bone formation around the
roughened implant surface [6,8]. Osteoblasts also
have been found to display a more mature phenotype
when grown on rougher surfaces [6]. Studies that
compared different types of roughened titanium sur-
faces concluded that rougher titanium plasma-
sprayed surfaces, corundum-blasted surfaces, and
sandblasted surfaces alone are inferior to surfaces
that are treated with a combination of sandblasting
and acid-etching [5,6,9].
Despite the convincing evidence that rough tita-
nium surfaces may enhance osseointegration and
implant stability, several studies have suggested that
a rough surface may contribute to plaque formation
and higher implant failure rates from peri-implantitis.
Tillmanns et al [9] experimentally induced peri-
implantitis in a canine model with three different
types of implant surfaces (smooth, blasted, and
hydroxyapatite coated). The results after 3 and
6 months revealed plaque accumulation and the same
amount of peri-implant bone loss among all implant
surfaces studied. The study concluded that the three
implant surfaces are equally susceptible to ligature-
induced peri-implantitis [9]. The conclusion is in
direct contrast to most experimental results that
compared implant surfaces. The finding seems to be
well supported that bacterial plaque formation
depends on the surface properties of the implant
material. A surface roughness (Ra) more than 0.2 mm
facilitates early plaque formation in an experimental
model but may be favorable to soft tissue sealing
around transmucosal abutments. Anything below this
Ra value (smoothening) seems to be ineffective in
reducing the amount of plaque formation, but it
131
prevents soft tissue attachment to the implant surface
[10]. Recently published results that evaluated differ-
ent surface coatings on rough surface implants con-
cluded that titanium nitride or zirconium nitride can
reduce the accumulation of plaque by coating the
underlying more reactive titanium surface (indepen-
dent of surface roughness), which may reduce peri-
implant mucositis and peri-implantitis [10].
Considerations with hydroxyapatite-coated
titanium implants
Hydroxyapatite is a bioactive material with osteo-
conductive properties [11]. Hydroxyapatite-coated
titanium alloy implants were first introduced in
1984 for use in restoring partially or totally edentu-
lous maxillas and mandibles [12]. Plasma-sprayed
hydroxyapatite-coated implants demonstrate greater
tolerance to unfavorable healing conditions and pro-
mote bone growth into gaps that measure less than
1 mm [11]. Early reports of faster osseointegration
(biointegration), a stronger bone-to-implant interface
(compared to titanium surfaces), and vertically direct-
ed bone growth along the implant surface were met
with optimism among the implant community
[12,13]. Short- and long-term studies have demon-
strated higher implant-to-bone contact in hydroxy-
apatite implants at 6 weeks when compared to
titanium implants. At 12 weeks, however, the tita-
nium implants displayed superior implant-to-bone
contact compared to hydroxyapatite-coated implants
and had an increase in total implant-to-bone contact
surface area after 1 year. The hydroxyapatite-coated
implants showed a reduction in the implant-bone
contact area over the same period with a significant
decrease in shear strength [12,14,15].
A combination of clinical observation and pro-
spective research suggests that hydroxyapatite-coated
implants may be more prone to enhanced plaque
growth and peri-implantitis because of the question-
able long-term quality of the hydroxyapatite-to-bone
bond and the affinity of microorganisms for the
hydroxyapatite surface [12]. Johnson reported sudden
and rapid bone loss around hydroxyapatite-coated
implants ‘‘after an initial period of apparent success’’
(Fig. 2A) [12]. Several clinical reports claim that
failure of hydroxyapatite-coated implants involves
‘‘significant morbidity and permanent destruction of
bone tissue,’’ whereas titanium implants tend to fail
with minimal loss of bone volume and typically
regenerate to the original dimension of the alveolus
(Fig. 2B) [12]. Despite the claims of higher long-term
failure rates with hydroxyapatite-coated implants,
132 R.G. Triplett et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 129–138
Fig. 2. (A) Hydroxyapatite-coated implants with significant bone and attachment loss. (B) Threaded, machined surface titanium
implants with significant attachment and bone loss.
many clinicians have found them valuable in situa-
tions that involve unfavorable bone quantity or quality
[16]. Recommendations and special considerations
regarding the salvage of infected hydroxyapatite-
coated implants are discussed in this article.
Retrograde peri-implantitis
Radiolucencies around the apical aspect of dental
implants have been attributed to several causes,
including contamination of the implant surface, over-
heating of bone, occlusal overload, preexisting bone
pathology, presence of residual root fragments or
foreign bodies in implant sites, lack of biocompati-
bility, placement of the implant in poor quality bone,
and drilling through the inferior border of the man-
dible or lingual cortex [17]. The term ‘‘retrograde
peri-implantitis’’ was first used to describe radio-
graphic periapical bone loss around a dental implant
without evidence of peri-implant soft tissue inflam-
mation [16]. It was proposed that a radiolucent lesion
found on plain film radiography was caused by
traumatic or premature implant loading that resulted
in microfractures of the peri-implant bone and sub-
sequent resorption. Because the microflora around
implants that suffered from retrograde peri-implantitis
have been shown to be similar to the microflora
around healthy implants, one can assume that infec-
tive failure is not experienced with this type of
implant lesion [3,16]. Because bacteria do not seem
to be a causative factor in this type of implant failure,
normal probing depths without bleeding are often
observed, and resolution of the periapical radiolu-
cency may be observed if traumatic loading of the
implant is relieved.
Diagnosis of the failing implant
and peri-implantitis
Much debate exists regarding the definition of
implant failure. The endeavor to define implant
failure should be preceded with a definition of
success. The First European Workshop on Periodon-
tology defined success as absence of implant mobil-
ity, an average radiographic marginal bone loss of
less than 1.5 mm during the first year of function and
less than 0.2 mm annually thereafter, and absence of
pain and paresthesia [2]. Because this definition was
based on the mean marginal bone loss around Bråne-
mark implants, it seems presumptuous to conclude
that different implants would behave in a similar
manner. Much attention has been given to the terms
‘‘ailing’’ and ‘‘failing’’ when referring to implant
health. It has been proposed that an ‘‘ailing’’ implant
demonstrates radiographic evidence of bone loss and
probing depths more than 5 mm that are stable when
reevaluated at 3 to 4 months. A ‘‘failing’’ implant
demonstrates increasing probing depths, suppuration
or bleeding when probed, and progressive bone loss
[18]. A failed implant no longer is osseointegrated or
never achieved osseointegration. These implants dis-
play peri-implant radiolucency caused by fibrous
tissue encapsulation, are clinically mobile, and dem-
onstrate dullness to percussion. Failed implants must
be removed to prevent chronic bone loss and the
possibility of osteomyelitis (Fig. 3) [18].
Esposito et al [19] concluded that radiographic
examination and mobility testing were the most
reliable parameters in determining the prognosis of
osseointegrated implants. Marginal bone loss around
the neck of the implant can be evaluated radiograph-
ically and by peri-implant probing. Reproducing the
 R.G. Triplett et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 129–138
Fig. 3. Failed implant with fibrous tissue encapsulation and
clinical mobility.
exact radiographic exposure geometry is difficult and
can lead to clinically significant variability. Despite
the latter finding, it seems that serial radiographs are
more reliable in monitoring peri-implant conditions
than probing, particularly in the setting of inflamed
peri-implant tissues and bony defects [19]. Digital
subtraction radiography eventually may prove to be
superior to traditional radiographic techniques in
evaluating subtle changes in peri-implant bone den-
sity, and it is highly recommended, if available
[19,20]. If no clinical evidence of inflammation is
present, radiographs should be obtained 1 year after
implant placement and not more than every 2 years
thereafter [1]. Radiographs should be taken more
frequently if clinical evidence of peri-implant inflam-
mation (increased probing depths) raises suspicion of
peri-implantitis [1].
Differences among implant systems make the task
of quantitatively defining the normal amount of
marginal bone loss difficult. Several studies have
demonstrated that marginal bone loss and biologic
width are determined largely by the implant system
and the host response. Gargiulo [21] demonstrated
that natural teeth have a normal biologic width of
2.73 mm. Hermann et al [22] evaluated the anatomic
location of the biologic width around various types of
dental implants and determined that the one-piece
(single-stage) nonsubmerged implant with a rough/
smooth border placed at or 1 mm apical to the
alveolar crest resulted in a biologic width that most
closely resembled that of natural teeth (2.84 mm).
They also concluded that the presence of a microgap
(component interface) in two-piece implant systems
significantly affects the level of crestal bone and soft
tissue dimensions. The presence of a microgap in
two-stage implant systems may play an important
role in the development of peri-implantitis because
133
bacterial contamination of this interface has been
demonstrated [22].
Peri-implant probing may provide valuable
information regarding implant health or progression
of disease. Healthy implants generally have probing
depths less than 4 mm, with interproximal probing
depths normally 0.5 to 1 mm more than the buccal
and lingual probing depths [20]. Peri-implant pockets
of 5 mm or more should be considered an indicator of
peri-implantitis because deep pockets have been
shown to harbor a microflora consistent with inflam-
mation and bone loss [1].
There has been much debate over the location of
the probe tip and the effect on the peri-implant tissues
during peri-implant probing in disease and health.
Several studies suggest that there is a resilient soft
tissue collar in peri-implant health and mucositis. The
tip of the probe may travel without impedance to the
alveolar crest in peri-implantitis, however [20]. It has
been demonstrated experimentally that the probe tip
penetrates apically to the laterally displaced junc-
tional epithelium with Brånemark implants, resulting
in the probe tip approaching the alveolar crest [18].
This is in contrast to the findings around ITI dental
implants, which demonstrate the probe tip location at
the apical termination of the junctional epithelium
(0.05 mm in healthy sites and 0.02 mm in diseased
sites) [19]. Issues regarding the practicality and
reproducibility of probing depths have brought the
entire practice of peri-implant probing into question.
Several studies have concluded that peri-implant
probing damages the peri-implant soft tissues, but
the magnitude and long-term effects have yet to be
determined [19].
It can be said that there is a positive correlation
between peri-implant probing depths and the degree of
peri-implant mucosal inflammation, but not necessar-
ily bone loss [19]. More important than a single
measurement at a single point in time is documentation
of progressive peri-implant bone loss as evidenced by
increasing probing depths. It is recommended that
baseline probing depths be acquired at the time of
prosthetic reconstruction to account for the predictable
marginal bone loss during the first several months after
implant placement. Assuming that probing has been
accepted by the surgeon as a method of providing
beneficial clinical information that exceeds the poten-
tial for harming the peri-implant tissues, the frequency
and interval must be tailored to each individual patient
based on compliance, oral hygiene, and other risk
factors. Probing attachment levels relate probing
depths to a fixed reference point on the implant or
abutment and provide valuable information regarding
attachment loss over time. Probing attachment level
134 R.G. Triplett et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 129–138
increases of 2 mm or more should be interpreted as
marginal bone loss [19].
In addition to providing information regarding
attachment level and bone loss, peri-implant probing
can demonstrate peri-implant inflammation clinically.
Bleeding on probing is another controversial issue
regarding the diagnosis of peri-implantitis. No cor-
relation has been shown between bleeding and the
histologic or radiographic changes associated with
peri-implant mucositis or peri-implantitis around
smooth surface threaded implants [19]. Another study
demonstrated that the absence of bleeding on probing
around ITI implants was associated with implant
health, whereas bleeding on probing correlated highly
with peri-implant mucositis and peri-implantitis [19].
Because the reason for these differences may be
caused by inconsistent probing forces or other fac-
tors, bleeding on probing is not scientifically sup-
ported as a method of diagnosing peri-implantitis
[19]. Several authors still maintain that ‘‘probing
depth measurements related to a fixed landmark on
the implant and examination of the bleeding tendency
of the peri-implant tissues seem to be well-suited for
the longitudinal monitoring of peri-implant stability’’
[1]. It has been recommended that nonmetallic probes
(eg, plastic) with a calibrated constant probing force
be used for more reliable, less traumatic measure-
ments. A probing force of 0.25 N has been recom-
mended by several authors to fulfill the previously
mentioned criteria [1,20]. Mombelli and Lang [1]
recommend the use of standardized probes, such as
the Audio Probe, titanium plasma-sprayed probe, or
the HAWE Click Probe, for consistent measurements.
Implant mobility in a previously healthy implant
should be considered a sign of failure. Even implants
that show significant bone loss usually are immobile if
any direct implant-to-bone contact remains. Several
devices and methods have been proposed for evalu-
ating implant mobility. Periotest (Siemens AG, Ben-
sheim, Germany) is a device used for measuring the
damping effect of the supporting tissues to a stand-
ardized force as an indicator of slight changes in
implant mobility. Although several studies have
reported its success in detecting subtle changes in
the bone-to-implant interface, its usefulness and accu-
racy are still being evaluated [1,18]. A torque wrench
that delivers a set amount of force is another method
for determining implant osseointegration. An implant
is considered to be osseointegrated if a torque of 10 to
20 Ncm is applied to the implant without resulting
mobility [18]. Although not scientifically supported,
the percussion test is reported to be a simple and
sensitive method of determining osseointegration. The
implant abutment interface is percussed with the blunt
end of an instrument and the sound is interpreted. A
‘‘ringing’’ sound in considered favorable for osseo-
integration, whereas a ‘‘dull’’ sound suggests fibrous
tissue encapsulation [18]. Finally, OSSTELL (Integra-
tion Diagnostics, Inc., Sävedalen, Sweden) is a new
Food and Drug Administration – approved device that
uses resonance frequency analysis to assess implant
stability. The reliability and usefulness of the device in
implant dentistry still are being evaluated.
A systematic approach should be used when
evaluating implants for possible disease. It is often
more prudent to evaluate the implant by assuming
disease and proving health because this mode of
reasoning tends to eliminate the possibility of false-
negative screening results. If probing is to be per-
formed at the recall visit, the new probing depths
should be compared to baseline measurements and
the overall trend observed. Attachment levels also
should be recorded because peri-implant pockets
potentially can remain normal as marginal bone loss
progresses and the attachment and marginal tissue
level follows. If no attachment loss is evident and
probing depths are normal, one can assume that the
implant is associated with a nonpathogenic micro-
flora and that the peri-implant tissues are not clin-
ically inflamed [1].
Implant salvage
The type of intervention for implant salvage
depends largely on clinical findings and implant
characteristics. Peri-implant mucositis is a reversible
process that often responds well to conservative,
noninvasive treatment because increased probing
depths are usually caused by soft tissue inflammation
and not crestal bone loss. Implants that show evi-
dence of mucosal inflammation, plaque, and accu-
mulation of calculus but lack suppuration and
probing depths more than 3 mm are often treated
effectively with mechanical débridement. Special
nonmetallic instruments should be used for débride-
ment to minimize surface defects and the theoretical
possibility of galvanic corrosion. Special rubber cups
and implant polishing paste can be used to remove
plaque [20]. Subgingival chlorhexidine irrigation may
be added to the regimen for cases in which probing
depths have increased to 4 to 5 mm and inflammation
and plaque/calculus deposits are noted. Generally,
treatment for 3 to 4 weeks with chlorhexidine as a
daily rinse or gel is required to achieve the desired
result [20].
Lang et al [20] recommend antibiotic treatment in
addition to mechanical débridement and antiseptic
 R.G. Triplett et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 129–138
treatment for peri-implant probing depths more than
6 mm. Tetracycline has been used for many years in
treating periodontally involved teeth because of its
antibiotic properties and facilitating effect on fibro-
blastic growth and attachment on root surfaces [16].
The decision to use tetracycline in implant therapy
may depend on the type of implant surface because
of the possibility of tetracycline altering the com-
position of the hydroxyapatite coating [16]. Treating
titanium implant surfaces with topical tetracycline
(50 mg/mL) for 3 minutes before regenerative tech-
niques has been reported as successful [23]. Actisite
(Alza, Palo Alto, CA) is a local delivery system that
consists of nonresorbable tetracycline fibers for local
use around periodontally involved teeth. It has been
supplanted largely by resorbable delivery systems.
Several case reports have documented the effective-
ness of the tetracycline fibers around infected dental
implants [20].
Some authors advocate systemic antibiotics
before and during implant salvage techniques. Three
recommended regimens are (1) clindamycin, 150 mg
orally three times a day, (2) doxycycline hyclate,
100 mg orally twice daily, and (3) amoxicillin with
or without clavulonic acid, 500 mg four times a day.
Some authors recommend the addition of metroni-
dazole if amoxicillin is selected for systemic treat-
ment. All regimens should be started 2 days before
implant salvage treatment and continued for 10 days
after [18].
