Periodontal disease(s) refers to the inflammatory processes that occur in the tissues surrounding the teeth in response to bacterial accumulations (dental plaque) on the teeth. Rarely do these bacterial accumulations cause overt infections, but the inflammatory response(s) which they elicit in the gingival tissue is ultimately responsible for a progressive loss of collagen attachment of the tooth to the underlying alveolar (jaw) bone, which, if unchecked, can cause the tooth to loosen and then to be lost. The resulting crevice between the tooth surface and the approximating epithelial surface is called the periodontal pocket. This pocket can extend from 4 to 12 mm and can harbor, depending on its depth and extent, from 10⁷ to almost 10⁹ bacterial cells (281). The gingival bleeding and attachment loss associated with this process is usually painless and is ignored by the individual. Often the first time that the individual is aware of the problem is when the dentist informs him or her of the presence of pockets measuring more than 4 mm in depth. For example, the individual in Fig. 1 came to the dental clinic seeking replacement of his missing front tooth and had to be told that he had advanced periodontal disease with many deep pockets, 5 mm in depth. This symptomless nature of periodontal disease is one of its defining characteristics.

View larger version (117K): [in this window] [in a new window]   FIG. 1.   Patient with advanced adult periodontitis. (A) Prior to treatment. The patient came to the clinic seeking replacement for missing tooth. He was not aware of having a periodontal problem, despite gingival bleeding and massive amounts of supragingival plaque and calculus about the teeth. Note the whitish material at junction where the teeth contact the gingival tissue, especially about the canine tooth on the lower right. (B) After the plaque and calculus have been removed. There has been considerable bone loss about certain teeth, which accounts for the exposed roots.

The prevalence of periodontal disease increases with age (36, 87, 88, 145, 217, 219) and as more people are living longer and retaining more teeth, the number of people developing periodontal disease will increase in the next decades. About 50% of the adult population has gingivitis (gingival inflammation without any bone loss about teeth and no pockets deeper than 3 mm) around three or four teeth at any given time, and 30% have periodontitis as defined by the presence of three or more teeth with pockets of 4 mm (9, 217). Between 5 and 15% of those with periodontitis have advanced forms with pockets of 6 mm (219). Another 3 to 4% of individuals will develop an aggressive form of periodontal disease, known as early onset periodontitis (EOP), between the ages of 14 and 35 years. Any medical condition that affects host antibacterial defense mechanisms, such as human immunodeficiency virus infection HIV (328), diabetes (219, 264), and neutrophil disorders (312), will predispose the individual to periodontal disease.

These prevalences suggest that about two million Americans younger than 35 years and another four million older than 35 years may have a form of periodontal disease that requires professional intervention. Traditional treatments reflect the premise that periodontal disease is due to the nonspecific overgrowth of any and all bacteria on the tooth surfaces and that the magnitude of the bacterial overgrowth on the teeth can be controlled by professional cleaning of the teeth at regular intervals. If these accumulations are not removed, various bacterial by-products and their cellular components such as lipopolysaccharides (LPS), antigens, or enzymes can provoke an inflammatory response in the gingival tissue. Undisturbed plaques often become calcified, forming dental calculus or tartar, which, if formed below the gingival margin, is often difficult to remove from the root surfaces without some form of surgical access.

The patient whose mouth is shown in Fig. 1A did not practice good oral hygiene, nor did he have a dentist clean his teeth at regular intervals, and so he accumulated massive amounts of plaque and calculus on his teeth. When his teeth were "cleaned," the gingival inflammation decreased (Fig. 1B). Because the residual depths of many pockets were 5 mm, large numbers of bacteria could still accumulate and not be accessible to normal tooth-cleaning procedures. Therefore, periodontal surgery was performed to reduce the pocket depths to 1 to 2 mm, so that the patient would be able, by good brushing techniques, to keep the bacterial load reduced, thereby preventing reoccurrence of inflammation and attachment loss. Professional toothcleanings were recommended every 3 to 4 months for the rest of his life. At no time during the patient's treatment was there a need for a bacteriological diagnosis of the type of bacteria present in this nonspecific overgrowth.

