Chlorhexidine has been widely used in clinical dentistry

as an antiplaque or anticariogenic additive for mouthrinses,

toothpastes, varnishes, etc. [Gjermo, 1989]. The retention
of chlorhexidine at plaque-inhibiting and antibacterial con-
centrations in the oral cavity is the determinant for its prac-
tical effect after oral application. Therefore, the concentra-
tions of salivary chlorhexidine have interested researchers
in an oral pharmacokinetic study.
Different analytical methods have previously been re-
ported for the quantitation of salivary chlorhexidine. A
method using
14
C-labeled chlorhexidine was used to evalu-
ate its retention in the oral cavity [Bonesvoll et al., 1974;
Bonesvoll and Gjermo, 1978]. However, the radioactive
technique is not ethically applicable to humans. UV spec-
troscopic [Jensen and Christensen, 1971] and fluorometric
methods [de Vries et al., 1991] need improvements in speci-
ficity and sensitivity for chlorhexidine and applicability to
all types of samples because they are batch methods. When
dealing with saliva samples, endogenous components in
saliva matrix other than analytes interfere with the quantita-
tive analysis. Among different methods, a chromatographic
technique is most suitable for the selective quantitation of
chlorhexidine in saliva. Analytical methods by high-per-
Original Paper
Caries Res 1999;33:156163
Received: April 14, 1998
Accepted after revision: July 7, 1998
High-Performance Liquid Chromato-
graphic Analysis of Chlorhexidine in
Saliva after Mouthrinsing
Fax +41 61 306 12 34
E-Mail karger@karger.ch
www.karger.com
1999 S. Karger AG, Basel
00086568/99/03320156 $17.50/0
Accessible online at:
http://BioMedNet.com/karger
Hironori Tsuchiya
Department of Dental Pharmacology
Asahi University School of Dentistry
1851 Hozumi, Hozumi-cho, Motosu-gun, Gifu 501-0296 (Japan)
Tel./Fax +81 58 329 1432, E-Mail hiro@dent.asahi-u.ac.jp
H. Tsuchiya T. Miyazaki S. Ohmoto
Department of Dental Pharmacology, Asahi University School of Dentistry, Gifu, Japan
Abstract
A high-performance liquid chromatographic method
was developed to quantify chlorhexidine in human
saliva. After addition of an internal standard and acidic-
deproteinization, saliva samples were chromato-
graphed by the reversed-phase ion-pair system using
pentanesulfonate as a counterion and the eluates were
detected by a diode array detector. Under optimal con-
ditions, the baseline separation of analytes was
achieved within 4.5 min without interference from com-
ponents in saliva matrix. The method showed high se-
lectivity to salivary chlorhexidine, quantitative range
(50.0 ng/ml50.0 g/ml), recovery (96.097.8%) and ana-
lytical precision (intra- and interassay CVs within
0.41%), permitting the effective application to both sali-
va and aqueous solutions. Chlorhexidine was found in
saliva at microgram per milliliter levels for at least 8 h
after mouthrinsing with 10 ml of an aqueous solution of
chlorhexidine (0.1 or 1.0 mg/ml) for 1 min. The concen-
trations of salivary chlorhexidine were reduced by in-
gesting the acidic beverage and food, but not by ingest-
ing the neutral beverage. The adsorption experiments in
artificial saliva revealed that chlorhexidine was signifi-
cantly adsorbed to saliva-coated hydroxyapatite, buccal
Key Words
Chlorhexidine High-performance liquid chromato-
graphy Quantitative analysis Retention Saliva
epithelial cells and mucin. The proposed method will be
a useful tool to study salivary chlorhexidine after oral
application and its retention in the oral cavity.
D
o
w
n
lo
a
d
e
d

b
y
:

U
n
iv
e
r
s
id
a
d
e

F
e
d
e
r
a
l
d
e

M
in
a
s


























1
5
0
.
1
6
4
.
1
1
0
.
1
2
7

-

7
/
2
3
/
2
0
1
4

2
:
3
4
:
5
0

P
M
formance liquid chromatography (HPLC) have recently
been reported for salivary chlorhexidine [Lam et al., 1993;
Pesonen et al., 1995]. However, the methodological prob-
lems to be solved have remained in these HPLC methods
concerning low recovery, precision and sensitivity.
Chlorhexidine is a cationic compound, which possibly
forms the ion-pair with an anionic counterion. The formed
ion-pair is subject to partition onto the solid phase in re-
versed-phase HPLC analysis, leading to the selective sepa-
ration of chlorhexidine from interfering compounds in sali-
va matrix. In the present study, a reversed-phase ion-pair
HPLC method using an alkyl sulfonate as a counterion was
developed to selectively and sensitively quantify chlorhexi-
dine in saliva and aqueous solutions. After optimization of
different analytical conditions, the proposed method was
applied to the quantitation of salivary chlorhexidine after
mouthrinsing. It was also applied to assess the adsorption
of chlorhexidine to hydroxyapatite, buccal epithelial cells
and proteins, which would account for the retention of
chlorhexidine in the oral cavity.
Materials and Methods
Saliva Collection and Sample Preparation
Whole saliva was collected from male subjects, aged 45 and 46
years, according to the standard guidelines for saliva collection
[Malamud and Tabak, 1993]. After an initial swallow, the subjects al-
lowed saliva to drain from their lower lips into collection tubes. At the
end of the collection, they expectorated residual saliva. The subjects
rinsed their mouths with 10 ml of an aqueous solution of chlorhexi-
dine diacetate for 1 min. The chlorhexidine concentration in rinsing
solutions was adjusted to either 1.0 or 0.1 mg/ml. Saliva (0.51 ml)
was collected at 0.25- to 8-hour intervals after mouthrinsing. All sub-
jects (n = 20 for rinsing with 1.0 mg/ml and n = 9 for rinsing with
0.1 mg/ml) abstained from beverage and food during a saliva collec-
tion period. To assess the influence of food ingestion on salivary
chlorhexidine, the same subjects (n = 10) rinsed their mouths on dif-
ferent days with 10 ml of an aqueous solution of chlorhexidine
(1.0 mg/ml) for 1 min and took a meal 2.25 h after mouthrinsing. To
compare salivary chlorhexidine between abstinence and beverage in-
gestion, three groups of subjects (n = 8 for each group), all aged 46
years, rinsed their mouths with 10 ml of an aqueous solution of
chlorhexidine (1.0 mg/ml) for 1 min. One group abstained from bev-
erage during a saliva collection period. The other groups drank 300 ml
of either green tea or orange juice 1.25 h after mouthrinsing.
Immediately after collection, the collected saliva was diluted with
10 mM HCl by defined dilution factors (2- to 5-fold diluted). A 250-
l aliquot of each saliva diluent was vortex-mixed for 3 s with 50 l
of 10% (v/v) HClO
4
containing 1,3-diphenylguanidine (30.0 g/ml)
as an internal standard. After centrifugation at 100,000 g for 5 min,
the supernatant was subjected to HPLC analysis. The concentrations
of salivary chlorhexidine were evaluated by means of quantitative
analyses in duplicate for each saliva sample.
HPLC Analysis
The HPLC system consisted of an LC-10ADVP liquid chromato-
graph (Shimadzu, Kyoto, Japan), a KMT-60A-II autosampler (injec-
tion volume of 40 or 100 l; Kyowa Seimitsu, Tokyo, Japan), a Shim-
pack CLC-C8 (M) column (250E4.6 mm ID, particle size of 5 m;
Shimadzu) placed in a thermo-controller (Kyowa Seimitsu) and an
SPD-M10AVP diode array detector (Shimadzu) controlled by an
FMV-5133D5 personal computer (Fujitsu, Tokyo, Japan). As the mo-
bile phase, 10 mM sodium pentanesulfonate solution in acetonitrile-
50 mMsodium phosphate buffer of pH 3.0 (47:53, v/v) was delivered
at a flow rate of 1.0 ml/min and at a column temperature of 50C.
Salivary chlorhexidine was identified by comparing the diode array
spectra and capacity factor of its peak obtained from saliva analysis
with those of standard chlorhexidine. It was quantified based on the
peak area ratio to 1,3-diphenylguanidine by reference to the calibra-
tion graph prepared as described below. The concentrations of
chlorhexidine in saliva were corrected by the recovery and dilution
factor.
Optimization of HPLC Conditions
Chlorhexidine and 1,3-diphenylguanidine (5.00 g/ml of each)
were chromatographed as described above by varying the mobile
phase composition: the type (carbon number of alkyl chains: 38) and
concentration (020 mM) of sodium alkyl sulfonates, the phosphate
buffer pH (24) and the acetonitrile concentration (3555%, v/v).
Chromatographic conditions were optimized based on the obtained
capacity factors. Results are expressed by the means of three different
determinations.
Analytical Evaluation
To evaluate the quantitative range, a calibration graph was pre-
pared by plotting the mean peak area ratios of chlorhexidine to 1,3-
diphenylguanidine in duplicate experiments against the known con-
centrations. Standard chlorhexidine (50.0 ng/ml50.0 g/ml) was
added to saliva diluents or water, and then the spiked solutions were
analyzed as described above.
To evaluate the recovery and analytical precision, replicate saliva
samples spiked with chlorhexidine were analyzed. Chlorhexidine
(0.50 or 5.00 g/ml) was added to saliva diluents which were pre-
pared 15 min after mouthrinsing with 10 ml of an aqueous solution of
chlorhexidine (1.0 mg/ml) for 1 min. The mean recovery (n = 7), and
the intra- (analyzed on the same day, n = 7) and interassay (analyzed
on different days, n = 4) coefficients of variation (CVs) were deter-
mined.
Adsorption of Chlorhexidine to Hydroxyapatite
Fifty milligrams of hydroxyapatite beads (particle size of 200 m;
Tonen, Tokyo, Japan) were washed with 2.0 ml of Katzs artificial
saliva: 0.3 mMCaSO
4
, 1.0 mMNaCl, 0.7 mMKCl, 0.4 mMKH
2
PO
4
and 0.4 mM Na
2
HPO
4
(pH 6.8) [Katz et al., 1986], thereafter cen-
trifuged at 800 g for 5 min. This washing treatment was repeated
twice. An experimental pellicle was prepared according to a previous
method [Steinberg et al., 1993]. Briefly, whole saliva collected from
different male subjects (n = 3), aged 45 and 46 years, was clarified by
centrifugation at 25,000 g for 20 min. Fifty milligrams of hydroxyap-
atite beads were incubated in 1.0 ml of clarified saliva for 45 min at
37C. Saliva-coated hydroxyapatite beads were similarly washed for
removal of unbound salivary components. Either saliva-coated or not
coated hydroxyapatite beads were vortex-mixed for 1 min in 1.0 ml of
a chlorhexidine solution (10.00 g/ml) in Katzs artificial saliva. Af-
HPLC Analysis of Salivary Chlorhexidine
Caries Res 1999;33:156163
157
D
o
w
n
lo
a
d
e
d

