Periodontal therapeutics: Current host‐modulation agents and future directions

Abstract With the recognition in the 1960s and 1970s of the periodontopathic importance of the microbial biofilm and its specific anaerobic microorganisms, periodontitis was treated as an infectious disease (more recently, as a dysbiosis). Subsequently, in the 1980s, host‐response mechanisms were identified as the mediators of the destruction of the collagen‐rich periodontal tissues (gingiva, periodontal ligament, alveolar bone), and the periodontopathogens were now regarded as the "trigger" of the inflammatory/collagenolytic response that characterizes actively destructive periodontitis. Also at this time a new pharmacologic strategy emerged, entitled "host‐modulation therapy", based on 2 major findings: (1) that the ability of tetracycline antibiotics to inhibit periodontal breakdown was due (in large part) to their previously unrecognized ability to inhibit the host‐derived matrix metalloproteinases (notably, the collagenases, gelatinases, macrophage metalloelastase), and by mechanisms unrelated to the antimicrobial properties of these medications; and (2) that nonsteroidal anti‐inflammatory drugs, such as flurbiprofen, again by nonantimicrobial mechanisms, could reduce the severity of periodontitis (however, the adverse effects of long‐term therapy precluded their development as safe and effective host‐modulatory agents). Additional mechanistic studies resulted in the development of novel nonantimicrobial formulations (Periostat® [now generic] and Oracea®) and compositions of tetracyclines (notably chemically modified tetracycline‐3) as host‐modulator drugs for periodontitis, arthritis, cardiovascular and pulmonary diseases, cancer, and, more recently, for local and systemic bone loss in postmenopausal women. Identification of the cation‐binding active site in the tetraphenolic chemically modified tetracycline molecules drove the development of a new category of matrix metalloproteinase‐inhibitor compounds, with a similar active site, the biphenolic chemically modified curcumins. A lead compound, chemically modified curcumin 2.24, has demonstrated safety and efficacy in vitro, in cell culture, and in vivo in mouse, rat, rabbit, and dog models of disease. In conclusion, novel host‐modulation compounds have shown significant promise as adjuncts to traditional local therapy in the clinical management of periodontal disease; appear to reduce systemic complications of this all‐too‐common "inflammatory/collagenolytic" disease; and Oracea® is now commonly prescribed for inflammatory dermatologic diseases.