This type of periodontal treatment is the standard of care in periodontal treatment and is based on the premise that if the bacterial overgrowth in dental plaque can be continuously suppressed by mechanical debridement, gingival and periodontal health will be maintained. It is the basis of the "plaque control" programs of organized dentistry and dentifrice manufacturers; as a public health effort, this approach has been very successful. One of the findings of the population-based dental surveys conducted in the last 20 years by the National Institute of Cranio-Facial and Dental Research has been the good overall periodontal health of U.S. citizens (217). However, as noted above, about 2 to 6 million people could require professional treatment, which would include "pocket elimination" surgery so as to gain access to plaque- and calculus-laden root surfaces. Since the average cost for full mouth periodontal surgery is about $4,000 to $5,000, and if 300,000 people (about 10%) actually received treatment, the projected cost could be more than one billion dollars. This would be an overwhelming liability for insurance companies and health care plans to cover. Accordingly very few, if any, dental insurance plans include the full cost of periodontal surgery, and it is not covered by Medicare or Medicaid. This out-of-pocket cost to the individual, plus patient concerns over the surgical procedures themselves (29, 187, 190), would discourage some individuals from seeking treatment.

Such chronic, asymptomatic periodontal infections may go unnoticed by the individual, as evidenced by the person whose mouth is shown in Fig. 1. Recent findings have indicated that chronic infections could serve as a source of inflammatory mediators, LPS, and other bioactive molecules that might contribute to the development of cardiovascular disease (56, 57, 188). Moderate increases in the level of C-reactive protein (CRP) in serum were predictive of new heart episodes in apparently healthy men (234). Others have shown that edentulism (all teeth missing) and periodontal disease are associated with elevated CRP levels in serum after controlling for established risk factors (267). In the case of periodontal disease, the magnitude of the association with CRP levels was comparable to that of chronic bronchitis and cigarette smoking and was strongest for individuals with no medical risk factors, i.e., healthy individuals (267).

Another mechanism by which periodontal bacteria could contribute to cardiovascular pathology relates to the antigenic similarity of certain bacterial proteins with host proteins. For example, subgingival plaque bacteria may share antigenic determinants, such as heat shock proteins (hsp), with host cells. Many host tissues, including the endothelial lining of blood vessels, produce hsp60 as they respond to certain stressors like high blood pressure and LPS. Xu, Wick, and coworkers have postulated that an autoimmune mechanism in which the host responds to foreign hsp60, such as bacterial hsp, could be important in the development of an atheroma (focal deposit of acellular, mainly lipid-containing material on the endothelial lining of arteries) (322, 335). The sera and inflamed gingival tissues of periodontal patients exhibited a positive antibody response to both the hsp produced by Porphyromonas gingivalis, i.e., GroEL hsp60, and to human hsp60 (291). Antibodies to GroEL hsp60 cross-reacted with human hsp60 and vice versa, suggesting that molecular mimicry between molecules of the bacteria and host could play a role in periodontal as well as humoral immune mechanisms. For example, antibodies against the hsps of P. gingivalis could react with human hsps exposed on the endothelium and produce cellular damage.

At least 14 of 17 studies of different design and rigor have provided statistical evidence for an association between periodontal disease and cardiovascular disease (27, 160, 188), raising the possibility that periodontal disease is a risk factor for cardiovascular disease. If so, periodontal disease, because it is both preventable and treatable, becomes a modifiable risk factor for cardiovascular disease. However, if periodontal disease is primarily due to the overgrowth of bacteria in the dental plaque, all individuals would need preventive treatment, since all individuals form dental plaque. If the focus were on the treatment of existing periodontal disease, the prospects for control would still not be good due to the projected costs of maintaining, with professional supervision, a clean mouth for a lifetime. Even if this cost could be met, the current standard of care, i.e., the debridement and surgical approach, fails in about 15 to 20% of treated individuals, the so-called refractory patients (111, 189, 190).

