J Periodontol • January 2010

Microbiologic Testing and Outcomes

of Full-Mouth Scaling and Root
Planing With or Without Amoxicillin/
Metronidazole in Chronic Periodontitis
Norbert Cionca,* Catherine Giannopoulou,* Giovanni Ugolotti,* and Andrea Mombelli*

Background: It has been suggested that use of systemic antibiotics should be limited to patients with spe-
cific microbiologic profiles. The main purpose of the present analysis was to study whether microbiologic
testing before therapy was of value in predicting which patients would specifically benefit from adjunctive
amoxicillin and metronidazole, given in the context of full-mouth scaling and root planing (SRP) within
48 hours.
Methods: This was a 6-month, single-center, double-masked, placebo-controlled, randomized longitu-
dinal study. Fifty-one patients received full-mouth periodontal debridement, performed within 48 hours;
then, 25 subjects received metronidazole, 500 mg, and amoxicillin, 375 mg, three times a day for
7 days, and 26 received a placebo (control group). Pooled microbiologic samples were taken from the deep-
est pocket at baseline in each quadrant before and 6 months after treatment. Six periodontal pathogens
were quantified by real-time polymerase chain reaction.
Results: Forty-seven patients were followed for 6 months. After treatment, test subjects had a substan-
tially lower mean number of persisting sites with probing depth >4 mm and bleeding on probing than did
control subjects (0.4 versus 3.0; P = 0.005; month 6). Aggregatibacter actinomycetemcomitans (previously
Actinobacillus actinomycetemcomitans) could not be detected in the antibiotic group after treatment. How-
ever, in the placebo group, three of six subjects positive for A. actinomycetemcomitans continued to be pos-
itive. Lower frequencies were also noted in the test group for Porphyromonas gingivalis (P = 0.013) and
Tannerella forsythia (previously T. forsythensis) (P = 0.007). However, even subjects testing negative for
A. actinomycetemcomitans at baseline had a significantly better primary clinical outcome if they received
the active drugs. The presence of six putative periodontal pathogens (A. actinomycetemcomitans, Fusobac-
terium nucleatum spp., P. gingivalis, Prevotella intermedia, Treponema denticola, and T. forsythia) quan-
tified prior to therapy was not correlated with the outcome of full-mouth SRP with or without amoxicillin
and metronidazole.
Conclusion: Excellent clinical results in the antibiotics group were obtained regardless of the presence or
absence of six classic periodontal periopathogens prior to treatment. J Periodontol 2010;81:15-23.
KEY WORDS
Actinobacillus actinomycetemcomitans; Aggregatibacter actinomycetemcomitans; amoxicillin;
chronic periodontitis; metronidazole; Porphyromonas gingivalis; root planing.

* Department of Periodontology, School of Dental Medicine, University of Geneva, Geneva, Switzerland.

