Volume 75 0 Number 8
`
`Subantimicrobial Dose Doxycycline
`Enhances the Efficacy of Scaling and Root
`Planing in Chronic Periodontitis:
`A Multicenter Trial
`
`Philip M. Preshaw,* Arthur F. Hefti,’r M. John Novak,it Bryan S. Michalowicz,§ Bruce L. Pihlstrom,“
`Robert Schoon'" Clarence L. Trummel,# John Dean,# Thomas E. Van Dyke,** Clay B. WalkerflJr
`and Mark H. Bradshawl’f
`
`
`Background: Previous studies have shown that subantimicrobial dose doxycycline (SDD) is of clinical ben-
`efit in the treatment of chronic periodontitis (CP). The aim of this study was to further assess the role of
`SDD as an adjunct to scaling and root planing (SRP) in the treatment of CP.
`Methods: A double—blind. randomized. placebo-controlled. multicenter clinical study was conducted to
`test the efficacy of SDD (20 mg doxycycline B.l.D.) in combination with SRP in subjects with moderate to
`severe CP. Two-hundred ten subjects were treated with a standardized episode of SRP and randomized to
`receive either adjunctive SDD or placebo for 9 months. Efficacy parameters included per-subject mean
`changes in clinical attachment level (CAL) and probing depth (PD) from baseline, and the total number of
`sites with attachment gains and probing depth reductions 22 mm and 23 mm from baseline.
`.
`Results: In periodontal sites with PD 4 to 6 mm and 27 mm (N = 209, intent-to-treat population), mean
`improvements in CAL and PD were greater following SRP with adjunctive SDD than SRP with placebo,
`achieving statistical significance in all baseline disease categories at month 9 (P <0.05). At month 9, 42.3%
`of sites in the SDD group demonstrated CAL gain 22 mm compared to 32.0% of sites in the placebo group
`(P <0.01). CAL gain 23 mm was seen in 15.4% of sites in the SDD group compared to 10.6% of sites in
`the placebo group (P<0.05). When considering the same thresholds of change in PD, 42.9% of sites in the
`SDD group compared to 31.1% of sites in the placebo group demonstrated PD reduction 22 mm (P <0.01),
`and 15.4% of sites in the SDD group compared to 9.1% of sites in the placebo group demonstrated PD reduc-
`tion 23 mm (P<0.01),
`Conclusion: Adjunctive subantimicrobial dose doxycycline enhances scaling and root planing. it results
`in statistically significant attachment gains and probing depth reductions over and above those achieved by
`scaling and root planing with placebo. J Periodontol 2004;75:1068-1076.
`KEY WORDS
`
`Adjunctive therapy; clinical trials, controlled; double-masked method; doxycycline/therapeutic use;
`multicenter studies; periodontitis/drug therapy; planing; scaling.
`
`
`Currently, Newcastle University School of Dental Sciences, Newcastle. U.K.-. pleviously. College oi Dentistry. The Ohio State University. Columbus, OH.
`Currently, Philips Oral Healthcare. lnc., Snoqualmie, WA; previously. College of Dentistry. The Ohio State University.
`Currently. Center for Oral Health Research, University of Kentucky. Lexington. KY: previously. (ii'iiversity of Pittsburgh School of Dental Medicine,
`Pittsburgh, PA.
`Oral lleallh Clinical Research Center. University of Minnesota, Minneapolis, MN.
`Currently. National institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda. MD; previously, Oral Health Clinical Research
`Center. Liniveisity of Minnesota.
`1] New York University, New York, NY.
`# University of Connecticut School of Dental Medicine. Farmington, CT.
`** Boston University. Boston. MA.
`11’ University of Florida. Gainesville, FL.
`H Covance, lnc., Princeton, NJ.
`
`The views expressed in this paper do not necessarily represent those of the National Institutes of Health or the United States government.
`
`
`
`

`

`] Periodontal
`' August. 2004
`
`
`l-lel'ti. Hovalt, ct ail.
`
`Prcshaw,
`
`acterial plaque is essential for the initiation of
`
`Behronic periodontitis (CP), but the characteristic
`
`clinical signs of CP (pocketing, loss of attacl'nnenl,
`alveolar bone destruction) occur primarily as a result
`of activation of host-derived immune and inflammatory
`defense mechanisms.l The development of an immune-
`inflammatory response in the susceptible individual
`results in the local production of a variety of proinflam-
`matory cytokines, enzymes, and mediators.2 Key among
`these are the matrix metalloproteinases (MMPs), a fam-
`ily of proteolytic enzymes produced by macrophages,
`tissue fibroblasts, junctional epithelial cells, and neutro-
`phils (PMNs). MMPs include collagenases, gelatinases,
`and metalloelastases and collectively have the capacity
`to degrade all components of the extracellular matrix
`of the periodontium. Destruction of collagen, the major
`structural protein in the periodontium,3 contributes to
`loss of attachment and alveolar bone resorption.
