Preservation of Liquid Drug Preparations for
Oral Administration
STEFAN SCHELER, SEBASTIAN SAUPE, ANGELA HERRE, ALFRED FAHR
Department of Pharmaceutical Technology, Friedrich Schiller University of Jena, Jena, Germany
Received 8 December 2008; revised 19 April 2009; accepted 23 April 2009
Published online 4 June 2009 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.21828
ABSTRACT: The aim of this study was to investigate functional interactions between
sorbic acid, alcoholic cosolvents, and pH with respect to pharmacopeial requirements for
the antimicrobial preservation of oral liquids. Twenty-seven test formulations of sugar-
free syrups with varying amounts of glycerol (18–36%), propylene glycol (4–21%), and
sorbic acid (0-0.15%) and with different pH values (5–8) were tested for antimicrobial
preservation efficacy after inoculation with spores of
Aspergillus niger. Multivariate
data analysis revealed that at pH 5 the minimum concentration of sorbic acid necessary
for a tenfold decrease of viable spores ranges between 0.08% and 0.10% exhibiting only
minor dependence on the cosolvents concentration. Various interactions between sorbic
acid and cosolvents could be observed and were discussed on basis of the degree of
dissociation and distribution of sorbic acid. All tested preparations, even those free of
sorbic acid, met the criteria for oral products with aqueous bases according to USP32-
NF27 and JP XV which claim no increase from the initial spore count at 14 and 28 days.
The EP requirement, not less than 1 log
10
reduction from the initial count at 14 days, was
only met by preparations with pH 5 and not less than 0.15% sorbic acid.
ß
2009 Wiley-Liss,
Inc. and the American Pharmacists Association J Pharm Sci 99:357–367, 2010
Keywords:
Aspergillus niger; chromatography; excipients; factorial design; formula-
tion; glycerol; preservation; propylene glycol; sorbic acid; stabilization
INTRODUCTION
Liquid preparations for oral use are commonly
supplied in multiple dose bottles. If such aqueous
preparations do not have antimicrobial properties
by themselves, preservation is necessary to prevent
the propagation of microorganisms and to limit
the risk of product contamination during usage.
Maximum limits for the counts of viable micro-
organisms per gram of the product are defined in
all main pharmacopeias. During the development
process of a liquid drug formulation, it has to
be proven that the preparation is sufficiently
protected against microbial growth by its ingre-
dients or by supplementation with appropriate
preservatives.
The effectiveness of an antimicrobial preserva-
tion is tested by preservatives-effectiveness tests
which are specified in all main pharmacopeias
(United States Pharmacopeia (USP),
1
European
Pharmacopeia (EP),
2
Japanese Pharmacopeia
(JP)
3
). The method requires an inoculation of
the product with a specified number of colony
forming units (cfu) of a variety of microorganisms
and monitoring the counts of viable microorgan-
isms during a 28-day period. For the interpreta-
tion of the results different product categories
are distinguished and for each of them different
specifications are defined concerning bacteria on
the one hand and yeast and moulds on the other.
USP and JP consider an oral product to be
Correspondence to: Stefan Scheler (Sandoz GmbH, Bio-
chemiestrasse 10, 6250 Kundl, Austria. Telephone: 43-5338-
200-2793; Fax: 43-5338-200-3238;
E-mail: stefan.scheler@sandoz.com)
Journal of Pharmaceutical Sciences, Vol. 99, 357–367 (2010)
ß
2009 Wiley-Liss, Inc. and the American Pharmacists Association
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
357
effectively preserved against yeast and moulds if
there is no increase from the initial count at
14 and 28 days. According to EP specifications the
tested sample complies with the requirements if
the microbial count is reduced at least by factor
10 within 14 days and if there is no increase
between days 14 and 28. In recent years,
investigations revealed that a considerable por-
tion of examined commercial products for oral use
did not comply with the pharmacopeial require-
ments for optimal preservation especially from
fungal contamination.
4,5
For this reason it was
investigated how typical excipients and pH of a
liquid oral preparation contribute to the anti-
microbial effect particularly in terms of obligatory
requirements.