doi: 10.1902/jop.2009.090390

15
Microbiologic Testing and Antibiotics Volume 81 • Number 1

A
beneficial effect of adjunctive systemic antibi- Analysis of the microbiologic data was carried out
otics on the clinical outcomes of non-surgical to assess whether the presence of a specific microbi-
periodontal treatment has been demon- ologic profile at baseline correlated with the outcome
strated.1,2 Nevertheless, questions remain with re- of full-mouth scaling and root planing (SRP) with or
gard to the optimal use of antibiotics and their specific without adjunctive A/M.
benefit in various clinical situations. To limit the
development of microbial antibiotic resistance in MATERIALS AND METHODS
general, and to avoid the risk for unwanted systemic This was a 6-month, single-center, randomized, pla-
effects in treated patients, a precautionary, restrictive cebo-controlled, parallel-design, and double-masked
attitude toward using antibiotics seems reasonable. trial. The protocol was approved by the Ethical Com-
Therefore, it was suggested to limit their use to patients mittee of the University Hospitals of Geneva, Geneva,
with advanced or aggressive disease or to subjects Switzerland and was authorized by the Swiss Agency
with specific microbiologic profiles.2 for Therapeutic Products (Swissmedic), Bern, Swit-
Quantitative assays for several putative periopath- zerland. Research was conducted according to the
ogens are available. These assays are sometimes re- principles outlined in the Declaration of Helsinki on
lied upon to guide the choice of therapeutic regimen, experimentation involving human subjects. Written
such as the use of adjunctive local or systemic antibi- informed consent was obtained from all subjects.
otics.3 A systematic review4 of microbial identification
Clinical Protocol
in the management of periodontal diseases included
The clinical protocol was presented in detail in a pre-
11 studies reporting a differential clinical response
ceding article.9 In brief, 51 systemically healthy pa-
depending on the detection or lack of specific organ-
tients, aged 25 to 70 years, with untreated moderate
isms. These studies generally attempted to demon-
to advanced periodontitis and ‡4 teeth with a probing
strate a correlation between the clinical outcome of
depth (PD) >4 mm, clinical attachment loss (AL) ‡2
a given therapy and the presence or absence of spe-
mm, and radiographic evidence of bone loss were re-
cific microbiologic profiles. However, surprisingly
cruited into the study between November 2006 and
little is known about the value of specific microbio-
December 2007. Exclusion criteria included preg-
logic information for selecting the optimal treatment.
nancy or lactation, systemic antibiotics taken within
Although the above-mentioned systematic review4
the previous 2 months, use of non-steroidal anti-
lists 13 articles reporting the use of microbial iden-
inflammatory drugs, confirmed or suspected intoler-
tification as an aid in treatment planning, the clinical
ance to 5-nitroimidazole-derivatives or amoxicillin,
advantage of any therapy chosen based on micro-
and subgingival SRP or surgical periodontal therapy
biologic testing, over adjunctive amoxicillin plus
in the previous year.
metronidazole (A/M) administration, remains to be
The clinical procedures were performed by two of
demonstrated. No trial has substantiated the exis-
the authors (NC and GU). The examiner (NC) re-
tence of microbiologically distinct patient populations
corded the medical history, obtained informed con-
that present better results if treated without these
sent, removed supragingival deposits, and gave oral
antibiotics. Although some reports5-7 suggested that
hygiene instructions. The subjects were recalled to
systemic antibiotics might only be effective in patients
assure that proper oral hygiene was established. Once
with a specific microbial profile, at least one study8 re-
the subjects had reached an appropriate level of
ported considerable clinical benefits of A/M, even in
plaque control, they were scheduled to receive sub-
subjects who tested negative for periodontal marker
gingival treatment within a maximum of 1 month.
organisms.
On the day of treatment, the examiner first re-
We recently reported the clinical results of a study
corded the periodontal parameters and took a pooled
on this patient population.9 Twenty-five test subjects
subgingival plaque sample by inserting a sterile paper
received amoxicillin, 375 mg, and metronidazole,
point to the bottom of the deepest pocket in each
500 mg, three times a day for 7 days, and 26 control
quadrant. Next, the operator (GU) treated at least half
subjects received a placebo.9 According to general
of the periodontally diseased teeth with thorough SRP
standards, good results were obtained in the control
to the depth of the pocket under local anesthesia. The
group. Patients in the test group benefited from an
operator first used ultrasonic instruments,† then
additional, statistically significant and clinically rele-
Gracey curets, and finally irrigated the pockets with
vant effect of the antibiotics and needed much less
a 0.1% aqueous solution of chlorhexidine. Subjects
additional treatment after 6 months. Microbiologic
were instructed to rinse the mouth twice daily during
testing was carried out in all subjects, but the patients
the next 10 days with 0.2% chlorhexidine. The remain-
for this trial were not selected based on microbiologic
der of the dentition was treated by the operator the
criteria, and treatment allocation was independent of
microbiology. † EMS, Nyon, Switzerland.