`The importance of MMPs in periodontal destruction is
`well established. Studies have demonstrated elevated col-
`
`lagenase levels in inflamed gingival tissues4 and in gin-
`gival crevicular fluid (GCF) from periodontal pockets.5
`Furthermore, increased GCF levels of activated neutrophil
`collagenase (MMP—8) are associated with degradation of
`periodontal tissues in progressive periodontitis.6 The regu-
`lation of MMPs involves a complex interplay between
`cytokines, prostaglandins, growth factors, hormones, and
`endogenous tissue inhibitors of MMPs (TlMPs).7 Proin-
`flammatory cytokines such as interleukin l-B (IL-1B) tend
`to upregulate MMPs, whereas TlMPs downregulate MMPs.
`An imbalance between activated MMPs and their inhibitors
`
`may result in the pathologic destruction of the extracellu-
`lar matrix observed in CP. Thus, any means of modulating
`production of MMPs in the host tissues is likely to have
`clinical benefit in the treatment of periodontal disease.
`Doxycycline, a member of the tetracycline family of
`antibiotics, has been shown to inhibit collagenase activ-
`ity in vitro8 and to reduce GCF collagenase levels in
`patients with CP.9'10 The anticollagenolytic properties
`of doxycycline are evident even when the drug is pre-
`scribed at subantimicrobial doses of 20 mg twice
`daily.11 Furthermore, the long-term use of subantimicro-
`bial dose doxycycline§§ (SDD; 20 mg doxycycline
`b.i.d.) is not associated with the development of anti-
`biotic resistance or detrimental shifts in the normal
`
`periodontal microflora.12'13
`The use of SDD as an adjunctive treatment for CP
`represents a new approach in the long-term manage-
`ment of this disease. The standard of care for CP is
`
`scaling and root planing (SRP), which has been shown
`to be efficacious in slowing or arresting periodontal dis-
`ease progression.”l5 Treatment outcomes following
`SRP may be compromised, however, particularly at sites
`with deep pockets and in patients with risk factors for
`
`moderate to severe CP.12 This study demonstrated that
`adjunctive SDD conferred benefit to patients, and
`improvements in clinical attachment level (CAL) and
`probing depths (PD) were significantly greater com-
`pared to patients who received placebo.
`The aim of the present study was to investigate fur-
`ther the role of SDD as an adjunct to SRP in the treat-
`ment of patients with moderate to severe CP.
`
`MATERIALS AND METHODS
`
`Study Design
`A nine-month, double-blind, randomized, placebo-
`controlled, multicenter, parallel group study was under-
`taken to evaluate the efficacy of SDD as an adjunct to
`SRP in the treatment of patients with moderate to
`severe CP. The study was conducted at six dental
`schools in the United States. Approval was granted by
`the respective institutional review boards of the study
`centers prior to study initiation. Written informed con-
`sent was obtained from all study subjects. Subjects
`were screened to confirm eligibility, and then underwent
`SRP at the baseline visit. Subjects were randomly allo—
`cated to receive either adjunctive SDD (doxycycline
`' hyclate 20 mg b.i.d.) or adjunctive placebo b.i.d. for
`thertreatment period ofithe study (nine months), com-
`mencing at baseline. Subjects returned to their respec-
`tive study center for evaluation 3, 6, and 9 months
`after baseline.
`
`Study Population
`Male and female patients 30 to 75 years of age with
`evidence of periodontitis (i.e., CAL and PD between
`5 mm and 9 mm with bleeding on probing [BOP]) in two
`sites in each of two quadrants were eligible for inclu-
`sion. Exclusion criteria were pregnancy or lactation;
`serious, chronic medical conditions (e.g., diabetes mel-
`litus, kidney, or liver disease); acute systemic infection;
`previous dental prophylaxis or periodontal treatment
`within 90 days of the baseline visit; requirement for
`antibiotic prophylaxis prior to dental procedures; use
`of non-tetracycline antibiotics within 6 weeks of base-
`line; use of tetracycline antibiotics within 3 months of
`baseline; hypersensitivity to tetracyclines; or the need
`for long-term (>2 weeks) daily antacid therapy with
`antacids containing aluminum, calcium, or magne-
`sium. Patients requiring chronic (2 weeks or more)
`antibiotic therapy or who participated in a periodon-
`tal clinical trial within 12 months of baseline were also
`excluded.
