Periodontology 2OW. Vol.

14, 1997, 216-248

Printed in Denmark . All rights reserved
Copyri ght 0 Munksgaard 1997
PERIODONTOLOGY 2000
ISSN 0906-6713
Advances in the pathogenesis of
periodontitis: summary of
developments, clinical
implications and future directions
ROY C. PAGE, STEVEN OFFENBACHER, HUBERT E. SCHROEDER,
GREGORY J . SEYMOUR &KENNETH S. KORNMAN
Purpose and rationale
The purpose of this volume of PERIODONTOLOGY 2000
is to present the basic concepts and facts constitut-
ing the current understanding of the pathogenesis of
human periodontitis based on the published litera-
ture. This chapter summarizes, organizes and as-
sembles the most important points in the preceding
chapters and, where useful, adds new information,
to further synthesize and develop new hypotheses,
concepts and ideas and to make clinical appli-
cations. Instead of providing original references, we
cite the other chapters in this volume or sections
thereof as references. Additional references are pro-
vided where new information not referenced in the
other chapters is presented. Our comments are root-
ed in the existing literature, but we freely speculate
beyond existing data and point out areas and direc-
tions for future research.
Theoretical approach
The paradigm of the pathogenesis of periodontitis is
shifting. Periodontitis is a family of related diseases
that differ in etiology, natural history, disease pro-
gression and response to therapy but have a com-
mon underlying chain of events. For example, the
histopathological and ultrastructural features and
pathways of tissue destruction as well as healing and
regeneration are very similar if not identical for all
forms of periodontitis. The same basic pathological
mechanisms underlie all forms of bacterially in-
duced periodontitis. These shared events are influ-
enced by disease modifiers, both genetic and en-
vironmental or acquired, which may differ from one
stage and form of disease to another. The clinical
picture observed is a result of the complex interplay
of these events and modifiers and the microbial
chalfenge. The severity and rate of progression of
disease feed back to influence the nature and magni-
tude of the microbial challenge by, for example, in-
fluencing the pH and availability of oxygen and vari-
ous nutrients in the periodontal pocket. This is a
new basic concept that was arrived at independently
by multiple observers. It provides the overall frame-
work upon which this volume has been constructed
(in this volume, see Fig. 1 in Page & Kornman (51)
and Salvi et al. (SO)).
Various forms of periodontitis with differing clin-
ical manifestations, natural histories and responses to
treatment result from differences in the microbial
etiology among groups and individuals (such as dif-
ferent species or combination of species) and/or fac-
tors that modify host response mechanisms or alter
innate susceptibility. The latter include but are not
limited to such systemic conditions as diabetes mel-
litus and compromised host defenses, hereditary fac-
tors and such environmental factors as cigarette
smoking and stress (in this volume, see: Hart & Korn-
man (23) and Salvi et al. (60)) (50). These modifying
factors account for the differences observed in differ-
ent periodontal conditions and they do so by inter-
fering with regulation, activation or inhibition of vari-
ous components of the mechanisms of host response,
216
Advances in the pathogenesis of periodontitis
tissue homeostasis and repair. An integral and essen-
tial part of this hypothesis and an aspect well sup-
ported by extensive evidence is that the components
of host response that normally provide protection are
the same ones that accomplish destruction.
The modifying influences may affect the age of
disease onset, the patterns of observed bone and
tissue destruction, the rates of disease progression,
the response to various kinds of therapy and the se-
verity and frequency of disease reoccurrence. These
modifying influences, such as some hereditary fac-
tors, may endure for life or may be present or absent
or vary in magnitude of effect at various times in life.
For example, individuals born with leukocyte ad-
hesion deficiency syndrome 1, in which the patients
are completely deprived of neutrophil protection,
manifest recurrent enduring acute infections, in-
cluding severe periodontitis, when they are very
young. In the late teenage years and beyond, how-
ever, the frequency of infections in these patients de-
creases dramatically for reasons that are not under-
stood but are probably related to amplification of
non-neutrophil-based aspects of host defense,
somewhat like development of collateral circulation
to heart muscle following coronary artery throm-
bosis. This idea fits with the known high degree of
redundancy among the components of the host de-
fense system.
This theoretical approach to the pathogenesis of
periodontitis permits a better understanding of why
individuals vary greatly in disease susceptibility, clin-
ical manifestations, rate of progression and thera-
peutic responsiveness. Of much greater importance,
it points the way to new and novel approaches to
prevention, diagnosis, treatment and prognosis. In-
deed, it points towards eventual control of this fam-
ily of diseases in many human populations. For ex-
ample, a polymorphism in the interleukin 1 (IL-1)
gene family has been identified that, when positive,
enhances the probability of having severe peri-
odontitis later in life by many times (31). Approaches
to successfully controlling and preventing these dis-
eases in individuals and in populations are now on
the horizon. Day-to-day practice is changing from
identifying disease presence and severity and apply-
ing measures to arrest progression to a new para-
digm consisting of regenerating destroyed tissue in
the individuals already affected and preventing dis-
ease and maintaining health in healthy individuals.
Prevention is based on assessment of risks that en-
hance disease susceptibility, progression and sever-
ity by affecting common, shared events in the patho-
genesis and applying measures to reduce the risk.
The microbial challenge
Periodontitis is an infectious disease
Periodontitis is an infectious disease, and the major
pathogens have been identified. Bacteria are essen-
tial but insufficient to cause disease. Host factors
such as heredity and environmental factors such as
smoking are equally as important as determinants
of disease occurrence and severity of outcome. The
interaction of these factors is well demonstrated by
recent research findings that demonstrate the com-
plexity of the interactions in multifactorial diseases
and include roles for specific bacteria and genetic
and environmental modifymg and risk factors. Re-
cent work has shown that individuals with localized
juvenile periodontitis, but not generalized juvenile
periodontitis, have elevated levels of immunoglob-
ulin G2(IgG2), which are genetically controlled (76).
The data suggest that a genetically determined en-
hanced immune response to Actinobacillus actino-
mycetemcomitans assists in limiting the disease to a
localized clinical expression. Smoking has been
shown to further compromise the individuals with
generalized disease by lowering the IgG2 response.
Interestingly, smoking did not reduce IgG2 in the lo-
calized juvenile periodontitis group. The specific
bacterial challenge of A. actinomycetemcomitans
therefore produces very different clinical disease
patterns that appear to be determined by combi-
nations of genetic factors and smoking.
Important new findings and perspectives related
to the microbial challenge are listed in Table 1. Most
cases of periodontitis are caused by a small number
of bacteria species. Although the subgingival flora
can harbor hundreds of species, serotypes and bio-
types, experts agree that, except for acute necrotiz-
ing periodontitis, Porphyrornorzas gingivalis, Bacter-
oides forsythus and A. actinomycetemcomitans cause
most cases of periodontitis (8). This is a major step
forward for several reasons. It permits tests to be de-
veloped for detecting the presence of specific peri-
odontopathic bacteria and provides a practical
rationale for monitoring patients and normal sub-
jects over time. As the capacity to identify early in
life the individuals who are highly susceptible for de-
velopment of severe periodontitis in later life is de-
veloping, monitoring the flora for these pathogens
becomes doubly important. Equally important, the
fact that only a few predominant species are in-
volved in causing human periodontitis makes it
more likely that a vaccine can be developed for pre-
vention in high-risk individuals, families and popu-
217
Page et al.
Table 1. The pathogenesis of periodontitis:
new findings and new perspectives on the
bacterial challenge
A limited number of specific bacteria are essential
to initiate disease and fuel progression but
insufficient to explain the prevalence and severity of
periodontitis.
environmental factors as smoking, are major
determinants of disease occurrence and severity.
Subgingival microbial plaque behaves as a biofilm.
Bacteria living as biofilms are difficult to eradicate
by antimicrobial means and are protected from
host defense mechanisms.
Biofilms are effectively managed by physical
disruption and removal.
In periodontitis, an enormous load of gram-negative
bacteria is located adjacent to the inflamed tissue.
Some of the recognized periodontal pathogens have
multiple genetically distinct clonal types. Some of the
clonal types are probably more virulent than others.
Bacterial transmission is common in family units.
The recognition of bacterial transmission and the
Host factors, including heredity and such
increased ability to identify individuals at high risk for
severe disease have major implications in preventing
and treating periodontal diseases.
Porphyromom gingivulis can impair the local
neutrophil response to itself and other
microorganisms in the plaque by blocking the
normal adhesion molecule response.
lations and for possible treatment of recalcitrant
cases.
Subgingival microbial plaque is a biofilm
Costerton et al. (9) have defined biofilm as matrix-
enclosed bacterial populations adherent to each
other and/or to surfaces or interfaces. Subgingival
microbial plaque adheres tightly to the tooth root
surface and manifests all the characteristics ex-
pected of biofilms (see Darveau et al. in this volume
(lo), pages 12-14). In addition, in contrast to other
biofilms, loosely adherent and unattached bacteria
are located between the biofilm and the gingival
tissue. The recent realization that subgingival mi-
crobial plaque is a biofilm and the rapidly growing
body of information about the nature of biofilms in
general is one of the most important conceptual ad-
vances in periodontal microbiology in many years.
Although the understanding of the behavior of bi-
ofilms is still minimal, the concept explains many
aspects of periodontal disease that were previously
obscure.
The behavior of bacteria in biofilms is remarkably
different than in planktonic cultures, and pathogen-
icity and virulence factor expression may be en-
hanced significantly. The characteristics of biofilms
Table 2. Characteristics of biofilm
Ecological communities evolved to allow the survival
The communities exhibit metabolic cooperativity.
There is a primitive circulatory system.
Numerous microenvironments have radically
of the community as a whole.
different pH, oxygen concentrations and electric
potentials.
Biofilms resist the usual host defenses.
Biofilms resist systemic or local antibiotics and
antimicrobial agents.
are listed in Table 2. Bacteria in biofilms may pro-
duce large numbers of proteins and other compon-
ents not expressed in culture. There is evidence that
they exchange information, and they build complex
structures that contain a primitive circulatory system
and microenvironments that vary greatly in pH, oxy-
gen availability and nutrient availability. Electrical
potential differences of more than 100 mV have been
measured in biofilms (9). Almost nothing is known
about the behavior of the periodontal pathogens I!
gingivalis, B. forsythus and A. actinomycetemcomit-
ans in biofilm structures. Elucidation of the structure
and nature of subgingival biofilms and the behavior
of periodontal pathogens residing in them provides
an exciting and fertile area for future research.
The fact that plaque is a biofilm may shed light
on why some bacterial species require the presence
of one or more other species for survival whereas, in
other cases, the presence of one given species pre-
cludes the presence of another. Evidence indicates
that discrete colonies of specific species may be
located in intimate relation to colonies of other spe-
cies and that metabolites and products may be mu-
tually exchanged.
Microbial plaque is notoriously resistant to the
normal host defense mechanisms. Gingival fluid,
which contains complement, antibodies and all the
other systems present in blood for preventing and
controlling infection, flows through the periodontal
pocket, continuously bathing the biofilm. Comple-
ment is known to be activated, and millions of ac-
tive, viable leukocytes, especially neutrophils mi-
grate continuously into the periodontal pockets and
contact subgingival plaque (see Dennison & Van
Dyke in this volume (11)). Nevertheless, the bacteria
survive and flourish; they spread laterally and apic-
ally along the root surface, causing tissue destruction
and pocket deepening. Such behavior is understand-
able in terms of the known nature of biofilms in
which bacteria are protected from host defenses.
21 8
Advances in the pathogenesis of periodontitis
These observations may help to explain why, in
some cases, the disease progresses or reoccurs even
in patients with high titers of serum and gingival cre-
vicular fluid opsonic antibodies specific for antigens
of their infecting pathogens. These observations may
also dampen enthusiasm for development of vac-
cines for periodontitis.
As with other biofilms, subgingival microbial
plaque is extraordinarily persistent and difficult to
eliminate. Because individual bacterial colonies are
protected by one another and by extracellular ma-
terial in which they embed themselves, they are un-
usually resistant to the effects of antibiotics and
other antimicrobial agents either applied locally or
administered systemically. This is probably why the
placement into periodontal pockets of devices that
deliver enormous concentrations of antibiotics that
would be expected to virtually sterilize the pocket
almost always yield disappointing results. It has
been well established that bacteria growing in bi-
ofllms are resistant to levels of antibacterial agents
several times greater than those that are inhibitory
in laboratory assays. The biofilm structure may also
account for the rebound in the pathogenic flora,
which usually occurs quite rapidly following the use
of antimicrobial agents.
Physical disruption and removal are effective ways
of dealing with biofilms. This is why scaling and root
planing is a remarkably effective treatment for peri-
odontitis. It has been an essential and central com-
ponent of periodontal therapy since the earliest
times and remains so today. Its effectiveness relative
to other forms of therapy is most certainly due to
the fact that it physically disrupts and removes the
microbial biofllm. Scaling and root planing is likely
to remain the central component of periodontal
therapy for the foreseeable future.
Periodontitis is strongly associated with major
systemic diseases
There is now strong evidence that periodontitis en-
hances the risk for various systemic diseases, includ-
ing atherosclerosis, coronary heart disease, stroke
and infants with low birth weight (4,481. Subgingival
biofilms on the roots of teeth are likely to be an im-
portant component of this enhanced risk. As noted
above, these biofilms are recalcitrant, they reoccur
following treatment and they contain numerous
gram-negative bacteria that continuously shed ves-
icles rich in lipopolysaccharide. A single pass of a
curette in a periodontal pocket may contain up to
lo8 cultivable bacteria. The biofilm load in a patient
with severe generalized periodontitis is large. If one
assumes the presence of 28 teeth and considers
tooth roots to be circular with an average diameter
of 10 mm and an average of 5 mm of biofilm on
each, the total mouth biofilm area would be about
72 cm2, or an area about the size of the back of an
adult human hand. This is an enormous burden of
gram-negative bacteria. Lipopolysaccharide and liv-
ing bacteria from the biofilms have ready access to
the connective tissues and circulation through the
pocket epithelium. Bacteremia occurs during treat-
ment by root planing (generally performed every 3-
4 months to control the disease) and even during
toothbrushing and chewing (4). The subgingival bi-
ofilms may provide a major and continuous source
of circulating lipopolysaccharide and even living
gram-negative bacteria, which directly affect vascu-
lar endothelium and account in part for the en-
hanced susceptibility to various systemic diseases
and the worsening of other diseases such as diabetes
mellitus.
Multiple clonal types
Another major advance in the area of microbiology
of periodontal diseases has been the development
and application of various procedures such as the
polymerase chain reaction and other DNA-based
techniques to detect extremely small numbers of
bacteria in complex samples and the demonstration
that most species of periodontopathic bacteria such
as I? gingivulis encompass a large number of genetic-
ally distinct clonal types (see Darveau et al. in this
volume (lo), page 25). I? gingivulis alone may have
50-100 or more. Very little is known about which of
these can cause periodontitis and which cannot. If
only a small portion are pathogenic, this could ac-
count for the many studies that have failed to dem-
onstrate a strong relationship between the presence
of the putative pathogenic species and active disease
and for the relatively high carrier rate in peri-
odontally normal individuals. Major tasks for the fu-
ture include identifying all the different clonal types
of I? gingivulis and the other major pathogens and
characterizing their pathogenicity and virulence.
Bacterial transmission
Development and application of the new DNA-
based techniques has also resolved other issues,
such as whether periodontopathic bacteria exist as
members of the normal oral flora in very small num-
bers that can greatly overgrow in disease-susceptible
219
Pane et al.
individuals or whether they are transmitted from
one individual to another. Several studies have
shown that transmission can and does occur among
individuals living in close contact, such as spouses
or siblings (see Darveau et al. in this volume (lo),
pages 25-26). The data indicate that when a patho-
genic species is present, all members of a given fam-
ily who are infected carry the same clonal type, and
generally only one clonal type is present in a given
pocket, patient, subject or family. The latter point,
however, is not yet well documented. Nothing is
known about the predominance of one clonal type
over another or about differing growth requirements
for the various clonal types.
These observations have major implications for
preventing and treating periodontal diseases. For ex-
ample, when one member of a family has peri-
odontitis, is periodontal therapy instituted to arrest
disease progression and eliminate the causative bac-
teria in the individual sufficient? Should all family
members be tested for the pathogen or pathogens?
Should all family members who are infected be
treated with the goal of eliminating the pathogen
from the entire family whether or not signs of peri-
odontitis are present? In other words, is the risk of
reinfection and recurrent active disease enhanced
when other periodontally normal family members
carry the pathogen? An approach of this sort may be
specifically implicated in families in which one or
more members may be highly genetically susceptible
to continuing or recurrent disease, such as those
with Papillon-Lefhre syndrome or any of several
neutrophil abnormalities. These critical questions
remain unanswered.
The unique role of l? gingivulis
Numerous potential virulence factors manifested by
periodontopathic bacteria have been identified and
studied. Ishikawa et al. (27) describe these in detail
in this volume (pages 79-85). Among these are toxins
that are lethal for leukocytes, lipopolysaccharide that
can activate and perpetuate tissue destruction, prod-
ucts that affect immune responses and cell growth,
and proteases that can destroy tissue and degrade
components important in host defense such as spe-
cific antibody and complement activation products.
None of these factors, however, appears to account
for important aspects of the pathogenesis: for ex-
ample, the very rapid overgrowth or bloom of such
pathogenic species as ? gingivalis in the subgingival
flora.
An entirely new mechanism that may account in
part for observed events was described recently.
Lipopolysaccharide from Escherichia coli and most
other gram-negative pathogenic species activates ex-
pression of E-selectin by vascular endothelial cells,
and this triggers binding of leukocytes, especially
neutrophils, and their emigration from the vessels to
the site of infection. This early event triggers an
acute inflammatory response and the formation of
an inflammatory cell infiltrate that, under favorable
conditions, can eliminate the infection. Darveau et
al. (10) demonstrate in this volume (page 21) that
lipopolysaccharide from P gingivalis does not have
the capacity to activate E-selectin expression by en-
dothelial cells; further, it blocks E-selectin expression
by other gram-negative bacteria and their lipopoly-
saccharide. P gingivulis lipopolysaccharide also fails
to elicit expression of IL-8 (a potent chemoattractant
for neutrophils), monocyte chemoattractant protein
1 and intercellular adhesion molecule 1 expression
by human endothelial cells, fibroblasts and epithelial
cells. In a mouse model of inflammation, it does not
elicit an immediate acute inflammatory response.