16
J Periodontol • January 2010 Cionca, Giannopoulou, Ugolotti, Mombelli

same way within 48 hours. Upon completion, each primer and probe sequences were selected with
subject received a package containing the test or software,i which checks the primer and probe sets
placebo medication; all packages were identical in for matching the guidelines that are recommended
appearance and were marked only with the subject for real-time PCR probes. The specificity was verified
number. Subjects in the test group received metroni- with purified genomic DNA from several bacterial and
dazole, 500 mg, and amoxicillin, 375 mg, to be taken fungal species as well as with human DNA. Even
three times per day for 7 days; subjects in the control closely related species, such as P. intermedia and
group received similarly looking placebos. The cap- P. nigrescens, did not show any cross-reactivity. The
sules were prepared by the pharmacy at University analysis is highly automated, has a detection limit of
Hospital Geneva. The treatment group was concealed 100 bacterial cells, and a linear range for quantifica-
to the patient, the clinical and laboratory staff, and the tion of at least seven orders of magnitude for each pa-
statistician (AM). rameter. The coefficient of variation for quantitative
The examiner recalled the subjects after 1 week and results is 15%.
after 3 and 6 months. Medical history, any concomitant
Statistical Analyses
medication, and any adverse events were recorded
Data were entered into a database and were checked
for each patient at each visit. The first post-treatment
for entry errors by two persons (NC and auxilliary
visit also served as a compliance control, as subjects
staff member). Average scores were calculated for
were asked to return any remaining medication. At
clinical parameters for each subject at each examina-
months 3 and 6, the examiner checked the oral hy-
tion by summing the scores and dividing by the num-
giene; reinforced it, if necessary; and removed supra-
ber of sites graded for that subject. The absolute
gingival calculus, if detected. A pooled subgingival
number of sites per subject with PD >4 mm and BOP
plaque sample was obtained at the same sites as
was the primary clinical outcome measure of the
before treatment. Residual periodontal pockets were
study (persisting pockets >4 mm with BOP are
not reinstrumented.
commonly perceived as needing further treatment).
Immediately before and 3 and 6 months after ther-
Secondary clinical outcomes included differences
apy, the following clinical parameters were recorded
between groups for changes in PD and clinical AL
at six sites of each tooth with a pocket >4 mm at
(clinical AL = PD+REC) at different initial PD ranges,
baseline: gingival index,10 PD, bleeding on probing
as well as the types and frequencies of adverse events.
(BOP), and recession (REC; positive if gingival mar-
For use in pocket-incidence analysis, the microbi-
gin located apical to the cemento-enamel junction,
ologic data for the six target organisms were dichoto-
negative if located coronal to the cemento-enamel
mized and expressed as detected or not detected.
junction). In addition, the presence or absence of
Quantitative aspects were inspected on scatter plots
plaque was recorded at six sites of all teeth by running
with logarithmic scale (negative samples were repre-
a probe across the site (plaque score).
sented as a count of 1), and mean counts were calcu-
lated. Differences among patients in the treatment
Microbiologic Procedures
groups were determined at each time point using
One sterile paper point was inserted to the bottom of
the Mann-Whitney U test. Backward stepwise logistic
the deepest pocket in each quadrant and left in place
regression analysis was used to determine the corre-
for 10 seconds. To obtain a pooled sample, the four
lation between the continued presence of PD >4 mm
paper points of the same subject were collected in
with BOP in at least one site (dependent variable)
one vial. The samples were numbered consecutively
at 6 months and the number of diseased sites at
and sent to a specialized laboratory‡ for the detection
baseline, the presence or absence of A. actinomyce-
and quantification of six periodontal marker organisms
temcomitans and P. gingivalis at baseline, smoking,
(Aggregatibacter actinomycetemcomitans [previously
and gender (predictor variables) separately for both
Actinobacillus actinomycetemcomitans], Fusobacte-
test and control groups.
rium nucleatum spp., Porphyromonas gingivalis, Pre-
One statistical program package was used for all
votella intermedia, Treponema denticola, and Tannerella
statistical analyses.¶ P values <0.05 were accepted
forsythia [previously T. forsythensis]) and total bacterial
for statistical significance.
load.
A commercially available real-time polymerase RESULTS
chain reaction (PCR)-based method,§ developed
The clinical results of this trial were presented previ-
and validated by the laboratory processing the sam-
ously.9 All 51 enrolled patients were followed up to
ples, was used for analysis. The details of the method
were described elsewhere.11-13 Primers and probes ‡ Carpegen, Münster, Germany.
§ Meridol, Perio Diagnostics, GABA International, Therwil, Switzerland.
were designed to match highly specifically to the ribo- i Primer Express, Applied Biosystems, Foster City, CA.
somal DNA of the six bacterial pathogens. The exact ¶ SPSS 16 for Mac OS X, SPSS, Chicago, IL.