`
`Procedures
`
`Full mouth CAL and PD measures were recorded by
`trained examiners at the screening visit using manual
`UHC-15 periodontal probes. Two qualifying quadrants
`
`

`

`Subantimicrobial Dose Doxycycilne Enhances Efficacy of SRP
`
`Volume 75 0 Number 8
`
`were selected using pre—established disease criteria
`(i.e., the presence of at least two non-adjacent sites at
`separate teeth with CAL and PD between 5 mm and
`9 mm with BOP). The oral soft tissue examination
`included assessments of the lips and tongue, hard and
`soft palate, gingiva, mucobuccal fold areas, buccal
`mucosa, and the floor of the mouth. The head and
`neck were also examined. Radiographs taken within
`6 months of enrollment were used to confirm the clin-
`ical diagnosis of CP, and rule out any periapical or
`endodontic pathosis. Demographic information, details
`of oral hygiene practices, and a smoking history were
`also recorded at the screening visit.
`Subjects returned for baseline assessments within
`2 weeks of the screening examination. Sites that had
`previously been selected on the basis of CAL and PD
`were further required to exhibit BOP for continued par-
`ticipation in the study. SRP was then performed until
`the crown and root surfaces were visually andfor tac—
`tilely free of all deposits with a time allowance of up
`to l hour per quadrant. Both ultrasonic and universai
`or area-specific curets were permitted, as was local
`anesthesia.
`
`- Following SRP, subjects were randomized to receive
`SDD 20 mg b.i.d. or placebo b.i.d. (lzl randomization).
`Randomization was carried out at each investigational
`site based on a randomly generated pre-defined sched-
`ule with fixed block size that was equal for each site.
`Subjects were instructed to take study medication once
`in the evening and once in the morning,
`1 hour prior to
`meals, at approximately l2—hour intervals. Study med-
`ication and placebo were identical in appearance and
`dispensed in 3-month quantities. Subjects received full-
`mouth clinical assessments at 3. 6. and 9 months, and
`were contacted monthly by telephone for the record—
`ing of adverse events. At month 9, a dental prophy—
`laxis was performed and the subjects were exited from
`the study.
`
`Assessments
`
`Prior to the study, all examiners were trained and prac-
`ticed in each of the clinical measurements. Succes-
`sive (duplicate) measurements were compared to
`ensure a standard deviation <1.0 mm for repeated
`within-subject PD and CAL measures. in order to min-
`imize the impact of interexaminer variability, the same
`examiner assessed a given subject for the duration of
`the study.
`Prior to statistical analyses, sites were stratified by
`the degree of PD at baseline: sites with PD 1 to 3 mm
`were considered normal; sites with PD 4 to 6 mm
`were considered mild to moderately diseased; and sites
`with baseline PD 27 mm were considered severely
`diseased.16
`
`sured at baseline and 3-month intervals by manual
`probing at six sites around each tooth. The CEJ-GM
`and PD measures were rounded up to the next whole
`millimeter. CAL was calculated from the PD and (EJ-
`GM measurements. Efficacy parameters included: i)
`the change in CAL from baseiine; 2) the change in PD
`from baseline; and 3} the number of sites demon-
`strating BOP. For those sites with a baseline PD 24 mm,
`additional efficacy parameters included the number
`of sites that demonstrated attachment gains 22 mm and
`23 mm and PD reductions 22 mm and 23 mm from
`baseline.
`
`Adverse events and use of concomitant medications
`were documented in subject diaries and recorded at
`each study visit and during monthly telephone con—
`tacts. Compliance with study drug therapy was recorded
`by counting the number of tablets dispensed and
`returned. Laboratory tests (blood chemistry and CBC)
`were conducted at the screening visit and month 9.
`Vital signs, oral soft tissue examinations, and exami-
`nations of the head and neck were performed at screen-
`ing; baseline; and at months 3, 6, and 9. A pregnancy
`test was performed at screening and at months 3, 6,
`and 9 for women of childbearing potential.