This scenario assumes that periodontal disease is the host inflammatory response to the bacterial accumulations on the tooth surface and that the types of bacteria present in the overgrowth is not important. But what would be the treatment options if the bacterial overgrowth always, or usually, resulted in the selection of a limited number of bacterial types in the plaque? Could this convergence on a common bacterial profile in disease-associated plaques be considered an infection, albeit a chronic one? If this were the actual situation, there would be a need to improve diagnostic capabilities beyond that associated with scoring plaque accumulations and measuring pocket depths with a pocket probe, so as to identify which individuals are "infected," and to focus treatment only on those individuals. Thus, the fundamental question in regard to periodontal pathology is whether the host is responding to the nonspecific overgrowth of bacteria on the tooth surfaces (inflammatory disease) or to the overgrowth of a limited number of species which produce biologically active molecules that are particularly proinflammatory or antigenic (infection).

The question of whether we are dealing with a disease should be answered by looking at how the host responds to the dental plaque. The dental plaque is unlike any other bacterial ecosystem that survives on the body surfaces, in that it develops on the nonshedding tooth surface and can form complex bacterial communities that may harbor over 400 distinct species and contain over 10⁸ bacteria per mg (200). The plaque is divided into two distinct types based on the relationship of the plaque to the gingival margin, i.e., supragingival plaque and subgingival plaque. The supragingival plaque is dominated by facultative Streptococcus and Actinomyces species, whereas the subgingival plaque harbors an anaerobic gram-negative flora dominated by uncultivable spirochetal species (44, 166). It is this gram-negative flora that has been associated with periodontal disease. Since many of its members derive some of their nutrients from the gingival crevicular fluid, a tissue transudate (35, 50) that seeps into the periodontal area, it is possible that their overgrowth is a result of the inflammatory process (146, 166).

Therefore, there is a distinction between the way the host responds to the supragingival plaque and its response to the subgingival plaque. The response to the supragingival plaque has been thoroughly studied in the experimental gingivitis model described below, whereas the response to the subgingival plaque remains under investigation. Does the host respond to the subgingival plaque as if it were an overgrowth of a bacterial community in which many members produce substances, such as LPS, that are particularly bioactive if they enter the approximating gingival tissue? Or does it respond to a plaque in which certain members produce more biologically active molecules, such as butyric acid (209) or hydrogen sulfide (233, 244), per cell or possess unique proteases, such as are found in P. gingivalis and Treponema denticola, which can degrade host molecules, creating a proinflammatory effect (84, 122, 144, 181, 304)? In either case, although bacteria are involved, it is not the scenario of a typical infection, as the offending bacteria generally remain outside the body, attached to the tooth.

Since bacteria are always present on the nonrenewing tooth surface, even healthy gingival tissue exhibits some inflammatory cells (258). These inflammatory cells increase in number as the bacterial plaque accumulates, and the tissue becomes edematous, reddens, and eventually bleeds. The initial sequence of events in the tissue adjacent to the newly forming plaque has been extensively documented through the use of an experimental gingivitis model in which volunteers are brought to an excellent level of gingival health through repeated cleanings. They then refrain from all oral hygiene procedures for a 3- to 4-week period, after which health is restored by resuming oral hygiene and dental cleanings (146). There is a highly reproducible relationship between plaque accumulation and gingivitis, which has been interpreted as proving that the plaque mass causes gingival inflammation (144).

The initial bacterial colonizers are predominantly streptococci (290), but within days, the bacterial community changes to one characterized by Actinomyces species and other bacterial types (138, 290, 300). Three-week-old plaque harbors a diverse flora, as evidenced by the isolation of 166 species from only four subjects and 96 plaque samples (198). As the plaque community ages, anaerobic species such as spirochetes are detected, but in small numbers (197, 198, 300), probably indicating that the oxidation-reduction potential of the plaque has decreased to levels where microaerophilic and moderately anaerobic species can ascend to numerical prominence in the plaque (155).

If the plaque mass is held constant, only certain plaques are associated with gingivitis, suggesting that it is the specific bacterial composition of the plaques, and not the bacterial numbers, which causes the gingivitis (161). At the time bleeding is noted, there is a significant proportional increase of Actinomyces viscosus and the appearance of Campylobacter and Prevotella species in the plaque. These latter species have nutritional requirements derived from the host, such as hemin and menadione, and from other microbes, such as formate. The emergence of these species at the time that bleeding is present suggests that these nutrients became available as a result of the tissue inflammation and bleeding. This proportional rearrangement of the flora, with a shift to gram-negative anaerobic species, portends the type of plaque flora which dominates in periodontal disease.