| INTRODUC TI ON
In this article we focus on a pharmacologic strategy for managing patients with chronic inflammatory periodontal disease. This strategy, termed "host-modulation therapy", was developed almost 3 decades ago by Golub et al 1,2 . To date, the only host-modulation therapy used clinically in the USA (approved by the US Food and Drug Administration) and beyond (Canada, Europe) is a nonantibiotic formulation of doxycycline, a member of the tetracycline antibiotics (discussed below). This seemingly counterintuitive approach arose from seminal discovery experiments more than 3 decades ago, [3][4][5][6][7] which resulted in an initial series of review articles a few years later that proposed the clinical use of this nonantibiotic formulation as a novel, safe, and effective therapeutic strategy as an adjunct to scaling and root planing. 1,2,6,7 This strategy has also been tested in surgical regimens of periodontal therapy. 9 As reviewed in several publications since, 10-16 2 major categories of host-modulation therapy have received the most attention. The first category modulates the host's inflammatory response either by inhibition 18 or, as described more recently, by resolution. [15][16][17][19][20][21] The second category (the main focus of this chapter) modulates the host's pathologic collagenolytic response in the soft tissues (gingiva and periodontal ligament), as well as the alveolar bone. It should be stressed that collagens in periodontal tissues, comprised mostly of type I but also other collagens, such as type III, are the major structural proteins of all of these soft and calcified tissues. In fact, this ubiquitous fibrous protein comprises over 90% of the organic matrix of the calcified periodontal tissues, the bone, and the cementum, and about 60% of the gingiva and periodontal ligament. 1,2 Regarding the first category of host-modulation therapies,2 approaches have been intensively investigated. The earliest studies involved nonsteroidal anti-inflammatory drugs, but this strategy has been rejected. In brief, the nonsteroidal anti-inflammatory drug that received the most attention in animal studies, and then in clinical trials, was flurbiprofen. Similar to other nonsteroidal anti-inflammatory drugs, flurbiprofen suppresses the host's inflammatory response, including its well-known mediators (eg, prostanoids, cytokines), but also inhibits osteoclast activity and bone resorption. 18 However, because of significant adverse events in long-term clinical trials testing nonsteroidal anti-inflammatory drug efficacy in periodontal patients, including a rebound effect of accelerated alveolar bone loss after cessation of this drug, 22 these compounds have not been approved for clinical use as a host modulator by governmental regulatory agencies in any country.
In contrast to nonsteroidal anti-inflammatory drugs, there is previously unrecognized multiple mechanisms of action of tetracycline antibiotics, as host modulators, were those published by the same group. 1,2 These research efforts focused on: (1) the ability of different formulations of subantimicrobial-dose doxycycline and compositions (eg, the chemically modified tetracyclines) as nonantimicrobial tetracyclines, to inhibit the pathologic breakdown of collagen-rich tissues, including the resorption of bone; and (2) to explain their mechanisms of action, which are outlined in Table 1 and are discussed briefly below. More than 10 years later, the field had expanded so dramatically in both dentistry and medicine that 2 issues of Pharmacological Research 33,34 were published following a number of related scientific symposia (eg, the New York Academy of Science, The Gordon Research Conferences, etc.) dedicated to the multiple clinical uses of these tetracycline-based host modulators (mostly doxycycline) for a number of "inflammatory/collagenolytic" diseases, including (but not limited to) periodontitis, arthritis, dermatologic diseases, cardiovascular and lung diseases, and cancer. In addition, the discovery of tetracyclines as host modulators for various medical as well as dental diseases was highlighted in 2 editorials published in JAMA. 35,36 As background, and of relevance to the drug-discovery experiments (for more detail, see Golub et al 1,2 ), a series of studies was designed to elucidate abnormalities in collagen structure, synthesis, crosslinking/maturation, and degradation, as a complication of experimental diabetes. These experiments led to the discovery that this hyperglycemic state upregulated the synthesis and secretion of the only host-derived neutral proteinases, the collagenases (ie, matrix metalloproteinase-1, -8, and -13), capable of degrading the triple-helical collagen molecule under physiologic conditions of pH and temperature. 4,37 Consistent with this observation, diabetes was also found to result in accelerated aging of collagen in connective tissues (including gingiva, skin, bone matrix), characterized by a reduction in the ratio of newly synthesized uncrosslinked (soluble) collagen to older highly crosslinked (insoluble) collagen. The production of excess collagenase provided a mechanism (in addition to decreased procollagen synthesis and excessive lysyl oxidase/crosslinking enzyme activity), the newly synthesized collagen is much more susceptible to TA B L E 1 Nonantibiotic tetracyclines as host modulators inhibit connective tissue breakdown: pleiotropic mechanisms of action a A. Extracellular mechanisms • Direct inhibition of activated matrix metalloproteinases in connective tissues, dependent on Zn ++ and Ca ++ binding by nonantibiotic tetracyclines • Inhibition of promatrix metalloproteinase activation by reactive oxygen species scavenging, independent of cation binding by nonantibiotic tetracyclines • Inactivation (by partial proteolysis) of promatrix metalloproteinases, dependent on the binding of cations by nonantibiotic tetracyclines • Indirect inhibition of serine proteinases (eg, elastase) by preventing the matrix metalloproteinase-mediated breakdown of serum alpha 1 -proteinase inhibitor (ie, alpha1-PI, also known as alpha1-antitrypsin) B. Cellular mechanisms • Decreased expression of inflammatory cytokines, nitric oxide, and phospholipase A 2 , thus suppressing promatrix metalloproteinase expression C. Proanabolic effects • "Upregulated" collagen synthesis, osteoblast activity, and bone formation a Modified from Golub et al. 2 collagenolytic attack than the older "leather-like" highly crosslinked collagen. Also consistent with the high level of diabetes-induced pathologically-excessive matrix metalloproteinase (collagenolytic) activity was the excessive urinary excretion of hydroxyproline (an amino acid unique to collagen molecules), compared with that in normal glycemic rats, and the accelerated breakdown of the collagen-rich periodontal (and other) tissues often seen in patients with diabetes. 4 To identify which of these 2 hypotheses was correct, the tetracycline antibiotic, minocycline, was administered to rats with diabetes to suppress the diabetes-induced altered crevicular microflora (hypothesis 1), and was found to reduce substantially the excessive host-derived collagenase activity (in the absence of any reduction of hyperglycemia). These observations initially appeared to support hypothesis 1. However, when the experiment was repeated using germ-free rats, maintained in germ-free environments, minocycline therapy again reduced the pathologically excessive gingival collagenase activity in the rats with diabetes back to normal rat levels.