`
`Statistical Analyses
`The study was designed to detect a difference of at
`least 0.25 mm in mean change from baseline PD
`betwaen treatment arms (two-tailed tests) with 90%
`power, at
`:: 0.05, and assuming standard error to be
`approximately 0.05 mm. lntent-to-treat analyses were
`performed on data obtained from all subjects having
`a baseline measurement and at least one efficacy mea-
`surement after starting treatment. A last-observation~
`carried-forward (LOCF) algorithm was used to impute
`missing data at each time point. For sites requiring
`additional periodontal therapy during the course of the
`study (those demonstrating 22 mm attachment loss
`at any point), the post-baseline observations recorded
`immediately prior to the locai therapy were analyzed
`using the LOCF strategy to ensure analyses were con-
`servative and not biased.
`Per-subject variables (i.e., change in CAL, change in
`PD, %BOP) were calculated at each time point (months
`3, 6, and 9) by averaging efficacy measurements within
`each subject in accordance with the baseline PD for
`each tooth site. Comparisons between groups were
`made using analysis of variance (ANOVA). Each ANOVA
`mode] analyzed the subject-level mean changes from
`baseline and inciuded factors for treatment group,
`investigational site, the treatment by investigational
`site interaction, and included the actual baseline vaiue
`of the parameter for each subject as a covariate. All
`tests of significance were two-sided; differences were
`
`

`

`j Periodontol
`
`° August 2004
`
`Preshaw, Hefti, Novak, et al.
`
`P values for the main effect of treatment group and
`least-squares estimates for the mean values for each
`treatment group.
`The number of periodontal sites that displayed
`attachment gains of 22 mm and 23 mm or PD reduc-
`tions of 22 mm and 23 mm from baseline were also
`
`determined. Separate analyses were done for all dis-
`eased sites (baseline PD 24 mm) and for deep sites
`(baseline PD 26 mm). The percentages of sites attain—
`ing these thresholds were calculated and compared
`between groups using a GEE model with adjustments
`for within-subject correlations among sites. Compli—
`ance with study medication was defined as the num—
`ber of pills taken as a percentage of the number that
`should have been taken based on the number of days
`between study visits.
`
`RESULTS
`
`A total of 210 subjects were enrolled into the study;
`107 subjects were randomized to the adjunctive SDD
`group and 103 to the placebo group. One subject who
`was enrolled and assigned to the placebo group quit
`the study before all baseline data were gathered. Since
`this subject had incomplete baseline data and no follow-
`up clinical data, the subject was excluded in the anal-
`ysis of efficacy outcomes, but not safety outcomes.
`There were no statistically significant differences
`between the treatment groups for mean age, gender,
`or racial distribution (Table 1). The treatment groups
`
`Table I.
`
`Demographic Characteristics
`of lntent-to-Treat Population
`
`SRP -l Placebo
`
`SR]? + SDD
`
`Characteristic
`
`(N = IUZ)
`
`(N = ID?)
`
`P
`
`also were similar with regard to other baseline charac-
`teristics: medical history, medication history, oral hygiene
`practices, and concomitant medications. One hundred
`fifty-seven subjects completed the study. The most fre-
`quent reason for discontinuation was “lost to follow-up."
`Seven subjects, all in the placebo group, discontinued
`due to adverse events (Table 2). Data from those sub-
`jects who prematurely withdrew were included in the
`final analyses using the LOCF algorithm.
`Baseline clinical parameters are shown in Table 3.
`With the exception of %BOP in the 4 to 6 mm base-
`line PD category, there were no significant differences
`in the baseline clinical characteristics of subjects in
`the two treatment groups.
`There was a statistically significant difference between
`the numbers of current smokers in the SDD group (41
`subjects, 38.3%) and the placebo group (26 subjects,
`25.5%) (P <0.05). This imbalance between the treat-
`ment groups with regards to smoking status occurred
`as a result of random allocation. Randomization was
`
`not stratified on smoking status.
`
`Table 2.
`
`Primary Reasons for Study Discontinuation
`
`Reason
`
`Adverse event
`
`Illness not related to drug
`Protocol deviation
`
`Lost to follow-up
`Treatment failure
`
`Other
`
`SRP + Placebo
`
`SRP + SDD
`
`7
`
`2
`I2
`
`l
`
`|
`0
`
`2
`
`O
`
`I
`5
`
`|
`
`l
`0
`
`|
`
`Age (ream)
`Mean
`
`Range
`Gender
`
`Male
`Female
`
`Ratefelhr Iicity
`VVlIiLe
`Black
`Asian
`l-iispanir.
`Tobacco use
`
`48
`
`'34 75
`
`48
`
`35-75
`
`58 (56.9%)
`44 (43. | %)
`
`67 (62.6%)
`40 (37.4%)
`
`70 (68.6%)
`l9 (l8.6%)
`6 (5.9%)
`7 (6.9%)
`
`76 (71.0%)
`l7 (l5.9%)
`5 (4.7%)
`9 (8.4%)
`
`0.7?’