The experimental gingivitis model is stopped for ethical reasons after 3 to 4 weeks of no oral hygiene, so as to prevent any deterioration of the subject's health towards developing a more anaerobic plaque flora similar to that found in the gingivitis associated with poor oral hygiene. This "neglect" gingivitis is the most common form of periodontal disease and is experienced by all individuals at some time. It is this gingivitis that progresses to periodontitis, usually about the molars (217), where it is more difficult to maintain oral hygiene. This transition from gingivitis to periodontitis is undocumented in humans, and the trigger for the conversion is not known. In animal models this trigger for conversion is rapid plaque accumulation after placement of a foreign body such as a silk ligature around the teeth in either dogs (130) or monkeys (257). An analogous situation in humans may occur when fibrous food is retained between the teeth or when a dental restoration is poorly contoured, creating an overhang where bacteria can accumulate at the junction of the filling and the tooth surface. When amalgam restorations were purposely placed with overhangs, the numbers of both spirochetes and black-pigmented Prevotella and Porphyromonas species increased in the adjacent plaque and gingival bleeding was observed (125).

The host mounts an inflammatory response in the approximating gingival tissue to bacterial accumulations on the teeth. This response prevents bacterial growth in the tissue; removes bacterial products such as antigens, LPS, and enzymes that have penetrated the tissue; and is associated with specific antibody formation that, in the case of Actinobacillus actinomycetemcomitans in the rare clinical condition known as localized juvenile periodontitis (LJP), appears to be protective (39, 60). However, the inflammatory response can also activate the matrix metalloproteases, which are the agents responsible for collagen loss in the tissues (288, 305). These latent collagenolytic enzymes can be converted to active forms by proteases and reactive oxygen species in the inflammatory environment (288), giving rise to elevated levels of interstitial collagenase in the inflamed gingival tissue (126, 305, 306). The resulting attachment loss deepens the sulcus, or depression, formed where the gingival tissues contact the tooth surface, thereby creating the periodontal pocket. By definition, this loss of attachment converts gingivitis to periodontitis.

The depth of the pocket reflects an inflammatory response that causes both the swelling of the gingival tissues at the top of the pocket and the loss of collagen attachment of the tooth to the alveolar bone at the bottom of the pocket. Good oral hygiene can reduce the inflammatory swelling (121, 134), but the attachment loss and accompanying bone loss is thought to be irreversible. Pockets tend to be stable in their depths, but some continue to extend toward the bottom of the tooth in either an intermittent (80) or gradual (113) fashion. As the pockets deepen, they provide a microbial niche where as many as 10⁸ to 10⁹ bacterial cells can accumulate (281). The deeper the pocket, the more isolated and inaccessible to oral hygiene procedures the subgingival plaque community becomes (321), so that these numbers remain relatively constant, with the slow microbial growth counterbalanced by the flushing of loosely adherent and dead organisms from the pocket by a tissue transudate called the gingival crevicular fluid (GCF) (35, 50).

The continuing presence of such large numbers of bacteria probably accounts for the varied host defense mechanisms against bacterial invasion and growth that can be found in the gingival tissues, i.e., high blood flow due to the presence of two separate arterial networks (258), large numbers of neutrophils in the GCF (24), large numbers of mononuclear cells in the epithelium (258), elevated immunoglobulin G (IgG) and IgA titers to specific bacterial species (61-63, 299), the formation of tissue defensins (319), and a high turnover rate of the gingival epithelium (258). This bacterial load requires additional defense mechanisms, one of which is the encasement of bacteria in hardened deposits known as dental calculus or tartar. Such deposits have been significantly associated with and have been suspected of contributing to periodontal disease (10, 46). However, subgingival calculus could be a consequence of the pocket, forming when bacterial cells at the base of the plaque die and then calcify (182). As such, calculus could be viewed as an effort by the host to prevent biologically active molecules like LPS from entering the gingival tissues, leaving only those microbes on its surface to provoke the approximating soft tissue. This interpretation would be supported by the observation in the 1988 to 1991 National Health and Nutrition Evaluation Survey (NHANES III) that 88% of sites with subgingival calculus did not bleed when gently touched with a dental instrument (37, 38, 217).