A typical composition of sugar-free syrup was
chosen as test preparation. The concentrations of
glycerol, propylene glycol, and sorbic acid were
varied stepwise between 18% and 36%, 4% and
21%, and 0% and 0.15%, respectively, whereas
fixed concentrations of sorbitol and hydroxyethyl
cellulose were employed in the tests. For sorbitol
decreases the solubility of sorbic acid in water
(0.11% in a 15% aqueous sorbitol solution at 208C)
the use of cosolvents is necessary to enable con-
centrations of sorbic acid up to 0.15%.
6
According to USP, EP, and JP the antimicrobial
properties against specified strains of each of
the following microorganisms are to be tested:
Pseudomonas aeruginosa, Staphylococcus aureus,
Candida albicans, and Aspergillus niger. Because
of the relatively high osmolarity of syrup pre-
parations only
A. niger was included in this study.
Among all mentioned microorganisms it has the
highest resistance against osmotic pressure and
thus is viable even in high concentrated solutions.
MATERIALS AND METHODS
Materials
All ingredients of the formulations, which were
glycerol 85% (Glycerinum Ph.Eur. 4.07), propy-
lene glycol (Propylenglycolum Ph.Eur. 4.00),
sorbitol solution 70% noncrystallizing (Sorbitum
solutum 70% Ph.Eur. 4.04), hydroxyethyl cellu-
lose (Natrosol
1
250 G Pharm. Ph.Eur. 4.07, Visc
Rota 274.8 mPa s), and sorbic acid (Acidum
sorbicum, Ph.Eur. 4.00) were purchased from
Caesar & Loretz GmbH (Hilden, Germany).
Sabouraud 4% glucose agar, Na
2
HPO
4
(Ph.Eur.,
USP), acetonitrile (gradient grade, LiChrosolv
1
),
methanol (gradient grade, LiChrosolv
1
), and
acetic acid (100%, p.a.) are from Merck KGaA
(Darmstadt, Germany). NaCl (>99.8%, KH
2
PO
4
(
99%, p.a.), casein peptone (peptone ex casein
tryptic digest), NaOH (
99%, p.a.) was from
Carl Roth GmbH
þCo (Karlsruhe, Germany).
Membrane filtration for determining the number
of viable spores was performed using a Milliflex
100 system (Millipore, Billerica, MA) with Milli-
flex Filter Funnel Units (100 mL, sterile, 0.45 mm
Durapore membrane) which were mounted on a
Milliflex Sensor II Twin-Head Pump for vacuum
filtration. Cultivation was performed after cou-
pling the filter units to solid media cassettes
(empty cassettes, ready-to-fill with culture med-
ium) which were filled with Sabouraud 4% glucose
agar (
¼agar medium C EP).
Preparation of the Test Formulations
Twenty-seven formulations were tested which
differed in glycerol/propylene glycol concentra-
tion, sorbic acid content, and pH according to
Table 1. The compounds were dissolved in purified
water and the appropriate pH of each buffered
solution was fine adjusted with NaOH. None of the
ingredients did show any microbial growth on
agar medium B and C (EP). Nevertheless, all
solutions were filtered through 0.2 mm mem-
branes and met the ‘‘test on sterility’’ according
to EP. The test formulation was dispensed into
injection vials under aseptic conditions. Samples
of 30 mL were prepared for the microbiological
tests and 100 mL samples were used for all experi-
ments in which additionally the concentration of
sorbic acid was determined.
Preparation of Spore Suspensions and
Determination of Their Concentrations
Spores of
A. niger, ATCC 16404 (purchased from
DSMZ, Braunschweig, Germany) were re-culti-
vated in potato glucose agar (DSMZ: Medium
129), portion wise transferred to a Cryobank
TM
System (Mast Diagnostica, Reinfeld, Germany)
and stored at
808C until further use.
For the preparation of the spore suspensions
cryoconservated
A. niger spores were grown 7 days
on Sabouraud 4% glucose agar. Ten milliliters of
sterile NaCl–Polysorbate 80 solution (Ritabate-80,
Rita Corporation, Crystal Lake, IL) was spread on
the surface of each agar plate. After 15 min, 7 mL
of the spore suspension was harvested from each
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
DOI 10.1002/jps
358
SCHELER ET AL.
Petri dish. Fifteen microliters of the suspension
were concentrated to a volume of 6 mL by
centrifugation (8000 rpm, 48C, 15 min). From this
stock suspension a series of dilutions was pre-
pared (10
1
, 5
10
2
, 10
2
, 5
10
3
, 10
3
, 10
4
,
10
5
, 10
6
). The apparent extinctions of the
dilutions 10
1
to 10
3
were determined at
660 nm (Beckman DU640 Spectrophotometer).