Peri-implantitis has been shown to be an endo-
toxin-mediated host response that progresses to
implant failure if not treated. Differences among
implant surface characteristics are important when
selecting appropriate interventional techniques for
implant salvage. Zablotsky et al [24] published a
landmark study in 1992 that compared the abilities
of various chemotherapeutic modalities to detoxify
endotoxin-contaminated implant surfaces (grit-blast-
ed titanium alloy and grit-blasted titanium alloy with
a hydroxyapatite plasma spray coating). The study
concluded that detoxification of hydroxyapatite-
coated implants is best accomplished with anhydrous
citric acid reconstituted to a 40% pH 1 (supersatu-
rated) solution that is applied to the implant surface
for 30 seconds to 1 minute [24]. This effect is caused
by a demineralization of the superficial hydroxyapa-
tite layer. The titanium grit-blasted surface was effec-
tively decontaminated by burnishing with saline or
citric acid for 1 minute. Air powder abrasives, such as
sodium bicarbonate mixed with sterile water, also
have been shown to be effective in decontaminating
implant surfaces [25]. Stannous fluoride treatment
seems to result in significantly greater levels of
135
endotoxin on both types of implant surfaces when
compared with controls, whereas treatment with
chlorhexidine gluconate, tetracycline HCl, hydrogen
peroxide, and chloramine T is less effective at
decreasing levels of endotoxin than saline burnishing
alone. Charge interactions have been proposed as the
reason for endotoxin having a greater affinity for
hydroxyapatite-coated surfaces than grit-blasted tita-
nium alloy surfaces [24]. Studies also have shown
that chlorhexidine and stannous fluoride can result in
the binding of endotoxin to the hydroxyapatite
implant surface because of the inherent charge char-
acteristics of these two compounds [16].
Surgical treatment should be considered for any
implant that displays radiographic evidence of pro-
gressive crestal bone loss that still has adequate
residual bony anchorage. If surgical treatment is to
be initiated, the patient first should be placed on
antibiotics, then a mucoperiosteal flap is elevated to
expose the defect, and the implant surface is decon-
taminated by one of the methods described previ-
ously. Any granulation tissue should be removed with
instruments that do not scratch or contaminate the
titanium surface so that it is more favorable for
regenerated tissue or osseointegration. Hydroxyapa-
tite implants should be inspected for evidence of
surface pitting, cracking, or color changes. If the
hydroxyapatite shows wear or contamination, the
entire layer should be removed mechanically and
the underlying titanium surface decontaminated in
the manner described previously.
Guided bone regeneration (GBR) also can be used
to treat osseous defects around failing implants (Fig. 4)
[16]. Decontamination of the diseased implant surface
is paramount if this technique is to be used. The
process for GBR involves placing a resorbable or
nonresorbable membrane over an osseous defect to
permit new bone growth into the defect while inhib-
iting soft tissue infiltration [26]. GBR has been per-
formed around peri-implant osseous defects with and
without grafting; however, bone fill and attachment
gain seem to be better achieved when a grafting
material is used in conjunction with a barrier mem-
brane [16]. Several different materials, including
demineralized freeze-dried bone, autologous bone,
and resorbable bovine-derived hydroxyapatite, have
been used in various forms as grafting material in
GBR. It has been proposed that an alloplast, such
as nonresorbable hydroxyapatite or bioactive glass,
be used if the implant surface is difficult to decontam-
inate because of the presence of vents, holes, or tor-
tuous osseous defects [18]. Alloplasts do not achieve
biologic healing in these cases but are effective in
filling bony defects and minimizing peri-implant
136 R.G. Triplett et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 129–138

Fig. 4. Guided bone regeneration around a failing implant. (A) Large bony defect of a rough surface implant with increased
probing depths and radiographic evidence of bone loss. (B) Bony defect covered with graft material after surface débridement
and decontamination with tetracycline HCl paste. (C) Resorbable collagen membrane prepared to cover the graft material and
necks of implants. (D) Collagen membrane in place. (E) Water-tight closure around implants and surgical site (some authors
advocate burying single-stage implant systems after this procedure). (Courtesy of Dr. Tinou Roncone.)

pockets [18]. Although studies have demonstrated that
GBR is effective around implants, questions have
been raised regarding the ability of the regenerated
bone to ‘‘re-osseointegrate’’ with the implant surface.
Persson et al [27] published a study investigating
whether ‘‘re-osseointegration’’ could be accomplished
after the treatment of peri-implantitis. The study con-
cluded that ‘‘re-osseointegration’’ could not be
achieved with a smooth (turned) surface but was
consistently successful with a decontaminated sand-
blasted, large grit, acid-etched surface [27]. If bone
loss has progressed to the apical one third of the
implant, removal is indicated because there is little
chance of successful salvage [25].
 R.G. Triplett et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 129–138
The membrane selected for GBR should be bio-
compatible and easy to place, maintain its original
shape, and be capable of functioning as a barrier for at
least 6 weeks [16]. A commonly used nonresorbable
material, expanded-polytetrafluorethylene, has pro-
vided good results for many years. Several resorbable
membranes recently have gained popularity, however,
because of a reduced number of complications from
infection or exposure and because they do not require
retrieval surgery. Although it is acceptable to leave a
membrane exposed during immediate implant place-
ment in fresh extraction sites, primary closure over the
membrane is necessary in implant salvage to prevent
contamination and infection of the involved area. If
GBR is to be used as a salvage procedure for implants,
it is imperative that a nonresorbable membrane remain
covered for as long as possible because premature
removal decreases the chances of success and results
in less-than-optimal bone fill [16]. Many authors
recommend a minimum of 6 weeks before removal
of a nonresorbable membrane [23]. If a resorbable
membrane is to be used, it should provide barrier
function for at least 6 weeks. It has been suggested
that all patients remain on an antiseptic mouth rinse,
such as chlorhexidine gluconate 0.12%, for 6 weeks or
until the membrane is retrieved to minimize the
chance of membrane infection [28]. If the membrane
becomes prematurely exposed, however, immediate
removal is indicated to prevent contamination of the
regenerating tissues [16]. Single-stage implant sys-
tems ideally should be submerged during GBR sal-
vage procedures to minimize contamination of the
membrane and regenerating tissues.
Summary
Peri-implantitis is a treatable disease that affects
functioning osseointegrated implants. Although
unfavorable mechanical loading may play a con-
tributing role, peri-implantitis seems to be mediated
primarily by the endotoxins from gram-negative
bacteria and the host response around the implant
site. Many patients who were previously considered
unfavorable candidates for implant therapy are
being treated successfully when certain treatment
considerations and implant maintenance programs
are implemented.
It is essential that implant surgeons have a firm
understanding of the favorable aspects of implant
design and the potential liabilities when these systems
are placed in unfavorable clinical conditions. Studies
have demonstrated that the same potential benefits of
implant materials and surface characteristics also may
137
lead to increased failure rates when these implants are
affected by peri-implantitis. In the presence of
adequate apical osseointegration, compromised
implants that present with peri-implantitis must
undergo thorough débridement and be decontami-
nated before any attempt at GBR. GBR may be
accomplished successfully with many different types
of membranes and grafting materials if the implant
surface is thoroughly decontaminated before regen-
erative therapy. Implant salvage is an important, yet
often ignored, component of clinical practice that can
prevent implant and prosthetic failure if the principles
of decontamination, biomodification, and guided tis-
sue regeneration are understood and followed.
References
[1] Mombelli A, Lang N. The diagnosis and treatment of
peri-implantitis. Periodontology 2000;17:63 – 76.
[2] Albrektsson T, Isidor F, et al. Consensus report of ses-
sion IV. In: Lang NP, Karring T, editors. Proceedings
of the First European Workshop on Periodontology.
London: Quintessence; 1994. p. 365 – 9.
[3] Löe H, Morrison E. Epidemiology of periodontal dis-
ease. In: Genco RJ, Goldman HM, Cohen DW, editors.
Contemporary periodontics. St. Louis: CV Mosby Co.;
1990. p. 106 – 16.
[4] Hammond BF, Genco RJ. Sensitivity of periodontal
organisms to antibiotics and other antimicrobial agents.
In: Genco RJ, Goldman HM, Cohen DW, editors. Con-
temporary periodontics. St. Louis: C.V. Mosby Co.;
1990. p. 161 – 9.
[5] Trisi P, Rao W, Rebaudi A. A histometric comparison
of smooth and rough titanium implants in human low-
density jawbone. Int J Oral Maxillofac Implants 1999;
14:689 – 98.
[6] De Leonardis D, Garg A, Pecora G. Osseointegration
of rough acid-etched titanium implants: 5-year follow-
up of 100 Minimatic implants. Int J Oral Maxillofac
Implants 1999;14:384 – 91.
[7] De Leonardis D, Garg A, Pecora G, et al. Osseointe-
gration of rough acid-etched implants: one-year fol-
low-up of placement of 100 Minimatic implants. Int
J Oral Maxillofac Implants 1997;12:65 – 73.
[8] Schwartz Z, Kieswetter K, Dean DD, Boyan BD.
Underlying mechanisms at the bone-surface interface
during regeneration. J Periodont Res 1997;32:166 – 71.
[9] Tillmans HW, Hermann JS, Tiffee JC, et al. Evaluation
of three different dental implants in ligature-induced
peri-implantitis in the beagle dog. Part II. Histology
and microbiology. Int J Oral Maxillofac Implants
1998;13:59 – 68.
[10] Gröbner-Schreiber B, Griepentrog M, Haustein I, et al.
Plaque formation on surface modified dental implants:
an in vitro study. Clin Oral Implants Res 2001;12:
543 – 51.
138 R.G. Triplett et al / Oral Maxillofacial Surg Clin N Am 15 (2003) 129–138
[11] Strnad Z, Strnad J, Povysil C, et al. Effect of plasma-
sprayed hydroxyapatite coating on the osteoconductiv-
ity of commercially pure titanium implants. Int J Oral
Maxillofac Implants 2000;15:483 – 90.
[12] Johnson BW. HA-coated dental implants: long-term
consequences. CDA Journal of the California Dental
Association 1992;20:33 – 41.
[13] Meffert RM, Block MS, Kent JN. What is osseointe-
gration? International Journal of Periodontics and Re-
storative Dentistry 1987;7:9 – 21.
[14] Cook SD, Kay JF, et al. Interface mechanics and his-
tology of titanium and hydroxyapatite-coated titanium
for implant applications. Int J Oral Maxillofac Implants
1987;2:15 – 22.
[15] Gottlander M, Albrektsson T. Histomorphometric stud-
ies of hydroxyapatite-coated and uncoated CP titanium
threaded implants in bone. Int J Oral Maxillofac Im-
plants 1991;6:399 – 404.
[16] Meffert RM. Periodontitis vs. peri-implantitis: the
same disease? The same treatment? Crit Rev Oral
Biol Med 1996;7:278 – 91.
[17] Scarano A, Di Domizio P, Petrone G, et al. Implant
periapical lesion: a clinical and histologic case report.
J Oral Implantol 2000;26:109 – 13.
[18] Martin RM, Carter JB, Barber HD. Surgical implant
failures. In: Fonseca R, Powers MP, Barber HD, edi-
tors. Oral and maxillofacial surgery: reconstructive and
implant surgery. Philadelphia: W.B. Saunders Co.;
2000. p. 275 – 308.
[19] Esposito M, Hirsch J-M, Lekholm U, et al. Biological
factors contributing to failures of osseointegrated oral
implants (I). Success criteria and epidemiology. Eur J
Oral Sci 1998;106:527 – 51.
[20] Lang NP, Wilson TG, Corbet EF. Biological compli-
cations with dental implants: their prevention, diagno-
sis and treatment. Clin Oral Implant Res 2000;11:
146 – 55.
[21] Garqinto AW, Wentz FM, Orban B. Dimensions and
relations of the dontogingival junction in humans.
Journal of Periodontology 1961;32:261 – 7.
[22] Hermann J, Buser D, Schenk RK, et al. Biologic width
around one- and two-piece titanium implants: a histo-
metric evaluation of unloaded nonsubmerged and sub-
merged implants in the canine mandible. Clin Oral
Implants Res 2001;12:559 – 71.
[23] Mellonig JT, Griffiths G, Mathys E, et al. Treatment of
the failing implant: case reports. International Journal
of Periodontics and Restorative Dentistry 1995;15:
385 – 95.
[24] Zablotsky MH, Diedrich DL, Meffert RM. Detoxifi-
cation of endotoxin contaminated titanium and hy-
droxyapatite-coated surfaces utilizing various
chemotherapeutic and mechanical modalities. Im-
plant Dentistry 1992;1:154 – 8.
[25] Jovanovic SA. The management of peri-implant break-
down around functioning osseointegrated dental im-
plants. J Periodontol 1993;64:1176 – 83.
[26] Rominger JW, Triplett RG. The use of guided tissue
regeneration to improve implant osseointegration.
J Oral Maxillofac Surg 1994;52:106 – 12.
[27] Persson LG, Berglundh T, Sennerby L, et al. Re-os-
seointegration after treatment of peri-implantitis at dif-
ferent implant surfaces: an experimental study in the
dog. Clin Oral Implant Res 2001;12:595 – 603.
[28] Lehmann B, Bragger U, Hammerle CHF, et al. Treat-
ment of an early implant failure according to the prin-
ciples of guided tissue regeneration (GTR). Clin Oral
Implant Res 1992;3:42 – 8.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 139 – 146
Management of posttraumatic soft tissue infections
A. Omar Abubaker, DMD, PhD
Department of Oral and Maxillofacial Surgery, School of Dentistry, Virginia Commonwealth University,
521 North 11th Street, PO Box 980566, Richmond, VA 23298, USA
Between 10 and 12 million traumatic wounds are
treated annually in emergency departments in the
United States [1,2]. More than 50% of these lacer-
ations are caused by blunt trauma. The others are
caused by sharp objects, such as metal, glass, and
wood. Only a small percentage of these wounds is
caused by mammalian and nonmammalian bites
[3,4]. Most of these lacerations occur on the face,
scalp, and arms, mostly in young men [2]. Because of
these locations, an important goal of management of
these wounds is to avoid infection, which can lead to
cosmetically and functionally unacceptable scars [5].
The current management of traumatic soft tissue
injuries incorporates many of the surgical principles
developed over the past century. These principles
include a thorough understanding of the pathophysi-
ology of wounding, the risk factors for infection, the
basic mechanisms by which posttraumatic sepsis
develops, and the appropriate methods for treatment
of these injuries and the prevention of complications.
This article reviews the defense mechanisms involved
in soft tissue wound healing, describes the risk factors
for posttraumatic wound infections, and discusses the
prevention and treatment of such infections.
Physiologic effects of wounding
The human body has evolved several defense
mechanisms body to protect itself from the micro-
biologic invasion that causes wound infection. These
mechanisms include efficient mechanical barriers to
bacteria and competent biologic protection mecha-
nisms. The mechanical barriers to bacterial invasion
E-mail address: Abubaker@vcu.edu
1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 6 8 - 7
include the epithelium of the skin and mucous
membranes. The skin also grants chemical protection
in the form of surface lipids, and the mucous mem-
branes provide protection by surface Ig A and an
acidic pH. The skin and oral mucosal surfaces are
also inhabited by normal flora that can compete with
potential microbial pathogens.
Biologic protection is provided in the form of
various internal mechanisms that are induced by local
tissue damage. These mechanisms are triggered when
the mechanical line of defense is violated. Once this
damage ensues, an intense chemical activity is trig-
gered, which involves activation of the kallikrein-
kinin system, the release of amines, and an increase in
vascular permeability that allows for influx of hu-
moral and cellular immunologic elements. These
elements are ultimately responsible for recognition
of the organisms involved and their subsequent
phagocytosis [6].
Traumatic wounds carry with them a higher
degree of contamination because the mechanical
protective features of the skin and mucous membrane
are disrupted, which allows direct invasion of micro-
organisms into the deeper tissues. The internal sys-
temic biologic mechanisms of host defense are also
compromised by the effects of the trauma. For
example, with major trauma there is a decrease in
cellular immune functions, intracellular killing, and
endothelial system function [7,8]. There is also a
decrease in humoral factors, such as the immunoglob-
ulins and the complement system [8]. If traumatic
shock develops, there is also a decrease in systemic
perfusion, which, in the presence of local tissue
damage, may reduce blood flow to the wounded area
and further compromise containment of invading
bacteria [9,10].
140 A.O. Abubaker / Oral Maxillofacial Surg Clin N Am 15 (2003) 139–146
Risk factors for posttraumatic wound infections
Studies on management of traumatic wounds have
shown that the rate of infection ranges between 2.5%
and 11.5% [11 – 13]. This rate is influenced by several
variables, which are often referred to as risk factors.
Some of these factors are related to the host, the
environment, and the type of wound, whereas others
are related to the techniques used to manage the
wounds [12,14 – 16]. Factors related to the wound
and patient include (1) size, configuration, and depth
of the wound, (2) location of the injury, (3) mech-
anism of injury, (4) type and amount of contamina-
tion, including presence of a foreign body, (5) time
between injury and wound closure, (6) care of the
wound between injury and definitive care, and (7) age
and systemic condition of the patient. The technical
risk factors for development of wound infection
include (1) inadequate débridement of foreign bodies,
further bacterial contamination, and presence of devi-
talized tissues, (2) inadvertent introduction of foreign
materials into the wound during cleansing, (3) inad-
equate hemostasis and failure to eliminate dead space,
which provides an environment for bacterial col-
onization, (4) using an excessive number of sutures
to close the wound, and (5) placing excessive tension
on the sutures used to approximate the tissue edges
and compromising local tissue perfusion.
Prevention of posttraumatic wound infection by
proper wound care
The greatest deterrent to posttraumatic wound
infection is a healthy wound. This status can be
accomplished by (1) thorough cleansing, (2) identifica-
tion, assessment, and atraumatic débridement of all
devitalized tissues, (3) removal of all foreign materials,
and (4) careful handling of the tissues. Evacuation of
any hematoma and obliteration of all potential dead
spaces by proper wound repair and approximation of
all tissue layers are also essential elements in the
prevention of soft tissue infections. Appropriate use
of prophylactic antibiotics and tetanus prophylaxis are
also important preventive measures [17].