| Therapeutic rationale
Golub et al 2 then designed 2 strategies of drug development to suppress connective tissue breakdown, including bone loss during periodontitis, and other dental and medical diseases, using nonantibacterial tetracyclines. Both strategies have recently been reviewed by this group. 13,24 In brief, the first strategy involved the custom formulation, then testing, of capsules containing standard levels of doxycycline (50 and 100 mg), which have antibiotic activity, as well as capsules containing lower levels of doxycycline (30, 20, and 10 mg), which are recognized to have non therapeutic antibiotic activity. In this program of design and testing, the 20 mg formulation in clinical studies was found to produce peak blood levels that were significantly lower than the >1.0 μg/mL needed for antibiotic activity. This low-dose formulation yielded peak blood levels of 0.3-0.6 μg/mL showed no difference compared with placebo capsules (vehicle only) in the emergence of an antibiotic-resistant microflora 47,50,51 and no difference in adverse events between the control and subantimicrobial-dose doxycycline groups over this prolonged period of time (Table 2). However, consistent with medical applications of subantimicrobial-dose doxycycline, the women on the 2-year regimen of the Food and Drug Administration-approved dental medication did show a statistically significant reduction in dermatologic adverse events ( Table 2) as well as statistically significant improvements in periodontal disease and systemic biologic parameters (Table 3).
Beneficial long-term results were also seen in clinical trials on patients with rheumatoid arthritis, 52 acne, 49 and other medical conditions, as discussed later in this review.

| CURRENT CLINI C AL HOS T-MODUL ATION AG ENTS: THER APEUTIC EFFI C AC Y AND SAFE T Y
As reviewed by our group and others, 1 25,26 The results of these studies are presented later in this review. ability of the medication to produce a therapeutic effect for some time after drug administration has stopped.