`
`035T
`
`090+
`
`Ex— or current
`smoker
`
`59 (57.8%)
`
`63 (58.9%)
`
`0.85Jr
`
`0.041
`4| (38.3%)
`26 (25.5%)
`Current smoker
`‘ Determined using analysis of variance with an adjustment for study center.
`
`Table 3.
`
`Baseline Clinical Characteristics
`
`of Intent-to-Treat Population (mean 1- SEM)
`
`Baseline PD/
`Parameter
`
`4-6 mm
`PD
`CAL
`%BOP
`
`27 mm
`PD
`CAL
`
`SRP + Placebo
`
`SRP + SDD
`
`P
`
`4.85 i 0.03
`4.74 i 009
`90.3 i |.4
`
`7.43 : 0.09
`7.|
`| : 0. l9
`
`
`
`4.83 i 0.02
`47| i 0.07
`86.2 i |.2
`
`7.56 i 0.07
`7.42 i O.l5
`
`0.56
`0.78
`0.03
`
`0.25
`0.2,0
`
`

`

`Subantimicrobial Dose Doxycycline Enhances Efficacy of SRP
`
`Volume 75 - Number 8
`
`Efficacy Outcomes
`The mean per-subject changes in CAL lrom baseline
`to month 9 are shown in Figures I and 2. Improve—
`ments in CAL from baseline were demonstrated in both
`groups. as would be expected following SRP. However,
`month 9 mean CAL gains were significantly greater
`with adjunctive SDD than with placebo in both disease
`categories (P <0.05).
`in sites with baseline PD 4 to
`6 mm. improvements from baseline were 35% greater
`with adjunctive SDD than with placebo ( | .2? mm ver»
`sus 0.94 mm, respectively. P 4.0.001).
`In sites with
`baseline PD 27 mm. improvements from baseline were
`
`3|% greater with adjunctive SDD than with placebo
`(2.09 mm versus L60 mm, respectively. P<0.05).
`The mean per-subject changes in PD from base-
`line to month 9 are shown in Figures 3 and 4. Again.
`as would be expected following SRP. Improvements in
`PD from baseline were observed in both groups. How-
`ever, month 9 mean PD reductions were significantly
`greater with adjunctive SDD than with placebo in both
`disease categories (P <0.05). in sites with baseline
`PD 4 to 6 mm.
`improvements from baseline were
`34% greater with adjunctive SDD than with placebo
`([29 mm versus 0.96 mm, respectively, P<0.001).
`
`2.4
`
`*
`
`2.2
`
`196:013
`2 .
`1 59 *0 11
`1.8
`
`*
`209t013
`
`l
`..
`1mm“) 15
`144 10 17
`
`C
`"’
`-—
`s. ”e
`._. E
`5 B,
`E E
`5 3
`H .DN
`E I“
`: E
`E
`
`1.6
`1_4
`1.2
`1
`0.8
`.
`
`
`
`0-4
`0-2 ‘
`o
`
`r_
`Month 3
`0
`' P <0 05 versus placebo
`
`-I-SRP + 300 (N: 107)
`-<>—SRP + placebo (N: 102)
`‘
`Month 6
`
`.
`Mon". 9
`
`1
`1.27 10.05 1.19t005
`
`I03t005
`
`
`-
`1139 ill 05
`o 94 10 05
`0 as it] 06
`
`1-4
`
`E A 1'2
`1
`0.3
`
`0.0
`
`0.4 -
`
`01
`
`0
`
`—l—SRP +SDD (N=»107)
`-(>-SRP +placebo (N: 102)
`
`l'I
`I
`Month 3
`Month 6
`0
`1 P <0 001 versus placebo
`
`_l
`Month 9
`
`
`
`1.4 -
`1.2
`
`1
`
`0.3
`
`0.4
`
`0.2
`
`T
`
`1291005
`
`
`
`
`09111305
`
`1031006
`
`'
`0.562006
`
`
`-I-SRP + SD) (N: 107)
`-<>-SRP + placebo (N = 102)
`
`_h
`I
`
`I
`
`Month 3
`0
`' P <0 05 versus placebo
`’r P <0 01 versus placebo.
`1P <0 001 versus placebo
`
`Month 6
`
`Month 9
`
`Figure l .
`Mean attachment gains ln sites Wlll'l baseline PD 4 to 6 mm. Subjecs
`received SRPat baseline and then either PU mg SDD or placebo b.i.d.