These observations suggest an unusual role for P
gingivulis in establishing subgingival biofilms and in
the bacterial colonization and overgrowth or bloom
of specific species. By shutting down the initial step
in the acute inflammatory process, I! gingivulis may
not only permit its own rapid growth because of the
absence of components of normal host defense but
it may also make possible the establishment and
growth of other species found in subgingival bi-
ofilms. These properties may be among the reasons
P gingivalis is so frequently associated with active
tissue destruction.
Host-parasite interactions are in
biological equilibrium
The combative character of the host-parasite inter-
action is presumed to account for the episodic na-
ture of periodontal disease progression. The perni-
cious sloughing of lipopolysaccharide from the mi-
crobial biofilm triggers the host release of catabolic
inflammatory mediators that destroy connective
tissues. Since the attachment apparatus, once lost,
has limited ability to regenerate spontaneously, the
destructive process appears unidirectional, with
greater microbial challenge resulting in more dis-
ease. At some point this destructive process usually
loses momentum and subsides, presumably due to
renewed host defenses, and thereby defines the epi-
220
Advances in the pathogenesis of periodontitis
~~
sodic nature of the disease process. One of the sim-
plest potential explanations for the loss of disease
momentum may be the loss of attachment itself,
which temporarily moves the tissue away from the
bacterial biofilm. This change in physical relation-
ships would be expected to reduce the concentration
of bacterial products in the tissue, perhaps suffi-
ciently to allow the host to recover. Once disease
progression is apparently stopped within the lesion,
a variable period of inactivity ensues during which
very limited repair occurs, only to be interrupted at
some time later by yet another episode of disease
activity. This process has been likened to a balance
or teeter-totter, with periods of hostile microbial
virulence overpowering host defenses and resulting
in clinical tissue destruction. The analogy is com-
pleted by a renewed host defensive antibody strategy
to limit the microbial damage, tipping the balance
temporarily in favor of the host.
Mathematically speaking, at a patient level a ran-
dom hit or first order model seems to reasonably ap-
proximate the distribution of attachment loss events
that occur among all sites within a patients mouth.
Thus, the distribution of attachment loss patterns of
disease appears to mimic a random selection of an
intraoral site for further disease progression. The fre-
quency rate at which these events occur within a pa-
tient, as determined by the first-order rate constant,
defines how many attachment loss episodes (or
hits) occur within a certain time period. Patients
with more severe disease appear to have a higher
hit-rate rather than a more prolonged or aggressive
period of disease progression at a particular site.
This suggests that host defenses act locally to rapidly
re-establish an equilibrium and bring the site into
remission. Although deeper sites have a somewhat
elevated risk of experiencing an episodic hit, the nu-
merical dominance of shallower sites means that a
patient with disease is more likely to experience the
episode at a shallow site. Thus, the random-hit
model is not unbiased toward randomness with re-
gard to site selection for disease progression, yet it
remains a reasonable approximation, since the
deep-site bias is relatively small, as the proportion
of deep sites is usually small.
These models of disease progression are based on
clinical observations and are not universally ac-
cepted. For example, the burst-like appearance of
attachment loss may be due to the poor resolution
of measurement tools. Indeed, in the last decade,
data have been reported using computer-assisted
probes and digital radiography, which lower the
threshold at which clinical change can be detected.
Even more sensitive markers of disease, such as bio-
chemical or microbiological markers, that are cap-
able of detecting subclinical changes in disease sta-
tus have been used. Models using these more sensi-
tive markers have provided a different perspective
on the interactive process than models based on
clinical observations.
Host-microbe interactions are interdependent.
The biofilm has co-evolved, maintaining a symbiotic
host-parasite interaction. Not all people or all
pockets are equally accommodating as hosts; for ex-
ample, deep sites provide a better growth environ-
ment than do less inflamed shallow sites. This com-
plex interaction can not be easily explained or
modeled as a unidirectional process, even using
multiple variables such as clinical signs, levels of mi-
crobial burden or others. We suggest that it more
closely mimics a closed system, like the cardiovascu-
lar system. In fact, the very nature of the regulatory
processes in the body allows the system to respond
to external stimuli yet keep the body within certain
boundaries. Thus the body survives external stimuli,
as long as they are within certain limits, because the
physiological process can convert these signals into
behaviors within a closed system. Affecting one
component of a closed system requires adaptive re-
sponses of many elements. For example, in the car-
diovascular system blood pressure, blood volume,
vessel resistance, cardiac output and stroke volume
are all linked interdependently, and changes in one
parameter, such as blood volume, affect most of the
other components within the system. In this ex-
ample, blood loss reduces pressure and output, lead-
ing to vessel constriction to raise blood pressure and
stroke volume and other compensatory responses in
an attempt to maintain equilibrium.
There is experimental evidence that the responses
in periodontal disease behave as if they were occur-
ring in a closed system. It has long been recognized
that bone resorption is a potent trigger for new bone
formation. One might expect that the equilibrium in
the balance of anabolic growth factors and catabolic
inflammatory mediators might regulate the stability
of the periodontal tissues in the presence of a mi-
crobial challenge. During bone loss, one might ex-
pect a disequilibrium, which would manifest as an
increase in inflammatory mediators and a decrease
in growth factors. However, at inflamed progressing
sites, both inflammatory mediators and anabolic
growth factors are simultaneously elevated relative
to noninflamed quiescent sites. The importance in
this observation is that this trait is a characteristic
of closed cyclic systems and not open-ended linear
221
Page et al.
models. Systems that are multifactorial with complex
feedback often do not have linear output patterns
but frequently display episodes that occur at non-
linear intervals. Such irregular episodes are the re-
sult of summing many nonlinear functions. This
may explain why attachment loss episodes have
been difficult to predict at a site level, and that, in
the future, neural networks or other nonlinear
modeling approaches (13) that can predict phase
shift behavior may prove to better predict the natural
history and clinical course of disease. Baxt (3) and
others have shown that these models are superior to
traditional linear modeling approaches in predicting
cardiac arrest, for example.
Further consideration of periodontal disease as a
closed system model with interdependence of host
factors and the biofilm can also lend insight into
therapeutic approaches. Anti-inflammatory agents
that are known to retard bone and attachment loss
might then be expected to influence the flora, per-
haps by diminishing nutrient availability, or by some
other indirect mechanism not yet characterized.
Periodontal surgery, which reduces the pocket depth
and impairs the anaerobic ecological niche, might
then be expected to have a long-term effect on the
microbial biofilm properties. Mechanical debride-
ment, which simultaneously lowers the local mi-
crobial burden and boosts the host local antibody
response, would not only be expected to result in
clinical improvement but would also be expected to
re-establish a new state of equilibrium for the sys-
tem. The closed behavior of the system would sug-
gest that perturbations are compensated for to reach
a state of equilibrium or homeostasis.
In complex systems with feedback, many different
equilibria may be achieved; otherwise the system
could not be controlled. Thus, changes in the flora,
such as the acquisition of new pathogens or changes
in the host such as increased inflammatory response
due to stress, would both change the equilibrium
and initiate a series of adaptive responses, such as
gingival recession, epithelial downgrowth or fibrosis.
For example, host stress that compromises neutro-
phi1 clearance would permit new penetration of mi-
crobial lipopolysaccharide and antigens, which
would stimulate an inflammatory response. The in-
flammatory process would both enhance nutrient
availability to the biofilm to promote the emergence
of nutrient-limited species and diminish antibody
production locally. The penetration of lipopolysac-
charide and bacterial antigens would re-initiate an
adaptive antibody response. The renewed round of
inflammatory monocytic tissue destruction would
222
ensue, until the neutrophil and antibody activity
could again bring the infection under control.
The pleiotropic role of IL-4 serves as a good ex-
ample of how adaptive responses might occur. IL-4
induces isotype switching in plasma cells to produce
IgG4, a more mature, high-avidity immunoglobulin
subclass produced late in the normal maturation se-
quence of antibody isotype switching. This antibody
is not highly opsonic and hence would not facilitate
clearance. On the other hand, IL-4 also serves to in-
hibit cytokine release from monocytic cells and can
induce apoptosis of monocytes that were previously
attracted and activated within the destructive period
of the lesion. At excessive levels, this interleukin
would decrease the levels of bone-resorbing cyto-
kines and levels of matrix metalloproteinases,
further reducing tissue destruction. High levels of IL-
4 also induce fibrosis and scarring - another hall-
mark of the periodontal lesion. Thus, periodontal
disease activity represents a joint or sequential acti-
vation of the inflammatory and the antibody com-
ponents of the immune response. The organisms are
affected by both sides of this response: first posi-
tively by inflammation and then negatively by anti-
body. This is more analogous to the interactions in a
closed system. This can be compared with ecological
studies of population shifts that can occur in co-
existing populations of predators and prey when
they are confined within a fixed territory. Saying that
this process is mediated entirely by the host or en-
tirely by the bacteria would simply propagate an-
other decade of misunderstanding. It would be like
a continuing argument about whether high blood
pressure is due to high cardiac output or high vascu-
lar resistance. A new appreciation for the interactive
process between the biofilm and the host can pro-
mote a concept of disease and health in which both
represent a state of biological equilibrium with dif-
fering levels of metabolic activity.
Assembling the players
During the early stages of developing gingival in-
flammation, all the players in the pathogenesis of
periodontitis are prepared for their roles in the
drama that is about to unfold. Resident junctional
epithelial cells and fibroblasts are notified and vari-
ous leukocytes and their subsets are assembled in
selected proportions and activated, and very large
varieties of messenger molecules that mediate cross-
talk among the cells and molecules that mediate
Advances in the pathogenesis of periodontitis
tissue destruction are produced (see Kornman et al.
(32) in this volume). Important new findings and
perspectives are listed in Table 3.
Summary of events: the process is iterative
Gingiva is a unique tissue. In contrast to other
tissues, under healthy noninflamed conditions the
small vessels of the subgingival plexus express ad-
hesion molecules such as E-selectin, and a continu-
ing stream of neutrophils exits the vessels and mi-
grates through the junctional epithelium into the
gingival sulcus. An estimated 530,000 leukocytes per
minute (predominantly neutrophils) migrate into
the oral cavity in periodontally healthy adults (55).
These cells reflect a normally functioning host de-
fense against a low level challenge, and tissue de-
struction does not ensue.
As the microbial challenge increases, clinically
manifest inflammation of the marginal gingiva be-
gins (63, 65). Bacterial components or their products
interact with the epithelium and penetrate into the
underlying connective tissue. The small blood ves-
sels immediately deep to the junctional epithelium
become inflamed, resulting in enhanced per-
meability and a great increase in the numbers of
neutrophils that exit the vessels and migrate through
the junctional epithelium and into the gingival sul-
cus. As this occurs, small quantities of collagen and
Table 3, The pathogenesis of periodontitis:
new findings and new perspectives on the
gingival tissue as a stage for the host-parasite
interactions
The cellular and molecular response to the evolving
bacterial challenge is iterative and involves constant
adjustment and regulatory feedback.
The epithelium plays an active sensing and signaling
role that is involved in specific leukocyte recruitment
and vascular permeability.
The bacterial activation of the junctional epithelium
initiates vascular endothelial responses that
initially involve primarily neutrophils migrating to
the sulcus.
The specific leukocyte populations in the tissue are
tightly regulated by cytokine and chemokine
responses to bacterial products that enter the tissue
and selectively activate certain cytokines and
chemokines.
Neutrophils do not dwell in the tissue but constitute
the majority of leukocytes in the sdcus, whereas
macrophages, lymphocytes and plasma cells form
the majority of cells in the tissue.
If the bacterial challenge is not controlled early,
resident tissue cells, such as fibroblasts, may become
activated by bacterial products and cytokines to
participate actively in tissue destruction.
other components of the perivascular extracellular
matrix are destroyed.
If the microbial challenge remains undisturbed, in
a matter of days marginal gingival inflammation or
gingivitis appears in most individuals. Subsequent
host defense activities may contain the challenge, re-
sulting in inflammation that is not worsening or im-
proving, or the microbial challenge may overcome
the host defense, allowing conditions to worsen.
When this occurs, supragingival plaque extends into
the gingival sulcus, disrupting the union between
the most coronal portion of the junctional epithel-
ium and the root surface. As a consequence, a shal-
low gingival pocket wi t h a typical pocket epithelium
forms. At this stage, there is no periodontal pocket,
as there is no attachment or bone loss. The presence
of the pocket epithelium, however, heralds an es-
calation in the progression toward periodontitis. The
pocket epithelia provide ready access for bacterial
substances to shower the connective tissues and to
activate the epithelial cells to express IL-8 and endo-
thelial cells of the inflamed small vessels to enhance
expression of adhesion molecules, resulting in en-
hanced emigration of leukocytes, edema and forma-
tion of an inflammatory cell infiltrate. The infiltrate
consists of monocytes and lymphocytes including B
cells and T cells of both the T-helper 1 (Thl) and
Th2 phenotype and emigrating neutrophils. Cyto-
kines produced by these cells and antigens from the
bacteria drive differentiation of B cells to specific
antibody producing plasma cells. A humoral im-
mune response that may or may not be protective
may be mounted. At any given point, the host de-
fenses may contain the microbial challenge, or the
challenge may be reduced, such as by tooth
cleaning, and the process reversed.
If containment does not occur, inflammation
worsens and is perpetuated, and connective tissue
and bone is then destroyed (Fig. 1). Macrophages ac-
tivated by lipopolysaccharide produce IL- l p, tumor
necrosis factor a, matrix metalloproteinases and
prostaglandin EZ. I L-I p and tumor necrosis factor a
activate resident fibroblasts to produce prosta-
glandin E2 and matrix metalloproteinases. Both acti-
vated cell types decrease the production of tissue in-
hibitors of metalloproteinases, resulting in greatly
increased relative levels of matrix metalloproteinas-
es. This destroys components of the extracellular
matrix, creating space for the enlarging inflamma-
tory cell infiltrate. The microbial biofilm may extend
apically and laterally. The epithelial cells activated by
lipopolysaccharide can produce matrix metalloprot-
einases, which can destroy attached collagen fibers
223
Pape et al.
Fig. 1. The monocyte/macrophage (M0) plays a central
role in the pathogenesis of periodontitis. Lipopolysac-
charide (LPS), the principal virulence component of
gram-negative periodontal pathogenic bacteria, is shed
from the biofihn and enters the gingival tissue. It binds to
lipid-binding protein (LBP), and this complex is recog-
nized by the CD-14 receptor expressed by the monocytel
macrophage. Receptor occupancy initiates transmem-
brane signals that activate the cells to synthesize and se-
crete prostaglandins, cytokines and matrix rnetalloprot-
einases (MMP). The activity of the activated cell is modu-
lated by hereditary factors and can be suppressed by
interferon y. Tumor necrosis factor a (TNFa) and I L-lP
at the apical terminus of the junctional epithelium,
allowing apical extension of the epithelium, forma-
tion of additional pocket epithelium and pocket
deepening. As this occurs, matrix metalloproteinases
mediate clinical attachment loss and prostaglandin
EZ mediates resorption of the alveolar bone (and in
some cases cementum and dentin), and the gingival
pocket progresses to become a periodontal pocket.
The events comprising the formation of gingival
and subsequently periodontal pockets as presented
above may appear unidirectional. This is not the
case; they are iterative, with cross-talk between the
microbial challenge and the host defenses that are
called into play. At any given point in the progression
bind to cell surface receptors of resident fibroblasts, initi-
ating signals to synthesize and secrete matrix metalloprot-
einase and prostaglandin E2 (PGE2). Matrix metalloprot-
einases mediate destruction of the extracellular matrix of
the gingiva and periodontal ligament, and prostaglandin
E2 mediates alveolar bone destruction. IL-ls and tumor
necrosis factor a directly mediate a minor portion of bone
loss. The diagram does not include proteases and in-
flammatory mediators produced by endothelial and epi-
thelial cells nor by the infiltrating neutrophils which con-
tribute significantly to the matrix metalloproteinase con-
centrations. Source: modified from Offenbacher et al. (47).
of events, host defenses may contain the challenge
and arrest or reverse progression, or the challenge
may overcome the defenses and enhance pro-
gression.
The junctional epithelium has sensing and
signaling effects on functions
The junctional epithelium has previously been view-
ed as more or less a bystander that provides a barrier
function or curtain separating the biofilm from the
connective tissue and does little else during the de-
velopment of periodontitis (see Kornman et al. (32)
in this volume, pages 34-37). Most of the real action
224
Advances in the pathogenesis of periodontitis
was thought to occur in the gingival and periodontal
ligament connective tissue and alveolar bone. This
now seems far from the case. The junctional epithel-
ium is engaged even prior to the onset of clinical
manifestations of gingivitis, when large numbers of
neutrophils migrate through the junctional epithel-
ium and into the mouth (63). The cells of the junc-
tional epithelium express intercellular adhesion mol-
ecule l and IL-8, a chemoattractant specific for neu-
trophils and for a small subset of lymphocytes.
Intercellular adhesion molecule 1 appears to be es-
pecially important to the neutrophil migration
across mucosal membranes. Antibody specific for in-
tercellular adhesion molecule 1 greatly inhibits neu-
trophil migration. Among the oral epithelia, neutro-
phi1 migration occurs through the junctional epithel-
ium and the pocket epithelium to which it gives rise;
and initially intracellular adhesion molecule 1 ex-
pression is limited to these epithelia. However, as in-
filtrate size increases, the junctional] sulcular and
gingival epithelia all may express intercellular ad-
hesion molecule 1 (18).
Neutrophil migration across the epithelial barrier
and the accessibility of antigen to the underlying
connective tissues may be affected by intraepithelial
mononuclear cells. These include cells likely to be
Langerhans or dendritic cells and specific subsets of
lymphocytes such as mucosal T cells and T cells
having a repertoire of receptors enriched for bac-
terial antigens. Although the role of these cells re-
mains obscure, their populations are greater at sites
of inflammation than at noninflamed sites, sug-
gesting that they may engage in limiting antigen
penetration and in antigen recognition, processing
and presentation.