| Safety and efficacy of subantimicrobial-dose doxycycline as an adjunct to nonsurgical and surgical periodontal therapy in humans
In early studies, Golub et al 6 administered subantimicrobial-dose doxycycline to patients with periodontitis for 2 weeks (control subjects received placebo capsules twice a day) and reported "that this regimen dramatically reduced the mammalian collagenase activity, not only in the gingival crevicular fluid, but also in the adjacent gingival tissues that were surgically excised for therapeutic purposes." However, subsequent clinical trials indicated that although increasing the regimen of subantimicrobial-dose doxycycline to 4 weeks also reduced periodontitis biomarkers, cessation of host-modulation therapy resulted in a rapid rebound of collagenase activity back to placebo levels. 10,13 As a result, the duration of host-modulation therapy as an adjunct to scaling and root planing was increased in later clinical trials by various groups 10,50,53,58,59 to regimens of 3, 6, and 9 months, and 1 and 2 years. This strategy provided solid evidence not only of prolonged clinical efficacy and safety, but also of substantivity, which is discussed below.
After early discovery studies demonstrated that short-term regimens of subantimicrobial-dose doxycycline inhibited inflammatory/collagenolytic biomechanisms of periodontal disease in humans, and without the emergence of antibiotic-resistant microorganisms, 2 longer-term studies were then carried out. 8 As an example of the impressive scope of some of these studies, Ciancio and Ashley 61 described a multicenter, placebo-controlled double-blind randomized clinical trial of 531 patients with chronic periodontitis to assess long-term efficacy of subantimicrobial-dose doxycycline.
In brief, clinical improvements were seen, over a 12-month time period, in pocket depth, clinical attachment level, bleeding on probing, alveolar bone height (assessed using subtraction radiography), and disease activity (defined as "incidences of rapid progression"). 61 Moreover, the 1-year administration of subantimicrobial-dose doxycycline produced a safety profile of adverse events that was the same as for the placebo (there was also no effect on diagnostic laboratory parameters, including liver and kidney function), and no microbial resistance was developed to tetracyclines or other antibiotics in the patients treated with subantimicrobial-dose doxycycline.
Subsequent clinical studies addressed the substantivity of hostmodulation therapy, namely the ability of subantimicrobial-dose doxycycline to produce therapeutic benefits for months after the patient has stopped taking the medication, which are now discussed.
Although Caton and Ryan 10 reported that a 1-month treatment with subantimicrobial-dose doxycycline was not sufficient to produce a long-term benefit, a 3-month regimen did produce prolonged improvement during the subsequent no-treatment phase of the study. Consistent with these findings, a 1-year double-blind placebo-controlled study of 190 participants with chronic periodontitis, which involved 5 dental centers, reported significant reductions in probing depth, gains in clinical attachment, and prevention of disease progression, over a 9-month time period of subantimicrobialdose doxycycline therapy. These clinical benefits were maintained for at least 3 months after the treatment was stopped.
This substantivity of clinical improvement was also seen in studies of patients with rapidly progressing severe periodontitis. In one such study, participants with severe disease were administered a 6-month regimen of subantimicrobial-dose doxycycline adjunctive to repeated scaling and root planing treatments. The dramatic improvement in clinical response, compared with repeated scaling and root planing plus placebo treatments, was maintained for at least 3 months after the cessation of host-modulation therapy. 14,48,62 In another study, a 3-month regimen of subantimicrobial-dose doxycycline adjunctive to nonsurgical therapy administered to smokers with severe periodontitis also produced significant substantivity. 53 Moreover, a demonstration of prolonged substantivity was provided by Emingil et al. 63 In this 1-year placebo-controlled double-blind study, the patients with periodontitis were administered adjunc-  Tables 2 and 3. 50,51,54,64-66 As shown in Table 2, there was no evidence of adverse events in the postmenopausal women taking subantimicrobial-dose doxycycline, adjunctive to nonsurgical therapy, for the 2-year study period compared with placebo. 66 However, 1 category did show a significant effect, but it cannot be described as an adverse event; namely, the women who were administered subantimicrobial-dose doxycycline showed a highly significant (P = .002) 90% reduction in dermatologic adverse events. As discussed below,