`l'or ‘9 months. The mean pensallleci changes from baseline and
`star-Maid errors are presented
`
`Figure 2.
`Mean attachment guns in srtes with baseline PD 27 mm. Subjects
`recelved S'Rl’al baseline and linen ertlier 20 mg SDD or placebo b.i.d.
`|for 9 months. The moon DEFS'le-fflt'l changes from baseline and
`standard errors are presented
`
`
`
`
`
`
`
`Meanprobingdepthreductions
`
`
`
`frombaseline(mm)
`
`2.4
`
`
`1‘2
`1881010
`2 .
`
`1
`l
`2311012
`2151012
`
`
`
`
`17610131
`
`l
`l/Imwfi
`1521016
`
`1.8
`
`1.6-
`1.4
`1.2 -
`
`
`
`0.8
`0.6 1
`0.4
`0.2
`
`-l—SRP + 500 (N: 107)
`-o-SRP + placebo (N = 102)
`
`w
`I
`
`.
`
`Month 3
`0
`1P <0 01 versus placebo
`
`Month 6
`
`Month 9
`
`Figure 3.
`Mean Pl) re'rluinons in euros mill baseline PD -l [a a rnrl'l. Sabiixts
`received SRP at basellne and then E'lil'lfi'l' 20 mg SUD or placebo blot
`
`Figure 4.
`Mean PD reductions in ones With baseline PD 27 mm. Subjects
`l'Flr‘f’lb’E'Cl SRP at baseline and lllc-n either 20 mg SDD or placebo bid.
`
`

`

`J Periodontol
`
`° August 2004
`
`Preshaw, Hefti, Novak, et al.
`
`In sites with baseline PD 27 mm, Table 4.
`PD reductions from baseline
`
`were 31% greater with adjunc-
`tive SDD than with placebo
`(2.31 mm versus 1.77 mm,
`respectively, P <0.01).
`The percentages of periodon-
`tal sites achieving thresholds of
`clinical improvement (CAL gains
`22 mm and 23 mm, and PD
`reductions 22 mm and 23 mm)
`at 9 months are shown in Table 4.
`
`Whether considering all sites
`(baseline PD 24 mm), or just
`deep sites (baseline PD 26 mm),
`it is clear that in both treatment
`
`groups, these thresholds of clin-
`ical improvement were achieved
`for a considerable number of
`
`Effect of Subantimicrobial Dose Doxycycline on Clinical
`Attachment Gains and Probing Depth Reductions at Month 9
`
`All Sites (baseline PD 24 mm)
`
`Deep Sites (baseline PD 26 mm)
`
`SRP + Placebo
`
`SRP + SDD
`
`SRP + Placebo
`
`SRl’ + SDD
`
`Change From Baseline
`
`(N = 2.964)
`% (N)
`
`(N = 3.465)
`% (N)
`
`(N = 998)
`% (N)
`
`IN _=_ Li 5| ).
`as (N)
`
`CAL gain 22 mm
`
`32.0 (949)
`
`42 31 (I .464)
`
`44.0 (439)
`
`583* (67 I)
`
`CAL gain 23 mm
`PD reduction 22 mm
`
`I06 (3 I5)
`
`I5 4* (532)
`
`20.0 (200)
`
`33.2’r (382)
`
`3|.l (923)
`
`42.9‘r (I.487)
`
`45.0 (449)
`
`62.3’r (7I7)
`
`PD reduction 23 mm
`
`36.71 (422)
`20.6 (206)
`I541 (535)
`9. I (27I)
`" P<0.05 and T P<0.01 compared to placebo (determined using GEE model with adjustment for within—subject
`correlations among sites).
`
`sites. However, in the adjunctive SDD group, the per-
`centage of sites achieving these thresholds of improve-
`ment was consistently significantly greater than
`observed in the placebo group (P <0.05). The bene-
`fits of SDD therapy were particularly evident when
`considering deep sites, with 62.3% of sites demon-
`strating PD reductions 22 mm (compared to 45.0% in
`the placebo group) and 36.7% of sites demonstrating
`PD reductions 23 mm (compared to 20.6% in the
`placebo group) (P <0.01).
`Compliance with study medication was high (>87%
`at all time points), with no statistically significant dif-
`ferences between treatment arms (P >0.05).