It turned out that the suspensions of dilution
step 10
5
have the optimum spore concentration
for counting the number of spores per mL after
growth on agar plates. One milliliter of this
suspension was spread on agar plates in triplicate
and after 2 days incubation at 238C the number of
star-like mycels was counted. From the average of
counts per mL of the 10
5
dilution the spore con-
centrations of the other dilution steps were calcu-
lated. A linear correlation between the apparent
extinction and the spore concentration could
be calculated with a correlation coefficient of
0.997.
Inoculation of the Test Formulations
A 1:100 dilution of the stock suspension was
prepared
and
the
spore
concentration
was
determined turbidimetrically. A concentration of
8
10
5
spores/mL was calculated from the appar-
ent extinction. This corresponds to 8
10
7
spores/
mL in the undiluted stock suspension which meets
the EP specification of about 10
8
spores/mL.
The test formulations were inoculated with 0.3
(30 mL samples) or 1.0 mL (100 mL samples) of the
spore suspension, respectively. Immediately after
Table 1. Composition of the Test Formulations
No.
Sorbitol
Sol. 70%
(Noncryst.)
% (w/v)
a
Hydroxyethyl
Cellulose
% (w/v)
Glycerol
% (w/v)
b
Propylene
Glycol %
(w/v)
b
Sorbic
Acid %
(w/v)
c
pH (Adjusted
With 0.1 M
Phosphate Buffer)
1
27.6
0.1
18.7
4.2
0
5
2
27.6
0.1
18.7
4.2
0
7
3
27.6
0.1
18.7
4.2
0
8
4
27.6
0.1
18.7
4.2
0.05
5
5
27.6
0.1
18.7
4.2
0.05
7
6
27.6
0.1
18.7
4.2
0.05
8
7
27.6
0.1
18.7
4.2
0.15
5
8
27.6
0.1
18.7
4.2
0.15
7
9
27.6
0.1
18.7
4.2
0.15
8
10
27.6
0.1
28.0
12.4
0
5
11
27.6
0.1
28.0
12.4
0
7
12
27.6
0.1
28.0
12.4
0
8
13
27.6
0.1
28.0
12.4
0.05
5
14
27.6
0.1
28.0
12.4
0.05
7
15
27.6
0.1
28.0
12.4
0.05
8
16
27.6
0.1
28.0
12.4
0.15
5
17
27.6
0.1
28.0
12.4
0.15
7
18
27.6
0.1
28.0
12.4
0.15
8
19
27.6
0.1
36.4
20.7
0
5
20
27.6
0.1
36.4
20.7
0
7
21
27.6
0.1
36.4
20.7
0
8
22
27.6
0.1
36.4
20.7
0.05
5
23
27.6
0.1
36.4
20.7
0.05
7
24
27.6
0.1
36.4
20.7
0.05
8
25
27.6
0.1
36.4
20.7
0.15
5
26
27.6
0.1
36.4
20.7
0.15
7
27
27.6
0.1
36.4
20.7
0.15
8
a
27.6% (w/v) Sorbitol sol. 70% non cryst.
¼ 15% (w/v) with respect to the content of solids.
b
Glycerol/propylene glycol 18.7/4.2, 28.0/12.4, 36.4/20.7% (w/v)
¼ Glycerol 85%/propylene glycol 18/4, 27/12, 35/20% (v/v).
c
0.05% (w/v)
¼ 4.5 mmol/l, 0.15% (w/v) ¼ 13.4 mmol/l.
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
PRESERVATION OF LIQUID DRUG PREPARATIONS FOR ORAL ADMINISTRATION
359
the inoculation, after 14 days, and after 28 days
three samples of 1 mL were drawn from each
vial and were diluted with a NaCl–Polysorbate
80–Peptone buffer solution in 1:10 steps until a
spore concentration of lower than 100 per mL
could be expected. One milliliter of this dilution
was mixed with 25 mL peptone buffer solution and
filtered through a Milliflex 100 unit. To remove
residues of the test formulation the membrane
was washed three times with 100 mL NaCl–
Polysorbate 80–peptone buffer solution. The
membranes were transferred to the solid media
cassettes with Sabouraud 4% glucose agar and
incubated at 238C. The grown mycels could be
counted after 2 days. If low counts were observed
the number of dilution steps was reduced at
the respective sample time. After 28 days the
pH value of each vial was measured. No changes
could be detected during the incubation.