Following extension of the biofilm subgingivally
into the gingival sulcus, the sloughing surface of the
junctional epithelium and, at a later stage, the
pocket epithelium are in direct contact with both the
microbial mass of the subgingival biofilm or sub-
stances such as lipopolysaccharide derived from it
and the gingival connective tissue. The junctional
epithelium is uniquely rich in neural elements; the
detailed function of these is unclear. The cells of the
junctional epithelium can sense the biofilm en-
vironment and respond by generating inflammatory
mediators that serve as messengers, transmitting in-
formation to the connective tissues, especially the
small vessels of the vascular plexus in the lamina
propria. I L-8 may be particularly important in facilit-
ating the transmigration of neutrophils and their ac-
cumulation at the surface of the biofilm. There is an
increasing gradient in an apicocoronal direction
guiding neutrophils to accumulate at the surface of
the biofilm. The cells express high-affinity receptors
that bind chemoattractant molecules at the low con-
centrations that exist some distance from the biofilm
and cause chemotaxis. Low-affinity receptors be-
come occupied and activate the cells near the bi-
ofilm where concentrations are high. Similarly, neu-
trophils express high and low affinity receptors for
the chemoattractant leukotriene B4.
As the biofilm extends apically between the junc-
tional epithelium and the tooth surface, the cells be-
come detached and additional pocket epithelium
forms. The epithelial cells continue to express inter-
cellular adhesion molecule 1 and become activated
and synthesize and secrete prostaglandin EP, IL-8
and matrix metalloproteinases. These products
along with inflammatory metabolites and compon-
ents released by the bacteria can initiate and per-
petuate the inflammatory response in the underlying
connective tissues.
For reasons that remain obscure but that may be
related to apical extension of the biofilm or to epi-
thelial cytokine production, basal cells of the junc-
tional epithelium begin to replicate and extend rete
ridges into the connective tissue and to migrate ap-
ically along the tooth root. Such migration and ex-
tension can only occur when the adjacent dense col-
lagen fibers attached to root cementum are de-
stroyed and removed to create space. The fact that
the epithelial ceus can produce matrix metalloprote-
inase provides a potential mechanism through
which pocket formation and deepening can occur.
The significance of junctional epithelium-derived
matrix metalloproteinase in pocket deepening re-
mains to be studied.
The leukocyte populations of the inflammatory
infiltrate are tightly regulated
One of the major contributions of Page & Schroeder
in 1976 (52) was the observation that the periodontal
lesion begins as an acute inflammation predomi-
nated by enhanced neutrophil emigration and for-
mation of an inflammatory cell infiltrate that soon
becomes dominated by small and medium-sized
lymphocytes, mostly T cells. This infiltrate then de-
velops into a fully evolved inflammatory infiltrate
dominated by B cells and plasma cells but also con-
taining T lymphocytes, macrophages and neutro-
phils. In 1976, the mechanisms governing infiltrate
formation, leukocyte and subset composition and
cell function were not understood. Further, in hind-
sight] the structural studies that had been performed
225
Page et al.
were inadequate, in part because the concept of dis-
ease-active and disease-inactive periodontal sites
had not yet been developed, and most studies of the
early lesion had been conducted on specimens from
adolescents (65). Major advances have since been
made in elucidating the details of leukocyte infil-
tration at sites of inflammation (see Kornman et al.
(32) in this volume, pages 37-43) (65).
Based on information now available, the staging
of the pathogenesis of periodontitis as initial, early,
established and advanced lesions that was done in
1976 remains basically correct, except that the early
lesion, characterized by a predominance of T
lymphocytes, is not a distinctly identifiable stage, as
gingival inflammation develops in adults. It is in-
stead observed during tooth eruption in children
and in the marginal gingiva of adolescents (65). In
adults, neutrophils and subsequently lymphocytes
and monocytes leave the circulation and accumulate
in the marginal gingiva as gingivitis and subsequent-
ly periodontitis develops. This process is tightly con-
trolled, and many of the details are now known. In-
teractions occur between the leukocytes and endo-
thelial cells of the postcapillary venules that result
in diapedesis and infiltrate formation. In the initial
stages this interaction may result from mast cell-de-
rived tumor necrosis factor a leading to the upregu-
lation of endothelial cell adhesion molecules. Mast
cell stimulation could occur as a result of possible
neural connections with intraepithelial Langerhans
cells (78).
As inflammation develops, specific and differing
populations of inflammatory cells accumulate in the
connective tissue and in the sulcus or pocket. Neu-
trophils do not dwell in the connective tissue, but
they constitute the majority of leukocytes in the sul-
cus or periodontal pocket, whereas macrophages,
lymphocytes and plasma cells form the majority of
cells in the connective tissues. Antigen-specific
memory and activated lymphocytes, aELa7 mucosal
lymphocytes, y6 T-cell receptor-positive cells and
CDla+antigen-presenting cells migrate selectively
into the gingival tissues, This is highly regulated and
is not a result of simple unregulated migration.
Mechanisms operate that are general for leuko-
cytes and others that confer specificity and selec-
tivity on diapedesis and infiltrate composition and
size (see Fig. 4 in Kornman et al. (32) in this volume).
The bacterial and inflammatory products that up-
regulate selective sets of adhesion molecules on
leukocytes and endothelial cells therefore determine
the migration kinetics of specific cell subpopulations
out of the vasculature and into the tissue.
In addition to the participation of specific ad-
hesion molecules, selective migration and accumu-
lation of leukocytes is determined by the recently
discovered chemokines, a family of low-molecular-
weight cytokines with potent and cell type-specific
chemoattractant properties. IL-8 is specific for neu-
trophils and a small population of lymphocytes,
monocyte chemoattractant protein 1 for monocytes
and other chemokines for other leukocytes.
As inflammation worsens, the pocket epithelium
becomes chemoattractant negative and neutrophil
migration through the epithelium and into the
pocket decreases (18). A possible decrease in the
gradient of chemoattraction may also occur, with
neutrophils dwelling and becoming activated within
the gingival tissue. This could be an important event
in the evolving capacity to destroy connective tissue.
Cell homing may also participate in infiltrate for-
mation. There is evidence from rodents that lymph-
oid cells specific for the antigens of periodontal
pathogens may specifically home from the circula-
tion to the gingival tissues. Eastcott et al. (12) sensi-
tized clones of lymphocytes to antigens of A. actino-
mycetemcomitans and infused them into mice.
When the animals were orally infected with the same
microorganism, the infused cells homed to the gin-
giva, but homing did not occur when the mice were
not infected. Whether such homing exists in humans
is not known.
It is these mechanisms that select for infiltrates
with differing proportions and total numbers of spe-
cific leukocytes and their subsets. Infiltrate compo-
sition and size are not static; infiltrates are dynamic
and change depending on changes in the subgingi-
val microbiota that drive infiltrate formation. The
properties and concentrations of factors activating
the endothelial cells and the nature and duration of
receptor expression are major determinants of infil-
trate composition and characteristics.
Resident tissue cells become active participants in
destruction
There is another major component. One of the most
significant advances in understanding the patho-
genesis of periodontitis over the last two decades has
been documentation that, in addition to the leuko-
cytes of the inflammatory infiltrate, the cells residing
in the normal periodontium, including fibroblasts,
junctional epithelial cells and vascular endothelium
(all of which function in the healthy periodontium
to maintain homeostasis), can be hijacked by ex-
posure to bacterial agents such as lipopolysacchar-
226
Advances in the pathogenesis of periodontitis
~~
ide, cytokines such as IL-1 or tumor necrosis factor
a or prostaglandins such as prostaglandin E2 and be-
come major participants in tissue destruction.
Well-documented events involving the gingival
fibroblasts illustrate this (Fig. 2). Under conditions of
health, the fibroblast genes for various collagens and
tissue inhibitors of matrix metalloproteinases are
turned on and functioning and the genes for matrix
metalloproteinases are turned off, as assessed by
measurement of various types of messenger RNA.
Genes for collagens and tissue inhibitors of metallo-
proteinases in fibroblasts in active periodontitis
lesions are turned off and those for matrix metallop-
roteinases are turned on. In addition, in the presence
of lipopolysaccharide, fibroblasts produce their own
inflammatory mediators, such as IL-lp. Further, fol-
lowing successful treatment, as the levels of bacterial
substance such as lipopolysaccharide, and proin-
flammatory cytokines such as IL-lP and tumor ne-
crosis factor a decrease, the reverse occurs. Whether
the same cells convert from an anabolic to a cata-
bolic state or whether the fibroblasts are replaced by
a different phenotype is not known. It is known that,
at an early stage of gingivitis, resident tissue fibro-
blasts are killed through what appear to be immuno-
pathological mechanisms, as seen in humans and
Fibroblast
/ Gene Activation \
Health
8
Metalloprotein&es
I
Cofiagen and other
extracellular matrix;
Inhibitors of
metalloproteinases
Tissue remodeling,
maintenance, healing,
regeneration
Tissue destru&on
*
Fig. 2. The role of resident fibroblasts in health and in
periodontitis. In situ hybridization and other studies have
shown that, in the healthy periodontiurn, the resident
fibroblasts genes for collagens and other extracellular
matrix components and tissue inhibitors of metalloprot-
einases are turned on and those for matrix metalloprot-
einases are turned off such that tissue remodeling, main-
tenance, healing and regeneration can occur. In active
periodontitis, genes for tissue inhibitors of metalloprot-
einases and components of the normal extracellular mat-
rix are turned off and those for matrix metalloproteinases
are turned on, permitting the resident fibroblasts to play
a major role in tissue destruction.
monkeys, resulting in significantly reduced popula-
tions (66, 69, 70). At subsequent stages, fibroblast
population sizes are restored. Whether these cells are
of the same phenotype as the cells present in the
healthy tissues remains unknown.
The concept of radius of effectiveness
Gross and histological features of periodontal lesions
in humans give clues as to how they may develop
and function. There has been considerable debate as
to the determinants of the frequency and cause of
intrabony lesions. Garant & Cho (14) suggested that
locally produced stimulators of bone resorption may
have an effective radius of action, and Waerhaug (77)
demonstrated that bone resorption occurs when mi-
crobial plaque approaches to within 0.5 to 2.0 mm
of the bone surface. Based on these and other obser-
vations (62), Page & Schroeder (53) postulated a
range of effectiveness of the subgingival plaque (bi-
ofilm) in generating alveolar bone loss of about 2.5
mm. They stated that when the bone surface has
been resorbed to about 2.5 mm apical or lateral to
the site of the bacteria, bone loss appears to cease
and bone production takes over, until it equals or
surpasses resorption. The inflammatory infiltrate
also plays a role. The closer the cells of the inflam-
matory infiltrate are to the bone, the more osteo-
clasts appear and the more bone is degraded (59,
64). Tal (73) provided data supporting this hypo-
thesis by measuring intrabony lesions in 344 inter-
proximal areas and 117 intrabony pockets in 84 pa-
tients. He observed intrabony lesions only rarely
when interdental distances were less that 2.6 mm.
Tho intrabony defects on adjacent teeth were ob-
served only when the interdental distances were
greater that 3.1 mm. AsSchroeder (62) pointed out,
lesions with dimensions much greater that 2.5 mm
are seen around single teeth not only in humans but
also in dogs and other species. There is evidence that
these are accounted for by the invasion of the tissue
by bacteria and abscess formation.
Some people develop gingivitis
and others periodontitis
For several decades periodontologists have con-
sidered the central question of how gingivitis pro-
gresses to periodontitis in the belief that the answer
will provide critical clues to aid in the prevention of
227
Page et al.
Table 4. The pathogenesis of periodontitis:
new findings and new perspectives on the
transition of gingivitis to periodontitis
Gingivitis is generally characterized biochemically
by an increase in leukotriene B4 in the gingival
crevicular fluid. This is a product of degranulating
neutrophils in the sulcus.
characterized by an increase in prostaglandin &,
IL-10 and tumor necrosis factor a, which represent
an activation of macrophages and lymphocytes
wi thi n the tissue.
Children do not demonstrate a macrophage and
lymphocyte response to bacterial plaque
accumulation, unless they have defective neutrophil
function.
In young adults involved in experimental gingivitis
studies, some have a neutrophil (leukotriene B,)
response, and some have a macrophage and
lymphocyte response.
The transition from gingivitis to periodontitis may
be monitored by the biochemical shift from
leukotriene B4 dominance to expression of
macrophage and lymphocyte activation products,
such as prostaglandin El.
Some individuals appear to have a very transient
clinical phase in the transition from gingivitis to
periodontitis. Others do not appear to undergo the
transition to eriodontitis. It may therefore be
practical to cgaracterize individuals as gingivitis or
periodontitis patients.
Later stages of gingivitis and periodontitis are
periodontitis. Important new findings and perspec-
tives are listed in Table 4.
Current evidence suggests that, whereas gingivitis
probably represents the early stage in the natural
history of initiating periodontitis, it may also exist as
an independent and stable clinical entity. The mi-
crobial composition of the biofilm associated with
both conditions seems to be quite similar; in fact,
gingivitis- and periodontitis-associated microbes are
commonly found even in adolescents. And since
periodontal pathogens can be transmitted, it would
appear that exposure to most periodontal pathogens
within a persons lifetime is a rather ubiquitous
event, but the biofilm must be sufficiently mature for
these late-blooming microbes to become resident
within the biofilm. In human populations such as
those in rural China (541, where oral care is minimal
or nonexistent, the prevalence of most periodontal
pathogens is about 90-95%. In fact, many or most of
these subjects are not infected with a few pathogens
but rather appear to harbor virtually all currently
recognized pathogens. This is somewhat surprising,
since the prevalence and severity of periodontal dis-
ease is diverse and not substantially greater than
that observed in populations in industrialized coun-
tries. Since bacteria are necessary for the initiation
~~ ~ ~
of periodontitis, these findings prompt the reassess-
ment of the contribution of various factors to the
clinical presentation of periodontitis.
It is now recognized that multifactorial diseases,
such as periodontitis and atherosclerosis, usually
have initiating factors, factors that permit or block
disease initiation and factors that modify clinical ex-
pression once the disease has been initiated. This bio-
logical model of multifactorial disease presents com-
plications in statistical models unless one explicitly
recognizes the underlying biological role of each fac-
tor. The process of disease initiation is being re-exam-
ined with recognition of the multifactorial nature of
periodontitis. The prominent pathogens, such as P
gingiualis and B. forsythus, in adult forms of peri-
odontitis will predictably emerge in the subgingival
plaque as the natural consequence of the undisturbed
maturation of the dental biofilm in individuals ex-
posed to routine social and family interchanges.
The accumulation of microbial plaque in children
and adolescents leads to an inflammatory response
referred to as gingivitis, the severity of which varies
among individuals. Experimental gingivitis in young
adults, which is caused by the overgrowth of gram-
positive species during the first 2-3 weeks following
the cessation of oral hygiene, presents clinically as
redness and edema that begins at a few interpapil-
lary and marginal areas and generally spreads to
more sites, as it slowly becomes more severe at in-
volved sites. During this period of time neutrophil
infiltration, epithelial proliferation and subepithelial
capillary proliferation dominate the lesion histolog-
ically. Biochemically, the barrage of neutrophil mi-
gration into the sulcus is marked by a rise in leuko-
triene B4 levels in gingival crevicular fluid, a product
of degranulating neutrophils (25). This reflects the
neutrophils actively engaging the bacteria in the sul-
cus, facilitating clearance. This represents the
earliest stage of the lesion that can be characterized
by neutrophil and keratinocyte activation facilitating
neutrophil recruitment and bacterial clearance.
Keratinocyte IL-8 is a potent signal for intracellular
adhesion molecule expression and neutrophil re-
cruitment. Neutrophil leukotriene B4 and comple-
ment C5a are vasoactive and sufficient for marginal
inflammation with little destruction of deeper con-
nective tissues.
As the biota shifts and becomes more gram-nega-
tive by 3-4 weeks, bleeding on probing begins to
rapidly increase concomitant with increases of
prostaglandin E2 in gingival crevicular fluid, a
marker of lipopolysaccharide penetration and mon-
ocytic activation (25). The ulceration from bleeding
228
Advances in the pathogenesis of periodontitis
on probing is classically considered to herald the
transition from gingivitis to periodontitis. Indeed,
the biochemical trigger of prostaglandin Ez also sug-
gests that neutrophil clearance capacity has been ex-
ceeded and lipopolysaccharide has penetrated into
the tissues, resulting in the release of prostaglandin
Ezr IL-1 and probably tumor necrosis factor a.
Prostaglandin E2 increases the chemotactic potency
of IL-8 by 10-100 fold (7), serving to exponentially
increase neutrophil recruitment. In contrast to neu-
trophil enzymes that are predominantly anti-
microbial in nature (lysozyme, lactoferrin, defensins
and P-glucuronidase), these mediators (prosta-
glandin EZ, IL-1 and tumor necrosis factor a) target
host cells to initiate tissue catabolism. These provide
signals that trigger fibroblastic apoptosis, matrix
metalloproteinase release and the expression of
homing receptors for lymphocytic recruitment.
What allows sufficient lipopolysaccharide to enter
the tissue and activate monocytes? One possibility
is that bacterial growth would bring high levels of
lipopolysaccharide in contact with the gingival
pocket, resulting in the formation of more pocket
epithelium, which would allow more lipopolysac-
charide to enter the tissues. This assumes, however,
that sufficient numbers of neutrophils are no longer
available to separate the bacteria from the epithel-
ium. Another possibility is that the emergence of or-
ganisms such as ? gingiuulis causes bacterial factors
to enter the tissue and interfere with the dynamics
of neutrophil flux into the sulcus (see Darveau et al.
(10) in this volume, page 21). This would reduce the
numbers of neutrophils in the sulcus and shift the
relative balance between neutrophils and the bac-
teria, allowing more bacterial expansion and more
lipopolysaccharide to enter the tissue. Other factors
that alter neutrophil function in the sulcus probably
also allow a rapid bacterial shift that overwhelms the
neutrophils. These factors appear to include intrinsic
neutrophil defects, acquired neutrophil modifiers
such as smoking and factors that alter antibody-neu-
trophil activity such as genetic variations in FcyII re-
ceptors (see Hart & Kornman (23) in this volume,
page 209).
The increasing levels of prostaglandin EZ, IL-1 and
tumor necrosis factor a represent the beginning of a
new stage in the pathogenesis that involves the re-
sponse of monocytes and plasma cells. In children
this stage does not generally occur, unless neutrophil
function is defective, such as that seen in leukocyte
adhesion deficiency syndrome 1. It also suggests that
children do not respond to lipopolysaccharide chal-
lenge with a monocytic and plasma cell infiltrate as
an inflammatory response to local challenge in the
same way as adults. The reason may be that, in gen-
eral, the cellular response to lipopolysaccharide with
resultant inflammatory mediator release is very weak
in children and increases with age. This is in part
due to the important role lipopolysaccharide in the
gut plays in the maturation of the immune system
in children, especially thymus-directed processes.