| THE CHEMI C ALLY MOD IFIED TE TR AC YCLINE S: DE VELOPMENT AND PRELIMINARY CLINIC AL S TUDIE S
Soon after the development of nonantibiotic formulations or dosage regimens of doxycycline (20 mg twice a day) as safe and effective host-modulator medications, a second strategy emerged involving the chemical modification of the tetracycline (or doxycycline or minocycline) molecule to eliminate its antibacterial activity selectively, but to retain or even enhance its matrix metalloproteinase-inhibitory (host-modulating) properties. As first described by Golub et al, 2,147 this involved the removal of a side-chain on the tetracycline molecule, the dimethylamino group at carbon-4 of the A ring, which is responsible for the drug's antibiotic activity, and the preservation of the zinc-and calcium-binding site (the beta-diketone moiety) at carbon-11 and carbon-12, which we identified as the site on the tetracycline molecule responsible for its matrix metalloproteinaseinhibitory activity. The elimination of this metal ion-binding site, by chemically converting the tetracycline molecule to its pyrazole analog, eliminated the drug's matrix metalloproteinase-inhibitory properties. 2 As a result of these, and additional chemical modifications, a series of nonantibacterial matrix metalloproteinase-inhibitory tetracyclines was developed. One of these, chemically modified tetracycline-3, was found, based on in vitro and animal studies, to be a potent matrix metalloproteinase inhibitor with significant therapeutic potential. 24 In vitro, chemically modified tetracycline-3 was in phase I and phase II clinical trials. In phase I trials, the maximum tolerable dose was determined to be 150 mg/d. When administered orally, the compound was found to be rapidly absorbed, producing a peak blood level of 2-5 μg/mL, with a prolonged serum half-life of approximately 40 hours. 75,77,78 In this study, chemically modified tetracycline-3 (CMT-3, or COL-3) was administered once a day to patients with AIDS-related Kaposi sarcoma, who exhibited a 44% response rate, reflecting a decrease of angiogenic lesions, which was associated with a suppression of matrix metalloproteinase-2 levels in the circulation. In a larger phase II clinical trial, a group of patients with Kaposi sarcoma, who received a daily oral dose of 50 mg of chemically modified tetracycline-3, showed a significant reduction of angiogenic lesions, whereas the individuals who were administered the higher dose (100 mg once a day) did not. One explanation is that the increased side effects of the higher dose resulted in decreased patient compliance. Significant reductions in plasma levels of matrix metalloproteinase-2 and matrix metalloproteinase-9 were observed, and the most common adverse events were photosensitivity and rash, which could be quite severe. 26,75 The same compound was found to reduce mortality in a pig and other animal models of an often-fatal lung disorder, acute respiratory distress syndrome, and showed evidence of efficacy in models of septic shock. [79][80][81][82] The only human study testing chemically modified tetracycline-3 on periodontal disease was a pilot trial, conducted by Ryan et al. 25

| NONANTIB I OTI C TE TR AC YCLINE S: ADDITIONAL CLINIC AL APPLIC ATIONS IN MEDICINE
Because collagen is the major structural protein in connective tis-

| Ophthalmology
Soon after the discovery of tetracyclines as matrix metalloproteinase inhibitors, one of the first medical applications was for noninfected or sterile corneal ulcers. 83 Perry reported that a patient with a sterile corneal ulcer, refractory to standard treatment, responded dramatically within 48 hours after initiating tetracycline therapy, with no recurrence long term (see Figure 1). In a followup case series involving 18 patients with persistent sterile corneal defects, most patients had healed lesions within 48 hours after initiating oral tetracycline therapy and all lesions were resolved within 2 weeks. 84 Given that various cell types in the cornea, including the epithelium, fibroblasts, macrophages, and neutrophils, produce a variety of matrix metalloproteinases (matrix metalloproteinase-1, -8, -2, -9), the response to tetracycline host-modulatory therapy, although profound, was not surprising. 85 Subsequently, both topical and systemic tetracycline therapies were effective in rabbit models of sterile corneal ulcers 86 and all three tetracyclines, including minocycline and doxycycline, were effective inhibitors of corneal collagenases. 87 Federici 88 reviewed a number of studies using tetracyclines for sterile corneal ulcers, chemical burns of the eye, and blepharitis of the eyelid, and concluded that these (and other ophthalmic diseases, such as cataract, macular degeneration, and diabetic retinopathy) may benefit from the nonantibiotic properties of tetracyclines and other novel matrix metalloproteinase inhibitors.

