Key words
Rumex acetosa
- Polygonaceae - proanthocyanidines - anti-adhesive -
Porphyromonas gingivalis
Abbreviations
API:
aproximal Plaque Index
P. g.
:
Porphyromonas gingivalis
RA1:
quantified proanthocyanidin-enriched extract from Rumex acetosa
SBI:
sulcular Bleeding Index
SD:
supragingival debridement
Introduction
Porphyromonas gingivalis (P. g. ) is one of the major pathogens associated with the onset and progression of periodontitis
[1 ]. It is viewed as a
“Keystone Pathogen” in altering the commensal oral microbiota towards a dysbiosis
that contributes to the detrimental clinical effects of periodontitis [2 ], [3 ]. P. g. is not only detected in patients with periodontal disease, but can frequently be
detected in periodontally healthy individuals, which may present a
microbiological risk factor for periodontitis [1 ], [3 ], [4 ]. Repeated professional supragingival plaque
removal markedly reduces bacterial counts but fails to eradicate P. g. from supra- and subgingival habitats [5 ]. The adhesion of P. g. to oral
surfaces, primarily mediated by fimbriae, hemagglutinins and gingipains [6 ], is pivotal for its colonization and consecutive transformation of the indigenous
healthy oral microbiome, specifically modifying the prevalence of other periodontal
pathogens, i.e. Aggregatibacter actinomycetemcomitans (A. a.), Treponema denticola (T. d.), Tannerella
forsythia (T. f.), Prevotella nigrescens (P. n.), Prevotella intermedia (P. i.), Eikenella
corrodens (E. c.), and caries associated bacteria like Streptococcus mutans (S.mutans) .
Inhibiting the adhesion of P. g. in the first instance presents a cytoprotective strategy and is a novel approach
in treating infections [7 ] and potentially
preventing periodontitis or caries.
Plant-derived polyphenols show potential in the treatment and prevention of periodontitis
[8 ]. Especially oligomeric proanthocyanidins (OPC), characterized by
oligomerized flavan-3-ol building blocks, exhibit strong anti-inflammatory and anti-adhesive
effects [9 ], [10 ]. OPC containing extracts
from green tea (Camellia sinensis ) [11 ], Myrothamnus flabellifolia
[12 ], cranberry fruits (Vaccinium macrocarpon )
[13 ], [14 ], [15 ], pomegranate fruits (Punica granatum L.) [16 ],
Rhododendron ferrugineum
[17 ] and Limonium brasiliense
[18 ] have shown to effectively inhibit the adhesion of
P. g. to human epithelial host cells. Moreover, an OPC-enriched and quantified extract
from common sorrel (Rumex acetosa L., RA) has been demonstrated in vitro to
significantly inhibit the adhesion of P. g.
[19 ] to human buccal cells in a dose-dependent manner of up to 90% [20 ].
Epicatechin-3-O -gallate-(4β -8)-epicatechin-3-O -gallate (procanidin-B2-di-gallate) has been identified as main antiadhesive compound,
which strongly interacts with the
bacterial virulence factor Arg-gingipain of P. g. , while the corresponding Lys-gingipain was hardly influenced. In silico docking studies revealed that the galloylation of this
compound is responsible for the fixation of the ligand to the Arg-gingipain [20 ]. Systematic preclinical investigation of RA indicated that the extract not only
interferes with the attachment of P. g. to eukaryotic host cells, but also inhibits the adhesion of Herpes simplex 1 [21 ] and influenza A virus to
host cells in vitro
[22 ]. As plant extracts are usually complex mixtures of different natural products the
phytochemical composition of RA has been
investigated in detail, indicating monomeric flavan-3-ols (epicatechin-3-O -gallate), oligomeric proanthocyanidins (epicatechin-(4β →8)-epicatechin (syn. procyanidin B2),
epicatechin-3-O-gallate-(4β →8)-epicatechin-3-O -gallate (syn. procyanidin B2-di-gallate), epicatechin-(β →8,2β →O → 7)-epicatechin-(4β →8)-epicatechin
(syn. cinnamtannin B1), and flavonoids (quercetin-3-O -glucuronide) as the main constituents [23 ]. The lead compound, interacting with the functionally
important regions of the main virulence factors of P. g. : Arg-gingipain and hemagglutinin, has been identified as epicatechin-3-O-gallate-(4β ,8)-epicatechin-3′-O-gallate [20 ].