Anesthesia
Before one begins definitive wound management,
adequate patient comfort should be ensured not only
to permit maximum wound repair but also to allow
thorough examination, cleansing, and débridement of
the wound. In most patients, comfort can be accom-
plished with local anesthesia (lidocaine with
1:100,000 or 1: 200,000 epinephrine). Inserting the
needle through the open part of the laceration can
decrease the pain of the injection [18,19]. Use of a
field block is beneficial in reducing distortion of the
operative site when accurate approximation of the
wound edges is necessary. If there are skin or tissue
flaps of doubtful viability, the use of lidocaine with-
out epinephrine is desirable to avoid further impairing
circulation [20].
Local anesthesia can be supplemented with anal-
gesics or sedatives, if necessary. General anesthesia is
preferable for uncooperative or combative patients
and for children. General anesthesia is also often
indicated in patients who require prolonged proce-
dures, as in extensive lacerations, lacerations that
require flap rotation or grafts for closure, and con-
comitant repair of facial bones fractures [21]. Patients
with confirmed or suspected involvement of impor-
tant structures, such as joints, nerves, or tendons, also
may be better treated in an operating room setting
[5,19,22].
Wound preparation
Once the patient is comfortable and the wound is
adequately anesthetized, a complete and methodical
cleansing of the wound should be achieved. The
process involves irrigation and débridement of the
wound before draping of the operative site in prep-
aration for wound closure. Irrigation is essential in
preventing infection because it removes debris, dirt,
microorganisms, and devitalized tissue from the
wound, which results in a reduction in infection rate
[3]. Irrigation with normal saline solution using a
50-mL syringe and a 16-gauge needle is adequate for
most lacerations [23,24]. In general, 250 to 500 cc of
solution provides adequate irrigation for small
wounds [24]. High-pressure irrigation is occasionally
indicated for large wounds and when a high degree of
contamination is present [17]. High-pressure irriga-
tion has been shown to decrease the bacterial count of
wounded tissues and decrease the rate of infection
[25]. Vigorous irrigation in general, however, and
high-pressure irrigation in particular may force debris
into the wound and can cause further tissue damage
[26,27]. For these reasons, high-pressure irrigation
should be used with discretion and reserved for
heavily contaminated wounds in which its benefits
may outweigh its risks [17]. Similarly, use of con-
centrated povidone-iodine, hydrogen peroxide, and
detergents may cause significant tissue damage and
should be avoided [28]. Some authors even ques-
tioned the value of any form of wound irrigation in
noncontaminated facial and scalp wounds [29].
 A.O. Abubaker / Oral Maxillofacial Surg Clin N Am 15 (2003) 139–146
Decontamination of the skin around the wound is
another step in preventing wound infection. Decon-
tamination can be accomplished effectively with mild
soap and water or with a providone iodine or hexa-
chlorphene solution. Use of chemical solutions should
be limited to the adjacent skin surface, and caution
should be used to avoid getting them into the wound
because these agents are noxious to tissues. Because
shaving may damage hair follicles and allow for
bacterial access and increase the risk of wound infec-
tion, it should be kept to a minimum. In areas such as
the eyebrows, it should be avoided totally [23].
Débridement is considered by many to be the
most important step in avoiding infection. One goal
of wound débridement is to remove all foreign
materials that may be attached to, or embedded in,
the edges of the laceration or abrasion. Removing
foreign material can be accomplished by thorough
irrigation of the wound, vigorous scrubbing, or even
surgical excision using a surgical blade or small
curette. The second goal of débridement is the
removal of devitalized tissue. The general rule
regarding débridement of devitalized tissue is that
all crushed or frayed edges that may become necrotic,
all obviously nonviable tissue, and all grossly con-
taminated tissue should be removed. Exceptions to
this rule involve certain anatomic areas, such as the
eyebrow and vermilion border of the lip, where such
débridement may compromise accurate realignment
of the tissues and in instances in which tissue
débridement would result in excessive tension on
the suture line. In these instances, it is preferable to
avoid extensive trimming of the tissue and to approx-
imate the irregular edges without tension [30]. In the
management of facial wounds, only minimal débride-
ment is generally required because of the excellent
blood supply. When tissue loss is present, it is
recommended to save all tissue that has a chance to
survive to provide the basis for secondary reconstruc-
tion at a later time [22].
Wound closure
Once the devitalized tissue has been removed, a
decision must be made regarding the timing and type
of wound closure. There are at least three major
choices [16]: (1) primary closure and healing by
primary intention, (2) leaving the wound open, treat-
ing it with frequent dressing changes, and allowing it
to heal by secondary intention and wound contrac-
ture, and (3) using delayed primary closure, in which
the wound is splinted in a position of rest with an
occlusive dressing and is closed in 3 to 5 days when it
is free of infection and necrotic tissue. The choice of
141
method is generally based on the assessment for risks
of infection. It is often difficult to determine which
treatment to use for a given wound, and the length of
the ‘‘golden period’’ within which it can be closed
primarily varies [31]. In general, there is a direct
correlation between the time from injury to closure of
the wound and the risk of infection. It is generally
accepted that wounds with a high risk of infection
should be closed as soon as possible (within the first
6 – 8 hours), whereas wounds with low risk of infec-
tion, such as those in the head and neck area, can be
closed primarily within the first 18 to 24 hours after
injury [17,24]. After 24 hours, for most wounds,
consideration should be given to packing them open
and performing a secondary repair 4 to 8 days later
[17,30].
Postoperative wound care
For most wounds, patients should be instructed to
keep them covered with a nonadherent dressing for at
least 24 to 48 hours to protect the wound from gross
contamination. After this period, the patient should
wash, but not scrub or soak, the wound. Once the
wound is left open, it should be cleaned two to three
times a day with a cotton applicator stick and hydro-
gen peroxide or soap and water. The patient also
should be instructed to place a topical antibiotic
ointment on the wound. Prospective and retrospective
studies have shown the value of topical antibiotic
agents in decreasing infection in certain wounds
[32,33]. Their benefit beyond day 5 of wound closure
remains controversial, however [19]. A topical anti-
biotic ointment also keeps the wound moist, which
speeds the rate of wound epithelialization [34].
The routine use of systemic prophylactic anti-
biotics to prevent infection of soft tissue wounds is
not recommended [35]. Because traumatic wounds
are, by definition, contaminated wounds, however,
some of them are prone to infection despite all
attempts to follow the principles of appropriate
wound care. In such instances, and when the con-
sequences of wound infection may be devastating,
prophylactic antibiotics may be indicated [16]. The
use of prophylactic antibiotics should be tailored
individually based on factors such as the degree of
bacterial contamination and the presence of the pre-
disposing risk factors discussed previously. Several
studies suggest that the administration of an antibiotic
to prevent infection of the soft tissues after trauma
may be of more value when given prophylactically
within the first 3 to 4 hours of the injury and
continued until the wound is sutured than when given
142 A.O. Abubaker / Oral Maxillofacial Surg Clin N Am 15 (2003) 139–146
for a therapeutic course postoperatively [16]. Some
authors recommend that prophylactic antibiotics be
continued for 24 hours [36], however, with additional
doses if the operative procedure is prolonged.
Tetanus continues to occur worldwide despite the
availability of an effective vaccine. Consideration
always should be given to administration of tetanus
prophylaxis at the time of initial wound examination.
A summary guide to tetanus prophylaxis is available
through the Centers for Disease Control (Table 1) [37].
Treatment of posttraumatic soft tissue
wound infections
Because traumatic wounds are caused by different
types of trauma, the resulting soft tissue damage often
ranges from blunt damage to penetrating and complex
tissue injuries. The spectrum of microbiologic soft
tissue contamination also varies and ranges from
simple wound colonization without invasion to frank
tissue infection and necrosis. Contamination of post-
traumatic wounds can occur from introduction of
endogenous normally nonpathogenic bacteria into
the wound, from missiles or other objects, or from
contact with various surfaces at the time of injury.
The variation in degree of tissue injury and microbial
invasion can result in a wide spectrum of wound
infections, including general inflammation, a cel-
lulitis or simple abscess, a diffuse inflammation and
spreading cellulitis with or without signs of systemic
toxicity, or a progressive necrotizing soft tissue
infection [16]. Once wound infection occurs, it results
in the immediate release of inflammatory cytokines
from the monocytes and macrophages, which causes
a delay in wound healing. These cytokines also may
result in the release of growth factors that stimulate
fibrosis and result in localized scar hypertrophy [20].
Table 1
Recommendations for prophylaxis against tetanus
After clean, minor wounds
Tetanus-diphtheria
History of tetanus immunization toxoid
Number of previous Yes
doses < 3 or not known
Number of previous doses  3
Timing of last dose
Within 5 y No
Within 5 – 10 y No
> 10 y ago Yes
a
Examples of these wounds include contaminated wounds, puncture wounds, avulsions, burns, and crush injuries.
Regardless of the mechanism of injury and the degree
of suspected wound contamination, once the signs
and symptoms of wound infection become clinically
evident, the goals of treatment should be to optimize
tissue perfusion and nutrition, remove devitalized
tissue, prevent further spread of the infection or
further tissue destruction, and achieve wound closure
[38]. To attain these goals, Fields et al described a
five-phase unified approach to treatment of posttrau-
matic soft tissue infections [6]. The phases of this
systematic approach that are applicable to wounds in
the head and neck are phase 1: early recognition of
infection; phase 2: rapid initiation of empiric anti-
biotics; phase 3: immediate surgical incision and
drainage or débridement, when necessary; and phase
4: early wound closure.
Early recognition
Early recognition and diagnosis are essential to
treating wound infections successfully and minimizing
morbidity [16]. The diagnosis of infection should be
based on the clinical signs and symptoms, the labo-
ratory findings, and identification of predisposing
factors. A definitive diagnosis can be made later based
on culture of the microorganisms and examination of
histologic specimens if débridement was performed.
Because a minor inflammatory reaction, pain, and
swelling are normal components of the early stages of
wound healing, recognition of infection based on
these cardinal signs can be difficult and wound
infections can be missed [16]. Changes in these signs
on frequent examination of the wound and obser-
vation of increased swelling, tenderness, induration,
discoloration, and fluctuation should increase the
index of suspicion [6,16,39]. Systemic signs of infec-
tion, such as malaise, fever, tachycardia, and leuko-
cytosis, may be present in patients with infected
After all other woundsa
Tetanus Tetanus-diphtheria Tetanus
immunoglobulin toxoid immunoglobulin
No Yes Yes
No No No
No Yes No
No Yes No
 A.O. Abubaker / Oral Maxillofacial Surg Clin N Am 15 (2003) 139–146
wounds. Severe infections also may be associated
with altered glucose metabolism, respiratory distress,
altered mental status, and hypotension [40]. These
signs and symptoms are nonspecific, however, espe-
cially in critically ill patients. When such signs and
symptoms and local signs of wound infection are
present, they are indicative of a serious problem and
may be associated with progression of the infection
into deeper tissue. Bacterial growth on culture pro-
vides essential information for treatment, although
this is not essential for the diagnosis of posttraumatic
wound infection. Because bacteria can colonize trau-
matic wounds even in the absence of infection,
interpretation of culture results must be conducted
with care. Finally, when there is an associated open
fracture, differentiation between infection that origi-
nates in soft tissues and infection that is related to the
fracture should be made by surgical exploration,
microscopic examination of tissue samples, culture
of organisms from the bone, or radiographic exam-
ination [41].
Also important in the recognition and diagnosis of
posttraumatic infection is differentiation between
superficial, or local soft tissue infection, and deep
infections that involve fascia and muscle. Risk fac-
tors for the presence of deep soft tissue infection
include systemic diseases, such as diabetes, immuno-
compromised status, underlying malignancy, and
local factors, such as gross contamination and delay
of wound closure.
Empiric antibiotics
The choice of empiric antibiotics should be based
on the location and depth of the wound and the
systemic status of the patient. For most posttraumatic
head and neck soft tissue infections, first-generation,
second-generation, and third-generation cephalospo-
rins remain the drugs of choice for Staphylococcus
aureus [42,43]. In most of the severe posttraumatic
soft tissue infections, a combination of antibiotics is
often used until data from culture and antibiotic
sensitivity testing become available on day 2 or 3 [6].
Fungal infections are uncommon in posttraumatic
wounds. Candida albicans and, less commonly, Phy-
comycetes are opportunistic organisms that may
cause secondary infection after systemic antibiotic
therapy, however. Fluconazole and amphotericin B
are the most commonly used agents for treatment of
these fungal infections [33,43]. Viral infections are
also rare in posttraumatic wound infections but can be
present in immunocompromised patients, patients
with major multisystemic trauma, and burn patients
[44]. The presence of these infections increases the
143
likelihood of sepsis from bacterial infection. Viral
infections should be treated with such agents as
systemic acyclovir.
Débridement
The sine qua non of treatment of wound infection
is to provide wide drainage of any purulent material
and débride all necrotic tissue [16]. Devitalized
tissue acts as a culture medium for bacteria, creates
an anaerobic environment, and impairs the cellular
and humoral immune defenses [6,45]. It is imper-
ative that débridement be performed if nonviable
tissue is present on the margins of the wound. In
instances in which tissue necrosis continues beyond
the time of injury, such as in blast gun shot wounds,
repeated débridement may be necessary. Frequent
cleansing of the wound and repeated dressing
changes should be instituted to assist in removal of
bacteria, exudate, and devitalized tissue. It is impor-
tant that a representative tissue specimen or a sample
of purulent discharge be submitted for microbiologic
identification of the offending organisms and for
diagnosis of the extent of tissue necrosis. These
specimens are best obtained from beneath the intact
skin, away from the wound, because bacteria recov-
ered from these sites are more likely to represent the
true pathogens and not part of the wound coloniza-
tion [6].
Wound closure
Wound closure should be performed when all the
infection is resolved and healthy granulation tissue is
present [6,46]. This goal usually can be accomplished
by secondary intention or primary closure if the
wound is small. Large wounds may require split-
thickness skin grafting or flaps, however.
Management of bite wounds
Each year 1% of all visits to emergency rooms
(approximately 300,000 visits) are related to bite
injuries [47,48]. Ninety percent of these injuries are
dog and cat bites, and the rest are human or other
animal bites. Although most of these bites involve the
hand, a significant percentage (16%) of dog bites are
in the face and scalp, whereas only a small percentage
(2%) of cat bites are in these regions [4,48].
Approximately 3% to 18% of dog bites and 28%
to 80% of cat bites become infected [48,49]. The risk
of infection is greatest for crush injuries, puncture
wounds, and wounds to the hand [47]. Human bites,
144 A.O. Abubaker / Oral Maxillofacial Surg Clin N Am 15 (2003) 139–146
although less common than dog and cat bites, were
believed to be more prone to infection than those
inflicted by animals. Such reports are biased by
emphasis on human bites of the hand that present
late with infection already present, however [50]. The
generally reported poor prognosis in such cases is
probably caused by the delay in treatment and loca-
tion of the wound rather than the cause of injury.
Human bites to the face, lips, and ears have a lower
risk of infection (less than 3%), than human bites
elsewhere (10% – 12%) if treated properly [50,51].
With proper wound care, the rate of dog bite infec-
tions in the head and neck is also as low as 1.4%,
even when prophylactic antibiotics were not used
[52]. Wound infection from cat bites to the face is
3%, whereas it is 19% in the hand and 18% in the
lower extremities [53].
The management of bite wounds remains some-
what controversial because there is considerable
variation in the patients studied and in wound
severity, and there is a relatively small number of
cases in the studies reported [47]. There is general
agreement, however, that the most effective method
for preventing infection of these wounds is through
proper initial management, which should include
thorough irrigation with a copious volume of normal
saline, débridement when necessary, and closure of
the wound when appropriate. The wound also
should be cultured. Radiographic examination of
the adjacent facial bones may be indicated in cases
in which a fracture is suspected. Proper selection
and administration of antimicrobial therapy, which is
based on the suspected flora and the risk of the
wound for infection, are also important elements
of treatment.
The most commonly isolated organism from
infected dog and cat bites is Pasteurella species,
50% and 75%, respectively [48,49]. Pasteurella canis
is the most common isolate from dog bites, whereas
P. multocida subspecies multocida and septica are the
most common isolates from cat bites. Other common
aerobic organisms isolated from dog and cat bites
include streptococci, staphylococci, Moraxella, and
Neisseria [48]. Common anaerobic organisms that
are isolated from mixed infections include Fusobac-
terium, Bacteroides, Porophyromonas, and Prevo-
tella. Isolates from human bites include S. aureus,
Eikenella corrodens, Haemophilus influenzae, and
beta-lactamase – producing oral anaerobic bacteria
[48,49].