`
`Safety Outcomes
`The percentage of subjects who experienced treatment-
`emergent adverse events (i.e., all causes) was similar for
`subjects treated with adjunctive SDD (59%) and placebo
`(61%) (Table 5). The most frequent events were infec-
`tion, headache, pain, and influenza. Adjunctive SDD
`was well tolerated, and there were no clinically mean-
`ingful differences in the numbers of treatment-emer-
`gent adverse events, including those events associated
`with the gastrointestinal (GI) tract and the genito—uri-
`nary tract, between the treatment groups. Given the
`known pharmacology of doxycycline, certain categories
`of adverse events are of particular interest, including
`adverse events related to the GI tract, skin (particularly
`rash, photosensitivity reactions, sunburn), hypersensi-
`tivity reactions, and infection (particularly enterocolitis,
`fever, and lymphadenopathies). There were no statis—
`tically significant differences in the frequency of adverse
`events in these categories between the SDD and
`placebo groups (P>0.05). No clinically meaningful dif-
`ferences among the placebo and SDD groups were
`
`Table 5.
`
`Adverse Events Reported by 25%
`of Patients
`
`Adverse Event
`
`Infection
`
`Headache
`
`Pain
`
`Influenza
`
`Back pain
`
`Abdominal pain
`
`Diarrhea
`
`Allergic reaction
`
`SRP + Placebo
`
`(N = l03)
`
`|8 (|8%)
`
`I5 (l5%)
`
`15 (I5%)
`
`IO (l0%)
`
`6 (6%)
`
`6 (6%)
`
`6 (6%)
`
`6 (6%)
`
`SRP + SDD
`
`(N = l07)
`
`I5 (|4%)
`
`|3 (|2%)
`
`9 (8%)
`
`6 (6%)
`
`5 (5%)
`
`4 (4%)
`
`4 (4%)
`
`3 (3%)
`
`Serious adverse events were reported for three sub-
`jects. The two serious adverse events in the placebo
`group were myocardial infarction and severe asthma.
`Both patients were hospitalized and both discontinued
`the study. The serious adverse event in the SDD group
`was chest tightening that resulted in hospitalization. This
`patient continued taking study medication throughout
`and completed the study successfully. None of these
`serious adverse events were considered related to study
`products.
`
`DISCUSSJON
`
`SRP is the standard of care in periodontics, and is the
`most common treatment in the management of CP. The
`
`

`

`Subantimicrobial Dose Doxycyciine Enhances Efficacy of SRP
`
`Volume 75 - Number 8
`
`instrumentation techniques are employed.16 Probing
`depth reductions of ~] .29 mm and clinical attachment
`gains of ~0.55 mm can be expected in moderately
`deep pockets (4 to 6 mm) following SRP over a simi-
`lar time period to that employed in this study. For deep
`sites (PD 27 mm), PD reductions of ~2.16 mm and
`attachment gains of ~1.19 mm can be expected”;-17
`In our study, using the same baseline probing depth cri-
`teria to categorize the data, very similar treatment
`responses were obtained following SRP alone (placebo
`group).
`In particular,
`the mean attachment gains
`reported in the placebo group were better than those
`reported in many studies of the effectiveness of SRP.16
`This favorable treatment response in the placebo group
`is attributed to the high quality of root surface instru-
`mentation that was undertaken as part of the study.
`Attachment gains reported in the placebo group were
`equivalent to or slightly better than those reported in
`many previous studies,16 whereas, in contrast, probing
`depth reductions were slightly less. We attribute this to
`the fact that a majority of subjects enrolled into our
`study were “long-term” periodontitis patients, having
`histories of previous periodontal treatment with limited
`success in the past. Anecdotally, in these patients, the
`opportunities for clinical
`improvements occurring
`because of further gingival shrinkage tend to be reduced,
`and gains in attachment may constitute the majority of
`any healing response.
`Improvements in CAL and PD were greater in the
`adjunctive SDD group than those seen in the placebo
`group at every follow-up examination, and achieved
`statistical significance at month 9 (P <0.05). The
`absolute magnitude of the improvements in PD and
`CAL was greater in sites with baseline PD 27 mm, com-
`pared to PD 4 to 6 mm. This is consistent with previ-
`ous observations that the magnitude of PD reduction
`and gains in attachment levels that can be expected
`following SRP are related to the initial probing depth
`measurement.16
`
`Mean changes in CAL and PD are useful summary
`statistics of changes occurring within the study popu-
`lation. While the differences between mean CAL and
`
`PD changes observed in the placebo and SDD groups
`may achieve statistical significance, these differences
`nevertheless can appear small. It is important to remem-
`ber that population means do not equate to what is
`directly measurable in an individual, or what may be
`occurring at any one periodontal site.18 Attention has
`recently focused on the issue of clinical significance as
`compared to statistical significance.19 A clinically sig-
`nificant improvement can be defined as a change that
`can easily be identified in an individual patient using
`routine diagnostic techniques at the chairside (for exam-
`ple, a detectable reduction of probing depth).20 Given
`
`ing depth that can reliably be detected in an individual
`periodontal site is of the order of approximately two
`standard deviations of repeated measures.22 A conser-
`vative assessment of categorical threshold data therefore
`uses CAL and PD changes of 22 mm to indicate clinically
`significant changes over time.