Another effective host defense mechanism is the highly vascularized gingival tissue, which presents an oxidative barrier to the penetration of the anaerobic flora from the dental plaque. The pO₂ of a 6-mm-deep pocket is about 13 to 15 -mm Hg (155), so that when bacteria living in that environment penetrate the gingival tissue and encounter a tissue pO₂ of 140 to 150 mm Hg, they are not likely to survive. While certain bacteria such as A. actinomycetemcomitans (47, 249), P. gingivalis (249), and spirochetes (243) can be detected within the tissue, they rarely are able to cause tissue necrosis. When necrosis does occur, as in noma or HIV-positive patients, the host is compromised by protein-calorie malnutrition (68) or T-cell deficiencies (204, 328). Conditions which cause a vasoconstriction of peripheral arterioles, such as smoking (30) and stress (73), are risk factors for periodontal disease, probably because the reduced blood flow allows some invading anaerobes to survive long enough in the tissues for them or their products to activate the latent interstitial collagenases. In this sense, the selection for anaerobes in the subgingival plaque may be beneficial to the host, since if facultative species were dominant, tissue invasion and necrosis might be more common. For example, anaerobic species are rarely isolated from bacteremias associated with dental procedures whereas facultative streptococci and Actinomyces species are frequent isolates (143, 148, 245), suggesting that they can survive in the gingival tissue long enough to enter the bloodstream.

The cited defense mechanisms are overwhelmingly beneficial to the host (258), with the only long-term detriment being a slow and intermittent loss of attachment of the teeth to the alveolar bone. A small percentage of tooth sites may show a burst of 2 to 3 mm of attachment loss (80), and a small percentage of individuals may show such rapid deterioration that they lose many of their teeth at an early age. In the former case, the incidence of "active" sites is so small and the bacterial flora is so variable that it is difficult to obtain a sufficient number of patients to demonstrate unequivocal differences between active and inactive sites (59, 101, 199, 294, 320). In the latter case, these rapidly progressing forms often occur within families, raising the possibility that there is a genetic component (213). For some rare inherited and chromosomal disorders, such as Papillon-Lefevre syndrome, Ehlers-Danlos syndromes, and Chédiak-Higashi syndrome, severe early-onset forms of periodontal disease are often characteristic of the syndrome and reflect a fundamental defect in epithelial cell, connective tissue, or leukocyte function (109, 287). Such enhanced deterioration is seen in other conditions with a genetic component, such as Down syndrome (4, 16, 49) and diabetes (219, 248, 250), especially diabetes among the Pima Indians (206).

In the majority of rapidly progressing forms, the genetic component is subtle and presumably manifests as an altered host response(s) to the bacterial flora (108). In one scenario, the host monocytes would overreact to a small bacterial challenge and produce large amounts of inflammatory mediators such as prostaglandins and cytokines (214). In another scenario, defects in leukocyte mobility and/or adhesion would cause a sluggish host response in the gingival environment (314), which results in bacterial overgrowth in the plaque and their presence in the tissue. This mechanism is supported by the in vitro finding that many patients with rapidly progressing periodontitis have neutrophils and/or monocytes that exhibit various chemotactic defects (23, 218). Neutrophils taken from patients with rapidly progressing periodontitis have a significantly lower expression of L-selectin (CD62L) (176), increased basal H₂O₂ production, and decreased L-selectin shedding. The last impairment, which correlated with increased interleukin-8 levels in plasma, could contribute to initial vascular damage (72). However, even with these genetic predispositions, a trigger or bacterial challenge from the plaque flora is needed to cause the tissue loss.

ARE WE DEALING WITH AN INFECTION?