Based on this spectrum of organisms, the empir-
ical therapy for animal bites should be directed
against Pasteurella, streptococci, staphylococci,
and anaerobes Commonly used empiric antibiotic
therapy for outpatients includes penicillin or diclox-
acillin, ampicillin, or a first-generation cephalo-
sporin [49,54,55]. A recent multicenter study of
infected animal bites concluded that optimal thera-
peutic agents for this purpose include a combination
of a b-lactam antibiotic and b-lactamase inhibitor, a
second-generation cephalosporin with anaerobic
activity, or combination therapy with either penicil-
lin and a first-generation cephalosporin or clinda-
mycin and fluoroquinolone [48]. Based on their in
vitro activity, azithromycin, trovafloxacin, and keto-
lide antibiotics are also effective against all of the
common aerobic and anaerobic isolates from these
bites [56 – 58]. For treatment of high-risk human
bites, empiric therapy includes penicillin and anti-
staphylococcal agents, such as dicloxacillin and
nafcillin [59]. An alternative to this combination is
oral amoxicillin-clavulanic acid, which provides
excellent in vitro coverage of the suspected patho-
gens [57,60]. Ertapenem also has an excellent
potency against the full range of animal and human
bite pathogens [58].
Use of prophylactic antibiotics for prevention of
bite wound infection remains an area of considerable
controversy. Some authors argue that such use is cost
effective in selected cases. These cases include
wounds seen more than 8 hours after injury, wounds
that affect the hand, wounds that involve bone and
joints, cat bite puncture wounds, especially near a
joint or joint prosthesis, and wounds in patients with
a compromised immune system. In such cases, the
use of prophylactic antibiotic may decrease the rate of
infection from 15% to 20% to approximately 5%
[47,49,54,61]. Excluding these instances, most
authors believe that prophylactic antibiotics are not
indicated, especially for patients with bite wounds in
the head and neck and patients who present within the
first few hours after injury and show no evidence of
infection [51,53,54,62,63]. This conclusion is based
on the fact that most studies that showed a benefit of
prophylactic antibiotics used diverse treatment proto-
cols for wounds that were not of similar severity.
Most of these studies also were biased by inclusion of
hand wounds, which have a higher incidence of
infection. Several studies have shown that with early
surgical intervention using irrigation, débridement,
and primary closure, the wound infection rate in the
face, with and without the use of prophylactic anti-
biotics, is similar and routine antibiotic prophylaxis is
not justified [52,53,55,64,65].
The timing of closure of bite wounds also con-
tinues to be controversial. In general, the time of
wound closure is influenced by the period between
injury and presentation for care. Researchers gen-
 A.O. Abubaker / Oral Maxillofacial Surg Clin N Am 15 (2003) 139–146
erally agree that bite wounds in the face, where
cosmesis is a major concern, can be closed primarily
or covered with skin graft when necessary—even
4 days after injury—with an acceptably low risk
for infection because of the rich blood supply
[51,63,66].
Summary
With improvements in surgical techniques and the
availability of effective systemic antibiotics, the
incidence of posttraumatic soft tissue wound infec-
tions is relatively low. When such infections occur,
however, they can result in significant morbidity.
These complications can be reduced by prompt initial
management of the wound and proper treatment if
infection occurs.
References
[1] Edlich RF, Rodeheaver GT, Thacker JG. Wounds, bites
and stings. In: Moore EE, Mattox KL, Feliciano DV,
editors. Trauma. 2nd edition. Norwalk (CT): Appleton
& Lange; 1991. p. 715 – 50.
[2] Stussman BJ. National Hospital Ambulatory Medical
Care Survey: 1994 emergency department summary.
Advance data from vital and health statistics. No
275. DHHS publication no. (PHS, 96 – 1250). Hyatts-
ville (MD): National Center for Health Statistics; 1996.
[3] Edlich RF, Rodeheaver GT, Morgan RF, Berman DE,
Thacker JG. Principles of emergency wound manage-
ment. Ann Emerg Med 1988;17:1284 – 302.
[4] Hollander JE, Singer AJ, Valentine S, et al. Wound
registry: development and validation. Ann Emerg
Med 1995;25:675 – 85.
[5] Singer JA, Hollander JE, Quinn JV. Evaluation of man-
agement of traumatic lacerations. N Engl J Med 1997;
337:1142 – 8.
[6] Fildes J, Bannon MP, Barrett J. Soft tissue infections
after trauma. Surg Clin North Am 1991;71:371 – 84.
[7] Howard RJ. Effect of brain injury, mechanical trauma
and operation on immune defense. Surg Clin North
Am 1979;59:199 – 211.
[8] Miller SE, Miller CL, Trunkey DL. The immune con-
sequences of trauma. Surg Clin North Am 1982;
62:167 – 81.
[9] Miles AA, Miles EM, Burke J. The value and duration
of defense reactions of the skin to the primary lodg-
ment of bacteria. Br J Exp Pathol 1957;38:79 – 84.
[10] Miles AA, Niven JSF. The enhancement of infection
during shock produced by bacterial toxins and other
agents. Br J Exp Pathol 1950;31:73 – 81.
[11] Dire DJ, Hogan DE, Riggs MW. A prospective evalu-
ation of risk factors for infections from dog bite
wounds. Acad Emerg Med 1994;1:258 – 66.
145
[12] Hollander JE, Singer AJ, Valentine SM, et al. Risk
factors in patients with traumatic lacerations. Acad
Emerg Med 2001;8:716 – 20.
[13] Stamou SC, Maltezou HC, Psaltopoulou T, et al.
Wound infection after minor limb laceration: risk fac-
tors and the role of antimicrobial agents. J Trauma
1999;46:1078 – 81.
[14] Cruse PJE, Foord R. A five-year prospective study of
23,649 surgical wounds. Arch Surg 1973;107:206 – 10.
[15] Fernandoz AM, Herruzo CR, Gomez-Sanch F, et al.
Four year study of the risk factors of surgical wound
infection in 5260 traumatological patients. Minerva
Med 1996;87:189 – 94.
[16] Molnar JA, Burke JF. Prevention and management of
infection in trauma. World Surg 1983;7:158 – 63.
[17] Curtin JW. Basic plastic surgical techniques in repair of
facial lacerations. Surg Clin North Am 1973; 53:33 – 46.
[18] Kelly A-M, Cohen M, Richards D. Minimizing the
pain of local infiltration anesthesia for wounds by in-
jection into the wound edges. J Emerg Med 1994;12:
593 – 5.
[19] Leach J. Proper handling of soft tissue in the acute
phase. Facial Plast Surg 2001;17:227 – 38.
[20] Key SJ, Thomas DW, Shepherd JP. The management
of soft tissue facial wounds. Br J Oral Maxillofac Surg
1995;33:76 – 85.
[21] Sacchetti A, Schafermeyer R, Geradi M, et al. Pediatric
analgesia and sedation. Ann Emerg Med 1995;23:
237 – 50.
[22] Dickinson JT, Jaquiss GW, Thompson JN. Soft tissue
trauma. Otolaryngol Clin North Am 1976;13:130 – 4.
[23] Singer AJ, Hollander JE, et al. Pressure dynamics of
various irrigation techniques commonly used in the
emergency department. Ann Emerg Med 1994;24:
36 – 40.
[24] Stuzin JM, Engrav LH, Buchler PK. Emergency treat-
ment of facial lacerations. Postgrad Med 1982;71:
88 – 94.
[25] Madden J, Edlich RF, Schauerhamer R, et al. Applica-
tion of principles of fluid dynamics to surgical wound
irrigation. Curr Top Surg Res 1971;3:85.
[26] Feng LJ, Eaton C. Soft tissue infection in lower ex-
tremity trauma. Clin Plast Surg 1986;13:735 – 45.
[27] Wheeler CB, Rodeheaver GT, Thacker JG, et al. Side
effects of high pressure irrigation. Surg Gynecol Obstet
1976;143:775 – 8.
[28] Oberg MS, Lindsey D. Do not put hydrogen peroxide
or povidone iodine into wounds! Am J Dis Child
1987;141:27 – 8.
[29] Hollander JE, Richman PB, Werblud M, Miller T,
Huggler J, Singer AJ. Irrigation in facial and scalp
lacerations: does it alter outcome? Ann Emerg Med
1998;31:73 – 7.
[30] Zook EG. The care of facial lacerations. J Fam Pract
1978;6:1089 – 95.
[31] Berk WA, Osbourne DD, Taylor DD. Evaluation of the
‘‘golden period’’ for wound repair: 204 cases from a
Third World emergency department. Ann Emerg Med
1988;17:496 – 500.
146 A.O. Abubaker / Oral Maxillofacial Surg Clin N Am 15 (2003) 139–146
[32] Halasz NA. Wound infection and topical antibiotics.
Arch Surg 1977;112:1240 – 4.
[33] Teepe RG, Koebrugge EJ, Lowick C, et al. Cytotoxin
effects of topical antimicrobial and antiseptic agents
on human keratinocytes in vitro. J Trauma 1993;35:
8 – 19.
[34] Dire DJ, Coppola M, Dwyer DA, Lorrette JJ, Karr JL.
A prospective evaluation of topical antibiotics for pre-
venting infections in uncomplicated soft-tissue
wounds repaired in the ED. Acad Emerg Med 1995;
2:4 – 10.
[35] Cummings P, Del Beccaro MA. Antibiotics to prevent
infection of simple wounds: a meta-analysis of
randomized studies. Am J Emerg Med 1995;13:
396 – 400.
[36] Robson MC, Heggers JP. Delayed wound closure
based on bacterial count. J Surg Res 1970;2:379 – 83.
[37] Centers for Disease Control. Diphtheria, tetanus and
pertussis: Recommendations for vaccine use and other
preventive measures. Recommendation of the Immu-
nization Practices Advisory Committee (ACIP). Mor-
bidity and Mortality Weekly Report 1991;40:1 – 22.
[38] Klein DG, Fritsch DE, Amin SG. Wound infection
following trauma and burn injuries. Critical Care Nurs-
ing Clinics of North America 1995;7:627 – 42.
[39] Esrig BC, Frazee L, Stephenson SFS, et al. The pre-
disposition to infection following hemorrhagic shock.
Surg Gynecol Obstet 1977;144:915.
[40] Pruitt BA, Yurt RW. Treating burn and soft tissue in-
fections. Infections in Surgery 1983;2:625 – 8.
[41] Garner JS, Harvis WR, Emori TG, et al. CDC defini-
tions for nosocomial infections. Am J Infect Control
1988;16:128 – 40.
[42] Seligman M, Martyn JAJ. Burn wound infections. In:
Martyn JAJ, editor. Acute management of the burned
patient. Philadelphia: WB Saunders; 1990. p. 288 – 305.
[43] Vitale P. Antimycotic and antifungal agents. In: Wil-
liams BR, Baer CL, editors. Essentials of clinical phar-
macology in nursing. Springhouse (PA): Springhouse
Corp.; 1994. p. 451 – 62.
[44] Kagan RJ, Naragi S, et al. Herpes simplex virus and
cytomegalovirus infections in burned patients. J Trau-
ma 1985;25:40 – 5.
[45] Haury B, Rodeheaver G, Vensko J, Edgerton MT, Ed-
lich RF. Débridement: an essential component of trau-
matic wound care. Am J Surg 1978;135: 238 – 42.
[46] Fratianne R, Papay F, Housini I, et al. Keratinocyte
allografts accelerate healing of split-thickness donor
sites: applications for improved treatment of burns.
J Burn Care Rehabil 1993;14:148 – 54.
[47] Goldstein EJC. Bite wounds and infection. Clin Infect
Dis 1992;14:433 – 8.
[48] Talan DA, Citron DM, Fredrick AM, et al. Bacterio-
logic analysis of infected dog and cat bites. N Engl J
Med 1999;340:8592 – 7.
[49] Goldstein EJC, Citron DM, Finegold SM. Dog bite
wounds and infection: a prospective clinical study.
Ann Emerg Med 1980;9:508 – 12.
[50] Callahan M. Prophylactic antibiotics in common dog
bite wounds: a controlled study. Ann Emerg Med
1980;9:410 – 4.
[51] Chen E, Hornig S, Shepherd SM, Hollander JE. Pri-
mary closure of mammalian bites. Acad Emerg Med
2000;7:156 – 61.
[52] Guy JR, Zook EG. Successful treatment of acute head
and neck dog bite wounds without antibiotics. Ann
Plast Surg 1986;17:45 – 58.
[53] Dire DJ. Cat bite wounds: risk factors for infection.
Ann Emerg Med 1991;20:973 – 9.
[54] Callahan M. Controversies in antibiotic choices for bite
wounds. Ann Emerg Med 1988;17:1321 – 30.
[55] Wolff KD. Management of animal bite injuries of the
face: experience with 94 patients. J Oral Maxillofac
Surg 1998;56:838 – 43.
[56] Goldstein EJC, Citron DM, Gerardo SH, Hudspeth M,
Merriam CV. Activities of HMR 3004 (RU 64004) and
HMR 3647 (RU 66647) compared to those of eryth-
romycin, azithromycin, clarithromycin, roxithromycin,
and eight other antimicrobial agents against unusual
aerobic and anaerobic human and animal bite patho-
gens isolated from skin and soft tissue infections in
humans. Antimicrob Agents Chemother 1998;42:
1127 – 32.
[57] Goldstein EJC, Citron DM, Hudspeth M, Gerardo SH,
Merriam CV. In vitro activity of BAY 12 – 8039, a new
8-methoxyquinolone, compared to the activities of 11
other oral antimicrobial agents against 390 aerobic and
anaerobic bacteria isolated from human and animal bite
sound skin and soft tissue infections in humans. Anti-
microb Agents Chemother 1997;41: 1552 – 7.
[58] Goldstein EJC, Citron DM, Merriam CV, et al. Com-
parative in vitro activity of ertapenem and 11 other
antimicrobial agents against aerobic and anaerobic
pathogens isolated from skin and soft tissue animal
and human bite wound infections. J Antimicrob Che-
mother 2001;48:641 – 6.
[59] Fallouji MA. Traumatic love bites. Br J Surg 1990;77:
100 – 1.
[60] Lindsey D, Christopher M, Hollenbach J, et al. Natural
course of human bite wound. J Trauma 1987;27: 45 – 8.
[61] Bunzli WF, Wright DH, Hoang AT, et al. Current man-
agement of human bites. Pharmacotherapy 1998;18:
227 – 34.
[62] Chidzonga M. Human bites of the face. S Afr Med J
1998;88:150 – 2.
[63] Donkor P, Bankas DO. A study of primary closure of
human bite injuries to the face. J Oral Maxillofac Surg
1997;55:479 – 81.
[64] Medeiros I, Saconato H. Antibiotic prophylaxis for
mammalian bites. Cochrane Database Syst Rev 2001;
2:CDO1738.
[65] Venter TH. Human bites of the face: early surgical
management. S Afr Med J 1988;74:277 – 9.
[66] Agrawal K, Mishra S, Panda KN. Primary reconstruc-
tion of major human bite injuries to the face. Plast
Reconstr Surg 1992;90:394 – 8.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 147 – 153

Management of infections associated with laser-assisted

cosmetic skin resurfacing
Robert A. Strauss, DDS, MD
Department of Oral and Maxillofacial Surgery, Virginia Commonwealth University, Medical College of Virginia,
PO Box 980566, Richmond, VA 23298, USA

The last decade has seen a large increase in the
number of cosmetic surgical procedures performed in
the United States. New operations, techniques, and
technologies seem to appear each year. Although the
exact reason for this increase is not totally clear,
cultural and social factors have played a role. Once
limited only to older women, cosmetic surgery is
currently sought by younger people, men, and even
children. Much controversy has been generated con-
cerning the wisdom of our society seeking the
performance of totally elective surgery. Proponents
make arguments based on psychological well-being
and the relative safety of these procedures, whereas
critics point out that even with all the technologic
advances, cosmetic surgery is still associated with a
low, but inevitable, degree of morbidity and even
occasional mortality.
Among the more recent advances in aesthetic facial
surgery has been the development of laser-assisted
cosmetic skin resurfacing (CSR), which has become
one of the most popular procedures performed by
maxillofacial cosmetic surgeons [1]. Similar in effect
to chemical peeling and dermabrasion (which also are
associated with the same incidence and types of
infections), CSR has gained in popularity because of
new advances in laser technology that allow precise,
targeted, tissue ablation and the fact that these devices
are relatively affordable and portable enough for
office-based use. Combining this technology with a
better understanding of the physiology of skin wound
healing and the availability of new scientifically based
pharmaceuticals for preoperative and postoperative
E-mail address: rastrauss@vcu.edu
skin care has made the procedure easily and safely
performed by various practitioners. Thousands of
these procedures have been performed with generally
excellent aesthetic results.
Although currently well accepted as a useful
surgical procedure in the cosmetic surgeon’s arma-
mentarium, CSR has had its share of controversy. As
with many technology-based procedures, the rapid
availability of the ever-changing types of hardware,
coupled with the need and desire of the practitioner to
generate a positive cash flow from these expensive
devices, has brought with it an immediate influx of
practitioners of varying backgrounds, surgical exper-
tise, and laser experience. Unfortunately, the rapid
advances in laser technology and widespread use of
CSR preceded the ability of researchers to provide a
clear understanding of the pathophysiology involved,
the mechanism of laser action, the indications and
contraindications for their use, and some of the
potential and inevitable complications that were to
be expected. As a consequence, CSR has been
associated with several serious and occasionally di-
sastrous complications.