| Dermatology
Early reports on tetracycline's efficacy in serious dermatologic lesions, soon after their matrix metalloproteinase-inhibitory properties were discovered, 1,2 focused on the hope that this safe collagenase inhibitor would be useful in a rare, and often fatal, skin disease, dystrophic epidermolysis bullosa. In this regard, White 89  The mechanism proposed involved tetracycline's inhibition of various matrix metalloproteinases (matrix metalloproteinase-2, matrix metalloproteinase-9, matrix metalloproteinase-13), which mediate the degradation of a basement membrane constituent, type XVII collagen. Another dermatologic application was described by Humbert et al. 90  In this oral disease, nonantibiotic doxycycline inhibited accelerated degradation of the hemidesmosomal protein, type XVII collagen, by matrix metalloproteinase-9, thus preventing blister formation.
Anecdotal observations indicated enhanced efficacy when this regimen was combined with a nonsteroidal anti-inflammatory drug, consistent with similar observations in rheumatology (discussed below).
However, the most immediate medical impact of nonantibiotic doxycycline has been on the common skin disorders, acne and rosacea. As described earlier in this chapter, a major clinical trial on postmenopausal women that focused on the long-term response of periodontal and systemic (osteopenia) bone loss to nonantimicrobial doxycycline resulted in a 90% (P = .002) reduction in dermatologic adverse events (eg, acne, rash) compared with the incidence of adverse events in individuals given a placebo ( Table 2).
As summarized by Monk et al, 49  lesions (papules, pustules, nodules) and decreased erythema. In a subsequent study, the same investigators 94 found that although nonantibiotic doxycycline was as efficacious as higher (100 mg) antibiotic-dose doxycycline, the lower dose produced significantly fewer adverse events. Subantimicrobial-dose doxycycline was also found to be significantly more effective in treating rosacea, even when combined with topical metronidazole. 95 As with rosacea, acne has also responded significantly to subantimicrobial- Animal studies (rats, dogs) on experimentally induced rheumatoid arthritis and osteoarthritis also provide further evidence of the efficacy of nonantibiotic tetracyclines as host modulators. In some of these studies, a chemically modified tetracycline (eg, chemically modified tetracycline-1; 4-dedimethylamino tetracycline) was tested and, when combined with a nonsteroidal anti-inflammatory drug (such as indomethacin, flurbiprofen), a synergistic therapeutic response was observed. 100,101 Subsequent studies demonstrated that the administration of doxycycline in human osteoarthritis produced these benefits in association with suppression of the matrix metalloproteinases, collagenase and gelatinase. 98

| Cardiovascular, pulmonary, and other medical disorders
The research and development of nonantibiotic tetracyclines as host modulators continues to expand into additional medical conditions. However, because these initial studies and biologic mechanisms have been reviewed extensively, 13,33,34 and because the results of longer-term randomized clinical trials that demonstrate the efficacy of host-modulation therapies on diseases such as cardiovascular disease, pulmonary disorders, etc., have yet to be reported, these topics will only be briefly discussed.