Determination of the Sorbic Acid Concentration
In order to prove that the concentration of sorbic
acid remains constant during the incubation
period, the chemical stability of sorbic acid was
determined under different storage conditions
of the formulations. These tests were performed
with formulations containing medium glycerol/
propylene glycol concentrations (28/12%), 0.05%
sorbic acid, and pH values of 5, 7, and 8,
respectively. In one series of samples the vials
were stored upright whereas those of another
series were stored upside down to investigate
possible adsorption effects of the stopper. In
a third experiment upright stored vials were
inoculated with spores for studying possible
decomposition or adsorption caused by the spores.
Sorbic acid was determined by HPLC with a
method described by Radus and Gyr.
7
Column:
Beckman Ultrasphere ODS, 5 mm spherical 8 nm
pore, 4.6 cm
15 cm, mobile phase: 0.01 M potas-
sium acetate in water/acetonitrile 70:30 (v/v), pH
was adjusted to 5.4 with acetic acid, flow rate:
2 mL/min, sample preparation: 1 mL sample was
diluted with mobile phase to 25 mL, injection
volume: 10 mL, detection: UV absorption at
256 nm.
The effect of
A. niger spores on the sorbic acid
concentration was tested at three different
pH values with the formulations 13, 14, and 15
(glycerol 28%, propylene glycol 12%, sorbic acid
0.05%, pH 5, 7, 8). One hundred milliliters of
each formulation was inoculated with 1 mL stock
suspension in triplicate. The glass bottles were
closed with rubber stoppers and aluminum caps
and were stored upright at 238C. Immediately
after inoculation, after 14 and after 28 days two
samples were drawn from each bottle for the
determination of the spore concentration and one
sample (1.5 mL) for the measurement of the sorbic
acid concentration. The samples were filtered
through a 0.22 mm membrane. The filters were
purged previously with another 2.5 mL of sample
solution avoiding losses of sorbic acid due to
adsorption to the surfaces.
Formulations 13–15 were also used to deter-
mine possible adsorption of sorbic acid to the glass
vials or absorption by the rubber stoppers.
One hundred milliliters of each formulation was
incubated in closed glass bottles in triplicate as
described before but without inoculation with
spore suspension. Three bottles of each pH were
stored upright and three bottles upside down.
Samples were withdrawn and analyzed as men-
tioned before.
In all experiments the sorbic acid concentrations
show a slightly decreasing trend with an average
drop of about 2% within 28 days. Because the
decrease is about the same in the upright and
the upside down stored vials it can be assumed that
there is no significant absorption by the stopper
material.
A. niger spores do not cause a concentra-
tion decline of sorbic acid. With respect to the
antimicrobial efficacy of the preparations the
decrease of about 2% can be regarded as negligible.
RESULTS AND DISCUSSION
Antimicrobial Effect of Glycerol and
Propylene Glycol
Glycerol and propylene glycol are known to have
antimicrobial effects at higher concentrations. It is
reported that the growth of
A. niger is inhibited by
glycerol concentrations larger than 50% and by
propylene glycol concentrations larger than 10%.
8
Other authors describe a fungistatic effect at a
propylene glycol concentration of 30% and a
fungicidal effect at
40%.
9
Due to a synergistic
effect the minimum concentration of glycerol
required for an antimicrobial effect is decreased
in mixtures with more than 20% sorbitol. By
contrast the antimicrobial activity of propylene
glycol is not enhanced by addition of up to 70%
sorbitol.
8
This indicates an osmotic effect of glycerol.
In a preliminary experiment the antimicrobial
effect of glycerol 85% and undiluted propylene
glycol upon
A. niger spores was tested. As Table 2
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
DOI 10.1002/jps
360
SCHELER ET AL.
shows glycerol 85% reduces the spore concentra-
tion by 1.5 log
10
steps after 14 days and by more
than 3 log
10
steps after 28 days. In propylene
glycol, the amount of viable spores is decreased
by more than 2.2 log
10
steps after 14 days and
more than 6 log
10
steps after 28 days, indicating a
more potent antimicrobial activity in comparison
to glycerol. This is in accordance with Herman
et al.