With time and much scientific research, many of
the unknowns that led to some of the initial problems
have been answered. Reasonable technique and patient
selection protocols, based on defensible science, have
decreased the incidence of complications dramatically.
They do occur, however, and the wise surgeon always
should be wary and diligent in looking for early
evidence of untoward intraoperative and postoperative
sequelae. Some of the complications seen to date
include scarring, hypopigmentation and hyperpigmen-
tation, blotchiness, pruritis, ectropion, and infection.
Of these complications, infection associated with CSR

1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 8 1 - X
148 R.A. Strauss / Oral Maxillofacial Surg Clin N Am 15 (2003) 147–153
has proved to be one of the more controversial areas in
terms of prophylaxis and treatment.
Normal healing of the skin
When a wound is created in skin, a series of
important biochemical and histopathologic changes
occurs. Shortly after injury, the adjacent basal cell
layer ceases its normal role of vertical migration
and maturation and begins the process of horizontal
migration to cover the exposed tissues. Any
adnexal epithelial structures that remain in the
wound also begin to form surface epithelium and
proliferate in all directions. Simultaneously, col-
lagen formation is accelerated in the wound to fill
the defect. The more rapid the reepithelialization of
the wound from external and internal sources, the
less collagen formation that occurs and the less
likelihood there is of any significant scarring. The
main goal in all of the skin resurfacing methods is
to remove the epidermis and papillary dermis and
preserve the reticular dermis with its intact epithe-
lial adnexal structures to allow for speedy internal
and external reepithelialization.
The skin is a fundamental physical barrier to the
entrance of microorganisms into the body and a
vitally important functioning member of the overall
immune system [2]. CSR, by its very design,
removes the entire epidermis and a significant por-
tion of the dermis, thereby altering the immune
mechanism and enhancing the ability of various
microorganisms to colonize the exposed surface.
For this reason a calculable incidence of infection
after CSR is to be expected.
Normal microbiology of the skin
Unlike the oral cavity, the skin of the face is
generally not a particularly good place for the
habitation of microorganisms. The skin, being dry
and having a slightly acidic pH, makes a hostile
environment, and the constant desquamation makes
it difficult for organisms to attain sufficient num-
bers to cause clinical infection [3]. Various bacteria
and yeasts consistently inhabit the skin in predict-
able and stable numbers, typically residing in the
stratum corneum or along the hair shafts and
follicles [4]. Patients with acute or chronic acne
present a much more diverse and complicated
microbiologic picture.
Because the epidermis is constantly exposed to the
environment, most of the bacteria found on skin are,
not surprisingly, aerobic or facultative anaerobes (eg,
Staphylococcus epidermidis and other Staphylococcus
species, Streptococcus viridans, and Corynebacterium
species). As would also be expected, the most common
fungal organism seen is Candida albicans, although
species of Pityrosporum, Malassezia, and Torulopsis
are also present. Not only does CSR result in loss of the
protective epidermal barrier but it also changes the
postoperative environment. This result, plus the me-
dicaments used after surgery, produces a dramatic
change in the local immune system and allows for
overgrowth of the normal inhabitants and infection by
organisms from adjacent areas, such as the oral cavity
and nares.
Bacterial infections
Numerous studies have examined the incidence
and microbiology of postoperative infections after
CSR [5 – 10]. The incidence seems to be variable,
ranging from near zero in some studies to alarming
rates in others, even with the use of prophylactic
antibiotics. Weinstein et al, in a study of more than
1900 patients, showed an overall rate of bacterial
infection of only 0.5% [10]. Similarly, in a study of
500 consecutive patients, Nanni and Alster reported
an overall rate of infection of less than 1% [7].
Conversely, Walia and Alster reported an infection
rate of 24% in a subset of their population who
received intraoperative and postoperative antibiotics
[9]. Gaspar et al also reported that 4 of 14 patients
(28%) not given prophylactic antibiotics also
developed postoperative bacterial infections [5]. Sev-
eral other studies have shown rates generally in the
range of 4% to 8% [6,8].
There seem to be multiple reasons for such
diverse findings, and understanding this is important
to critically analyzing the literature in this area. One
of the main obstacles to studying the incidence of
postoperative infections is the difficulty in defining
the term ‘‘infection.’’ It is variably defined in
different studies as a positive culture associated with
signs and symptoms, a positive culture regardless of
signs and symptoms, or the presence of clinical
signs and symptoms with or without any culture
results. Even more nonspecific, some studies defined
the patient as being ‘‘infected’’ whenever the cli-
nician, in his or her judgment, decided to institute
systemic antibiotics regardless of the culture results
[11]. Such variations render interstudy comparisons
difficult or impossible.
Other reasons for the diversity of findings likely
include the wide variations in the use of prophy-
 R.A. Strauss / Oral Maxillofacial Surg Clin N Am 15 (2003) 147–153
lactic antibiotics, the differences in postoperative
dressing management, and even when the study
was performed, because differences in these ever-
evolving techniques and technologies must play a
role in the incidence of infection. Other factors that
have been shown to make significant and dramatic
differences in infection rate are the size and location
of the operative site. Full-face resurfacing seems to
have a much higher incidence of infection than
resurfacing restricted to only one or two cosmetic
units. Anunt et al, in their study of 396 patients,
showed an eightfold decrease in infection when only
partial resurfacing was performed, and no infections
occurred in patients who received resurfacing of
only a single cometic unit [8].
The clinical signs and symptoms of a postoper-
ative infection after CSR are well recognized
[7,9,10]. Persistent or increasing pain (especially if
the quality of the pain differs from the initial ‘‘sun-
burn’’ pain normally felt for the first 48 hours),
burning, erythema (localized, worse than the sur-
rounding operated skin, and often showing crusting),
pruritis that lasts more than 72 hours, and a yellow
exudate are common findings. Papules, pustules, and
erosions also may be seen. There may be an asso-
ciated fever, but this not a consistent finding.
Depending on the microbes involved, a foul odor
may be noted by the patient. Although postoperative
bacterial infections may be obvious and overt, one
should remember that some individuals present with
only subtle findings, such as prolonged patchy erythe-
ma. Given the consequences that can result and the
success rate of early treatment, the astute clinician
always should suspect an infection when any finding
is out of the ordinary or there is an atypical complaint
from the patient after a CSR procedure and appro-
priately rule out this possibility.
The microbiology of postoperative bacterial
infections varies and can be altered by many
factors, such as the use of prophylactic antibiotics
and the type of postoperative management (eg,
closed versus open dressings). Because the overall
infection rate is relatively low, no large-scale stud-
ies have shown a statistically significant classifica-
tion of the organisms most likely involved in CSR.
A review of the literature shows that the most
prevalent organisms are Staphyloccocus aureus, S.
epidermidis, and Pseudomonas aeruginosa, but
occasionally Enterobacter, Escherichia coli, Pro-
teus, and Serratia species are found. These orga-
nisms are similar to the organisms typically
responsible for burn wound infections, because
the wound created by CSR is essentially a par-
tial-thickness burn [12].
149
Factors that affect the rate of bacterial infection
One interesting factor regarding postoperative
infection is the time of presentation. Although most
infections occur in the first 3 to 10 days after surgery,
when the wound has not yet epithelialized [8,13],
some studies have shown that infections can occur
several weeks after surgery [14]. This occurrence can
be explained by the concept that there may be at least
two different mechanisms of infection. One is related
to the immediate loss of the protective immunity
supplied by the epidermal covering and the type of
initial postoperative management, and the other is
related to the long-term medicaments applied to the
surgical site.
Another factor that may affect the incidence and
treatment of bacterial infection is the change in
local immunity caused by previous operations on
the face. Repeated procedures, such as CSR or
chemical peeling, especially when there is diffuse
scarring, can lead to infections that may be refrac-
tory to the normal first-line antibiotics of choice.
Although most postoperative bacterial infections are
community acquired, patients who undergo surgery
in the hospital environment or who have had recent
hospitalizations must be considered at risk for
nosocomial infections, such as those caused by
methicillin-resistant S. aureus or Pseudomonas.
Finally, the role of the immunosuppressive effects
of short-term systemic or topical steroids in the
development of postoperative infections has yet to
be evaluated.
One of the most controversial areas in CSR is
the choice of open versus closed dressings for
postoperative management. The advantage of open
dressings (the use of topical creams or ointments to
maintain a moist environment through the course of
reepithelialization) is that the patient and doctor
have direct visual access to the wound, can perform
hygiene procedures often, and can recognize infec-
tions early. The advantages of closed dressings (the
use of occlusive and semiocclusive materials to
cover the wound) are that they eliminate the need
for the patient to be responsible for cleaning the
wounds in the immediate postoperative period and
they are associated with less discomfort. Researchers
have questioned whether the warm, moist, enclosed
environment created under such dressings engenders
the growth of infective organisms. Although some
authors contend that there is an increased incidence
of infection when occlusive dressings are used [15],
others claim that there is no difference [16]. It is
clear that if the dressings are used, use should be
limited to 48 hours, followed by either removal or
150 R.A. Strauss / Oral Maxillofacial Surg Clin N Am 15 (2003) 147–153
thorough wound cleansing and replacement [17].
Even in cases in which a higher incidence of
infection followed the use of closed dressings,
changing or eliminating the dressing early either
dramatically decreased or eliminated the problem
[16,17]. Whether real or not, the general feeling
that these dressings may increase the infection rates,
along with the tendency toward smaller unit CSR
(rather than full faces), has swung the pendulum for
many surgeons in favor of open dressings.
Treatment of bacterial infections
Once a patient’s infection by clinical examination
has been determined, it is imperative that the treat-
ment be immediate, aggressive, and appropriate. This
response is vital to limiting the chance of long-term
scarring or other sequelae. Empirical treatment
should be started, but culture and antibiotic sensitiv-
ity testing, a Tzanck stain, and a KOH preparation
should be performed as soon as possible to guide
therapy should the infection not respond to the initial
course of drugs. Because many CSR-related infec-
tions are mixed in nature (ie, they include either
multiple bacterial organisms or combined bacterial
and fungal or viral organisms), some authors recom-
mend beginning treatment with an antibiotic, an
antifungal, and an antiviral agent [8]. These choices
can be refined when the results of the cultures and
stains are obtained.
The choice of an empirical antibiotic always
should be based on the likely organisms involved
in the infection. Given the proclivity for staphylo-
coccal infection, a drug that is effective against
treating gram-positive cocci, such as a cephalo-
sporin, is logical, but because there is also a high
incidence of Pseudomonas and other gram-negative
infections, most authors recommend a broader spec-
trum agent, such as ciprofloxacin, 500 to 750 mg
twice daily for 7 days [8,9]. Topical antibiotics are
less effective and may alter the local flora, which
results in the production of resistant organisms. They
should not be used in place of systemic antibiotics
when any significant infection exists. Topical agents,
such as mupiricin ointment, may be added to sys-
temic treatment for local management of cutaneous
erosions, however. When considering therapy, one
should remember that patients admitted to the hos-
pital for intravenous antibiotics or hydration are at
greater risk for developing methicillin-resistant S.
aureus infections or infections caused by drug-re-
sistant strains of Pseudomonas, and repeated cultur-
ing is appropriate [18].
Treatment of viral infections
Among the infections seen after CSR, viral infec-
tions are probably the most common. This finding is
not unexpected given the prevalence of herpes sim-
plex virus-1 (HSV-1) in the population and the
proximity of the operative sites to the most common
place of hibernation for that organism. The virus lies
dormant and quiescent in the trigeminal ganglion but
may be activated at any time throughout life by a
chemical, mechanical, or traumatic stimulus or when
the immune system is compromised significantly.
Widespread herpetic infections, with subsequent
scarring of the face, have been well documented in
the past in burn patients and persons who undergo
dermabrasion and chemical peeling [19,20]. It is not
surprising that patients who undergo perioral CSR
are at high risk for an HSV reactivation. One also
should remember that although HSV-1 is the most
common cause of such infections, HSV-2 (typically
associated with genital lesions and transmitted by
sexual contact) also can be the causative agent in
some patients.
Herpes simplex virus infection is not an insignifi-
cant complication. The incidence of HSV reactivation
after dermabrasion and chemical peeling, without the
use of antiviral agents, has been reported to be as high
as 50% [21]. Before the introduction of effective and
specific antiviral agents, trigger areas such as the
perioral area were purposely avoided during chemical
peeling or dermabrasion to decrease the risk of an
HSV outbreak. Alternatively, patients with chronic or
recurrent HSV were considered unacceptable candi-
dates for these types of procedures [22].
Reactivation of the perioral HSV is manifested by
severe burning pain in the operated site that is
different from the usual stinging felt for the first
48 hours after surgery. Commonly the pain occurs
just after the cessation of antiviral therapy, but in
patients who do not receive prophylactic antibiotics,
it occurs within the first 7 to 10 days before reepi-
thelialization. Within 24 hours of the initial infection,
the patient develops a dermatomal erythema followed
by the typical vesicular lesions associated with HSV.
As the vesicles rupture, the area develops ulcerated,
weepy patches. The infection may be difficult to
diagnose because the entire surgical site is already
denuded by the CSR. An HSV infection should be
considered in any patient who develops fever or
malaise within the first week after surgery. Because
of the large amount of deepithelialized skin (loss of
the normal protective epidermal barrier), the infection
can spread rapidly to any or all of the treated areas,
with potentially disastrous consequences.
 R.A. Strauss / Oral Maxillofacial Surg Clin N Am 15 (2003) 147–153
Initially, a history of recurrent perioral HSV
infection was considered a true contraindication for
CSR. With the use of prophylactic antiviral therapy
this is no longer the case. Much controversy has been
generated in the literature about which drugs to
administer and what protocol to follow. Questions
regarding when to begin therapy, how much to give,
and when to cease therapy are still not clear and are
being investigated.
Because it was the first modern antiviral agent to
be commercially available, acyclovir has been used
extensively on patients who undergo CSR proce-
dures. Perkins and Sklarew showed that with typical
low-dose prophylaxis of 600 mg per day, break-
throughs were still common (8.3%, although not
nearly as common as the 50% reactivation rate they
found in untreated patients) [21]. Weinstein et al
showed that at 1000 mg per day the rate dropped to
0.3% [10]. When Perkins and Sklarew began their
prophylactic regimen 2 days before surgery at an
increased dose of 2400 mg per day and continued it
for 2 weeks after surgery, their reactivation rate
dropped to zero [20].
Although acyclovir has been shown to be effec-
tive, its low absorption and bioavailability necessi-
tate high dosages for long periods. Newer agents,
such as famcyclovir and valacyclovir, have much
better bioavailability and are probably better
choices [23,24]. Valacyclovir, 500 mg twice daily,
started the day of or the day before surgery and
continued for 14 days postoperatively has been
shown to be 100% effective in eliminating HSV
reactivation in CSR patients [25]. Although patients
with a history of recurrent HSV outbreaks or at
least a previous single outbreak have a documented
higher risk of infection if they do not receive
prophylactic antivirals, patients who deny such a
history should not be considered at low risk.
Several studies have demonstrated a high incidence
of apparent or proven HSV reactivation in patients
who denied any prior history of such lesions
[25,26].
When reactivation does occur, recognition is of
paramount importance. Tzanck stains or viral cul-
tures may be useful in making the diagnosis, but
given the potential for disastrous sequelae and the
low incidence of adverse reactions to the drugs,
the initiation of high-dose therapy should not be
delayed pending the results of these tests. Acyclovir
should be increased to 4000 mg per day in five
divided doses, and valacyclovir should be increased
to 1000 mg three times per day. This high-dose
therapy should be continued for at least 2 weeks
after the lesions disappear.
151
Treatment of candidal infections
Fungal infections also have been reported after
CSR [1,9,27]. The postoperative environmental
conditions of the face (ie, warm, moist, denuded
epithelium) provide an ideal growth medium for
fungal growth. To date, Candida organisms are
isolated almost exclusively. One should note, how-
ever, that several authors have found that candidal
infection is often associated with bacterial and viral
infections [8,9].
Candidal infection after CSR should be suspected
when the skin develops a fine, white film with an
undersurface that bleeds when wiped gently with
gauze. Pain, burning, and delayed healing are also
typical signs. In some cases, yeast pustules also may
form. When Candida is suspected, a KOH prepara-
tion and fungal cultures confirm the diagnosis.
The treatment for fungal infection is straightfor-
ward. The patient should be placed on systemic
antifungal therapy using 200 mg per day of either
fluconazole or itraconazole. A course of 2 to 4 days
seems to be adequate to resolve most fungal infec-
tions. Topical antifungal agents, such as ketoconazole
cream (Nizoral), may be used to complement the
systemic therapy. Some authors have suggested using
prophylactic fluconazole therapy for 1 day before
surgery in women who have a history of chronic
vaginal yeast infections [27]. At least one study has
suggested that prophylactic fluconazole therapy,
300 mg per day given between 3 and 8 days post-
operatively, may speed reepithelialization, although
this report remains controversial [28]. Manuskiatti et
al reported a 1.8% to 2.2% incidence of yeast
infection in 356 consecutively operated patients, but
this rate dropped to zero with the use of prophylactic
antifungal therapy [6]. Each clinician must determine
whether this low incidence of infection warrants the
use of prophylactic therapy.
Prevention of infections
Because of the grave sequelae that can accompany
infections after CSR, much emphasis has been placed
on prevention protocols and treatment protocols [29].