The traditional use of common sorrel (R. acetosa ) for the treatment of inflammatory diseases in the mouth has been described over
centuries. However, no scientific data for
rationalizing this ethnopharmacological use have been recorded. Not surprisingly,
controlled clinical trials studying the anti-inflammatory and anti-adhesive effects
are still missing.
Therefore, a mouth rinse, containing 0.8% (w/w) of a quantified proanthocyanidin-enriched
extract of RA was developed (RA1) and evaluated in a randomized controlled pilot-trial
for its
microbiological, clinical and cytological effects in systemically healthy individuals
with a Community Periodontal Index of Treatment Needs CPITN ≤ 2 [24 ] and
harboring P. g. intraorally.
Results
The microbiological, clinical and cytopathological effects of either RA1 or the placebo
mouth rinse in combination with supragingival debridement, were evaluated in a randomized
controlled
trial with a follow-up period of 14 days.
42 out of 249 subjects fulfilling the clinical inclusion criteria were found to harbor
P. g. intraorally and were enrolled into the study. Prior to baseline visit, 4 subjects
withdrew
their participation due to scheduling difficulties and 1 subject had to be excluded
due to systemic antibiotic therapy. Furthermore, 2 subjects dropped out during the
study as they cancelled
or did not attend the scheduled appointment. A detailed overview of the participant
study flow is presented in [Fig. 1 ] and the final group distribution and
demographic characteristics in [Table 1 ]. In terms of age, gender, smoking status, number of teeth and clinical parameters
(API and SBI), the study subjects were
equally distributed across the two groups.
Fig. 1 Flow chart of the study design and participants according to CONSORT 2010.
Table 1 Group characteristics. Demographic and clinical characteristics of the included participants
at baseline V2.
Test
Placebo
p-value
SD: standard deviation
Number of participants
18
17
Χ²= 0.866
Age (mean ± SD)
44.83 ± 15.33
41.41 ± 13.31
Mann-Whitney-U-test p = 0.546
Gender F (female), M (male)
F 12, M 6
F 13, M 4
Fisher test p = 0.71
Current smoker
2
2
Fisher test p = 1.000
Number of teeth (mean ± SD)
27.44 ± 2.04
27.88 ± 2.42
Mann-Whitney-U-test p = 0.782
API V2 (mean % ± SD)
46.56 ± 24.32
49.24 ± 22.24
Mann-Whitney-U-test p = 0.708
SBI V2 (mean % ± SD)
28.28 ± 19.37
35.35 ± 23.66
Mann-Whitney-U-test p = 0.351
No significant differences were detected between the test and control group for any
of the tested microorganisms at any time point.
[Fig. 2 ] shows P. g. as a relative fraction of the total biofilm count according to the different reference
dates. In the control group, a significant
reduction of P.g . from baseline to day 2 (p < 0.01), day 4 (p < 0.01) and day 7 (p < 0.01) was found,
whereas there was no significant change in the test group. The boxplot
diagram shows an initial reduction for both groups until day 4, followed by a trend
towards an increase and a higher distribution of the individual values within the
test group. No significant
changes in other putative periodontal pathogens ([Fig. 3 a ]) or commensal bacteria ([Fig. 3 b ]) were found in neither group 7 nor 14
days after debridement.
Fig. 2 Prevalence of P. g. (relative fraction of the total plaque count) according to the time of measurement
(V1 = screening, V2 = baseline, V3 = day 2, V4 = day 4, V5 =
day 7, V6 = day 14) as a boxplot showing 0th, 25th, median, 75th and 100th percentile,
non-linear display of the Y axis to the exponent 0.2. Significant changes between
the V2 – V6
time-points compared to the V1 time-point are marked with an asterisk symbol.
Fig. 3 a Prevalence of 8 oral pathogenic microorganism (relative fraction of the total plaque
count) according to the time of measurement (V1 = screening, V2 = baseline, V3
= day 2, V4 = day 4, V5 = day 7, V6 = day 14) as a boxplot showing 0th, 25th, median,
75th and 100th percentile, non-linear display of the Y axis to the exponent 0.2. Significant
changes
between the V2 – V6 time-points compared to the V1 time-point are marked with an asterisk
symbol.