17
Microbiologic Testing and Antibiotics Volume 81 • Number 1

month 3; 47 were followed up to month 6. The 47 par-
ticipants completing the study had a mean age of 50.5
years, 59% were females, 34% were smokers, and
92% were white. Differences between the groups were
not significant.
A total of 3,126 sites (six sites each on 521 peri-
odontally diseased teeth with a pocket >4 mm at base-
line) were clinically monitored at baseline and at 3
and 6 months by a masked, independent investiga-
tor (NC). Subjects in the placebo group had a mean
PD of 4.4 – 0.4 mm, 62.5% – 13.3% of all sites had
BOP, and 29.9 – 17.1 sites had PD >4 mm and BOP.
The subjects in the test group had a mean PD of
4.3 – 0.4 mm, 64.2% – 10.7% of all sites had BOP,
and 26.9 – 15.6 sites had PD >4 mm and BOP. Differ-
ences between the groups were not significant.
As reported previously,9 a significantly better pri-
mary clinical outcome was obtained after 3 and 6
months in subjects treated with full-mouth SRP plus
A/M compared to those receiving placebo. In the
placebo group, a mean of 4.4 and 3.0 pockets >4 mm
with BOP were noted after 3 and 6 months, respec-
tively, whereas 1.3 sites were detected in the test
group after 3 months (P = 0.02), and only 0.4 pockets
were present after 6 months (P = 0.005).
Differences in the frequencies of the six monitored
microbiologic parameters in pooled samples from test
and control patients were not significant at baseline
(Table 1). Three subjects in the test group were pos-
itive for A. actinomycetemcomitans before treatment.
The organism could not be detected after treatment
with antibiotics. However, in the placebo group, of
six subjects positive for A. actinomycetemcomitans,
three continued to be positive 6 months later. At
month 3, the frequencies of detection were signifi-
cantly lower in samples from subjects in the test group
for A. actinomycetemcomitans (P = 0.043), P. gingiva-
lis (P = 0.013), and T. forsythia (P = 0.007). At month
6, differences in the frequencies of detection did not
reach a level of significance.
The primary outcomes of therapy were determined
specifically in subjects positive or negative for A.
actinomycetemcomitans and P. gingivalis. Table 2
shows the differential primary outcome at month 6
with respect to whether subjects tested positive or
negative for A. actinomycetemcomitans at baseline
and whether they received the antibiotics or placebo.
A majority of 38 subjects, 18 in the placebo group and
20 in the test group, were negative for A. actinomyce-
temcomitans at baseline. None of these 38 persons
tested positive for A. actinomycetemcomitans at
month 3 or 6. Thus, although these subjects were neg-
ative for A. actinomycetemcomitans throughout the
study, if they received A/M, they had a significantly
better primary clinical outcome than if they were
treated with placebo. At month 6, the subjects nega-
tive for A. actinomycetemcomitans who were treated
with antibiotics had significantly smaller mean PDs
(3.0 – 0.2 mm) and fewer mean residual pockets >4
mm with BOP (0.5 sites) than subjects in the placebo
group (3.2 – 0.3 mm, P = 0.02; 3.9 sites, P = 0.002).
Table 3 shows the differential primary outcome at
month 6 with respect to whether subjects received
the antibiotics or placebo and whether they tested
positive or negative for P. gingivalis at baseline. Pa-
tients treated with antibiotics tended to show better
results with regard to the persistence of sites with
PD >4 mm and BOP, irrespective of whether they

Table 1.
Frequency of Six Microorganisms and Their Mean Counts in Positive Samples Before and
After Treatment (n = 47 subjects)

Placebo Group Test Group

Baseline Month 3 Month 6 Baseline Month 3 Month 6

MO Freq C Freq C Freq C Freq C Freq C Freq C

Aa 25 310 13* 34 13 25 13 258 0* 0 0 0

Fn 96 1,013 96 442 88 386 100 610 87 81 83 170

Pg 83 4,598 54* 2,159 67 2,598 78 3,817 22* 413 39 258
Pi 67 605 42 394 50 1,056 78 1,036 39 517 30 921

Td 83 2,862 75 1,419 67 1,767 96 2,558 57 420 61 333

Tf 96 2,311 83* 965 75 1,362 96 3,151 43* 185 52 239
MO = microorganism; Freq = frequency (%); C = counts; Aa = A. actinomycetemcomitans; Fn = F. nucleatum spp.; Pg = P. gingivalis; Pi = P. intermedia; Td = T.
denticola; Tf = T. forsythia.
* Statistically significant difference between the test and control groups (P <0.05).