Top Previous Next References

Nonspecific Plaque Hypothesis

The bacterial composition of the plaque has long defied comprehension and has led to the concept that it is the nonspecific overgrowth of any or all bacteria that causes dental disease. This tradition dates back to the 19th century, when Willoughby Miller (192), a student of Koch, had hoped to identify one or several bacterial species as being responsible for dental decay (caries). However, given the limited taxonomic data on the oral bacterial species and a complete ignorance of distinct microbial niches within the oral cavity, he concluded that caries was bacteriologically nonspecific. Miller and others reasoned that because acid demineralizes the tooth and all plaque bacteria produce acid, then all bacteria contribute to decay, especially when bacteria accumulated on tooth surfaces that were difficult to keep clean. Corrective treatment required that bacterial accumulations be eliminated and/or reduced in the retentive sites on the top of the teeth (occlusal surfaces) and between the teeth by daily toothbrushing and frequent dental cleanings or prophylaxis. Dentifrices with abrasive pumices were introduced, and the objectionable taste of the pumice was sweetened with sucrose! Mouths became cleaner, but the prevalence of dental decay approached 100% and the severity of decay resulted in tooth extractions in almost everyone, with complete tooth loss, i.e., edentulism in about 60% of older individuals (329). Dental decay became a public health problem, and only when fluoride was introduced into drinking water and dentifrices (coupled with the quiet removal of sugar) did it abate.

The treatments that were ineffective in caries control, i.e., brushing, flossing, and dental "prophylaxis," however, reduced gingivitis and became the definitive treatment modalities for the prevention of periodontal disease (217). However, an estimated 2 million to 6 million people in the United States still have periodontitis (9, 217). The treatment of this more advanced condition remained essentially the same as for prevention, namely, the debridement of the tooth at periodic intervals over a lifetime by the hygienist or dentist, supplemented, when there are deep pockets, by surgical procedures that eliminate or greatly reduce the depth of the pocket. This debridement approach is based on the premise that bacterial overgrowth per se is the cause of periodontitis.

Bacterial complexity of dental plaque. No one knows how many bacterial species, ribotypes, and serotypes coexist in the dental plaque, but the number is very large. Moore and Moore isolated 509 species from only 300 individuals, with most being previously undescribed species (200). Most cultivable species were present in such low proportions that it was difficult to associate the overgrowth of any single species with periodontal disease. With the advent of PCR technologies, many new uncultivable species are being identified. For example, when a single plaque sample was screened with a treponeme-specific oligonucleotide probe, 23 species, including 19 new species, were identified from 81 sequenced spirochetal clones (43). Although the levels of the cultivable spirochetes, T. denticola and T. vincentii, increased in plaques from diseased sites, these organisms were not the most numerous spirochetal types present in diseased sites (201).

Nonspecific mechanisms. Many of the proposed mechanisms by which bacteria provoke the inflammatory response are nonspecific. If the host responds in diverse ways to LPS which enters the tissue, would the LPS be more likely to come from any gram-negative species rather than from a uniquely specific organism? The prostaglandin E₂ levels observed in the GCF are highly correlated with levels of prostaglandin E₂ secreted by peripheral blood monocytes in vitro in the presence of bacterial endotoxins (216). The levels in GCF increase with the severity of the clinical condition, suggesting that the more bacterial endotoxin in the plaque, the more prostaglandin E₂ that will be secreted. Most subgingival species are proteolytic and produce an array of volatile fatty acids including butyrate, propionate, and isobutyrate (85, 208, 261), as well as sulfides such as hydrogen sulfide and methyl mercaptan, as metabolic by-products (226, 227). All these compounds could be cytotoxic to the gingival tissue, based on tissue culture studies in which propionate and butyrate inhibited the growth of cultured epithelial and endothelial cells and fibroblasts (114, 123) and in which sulfides inhibited epithelial cells (233, 244).

Treatment based on the nonspecific plaque hypothesis. While the above interpretations of the data are reasonable, the resulting treatment paradigm mandates that the flora be suppressed either continuously or periodically by mechanical means, so as to maintain bacterial levels compatible with gingival health. When this traditional debridement approach fails, patients are considered to have a refractory periodontitis and are suspected of either having a subtle genetic predisposition to periodontal disease or being noncompliant in their oral hygiene practices (212, 325). For these individuals, antimicrobial agents are chosen which kill as many bacterial types as possible. This encourages the usage of broad-spectrum agents such as tetracycline (110, 132) or the combination of agents such as amoxicillin and metronidazole (316). Since the plaque flora would have to be suppressed either continuously or periodically, this approach could lead to the overuse of these agents. Consider the following quote taken from a study evaluating the ability of clindamycin to control the deteriorating situation found in refractory patients: "During the year prior to entering the study, each patient had received antibiotics as part of the periodontal treatment. Tetracycline therapy ranged from one week to one year duration with most patients receiving four or five administrations of 250 mg qid for 10 to 14 days. Every patient was also treated with at least one other antibiotic and these included penicillin V, ampicillin, augmentin, erythromycin, cephalexin or metronidazole" (81). In another study, nine patients refractory to the debridement and surgical approach had been treated with either penicillin, tetracycline, minocycline, or metronidazole (309). In still another study, 17 different antimicrobial regimens were used by 23 clinicians in the treatment of recurrent (refractory) periodontitis (127).