Indeed, germ-free animals fail to develop normal T-
cell immunity if lipopolysaccharide is not present,
illustrating the key role of lipopolysaccharide as a
co-evolutionary signal for the maturation of the im-
mune system (2). This nonresponsiveness in
children may in part account for the apparent resist-
ance children have to periodontitis.
In young adults with experimental gingivitis, not
all patients behave the same biochemically or clin-
ically. The maturation of the biofilm by 4 weeks and
acquisition of gram-negative anaerobes presents a
challenge that escapes neutrophil clearance in some
patients and results in the penetration of lipopoly-
saccharide. This is suggested by data showing that
some experimental gingivitis patients have prosta-
glandin E2 levels by 4 weeks identical to those in un-
treated adult periodontitis. Other patients have low
levels of prostaglandin E2 that have not increased
from baseline. The central question then is whether
these different responses both represent gingivitis.
Instead it appears to represent a time-dependent es-
calating response to the lipopolysaccharide burden
that is absent in some patients. The magnitude of
the clinical severity would be directly related to the
magnitude of the inflammatory response (that is, the
amount of prostaglandin EZ, IL-1 and tumor necrosis
factor a produced). If gingivitis represents a neutro-
phi1 lesion, then this rise in prostaglandin E2 does
not represent experimental gingivitis but rather ex-
perimental periodontitis that begins with gingival in-
flammation, a nonprotective neutrophil response
and rapid lipopolysaccharide triggering of the in-
flammatory mediators of periodontitis.
In nonsusceptible patients, a protective response
by neutrophils and antibodies limits the extent (and
severity) of attachment loss. In susceptible patients,
neutrophil clearance is less protective and the extent
and severity of disease are governed by the magni-
tude of the host inflammatory response to the mi-
crobial challenge, especially the response to lipo-
polysaccharide. The magnitude of the inflammatory
response is driven by microbial antigens and con-
trolled by T cells. The T-cell response up- or down-
regulates the inflammatory component, plasma cell
differentiation and antibody production based on
229
Pane et al.
the cytokine profile. Antibody is critical to enhance
neutrophil clearance, which reduces the lipopolysac-
charide and antigenic challenge. Thus, the relative
balance between the inflammatory response and the
protective antibody response is regulated by the mix
of cytokines that is produced primarily by macro-
phages and T cells.
Studies of the natural history of new episodes of
periodontal attachment loss show that nonsuscep-
tible patients with existing gingivitis are not at sig-
nificant risk for attachment loss (37). Susceptible pa-
tients with periodontitis are at high risk for new
attachment loss ( 22) . Thus, established gingivitis
does not appear to readily make the transition to
periodontitis. Stated another way, in some suscep-
tible individuals the gingivitis state that precedes
periodontitis appears to have a very transient clin-
ical presentation that is quickly passed along the
evolving path to periodontitis. Further, all sites in
these susceptible patients with periodontitis remain
at risk of further attachment loss, even after treat-
ment. In patients of this type, poor plaque control
and gingival inflammation will virtually always result
in further periodontal breakdown.
In the last decade new appreciation has been
gained of the essential role of bacteria in initiating
periodontitis, but the severity of disease is deter-
mined by the magnitude and quality of the host re-
sponse to the microbial biofilm. An argument can be
made that gingivitis and periodontitis share the
same causal factors but represent different clinical
manifestations of the same fundamental disease
process due to differences in how the host responds
to the microbial challenge. For example, in simplistic
terms the patient plays host to the microbiota and
serves to incubate the common oral flora. The bi-
ofilm follows a predictable maturation process be-
ginning with gram-positive rods and cocci and fol-
lowed by gram-negative anaerobes. Some patients
serve as feeble incubators due to recurrent mechan-
ical disruption of the dental biofilm and effective
host neutrophil defenses. As a result, the natural
maturation of the biofilm is stifled or perhaps ar-
rested and such individuals would present clinically
as gingivitis patients who may have a site or two with
bleeding on probing and some loss of attachment.
Even after many years of exposure from exogenous
sources of periodontal pathogens, such as P gingi-
ualis, the organisms can never really establish well
enough in the environment provided by the host to
emerge as dominant members of the biofilm. Thus,
the patient would remain nonsusceptible.
In contrast, another patient presented with the
same microbial biofiim may serve as a more nurtur-
ing and less combative incubator (see Hart & Korn-
man (23) in this volume), permitting the gram-nega-
tive, black-pigmented, anaerobic species that require
host-derived nutrients and growth factors such as
peptides and heme to occupy the biofilm. Any ex-
ogenous exposure to periodontal pathogens in this
host could cause further attachment loss. Further,
this host environment will never permit stable gingi-
vitis, since the host is more reactive to the bacterial
challenge and the host conditions will drive the
natural maturation of the biofilm, leading to the es-
tablishment of a lush gram-negative biota contain-
ing high levels of periodontopathic bacteria. Gingi-
vitis probably preceded periodontitis in susceptible
patients, but the distinction between the two con-
ditions is probably blurred and transient at best.
Most importantly, the gingivitis that is prodromal to
periodontitis in this circumstance would follow a
distinctly different clinical course than nonpro-
gressing gingivitis. In this context gingivitis and peri-
odontitis really represent two different clinical mani-
festations of the same pathological process. The dif-
ference in clinical manifestations is due to intrinsic
differences in the host that modify simultaneously
both the destructive response to the bacteria as well
as the natural growth and maturation potential of
the biofilm (in this volume, see: Hart & Kornman (23)
and Salvi et al. (60)). Perhaps the wrong question is
being posed in attempting to understand the tran-
sition from gingivitis to periodontitis, since both
conditions may represent the same pathogenic pro-
cess with varying clinical expression in different
people with differing levels of susceptibility.
If this concept is proven valid, it may have some
important ramifications. Individuals susceptible to
periodontitis would have a minimal or no discern-
ible gingivitis phase but go rapidly into initiation of
periodontitis. In such subjects, preventing gingivitis
as a strategy for preventing periodontitis may not be
effective. These individuals may require more ex-
treme control of bacteria andlor host modulation
therapy. In subjects not susceptible to periodontitis,
the gingivitis phase may be so stable that preventing
gingivitis is not necessary to prevent the initiation of
periodontitis. In addition, therapies that block peri-
odontal disease progression may not affect gingivitis.
The effects of nonsteroidal anti-inflammatory drugs
on the progression of periodontal disease may be
such an example, since they are effective in prevent-
ing attachment and bone loss but have only a small
effect on gingival inflammation in susceptible pa-
tients or in experimental gingivitis.
230
Advances in the pathogenesis OJ periodontitis
The cytokine network and
prostaglandins
Characteristics of the cytokine network
The cytokines comprise a large family of polypep-
tides produced in the periodontium by infiltrating
leukocytes and by resident fibroblasts, junctional
epithelial cells, vascular endothelium, mast cells and
bone cells. The cytokines comprise the major regu-
lators of the immunoinflammatory response charac-
teristic of periodontitis, and they are the major de-
terminants of the outcome. The interleukins are a
subfamily of cytokines designated interleukin- 1 (IL-
1) through interleukin-15 (IL-15). The chemokines
comprise an additional subfamily of smaller pep-
tides that serve as chemoattractants for specific
leukocytes. For example, IL-8 is specific for neutro-
phils and monocyte chemoattractant protein 1 is
specific for monocytes, whereas other chemokines
are specific for other mononuclear cells. Other cyto-
kines that are important in the immunoinflamma-
tory response in periodontitis include interferon-y
(produced by activated T cells, monocytes, natural
killer cells) and transforming growth factor p (pro-
duced mostly by monocytes and macrophages).
Cytokines can induce one another and, in some
Table 5. The pathogenesis of periodontitis:
new findings and new perspectives on cytokines
and prostaglandins
Thecytokines, chemokines and prostaglandins
appear to comprise the major regulators of the
imunoinflammatory responses that characterize
periodontitis.
Different cytobneytterns haw been observed in
tissues from erio onhtis sites, suggesting that
specific &tory patterns may be Invoked in
stable and active lesions.
The disease active lesions do not appear to have
cytokine patterns that are characteristic of either a
Thl or a ThZ pattern.
The proinflammatory cytokines K - l p and tumor
necrosis factor a have been strongly associated
with periodontltis.
Irost@andin Ez is E primary mediator of tissue
destruction inperiodontitis.
Prostaglandin Ez levels are elevated in iridividuals
with high susceptibility for severe periodontitis.
Prostaglandin Fq levels are determined by the
pmstagiandin endoperoxide synthase 2
(cyclooxygenase 2) enzyme, which is W y
regulated by lipopolysaccharide, and various
cytokines, including most prominently IL-10 and
tumor necrosis factor a.
~ _ _ ~ . ~~ .~ ~ ~~ ~ ~~
~_ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ ~_
cases, such as IL-lp, perpetuate production of them-
selves. Important new findings and perspectives are
listed in Table 5.
The cytokines act at very low concentrations gen-
erally in the range of 10-lo to lo- M. They serve
as messengers that mediate cross-talk among cells
by binding to high-affinity transmembrane recep-
tors. Some cytokines such as IL-1 and tumor ne-
crosis factor a are proinflammatory, whereas others,
such as IL-10, suppress the inflammatory response.
Cytokines can act synergistically or antagonistically
or be additive or subtractive. They function as a cas-
cading network that manifests unusual complexity,
redundancy and pleotrophism. Many of the func-
tions of cytokines depend on concentration. These
features of the network are difficult to explain but
may be accounted for by the following explanation.
Bacteria have very short generation times that per-
mit frequent mutation. As a consequence, they can
evolve a variety of mechanisms to avoid and ma-
nipulate the host defense mechanisms. Further,
whereas an encounter between the host defenses
and a single species may be reasonably well defined,
in periodontitis the ecosystem is exceedingiy com-
plex, and co-evolutionary processes between diverse
bacteria species and the host most likely have al-
ready produced a variety of host molecules to cope
with the bacterial challenge. The complexity, redun-
dancy and pleotrophism of the cytokine network
may provide a mechanism to finely tune the host
response such that it properly limits the bacterial
challenge but is modulated to limit tissue destruc-
tion and allow repair.
The Thl and Th2 paradigm: the cytokine profile
directs the nature of the host response and
regulates connective tissue homeostasis
In the mouse, CD4+helper T cells fall into two dis-
tinct categories designated Thl and Th2 cells based
on the pattern of cytokines they produce (in this vol-
ume, see: Gemmell et al. (17), pages 113-115 and
Ishikawa et al. (271, pages 90-91). It has been as-
sumed that naive antigen-specific T-helper cells,
designated Tho, differentiate on repeated stimula-
tion into Thl and Th2. This concept has been ques-
tioned and may not be true in humans. Thl clones
are generally characterized by the production of in-
terferon y. A predominantly Thl infiltrate is thought
to result in a cell-mediated or delayed-type hyper-
sensitivity response with the activation of macro-
phages, the production of proinflammatory cyto-
kines, including IL-1 and tumor necrosis factor a.
231
Page et al.
Interferon y stimulates neutrophils and appears to
be essential for neutrophils to control bacterial in-
fection. On the other hand, IL-1 and other proin-
flammatory cytokines are found in inflamed peri-
odontal tissues at significantly elevated levels, and
these levels return to normal following successful
therapy. Monocytes from patients with severe peri-
odontitis and those with high susceptibility produce
more IL- 1 before and following stimulation than
cells from periodontally normal individuals (15, 16,
58). A major role for I L-l is its direct action on the
promoter region of the matrix metalloproteinase
gene, resulting in enhanced matrix metalloproteina-
se production. Individuals susceptible to severe
adult periodontitis later in life manifest a unique
polymorphism in the IL-1 gene family that results in
the production of approximately four times more IL-
l p than individuals not having that genetic trait, and
their likelihood of developing severe disease in-
creases significantly (31). IL-6 may also be important
in periodontitis, as it can induce the formation of
multinucleated cells that can resorb bone.
Th2 clones are generally characterized by the pro-
duction of IL-4, IL-5 and IL-6. Both Thl and Th2
cells produce other cytokines such as IL-10 and IL-
13, tumor necrosis factor a and granulocyte-macro-
phage colony-stimulating factor. A predominantly
Th2 infiltrate is considered to drive the differen-
tiation of B cells into antibody-producing plasma
cells, which may be protective or nonprotective de-
pending on the nature of the antibody produced. A
strong Th2 response can inhibit a Thl response and
vice versa. For example, IL-4 can inhibit the Thl re-
sponse, and interferon y can inhibit the Th2 re-
sponse. Whether a given response is Thl or Th2 or
mixed is determined by several factors. These in-
clude the nature of the activating antigen, the route
of antigen administration and antigen concen-
tration, the identity of the antigen-presenting cell,
major histocompatibility complex molecules and the
nature of the interaction with the T-cell receptor. The
nature of the antigen-presenting cell, however, may
be of fundamental importance.
Whether the Thl and Th2 paradigm exists in
humans as it is manifested in mice remains un-
proven and controversial. This is a key question for
periodontal pathogenesis since the Thl and Th2
paradigm could be a major determinant of the pro-
gression from gingivitis to periodontitis, of the tran-
sition from a stable nonprogressive periodontitis
lesion to a burst of destructive activity and of refrac-
tory or recurrent disease. The Thl and Th2 paradigm
has been studied by measuring the levels of various
232
cytokines in gingival crevicular fluid and extracts of
healthy and diseased gingival tissue, determining
levels of cytokine messenger RNA in the tissue using
immunocytochemical techniques or polymerase
chain reaction on tissue extracts and by measuring
cytokine production by mononuclear cells harvested
from gingival tissue. Based on their work and pub-
lished studies by others, Gemmell et al. (17) postu-
late in this volume that the Thl response is associ-
ated with stable periodontal lesions, whereas a Th2
response leads to the production of nonprotective
antibody and disease progression. They have further
suggested that a mixed Thl and Th2 lesion in which
both IL-4 and interferon y are present may, depend-
ing on the relative concentration of each cytokine,
lead to the production of protective antibody. Other
investigators have postulated the opposite, and still
others have observed no skewing toward either Thl
or Th2. The data appear to be inconclusive. The pat-
tern of cytokine production may result from a dis-
tinct cell phenotype, a specific state of cell differen-
tiation or a transitional state. Which, if any, of these
is correct is currently unclear.
The controversial nature of the Thl and Th2 para-
digm studies may result from the complexity of the
experiments. Humans vary enormously in suscepti-
bility to periodontitis and rates of disease pro-
gression. Tissue destruction appears to occur in site-
specific bursts of activity that are highly variable and
seem to be infrequent in most patients but common
in others. There are no methods for distinguishing
between disease-active and -inactive sites except by
rather crude methods of assessing clinical attach-
ment loss or alveolar bone destruction over time.
These factors have not been adequately controlled
nor have they been taken into account in the Thl
and Th2 studies. The data appear to support the
conclusion that Tho, Thl and Th2 clones participate
in periodontitis, and there is little evidence that any
single T-helper cell subset is associated with any spe-
cific clinical status. As stated by Gemmell et al. in
this volume (17), a focus on which cytokines are
present, their local concentrations and hence their
biological activity may be more fruitful than con-
tinuing to focus on helper cell clonal type.
This makes it clear that the levels of various cyto-
kines are the chief determinants of the progression
or suppression of periodontitis. What remains ob-
scure is the identity of the agents and the mechan-
isms that determine which cytokines or antagonists
are produced, when and how long they are produced
and their concentrations. Further elucidation of the
details of the cytokine cascade will indicate new ap-
Advances in the pathogenesis of periodontitis
proaches to prevention, diagnosis and treatment of
periodontitis.
Arachidonic acid metabolites are key components
of the pathogenesis of periodontitis
Arachidonic acid metabolites, especially prosta-
glandin EL, are important mediators of the immuno-
inflammatory response in periodontitis (see Gem-
me11et al. (17) in this volume, pages 126-129). Al-
though prostaglandins can be produced by most
nucleated cells and by blood platelets, the major
source in inflamed gingiva is activated macrophages
and fibroblasts. Arachidonic acid is generated by the
action of phospholipase A2 on cell phospholipids. It
is a substrate for the cyclooxygenase enzyme sys-
tems (cyclooxygenase 1 and cyclooxygenase 2, also
referred to as prostaglandin endoperoxide synthase 1
and 2) that produce a large family of prostaglandins,
thromboxane and prostacycline. It is also a substrate
for the lipoxygenases that give rise to the leuko-
trienes, the most important one of which (for peri-
odontal disease) is leukotriene Bq.
Prostaglandins and leukotrienes are major in-
itiators and perpetuators of inflammation, and
prostaglandin E2 is the major mediator of patholog-
ical alveolar bone destruction (see Schwartz et al.
(61) in this volume, pages 166-168). Prostaglandins,
especially prostaglandin EZ, are present in gingival
crevicular fluid from disease active periodontal sites
at concentrations five or more times greater than in
samples from known inactive sites, and whole-
mouth scores for prostaglandin E2 have been re-
ported to be a marker for concurrent or future dis-
ease activity. Successful therapy markedly decreases
prostaglandin E2 levels.
Mononuclear cells from patients highly suscep-
tible to severe periodontitis release significantly
more prostaglandin E2 than do cells from resistant
individuals. Release is stimulated by lipopolysac-
charide and by cytokines such as IL-1p. Prosta-
glandin E2 participates along with cytokines in regu-
lating IgG production. High concentrations inhibit
and low concentrations acting synergistically with
IL-4 enhance IgG production. Prostaglandin E2 has
major catabolic effects on gingival and periodontal
ligament fibroblasts. It decreases DNA synthesis and
cell growth and the production of collagen and non-
collagen proteins. Prostaglandin Ez also inhibits
fibroblast IL-6 production induced by TL-1 p or tumor
necrosis factor a. In both experimental periodontitis
in monkeys and spontaneously occurring peri-
odontitis in dogs, a rise in prostaglandin E2 is associ-
ated with a worsening clinical status. Rising prosta-
glandin E2 levels can be inhibited in experimental
animals and humans by systemic or topical appli-
cation of nonsteroidal anti-inflammatory drugs,
which also inhibit bone loss. The cytokine cascade
and arachidonic acid pathways present numerous
opportunities to intervene therapeutically in the
progression of periodontitis (see Gemmell et al. (17)
in this volume, pages 129-132).