| Diabetes and postmenopausal bone loss
The impact of nonantibiotic tetracycline host modulators (chemically modified tetracyclines and subantimicrobial-dose doxycycline) on two major medical conditions, namely diabetes and postmenopausal bone loss (oral and skeletal), has been summarized in several articles, including a recent review by our group. 24 Historically, the concept of host-modulation therapy originated from early experiments by our group targeting the impact of type 1 diabetes on periodontal inflammation, collagen turnover, and bone loss. 2,[4][5][6] The breakthrough occurred when tetracycline therapy was found to modulate the host response effectively even in germ-free rats with diabetes. Subsequent studies by Ryan et al 68 found similar periodontal and systemic biologic changes in rats with type 2 diabetes treated with host-modulating tetracyclines.
One long-held controversy in this field involved the reported efficacy of nonsurgical periodontal therapy (scaling and root planing combined with oral hygiene instruction and antiseptic mouthrinses) in reducing hyperglycemia in patients with diabetes. 110,111 As recently reviewed, 24 this was subsequently challenged in a multi-institutional clinical trial by Engebretson et al, 112 who did not find any significant improvement in glycated hemoglobin levels as a result of nonsurgical periodontal therapy. The same author reported that although scaling and root planing alone, or combined with adjunctive antibiotic therapy, did not significantly reduce glycated hemoglobin levels, when a 3-month regimen of host-modulation therapy (nonantibiotic doxycycline, 20 mg twice a day) was added to the scaling and root planing treatment protocol, the level of glycated hemoglobin was reduced. 58 Clearly this important topic in dentistry and medicine needs further investigation to confirm whether subantimicrobial-dose doxycycline, with prolonged administration, can significantly reduce hyperglycemia; and whether host-modulating periodontal therapy should be incorporated into the management protocol for diabetes, which is an increasingly prevalent and devastating disease worldwide.
The final topic to be discussed, before addressing newer host- A preliminary 1-year clinical trial was then carried out on postmenopausal women who were diagnosed with both local (periodontitis) and systemic (osteopenia/osteoporosis) bone loss.
Subantimicrobial-dose doxycycline, compared with treatment with placebo capsules, as an adjunct to periodontal maintenance therapy, reduced alveolar bone loss and suppressed the progressive loss of clinical attachment. 50 The totality of the in vitro, animal, and preliminary clinical study just described, led to a major bi-institutional

| Introduction
Although the discovery of the unexpected property of tetracyclines as matrix metalloproteinase inhibitors has been translated into novel first-generation (Periostat ® , Oracea ® ) and secondgeneration (chemically modified tetracycline-3) pharmaceutical agents, a third generation of related host-modulating therapeutics is now being developed (Figure 2). Based on our previous identification of the matrix metalloproteinase-inhibiting active site on the 4-phenolic-ring tetracyclines, 1,2,24 we began to focus on diphenolic compounds with the same cation (Zn ++ , Ca ++ )-binding site, the beta-diketone moiety. These included the bis-aroyl methanes and curcumin. The bis-aroyl methane compounds exhibited positive, but weak, matrix metalloproteinase-inhibitory efficacy in vitro, namely high IC 50 (μm) values, and were quickly abandoned. 30,31 Curcumin, a dietary herbal ingredient derived from turmeric, has historically been advocated as a safe and effective treatment for a variety of diseases. [30][31][32] However, it has long been known that this compound's insolubility and poor absorption by the oral route has limited its clinical application. [123][124][125] Recently, our Stony Brook University laboratories (Oral Biology and Chemistry departments, Dr. Francis Johnson) have synthesized, developed, and tested a series of novel chemically modified curcumins with various side chains added to the carbon-4 position of this biphenolic compound ( Figure 2). Based on these studies, a lead triphenolic compound, a triketonic phenylaminocarbonyl curcumin-2.24, was identified, which has been tested in vitro, in cell and tissue culture, and in vivo using various animal species, including mice, rats, rabbits, and, most recently, dogs. To explore further the potential of our lead compound as a host modulator, chemically modified curcumin-2.24 was orally administered to 2 different animal models of severe periodontal breakdown [27][28][29] . In both the locally (microbial endotoxin/lipopolysaccharide) and