`In our study, adjunctive SDD resulted in a statistically
`greater proportion of periodontitis sites undergoing
`22 mm gains in CAL and 22 mm reductions in PD com-
`pared to placebo (P <0.05) (Table 4). When using a
`threshold of 23 mm changes in CAL and PD to represent
`clinical significance, adjunctive SDD still resulted in a
`statistically greater proportion of periodontitis sites
`achieving these thresholds compared to placebo
`(P <0.05). The benefits of SDD therapy (in terms of
`percentages of sites achieving such thresholds of
`change) were particularly apparent in deep periodon-
`tal sites (baseline PD 26 mm). For example, nearly
`twice as many deep sites in the SDD group (36.7%)
`demonstrated PD reductions 23 mm compared to the
`placebo group (20.6%). Furthermore, nearly two-thirds
`of the deep sites in the SDD group (62.3%) demon-
`strated PD reductions 22 mm. We feel that these ben-
`efits reveal an enhanced treatment response when using
`SDD compared to placebo, and represent a clear and
`clinically significant benefit to patients.
`Past and current smoking is known to have an effect
`on periodontal health and the outcome of therapeutic
`intervention.23 There were more current smokers in the
`SDD group (38.3%) compared with the placebo group
`(25.5%) (P <0.05), yet a better treatment response was
`observed in the SDD group compared to placebo sup-
`porting the notion that the benefits of adjunctive SDD
`are not limited to non-smokers. The proportion of sub-
`jects who reported either past or current use of tobacco
`was almost identical in each treatment group even
`though this factor was not used to stratify the randomiza-
`tion. Hence, the effect of tobacco use does not bias the
`general conclusions of this study. An adjustment to the
`analysis for smoking status was not performed because
`of the difficulty in interpreting least-square mean results
`adjusted both for continuous and bivariate factors.
`Further analysis of the magnitude of benefit of SDD in
`smokers and non-smokers is being conducted in this
`and similar studies and will be reported separately.
`Our results are consistent with previous studies
`which indicate that SDD is well tolerated with minimal
`unwanted effects.12'24 The percent of subjects who expe-
`rienced adverse events was similar for both treatment
`groups. Adjunctive SDD was well tolerated and there
`were no clinically meaningful differences in the numbers
`of treatment-emergent adverse events between treatment
`groups. The number of subjects who discontinued early
`(for whatever reason) was disappointing, particularly
`
`

`

`J Periodontol
`
`- August 2004
`
`Preshaw, Hefti, Novak, et al.
`
`questions the interpretation of the results as favoring
`SDD, as an imbalance of early discontinuations in the
`opposite direction would be required to bias the study
`in favor of SDD. The uneven discontinuation rate is
`
`either due to chance alone, or the benefits of SDD result-
`ing in greater compliance with the study protocol, includ-
`ing returning for assessments as per the study schedule.
`The use of adjunctive SDD was recently reported in
`a 9-month, randomized, placebo-controlled clinical trial
`of severe, untreated, generalized periodontitis (defined
`by CAL 25 mm at >30% of sites in patients S45 years
`old).25 All subjects received full supra- and subgingival
`debridement over four l-hour sessions, followed by a
`periodontal maintenance recall program. Clinically and
`statistically significant improvements in probing depths
`were reported in the adjunctive SDD group compared
`to the placebo group. In the SDD group, pockets 27 mm
`at baseline were reduced by an average of 3 mm at
`month 9. These authors suggested that the combination
`of a systemic host modulator (i.e., SDD) together with
`topical antimicrobial agents (e.g., chlorhexidine, or local
`delivery antibiotic systems) might confer yet further
`clinical benefits to patients with periodontitis.25 In this
`study, the reductions in probing depths that occurred
`were mainly due to recession, and not. gains in CAL.
`This contrasts with the present study in which signifi-
`cant gains in CAL were observed following treatment,
`and which constituted the majority of the improvements.
`This difference betwe