18
J Periodontol • January 2010 Cionca, Giannopoulou, Ugolotti, Mombelli

Table 2. There was not a signifi-

cant impact of microbial
Detection of A. actinomycetemcomitans at Baseline and
parameters at baseline in
After 6 months and Mean Number of Pockets >4 mm With the test or the control
BOP Per Subject group, i.e., the presence
or absence of A. actinomy-
Placebo Group (n) Test Group (n) cetemcomitans and P. gin-
givalis before treatment.
Sites With PD Sites With PD
Figure 1 shows the
>4 mm and >4 mm and
total bacterial loads be-
Subjects Time Point Subjects BOP Subjects BOP
fore and 6 months after
All subjects Baseline 24 29.9 23 26.9 therapy. All subjects with
M6 24 3.0 23 0.4 lower counts after treat-
Subjects with Aa at baseline Baseline 6 25.0 3 21.0 ment are below the di-
M6 Aa -
3 0.7 3 0 agonal line. Although,
M6 Aa +
3 0 0 – visually, more red than
blue dots are located be-
Subjects with no Aa at baseline Baseline 18 31.6 20 27.8 low the diagonal line, we
M6 Aa- 18 3.9 20 0.5
found no significant dif-
M6 Aa+ 0 – 0 –
ference between test and
M6 = month 6; Aa = A. actinomycetemcomitans; - = negative; + = positive; – = no data. control subjects for total
bacterial load.
Figure 2 shows a scat-
Table 3. ter plot for the quantitative
Detection of P. gingivalis at Baseline and After 6 Months and data for A. actinomyce-
Mean Number of Pockets >4 mm With BOP Per Subject temcomitans, P. gingivalis,
F. nucleatum spp., and
T. forsythia. For A. actino-
Placebo Group (n) Test Group (n)
mycetemcomitans, most
Sites With PD Sites With PD dots are found on the hor-
>4 mm and >4 mm and izontal base, indicating no
Subjects Time Point Subjects BOP Subjects BOP detection after therapy.
The small number of posi-
All subjects Baseline 24 29.9 23 26.9
tive samples precluded
M6 24 3.0 23 0.4
a formal statistical testing
Subjects with Pg at Baseline 20 31.7 18 29.0 of the quantitative aspect.
baseline M6 Pg- 4 1.0 9 0.2 P. gingivalis, F. nucleatum
M6 Pg+ 16 3.9 9 0.6 spp., and T. forsythia at
Subjects with no Pg at Baseline 4 21.0 5 18.8 month 6 were significantly
baseline M6 Pg- 4 1.5 5 0.6 lower in test samples
M6 Pg+ 0 – 0 – than in control samples.
M6 = month 6; Pg = P. gingivalis; - = negative; + = positive; – = no data. Quantitative differences
between the test and con-
trol groups were not signif-
icant for P. intermedia and
tested positive or negative before treatment. Upon fur- T. denticola (data not shown).
ther inspection, it was determined that all nine subjects
initially negative for P. gingivalis remained negative DISCUSSION
after treatment and they were also negative for A. acti- The primary aim of the present analysis is to deter-
nomycetemcomitans before and after treatment. The mine if microbiologic testing before therapy identified
individual results for these subjects, who according to subjects with chronic periodontitis who would benefit
general consensus should not be treated with anti- from adjunctive A/M, when given in the context of full-
biotics, are shown in Table 4. Because of the small mouth SRP within 48 hours. In the evaluation of the
sample size, a statistical analysis of the differences usefulness of any diagnostic test, the prevalence of
between those treated with antibiotics and placebo the condition (in the present study, the presence or
was not possible. absence of target microorganisms above detection

19
Microbiologic Testing and Antibiotics Volume 81 • Number 1

Table 4.
Individual Outcomes for the Nine Subjects Negative for A. actinomycetemcomitans
and P. gingivalis

Sites With PD >4 mm Sites With PD >4 mm Adverse Events Reported

Antibiotics Gender Smoking and BOP at Baseline (n) and BOP at 6 Months (n) by Subject
No Female No 10 0 None
No Male No 28 0 None

No Male Yes 11 1 Diarrhea

No Female Yes 35 5 Nausea
Yes Female Yes 15 0 Nausea, diarrhea
Yes Female No 21 0 None

Yes Female Yes 10 1 Stomach burning

Yes Female No 24 1 None
Yes Female No 24 1 None

high hazard or benefit of particular

interventions in subjects who are truly
negative for F. nucleatum spp., T.
forsythia, or T. denticola, testing for
these three organisms is not justified.
The presence of these organisms
should simply be associated with the
diagnosis of chronic periodontitis, as
demonstrated previously.14
In the placebo group, the detection
frequency of A. actinomycetemcomi-
tans and P. gingivalis was 25% and
83%, respectively. These percentages
are close to those calculated in a sys-
tematic review,15 in which 28% of
subjects with a clinical diagnosis of
chronic periodontitis were positive for
A. actinomycetemcomitans, and 71%
were positive for P. gingivalis. Five of
23 subjects treated with antibiotics in
the present study were negative for P.
Figure 1.
Total bacterial loads before and 6 months after therapy. Each dot represents one subject. gingivalis at baseline. These subjects
Arrowheads indicate mean values. ‘‘Before’’ represents baseline. E = exponent; P = placebo; were also negative for A. actinomyce-
A = antibiotics; D NS = difference not significant. temcomitans. According to current con-
sensus, if a microbial test had been
performed before SRP, these subjects
levels) and the consequences of a positive or nega- would not have been treated with antibiotics. In a com-
tive test should be put into perspective. mentary on our first article,16 in which we presented
As can be seen in Table 1, in the present study (a the general clinical findings of this trial,9 it was stated
pooled subgingival plaque sample was obtained from that treatment of patients negative for P. gingivalis
the deepest pocket in each quadrant), the frequency with antibiotics should be considered overtreatment.
of detection of three of the six tested microorganisms This statement was based on results from a study6
(F. nucleatum spp., T. forsythia, and T. denticola) was by the same investigators involving 49 subjects who
high. In the absence of evidence for an especially were treated with full-mouth SRP plus A/M or placebo.