Fig. 3 b Prevalence of 4 commensal bacteria (relative fraction of the total plaque count)
according to the time of measurement (V1 = screening, V2 = baseline, V3 = day 2, V4
= day 4, V5 = day 7, V6 = day 14) as a boxplot showing 0th, 25th, median, 75th and
100th percentile, non-linear display of the Y axis to the exponent 0.2. Significant
changes between the
V2 to V6 time-points compared to the V1 time-point are marked with an asterisk symbol.
High throughput 16S rRNA gene sequence analysis was performed on one control (SA120)
and one test subject (SA088), at each time-point (V1 – V6), to obtain a representative
ʼsnap-shotʼ of the
bacterial microbiome in each subject group (results are presented in the Supporting
Information).
No significant differences in the assessed clinical parameters were found between
test and control group at any time point throughout the observation period. However,
in the test group the
API was significantly reduced from baseline to day 7 (p < 0.05) and day 14 (p < 0.01)
([Fig. 4 ]) while no significant changes over time in the API were
observed in the control group. The modified SBI was significantly reduced compared
to baseline in the test group at day 7 (p < 0.001) and 14 (p < 0.01) and in the control
group at day 7
(p < 0.01), see [Fig. 4 ]. Further, the boxplot diagram illustrates a narrow scatter of the individual values
for both groups at day 7 and day 14.
Fig. 4 API and SBI in percentage (0.4 = 40%) according to the time of measurement (V1 =
screening, V2 = baseline, V3 = day 2, V4 = day 4, V5 = day 7, V6 = day 14) as a boxplot
showing 0th, 25th, median, 75th and 100th percentile. Significant changes between
the V2 to V6 time-points compared to the V1 time-point are marked with an asterisk
symbol.
Investigation of cytopathology indicated that neither the verum nor the placebo mouth
rinse caused dysplastic changes of oral mucosal cells.
No severe adverse events were recorded during the study period. In the test group,
minor complaints including a transient brown discoloration of the tongue (n = 8),
a dry mouth feeling after
rinsing (n = 4), an unpleasant alcoholic taste of the mouth rinse (n = 4) and a white
cloudy consistence of the saliva after rinsing (n = 4) were reported. None of the
participants considered
breaching the rinsing protocol and the observations vanished completely at day 14.
Discussion
This study is the first to examine the intraoral effects of a proanthocyanidin-enriched
and standardized extract from R. acetosa L. on a microbiological, cytopathological and clinical
level. RA1 was found to be safe without any observed impact on the composition of
the intraoral microbiota. However, RA1 did not significantly reduce the intraoral
prevalence of
P. g. .
The early microbial succession in redeveloping dental biofilms in periodontally healthy
and diseased subjects after professional debridement and followed by 7 days of no
oral hygiene has been
assessed before [25 ]. In the study of Teles et al. (2012) [25 ], the values for P. g. within the supragingival biofilm samples
showed no significant difference directly after the debridement and 1-day post. Whereas
2, 4 and 7 days later the values were significantly reduced. The same pattern can
be seen in the placebo
group of this study with a significant reduction of P. g. at day 2, day 4 and day 7. In contrast, P. g. values for the test group did not show any significant reduction. Looking
at day 7 in the boxplot diagram (see [Fig. 2 ]) it even shows an increased variability of the individual values. Noting that the
P. g. levels vary
differently in those two groups, across the time-points, it can be concluded that
RA1 has an effect on P. g. in vivo during this period. Theoretically, P. g. might initiate
regulatory or compensatory mechanisms, e.g., changing its proliferation rate or increasing
the expression of adhesion proteins, to counteract the adhesion inhibition by RA1.
In vitro
studies on the influence of RA on the expression of rgpA for Arg-gingipain, kgp for Lys-gingipain and fimA for fimbrillin demonstrated constant gene expression rates in
P. g. liquid cultures [20 ]. It is known that the gene expression of P. g. differs when grown in a biofilm or planktonic culture [26 ]. Hence, additional in vitro studies should investigate the effects of RA1 on gene expression levels when P. g. is biofilm associated.
The prevalence of the other pathogenic bacteria tested, A. a., T. d., P. n. and S. mutans, parallels those described by Teles et al. (2012) [25 ].