Mechanisms of tissue destruction
in periodontitis
One of the highlights of the past two decades of peri-
odontal research is the major progress made in
understanding the basic mechanisms by which bac-
teria present in subgingival biofilms and their prod-
ucts and components mediate the pathological
changes characteristic of periodontitis. These path-
ways are now understood in sufficient detail to begin
to devise treatments that alter events and outcomes.
Alterations in the junctional epithelium have already
been considered, and the focus now turns to path-
ological alterations in components of the extracellu-
lar matrix of the gingiva and periodontal Iigament
and resorption of alveolar bone. Important new
findings and perspectives are listed in Table 6.
Table 6. The pathogenesis of periodontitis:
new findings and new perspectives on the
mechanisms of tissue destruction
Periodontitis involves the destruction of bone and
connective tissues, including collagens,
proteoglycans, and other components of the
extraceuular matrix
4 The tissue destruction is not unidirectional, but
rather is an iterative process that is constantly
being adjusted by the host-bacterial interactions.
determined by the balance of matrix
metalloproteinases and their inhibitors.
Thebalance of matrix metalloproteinases and
inhibitors is regulated locally by exposure to IL- l a,
IL-lp, IL-10, transforming growth factor p and
lipopolysaccharide.
Alveolar bone destruction in periodontitis is a result
of uncoupling of the normally tightly coupled
processes of bone resorption and bone formation.
Tissue prostaglandin &, IL-1s and, to a lesser extent,
tumor necrosis factor a appear to mediate bone
resorption in periodontitis. IL-6 may also be
involved.
Circulating factors, including the steroid hormones,
parathyroid hormone, calcitonin, and vitamin D3,
regulate the overall bone remodeling process.
The destruction of the extracellular matrix is
233
Page et al.
Summary of the process
The changes in the connective tissues consist pre-
dominantly of destruction of the collagens, proteo-
glycans and other components of the extracellular
matrix. Evidence of destruction is first seen in the
region of the vascular plexus just deep to the junc-
tional epithelium. As disease worsens, the area of in-
flamed connective tissue enlarges and eventually
comes to occupy virtually all the marginal gingiva.
The inflammatory infiltrate does not generally ex-
tend into the periodontal ligament. Instead, the
extracellular matrix of the ligament is destroyed as
alveolar bone is resorbed. As the disease progresses,
the fibrous connective tissue is replaced with a
dense infiltrate of inflammatory cells, and alveolar
bone is destroyed. Destruction of the extracellular
matrix is generally viewed as pathological and unde-
sirable, but it is an integral and essential part of the
immunoinflammatory process since it is necessary
to create space for occupation by the enlarging
populations of infiltrating leukocytes. In advanced
lesions of long standing, manifestations of fibrosis
and the scarring characteristic of chronic inflamma-
tory diseases can be seen histologically and clin-
ically. Specific species of bacteria in the subgingival
biofilm on the roots of the teeth ultimately initiate
the destruction of the extracellular matrix and al-
veolar bone, and some species such as l? gingiualis
produce families of potent enzymes, including col-
lagenase. However, virtually all the collagenases
found in periodontally diseased tissues derive from
host cells, not the bacteria. The bacteria drive de-
struction by attracting and activating host cells that
carry out the observed destruction.
Although most of the extracellular matrix of the
gingiva and periodontal ligament is destroyed and
variable amounts of alveolar bone resorbed in ad-
vanced periodontitis, the progress of the disease is
not one dimensional or unidirectional. I t is iterative
and constantly being adjusted as a result of multiple
and changing microbial challenges and multiple lo-
cal and systemic host defenses. Thus, in addition to
the destructive activity, manifestations of attempts
at regeneration and healing are also apparent in his-
tological sections. Destruction waxes and wanes
with increases and decreases in the extent of immu-
noinflammation, changes in the numbers and pro-
portions of various inflammatory cells and their sub-
sets in the tissue and fluctuation in the levels of vari-
ous cytokines, prostaglandins and effector and
inhibitor molecuies such as the matrix metalloprot-
einases and the tissue inhibitors of matrix metallop-
roteinases. The iterative nature of the pathogenesis
is reflected clinically by periods of apparent quiesc-
ence of various lengths followed by bursts of de-
structive disease activity.
Although many details are lacking, the overall
events that result in connective tissue destruction
and, to a lesser extent, alveolar bone resorption are
reasonably well understood. Antigenic and other
components of periodontopathic bacteria, especially
lipopolysaccharides, stimulate the infiltrating leuko-
cytes and resident cells to produce cytokines and
prostaglandins, and they induce production of the
matrix metalloproteinases and other proteolytic en-
zymes and additional mediators by monocytes and
macrophages, resident fibroblasts, junctional epi-
thelium and endothelium. The infiltrating neutro-
phils also participate. Collectively, these molecules
mediate destruction of the connective tissue and al-
veolar bone. Anillustration of the overall functioning
of these events is presented in Fig. 1.
Destruction of the extracellular matrix of the
gingiva and periodontal ligament is determined by
the balance of matrix metalloproteinases and their
inhibitors
The matrix metalloproteinases comprise a large fam-
ily of related gene products with extensive homo-
logies (see Table 1 and Fig. 2 in Reynolds & Meikle
(57) in this volume). These enzymes are designated
currently as matrix metalloproteinases 1 through 17,
and the list is still growing. They fall into classes in-
cluding collagenases, gelatinases, stromelysins, ma-
trilysins, metalloelastase and membrane-bound
matrix metalloproteinases. All the matrix metallop-
roteinases are produced in latent precursor form and
require activation by plasmin or plasmin-like en-
zymes. Their activities depend on divalent cations.
Collectively, these enzymes have the capacity to de-
grade all the components of the extracellular matrix.
In the periodontium, matrix metalloproteinases are
produced by macrophages and resident tissue
fibroblasts, junctional epithelial cells and neutro-
phils activated by bacterial substances such as lipo-
polysaccharide, proinflammatory cytokines and
prostaglandins (Fig. 1). Because of their large num-
bers, the fibroblasts are a major source. Neutrophils
carry a pre-made matrix metalloproteinase that can
be released. Because of their continuous influx, they
too are a major source. At very early stages of in-
flammation, neutrophil matrix metalloproteinase
predominates. Fibroblasts can also express a mem-
brane-bound matrix metalloproteinase that could
234
Advances in the pathogenesis of periodontitis
~~~
function to provide the initial attack on intact native
collagen fibers in the absence of an inflammatory
infiltrate.
As might be expected for enzyme activities of high
destructive potential, matrix metalloproteinase pro-
duction and activity are tightly controlled and regu-
lated. Tissue inhibitors of matrix metalloproteinases
play a major role, with tissue inhibitor of metallopro-
teinase 1 being the most important. Tissue inhibitors
of metalloproteinases bind to matrix metalloprotein-
ases and irreversibly inhibit enzyme activity. The ex-
tent of degradative activity in the tissue is largely a
function of the balance between the levels of matrix
metalloproteinases and tissue inhibitors of metallo-
proteinases. It has been suggested that tissue degra-
dation only occurs at locations where the levels of
tissue inhibitors of metalloproteinases are low (see
Reynolds & Meikle (57) in this volume, pages 147-
149). Fibroblasts and macrophages are the major
sources of tissue inhibitors of metalloproteinases as
well as matrix metalloproteinases, and the nature of
the message received by these cells is a major deter-
minant of outcome.
The overall regulation of matrix metalloproteinase
and tissue inhibitors of metalloproteinase produc-
tion in the periodontium is complex and involves
cytokines, hormones, growth factors, prostaglandins
and bacterial factors such as lipopolysaccharide (see
Reynolds & Meikle (57) in this volume, page 149).
Gingival fibroblasts produce prostaglandin E2 but
not matrix metalloproteinase following exposure to
lipopolysaccharide; they produce matrix metallopro-
teinase following exposure to I L-lp (Fig. 1). In gen-
eral, proinflammatory cytokines such as IL-lP up-
regulate matrix metalloproteinase and downregulate
tissue inhibitors of metalloproteinase production,
whereas other cytokines such as IL-10 downregulate
matrix metalloproteinase and upregulate tissue in-
hibitors of metalloproteinase. IL- 1 p upregulation ap-
pears to occur through stimulation of nuclear factor
KB proteins and thereby acts on the promoter region
of the matrix metalloproteinase gene. Transforming
growth factor p is especially important in downregul-
ating matrix metalloproteinase and upregulating
tissue inhibitors of metalloproteinase. Cell-associ-
ated I L-l a is a potent stimulator of matrix metallop-
roteinase release by other cells. These observations
demonstrate that the balance of cytokines in the gin-
gival tissue is the major determinant of whether
tissue degradation occurs or homeostasis is main-
tained, and IL-1s plays a key role.
There is strong evidence that IL-1-induced matrix
metalloproteinases participate in degrading the
extracellular matrix in periodontitis and that tissue
inhibitors of metalloproteinase, IL-10 and trans-
forming growth factor p participate in resolving in-
flammation and maintaining homeostasis. Gingival
crevicular fluid from sites of active disease have elev-
ated levels of IL-lP and matrix metalloproteinase
and low Ievels of tissue inhibitors of metalloprotein-
ase, and the same is true for diseased versus healthy
gingival tissue. Cultures of peripheral blood mono-
cytes exposed to lipopolysaccharide produce IL- 1 s
and matrix metalloproteinases, and cells from indi-
viduals who are highly susceptible to periodontitis
are more responsive than are cells from resistant in-
dividuals. Successful treatment results in a signifi-
cant decrease in leveIs of matrix metalloproteinase
and I L-1p in the tissue and gingival crevicular fluid
and in increased levels of tissue inhibitors of metal-
loproteinase.
Several approaches for treating periodontitis by
decreasing the degradation of the extracellular mat-
rix are apparent (see Reynolds & Meikle (57) in this
volume, pages 152-153). The most studied of these
is the use of inhibitors of matrix metalloproteinase
activity based on the fact that activity depends on
divalent cations. Chemically modified tetracycline
and low-dose doxycycline inhibit matrix metallopro-
teinase activity by chelating divalent cations. Low
doses prevent collagen breakdown and alveolar bone
loss in animal studies and result in reduction in
pocket depth and clinical attachment loss in human
clinical trials. These drugs appear to work by binding
enzyme-associated calcium ions, making the en-
zyme more susceptible to proteolysis. Hydroxamic
acid derivatives also inhibit matrix metalloproteina-
se activity, but their usefulness in periodontitis has
not been studied. The isothiazolones can inhibit car-
tilage proteoglycan degradation without decreasing
synthesis. These drugs appear to target the acti-
vation step of matrix metalloproteinase without
being active against the active enzyme. There has
been no work on the use of inhibitors of phos-
pholipase A2, which controls the rate-limiting step in
prostaglandin production.
Mechanisms of destruction of alveolar bone
Progress in elucidating details of the mechanism of
bone resorption and regeneration in periodontitis
has lagged behind considerably the progress related
to soft tissue. Far less progress has been made in
understanding the basic events accounting for al-
veolar bone loss in periodontitis. Certain facts are
well established (see Schwartz et al. (61) in this vol-
235
Page et al.
ume, pages 160-164). Bone loss in periodontitis is a
result of an uncoupling of the normally tightly
coupled processes of bone resorption and bone for-
mation; enduring pathogenic levels of prostaglandin
E2, IL- 10 and, to a lesser extent, tumor necrosis fac-
tor a mediate the loss. IL-6, a cytokine that mediates
the formation of multinucleated cells that resorb
bone, may also be involved. It is also well docu-
mented that the activities of osteoblasts and osteo-
clasts and their precursor cells are highly integrated
and coordinated with one another and with osteo-
cytes. There appear to be many feedback loops
among these cells, with osteoblasts producing fac-
tors that enhance osteoclasts and vice versa. Details
of these cell-cell and matrix-cell interactions even in
normal bone remain obscure. All these are regulated
in an overall long-term sense by circulating factors
such as steroid hormones, parathyroid hormone,
calcitonin and vitamin DS. Fine tuning is achieved
by events occurring locally in bone such as the re-
lease of IL-l p, prostaglandin EZ, IL-6 and various
other locally produced proteins and factors such as
bone morphogenic proteins, fibroblast growth fac-
tor, transforming growth factor p, platelet-derived
growth factor and epidermal growth factor released
from resorbing bone matrix. These enhance the
attachment and replication of various bone cells and
regulate their activity. Virtually nothing is known
about the effect on these events of substances de-
rived from infecting periodontopathic bacteria or
the large families of cytokine and regulatory mol-
ecules produced during active disease progression.
A top priority for further periodontal research must
be elucidating the details of normal and abnormal
bone physiology, much as has already occurred with
the mechanisms of destruction and reconstitution of
the extracellular matrix.
There is strong evidence that prostaglandin E2 is
the major mediator of alveolar bone loss in peri-
odontitis and that IL-lp, tumor necrosis factor a and
IL-6 also play important roles (see Schwartz et al.
(61) in this volume, pages 166-168). These mediators
can be produced by resident fibroblasts and leuko-
cytes in the inflammatory infiltrate, especially
macrophages. They are found in the inflamed gingi-
val tissue and gingival crevicular fluid in elevated
concentrations, and these concentrations decrease
following successful therapy. Inhibition of prosta-
glandin E2 production suppresses or blocks alveolar
bone destruction in both experimental animals and
spontaneously occurring periodontitis in humans.
Nevertheless, details of how this group of mediators
actually functions in normal bone in the periodon-
236
~
tium and in bone destruction in periodontitis re-
main elusive, and they are sometimes contradictory.
For example, there is evidence that prostaglandin E2
actually stimulates bone apposition in vivo in some
situations. The roles of calcitonin and vitamin D3 as
well as factors released locally by bone resorption
that inhibit resorption and stimulate formation, such
as transforming growth factor p, matrix metalloprot-
einase, fibroblast growth factor, and insulin-like
growth factor, remain obscure.
Agehas dramatic effects on bone: both men and
women lose bone mass after age 25-30 years. The
prevalence and severity of periodontitis begin to in-
crease considerably after these same ages. Gender
and hormone levels also appear to affect bone sta-
tus. Both estrogens and androgens affect bone sta-
tus, but loss of bone mass is greatly accelerated
among postmenopausal women not on estrogen re-
placement therapy. Nevertheless the prevalence and
severity of periodontitis are greater among aging
men than among aging women. The relationship be-
tween osteoporosis and periodontitis remains ob-
scure.
Regeneration of periodontal attachment including
alveolar bone, cementum and functional peri-
odontal ligament has rapidly become a frequently
attempted treatment procedure known as guided
tissue regeneration. Factors known to be related to
bone regeneration including bone morphogenic pro-
teins, platelet-derived growth factor, insulin-like
growth factor 1 and transforming growth factor p
have to be assessed in experimental animals. The
role of these factors in the mechanisms underlying
regeneration and the factors involved in its success
or failure remain unknown.
A wide range of potential approaches exist for al-
tering alveolar bone destruction in periodontitis (in
this volume, see: Schwartz et al. (61), pages 166-168
and Reynolds & Meikle (57), pages 152-153). Prosta-
glandin E2 is well documented to participate in the
immunoinflammatory response in periodontitis in
mediating connective tissue alterations and bone re-
sorption, and the nonsteroidal anti-inflammatory
drugs effectively inhibit these activities. Neverthe-
less, these drugs are not currently used in treating or
preventing periodontitis. When they are adminis-
tered systematically over long time periods, nephro-
toxicity and other undesirable side effects can occur.
There is strong evidence that their use topically in
oral rinses or toothpaste is effective in slowing or ar-
resting periodontitis, although more clinical trial
data are needed.
The prostaglandin family of mediators is pro-
Advances in the pathogenesis of periodontitis
duced mainly by the cyclooxygenase 1, which is ex-
pressed constitutively. At sites of inflammation,
cyclooxygenase 2, a recently discovered pathway that
is not produced constitutively, can be induced by
lipopolysaccharide, IL- 1 p and tumor necrosis factor
a (82). Cyclooxygenase 2 is attracting considerable
attention as a potential target for intervention since
blocking cyclooxygenase 2 could significantly reduce
the production of prostaglandins specifically at sites
of inflammation and tissue destruction. Radicicol, a
fungal antibiotic, inhibits cyclooxygenase 2 probably
by inhibiting protein tyrosine kinase. Substrate ara-
chidonic acid is made available by the action of
phospholipase A2 on cell phospholipids. Cortico-
steroids inhibit phospholipase A*, probably through
lipocortin, and can thereby block the production of
prostaglandins and leukotrienes. Whether any of
these drugs will prove useful in controlling peri-
odontitis remains to be seen. There is a new family
of drugs known as the cytokine-suppressing anti-in-
flammatory drugs (see Reynolds & Meikle (57) in this
volume, pages 152-153). The prototype is SKF86002.
These drugs selectively inhibit the protein kinase
family designated RK38, or CS binding protein. This
family is activated by lipopolysaccharide, proin-
flammatory cytokines or physicochemical stress. Ac-
tivation is required for production of IL-1, tumor ne-
crosis factor a and probably prostaglandins. Cyto-
kine-suppressing anti-inflammatory drugs strongly
inhibit cyclooxygenase 2. The potential role of these
new agents in treating periodontitis will undoubt-
edly be an area of active investigation in the future.
The transition from a stable
periodontal site to active disease
progression
Prior to the 1980s, the presence of periodontal
pockets was equated with active periodontitis. Sev-
eral longitudinal studies conducted in the 1980s (29,
35, 36) modified this view. These studies demon-
strated clearly that stable, disease-inactive as well as
progressing disease-active periodontal pockets exist.
Most periodontal sites (usually more than 90%) in
most adult periodontitis patients are disease inactive
and nonprogressing, and the occurrence of disease
active sites is relatively rare and episodic. Most dis-
ease-active sites occur in a very small proportion of
the patients. These observations focused attention
on the nature of the transition from a stable peri-
odontal site to active disease progression. A clear
understanding of this transition could shed con-
siderable light on better approaches to prevention
and therapy.
The transition from a stable to a disease-active
site may or may not be the same as the transition
from gingivitis to periodontitis. Indeed, as discussed
in the previous section, gingivitis lesions and peri-
odontitis lesions may differ considerably from the
very beginning of the microbial challenge. This
would not be the case for the transition from an al-
ready existing but stable periodontal pocket to a dis-
ease-active site. If periodontal disease is to progress,
the microbial challenge, subgingival biofilms, must
continue to be present, although in some individuals
the magnitude of the challenge may be surprisingly
small in quantity. In addition. there must be per-
sisting inflammation and destruction of the com-
ponents of the extracellular matrix manifested as
loss of clinical attachment and a net loss of alveolar
bone. Pockets may or may not become deeper.