20
J Periodontol • January 2010
Figure 2.
Scatter plots for the quantitative assessment of A. actinomycetemcomitans, P. gingivalis, F. nucleatum spp., and T. forsythia. Arrowheads indicate mean
values. All baseline differences not significant. NT = not tested.
In their trial, approximately half of the subjects in each
group were culture positive for P. gingivalis. These pa-
tients showed significantly better results when treated
with antibiotics than when treated with placebo. How-
ever, the percentage of pockets >4 mm decreased
very similarly in subjects negative for P. gingivalis
who were treated with antibiotics (from 40% to 12%
in subjects negative for P. gingivalis and from 46%
to 11% in subjects positive for P. gingivalis). Another
study17 contradicts their findings; the results were
interpreted to show that A/M adversely affected the
clinical outcome of patients harboring P. gingivalis
but not those harboring A. actinomycetemcomitans.
Because of this controversy, the data for each of the
nine patients negative for P. gingivalis in our trial are
listed in Table 4.
Despite the potential to consider treatment of sub-
jects negative for P. gingivalis with antibiotics over-
treatment, such treatment was clinically successful
in all five subjects concerned, whereas in the four sub-
Cionca, Giannopoulou, Ugolotti, Mombelli
jects negative for P. gingivalis who were not treated
with antibiotics, one was left 6 months later with five
pockets needing further therapy. Therefore, the fre-
quency and potentially negative effects of the anti-
biotics should be contrasted against the potential
deleterious health consequences of not rapidly sup-
pressing a periodontal infection and the inconve-
nience, discomfort, and financial consequences of
further therapy.
Because of A/M’s proven ability to suppress peri-
odontal pathogens, such as A. actinomycetemcomi-
tans, from periodontitis lesions and other oral
sites,7,18-25 it is the first choice of many clinicians, es-
pecially for the treatment of advanced A. actinomyce-
temcomitans-associated periodontitis. Effectively, all
subjects positive for A. actinomycetemcomitans in
the antibiotic group of our trial became negative for
A. actinomycetemcomitans, whereas three subjects
in the placebo group remained positive. The question
of whether provision of A/M could be considered
21
Microbiologic Testing and Antibiotics
overtreatment for those who were negative for A. ac-
tinomycetemcomitans was addressed by repeating
the analysis presented in our earlier article,9 this time
excluding the nine subjects who were positive for A.
actinomycetemcomitans at baseline. Rechecking
the clinical differences between patients in the anti-
biotics and placebo groups revealed the same sig-
nificant differences as those reported previously:
systemic A/M significantly improved the 3- and
6-month clinical outcomes of full-mouth non-surgical
periodontal debridement. These results are in agree-
ment with another placebo-controlled study,8 in
which patients treated with A/M showed significantly
better clinical outcomes after non-surgical therapy
than those treated with placebo, despite the fact that
only 27% of the subjects were positive for P. gingivalis,
and only 11% were positive for A. actinomycetemco-
mitans. Although A/M proved efficient in suppressing
A. actinomycetemcomitans below detection limits, it
may be concluded that testing for A. actinomycetem-
comitans among subjects with a clinical diagnosis of
chronic periodontitis was of no predictive value in
identifying those subjects who would benefit from
antibiotics.
Based on current knowledge of the genetic and
phenotypic diversity of A. actinomycetemcomitans
and its distribution in various populations and co-
horts26 and a large prospective study,27 one distinct
subpopulation of A. actinomycetemcomitans, known
as the JP2 clone,28 displays the properties of a true
pathogen in at least one group of humans of North
and West African descent. Therefore, one may raise
the question of whether it would be better to test
patients specifically for JP2 in the future. However,