Interestingly, the results seem to differ for T. f. . Whereas the placebo group follows the same significant reductive trend as seen in
the study from Teles et al. (2012) [25 ], the values for the test group seem to stay at a constant level. In supra- and subgingival
biofilms P. g. and T. f. cells may associate with one
another [27 ], [28 ] and both have a synergistic effect in causing periodontal abscesses [29 ]. Under
limited nutrient supply T. f. stimulates the growth of P. g. , and gingipains play an important role in the uptake and digestion of growth-promoting
factors [29 ]. T. f. likewise has the ability to adhere and invade into epithelial cells and the presence
of P. g. enhances this process [30 ]. For the commensal bacteria S. sang., Veill., S. mitis and Act. the values of the test and placebo groups are akin to those reported by Teles et
al. (2012) and no
intergroup differences could be detected. In vitro tests have shown that proanthocyanidins selectively target gram-negative periodontal
pathogenic strains while preserving the viability
of the beneficial commensal Streptococcus salivarius
[31 ]. RA1 and the respective placebo do not cause a significant overgrowth of pathogenic
bacteria or a
significant reduction of commensal bacteria. The control and test mouth rinses have
no negative effect on the qualitative composition of the supragingival biofilm and
the exemplary microbiome
analysis of one test and one control subject validated this.
PCR does not differentiate between viable and dead bacteria. If a mouth rinse has
a fast therapeutic effect on reducing the amount of living pathogens, using a molecular
genetic analysis via
real time quantitative PCR would have been a desired method [32 ]. RA1 was not designed with an antibacterial concentration. Therefore, determination
of the
viability of P. g. was not requisite. For this study a baseline professional debridement was performed
to disrupt the existing biofilm and to enable RA1 to directly interact with
P. g. and other bacteria. In general, supragingival debridement reduces the basic bacteria
count to a uniform level and after two days the total bacteria count is expected to
correlate
with the initial values again [25 ], [33 ]. Hence, any anti-adhesive and biofilm inhibitory effect of RA1 should be reflected
in changes
of clinical and microbiological parameters after 2 days, as it has been demonstrated
with pomegranate and green tea-containing mouth rinses [16 ], [34 ]. The professional debridement induced the overall reduction of the API values in
both groups at the day 7 time-point. At day 14, the API for the placebo group
reverted to the baseline values. In contrast, the test groupʼs proximal biofilm levels
continued to stay significantly decreased even 7 days after cessation of RA1. It seems
that RA1 has a
prolonged inhibitory effect on biofilm formation. In contrast to this finding, long
term studies with a cranberry mouth rinse over 6 weeks failed to prove a biofilm inhibitory
effect but
noticed a reduction in the total bacteria count [35 ]. Further long-term studies are needed to determine if bacteria build up a tolerance
against the mouth rinse,
and for how long RA1 remains effective after cessation.
Using a CPITN ≤ 2 as inclusion criteria implies that participants with gingivitis
have been included into the study. The SBI values at baseline signal a modest level
of gingival inflammation
in both groups and those numbers are coherent with the amount of biofilm detected
in the API at the same time-point. In systemically healthy patients, debridement improves
clinical signs and
symptoms of gingival inflammation within 14 days. Therefore, the main cause of the
significant reduction of the sulcular bleeding at day 7 in both groups seems to be
the removal of calculus
and biofilm at the baseline visit. In vitro , RA has been shown to significantly reduce IL-6 release and therefore act cytoprotective
[19 ]. As the SBI in the
test group on day 14 continued to stay on a significantly reduced level, this could
either be based on the lower biofilm levels or on the reduced level of the proinflammatory
cytokine IL-6.
Future studies involving a larger number of participants will be required to determine
whether RA1 reduces sulcular bleeding levels or the plaque formation by a significant
amount. In
addition, an experimental gingivitis study model (or similar) may be highly informative
for further elucidating the potential influence of RA1 on proinflammatory cytokines
[36 ], [37 ].
To validate the effectiveness of a mouth rinse and the method of testing, a test substance
should be assessed against a positive and a negative control rinsing agent. 0.2% chlorhexidin
digluconate is the ʼgold-standardʼ due to its unspecific bacteriostatic and subsequent
plaque inhibiting effect. Nevertheless, the focus of this study was to determine the
anti-adhesive
properties of RA1 versus a placebo and not the antimicrobial effect. Hence, it was
decided not to test against a positive control group using chlorhexidin digluconate.