Theoretically, the characteristics of disease pro-
gression may be expected to consist of the continu-
ing presence of lipopolysaccharide and other bac-
terial components, an inflammatory cell infiltrate,
high levels of proinflammatory cytokines including
IL-1p and tumor necrosis factor a and low levels of
IL-10 and transforming growth factor p, cytokines
that suppress the immunoinflammatory response,
low levels of tissue inhibitors of metalloproteinases
and high levels of matrix metalloproteinases and
prostaglandin EP. Macrophages, fibroblasts and to a
lesser extent other resident and inflammatory cells
activated by lipopolysaccharide and cytokines, es-
pecially IL-lf3 and tumor necrosis factor a, produce
matrix metalloproteinases and prostaglandin EZ, and
can maintain them at pathogenic levels. The ques-
tion now becomes, what factors and events regulate
and determine production and levels of these me-
diators? The following represents speculation on
mechanisms that can explain disease progression.
Loss of the chemotactic gradient
In a stable periodontal pocket, normal host defense
contains the microbial challenge. Leukocytes, mostly
neutrophils, leave the small vessels and migrate
through the connective tissue and pocket epithelium
to form a wall of viable cells between the biofilm
and the pocket wall (see Fig. 7 in Kornman et al. (32)
in this volume). This emigration is mediated by a
gradient of chemoattractant molecules consisting of
IL-8 produced by the epithelial cells and peptides of
the N-formyl-methionyl-leucyl-phenylalanine type
237
Page et al.
emanating from the biofilm. The leukocytes follow
this gradient and do not become activated until their
low-affinity receptors became occupied by high con-
centrations of mediators they encounter as they
reach the biofilm. Upon activation, these cells phag-
ocytose and kill bacteria and release chemoattrac-
tants and killing agents such as oxygen intermedi-
ates. Under conditions that disrupt the chemoat-
tractant gradient, neutrophils recognize no guidance
system and accumulate within the connective tissue,
where they become activated and release enzymes
such as matrix metalloproteinase that destroy the
tissue, resulting in attachment loss and bone loss.
Further, loss of the gradient can preclude contain-
ment of the microbial challenge by neutrophils and
antibody and can allow lipopolysaccharide to
shower the connective tissues. Gemmell et al. (18)
have provided evidence for loss of the chemoattract-
ant gradient by showing that, as inflammation
worsens, pocket epithelium may become negative
for intercellular adhesion molecule 1, and neutrophil
migration through the epithelium and into the
pocket decreases. Loss of the gradient could also re-
sult from changes in the composition of the biofilm
or production of high levels of chemoattractant
within the inflamed connective tissue. This hypo-
thesis has not been proven but appears to merit
further investigation.
Catabolic activities of resident cells
The resident tissue fibroblasts and junctional epi-
thelial cells may be hijacked by the bacterial chal-
lenge and convert from their normal activities of cre-
ating and maintaining tissue homeostasis and health
to destroying the structures they initially created and
maintained. There is evidence for this hypothesis, as
presented on pages 226-227 of this chapter. The
combinations and concentrations of bacterial sub-
stances (lipopolysaccharide) and various mediators
that cause the resident cells to convert from an ana-
bolic to an catabolic state may herald the transition
from a stable to a disease active lesion.
Thl and Th2 lymphocyte phenotype
The nature of the Thl and Th2 responses may be a
major determinant of the progression to destructive
periodontitis. Although studies aimed at determin-
ing the possible association between the Thl and
Th2 lymphocyte phenotype ratios and disease status
have been performed, the results are not conclusive.
In the active periodontitis lesion, Th2 cells appear to
participate increasingly such that both Thl and Th2
like cells are probably involved. In this context, pro-
gression of the disease is probably related to the bal-
ance and levels of both Thl and Th2 cytokines. High
levels of I L-lP and tumor necrosis factor a would
drive active disease, whereas high levels of IL-10
would downregulate inflammatory cytokines, such
as I L-lP production by macrophages, and hence re-
duce the potential for tissue destruction.
Anergy
Recent data from experimental animals suggest that
anergy may be an important component of both the
transition from gingivitis to periodontitis as well as
progression from a stable to an active lesion. Presen-
tation of antigen by nonprofessional antigen-pre-
senting cells such as epithelial or endothelial cells,
results in anergy (24, 75). The mechanism of this
anergy is unknown, but it is well established that
junctional and pocket epithelial cells express major
histocompatibility complex class I1 antigens (67) and
hence could be capable of antigen presentation,
which results in anergy and possible apoptosis of the
Thl cells and thereby allows cells with a Th2 cyto-
kine profile to emerge. Endothelial cells, which in
humans constitutively express major histocompat-
ibility complex class 11, could also present antigen
leading to anergy (75). In this context, regulation of
the immune response may reside at the level of anti-
gen presentation such that presentation by pro-
fessional antigen-presenting cells (such as dendritic
cells and macrophages) results in stimulation of de-
structive cytokines while presentation by nonpro-
fessional antigen-presenting cells (such as epithelial
or endothelial cells) results in anergy and no tissue
destruction. There seems little doubt that levels of
proinflammatory cytokines, matrix metalloprotein-
ases and prostaglandin E2 increase greatly and tissue
inhibitors of metalloproteinases decrease in tissues
manifesting periodontitis relative to those with gin-
givitis. Additional studies need to be performed to
test this concept.
Leukocyte emigration
Suppression of leukocyte diapedesis and migration
may be involved in determining whether disease
progression occurs. Administration of anti-neutro-
phi1 serum or drugs that cause severe neutropenia
to rats or dogs results in the rapid onset and pro-
gression of periodontal tissue destruction (in this
volume, see: Ishikawa et al. (27) and Dennison & Van
238
Advances zn the pathogeneszs oJ perzoaonnru
Dyke (11) page 67) (1). The release of soluble E-se-
lectin by endothelial cells and the shedding of L-se-
lectin can modulate the diapedesis of leukocytes
such as neutrophils. Decreased L-selectin levels have
been reported for neutrophils from gingival capillary
blood from patients as compared with cells from pe-
ripheral blood (33). These differences were not ob-
served in neutrophils from healthy subjects, sug-
gesting a change in the selectin-mediated ability of
neutrophils to interact with endothelial cells in the
chronically inflamed gingiva. Bacterial species in the
biofilm produce large quantities of polyamines that
can inhibit neutrophil migration (79) and perhaps
most importantly, I? gingiudis has been shown to
block multiple aspects of the endothelial signaling
process critical to local recruitment of neutrophils,
which suppresses the primary host defense system
protecting against the bacterial challenge. The enor-
mous importance of neutrophils in fending off the
pathogenic effects of the subgingival bioflm has
been amply demonstrated (see Dennison & Van
Dyke in this volume (1 I) ) .
Changes in the composition of the biofilm and
invasion
The transition from gingivitis to periodontitis may
be related to changes in bacterial composition of the
biofilm or invasion of the tissue by bacteria from the
biofilm. In other words, the bloom of certain species
in the flora may be causally related to the burst of
destructive activity characteristically observed in
periodontitis.
Environmental and acquired risk
factors
A new paradigm for periodontal diseases is emerg-
ing. Progress in understanding these diseases can be
summarized more or less by decade. In the 1960%
the major development was the demonstration that
bacteria in dental plaque cause human gingivitis and
periodontitis (30, 34, 38). In the 1970% specific spe-
cies of predominantly gram-negative, anaerobic bac-
teria were associated with these diseases (28, 44, 71,
7 4 , and a wealth of evidence was presented that the
immune system of people with periodontitis could
recognize the antigens of these infecting bacteria
(28). The 1980s saw strong documentation of the as-
sociation of specific bacteria with active tissue de-
struction, beginning characterization of the immune
response to antigens and mitogens of the infecting
bacteria and beginning elucidation of the role of
cytokines and prostaglandins in the pathogenesis
(19, 21, 49, 56, 68). It was also during the 1980s that
the site-specific nature of periodontitis was demon-
strated (72), that the prevalence of disease, especially
severe disease, was shown to be much lower than
previously believed (61, and that progression of
tissue destruction at specific sites was episodic and
relatively infrequent (29, 35,361. Significant progress
was made in elucidating the pathways of connective
tissue destruction and alveolar bone resorption (5).
Prior to the 1980% periodontitis was thought to be
universally prevalent in human adults by early
middle age. Because of this, no thought was given to
important roles for factors other than bacteria. The
role and importance of bacteria in the etiology of
periodontitis dominated thought in the 1970s and
1980s, and almost all preventive measures and treat-
ments were targeted at eliminating or at least con-
trolling the bacterial challenge. Consideration of a
major role for the host was generally an afterthought
and played no role in diagnostic and therapeutic de-
cisions. This began to change dramatically in the
1980s with the demonstration that the prevalence of
periodontitis was in fact much lower than previously
thought: disease of sufficient severity to cause loss
of teeth affects only about 15% of adults in most
countries (6).
The 1990s are dramatically different; research has
discovered that bacteria are essential but insufficient
for disease, that bacteria account for a relatively
small portion of the variance in susceptibility for dis-
ease expression (about 20% by some estimates, and
tobacco smoking accounts for more) (20), and that
hereditary factors alone can account for up to rough-
ly 50% of the variance (42,431. This is the decade of
the paradigm shift; this is the decade of the host and
disease modifiers (see Fig. 1 in Page & Kornman (51)
in this volume on page 10). In addition, this shift
provides a new perspective on the distinctly different
roles of the bacteria, the host and risk factors and
indicators in the disease process.
So far this chapter has focused on the nature of
the microbial challenge, assembling the cells and
systems that participate in periodontitis, the role of
the immune system, complement and the phago-
cytic cells, regulation by the cytokine network and
mechanisms by which the connective tissues of the
gingiva and periodontal ligament are destroyed and
alveolar bone resorbed. These components of the
pathogenesis are shared by all forms of periodontitis
(47, 48). The next section focuses on the factors that
239
Page et al.
influence the shared common events. Important
new findings and perspectives are summarized in
Table 7. In this volume, Salvi et al. (60) discuss fac-
tors that can affect shared events in the pathogenesis
in a manner that hastens the onset of disease, in-
creases its rate of progression or severity or causes
the disease to be refractory to treatment or to reoc-
cur. These include diabetes mellitus, tobacco smok-
ing, stress, HIV infection and osteoporosis. Other
factors reported in the literature are alluded to but
not discussed in detail. These include several ad-
ditional diseases and conditions that compromise
host defense, severity of disease at baseline, age,
presence of pathogenic bacteria in the oral flora,
socioeconomic status, educational level, oral hy-
giene and frequency of visits to the dentist. Heredi-
tary factors are considered in a separate section.
Diabetes mellitus
Diabetes mellitus is well documented as a risk factor
for periodontal disease; diabetics have an odds ratio
of about 2-3 relative to non-diabetics. Individuals
with diabetes mellitus have an increased prevalence
of periodontitis and respond poorly to therapy. Peri-
Table 7. The pathogenesis of periodontitis:
new findings and new perspectives on the
environmental and acquired risk factors
Periodontitis appears to behave as a multifactorial
disease in which multiple genetic and environmental
factors interact to produce the disease and modify
its clinical expression. One prime example of these
interactions is the IgG2 levels in early-onset
periodontitis, which appear to be influenced by
both genetic factors and smoking.
systemic diseases, including cardiovascular disease,
diabetes and pre-term low-birth-weight delivery.
Diabetes is a well-documented risk factor for
periodontitis.
Many diabetics manifest an upregulated monocyte
phenotype that is characterized by expression of
high levels of prostaglandin E2, IL-10 and tumor
necrosis factor a in response to a lipopolysaccharide
challenge.
Diabetes is known to alter collagen metabolism and
the basement membranes of small vessels.
Smoking is a well-documented risk factor for
periodontitis.
Smoking inhibits neutrophil function, suppresses
IgG2 antibody response, enhances rnonocyte
release of IL-1s and may directly affect osteoblast
function.
Psychosocial stress may be associated with an
increased severity of periodontitis.
HIV infection is a risk factor for an ulcerative form
of periodontitis.
Periodontitis is strongly associated with major
odontitis may be more severe in patients with dia-
betes of long duration and with poor metabolic con-
trol. There is evidence that diabetics with excellent
metabolic control are no more susceptible to severe
periodontitis than non-diabetics (see Salvi et al. (60)
in this volume). The diabetic state does not appear
to affect the composition of the subgingival biofilms;
instead, numerous aspects and steps in the shared
pathogenesis can be affected and thereby suscepti-
bility enhanced and severity worsened (see Salvi et
al. (60) in this volume, pages 181-183). Nevertheless,
diabetic individuals vary substantially; the diabetic
state may have multiple manifestations and conse-
quent effects on periodontitis susceptibility in some
individuals and none in others. This great variation
may be genetically based.
The mechanisms by which diabetes enhances the
risk for severe periodontitis are poorly understood,
although some diabetics manifest pathological
changes that may be related (see Salvi et al. (60) in
this volume, pages 181-183). Neutrophil adhesion,
chemotaxis, phagocytosis and killing can be im-
paired and the patient deprived of the major host
defense against microbial challenge. Many diabetics
manifest an upregulated monocyte phenotype re-
sulting from metabolic or genetic effects. Such cells
are proinflammatory in that they produce more IL-
l p, tumor necrosis factor a and prostaglandin E2
both unchallenged and in response to lipopolysac-
charide challenge than normal. The gingival tissues
and gingival crevicular fluid contain elevated con-
centrations of these mediators, and their presence
can elevate levels of matrix metalloproteinase, en-
hance tissue destruction and increase disease sever-
ity. The connective tissues in diabetics may manifest
pathological change. Collagen production by fibro-
blasts in the periodontal ligament and gingiva may
decrease and the production of gingival matrix
metalloproteinase may increase. Matrix metalloprot-
einase activity in gingival crevicular fluid can be el-
evated, and gingival fibroblasts in culture synthesize
less collagen than normal. The excess matrix metal-
loproteinase appears to derive from neutrophils.
Under hyperglycemic conditions, collagen is nonen-
zymatically glycosylated and its solubility and turn-
over rates change. Diabetics also have impaired pro-
duction of bone matrix components by osteoblasts.
A fundamental lesion is thickening of the base-
ment membranes of small vessels with nonenzy-
matic glycosylation of proteins and accumulation of
deposits within the vessel wall and on the luminal
surfaces. These changes narrow the vessel lumen
and interfere with transport across the vessel wall.
240
Advances in the pathogenesis of periodontitis
Patients vary greatly in the development of such
lesions; they are seen most frequently in individuals
having a genetic predilection, diabetes of long dur-
ation and poor metabolic control of hyperglycemia.
Advanced glycation end products induce oxidant
stress in periodontal tissues, and this may be related
to enhanced periodontal disease. Immune regula-
tion may also be affected. In subjects with type 1
diabetes there is a clear association with HLA-BB
and B15, and about 95% of such patients manifest
DR3or DR4 or both.
Of the possibilities described 'above, the elevated
levels of IL-1, tumor necrosis factor a, prostaglandin
E2 and matrix metalloproteinases are likely to have
the greatest influence on the common shared events
in the pathogenesis of periodontitis.
Tobacco smoking
A wealth of data demonstrate that cigar tte smoking
is the known environmental risk mos 9 strongly as-
sociated with periodontitis, and especially severe
periodontitis (see Salvi et al. (60) in this volume,
pages 183-184). Indeed, in some studies, the impact
of smoking actually outweighs the effect of the
pathogenic bacteria as a determinant of outcome.
Smoking has marked effects on the clinical features
of periodontitis. The gingival tissue tends to be
hyperkeratotic and fibrotic with thickened margins
and tends to manifest minimal erythema and edema
relative to disease of equal severity in nonsmokers.
Attachment loss is generally more severe in the an-
terior regions, especially on the palatal aspects of the
maxillary anterior teeth. The disease is more severe
and widespread relative to a nonsmoking cohort of
comparable age. There may be no positive associ-
ation between periodontal status and high plaque
and calculus scores. Surgical and nonsurgical ther-
apies are less effective in smokers than in non-
smokers, and the disease is more likely to recur.
Both locally and systemically induced effects on
the periodontium have been described (see Salvi et
al. (60) in this volume, page 185). Smoking tends to
mask gingival inflammation by constricting the
blood vessels of the gingiva, as well as the coronary
arteries and vessels of the placenta. Smoking affects
leukocyte function. The functional activity of both
salivary and tissue neutrophils is suppressed. Smok-
ing has major effects on immune responsiveness. A
negative significant association has been shown be-
tween smoking and serum antibody levels to certain
periodontal bacteria. The effect may be specific.
Smoking suppresses production of the IgG2 subclass
of immunoglobulin both in periodontitis patients
and periodontally normal individuals. Levels of IgG2
antibody specific for antigens of A. actinomycetem-
comituns are significantly lower among smokers with
adult periodontitis and rapidly progressive peri-
odontitis than among nonsmokers with the same
diseases. Since the predominant serum antibody re-
sponse to the major periodontal pathogens is mostly
of the IgG2 subclass, suppression of IgG2 production
may be a primary mechanism through which smok-
ing enhances susceptibility to severe periodontitis.
Smoking appears to have other effects on immune
regulation: T-cell subset ratios appear to be altered
in smokers. Nicotine also increases the lipopolysac-
charide-induced release of IL-1 p by monocytes.
Smoking may directly affect connective tissues
and bone. Cytotoxic substances such as nicotine and
its major metabolite, cotinine, can be detected in
saliva and crevicular fluid and in serum and urine,
demonstrating their systemic availability. These sub-
stances rapidly penetrate epithelium and affect
fibroblasts. Smoking decreases the intestinal absorp-
tion of calcium and may thereby affect osteoblast
function and increase bone loss. Toxic substances
from tobacco smoke can coat the root surfaces of
periodontally involved teeth and interfere with post-
surgical healing. Finally, there is evidence that smok-
ing may affect the composition of the subgingival
flora and enhance subgingival infection. Smoking al-
ters the short-term oxidation-reduction potential in
plaque and may thereby result in an increased pro-
portion of anaerobic bacteria.
Because cigarette smoking has extraordinary ef-
fects in enhancing the susceptibility to severe peri-
odontitis, which hinders successful therapy, smoking
cessation is now becoming an important part of suc-
cessful prevention and management of peri-
odontitis.