To change the treatment paradigm from the nonspecific reduction of plaque mass to one based on principles of antibacterial management of infections will require convincing evidence that it is the overgrowth of a limited number of bacterial species in the dental plaque that can produce the destructive inflammatory changes in the gingival tissue. Such evidence was able to change the paradigm concerning the bacterial etiology of dental decay (147).

In describing dental decay, Miller was correct in recognizing that retentive sites on the tooth surface are predisposed to decay, but he had no way of knowing that cariogenic organisms, such as the mutans streptococci and lactobacilli, are selected for in these stagnant environments. After a brief exposure to dietary sucrose, the plaque pH will quickly drop to about 5.0 to 5.5. While most supragingival plaque bacteria produce acid, they are less active at these pHs and may cease to grow. However, at this pH the tooth hydroxyapatite begins to dissolve (demineralize), serving as a buffer which allows the mutans streptococci and lactobacilli, due to their aciduricity, to survive and their numbers to increase (149). If the demineralization process is not reversed by the remineralizing potential of saliva, the mineral lost from the tooth first appears as a white spot on the tooth surface and then progresses to dental decay. Thus, the caries process reflects a selection of plaque organisms that can survive in the low-pH environment occasioned by frequent access to sucrose. The nonspecific plaque hypothesis was successfully challenged when studies in germfree animals showed that most acidogenic bacteria were not cariogenic (71) and when Keyes demonstrated the infectious and transmissible nature of dental decay in animal models (118).

If dental decay was a specific infection, why could periodontal disease not also be a specific infection resulting from the selection of bacteria that can grow in the stagnant pocket environment, using nutrients which leak into the pocket in the GCF as the result of the microbes' production of proinflammatory molecules? In the past 25 years, over 200 studies have compared the flora of disease-associated plaques with the flora found in plaques associated with periodontal health. The results have generally shown a limited number of bacterial species, mainly gram-negative anaerobes, to be significantly associated with periodontal disease (see Tables 2 to 5). These findings have not changed the prevailing treatment philosophy in periodontal disease, because of the powerful legacy of the nonspecific plaque hypothesis in dictating treatment protocols that have become the standard of care in clinical dentistry. It is difficult to change a treatment approach whose 80% level of effectiveness is accepted by the clinician (111, 189, 190) and which provides the economic infrastructure of clinical periodontology.

There are certain periodontal conditions that do not conform to the "bacterial overgrowth hypothesis," such as LJP in which the teeth are "clean," and acute necrotizing ulcerative gingivitis (ANUG), which has a sudden onset. Both of these conditions are rare, affect young individuals, and seem to involve a host component (defective neutrophils in LJP [74] and psychological stress in ANUG [51]). However, more importantly, they appear to be specific bacterial infections which can be successfully managed by antimicrobial treatments directed at specific bacterial species (58, 131). If the bacterial flora in these exceptional conditions resembles that found in the more common forms of periodontal disease, then an antimicrobial treatment approach may extend to other forms of periodontal disease.

(i) Localized juvenile periodontitis. LJP occurs among teenagers, with a prevalence of about 1 to 5 in 1,000 (36, 217, 310). In its classic form, the pathology is confined to the teeth that erupt in the mouth at about 6 years of age, i.e., first molars and incisors, although deep pockets are usually not discovered until after puberty (344). These pockets are notable for small amounts of plaque and the absence of calculus, so that a "dirty mouth" could not be evoked to explain the observed attachment and bone loss. Metabolic disorders in calcium metabolism were suspected but never documented. The condition was considered degenerative and was labeled as periodontosis, and the involved teeth were often extracted (344). However, when plaques associated with these lesions were shown to contain mostly unknown species, one of which was subsequently identified as A. actinomycetemcomitans, a specific microbial etiology was suspected (207, 344). Traditional debridement and surgical procedures, combined with short-term usage of locally delivered antimicrobial agents and systemic tetracycline, resulted in the retention of teeth that formerly were considered hopeless (131, 186, 275). Thus, a new treatment paradigm was introduced, namely, that LJP was a treatable bacterial infection.