This has proved to be one of the most controversial
issues in the still-evolving field of facial rejuvenation.
The wisdom of prophylactic antibiotic, antiviral, and
antifungal agents, the use of closed versus open
dressings, and various postoperative wound hygiene
techniques have come under scrutiny. The use and
type of prophylactic antibiotics, in particular, have
become sources of discussion.
152 R.A. Strauss / Oral Maxillofacial Surg Clin N Am 15 (2003) 147–153
Prophylactic antibiotics are considered appropriate
to prevent infection either when the risk is exception-
ally high or when the ramifications of the infection are
grave. The benefits of the therapy must outweigh the
risks of adverse effects. CSR is considered a ‘‘clean-
contaminated’’ procedure because of the intimate
proximity of the surgical site to contaminated areas,
such as the oral and nasal cavities. In general, it is
anticipated that ‘‘clean-contaminated’’ surgery has an
expected infection rate of approximately 10%. The
actual infection rate of CSR procedures reported in the
literature ranges from 1% to 9%. With improved
posttreatment protocols currently being used, the
current rate is likely at the low end of this range.
Despite the low incidence, many surgeons still pro-
vide prophylactic coverage to their patients [30,31].
Although at least one study has shown that using
antibiotics can lead to a decreased incidence of
infection (decreased from 8.2% to 4.3% using cipro-
floxacin and open dressings) [6], several other studies
that showed similar results either were based on small
samples [13] or involved retrospective studies with
multiple variables (such as the type of postoperative
dressing applied) that could have affected the infec-
tion rate [8]. If one decides to use prophylactic anti-
biotics, it is wise to use narrow-spectrum agents.
The reasons for not giving prophylactic antibiotics
are also compelling, including the cost of the drugs,
possible allergic or other untoward reactions, altera-
tion of the normal skin flora, and selection of resistant
organisms or development of superinfection. Several
studies also have shown no decrease in infection rate
with the use of prophylactic therapy [9,32,33].
The use of topical antibiotics also has been
evaluated as a means of postoperative management
and for prophylaxis. Postoperative use of antibiotic
ointments has been associated with the development
of milia, selection of resistant organisms, and a
significant risk of contact dermatitis, especially in
deepithelialized, laser-abraded skin [13,34]. The use
of antibiotics is probably best avoided. Mupirocin, an
intranasal, topical, antistaphylococcal spray, also has
been tested for prophylaxis without success [6].
Although the evidence seems to indicate that use
of prophylactic antibiotics after CSR probably is not
necessary, the use of closed dressings may alter this
paradigm. In the opinion of many authors (even
authors opposed to antibiotic use with open dress-
ings), the surface environment associated with closed
dressings increases the risk of infection sufficiently to
warrant the use of prophylactic agents [14]. Although
the use of prophylactic agents to prevent bacterial or
fungal infection has generated some controversy,
there is no confusion as to the necessity for preoper-
ative and postoperative antiviral therapy. This neces-
sity is well documented and well accepted.
One area of interest that is often overlooked but
must be considered in any discussion of postoperative
infection is wound hygiene. Changing or eliminating
closed dressings early, meticulous wound cleansing,
and strict hand washing are imperative. The patient
also should be instructed carefully on the avoidance
of ‘‘double dipping’’—using the fingers or the hands
to apply postoperative topical agents and then placing
them back into the container for more material. This
practice quickly contaminates the container and leads
to re-inoculation. Contamination can be avoided by
using fresh cotton swabs and removing the entire
amount of agent needed at one time.
Summary
Infections after CSR are to be expected in a small
number of patients, and they can be bacterial, viral, or
fungal. Despite the low incidence, these infections
can have devastating sequelae. When they occur, they
should be investigated promptly and thoroughly to
determine their cause and the best course of treat-
ment. The use of prophylactic antibiotics and anti-
fungal agents is still controversial, and their use
seems in many cases to be dictated primarily by legal
and patient demands rather than good science. It is
hoped that future studies will better elucidate when
and if these agents are necessary and what drugs
should be used. Conversely, antiviral therapy is
considered vital in the prevention of herpetic out-
breaks, and it has been well documented that failure
to use such agents leads to an unacceptably high
incidence of infection and subsequent scarring, even
in patients with no previous history of HSV. Finally,
there is still much to be learned regarding the most
ideal posttreatment protocols to minimize the inci-
dence of infection. Proponents of closed and open
dressings must demonstrate convincing evidence to
show that one regimen is better than the other.
References
[1] Fitzpatrick RE. Laser resurfacing of rhytides. Dermatol
Clin 1997;15:431 – 47.
[2] Gilchrest BA. Skin aging and photoaging. Dermatol
Nurs 1990;2:79 – 84.
[3] Gallis HA. Normal flora and opportunistic infections.
In: Joklik WK, Willet HP, Amos DB, et al, editors.
Zinsser microbiology. 19th edition. Norwalk (CT):
Appleton and Lange; 1988. p. 337 – 42.
 R.A. Strauss / Oral Maxillofacial Surg Clin N Am 15 (2003) 147–153
[4] Schuster GS. Microbiology of the orofacial region. In:
Topazian RG, Goldberg MH, Hupp JR, editors. Oral
and maxillofacial infections. 4th edition. Philadelphia:
W.B. Saunders Co.; 2002. p. 30 – 42.
[5] Gaspar Z, Vincuillo C, Elliot T. Antibiotic prophylaxis
for full face laser resurfacing: is it necessary? Arch
Dermatol 2001;137:1251 – 2.
[6] Manuskiatti W, Fitzpatrick RE, Goldman MP, et al.
Prophylactic antibiotics in patients undergoing laser
resurfacing of the skin. J Am Acad Dermatol 1999;
40:77 – 84.
[7] Nanni CA, Alster TS. Complications of carbon dioxide
laser resurfacing: an evaluation of 500 patients. Der-
matol Surg 1998;24:315 – 20.
[8] Sriprachya-Anunt S, Fitzpatrick RE, Goldman MP, et
al. Infections complicating pulsed carbon dioxide laser
resurfacing for photoaged facial skin. Dermatol Surg
1997;23:527 – 35.
[9] Walia S, Alster TS. Cutaneous CO2 laser resurfacing
infection rate with and without prophylactic antibiot-
ics. Dermatol Surg 1999;25:857 – 61.
[10] Weinstein C, Pozner JN, Ramirez OM. Complications
of carbon dioxide laser resurfacing and their preven-
tion. Aesthetic Surgery Journal 1997;17:216 – 22.
[11] Futoryan T, Grande D. Postoperative wound infection
rates in dermatologic surgery. Dermatol Surg 1995;21:
509 – 14.
[12] Husain MT, Karim QN, Tajuri S. Analysis of infection
in a burn wound. Burns 1989;15:299 – 302.
[13] Ross EV, Amesbury EC, Barile A, et al. Incidence of
postoperative infection or positive culture after facial
laser resurfacing: a pilot study, a case report, and a
proposal for rational approach to antibiotic prophy-
laxis. J Am Acad Dermatol 1998;39:975 – 81.
[14] Christian MM, Behroozan DS, Moy RL. Delayed in-
fections following full-face CO2 laser resurfacing and
occlusive dressing use. Dermatol Surg 2000;26:32 – 6.
[15] Lawrence N, Coleman WP. Postoperative care after
laser resurfacing. In: Coleman WP, Lawrence N, edi-
tors. Skin resurfacing. Baltimore: Williams and Wil-
kins; 1998. p. 189 – 95.
[16] Newman JP, Fitzgerald P, Kock RJ. Review of closed
dressing after laser resurfacing. Dermatol Surg 2000;26:
562 – 71.
[17] Goldman MP. Pre- and postoperative care of the laser
resurfacing patient. International Journal of Aesthetic
and Restorative Surgery 1997;5: 46 – 9.
[18] Bellman B, Brandt FS, Holtmann M, et al. Infection
with methicillin-resistant Staphylococcus aureus after
carbon dioxide resurfacing of the face: successful treat-
ment with minocycline, rifampin, and mucipirin oint-
ment. Dermatol Surg 1998;24:279 – 82.
[19] Kagan RJ, Naraqi S, Matsuda T, et al. Herpes simplex
153
virus and cytomegalovirus infections in burned pa-
tients. J Trauma 1985;25:40 – 5.
[20] Rappaport MJ, Kramer F. Exacerbation of facial herpes
simplex after phenol facial peels. J Dermatol Surg On-
col 1984;10:57 – 8.
[21] Perkins SW, Sklarew EC. Prevention of facial herpetic
infections after chemical peel and dermabrasion: new
treatment strategies in the prophylaxis of patients
undergoing procedures of the perioral area. Plast Re-
constr Surg 1996;98:427 – 33.
[22] Farber GA. Chemical peeling in the treatment of cer-
tain cosmetic defects and diseases of the skin. In:
Burks JW, editor. Dermabrasion and chemical peeling.
Springfield (IL): Charles Thomas Publishing; 1979.
p. 209 – 26.
[23] Alster TS, Nanni CA. Famciclovir prophylaxis of her-
pes simplex virus reactivation after laser skin resurfac-
ing. Dermatol Surg 1999;25:242 – 6.
[24] Wall SH, Ramey SJ, Wall F. Famciclovir as antiviral
prophylaxis in laser resurfacing procedures. Plast Re-
constr Surg 1999;104:1103 – 8.
[25] Gilbert S, McBurney E. Use of valacyclovir for herpes
simplex virus-1 (HSV-1) prophylaxis after facial resur-
facing: a randomized clinical trial of dosing regimens.
Dermatol Surg 2000;1:50 – 4.
[26] Rendon-Pellerano MI, Lentini J, Eaglestein WE, et al.
Laser resurfacing: usual and unusual complications.
Dermatol Surg 1999;5:360 – 6.
[27] Kilmer SL. Laser resurfacing complications: how to
treat them and avoid them. International Journal of
Aesthetic and Restorative Surgery 1997;5:41 – 5.
[28] Conn H, Nanda VS. Prophylactic fluconazole pro-
motes reepithelialization in full-face carbon dioxide
laser skin resurfacing. Lasers Surg Med 2000;26:
201 – 7.
[29] Strauss RA, McMunn W, Gregory BA. Cosmetic skin
resurfacing. Selected Readings in Oral and Maxillofa-
cial Surgery 2001;9:1 – 32.
[30] Friedman PM, Geronemus RG. Antibiotic prophylaxes
in laser resurfacing patients. Dermatol Surg 2000;26:
695 – 6.
[31] Kauvar ANB. Laser skin resurfacing: perspectives at
the millennium. Dermatol Surg 2000;26:174 – 7.
[32] Alster TS. Against antibiotic prophylaxis for cutaneous
laser resurfacing. Dermatol Surg 2000;26:697 – 8.
[33] Gaspar Z, Apfelberg DB. Summary of the 1997
ASAPRS/ASPRS laser task force survey on laser re-
surfacing and laser blepharoplasty. Plast Reconstr Surg
1998;101:511 – 8.
[34] Waldorf HA, Kauvar ANB, Grossman MC, et al. Skin
resurfacing of fine to deep rhytids using a char free
carbon dioxide laser in 47 patients. Dermatol Surg
1995;21:940 – 6.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 155 – 160
The use of prophylactic antibiotics for the prevention of
postoperative infections
Daniel M. Laskin, DDS, MS
Department of Oral and Maxillofacial Surgery School of Dentistry, Virginia Commonwealth University,
Richmond, VA 23298-0566, USA
Despite use of the best surgical techniques, some
operations still carry a high risk of wound infection.
Basic and clinical studies have shown that this risk
can be reduced by the administration of prophy-
lactic antibiotics. There are certain inherent risks
associated with the use of these agents, however,
such as toxic and allergic reactions, emergence of
resistant bacteria, drug interactions, and superinfec-
tions. Moreover, prophylactic antibiotics do not
prevent all postoperative infections. For these reas-
ons, their use should be based on an understanding
of certain basic principles. This article reviews these
principles and discusses their application in specific
clinical situations.
Principles of prophylactic antibiotics
Antibiotic prophylaxis has been defined as the
preoperative use of antibiotics to prevent infection.
Prophylactic use of antibiotics is in contradistinction
to the therapeutic use of antibiotics, which are given
to treat an already existing infection. Not only are the
purposes different but also the manner in which the
drugs should be administered for an optimum effect
are different. A review of the literature clearly indi-
cates that, in many instances, antibiotics are given
postoperatively rather than preoperatively, although it
is still referred to as a prophylactic procedure. This
difference must be taken into consideration when
evaluating the conclusions that are reached and the
E-mail address: dmlaskin@vcu.edu
1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 2 - 3 6 9 9 ( 0 2 ) 0 0 0 6 7 - 5
recommendations that are made by the authors. It also
may account for some of the contradictory findings
that have been reported in the literature.
As the result of the basic studies of Miles and
Burke et al [1,2] and several well-controlled clinical
studies, the principles of proper antibiotic prophylaxis
have been established for general surgery, and they
are applicable to the field of oral and maxillofacial
surgery. These principles are (1) that the intended
procedure must carry a significant risk of postoper-
ative infection, (2) that the correct antibiotic must be
selected, and (3) that the antibiotic is administered
properly. To these principles one may add a fourth
principle: not to rely solely on prophylactic anti-
biotics to prevent postoperative infections.
Risk of infection
Although such factors as age of the patient,
nutritional status, nature of any underlying disease,
presence of necrotic tissue, and a decreased blood
supply can increase the potential for postoperative
infection, perhaps the most important factor is wound
contamination. Altemeier et al [3] developed a clas-
sification of general surgical wounds relating con-
tamination to the risk of infection. In this
classification, operative wounds are categorized as
clean, clean-contaminated, contaminated, or dirty,
with contamination resulting from planned or
unplanned entry into the respiratory, gastrointestinal,
or genitourinary tracts. Because of the difference in
host response to such contamination and oral con-
tamination, this classification cannot be used for
intraoral wounds. It can be modified, however, to
156 D.M. Laskin / Oral Maxillofacial Surg Clin N Am 15 (2003) 155–160
establish the indications for prophylactic antibiotics
in other forms of oral and maxillofacial surgery.
According to this modified classification, extraoral
wounds are still categorized in terms of increasing risk
of infection as clean, clean-contaminated, contami-
nated, and dirty. However, the definitions must be
altered to fit the regional differences. Clean surgical
wounds, which have a low infection rate, usually
involve no significant tissue trauma or inflammation,
the incision is closed primarily, the wound is not
drained, and there is no communication with the oral
cavity. Clean-contaminated surgical wounds are sim-
ilar to clean wounds except that there is a greater risk of
infection because communication with the oral cavity
occurs. Contaminated wounds are fresh traumatic
injuries that involve the oral cavity, whereas dirty
wounds are traumatic injuries with delayed treatment
that communicate with the oral cavity and contain
devitalized tissue or foreign bodies.
Clean surgical wounds do not require antibiotic
prophylaxis; however, it should be used in clean-
contaminated wounds. Contaminated wounds usually
can be managed with preoperative prophylactic anti-
biotics if there are no other significant risk factors.
Otherwise, patients with such wounds also should
receive postoperative prophylactic antibiotics. Patients
with dirty wounds, which are already infected, require
preoperative and postoperative therapeutic antibiotics.
The decision not to use antibiotics or to use only
preoperative prophylactic antibiotics is always based
on the assumption that the patient does not have any
significant medical risk factors that could affect the
body’s humoral and cellular defense mechanisms.
These risk factors include poorly controlled diabetes,
end-state renal disease and uremia, severe alcoholism,
an immunocompromising disease (eg, leukemia, lym-
phoma, or an advanced malignant neoplasm), and the
use of chemotherapeutic agents or other immunosup-
pressive drugs. In such patients, one is concerned not
only with preventing the initial contamination of the
wound but also with the possibility of recontamina-
tion during the postoperative period. Therefore, post-
operative prophylaxis is also indicated.
Antibiotic selection
Once it has been determined that prophylactic
antibiotics are indicated in a particular patient, the
second principle involves making the appropriate
selection. The choice of antibiotic is determined
largely by its effectiveness against the pathogens
commonly encountered at the specific surgical site.
It need not be able to eradicate every potential
pathogen that may be encountered. Expanded anti-
biotic coverage serves only to increase the rate of
bacterial resistance and the development of super-
infections. The most commonly encountered orga-
nisms from oral contamination are the streptococci,
anaerobic gram-positive cocci, and anaerobic gram-
negative rods, whereas contamination from the sinus
and nose may include Haemophilus influenzae, dip-
theroids, and peptostreptococci [4]. When the skin is
involved, the presence of Staphylococcus aureus and
epidermidis also must be considered.
Although the first consideration is that the anti-
biotic selected is effective against the major contam-
inating organisms, it also must be nontoxic and
relatively nonallergenic, bactericidal, capable of
achieving therapeutic tissue concentrations, and have
a long half-life so that redosing is generally not
required during the procedure. The antibiotic that
currently still fulfills these requirements best is peni-
cillin. When skin contaminants are a concern, how-
ever, a first-generation cephalosporin, such as
cefazolin, can be used because of its effectiveness
against most staphylococci.