Psychosocial stress
Clinicians have long suspected that stress is import-
ant in susceptibility to periodontitis and in response
to therapy. Research is this area is still in its infancy,
and documentation of an association between stress
and the common forms of periodontitis is lacking.
Data support an association between stress and
acute necrotizing ulcerative gingivitis, and stressed
acute necrotizing ulcerative gingivitis patients mani-
fest depressed neutrophil chemotaxis and phago-
cytosis as well as reduced lymphocyte proliferation
in response to nonspecific mitogen stimulation.
Stress, distress and coping behaviors may be associ-
241
-
Page et al.
ated with an increased severity of periodontal tissue
destruction. Salvi et al. (60) have evaluated the data
thoroughly and conclude in this volume (pages 185-
186) that, although stress is a potential risk factor, it
has not been adequately evaluated.
Cellular and molecular interactions between neur-
oendocrine and immune systems have been well
documented (see Salvi et al. (60) in this volume,
pages 186-187). Corticosteroids inhibit several in-
flammatory cells, including macrophages, neutro-
phils, eosinophils and mast cells. These are mediated
by suppressing cytokines such as IL-1, IL-2, IL-3, IL-
4, tumor necrosis factor a, interferon y and granul-
ocyte-macrophage colony-stimulating factor and by
inhibiting arachidonic acid metabolism and the pro-
duction of prostaglandins and leukotrienes. De-
pressed immune responsiveness resulting from
stress has been postulated as one of several factors
involved in destructive periodontal diseases.
HlV infection
A form of periodontitis designated as necrotizing ul-
cerative periodontitis has been described in patients
with HIV infection or AIDS. The disease is painful,
rapidly destructive and frequently does not respond
to treatment. The frequency of this form of peri-
odontitis seems very low. Nevertheless, AIDS is
clearly a risk factor for this form of periodontitis (see
Salvi et al. (60) in this volume, pages 188-189).
The subgingival flora is similar to that seen in
adult periodontitis, although Candida albicans and
enteric bacterial species may be present in the sub-
gingival biofilms. Patients may manifest decreased
neutrophil chemotaxis, phagocytosis and bacterial
killing and impaired Fc receptor-specific clearance.
Monocyte chemotaxis may also be abnormal. The
T4:T8 lymphocyte ratio and the absolute numbers of
T4helper cells may be decreased. Levels of IL- 1 p in
gingival crevicular fluid are elevated. These abnor-
malities taken together would be expected to result
in inadequate host defense against the microbial
challenge and cytokine-mediated enhancement of
matrix metalloproteinase-mediated connective
tissue destruction and prostaglandin E-mediated al-
veolar bone loss.
Genetic risk factors for periodontitis
Evidence for genetic susceptibility
Genetic factors that act on and modify host re-
sponses to the microbial challenge are major deter-
minants of susceptibility to periodontitis and the
rates and extent of disease progression and severity.
New findings and perspectives are listed in Table 8.
Evidence for genetic susceptibility comes from three
sources: the association of periodontitis with specific
genetically transmitted disease traits, twin studies of
adult periodontitis and linkage and segregation
studies of families with early-onset forms of peri-
odontitis. Among the genetically transmitted disease
traits are neutropenia, trisomy 21 and Papillon-Le-
f h e syndrome, which confer abnormalities in neu-
trophil number or function and thereby greatly en-
hance susceptibility to periodontitis. Twin studies of
adult periodontitis have shown greater concordance
for periodontitis susceptibility between monozygotic
twins than between dizygotic twins and demon..
strated that heredity accounts for about 50% of thc
enhanced risk for severe periodontitis. Studies on
families with early-onset periodontitis have shown
one or more autosomal major gene loci linked to en-
Table 8. The pathogenesis of periodontitis:
new findings and new perspective on the genetic
risk factors
Periodontitis appears to behave as a multifactorial
disease in which multiple genetic and environmental
factors interact to produce the disease and modify
its clinical expression. One prime example of these
interactions may be seen in the IgG2levels in early-
onset periodontitis, which appear to be influenced
by both genetic factors and smoking.
Genetics has been shown to influence both early-
onset periodontitis and adult forms of periodontitis.
There are currently five promising candidates for a
genetic influence on periodontitis.
Given the critical protective role neutrophils play in
periodontitis, genetic defects in neutrophil
function would be expected to alter the disease.
IgG2 subclass of antibody is prominent in both
early-onset periodontitis and adult periodontitis.
IgG2 levels are known to be genetically regulated at
the G2M23 locus, which has been associated with
early-onset periodontitis. IgG2 levels are also
reduced in some smokers.
A polymorphism in the genes coding for the FcyII
receptor on neutrophils has been associated with
early-onset periodontitis and with phagocytic
function of neutrophils in conjunction with IgG2
antibodies.
A combination of two polymorphisms in the IL-1
genes has been associated with more severe
periodontitis in adults.
Given the apparent role of prostaglandin E2 in
periodontitis, genes controlling the prostaglandin
endoperoxide synthase enzymes may be good
candidates for genetic influences on periodontitis.
A recent linkage analysis in early-onset periodontitis
families identified a disease associated with the
physical region on chromosome 9q32-33, which
contains the gene for prostaglandin endoperoxide
svnthase- I .
242
Advances in the pathogenesis of periodontitis
hanced susceptibility for periodontitis (see Hart &
Kornman (23) in this volume, pages 207-21 1).
Candidate genes for enhanced periodontitis
The genetic traits that may confer enhanced suscep-
tibility to periodontitis are listed in Table 9 and dis-
cussed below (see Hart & Kornman (23) in this vol-
ume, pages 206-211).
Abnormalities in neutrophil function have been
demonstrated in approximately 75% of individuals
with juvenile periodontitis. Cells from about 75% of
these individuals manifest suppressed chemotaxis
and phagocytosis in vi m. In addition, the cells are
unable to mobilize extracellular calcium normally,
and intracellular signaling pathways appear to be
abnormal. These abnormalities could hamper the
critical role the phagocytic cells play in fending off
the microbial challenge and thereby account for en-
hanced disease susceptibility (see Hart & Kornman
(23) in this volume, pages 207-211) (26).
Serum IgG2 antibody appears to be the major
serum antibody produced in response to periodontal
infection. The capacity to produce IgG2 is genetically
determined and varies greatly among individuals
(see Hart & Kornman (23) in this volume, pages 207-
209). Studies on more than 100 families with early-
onset periodontitis have demonstrated a linkage of
disease susceptibility to the human leukocyte anti-
gen region of chromosome 6 containing the immune
response genes (81). Segregation analysis of such
families has revealed a major locus accounting for
approximately 62% of the variance for IgG2 produc-
tion (39, 40). Thus, a genetically determined reduced
capacity to produce IgG2 during the course of peri-
odontal infection may in part account for the en-.
hanced disease susceptibility.
The phagocytic cells, especially the neutrophils and
specific antibody, comprise the major host defense
against periodontal infection. These cells function by
Table 9. Genetic traits that may confer enhanced
susceptibility for periodontitis
Abnormal phagocyte function
Reduced capacity 10 produce IgG2. chromosome 6
hFc-y-IUIa polymorphism
hmor necrosis factor a polymorphism,
Variable rnonocyte and macrophage function
IL-10 polymorphism, chromosome 2q13
Prostaglandin endoperoxide synthase 1 gene,
chromosome 6
chromosome 9q32.33
- - ___. . . - - . __
accumulating at the sites of challenge by chemotaxis,
where they phagocytose and kill microorganisms that
have been opsonized. The major opsonin is specific
antibody, which is mostly IgG2 in periodontitis pa-
tients. The phagocytic cells express a family of cell
surface receptors that recognize bacteria coated with
specific antibody by binding the Fc region and there-
by provide the recognition mechanism for phago-
cytosis and killing. More than one receptor can recog-
nize bacteria opsonized by IgGl and IgG3, but only
one receptor, designated hFc-y-RIIa, recognizes bac-
teria coated wi t h IgG2. This receptor is polymorphic
at amino acid reside number 131, where either histi-
dine or arginine may be present (see Hart & Kornman
(23) in this volume, page 209). When the receptor is
homozygous for histidine, the receptor affinity is high;
when the receptor is homozygous for arginine it is low,
and when the receptor is heterozygous it is of inter-
mediate affinity. Neutrophils from individuals who
are homozygous for histidine effectively phagocytose
and kill bacteria opsonized by IgG2. Individuals who
are homozygous for arginine or those who are hetero-
zygous may manifest enhanced susceptibility to in-
fections by encapsulated bacteria such as Huerno-
philus infuenzue and to meningococcal disease. The
same may be true for periodontitis sites, since the pre-
dominant humoral response for periodontal mi-
crobial infection is IgG2. Therefore, even though a
given individual may produce high titers of high-af-
finity IgG2 antibody, phagocytosis and killing of in-
fecting bacteria may not occur if the low-affinity re-
ceptor is expressed.
Monocytes and macrophages are key in the patho-
genesis of periodontitis because they are activated
by lipopolysaccharide and proinflammatory cyto-
kines and can produce large amounts of tumor ne-
crosis factor a, IL-l p, prostaglandin E2 and matrix
metalloproteinase (Fig. 1). Cell responses are genet-
ically determined and unique for individuals. Vari-
ation in responsiveness among individuals is pro-
found and stable over time. Blood monocytes from
individuals who are susceptible to severe peri-
odontitis release significantly more tumor necrosis
factor a, I t - l p and prostaglandin E2 than monocytes
from individuals who are resistant (15, 16, 45, 58).
Based on these observations, Offenbacher (46) pro-
posed a model for the pathogenesis of periodontitis
in which susceptible individuals manifest a hyperre-
sponsive monocyte phenotype relative to individuals
who are resistant.
The proinflammatory cytokines tumor necrosis
factor a and IL-1p and prostaglandins (especially
prostaglandin E2) are extremely important in the
243
Page et al,
Fig. 3. In sicu hybridization studies and analyses per-
formed on gingival crevicular fluid and tissue extracts
have shown that periodontal health is characterized by
low levels of proinflammatory cytokines, prostaglandin E2
(PGE,) and matrix metalloproteinases (MMPs) and high
levels of tissue inhibitors of metalloproteinases (TIMP)
and cytokines that suppress the immunoinflammatory re-
sponse. The opposite is observed in active periodontitis.
TNF-a: tumor necrosis factor a; INF-1: interferon y; TGF-
f.k transforming growth factor p.
pathogenesis of periodontitis (Fig. 3). The levels of
these molecules and of matrix metalloproteinase are
high at sites of active tissue destruction and low at
healthy and successfully treated sites. The tumor ne-
crosis factor a gene has been mapped to chromo-
some 6, and a polymorphism exists in the gene pro-
moter region that results in greatly increased tumor
necrosis factor a production (41, 80). Individuals ex-
pressing this polymorphism manifest enhanced sus-
ceptibility to certain infections. Although this poly-
morphism has not yet been linked directly to suscep-
tibility to periodontitis, there is a strong possibility
that such a link does exist. Similarly, the IL-1 gene
family located on chromosome 2q13 is polymorphic.
Individuals with one of the specific genotypes pro-
duce about four times more IL-1p than do genotype-
negative individuals, and they are about 20 times
more likely to develop severe adult periodontitis at
ages beyond 40 years than are those who are gene
negative (31). Finally, studies on more than 100 fam-
ilies with early-onset periodontitis have demon-
strated a linkage with chromosome 9q32,33 contain-
ing the gene for cyclooxygenase 1 (prostaglandin en-
doperoxide synthetase 1) (81). This is the enzyme
system responsible for producing prostaglandin EZ.
The linkage of the gene to enhanced disease suscep-
tibility is most likely related to enhanced levels of
prostaglandin E2 production. One or more of these
genetic abnormalities may account at least in part
for the hyperresponsive monocyte phenotype sug-
gested by others (46, 47).
The pathogenesis of periodontitis:
a new paradigm
Periodontitis is a family of related chronic inflam-
matory diseases, all of which are bacterial infections,
The bacteria most well documented to cause these
diseases include I! gingivalis, B. forsythus and A. acti-
nornyceterncornitans ( 8) . Other species may be in-
volved in causing periodontitis, but they act more
indirectly than these species, and other species such
as enteric bacteria may be involved in special cases.
Nevertheless, these species are the primary patho-
gens in most cases. Bacteria are necessary but insuf-
ficient to cause periodontitis. In studies appropriate
for multivariate analysis, the bacterial component
accounts for a relatively small proportion (roughly
20%) of the variance in disease expression. Host fac-
tors are equally or more important than the bacteria
in determining disease development and outcome.
The complex interplay between the bacterial chal-
lenge and innate and acquired host factors deter-
mines the outcome.
Wepropose a new paradigm for the pathogenesis
of periodontitis (see Fig. 1 in Page & Kornman (51)
in this volume). This paradigm is based on advances
in knowledge in three specific areas. First, the sub-
gingival microbial flora is highly organized into bi-
ofilms and manifests the characteristics of biofilms.
The bacteria are largely protected from the host de-
fenses and are highly resistant to chemotherapeutic
agents. Physical disruption by scaling and root plan-
ing is an effective treatment. Second, major ad-
vances have been made at the cellular, molecular
and genetic levels in understanding the pathways
and mechanisms by which the bacteria present in
these biofilms initiate and perpetuate the immuno-
inflammatory response that destroys the connective
tissue of the gingiva and periodontal ligament and
resorbs the alveolar bone. It is now known that these
pathways underlie and are shared by all forms of
chronic marginal periodontitis. Third, although bac-
teria cause periodontitis and are essential for disease
to occur, bacteria alone are insufficient. A suscep-
tible host is required. Acquired and environmental
risk factors such as tobacco smoking as well as gen-
etically transmitted traits modify the shared path-
ways by which bacteria cause tissue destruction, and
they determine disease susceptibility, onset, pro-
gression, severity and outcome. This new way of
244
Advances in the pathogenesis of periodontitis
looking at the pathogenesis of periodontitis opens
entirely new approaches to preventing and manag-
ing this family of diseases.
Although many of the details of the pathogenesis
of periodontitis are complex and not yet well under-
stood, certain aspects stand out. Perhaps most im-
portantly, the dynamic events of pathogenesis are
determined primarily by the signaling and regulating
molecules that direct cellular function. Elevated
levels of bacterial substances characterize peri-
odontitis, especially active progressing lesions. The
main substances are lipopolysaccharide, the proin-
flammatory cytokines (IL-1, tumor necrosis factor a
and, to a lesser extent, IL-61, prostaglandins (es-
pecially prostaglandin E2) and the matrix metallop-
roteinases. In marked contrast, an absence of or very
low levels of bacterial substance, especially lipopoly-
saccharide, the presence of cytokines that suppress
the immunoinflammatory response (IL-10 and
transforming growth factor J3) and tissue inhibitors
of metalloproteinases combined with low levels of
prostaglandin E2 and matrix metalloproteinases
characterize stable and resolving periodontal lesions
and sites of periodontal health (Fig. 3). These are
central features of the pathogenesis shared by all
forms of periodontitis. Environmental, acquired and
genetic risk factors are major determinants of the
presence and concentration of specific antibodies,
cytokines, prostaglandins and proteases. Many of
these risk factors can be modified to prevent or alter
disease onset or progression or used as tests to de-
termine disease susceptibility.
Since it was discovered in the 1960s and following
years that periodontitis is infectious, therapy and
prevention have been targeted almost exclusively at
the bacteria and eliminating or reducing them to
numbers below the threshold levels required to in-
itiate and perpetuate disease. This approach, es-
pecially physical disruption and removal of the sub-
gingival biofilms by procedures such as scaling and
root planing, will remain a central and essential part
of periodontal therapy. Nevertheless, the new para-
digm modifies the traditional approach to account
for host and risk factors.
The presence and concentrations of factors char-
acterizing destructive periodontitis lesions vary
greatly among individuals, and many are determined
genetically. Environmental and acquired factors such
as tobacco smoking also affect these factors signifi-
cantly, either directly or indirectly. These obser-
vations point the way to new approaches for preven-
tion, diagnosis and treatment. Approaches that sup-
press the levels of the agents that characterize active
disease and that enhance the levels of those that
characterize health may be expected to be effective.
Interventions that reduce the levels or activities of
I L-lp, tumor necrosis factor a, prostaglandin E2 and
matrix metalloproteinases and those that enhance
the levels or activities of IL-10, transforming growth
factor p and tissue inhibitors of metalloproteinases
could serve as the basis for effective treatment. The
various chapters in this volume have suggested sev-
eral such approaches.
Under the new paradigm, periodontics is rapidly
changing from diagnosing and treating existing dis-
ease to prevention and health promotion. Reduction
of risk becomes the primary objective of inter-
vention for individuals and populations. Identifymg
the factors that place individuals and groups at en-
hanced risk and managing risk as a means of preven-
tion and treatment are of ever increasing import-
ance. Some risk factors are immutable to change;
others are not. For example, the data show that
smoking cessation alone can reduce the risk of se-
vere periodontitis remarkably. The risk imposed by
diabetes, although immutable, can be significantly
reduced or eliminated by measures ensuring the
highest level of metabolic control.
The most dramatic potential for future control or
virtual elimination of periodontal disease may ema-
nate from advances in understanding the role of her-
edity in determining susceptibility to and the sever-
ity of disease. Although this field is still in its infancy,
enough is known already to glimpse the future. Sev-
eral studies are linking susceptibility to early-onset
periodontitis to the human leukocyte antigen region
of chromosome 6, the site of the genes regulating the
IgG2 humoral immune response and the production
of tumor necrosis factor a. IgG2 is the major anti-
body class produced in response to periodontal
pathogens. The data indicate that a major locus,
possibly at that location, can account for 62% of the
variance in IgG2 levels. If this is the case, a measure
of the capacity of an individual to mount an IgG2
response could be a major indicator of disease sus-
ceptibility, severity and progression. Several genetic
polymorphisms in the tumor necrosis factor gene
family are known to exist. Polymorphisms in t he tu-
mor necrosis factor genes associated with suscepti-
bility to periodontitis can be searched for and prob-
ably exist.
The discovery of polymorphisms in the genes for
the hFc-y-RIIa receptor on the phagocytic cells
further elucidates the possible genetic basis for sus-
ceptibility to all forms of periodontitis. This poly-
morphism results in the expression of receptors of
245
Pane et al.
high, low or intermediate affinity. Since the hFc-y-
RIIa receptor is the only one that recognizes bacteria
opsonized with IgG2, expression of the low-affinity
receptor may be expected to significantly enhance
susceptibility to periodontal pathogens and to severe
periodontitis, regardless of the capacity of the
affected individuals to produce high levels of biolo-
gically effective antibody.