It has already been demonstrated that proanthocyanidins are non-toxic to gastric mucosal
cells [38 ]. Nevertheless, mouth rinse formulas can have a mutagenic and
cytotoxic effect on mucosa cells [39 ]. Therefore, assessing the safety of RA1 on a cytologic level was indispensable.
Brush biopsies with a negative cytology and a
lack of DNA aneuploidy can exclude high-grade oral epithelial cell dysplasia or squamous
cell carcinoma avoiding invasive diagnostic biopsies [40 ]. In this study
none of the participants of either group showed dysplastic changes of the buccal mucosa
cells. It follows, RA1 is non-cytotoxic and safe to use over a 7-day period.
Proanthocyanidins are commonly linked to a bitter taste, an astringent sensation and
the formation of complexes (precipitates) of salivary proteins + procyanidin [41 ]. The participantʼs compliance to the rinsing protocol could have been affected by
this. A few members of the test group reported only minor complaints, which did not
affect their
quality of life during the study period or prevented from adhering to the rinsing
protocol. All experienced sensations can be related to the proanthocyandinsʼ characteristics.
Therefore, the
test mouth rinses can be considered non-cytotoxic over a clinical usage period of
7 days and in addition to that is well tolerated by the study participants.
Although the RA1 mouthwash did not show any therapeutic effects by intergroup comparisons,
the intragroup comparison indicates anti-inflammatory effects and interference with
biofilm
formation. This and the lack of any cytotoxic effects may support the notion of RA1
as a safe cosmetic adjunct following professional debridement. However, long-term
intervention studies are
needed to fully appraise the anti-inflammatory and anti-plaque effects of R. acetosa .
Material and Methods
Study subjects
This study has been carried out in accordance with the ethical principles of the World
Medical Association Declaration of Helsinki (version 2008) and was independently reviewed
and approved
by the institutional review board of the Medical Faculty of the Heinrich-Heine-University,
Düsseldorf, Germany (approval number #3786, approval date 16/03/2012). The clinical
trial was
registered at www.clinicaltrials.gov under the number NCT02039648. All participating individuals signed an informed consent
form. Inclusion
criteria were age ≥ 18 years, systemically healthy, a CPITN ≤ 2 and tested positive
for P. g. (see below for experimental details). Exclusion criteria were age < 18 years,
pregnancy or nursing, allergies to several mouthwash components, soft tissue lesions
(e.g. leucoplakia, lichen planus), any systemic conditions that require an antibiotic
prophylaxis for
routine dental procedures (e.g. endocarditis), antibiotic therapy in the previous
6 months, any systemic disease (e.g. diabetes or immunological disorders) or long-term
medication, that
might have potential influence on the immune response, history of periodontal disease
and a CPITN ≥ 3.
Study design
The microbiological, clinical and cytopathological effects of either RA1 or the placebo
mouth rinse in combination with supragingival debridement, were evaluated in a two-arm
parallel
randomized controlled clinical trial with a follow-up period of 14 days. The participants,
the examiner and the outcome assessors were blinded to the group assignment.
The extract RA1 was prepared from the herbal material of R. acetosa L., as previously described in detail [20 ]. In principle dried R. acetosa
herbal material (A. Galke, batch 14593) was exhaustively extracted with acetone/water
(7 : 3 v/v), followed by evaporation of the organic solvent of the extract in vacuo , removal of
precipitated chlorophyll by filtration, removal of lipophilic compounds by extraction
with heptan and lyophilisation of the aqueous phase to yield 9.3% (w/w) of the extract
RA1. For
analytical quality control a ICH-2 validated UHPLC method [20 ] was used for identification and quantitation of the marker compounds
epicatechin-3-O -gallate (17.9 mg/g extract RA1), epicatechin-3-O -gallate-(4β -8)-epicatechin-3-O -gallate (syn. procyanidin-B2) (3.8 mg/g),
epicatechin-3-O -gallate-(4β -8)-epicatechin-3-O -gallate (syn. procyanidin-B2-di-gallate) (11.9 mg/g),
epicatechin-(4β -8,2β -O -7)-epicatechin-(4β -8)-epicatechin (syn. cinnamtannin B1)) (6.0 mg/g) and quercetin-3-O -glucuronide (24.0 mg/g) [20 ].