b
y
:

U
n
iv
e
r
s
id
a
d
e

F
e
d
e
r
a
l
d
e

M
in
a
s


























1
5
0
.
1
6
4
.
1
1
0
.
1
2
7

-

7
/
2
3
/
2
0
1
4

2
:
3
4
:
5
0

P
M
ter centrifugation at 800 g for 5 min, a 250-l aliquot of the super-
natants was mixed with 50 l of a 1,3-diphenylguanidine solution
(30.0 g/ml), followed by HPLC analysis. Results are expressed by
the mean B SE of seven different determinations for each subject.
Adsorption of Chlorhexidine to Buccal Epithelial Cells
Buccal epithelial cells collected from different male subjects
(n = 3), aged 45 and 46 years, were washed with 5.0 ml of Katzs ar-
tificial saliva and centrifuged at 10,000 g for 20 min. This washing
treatment was repeated twice. Either washed or not washed cells were
suspended in Katzs artificial saliva to 2.0E10
6
cells/ml based on an
optical density at 550 nm [Webb et al., 1995]. Chlorhexidine
(10.00 g/ml) was added to 1.0 ml of each cell suspension and vortex-
mixed for 1 min. The mixture was ultrafiltered by a Centricon-10
(cutoff molecular weight of 10,000; Amicon, Danvers, Mass., USA).
After centrifugation at 5,000 g for 20 min, a 250-l aliquot of the ul-
trafiltrates was mixed with 50 l of a 1,3-diphenylguanidine solution
(30.0 g/ml), followed by HPLC analysis. Results are expressed by
the mean B SE of seven different determinations for each subject.
Adsorption of Chlorhexidine to Proteins
Chlorhexidine (10.00 g/ml) was vortex-mixed for 1 min in
1.0 ml of Katzs artificial saliva containing mucin (0.110 mg/ml;
Wako, Tokyo, Japan) or albumin (0.110 mg/ml; Nacalai Tesque, Ky-
oto, Japan). The mixture was ultrafiltered by a Centricon-10, and then
a 250-l aliquot of the ultrafiltrates was analyzed as described above.
Results are expressed by the mean B SE of four different determina-
tions.
Statistical Analysis
The data obtained from mouthrinsing and adsorption experiments
were evaluated by Students t test.
Results
HPLC Analysis of Chlorhexidine
When using a reversed-phase C8 column, the capacity
factors of chlorhexidine and 1,3-diphenylguanidine in-
creased by adding different alkyl sulfonates to the mobile
phase (fig. 1). Such an increase depended on the length of
alkyl chains (carbon number varying from 3 to 8). The re-
tention of both analytes was enhanced with increasing
concentrations of sodium pentanesulfonate chosen as a
counterion (fig. 1). As a compromise between rapid separa-
tion and high resolution, the concentration of 10 mM was
used for sodium pentanesulfonate, which provided the reso-
lution factor greater than 2.5 between analytes. The other
HPLC conditions and the pretreatment conditions for saliva
samples were optimized as described in Materials and
Methods. Under optimal conditions, the baseline separa-
tion of chlorhexidine and 1,3-diphenylguanidine was com-
pleted within 4.5 min (fig. 2). In application to saliva sam-
ples, components from saliva matrix never interfered with
the separative analysis of both analytes. The ingestion of
any beverage and food provided no peaks other than those
obtained in abstinence conditions, indicating that neither
drinking nor eating influences the quantitation of salivary
chlorhexidine. All the peaks obtained after mouthrinsing
showed the same diode array spectra and capacity factors as
those of standard chlorhexidine.
158
Caries Res 1999;33:156163
Tsuchiya/Miyazaki/Ohmoto
Fig. 1. Effects of the carbon number of alkyl
sulfonates (a) and the concentration of sodi-
um pentanesulfonate (b) used as a counterion
on the capacity factors.
a b
D
o
w
n
lo
a
d
e
d