Similarly, the observation of linkage between sus-
ceptibility to early-onset periodontitis and the region
on chromosome 9 encoding for cyclooxygenase 1 is
seminal and more definitively documents the sig-
nificance of prostaglandin E2 in the pathogenesis of
periodontitis. This gene encodes for the constitut-
ively produced cyclooxygenase enzyme system re-
sponsible for production of prostaglandin EZ, the
major mediator of alveolar bone destruction in peri-
odontitis. Location and characterization of the gene
for cyclooxygenase 2, which is activated by I L-1
specifically in inflammatory diseases such as peri-
odontitis, may have even more important impli-
cations.
IL-1 and especially IL-1P is unquestionably im-
portant in the pathogenesis of periodontitis. A poly-
morphism in the IL-1 gene family associated with
high susceptibility to severe adult periodontitis at
age 40 and beyond has been identified, and a labora-
tory test to detect the polymorphism has been de-
veloped (31). Individuals who do not smoke and test
positively produce approximately 4-fold more IL- 1 p
in response to lipopolysaccharide, substantially in-
creasing the probability that they will develop severe
periodontitis relative to those who test negatively.
Smoking and the IL-1P gene test could identify more
than 85% of the individuals with enhanced suscepti-
bility for severe periodontitis.
These observations point the way to the future.
We envision the development and availability of
relatively simple, inexpensive tests for the poly-
morphism in the I L-l P gene family, for hFc-y-RIIa
phagocyte receptor affinity and for the capacity to
mount an effective IgG2 humoral immune response.
Weenvision a time when young people can be tested
and the results combined with other potential risk
factors to accurately assess their risk of periodontitis.
Young individuals could probably be accurately dif-
ferentiated into those who are almost certain to de-
velop periodontitis in later life and those who are
not. Susceptible individuals could be monitored, for
example, using DNA probes to detect early coloniza-
tion by periodontal pathogens. When colonization
occurs, it could easily and inexpensively be elimin-
ated. Elimination of periodontitis as a significant
disease in some populations now seems to be poss-
ible.
References
1. Attstrom R, Schroeder HE. Effect of experimental neutrop-
enia on initial gingivitis in dogs. Scand J Dent Res 1979: 87:
2. Babb J L, Kiyono H, McGhee JR, Michaelick S. LPS regula-
tion of the immune response: suppression of immune re-
sponses to orally administered T-independent antigens. J
Immunol 1981: 127: 1052-1057.
3. Baxt WG. Complexity, chaos and human physiology: the
justification for non-linear neural computational analyses.
Cancer Lett 1994: 72: 85-93.
4. Beck J , Garcia R, Heiss G, Vokonas PS, Offenbacher S. Peri-
odontal disease and cardiovascular disease. 1 Periodontol
5. Birkedal-Hansen H. Role of cytokines and inflammatory
mediators in tissue destruction. J Periodont Res 1993: 28:
500-510.
6. Brown LJ, Ltje H. Prevalence, extent, severity and pro-
gression of periodontal disease. Periodontol 2000 1993: 2:
7. Colditz IG. Effects of exogenous prostaglandin E2 and ac-
tinomycin D on plasma leakage induced by neutrophil ac-
tivating peptide-l/IL-8. Immunol Cell Biol 1990: 68: 397-
403.
8. Consensus report on periodontal diseases: pathogenesis
and microbial factors. Ann Periodontol 1996: 1: 926-932.
9. Costerton JW, Lewandowski Z, DeBeer D, Caldwell D, Korb-
er D, J ames G. Biofilms, the customized microniche. J Bac-
terial 1994: 176: 2137-2142.
10. Darveau RP, Tanner A, Page RC. The microbial challenge in
periodontitis. Periodontol 2000 1997: 14: 12-32.
11. Dennison DK, Van Dyke T. The role of phagocytic cells,
antibody and complement in periodontal health and dis-
ease. Periodonto12000 1997: 1997: 14: 54-78.
12. Eastcott JW, Yamashita K, Taubman MA, Harada Y, Smith
DJ . Adoptive transfer of cloned T helper cells ameliorates
periodontal disease in nude rats. Oral Microbiol Immuno
13. Fink PK, Herren LT. Modeling disease processes for drug
development: bridging the gap between quantitative and
heuristic models. In: Charnes J M, Morrice DJ , Brunner DT,
Swain J J , ed. Proceedings of the 1996 Winter Simulation
Conference. San Diego: International Society for Computer
7-23.
1996 67: 1123-1137.
57-7 1.
1994: 9: 284-289.
14.
15.
16.
17.
Simulation, 1996.
Garant P, Cho MI. Histopathogenesis of spontaneous peri-
odontal disease in conventional rats. I. Histometric histo-
logic study. J Periodont Res 1979: 14: 297-309.
Garrison SW, Holt SC, Nichols FC. Lipopolysaccharide-
stimulated PGEz release from human monocytes. J Peri-
odontol 1988: 59: 684-687.
Garrison SW, Nichols FC. LPS-elicited secretory responses
in monocytes: altered release of PGEp but not IL-lp in pa-
tients with adult periodontitis. 1 Periodont Res 1989: 24:
88-956.
Gemrnell E, Marshall RI, Seymour GJ . Cytokines and
prostaglandins in immune homeostasis and tissue destruc-
246
Advances in the pathogenesis of periodontitis
tion in periodontal disease. Periodonto12000 1997: 14: 112-
143.
18. Gemmell E, Walsh LJ, Savage NW, Seymour GJ. Adhesion
molecule expression in chronic inflammatory periodontal
disease. J Periodont Res 1994: 29: 46-53.
19. Genco RJ . Host responses in periodontal diseases: Current
concepts. J Periodontol 1992: 63: 3338-355.
20. Grossi SG, Zarnbon J J , Ho AW, Koch G, Dunford RG, Macht-
ei EE, Norderyd OM, Genco RJ , Assessment of risk for peri-
odontal disease. l. risk indicators for attachment loss. J
Periodontol 1994: 65: 260-267.
21. Gunsolley JC, Burmeister JA, Tew J G, Best AM, Ranney RR.
Relationship of serum antibody to attachment level pat-
terns in young adults with juvenile periodontitis or gener-
alized severe periodontitis. J Periodontol 1987: 58: 314-320.
22. Haffajee AD, Socransky, SS, Lindhe J , Kent RL, Okamoto H,
Yoneyama T. Clinical risk indicators for periodontal attach-
ment loss. J Clin Periodontol 1991: 18: 117-125.
23. Hart TC, Komman KS. Genetic factors in the pathogenesis
of periodontitis. Periodontol 2000 1997: 14: 202-215.
24. Haverson K, Stokes CR, Bailey M. Immunophenotypic
study of cell populations in the pig gut lamina propria. Im-
munol Cell Biol 1997: 75(suppl 1): A86, abstr W2.5.23.
25. Heasman PA, Collin J G, Offenbacher S. Changes in crevic-
ular fluid levels of interleukin l p, leukotriene B4, prosta-
glandin E2, thromboxane B2 and tumor necrosis factor a in
experimental gingivitis in humans. J Periodont Res 1993:
26. Hemmerle J , Frank RM. Bacterial invasion of periodontal
tissues after experimental immunosuppression in rats. J
Biol Buccale 1991: 19: 271-282.
27. lshikawa I, Nakashima K, Koseki T, Nagasawa T, Watanabe
H, Arakawa S, Nitta H, Nishihara T. Induction of the im-
mune response to periodontopathic bacteria and its role in
the pathogenesis of periodontitis. Periodontol 2000 1997:
28. lvanyi L, Lehner T. Stimulation of lymphocyte transform-
ation by bacterial antigens in patients with periodontal dis-
ease. Arch Oral Biol 1971: 1 6 1117-1121.
29. J enkins WMM, MacFarlane TW, Gilmour WH. Longitudinal
study of untreated periodontitis. Clinical findings. J Clin
Periodontol 1988: 15: 324-330.
30. J ordan W, Keyes PH. Aerobic, gram-positive filamentous
bacteria as etiologic agents of experimental periodontal
disease in hamsters. Arch Oral Biol 1964: 9: 401414.
31. Kornman KS, Crane A, Wang H-Y, di Giovine FS, Newman
MG, Pirk FW, Wilson TG J r, Higginbottom FL, Duff GW. The
interleukin 1 genotype as a severity factor in adult peri-
odontal disease. J Clin Periodontol 1997: 24: 72-77.
32. Kornman KS, Page RC, Tonetti MS. The host response to
the microbial challenge in periodontitis: assembling the
players. Periodontol 2000 1997: 1 4 33-53.
33. Krugluger W, Nell A, Solar P, Matejka M, Doltz-Nitulescu G.
Influence of sE-selectin and L-selectin on regulation of cell
migration during chronic periodontitis. J Periodont Res
34. Lindhe 1, Hamp SE, Loe H. Experimental periodontitis in
the beagle dog. 1 Periodont Res1973: 81: 1-10.
35. Lindhe 1, Haffajee AD, Socransky SS. Progression of peri-
odontal disease in adult subjects in the absence of peri-
odontal therapy. J Clin Periodontol 1983: 10: 433-442.
36. Lindhe J , Okomoto H, Yoneyama T, Haffajee AD, Socransky
SS. Longitudinal changes in periodontal disease in un-
treated subjects. J Clin Periodontol 1989: 16: 662-670.
28: 241-247.
1997: 14: 79-111.
1995: 30: 198-203.
37. Listgarten MA, Schifter CC, Laster L. 3-year longitudinal
study of the periodontal status of an adult population with
gingivitis. J Clin Periodontol 1985: 12: 225-238.
38. Loe HI, Theilade E, J ensen SB. Experimental gingivitis in
man. J Periodontol 1965: 36: 177-187.
39. Marazita ML, Burmaster JA, Gunsolley JC, Koertge TE, Lake
K, Schenkein HA. Evidence for autosornal dominant in-
heritance and race-specific heterogeneity in early-onset
periodontitis. J Periodontoll994: 65: 623-630.
40. Marazita ML, Lu H, Cooper ME, Quinn SM, Zhang 1,
Bermeister JA, Califano Iv, Pandey J R Schenkein HA, Tew
J G. Genetic segregation analysis of serum IgG2 levels. Am J
Hum Genet 1996: 58: 1042-1049.
41. McGuire W, Hill AV, Ahopp CE, Greenwood BM, Kwiatkow-
ski D. Variation in the TNF-a promoter region associated
wi th susceptibility to cerebral malaria. Nature 1995: 371:
42. Michalowicz BS, Aeppli D, Virag JG, Klump DG, Hinrichs
J E, Segal NS, Bouchard TJ Jr, Philstrom BL. Periodontal
findings in adult twins. J Periodontol 1991: 62: 293-299.
43. Michalowicz BS. Genetic and inheritance considerations in
periodontal disease. Curr Opin Periodontol 1993: 1: 11-176.
44. Newman MG, Socransky SS, Savitt ED, Propas DA, Craw-
ford A. Studies of the microbiology of periodontosis. J Peri-
odontol 1976: 47: 373-379.
45. Nichols FC, Garrison SW, Davis HW. Prostaglandin E2 and
thromboxane B2 release from human monocytes treated
with lipopolysaccharide. J Leukoc Biol 1988: 44: 376-384.
46. Offenbacher S. Periodontal diseases: pathogenesis. Ann
Periodontol 1996: 1: 879-925.
47. Offenbacher S, Heasman PA, Collins J G. Modulation of host
PGE2 secretion as a determinant of periodontal disease ex-
pression. J Periodontol 1993: 64: 432444.
48. Offenbacher S, Katz V, Fertik G, Collins J , Boyd D, Maynor
G, McKaig R, Beck J. Periodontal infection as a possible risk
factor for preterm low birth weight. J Periodontol 1996: 67:
49. Page RC. The role of inflammatory mediators in the patho-
genesis of periodontal disease. J Periodont Res 1991: 26:
230-242.
50. Page RC, Beck JD. Risk assessment for periodontal diseases.
Int Dent J (in press).
51. Page RC, Kornman KS. The pathogenesis of human peri-
odontitis: an introduction. Periodontol 2000 1997: 14: 9-11.
52. Page RC, Schroeder HE. Pathogenesis of inflammatory
periodontal disease. A summary of current work. Lab Invest
53. Page RC, Schroeder HE. Periodontitis in man and other
animals. A comparative review. Basel: S. Karger, 1982.
54. Papapanou PN, Baelum V, Luan W-M, Madianos PN, Chen
X, Fjerskov 0, Dahlin G. Subgingival microflora in adult
Chinese: prevalence and relation to periodontal disease
progression. J Periodontol (in press).
55. Raeste AM, Tapanila T, lbpokka R. Leukocyte migration
into the healthy dentulous mouth. J Periodont Res 1977:
12: 444-449.
56. Ranney RR, Yanni NR, Burmeister J A, Tew J G. Relationship
between attachment loss and serum antibody to Actino-
bacillus actinomyceremcornitans in adolescents and young
adults having severe periodontal destruction. J Periodontol
57. Reynolds J J , Meikle MC. Mechanisms of connective tissue
matrix destruction in periodontitis. Periodonto12000 1997:
14: 144-157.
508-510.
1103-1 113.
1976: 33: 235-249.
1982: 53: 1-7.
24 7
Page et al.
58. Richards D, Rutherford RB. The effects of interleukin-1 on
collagenolytic activity and prostaglandin E secretion by hu-
man periodontal ligament and gingival fibroblasts. Arch
Oral Biol 1988: 33: 237-243.
59. Rowe DJ , Bradley LS. Quantitative analysis of osteoclasts,
bone loss and inflammation in human periodontal disease.
J Periodont Res 1981: 16: 13-19.
60. Salvi GE, Lawrence HP, Offenbacher S. Beck JD. influence
of risk factors on the pathogenesis of periodontitis. Peri-
odontol2000 1997: 14 173-201.
61. Schwartz Z, Goultschin J , Dean DD. Boyan BD. Mechan-
isms of alveolar bone destruction in periodontitis. Peri-
odontol2000 1997: 14: 158-172.
62. Schroeder HE. Discussion: pathogenesis of periodontitis. J
Clin Periodontol 1986: 13: 426-430.
63. Schroeder HE. The junctional epithelium: origin, structure,
and significance. A review. Acta Med Dent Helv 1996: 1:
155-167.
64. Schroeder HE, Lindhe J . Conditions and pathologic fea-
tures of rapidly destructive, experimental periodontitis in
dogs. 1 Periodontol 1980: 51: 6-19.
65. Schroeder HE, Listgarten MA. The gingival tissues: the
architecture of periodontal protection. Periodontol 2000
66. Schroeder HE, Page RC. Lymphocyte-fibroblast interactins
in the pathogenesis of inflammatory gingival disease. Ex-
perientia 1972: 28: 1228-1230.
67. Seymour GI, Gemmell E, Walsh LJ , Powell RN. Immunohis-
tological analysis of experimental gingivitis in humans. Clin
Exp Imrnunol 1988: 71: 132-137.
68. Seymour GI. Importance of the host response in the peri-
odontium. J Clin Periodontol 1991: 18: 421-426.
69. Simpson DM, Avery BE. Histopathologic and ultrastruc-
tural features of inflamed gingiva in the baboon. 1 Peri-
odontol 1974: 45: 500-510.
70. Simpson DM, Avery BE. The baboon as a model system for
the study of periodontal disease: clinical and light micro-
scopic observations. J Periodontol 1973: 44: 675-686.
71. Slots I , Reynolds HS, Genco RJ. Actinobacillus actinomyce-
terncomituns in human periodontal disease: a cross-sec-
tional microbiological investigation. Infect Immun 1980:
1997: 13: 91-120.
29: 1013-1020.
72. Socransky SS, Haffajee AD, Goodson J M, Lindhe 1. New
concepts of destructive periodontal disease. J Clin Peri-
odontol 1984: 11: 21-32.
73. Tal H. Relationship between interproximal distance of roots
and the prevalence of intrabony pockets. 1 Periodoniol
1984: 55: 604-607.
74. Tanner ACR, Haffer C, Bratthall GT, Visconti RA, Socransky
SS. A study of the bacteria associated with advancing peri-
odontitis in man. I Clin Periodontol 1979: 6: 278-307.
75. Taubman MA, Kawai T, Watanabe H, Eastcott JW, Smith DJ .
Cytokinelendothelial regulation of T lymphocyte trans-
migration produces anergy: a protective mechanism in
periodontal disease. Immunol CeU Biol 1997: 75(suppl 1):
A6, abstr S1.1.5.
76. Tew J G, Zhang J -B, Quinn S, Tangata S , Nakashima K, Gun-
solley J C, Schenkein HA, Califano JV Antibody of the IgG2
subclass, Actinobacillus actinomycetemcomitans and early-
onset periodontitis. J Periodontol 1996: 67(suppl): 317-322.
77. Waerhaug J . Theangular bone defect and its relationship
to trauma from occlusion and down-growth of subgingival
plaque. J Clin Periodontol 1979: 6 61-82.
78. Walsh LJ , Murphy GF. Langerhans cells are intimately as-
sociated wi th intra-epithelial nerves. J Dent Res 1992: 71:
993.
79. Walters ID, Miller TJ , Cario AC, Beck FM, Marucha PT. Poly-
mi nes found in gingival fluid inhibit chemotaxis by hu-
man polymorphonuclear leukocytes i n uitro. J Periodontol
1995: 66: 274-278.
80. Wilson AG, di Giovine FS, Duff GW. Genetics of tumor ne-
crosis factor alpha in autoimmune, infectious and neoplas-
tic diseases. J Inflammation 1995: 45: 1-12.
81. Yang S, Sun C, Gillanders E, Wang Y-F, Duffy D, Bock C,
Freas-Lutz D, Zhang Y-I, Lopez N, Schenkein H, Deihl S.
Genome scan for susceptibility loci to the complex disorder
early onset periodontitis. AmJ Hum Genet 1996: 59: 1386
(abstr).
82. Yucel-Lindberg T, Ahola H, Nilsson S, Carlstedt-Duke J , Mo-
deer T. Interleukin-I p induces expression of cyclooxygen-
ase-2 mRNA in human gingival fibroblasts. Inflammation
1995: 19: 549-560.
248