Antibiotic administration
Although one selects the appropriate antibiotic for
prophylaxis, its ultimate effectiveness in preventing
infection still depends on proper administration. This
decision varies according to whether one is dealing
with normal patients, in whom there is only concern
about the initial bacterial contamination that can
occur during the operation, or patients with other
significant risk factors that decrease the body’s
defenses and retard healing, which makes them more
susceptible to recontamination during the healing
period. In the first instance, the antibiotic must be
administered intravenously or intramuscularly within
30 minutes of the incision time at twice the thera-
peutic dose [5]. For penicillin, the dose is 2 million
U, and for cefazolin it is 1 g. This dose generally
provides adequate coverage for up to 4 hours. If the
procedure is prolonged, however, it is advisable to
administer additional doses every 4 hours until the
operation is completed. In patients who are medically
compromised or immunosuppressed, it may be advis-
able to continue prophylactic oral antibiotics until
biologic sealing of the wound has occurred. No data
exist on how long this period should be or even if it
prevents postoperative infections. In fact, one must be
concerned about promoting the emergence of resist-
ant strains and causing superinfections. In a wound
that is healing by primary intention, the period should
 D.M. Laskin / Oral Maxillofacial Surg Clin N Am 15 (2003) 155–160
probably be no longer than 48 hours postoperatively
and in a wound healing by secondary intention it
should be no longer than 3 to 5 days, at which time a
biologic seal should have occurred.
Adjunctive procedures
Although the correct use of prophylactic anti-
biotics significantly reduces the incidence of post-
operative infections, their occurrence can be reduced
further by careful attention to proper surgical tech-
nique. Proper technique involves adequate cleansing
of the surgical site, strict adherence to sterile tech-
nique, avoiding tissue trauma, and minimizing oper-
ating time. In general, the risk of infection has been
shown to increase with each hour of surgery. Shav-
ing the skin in the surgical site with a razor the
evening before the operation also has been shown to
increase the infection rate, probably because of
bacterial proliferation in the areas of minor trauma
produced by shaving [6]. When shaving is neces-
sary, it should be done just before preparation of the
surgical site.
The use of drains also may contribute to post-
operative infections. When drains are necessary in
noninfected wounds, the closed-suction type is prefer-
able to open drains. They should not be placed through
the operative incision and should be removed as soon
as possible. Paying proper attention to these details
during management of the surgical patient greatly
improves the effectiveness of antibiotic prophylaxis.
Specific applications in oral and
maxillofacial surgery (Table 1)
Exodontia and dentoalveolar surgery
Although intraoral surgical wounds are contami-
nated by the oral flora, the ability of patients
normally to tolerate this bacterial population and
the excellent blood supply to the oral tissues allow
such wounds to be managed similarly to clean
wounds. It is unnecessary to use prophylactic anti-
biotics to prevent infection when performing most
types of exodontia and dentoalveolar surgery unless
there are other contributing risk factors (Table 1). If
the procedure involves the maxillary sinus or nasal
cavity, however, this can result in cross-contamina-
tion with new organisms, and prophylactic antibi-
otics should be used except when there is already
infection in these areas. In the latter instance, thera-
peutic antibiotics are indicated.
157
Impacted third molar surgery
Although many practitioners routinely prescribe
antibiotics for patients who have impacted third
molars removed [7,8], they are generally prescribed
postoperatively. This practice does not accomplish
the true purpose of using prophylactic antibiotics:
having a high tissue concentration present at the
time of surgery when the wound is exposed to
bacterial contamination. It is not surprising that no
convincing data show a significant reduction in
postoperative infections after third molar removal
when antibiotics are used in this manner [7,9]. The
fact that such infections are relatively uncommon
also may be a contributing factor to such findings
[10,11]. Most studies that report high postoperative
infection rates [12] include alveolar osteitis (dry
socket) and fascial space infections. In such stud-
ies, however, there is still not a significant reduc-
tion in either type of problem, whether the
antibiotics are given postoperatively or preopera-
tively [13 – 15].
Although isolated cases of severe fascial space
infections after third molar removal have been
reported [16], the incidence is too low to justify the
routine use of prophylactic antibiotics. Such treat-
ment should be reserved only for patients with
significant medical risk factors for infection [1,17].
In these cases it should be given immediately before
surgery and for 3 to 5 days after surgery to provide an
adequate period of coverage.
Dental implants
Few studies have been conducted on the effect of
antibiotics on the infection rate after implant place-
ment. Two studies in which no control group was
used [18,19] concluded that preoperative prophylactic
antibiotics were effective in preventing postoperative
infections. In a subsequent, better controlled study
[20], however, no difference in postoperative infec-
tions or implant failure was found between the
two groups.
In a more recent, large, multicenter study, Dent et
al [21] analyzed implant failures (which probably
included some cases with infection) and found sig-
nificantly fewer failures up to stage 2 surgery, when
high-dose preoperative antibiotics (penicillin in 69%
of cases) were administered. These findings were
confirmed in a follow-up study at 36 months in the
same patient population (4.6% versus 10% failure)
[22]. Thus, there seems to be a benefit in using
prophylactic antibiotics in dental implant patients.
158 D.M. Laskin / Oral Maxillofacial Surg Clin N Am 15 (2003) 155–160
Table 1
Indications for prophylactic antibiotics
Procedure
Exodontic and dentoalveolar surgery
Impacted third molar surgery
Dental implants
Orthognathic surgery (extraoral approach)
Orthognathic surgery (intraoral approach)
Mandibular fractures (no oral communication)
Mandibular fractures (oral communication)
Facial bone fractures
Soft tissue trauma (oral injuries)
Soft tissue trauma (clean extraoral lacerations)
Soft tissue trauma (blunt trauma, gunshot wounds,
bites, orocutaneous communication)
Major head and neck surgery
a
High risk for infection refers to such factors as poor nutritional status, complicating medical problems, presence of necrotic
tissue or foreign bodies, and decreased blood supply to the region.
Orthognathic surgery
Orthognathic surgery performed via an extraoral
approach is considered a clean procedure and pro-
phylactic antibiotics should not be necessary unless
communication with the mouth is anticipated [23].
Intraoral procedures and procedures that involve the
maxillary sinus and nasal passages are clean-contam-
inated operations, and short-term prophylactic anti-
biotics have been shown to reduce the postoperative
infection rate [24,25]. There seems to be no advan-
tage in prolonged postoperative antibiotic administra-
tion [25,26]. In one study in which a 5-day regimen
was shown to be better than a 1-day regimen [27], as
pointed out by Abubaker [24], the difference was
caused by the difference in the criteria used to
establish wound infection.
Mandibular fractures
Patients with condylar process fractures treated
by either open or closed reduction require no pro-
phylactic antibiotics. The same is true for fractures in
other non – tooth-bearing areas that are not in com-
munication with the mouth, because these are all
clean wounds. In patients with compound man-
dibular fractures, however, which are contaminated
wounds, studies have shown that the use of anti-
Antibiotic regimen Exceptionsa
None High risk for infection; communication
with sinus or oral cavity
None High risk for infection
Preoperative Also postoperative prophylaxis when
high risk of infection
None Preoperative when anticipate possible
oral contamination
Preoperative and None
1-day postoperative
None None
Preoperative and Use prophylactic antibiotics 3 – 5 days
12 hours postoperative postoperatively when treatment delayed
Preoperative None
None Preoperative when high risk for infection
None Preoperative when high risk for infection
Preoperative Postoperative prophylaxis for 3 – 5 days
when high risk for infection
Preoperative Postoperative prophylaxis 3 – 5 days when
packs or drains used; lack of watertight
closure; high risk for infection
biotics is effective in reducing postoperative infec-
tions [28,29]. In most of these studies, however, the
antibiotics were given not only preoperatively but
also for a long period postoperatively. More recent
investigations [28,30] have shown that prophylactic
antibiotics given preoperatively and for no longer
than 12 hours postoperatively are just as effective as
long-term use in preventing postoperative infections.
These findings apply only to fractures that are treated
shortly after the injury has occurred. Fractures for
which there is delayed treatment should be consid-
ered dirty wounds, and such patients should receive
therapeutic antibiotics postoperatively.
Facial bone fractures
Although it has been suggested that any mid-
facial fracture compounded into the mouth, nose,
or paranasal sinuses requires antibiotic coverage
[31,32], other studies [28,33] have shown that it
may not be necessary. Because one of these studies
[33] was not well controlled and the other [28] had a
relatively small number of cases, the issue remains
unresolved. Based on the fact that such compound
fractures communicate with a contaminated cavity,
they should be considered as clean-contaminated
wounds and preoperative prophylactic antibiotics
should be used.
 D.M. Laskin / Oral Maxillofacial Surg Clin N Am 15 (2003) 155–160
Soft tissue trauma
Patients with traumatic injuries that involve the
oral mucosa, gingiva, or tongue do not require
prophylactic antibiotics because such wounds,
although contaminated, generally heal without infec-
tion. Simple extraoral lacerations from relatively
clean objects that are closed within 4 hours also have
a low infection rate and do not require prophylactic
antibiotics [34]. Extraoral soft tissue injuries, such as
those caused by blunt trauma, gunshot wounds, and
bites and injuries that involve orocutaneous commu-
nication, fall into either the category of clean-con-
taminated or contaminated wounds, and the patient
should receive pretreatment antibiotic prophylaxis.
If the wounds are extremely dirty, the patient also
should receive postoperative therapeutic antibiotics.
Major head and neck surgery
Researchers generally agree that patients who
undergo major surgical procedures in the head and
neck region, such as oncologic and reconstructive
surgery, should receive preoperative prophylactic
antibiotics [31,35 – 39]. There is a question regarding
how long they should be used postoperatively. Sev-
eral studies [31,35,36,38,39] that have shown that
there is no advantage in extending the prophylactic
antibiotics beyond 1 day after surgery in such cases
unless there are packs or drains in the wound or it is
not possible to obtain a watertight closure and there is
prolonged leakage of saliva into the wound [40].
Summary
Although prophylactic antibiotics do not prevent
all postoperative infections, they can reduce the
incidence significantly when administered correctly.
However, they should be used only in patients in
whom the surgical procedure or the medical condition
puts them at a high risk of developing such infec-
tions. As a general rule, the anticipated risk should
exceed 10% [41]. Use of antibiotics in low-risk cases
in an attempt to prevent postoperative infections,
especially when used for prolonged periods, can
result in adverse drug effects, superinfections, and
the emergence of resistant strains without providing
any significant benefits. As Furstenburg [42] stated so
succinctly more than 50 years ago: ‘‘With the dis-
covery of penicillin, its allied agents, and their mass
production, clinicians have employed them in differ-
ent forms and by various methods for almost every
illness in the category of medicine. The urge to
159
administer appears to transcend therapeutic rationale
and provokes the use of antibiotics for purposes often
obscure and irrelevant. It is only natural that great
inventions and discoveries should gain wide acclaim
and enthusiastic endorsement and enjoy universal
acceptance far in excess of which they merit.’’
References
[1] Miles AA, Miles EM, Burke J. The value and duration
of defense reaction of the skin to primary lodgment
of bacteria. British Journal of Experimental Pathology
1957;38:79 – 86.
[2] Burke JF. The effective period of preventive antibiotic
action in experimental incisions and dermal lesions.
Surgery 1961;50:161 – 8.
[3] Altemeier WA, Burke JF, Pruitt Jr BA, et al. Manual on
control of infection in surgical patients. 2nd edition.
Philadelphia: JB Lippincott; 1984. p. 26 – 9.
[4] Gagliano NC, Leisure GS. Perioperative pharmacol-
ogy. In: Stone OJ, Bogdonoff DL, Leisure GS, et al,
editors. Perioperative care: anesthesia, medicine, and
surgery. St. Louis: Mosby; 1998. p. 486.
[5] Peterson LJ. Principles of antibiotic therapy. In: Topa-
zian RG, Goldberg MH, editors. Oral and maxillofacial
infections. 3rd edition. Philadelphia: WB Saunders;
1994. p. 160 – 97.
[6] Cameron JL. Current surgical therapy. 7th edition.
St. Louis: Mosby; 2001. p. 1281.
[7] Capuzzi P, Montebugnoli L, Vaccaro MA. Extraction
of impacted third molars: a longitudinal, prospective
study. Oral Surg Oral Med Oral Pathol 1994;77:
341 – 3.
[8] Piecuch JF, Arzadon J, Lieblich SE. Prophylactic anti-
biotics for third molar surgery: a supportive opinion.
J Oral Maxillofac Surg 1995;53:53 – 60.
[9] Chiapasco M, Cicco LD, Marrone G. Side effects and
complications associated with third molar surgery. Oral
Surg Oral Med Oral Pathol 1993;76:312 – 20.
[10] Goldberg MA, Nemarich AN, Marco WP. Complica-
tions after mandibular third molar surgery: a statistical
analysis of 500 consecutive procedures in private prac-
tice. J Am Dent Assoc 1985;111:277 – 9.
[11] Osborn TP, Frederickson G, Small IA, et al. A prospec-
tive study of complications related to mandibular third
molar surgery. J Oral Maxillofac Surg 1985;43:767 – 9.
[12] Mitchell DA. A controlled clinical trial of prophylactic
tinidazole for chemoprophylaxis in third molar surgery.
Br Dent J 1986;160:284 – 6.
[13] Curran JB, Kennett S, Young AR. An assessment of
the use of prophylactic antibiotics in third molar sur-
gery. Int J Oral Surg 1974;3:1 – 6.
[14] Happonen R-P, Backstrom A-C, Ylipaavalniemi P.
Prophylactic use of phenoxy-methylpenicillin and tini-
dazole in mandibular third molar surgery: a com-
parative placebo controlled clinical trial. Br J Oral
Maxillofac Surg 1990;28:12 – 5.
160 D.M. Laskin / Oral Maxillofacial Surg Clin N Am 15 (2003) 155–160
[15] Thomas DW, Hill CM. An audit of antibiotic prescrib-
ing in third molar surgery. Br J Oral Maxillofac Surg
1997;35:126 – 8.
[16] Indresano AT, Haug RH, Hoffman MJ. The third molar
as a cause of deep space infections. J Oral Maxillofac
Surg 1992;50:33 – 5.
[17] MacGregor AJ. Reduction in morbidity in the surgery
of the third molar removal. Dent Update 1990;17:
411 – 4.
[18] Larsen P, McGlumphy E. Antibiotic prophylaxis for
placement of dental implants. J Oral Maxillofac Surg
1993;51:194.
[19] Peterson L, McGlumphy E, Halikas L. Long-term anti-
biotic prophylaxis is not necessary for placement of
dental implants. J Oral Maxillofac Surg 1996;54:76.
[20] Gynther G, Kondell P, Moberg L, et al. Dental implant
installation without antibiotic prophylaxis. Oral Surg
Oral Med Oral Pathol Oral Radiol Endod 1998;85:
509 – 22.
[21] Dent CD, Olson JW, Farish SE, et al. Influence of
preoperative antibiotics on success of endosseous im-
plants up to and including Stage 2 surgery: a study
of 2,641 implants. J Oral Maxillofac Surg 1997;55:
19 – 24.
[22] Laskin DM, Dent CD, Morris HF, et al. The influence
of preoperative antibiotics on success of endosseous
implants at 36 months. Ann Periodontol 2000;5:
166 – 74.
[23] Zallen RD, Strader RJ. The use of prophylactic anti-
biotics in extraoral procedures for mandibular prog-
nathism. J Oral Surg 1971;29:178 – 9.
[24] Abubaker AO. Antibiotic prophylaxis in orthognathic
surgery: a 1-day versus 5-day regimen [discussion].
J Oral Maxillofac Surg 1999;57:230 – 2.
[25] Ruggles JE, Hann JR. Antibiotic prophylaxis in intrao-
ral orthognathic surgery. J Oral Maxillofac Surg
1984;42:797 – 801.
[26] Fridrich KL, Partnoy BE, Zeitler DL. Prospective anal-
ysis of antibiotic prophylaxis for orthognathic surgery.
Int J Adult Orthod Orthognath Surg 1994;9:129 – 31.
[27] Bentley KC, Head TW, Aiello GA. Antibiotic prophy-
laxis in orthognathic surgery: a 1-day versus 5-day regi-
men. J Oral Maxillofac Surg 1999;57:226 – 30.
[28] Chole RA, Yee J. Antibiotic prophylaxis for facial
fractures: a prospective, randomized clinical trial. Arch
Otolaryngol Head Neck Surg 1987;113:1055 – 6.
[29] Zallen RD, Curry JT. A study of antibiotic usage in
compound mandibular fractures. J Oral Surg 1975;33:
431 – 4.
[30] Abubaker AO, Rollert MA. Postoperative antibiotic
prophylaxis in mandibular fractures: a preliminary
randomized, double-blind, and placebo-controlled clin-
ical study. J Oral Maxillofac Surg 2001;59:1415 – 9.
[31] Eschelman LT, Schleuning AJ, Brummett RE. Prophy-
lactic antibiotics in otolaryngologic surgery: a double-
blind study. Transactions of American Academy Op-
thalmology and Otolaryngology 1971;75:387 – 94.
[32] Zallen RD, Black SL. Antibiotic therapy in oral and
maxillofacial surgery. J Oral Surg 1976;34:349 – 51.
[33] Paterson JA, Cordo Jr VA, Stratigas GT. An examina-
tion of antibiotic prophylaxis in oral and maxillofacial
surgery. J Oral Maxillofac Surg 1970;28:753 – 9.