b
y
:

U
n
iv
e
r
s
id
a
d
e

F
e
d
e
r
a
l
d
e

M
in
a
s


























1
5
0
.
1
6
4
.
1
1
0
.
1
2
7

-

7
/
2
3
/
2
0
1
4

2
:
3
4
:
5
0

P
M
When using 1,3-diphenylguanidine as an internal stan-
dard, the calibration graph showed a good linearity in con-
centrations from 50.0 ng/ml to 50.0 g/ml of chlorhexidine.
The regression equation was found to be y = 0.2922x +
0.0056 (r
2
= 0.9999). There was no difference in a slope of
the calibration graph between saliva and aqueous solutions
added with chlorhexidine of the same concentrations. In
spiking experiments, the mean recovery, and the intra- and
interassay CV were 97.8, 0.22 and 0.32% for spiking with
5.00 g/ml, and 96.0, 0.33 and 0.41% for spiking with
0.50 g/ml, respectively.
All analytical procedures including saliva pretreatment
and HPLC separation were completed within 15 min per
sample.
Salivary Concentration of Chlorhexidine after
Mouthrinsing
The concentrations of salivary chlorhexidine decreas-
ingly changed with time after mouthrinsing (table 1). In ab-
stinence from beverage and food, chlorhexidine maintained
the salivary concentrations of 1.97B0.38 and 0.48B0.10
g/ml even 8 h after mouthrinsing with chlorhexidine of 1.0
and 0.1 mg/ml, respectively. By taking a meal 2.25 h later,
however, salivary chlorhexidine significantly decreased as
seen in saliva 4, 6 and 8 h after mouthrinsing (pd0.01 for
all, vs. abstinence).
The concentrations of salivary chlorhexidine were influ-
enced by the kind of ingested beverages (fig. 3). Although
no difference was found between abstinence and green tea
ingestion, drinking orange juice resulted in a significant de-
crease of salivary chlorhexidine 1.5, 2, 3, 4, 5 and 6 h after
mouthrinsing (pd0.01 for all, vs. abstinence). The pH of
green tea and orange juice was 6.95 and 3.85, respectively.
Adsorption of Chlorhexidine to Hydroxyapatite
Chlorhexidine in artificial saliva decreased to 7.86B0.06
g/ml by treating with hydroxyapatite (pd0.01, vs. the
original concentration of 10.00 g/ml). When coating hy-
droxyapatite with saliva of 3 subjects, the concentrations
were further reduced to 5.15B0.28, 5.73B0.20 and 7.09B
0.05 g/ml (pd0.01 for all subjects, vs. without saliva-
coating).
Adsorption of Chlorhexidine to Buccal Epithelial Cells
When treating with washed cells of 3 subjects, chlorhex-
idine in ultrafiltrates decreased to 7.95B0.06, 8.21B0.14
and 9.64B0.11 g/ml (pd0.01 for all subjects, vs. the orig-
inal concentration of 10.00 g/ml). The treatment with not
HPLC Analysis of Salivary Chlorhexidine
Caries Res 1999;33:156163
159
Fig. 2. High-performance liquid chromatograms obtained from standard and saliva samples. After mouthrinsing with
10 ml of an aqueous solution of chlorhexidine (1.0 mg/ml) for 1 min, saliva was collected 2 and 4 h later. Saliva col-
lecting conditions: abstinence from beverage and food during a saliva collection period, orange juice ingestion after
1.25 h and food ingestion after 2.25 h.
D
o
w
n
lo
a
d
e
d

b
y
:

U
n
iv
e
r
s
id
a
d
e

F
e
d
e
r
a
l
d
e

M
in
a
s


























1
5
0
.
1
6
4
.
1
1
0
.
1
2
7

-

7
/
2
3
/
2
0
1
4

2
:
3
4
:
5
0

P
M
160
Caries Res 1999;33:156163
Tsuchiya/Miyazaki/Ohmoto
Fig. 3. Influence of beverage ingestion on concentrations of salivary
chlorhexidine after mouthrinsing with 10 ml of an aqueous solution of
chlorhexidine (1.0 mg/ml) for 1 min. The subjects (n = 8 for each ex-
periment) abstained from beverage during a saliva collection period
or ingested either green tea or orange juice 1.25 h after mouthrinsing.
Statistically significant from abstinence: *pd0.01.
Fig. 4. Chlorhexidine concentrations in ultrafiltrates obtained after
mixing with mucin or albumin of different concentrations. Statistical-
ly significant from albumin treatment: *pd0.01.
Table 1. Mean (B SE) concentrations of salivary chlorhexidine after mouthrinsing with 10 ml of an aqueous solution
of chlorhexidine (1.0 or 0.1 mg/ml) for 1 min and difference between dietary abstinence and food ingestion
Chlorhexidine in saliva, g/ml
Time after conditions for saliva collection
mouthrinsing, h
rinsing with 1.0 mg/ml rinsing with 0.1 mg/ml rinsing with 1.0 mg/ml
abstinence abstinence food ingestion 2.25 h after rinsing
(n = 20) (n = 9) (n = 10)
0.25 21.56B4.00 3.95B0.29 21.18B3.29
0.5 17.32B2.68 2.48B0.23 16.78B2.56
1 12.42B1.01 1.95B0.21 11.59B1.14
1.5 8.42B2.03 2.04B0.08 8.53B1.63
2 8.84B1.61 1.94B0.32 8.47B1.31
4 6.63B1.31 0.98B0.13 1.80B0.25*
6 2.50B0.24 0.22B0.01 0.75B0.05*
8 1.97B0.38 0.48B0.10 0.60B0.07*
Statistically significant from abstinence: *pd0.01.
D
o
w
n
lo
a
d
e
d

b
y
:

U
n
iv
e
r
s
id
a
d
e

F
e
d
e
r
a
l
d
e

M
in
a
s


























1
5
0
.
1
6
4
.
1
1
0
.
1
2
7

-

7
/
2
3
/
2
0
1
4

2
:
3
4
:
5
0

P
M
washed cells more significantly reduced the concentrations
to 5.20B0.08, 7.95B0.06 and 8.01B0.09 g/ml (pd0.01 for
all subjects, vs. washed cells).
Adsorption of Chlorhexidine to Proteins
By mixing with mucin or albumin in artificial saliva,
chlorhexidine in ultrafiltrates decreased with increasing
both protein concentrations (fig. 4). Mucin showed the
greater reduction in chlorhexidine concentrations than albu-
min (pd0.01 for all concentrations over 0.5 mg/ml), indi-
cating that chlorhexidine is more easily adsorbed to mucin
than to albumin.
Discussion
Chlorhexidine has been added to various delivery vehi-
cles such as mouthrinses, toothpastes, varnishes, dental
floss, etc. for the prevention of plaque formation [Gjermo,
1989; Huizinga et al., 1990]. Since the content of chlorhex-
idine in saliva is important for its practical effect [Bonesvoll
et al., 1974; de Vries et al., 1991], an oral pharmacokinetic
study needs the quantitation of salivary chlorhexidine after
oral application. In the present study, an HPLC method has
been developed by optimizing different analytical condi-
tions. The proposed method has allowed the quantitative
analysis selective for chlorhexidine in saliva by employing
the reversed-phase ion-pair chromatographic system.
Chlorhexidine is a dicationic compound with pKa of 2.2
and 10.3 [Lam et al., 1993], therefore it is primarily ionized
in the mobile phase to form the ion-pair with anionic alkyl
sulfonates. Chlorhexidine and 1,3-diphenylguanidine have
been revealed to increase their capacity factors with in-
creasing alkyl chain lengths and concentrations of the coun-
terions (fig. 1). Such changes in chromatographic behavior
indicate that both cationic analytes in an ion-pair form are
specifically retained onto the solid phase. Salivary compo-
nents not forming the ion-pair are eluted from an HPLC col-
umn earlier than two analytes, achieving the separative
analysis of salivary chlorhexidine (fig. 2).
Previous analytical methods for chlorhexidine used UV
spectroscopy based on its absorption at 255 nm [Jensen and
Christensen, 1971] and fluorometry based on its complex
formation with eosin [de Vries et al., 1991]. However, the
UVspectroscopic method is not applicable to chlorhexidine
in saliva because salivary proteins also absorb in that UV
region. Although the fluorometric method was applied to
aqueous solutions, centrifuged saliva and whole saliva, not
only slopes of the calibration graphs vary among samples
but also their linearity fails at lower concentrations of
chlorhexidine in saliva samples, indicating that the fluoro-
metric method is not suitable for the reliable quantitation of
chlorhexidine in whole saliva. The fluorescent response al-
so significantly varies depending on the time of saliva col-
lection. These large variations suggest the influence of sali-
vary components on the quantitation. In contrast, the
present separative method is not different in the quantitative
response between saliva and aqueous solutions. Its speci-
ficity to chlorhexidine is also superior to UV spectroscopic
[Jensen and Christensen, 1971] and fluorometric methods
[de Vries et al., 1991] as seen in chromatographic results
(fig. 2), with which neither salivary components nor inges-
tion of food and beverage interfere. While the quantitative
range is only microgram per milliliter levels for these meth-
ods, the proposed method allows the quantitation for
chlorhexidine of 50.0 ng/ml to 50.0 g/ml. The qualitative
information on identification of salivary chlorhexidine is al-
so available from the diode array spectra and retention time.
Concerning the quantitation of chlorhexidine, the UV spec-
troscopic method is confined to the direct analysis of aque-
ous solutions. The fluorometric method is suitable for aque-
ous solutions, but only for a part of saliva samples.
Applicability of the present HPLC method is independent
of the type of saliva samples.
Two HPLC methods have recently been applied to the
analysis of salivary chlorhexidine [Lam et al., 1993; Peso-
nen et al., 1995]. The present HPLC method is comparable
in quantitative sensitivity to the former and is more sensi-
tive than the latter. The recovery (96.097.8%) of chlorhex-
idine from saliva samples is superior to those of both meth-
ods. The use of an internal standard, 1,3-diphenylguanidine,
has provided higher analytical precision (intra- and interas-
say CVs ranging from 0.22 to 0.41%) than the previous
HPLC methods (ranging from 1.7 to 5.8% or from 4.7 to
9.0%) without using any internal standards. All procedures
including pretreatment and HPLC separation are completed
in a short time, which is substantially the same as that need-
ed for a batch method [de Vries et al., 1991]. The proposed
method is effectively applicable to multisample routine
analysis for all types of samples.
Its application to mouthrinsing experiments has revealed
that the significant amounts of chlorhexidine continue to be
present in saliva for a long time after mouthrinsing (table 1).
When abstaining from beverage and food, the concentra-
tions of salivary chlorhexidine have maintained microgram
per milliliter levels up to 8 h after mouthrinsing with
chlorhexidine of 0.11.0 mg/ml. Such a long-lasting reten-
tion in the oral cavity agrees with the previous results of
oral application of chlorhexidine [Jensen and Christensen,
1971; Bonesvoll et al., 1974; Pesonen et al., 1995].
HPLC Analysis of Salivary Chlorhexidine
Caries Res 1999;33:156163
161
D
o
w
n
lo
a
d
e
d

b
y
:

U
n
iv
e
r
s
id
a
d
e

F
e
d
e
r
a
l
d
e

M
in
a
s


























1
5
0
.
1
6
4
.
1
1
0
.
1
2
7

-

7
/
2
3
/
2
0
1
4

2
:
3
4
:
5
0

P
M
162
Caries Res 1999;33:156163
Tsuchiya/Miyazaki/Ohmoto
While a number of factors affect the retention of
chlorhexidine in the oral cavity [Walton et al., 1994], food
ingestion has significantly decreased salivary chlorhexidine
(table 1). However, the decreasing effects of beverage in-
gestion seem to depend on the pH of ingested beverages
(fig. 3). The concentrations of salivary chlorhexidine are re-
duced by drinking the acidic beverage, orange juice, where-
as those are not different from abstinence when drinking the
neutral beverage, green tea. The pH in the oral cavity influ-
ences the concentrations of salivary chlorhexidine and the
lower pH reduces the retention [Walton et al., 1994]. Adif-
ference in retention between orange juice and green tea in-
gestion suggests that the negatively charged binding sites
for chlorhexidine are less available in the acidic oral envi-
ronment, although the possibility that the acidic stimulation
by orange juice enhances the salivary secretion to dilute
chlorhexidine in saliva immediately after ingestion is not
excluded.
The antibacterial activity of chlorhexidine is primarily
responsible for its antiplaque effect [Jenkins et al., 1988;
Walton et al., 1994]. The minimum inhibitory concentra-
tions of chlorhexidine to cariogenic bacteria are estimated
as 0.86.3 g/ml for mutans streptococci, d0.40.8 g/ml
for Actinomyces, 3.1 g/ml for Lactobacillus and
1.66.3 g/ml for other oral streptococci [Tsuchiya et al.,
1994]. Chlorhexidine is presumed to keep its salivary con-
centrations almost corresponding to such effective concen-
trations for at least up to 68 h after mouthrinsing with
chlorhexidine of 0.11.0 mg/ml.
Chlorhexidine remains in saliva for a longer time com-
pared with the other agents such as green tea catechins
which have been added to mouthrinses, candy, chewing
gum, food, etc. for the purpose of plaque or caries preven-
tion [Tsuchiya et al., 1997]. Its overwhelmingly retentive
property raises the question of how chlorhexidine maintains
the salivary concentrations. The affinity for different oral
sites would answer this question [Bonesvoll and Gjermo,
1978; Jenkins et al., 1988; Walton et al., 1994].
Concentration changes of chlorhexidine were deter-
mined after mixing with hydroxyapatite, buccal epithelial
cells and proteins in artificial saliva, followed by centrifu-
gation or ultrafiltration. Chlorhexidine in a free form re-
mains in the obtained supernatants, whereas the bound
chlorhexidine precipitates with hydroxyapatite. The treat-
ment with hydroxyapatite has significantly reduced the con-
centrations of chlorhexidine in supernatants, indicating the
adsorption of chlorhexidine to hydroxyapatite. Chlorhexi-
dine was mixed in the cell suspensions or protein solutions
and subjected to ultrafiltration, by which the bound
chlorhexidine remains in the retentates but the free
chlorhexidine is filtered with the constant concentrations in
ultrafiltrates and retentates [Sebille et al., 1990]. The con-
centrations of chlorhexidine in ultrafiltrates have been sig-
nificantly reduced by treating with buccal epithelial cells,
indicating that chlorhexidine is adsorbed to the cells. When
chlorhexidine was treated with saliva-coated hydroxyap-
atite and not washed buccal epithelial cells, its concentra-
tions in supernatants and ultrafiltrates were further reduced,
suggesting that the presence of proteins on pellicles and cell
surfaces is favorable for the adsorption of chlorhexidine.
Since chlorhexidine is monocationic at physiological pH,
its adsorption is considered to be due to an interaction with
negatively charged binding sites of salivary proteins [Wal-
ton et al., 1994]. Chlorhexidine was mixed in the artificial
saliva containing mucin or albumin, and then its concentra-
tions in ultrafiltrates were determined to study which sali-
vary proteins participate in the binding of chlorhexidine.
Chlorhexidine has more significantly decreased by mixing
with mucin than with albumin (fig. 4), indicating that mucin
is the better adsorbent for chlorhexidine. Considering an
acidic moiety in mucin, salivary mucins are very likely to
be responsible for the adsorption of chlorhexidine [Cohen
and Levine, 1989; Walton et al., 1994]. The major salivary
proteins are composed of mucins [Cohen and Levine,
1989], which form the coatings covering tooth enamel and
epithelial cells [Levine, 1993; Slomiany et al., 1996]. After
adsorption to the surfaces of the tooth, oral mucosa and pel-
licles during mouthrinsing, the adsorbed chlorhexidine ap-
pears to be gradually released from these reservoirs depend-
ing on lowering concentrations in the oral environment to
maintain the salivary concentrations [Jenkins et al., 1988;
Walton et al., 1994].
The present study has provided the selective and sensi-
tive method to quantify chlorhexidine in saliva and aqueous
samples, which is effectively usable for mouthrinsing and
adsorption experiments to investigate the retention in the
oral cavity and the mechanism underlying it.
Acknowledgements
This study was supported in part by a Grant-in-Aid for Scientific
Research (C) from the Ministry of Education, Science, Sports and
Culture of Japan.
D
o
w
n
lo
a
d
e
d

b
y
:

U
n
iv
e
r
s
id
a
d
e

F
e
d
e
r
a
l
d
e

M
in
a
s


























1
5
0
.
1
6
4
.
1
1
0
.
1
2
7

-

7
/
2
3
/
2
0
1
4

2
:
3
4
:
5
0

P
M
HPLC Analysis of Salivary Chlorhexidine
Caries Res 1999;33:156163
163
References
Bonesvoll P, Gjermo P: A comparison between
chlorhexidine and some quaternary ammoni-
um compounds with regard to retention, sali-
vary concentration and plaque-inhibiting effect
in the human mouth after mouth rinses. Arch
Oral Biol 1978;23:289294.
Bonesvoll P, Lkken P, Rlla G, Paus PN: Reten-
tion of chlorhexidine in the human oral cavity
after mouth rinses. Arch Oral Biol 1974;19:
209212.
Cohen RE, Levine MJ: Salivary glycoproteins; in
Tenovuo JO (ed): Human Saliva: Clinical
Chemistry and Microbiology. Boca Raton,
CRC Press, 1989, vol 1, pp 101130.
Gjermo P: Chlorhexidine and related compounds.
J Dent Res 1989;68(special issue):16021608.
Huizinga ED, Ruben J, Arends J: The effect of an
antimicrobial-containing varnish on root dem-
ineralization in situ. Caries Res 1990;24:130
132.
Jenkins S, Addy M, Wade W: The mechanism of
action of chlorhexidine: A study of plaque
growth on enamel inserts in vivo. J Clin Peri-
odontol 1988;15:415424.
Jensen JE, Christensen F: A study of the elimina-
tion of chlorhexidine from the oral cavity using
a new spectrophotometric method. J Periodon-
tal Res 1971;6:306311.
Katz S, Park KK, Stookey GK, Schemehorn BR:
Development and initial testing of a model for
in vitro formation of pit and fissure caries.
Caries Res 1986;20:424428.
Lam YWF, Chan DCN, Rodriguez SY, Lintakoon
JH, Lam T-H: Sensitive high-performance
liquid chromatographic assay for the determi-
nation of chlorhexidine in saliva. J Chromatogr
1993;612:166171.
Levine MJ: Salivary macromolecules; in Malamud
D, Tabak L (eds): Saliva as a Diagnostic Fluid.
New York, New York Academy of Sciences,
1993, pp 1116.
Malamud D, Tabak L: Guidelines for saliva
nomenclature and collection; in Malamud D,
Tabak L (eds): Saliva as a Diagnostic Fluid.
New York, New York Academy of Sciences,
1993, pp xixii.
Pesonen T, Holmalahti J, Pohjola J: Determination
of chlorhexidine in saliva using high-perfor-
mance liquid chromatography. J Chromatogr B
1995;665:222225.
Sebille B, Zini R, Madjar C-V, Thuaud N, Tille-
ment J-P: Separation procedures used to reveal
and follow drug-protein binding. J Chromatogr
1990;531:5177.
Slomiany BL, Murty VL, Piotrowski J, Slomiany
A: Salivary mucins in oral mucosal defense.
Gen Pharmacol 1996;27:761771.
Steinberg D, Kopec LK, Bowen WH: Adhesion of
Actinomyces isolates to experimental pelli-
cles. J Dent Res 1993;72:10151020.
Tsuchiya H, Sato M, Iinuma M, Yokoyama J,
Ohyama M, Tanaka T, Takase I, Namikawa I:
Inhibition of the growth of cariogenic bacteria
in vitro by plant flavanones. Experientia 1994;
50:846849.
Tsuchiya H, Sato M, Kato H, Okubo T, Juneja LR,
Kim M: Simultaneous determination of cate-
chins in human saliva by high-performance
liquid chromatography. J Chromatogr B 1997;
703:253258.
de Vries J, Ruben J, Arends J: Determination of
chlorhexidine in saliva and in aqueous solu-
tions. Caries Res 1991;25:410414.
Walton JG, Thompson JW, Seymour RA: Textbook
of Dental Pharmacology and Therapeutics, ed
2. Oxford, Oxford University Press, 1994, pp
147149.
Webb BC, Willcox MDP, Thomas CJ, Harty DWS,
Knox KW: The effect of sodium hypochlorite
on potential pathogenic traits of Candida albi-
cans and other Candida species. Oral Microbi-
ol Immunol 1995;10:334341.
D
o
w
n
lo
a
d
e
d

b
y
:

U
n
iv
e
r
s
id
a
d
e

F
e
d
e
r
a
l
d
e

M
in
a
s


























1
5
0
.
1
6
4
.
1
1
0
.
1
2
7

-

7
/
2
3
/
2
0
1
4

2
:
3
4
:
5
0

P
M