| Discovery and therapeutic rationale
the systemically (diabetes)-induced rat models of disease, chemically mitogen-activated protein kinase pathways. 29 These data further supported the potent ability of this chemically modified curcumin to modulate intra-and extracellular mechanisms that are associated, at least in part, with the zinc-binding characteristics of this and related compounds. 29 This pattern of efficacy is also consistent with the pleiotropic characteristics of the zinc-binding nonantimicrobial tetracyclines. 2,24 The potency of chemically modified curcumin-2.24 therapy has also been demonstrated in a challenging, but clinically relevant, dental/medical situation using an animal model that mimics a patient with poorly controlled diabetes and severe periodontitis. Once again, chemically modified curcumin-2.24 did not reduce the severe hyperglycemia (at least during the 1-month experimental protocol), yet completely prevented alveolar bone loss, which was induced by multiple lipopolysaccharide injections into the gingiva of rats with severe diabetes. 29 Addressing the causality of this decreased periodontal breakdown, this novel therapeutic also suppressed the excessive levels of pro-and activated matrix metalloproteinase-2 and matrix metalloproteinase-9 (gelatinases/type IV collagenases) in the diseased gingival tissues, as well as the activated (lower molecular weight) form of the dominant collagenase, matrix metalloproteinase-8, in these oral tissues. The dominant cytokines in the inflamed gingiva, interleukin-1beta and interleukin-6 (tumor necrosis factor-alpha was not detected), were also reduced locally by this treatment, and similar beneficial biomarker changes were seen systemically in the circulation, as well as in skin.
Collagen atrophy and excessive collagen crosslinking (ie, "leather-like" texture) were also normalized in the animals with diabetes that were treated with chemically modified curcumin-2.24. This is of particular relevance to a common medical complication in patients with diabetes ie, impaired healing of wounds in skin, gingiva, and other tissues 32 .
The clinical complications (adverse events) associated with the severely hyperglycemic state in this rat model of type 1 diabetes, which included bleeding under the nails, inflammation of the sclera, and impaired wound healing, were also prevented by this treatment. 28,29 It should be noted that impaired wound healing in skin, especially in the lower extremities, can be a lifethreatening complication in patients with diabetes and is a major cause of limb amputation in humans. 128,129 This important complication and other serious medical disorders, and their response to chemically modified curcumin-2.24, are also discussed in the next section.

| Wound healing
The novel host-modulation therapy highlighted in this review was recently found to be exceptionally effective in normalizing, not just improving, the healing of skin wounds categorized as standard and severe in rats with type 1 diabetes. 32 Of particular interest, the therapeutic response occurred even in the presence of severe hyperglycemia (blood glucose levels of 500 mg/dL) and was equally effective when applied either topically or systemically. A 1% suspension of chemically modified curcumin-2.24, applied topically, produced optimal results (assessed using clinical, biochemical and histomorphometric measurements) based on a significant reduction of pathologically excessive matrix metalloproteinase-8 (collagenase-2) and hydroxyproline levels in wound tissue. An additional mechanism currently being investigated is the ability of chemically modified curcumin-2.24 to normalize a key resolvin (RvD1; a docosahexanoic acid derivative) and interleukin-10 (an anti-inflammatory cytokine), which are both underexpressed in macrophages isolated from severely hyperglycemic rats with diabetes. 130

| Arthritis
In contrast to the tetracycline-based host modulators that have been studied extensively in human and experimental rheumatoid arthritis and osteoarthritis, the chemically modified curcumins are just beginning to receive attention. In this regard, Katzap    Regarding resveratrol's therapeutic potential, significant issues remain, including poor bioavailability associated, in part, with rapid urinary excretion. 141

| CON CLUD ING COMMENTS AND FUTURE DIREC TIONS
With the increasing recognition of the overarching importance of the host's inflammatory/collagenolytic responses as the "driver" of tissue breakdown during periodontitis, new treatment paradigms are emerging. Accordingly, this chapter has highlighted the 2 mostdeveloped host-modulation therapy strategies to date. However, the specialty of periodontics has increasingly incorporated the placement of dental implants in its treatment strategies. This reflects, at least in part, its (and the public's) recognition of the significant enhancement of the dentition's function and esthetics by this technology.
However, a surprisingly high incidence of peri-implant mucositis and peri-implantitis (19%-65%) has been recognized as a significant