(ii) Acute necrotizing ulcerative gingivitis. ANUG tends to occur in young individuals and is characterized by a sudden onset, acute pain, and necrosis of the tissue between the teeth, i.e., the interdental papilla. Vincent, in the late 19th century, described it as a fusospirochetal infection due to the prominence of these organisms in smears of material removed from the lesion. It is the trench mouth of World War I and has a long history of association with military personnel and other individuals under stress (67, 76). Bacteria, especially spirochetes and including an extremely large spirochete with more than 20 axil fibrils inserted at each end, can be found in the tissue in advance of the necrosis (136). This large spirochete has never been cultured and is seen in plaque samples associated with disease (200). Quantitative bacteriological studies, comparing diseased and healthy sites in the same patient revealed a significant increase in the number of spirochetes and Prevotella intermedia in the diseased sites (68, 162).

Is periodontal disease an anaerobic or a microaerophilic infection? If LJP and ANUG could be treated as bacterial infections, could other forms of periodontal disease also be associated with specific bacterial types and treated in the same way? The older literature identified anaerobic organisms such as spirochetes and black-pigmented Bacteroides species (now classified as Porphyromonas and Prevotella species), as putative periodontal pathogens (166, 175, 247). The importance of anaerobes was reinforced by microscopic examination of plaque samples which showed spirochetes increasing, both in numbers and in proportions, as the clinical condition worsened (133, 139, 164, 201, 239) and by culture studies which showed increased proportions of Prevotella and Porphyromonas species in most forms of periodontal disease (164, 270, 294). With the identification of A. actinomycetemcomitans as a putative periodontal pathogen, emphasis shifted from anaerobes to this microaerophilic species. Selective media were developed that allowed its detection even when it was outnumbered 1,000:1 by other plaque species (185, 269). On the basis of its prevalence in plaque samples, it was implicated as a putative periodontal pathogen in refractory periodontitis, EOP, and the rapidly progressive lesion (59, 99, 272, 294, 314).

(i) Early-onset periodontitis (aggressive periodontitis). Our description of aggressive forms of periodontitis will include bacteriological studies on individuals 35 years or younger, with the prototype clinical example being LJP (Table 2). LJP is unique among periodontal clinical entities in that it can be unequivocally identified by its occurrence about molar and incisor teeth in young individuals in the absence of obvious plaque and calculus accumulations. However, periodontitis is also observed in other young individuals in association with plaque and calculus (7) and results in tooth loss before the age of 20 years. The prevalence of these early-onset forms varies from about 0.05 to 0.2% in European children, from 0.65 to 2.3% in U.S. children, and up to 3.7% in Brazilians (145). Some of the variability in prevalence depends on the criteria used to define disease. For the United States in 1986 to 1987, 70,000 adolescents were estimated to have LJP, another 17,000 were estimated to have a more generalized destructive form involving more teeth (called generalized juvenile periodontitis), and another 212,000 were estimated to have what was defined as an incidental loss of attachment, with one or more teeth exhibiting an attachment loss of 3 mm (145). In addition, there is an aggressive clinical entity seen in young adults between 18 and 35 years of age, which is known by several names, i.e., generalized EOP, generalized destructive periodontitis, severe periodontitis, or rapidly progressive periodontitis, and is estimated to affect about 1,780,000 Americans (217).

(ii) Adult periodontitis (chronic periodontitis). The numerically more prominent clinical condition is the periodontitis found in adults older than 35 years, which may be the most common chronic infection among Americans (324). Bacteriological investigations of AP were hampered for many years by the labor-intensive cost of anaerobic culturing procedures, so that many early studies included very few patients and/or plaque samples. With limited numbers of samples to compare, it was difficult to recognize any meaningful pattern of bacterial specificity.