Mouth rinse solutions, containing 0.8% (w/w) of RA1 were manufactured in a facility,
which has been approved for manufacture of pharmaceuticals. The detailed composition
of the verum and
placebo preparations are displayed in [Table 2 ]. All ingredients complied with the specifications of European Pharmacopeia. Food
coloring and aroma were added to
the placebo mouth rinse to mimic the taste, smell, and color of the test solution.
Table 2 Ingredients RA1 verum and placebo mouth rinse.
ingredient RA1
specification
mass fractions (g)
function
RA
0.8
active ingredient
calcium-L-ascorbate
Ph. Eur. 6.0
0.1
stabilizer, antioxidant
ethanol 96%
Ph. Eur.
6.4
solubilizer
glycerol 85%
Ph. Eur. 6.0
15
solubilizer
potassium sorbate
Ph. Eur. 6.0
0.47
preservative
citric acid, anhydrous
Ph. Eur. 6.0
0.014
pH-stabilizer
peppermint aroma
Symrise AG, Holzminden, Germany
0.2
aroma
water, sterile, pyrogen free
B. Braun AG, Melsungen, Germany
ad 100
solvent
ingredient placebo
specification
mass fractions
function
calcium-L-ascorbate
Ph. Eur. 6.0
0.1
stabilizer, antioxidant
ethanol 96%
Ph. Eur.
6.4
solubilizer
glycerol 85%
Ph. Eur. 6.0
15
solubilizer
potassium sorbate
Ph. Eur. 6.0
0.47
preservative
citric acid, anhydrous
Ph. Eur. 6.0
0.045
pH-stabilizer
peppermint aroma
Symrise AG, Holzminden, Germany
0.2
aroma
plain orange aroma
Symrise AG, Holzminden, Germany
0.25
aroma
Symcolor caramel EBC30500
Symrise AG, Holzminden, Germany
0.1
EU-approved food colouring E 150 C
water, sterile, pyrogen free
B. Braun AG, Melsungen, Germany
ad 100
solvent
The rinses were stored in identical containers and labeled with numbers following
a randomly assigned allocation ratio of 50 : 50 by a staff-member of the Institute
of Pharmaceutical
Biology and Phytochemistry, University of Münster, Germany. The mouth rinses were
handed to the subjects in ascending sequence of their label number following the baseline
date.
The treatment allocation was unknown to the examiner, the study participants, microbiologist
and cytopathologist. A code break for the randomization was stored in individually
sealed
envelopes on site in case of adverse events. The randomization scheme was decoded
for statistical analyses.
Intervention
At baseline, the participants received a supragingival debridement (SONICflex 2003 L
with Paro instrument tip No. 60, KaVo) and polishing (AIR-FLOW MASTER with AIR-FLOW
Powder Perio and
Classic, EMS). Standardized instructions were given, in verbal and written form, to
rinse 3 times per day with 10 ml of the assigned mouth rinse for 7 days in addition
to their oral hygiene
procedures. No specific oral hygiene instructions were provided. Supragingival biofilm
samples were taken at pre-determined visits (V1 = screening, V2 = baseline after debridement,
V3 = 2
days, V4 = 4 days, V5 = 7 days and V6 = 14 days after baseline) and P. g. was identified and quantified by real-time polymerase chain reaction (qrt-PCR) using
previously described
primers (Supplementary Data Table 1S ). In addition, the amounts of eight species of potentially-pathogenic oral bacteria:
Aggregatibacter actinomycetemcomitans (A. a.), Treponema
denticola (T. d.), Tannerella forsythia (T. f.), Prevotella nigrescens (P. n.), Prevotella
intermedia (P. i.), Eikenella corrodens (E. c.) and Streptococcus mutans (S.mutans) ; and
four species of oral commensal bacteria: Streptococcus sanguinis (S.sang.), Streptococcus mitis (S.mitis), Veillonella parvula
(Veill.) and Actinomyces viscosus (Act.) were measured
during the entire study period using an analogous qrt-PCR. Furthermore, clinical parameters,
the Approximal Plaque Index (API) [42 ] and the modified Sulcular
Bleeding Index (SBI) [43 ], were recorded at baseline before debridement, at day 7 and day 14. Oral brush biopsy-based
cytology was chosen for the non-invasive
investigation of oral epithelial dysplasia. At baseline and day 7 cellular material
was collected by rotating a brush (Gynobrush Plus, Heinz Herenz) 10-times over the
left and right buccal
mucosa. The cells were transferred onto a microscopic slide and fixated (Merckofix,
Merck).