because no interventional study has demonstrated
a specific advantage of treatment other than A/M in
cases positive for JP2, the usefulness of such a test
remains to be determined. The same holds true in
principle for any other specific test of particular pro-
perties of members of the subgingival flora. In addition,
given the large diversity of the microbiota associated
with all forms of periodontitis and the complex in-
teraction among the members of the flora, the con-
cept of ‘‘surgical excision’’ of only the bad bug, guided
by sophisticated diagnosis, may be illusionary. It is
possible that A/M has an effect beyond the factors
studied using bacterial culture and currently avail-
able tests, maybe even a non-antibacterial one.
CONCLUSIONS
In patients with a clinical diagnosis of chronic peri-
odontitis, A/M significantly enhanced the already ex-
cellent effects of full-mouth SRP, thereby reducing the
need for further therapy, which, according to current
treatment concepts, would frequently be surgical in
nature. In these patients, testing for A. actinomyce-
22
Volume 81 • Number 1
temcomitans, F. nucleatum spp., P. gingivalis, P. in-
termedia, T. denticola, and T. forsythia was of no
predictive value in identifying subjects who would
benefit from A/M.
The present study does not provide evidence for
the indiscriminate use of just any antibiotic in any peri-
odontal patient. It is reasonable to prescribe antibi-
otics restrictively, given the widespread development
of antibiotic resistance as well as other potentially
deleterious factors. Furthermore, it is reiterated that
in the present study, antibiotics were not tested as
an alternative to thorough debridement and proper
oral hygiene.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the support of
GABA International, Therwil, Switzerland, which pro-
vided the microbiologic analyses free of charge. The
authors report no conflicts of interest related to this
study.
REFERENCES
1. Haffajee AD, Socransky SS, Gunsolley JC. Systemic
anti-infective periodontal therapy. A systematic re-
view. Ann Periodontol 2003;8:115-181.
2. Herrera D, Sanz M, Jepsen S, Needleman I, Roldán S.
A systematic review on the effect of systemic anti-
microbials as an adjunct to scaling and root planing
in periodontitis patients. J Clin Periodontol 2002;
29(Suppl. 3):136-159; discussion 160-162.
3. Shaddox LM, Walker C. Microbial testing in periodon-
tics: Value, limitations and future directions. Periodon-
tol 2000 2009;50:25-38.
4. Listgarten MA, Loomer PM. Microbial identification in
the management of periodontal diseases. A system-
atic review. Ann Periodontol 2003;8:182-192.
5. Wolf H, Bolivar I. Microbiological diagnosis in peri-
odontology. A therapy-supporting measure (in Ger-
man). Swiss Dent 2002;23:11-18.
6. Winkel EG, van Winkelhoff AJ, Timmerman MF, Van
der Velden U, Van der Weijden GA. Amoxicillin plus
metronidazole in the treatment of adult periodontitis
patients. A double-blind placebo-controlled study.
J Clin Periodontol 2001;28:296-305.
7. Flemmig TF, Milián E, Kopp C, Karch H, Klaiber B.
Differential effects of systemic metronidazole and
amoxicillin on Actinobacillus actinomycetemcomi-
tans and Porphyromonas gingivalis in intraoral hab-
itats. J Clin Periodontol 1998;25:1-10 (erratum 1998;
25:437).
8. Rooney J, Wade WG, Sprague SV, Newcombe RG,
Addy M. Adjunctive effects to non-surgical therapy of
systemic metronidazole and amoxycillin alone and
combined. A placebo controlled study. J Clin Peri-
odontol 2002;29:342-350.
9. Cionca N, Giannopoulou C, Ugolotti G, Mombelli A.
Amoxicillin and metronidazole as an adjunct to full-
mouth scaling and root planing of chronic peri-
odontitis. J Periodontol 2009;80:364-371.
10. Löe H, Silness J. Periodontal disease in pregnancy. I.
Prevalence and severity. Acta Odontol Scand 1963;
21:533-551.
J Periodontol • January 2010
11. Jervøe-Storm PM, Alahdab H, Koltzscher M, Fimmers
R, Jepsen S. Comparison of curet and paper point
sampling of subgingival bacteria as analyzed by real-
time polymerase chain reaction. J Periodontol 2007;