[34] Dellinger ED. Antibiotic use. In: Trunkey DD, Lewis
Jr FR, editors. Current therapy of trauma. 3rd edition.
St. Louis: Mosby; 1991. p. 47.
[35] Conover M, Kaban LB, Mulliken JB. Antibiotic pro-
phylaxis for major maxillocraniofacial surgery. J Oral
Maxillofac Surg 1985;43:865 – 70.
[36] Fee Jr WE, Glenn M, Handen C, et al. One day versus
2 days of prophylactic antibiotics in patients under-
going major head and neck surgery. Laryngoscope
1984;94:612 – 4.
[37] Johnson JT, Myers EN, Thearle PB, et al. Antibiotic
prophylaxis for contaminated head and neck surgery.
Laryngoscope 1984;94:46 – 51.
[38] Johnson JT, Schuller DE, Silver E, et al. Antibiotic
prophylaxis in high-risk head and neck surgery: one-
day vs five-day therapy. Otolaryngol Head Neck Surg
1986;95:554 – 7.
[39] Mombelli G, Coppens I, Dor P, et al. Antibiotic pro-
phylaxis in surgery for head and neck cancer: com-
parative study of short and prolonged administration
of carbenicillin. J Antimicrob Chemother 1987;7:
667 – 71.
[40] Shapiro M. Prophylaxis in otolaryngologic surgery
and neurosurgery: a critical review. Rev Infect Dis
1991;13(Suppl 10):S858 – 68.
[41] Burkett HG, Quick RG, Gott DT. Essential surgery:
problems, diagnosis and management. 2nd edition. Ed-
inburgh: Churchill-Livingstone; 1998. p. 70.
[42] Furstenburg AC. Antibiotics in the treatment of disease
of the ear, nose, and throat. Ann Otol 1949;58:5 – 17.
 Oral Maxillofacial Surg Clin N Am 15 (2003) 161 – 165
Index
Note: Page numbers of article titles are in boldface type.
A
Abscesses, retropharyngeal, versus cellulitis, imaging
of, 47
Actinomycosis, cervicofacial. See Cervicofacial acti-
nomycosis.
Acyclovir, for viral infections, due to laser-assisted
skin resurfacing, 155
AIDS. See also HIV infection; Immunocompro-
mised patients.
and oropharyngeal candidiasis, 93 – 94
immune defects in, 107
Airway management, in head and neck
infections, 107
Alcoholism, immune defects in, 106
Aminoglycosides, for head and neck infections, 26
Amoxicillin, for head and neck infections, 23
Amphotericin B, for oropharyngeal candidiasis,
97 – 99
Antibiotics, for bite wounds, 146 – 147
for head and neck infections. See Head and
neck infections.
for infections due to laser-assisted skin resurfac-
ing, 154, 156
for peri-implantitis, 135
for post-traumatic soft tissue infections,
143 – 144, 145
to prevent maxillofacial infections. See Maxillo-
facial infections.
Antiseptic agents, for oropharyngeal candidiasis, 97
Avelox, for head and neck infections, 33034
Azoles, for oropharyngeal candidiasis, 97, 98
B Cytochrome P450 system, and antibiotic
Bacterial infections, laser-assisted skin resurfacing
and, 152 – 154
1042-3699/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S1042-3699(02)00097-3
Bisphosphonates, for diffuse sclerosing osteomyelitis,
of jaws, 76
Bite wounds, management of, 145 – 147
Bone penetration, by antibiotics, 21
C
Candidal infections, laser-assisted skin resurfacing
and, 155
oropharyngeal. See Oropharyngeal candidiasis.
Cellulitis, versus retropharyngeal abscesses, imaging
of, 47
Cerebrospinal fluid penetration, by antibiotics, 21, 23
Cervicofacial actinomycosis, 51 – 58
clinical features of, 55 – 56
diagnosis of, 56 – 57
histopathology of, 54 – 55
historical aspects of, 53
management of, 57 – 58
medicolegal considerations in, 58
microbiology of, 54
natural history of, 54
Chemotherapy, and oropharyngeal candidiasis, 93
Clotrimazole, for oropharyngeal candidiasis, 97, 98
Computed tomography, of maxillofacial infections.
See Maxillofacial infections.
Condyloma acuminatum, dental issues in, 86 – 87
Corticosteroids, for recurrent aphthous stomatitis. See
Recurrent aphthous stomatitis.
Cutaneous maxillofacial infections, microbiology of,
11 – 12
metabolism, 27
Cytomegalovirus, dental issues in, 84 – 85
162 Index / Oral Maxillofacial Surg Clin N Am 15 (2003) 161–165

D G
Deep neck infections, recurrent, imaging of, 47 Gemifloxacin, for head and neck infections, 33034
Dental implant surgery, prophylactic antibiotics in, Guided bone regeneration, for peri-implantitis,
161 – 162 135 – 137
Dentoalveolar surgery, prophylactic antibiotics
in, 161 H
Denture wearers, oropharyngeal candidiasis in, 94, 95 Head and neck, necrotizing fasciitis of. See
Necrotizing fasciitis.
Diabetes mellitus, and oropharyngeal candidiasis, 93
Head and neck cancer, and oropharyngeal
immune defects in, 105 – 106
candidiasis, 93
Diffuse sclerosing osteomyelitis, of jaws, 69 – 78
Head and neck infections. See also
case report of, 77 – 78
Maxillofacial infections.
clinical features of, 72
antibiotics for, 17 – 38
differential diagnosis of, 71
fluoroquinolones, 33 – 34
etiology of, 71
for fungal infections, 36 – 37
histology of, 73
for odontogenic infections, 35 – 36
imaging of, 72 – 73
for osteomyelitis of jaw, 36
laboratory tests for, 73 – 74
for sinus infections, 36
management of, 75 – 77
ketolides, 34 – 35
microbiology of, 74 – 75
oxazolidinones, 34
pathogenesis of, 71
pristinamycins, 35
Disodium clodrenate, for diffuse sclerosing osteo- resistance to, mechanisms of, 18
myelitis, of jaws, 76 molecular biology of, 17
prevention of, 18 – 19
selection of, adverse reactions in, 26
allergy or intolerance in, 19 – 20
E antimicrobial spectrum in, 21
cost in, 27, 30, 33
Enterocolitis, periodontal surgery and, 127
drug interactions in, 26 – 27
Enteroviruses, dental issues in, 87 – 88 for usual pathogens, 19
in children, 26
Epstein-Barr viruses, dental issues in, 83 – 84
in immunocompromised patients, 20 – 21
Exodontia, prophylactic antibiotics in, 161 in pregnancy, 26
pharmacokinetics in, 23, 26
previous antibiotic therapy and, 21
tissue distribution in, 21, 23
prevention of, antibiotics in, 163
F
Facial bone fractures, antibiotics for, 162 – 163 Hepatitis C virus infections, dental issues in, 88
Fluconazole, for candidal infections, due to laser- Herpes simplex virus infections, dental issues in,
assisted skin resurfacing, 155 82 – 83
for oropharyngeal candidiasis, 97 laser-assisted skin resurfacing and, 154 – 155
maxillofacial, microbiology of, 12
Fluoroquinolones, for head and neck
infections, 33034 Herpesvirus infections, dental issues in, 82 – 83,
85 – 86
Fractures, facial bone, antibiotics for, 162 – 163
mandibular, antibiotics for, 162 HIV infection. See also AIDS;
Immunocompromised patients.
Frontal sinusitis, microbiology of, 10 – 11 and oropharyngeal candidiasis, 93
Fungal infections, laser-assisted skin resurfacing dental issues in, 81 – 82
and, 155 head and neck infections with, antibiotics for, 20
of head and neck, antibiotics for, 36 – 37 immune defects in, 106 – 107
 Index / Oral Maxillofacial Surg Clin N Am 15 (2003) 161–165
Human papillomavirus, dental issues in, 86 – 87
Hyposalivation, and oropharyngeal candidiasis, 94
I prevention of, 151 – 152
Immunocompromised patients. See also AIDS;
HIV infection.
head and neck infections in, 103 – 110
AIDS patients, 107
airway management in, 107
alcoholic patients, 106
antibiotics for, 20 – 21, 107 – 108
diabetic patients, 105 – 106
HIV-positive patients, 106 – 107
versus normal immune system, 103 – 105
Impacted third molar surgery, prophylactic antibiotics
in, 161
Infections, head and neck. See Head and
neck infections.
laser-assisted skin resurfacing and. See
Laser-assisted skin resurfacing.
maxillofacial. See Maxillofacial infections.
periodontal. See Periodontal infections.
post-traumatic soft tissue. See Post-traumatic soft
tissue infections.
Infectious rhinitis, microbiology of, 10
Itraconazole, for oropharyngeal candidiasis, 97
J technique for, 41 – 43
Jaws, diffuse sclerosing osteomyelitis of. See Diffuse
sclerosing osteomyelitis.
osteomyelitis of, antibiotics for, 36
K acute maxillary sinusitis, 9 – 10
Kaposi’s sarcoma, dental issues in, 96
Ketek, for head and neck infections, 34 – 35
Ketoconazole, for oropharyngeal candidiasis, 97
Ketolides, for head and neck infections, 34 – 35
Koebner phenomenon, in recurrent aphthous
stomatitis, 114
L trauma-related, 7 – 9
b-Lactam antibiotics, for head and neck infections,
19 – 20
163
Laser-assisted skin resurfacing, and infections,
147 – 153
bacterial, 148 – 152
candidal, 152
fungal, 152
versus normal skin healing, 150
versus normal skin microbiology, 150
viral, 152 – 153
Leukoplakia, dental issues in, 83 – 84
oropharyngeal candidiasis and, 95
Lichen planus, oral, dental issues in, 88
Linezolid, for head and neck infections, 34
M
Magnetic resonance imaging, of maxillofacial
infections. See Maxillofacial infections.
Mandibular fractures, antibiotics for, 162
Maxillary sinusitis, microbiology of, 9 – 10
Maxillofacial infections. See also Head and
neck infections.
imaging of, 39 – 49
computed tomography in, interpretation of, 44
necrotizing fasciitis, 48 – 49
retropharyngeal abscess versus cellulitis, 47
septic thrombophlebitis, 47 – 48
sialadenitis, 48
sinusitis, 49
magnetic resonance imaging in, interpretation
of, 44 – 45
sinusitis, 49
technique for, 43 – 44
microbiology of, 1 – 15
chronic maxillary sinusitis, 10
cutaneous infections, 11 – 12
frontal and sphenoid sinusitis, 10 – 110
herpes simplex infections, 12
infectious rhinitis, 10
odontogenic infections, 4 – 7
antibiotic-resistant, 7
culture and antibiotic sensitivity in, 4
isolates per patient in, 4
patient characteristics in, 4
polymicrobial, 4 – 6
prevention of, antibiotics in, 155 – 160
adjunctive procedures with, 158
164 Index / Oral Maxillofacial Surg Clin N Am 15 (2003) 161–165

administration of, 157 – 158 Nontuberculous mycobacterial infections,

choice of, 157 maxillofacial, imaging of, 47
for dental implants, 158 – 159
Nystatin, for oropharyngeal candidiasis, 96 – 97
for exodontia and dentoalveolar
surgery, 158
for facial bone fractures, 159 – 160 O
for impacted third molar surgery, 158
Odontogenic infections, antibiotics for, 35 – 36
for major head and neck surgery, 160
imaging of, 45, 47
for mandibular fractures, 159
microbiology of. See Maxillofacial infections.
for orthognathic surgery, 159
for post-traumatic soft tissue Oral contraceptives, failure of, antibiotics and, 27
infections, 160
Oral mucosa, disruption of, and oropharyngeal can-
principles of, 156
didiasis, 94
risk of, 156 – 157
versus normal microflora, 2 – 3 Oropharyngeal candidiasis, 91 – 102
cutaneous, 2 – 3 clinical features of, 94 – 95
nasal and paranasal, 3 diagnosis of, 95
oral, 3 management of, 95 – 99
viral, 79 – 89 amphotericin B in, 97 – 99
cytomegalovirus, 84 – 85 antiseptic agents in, 97
enteroviruses, 87 – 88 choice of, 95 – 96
Epstein-Barr viruses, 83 – 84 combination therapy in, 99
hepatitis C virus, 88 nystatin in, 96 – 97
herpes simplex, 82 – 83 polyene antifungal agents in, 96 – 99
herpesvirus 8, 86 resistance to, 98
herpesviruses 6 and 7, 85 susceptibility testing in, 95
HIV infection, 81 – 82 risk factors for, 92 – 94
human papillomavirus, 86 – 87 local and oral, 94
Kaposi’s sarcoma virus, 86 microbial, 93
varicella-zoster virus, 85 systemic, 93 – 94
systemic, 95
Miconazole, for oropharyngeal candidiasis, 97
Orthognathic surgery, prophylactic antibiotics in, 162
Mononucleosis, dental issues in, 83 – 84
Osteomyelitis, diffuse sclerosing. See Diffuse
Moxifloxacin, for head and neck infections, 33034
sclerosing osteomyelitis.
Mycobacterial infections, maxillofacial, imaging of jaw, antibiotics for, 36
of, 47
Oxazolidinones, for head and neck infections, 34

N
Nasal sinus infections, microbiology of. See P
Maxillofacial infections. Paranasal sinus infections, microbiology of. See
Maxillofacial infections.
Necrotizing fasciitis, of head and neck, 59 – 67
classification of, 63 Penicillins, for cervicofacial actinomycosis, 57
clinical features of, 65 for head and neck infections, 19, 36
diagnosis of, 65 – 66 cerebrospinal fluid penetration by, 21, 23
historical aspects of, 61 – 63 pharmacokinetics of, 23
imaging of, 48 – 49 Peri-implantitis, 129 – 138
management of, 66 – 67
diagnosis of, 132 – 134
results of, 67 – 68
implant salvage in, 134 – 137
microbiology of, 63
antibiotics in, 135
pathogenesis of, 63 – 65
guided bone regeneration in, 135 – 137
NewTom imaging, of maxillofacial infections, 43 microbiology of, 129 – 130
 Index / Oral Maxillofacial Surg Clin N Am 15 (2003) 161–165 165

retrograde, 132 Retropharyngeal abscesses, versus cellulitis, imaging

titanium implants and, 130 – 131 of, 47
hydroxyapatite-coated, 131 – 132 Rhinitis, infectious, microbiology of, 10
Periodontal infections, 123 – 128
diagnosis of, 124 – 125
in adults, 124 – 125
S
in children, 124 Salivary gland function, impaired, and oropharyngeal
management of, 125 – 128 candidiasis, 94
complications of, 127 Septic thrombophlebitis, maxillofacial, imaging of,
Plain films, of peri-implantitis, 133 47 – 48
Sialadenitis, imaging of, 48
Polyene antifungal agents, for oropharyngeal
candidiasis, 96 – 99 Sinusitis, antibiotics for, 36
Post-traumatic soft tissue infections, 139 – 146 imaging of, 49
microbiology of. See Maxillofacial infections.
antibiotics for, 143 – 144, 145, 163
bite wounds and, 145 – 147 Skin resurfacing, laser-assisted. See Laser-assisted
antibiotics for, 146 – 147 skin resurfacing.
débridement of, 143, 145 Soft tissue infections, post-traumatic. See
postoperative care for, 143 – 144 Post-traumatic soft tissue infections.
prevention of, 142 – 143
risk factors for, 142 Sphenoid sinusitis, microbiology of, 10 – 11
wound closure in, 143, 145 Squamous cell carcinoma, oral, dental issues in, 87
wounding mechanism in, 141 – 142 Synercid, for head and neck infections, 35
Pregnancy, head and neck infections in, antibiotics
for, 26
T
Pristinamycins, for head and neck infections, 35 Telithromycin, for head and neck infections, 34 – 35
Tetracycline, for peri-implantitis, 135

Q Thalidomide, for recurrent aphthous stomatitis, 120

Quinupristin/dalfopristin, for head and neck
infections, 35
R and failure of, 131 – 132
Recurrent aphthous stomatitis, 111 – 122
clinical features of, 114 – 116
etiology of, 111
genetic factors in, 112
histopathology of, 116
immunologic factors in, 112 – 113
management of, 116 – 120
burst therapy in, 119 – 120
corticosteroid injections in, 119
corticosteroid rinses in, 118 – 119
corticosteroid syrup in, 119 – 120
objectives, principles, and caveats of, 117
thalidomide in, 120
topical corticosteroids in, 118 – 119
microbiology of, 113
nutritional factors in, 113 – 114
pathogenesis of, 111 – 112
Third molar surgery, prophylactic antibiotics in, 161
Thrombophlebitis, septic, maxillofacial, imaging of,
47 – 48
Titanium implants, hydroxyapatite-coated, success
success and failure of, 130 – 131
Trauma, maxillofacial, and infections, 7 – 9
Tuberculous infections, maxillofacial, imaging of, 47
V
Varicella-zoster virus, dental issues in, 85
Verruca vulgaris, dental issues in, 87
Verrucous leukoplakia, dental issues in, 87
Viral infections, laser-assisted skin resurfacing and,
154 – 155
Z
Zyvox, for head and neck infections, 34