On each visit medical health and medication updates, compliance and occurrence of
adverse events were checked. For consistency and reliability purposes enrollment and
all clinical
interventions were performed by the same experienced periodontist in the facilities
of the Department of Periodontics, Heinrich-Heine-University, Düsseldorf, Germany.
Microbiological analysis
Sample collection One tooth per sextant was selected, based on the presence of biofilm during the screening
visit. At screening, baseline, day 2, 4, 7 and 14 supragingival biofilm
samples were obtained from the same teeth with a sterile hand instrument (Scaler 7835,
Storz am Mark) and by swiping an additional cotton swab over the tongue, palate, cheeks
(left and
right) and floor of the mouth. The collected samples were immediately frozen at − 80 °C
until further analysis.
DNA extraction and qrtPCR DNA isolation was processed with the innuPREP DNA Mini Kit (Analytik Jena AG) according
to the manufacturerʼs instructions. For qrtPCR the maxima SybrGreen
Kit (Thermo Fisher Scientific) on CFX 96 BioRad was used. The bacteria specific primers
and cycle settings are listed in Supplementary Data Table 1S .
Bacterial microbiome analysis 16 s rRNA gene sequence analysis was performed on DNA isolated from one randomly
selected subject from each of the test and control groups. Detailed
information is provided in the Supporting Information.
Cytopathological analysis
Oral brush biopsies were examined by an experienced cytopathologist at the Institute
of Cytopathology (Heinrich-Heine-University, Düsseldorf, Germany) using the method
of Dr. Papanicolaou
and interpreted based on the Düsseldorf nomenclature [44 ].
Statistical methods
Sample size calculation Because of the innovative character of the mouth rinse no data were available in
literature on the estimated quantitative difference in P. g. or any
other bacteria prevalence. Hence, the calculations were based on a large effect with
a dichotomous distribution. To detect a 60% difference in the P. g. prevalence with a power of 80%
and a p ≤ 0.05, a minimum sample size of 15 persons per group was required. A 25%
drop-out rate was considered as participants might have difficulty complying with
the high number of visits
within the 14-day study period [45 ]. Additionally, no data existed on the taste quality of the tested mouth rinses and
potential effects on the compliance to
adhere to the study protocol. Accordingly, a sample size of 40 participants was determined
for the overall study population (20 per group).
Outcomes The primary outcome measure of this study was the change of the intraoral prevalence
of P. g. between the test and control group at baseline, day 2, day 4, day 7 and
day 14. Secondary outcomes included the change of the API and SBI values and the cytopathological
appearance of the mucosal tissue. The change of prevalence of the 11 other oral
microorganisms was considered as additional outcome parameters. Microbiome analysis
of one test and one control subject was done as a safety precaution to detect for
any alterations in the
composition of the oral microflora.
Statistical analysis The statistical analysis was accomplished by IBM SPSS Statistics 21 (IBM). For a
simple descriptive analysis continuous data were calculated as mean values with
their standard deviation. Due to the small group size non-parametric testing methods
were used. The Fisher exact test was performed to assess the association between demographic
variables of
the test and control group. Every comparison of the test- and the control-group regarding
the clinical and microbiological outcomes was performed using the Mann-Whitney-U-test.
Multiple
comparisons of time variables were conducted by the Friedman-test and their significant
data were confirmed by the post-hoc Wilcoxon rank test with a Bonferroni adjusted
alpha. The level of
significance was set at p ≤ 0.05.
Competing Financial Interests Statement
Competing Financial Interests Statement
This study was financially supported by the European Union (ZIEL.2.NRW program) with
funding to TB (grant number 300 262 502) and AH (grant number w1006sb014). Residual
expenses were borne by
intramural annual budgets from the University of Düsseldorf to TB and the University
of Münster to AH. The study design, data collection and analysis, as well as the intention
to publish and
its completion have not been influenced by the funders.
Contributorsʼ Statement
S. S. conducted the clinical trial, clinical sample and data acquisition and drafted
the manuscript. S. B. and A. H. manufactured the test and placebo solution and were
responsible for the
randomization. Sample processing and bioinformatics analysis was completed by U. P.,
R. W. and R. T. The conception, study design and protocol, data analysis and interpretation
were performed
by S. S., A. K., U. P., T. F. and partially T. B. All authors have critically revised
and approved the final manuscript.