78:909-917.
12. Jervøe-Storm PM, AlAhdab H, Semaan E, Fimmers R,
Jepsen S. Microbiological outcomes of quadrant ver-
sus full-mouth root planing as monitored by real-time
PCR. J Clin Periodontol 2007;34:156-163.
13. Jervøe-Storm PM, Koltzscher M, Falk W, Dörfler A,
Jepsen S. Comparison of culture and real-time PCR for
detection and quantification of five putative periodon-
topathogenic bacteria in subgingival plaque samples.
J Clin Periodontol 2005;32:778-783.
14. Socransky SS, Haffajee AD. Periodontal infections. In:
Lindhe J, Lang NP, Karring T, eds. Clinical Periodon-
tology and Implant Dentistry, vol. 1. Oxford, U.K.:
Blackwell Munksgaard; 2008:207-267.
15. Mombelli A, Casagni F, Madianos PN. Can presence or
absence of periodontal pathogens distinguish between
subjects with chronic and aggressive periodontitis? A
systematic review. J Clin Periodontol 2002;29(Suppl.
3):10-21; discussion 37-38.
16. van Winkelhoff AJ, Winkel EG. Antibiotics in peri-
odontics: Right or wrong? J Periodontol 2009;80:
1555-1558.
17. Flemmig TF, Milian E, Karch H, Klaiber B. Differential
clinical treatment outcome after systemic metronida-
zole and amoxicillin in patients harboring Actinobacil-
lus actinomycetemcomitans and/or Porphyromonas
gingivalis. J Clin Periodontol 1998;25:380-387.
18. Goené RJ, Winkel EG, Abbas F, Rodenburg JP, van
Winkelhoff AJ, de Graaff J. Microbiology in diagnosis
and treatment of severe periodontitis. A report of four
cases. J Periodontol 1990;61:61-64.
19. Pavicic MJ, van Winkelhoff AJ, de Graaff J. In vitro
susceptibilities of Actinobacillus actinomycetemco-
mitans to a number of antimicrobial combinations.
Antimicrob Agents Chemother 1992;36:2634-
2638.
20. Pavicic MJ, van Winkelhoff AJ, Douqué NH, Steures
RW, de Graaff J. Microbiological and clinical effects of
metronidazole and amoxicillin in Actinobacillus acti-
nomycetemcomitans-associated periodontitis. J Clin
Periodontol 1994;21:107-112.
21. van Winkelhoff AJ, Rodenburg JP, Goené RJ, Abbas F,
Winkel EG, de Graaff J. Metronidazole plus amoxicillin
Cionca, Giannopoulou, Ugolotti, Mombelli
in the treatment of Actinobacillus actinomyce-
temcomitans associated periodontitis. J Clin Peri-
odontol 1989;16:128-131.
22. van Winkelhoff AJ, Tijhof CJ, de Graaff J. Microbio-
logical and clinical results of metronidazole plus
amoxicillin therapy in Actinobacillus actinomycetem-
comitans-associated periodontitis. J Periodontol 1992;
63:52-57.
23. Berglundh T, Krok L, Liljenberg B, Westfelt E, Serino
G, Lindhe J. The use of metronidazole and amoxicillin
in the treatment of advanced periodontal disease. A
prospective, controlled clinical trial. J Clin Periodontol
1998;25:354-362.
24. Dannewitz B, Pohl S, Eickholz P, Kim TS. Clinical and
microbiological effects of a combined mechanic-anti-
biotic therapy in subjects with Actinobacillus actino-
mycetemcomitans-associated periodontitis. Am J
Dent 2007;20:153-156.
25. Mombelli A, Schmid B, Rutar A, Lang NP. Local
antibiotic therapy guided by microbiological diagnosis.
Treatment of Porphyromonas gingivalis and Actino-
bacillus actinomycetemcomitans persisting after
mechanical therapy. J Clin Periodontol 2002;29:
743-749.
26. Kilian M, Frandsen EV, Haubek D, Poulsen K. The
etiology of periodontal disease revisited by popula-
tion genetic analysis. Periodontol 2000 2006;42:
158-179.
27. Haubek D, Ennibi OK, Poulsen K, Vaeth M, Poulsen S,
Kilian M. Risk of aggressive periodontitis in adolescent
carriers of the JP2 clone of Aggregatibacter (Actino-
bacillus) actinomycetemcomitans in Morocco: A pro-
spective longitudinal cohort study. Lancet 2008;371:
237-242.
28. Tsai CC, Shenker BJ, DiRienzo JM, Malmud D, Taichman
NS. Extraction and isolation of a leukotoxin from
Actinobacillus actinomycetemcomitans with poly-
myxin B. Infect Immun 1984;43:700-705.
Correspondence: Dr. Norbert Cionca, Department of
Periodontology, School of Dental Medicine, University of
Geneva, Rue Barthélemy-Menn 19, CH-1205 Geneva,
Switzerland. Fax: 41-22-379-4032; e-mail: Norbert.Cionca@
unige.ch.
Submitted July 8, 2009; accepted for publication August
24, 2009.
23