10
who described a relationship between
molecular asymmetry of short chain alcohols and
glycols and their antimicrobial activity.
Antimicrobial Effects of the Test Formulations
To evaluate the effect of glycerol and propylene
glycol in combination with other excipients of
liquid preparations for oral use, a typical for-
mulation was designed and modified with respect
to the glycerol/propylene glycol concentration, the
concentration of an additional preservative (sorbic
acid), and the pH value. The spore count of each
test formulation was determined immediately
after inoculation, after 14 days, and after 28 days.
Each result was extrapolated to the initial volume
in order to compensate the volume loss caused by
the withdrawal of the samples. The spore con-
centrations were calculated as the average of
three samples. The standard deviations of the
percental spore count reductions were calculated
according to the rules for the propagation of
random errors. If no spores could be detected in
the analyzed dilution the spore count was quoted
to be smaller than the value calculated on the
basis of 1 spore per mL in the diluted sample.
Immediately after inoculation an average spore
concentration of 1.3
10
6
per mL was measured.
Table 3 shows that without the addition of preser-
vatives neither glycerol and propylene glycol up to
36% and 21%, respectively, nor the osmotic effect
of about 15% sorbitol (
¼21.4% of sorbitol solution
70%) has a sufficient antimicrobial activity to
reduce the initial spore concentration by the factor
10 within 14 days.
Amongst the unpreserved formulations only the
alkaline solution with the highest glycerol/propy-
lene glycol content is able to reduce the spore
concentration by 1 log step within 28 days. Even
with an addition of 0.05% sorbic acid a tenfold
reduction of viable spores within 14 days cannot
be achieved. However, in interaction with high
glycerol/propylene glycol concentrations (36%/
21%) preservation with 0.05% sorbic acid reduces
the spore concentration by at least 1 log step after
28 days at all tested pH values. At pH 5 a more
than tenfold reduction (28 days) can be obtained
even with medium glycerol/propylene glycol con-
centrations (28%/12%) in the presence of 0.05%
sorbic acid. With 0.15% sorbic acid all acidic
formulations (pH 5) showed a more than tenfold
spore reduction after 14 days irrespectively of
their glycerol/propylene glycol concentration and
the reduction exceeded at least one further log
step after 28 days. The formulation with high
glycerol/propylene glycol concentrations (36%/
21%) and 0.15% sorbic acid decreased the spore
concentration by about 1 log step after 14 days
even at pH 8.
These findings are in accordance with the
reported
minimum
inhibitory
concentration
(MIC) of sorbic acid needed to completely inhibit
the germination of
A. niger spores. The MIC is
described to depend on the spore concentration
and amounts to 4.5 mmol/L (0.05%) for 10
5
spores
per mL at pH 4.0.
11
This value can also be
interpolated from data given by Razavi-Rohani
and Griffiths
12
who determined the MIC in
dependence of the pH but used a mycelial
inoculum. For pH values of 3.5, 5.6, and 7.0 they
reported MICs of 2.1, 12.0, and >28.3 mmol/L
(0.02, 0.13, >0.32%), respectively. These preser-
vative concentrations necessary for growth inhi-
bition were compared with those required for
inactivation of
A. niger spores. The response
surface method described below was helpful to
interpolate the spore count reduction for each pH
and each sorbic acid concentration. It can be
concluded that the minimal inhibitory concentra-
tions reported for different pH values correspond
to a 14 days spore count reduction of about
1–1.5 log
10
steps in case of the tested formulations.
After a storage period of 28 days the correspond-
ing spore count reduction rates were calculated to
about 2–4 log
10
steps.
The straight lines in the semilogarithmic
scaled diagram in Figure 1 indicate (shown
for the example of all preparations with a
sorbic acid concentration of 0.15% and pH 8)
Table 2. Reduction (Log
10
Steps) of Viable Spore
Count of
A. niger After 14 and 28 days Incubation With
Glycerol or Propylene Glycol (SD, Standard Deviation)
14 Days
28 Days
Log
10
SD Log
10
Log
10
SD Log
10
Glycerol 85%
1.480
0.253
3.230
0.192
Propylene glycol >2.222
>6.004
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
PRESERVATION OF LIQUID DRUG PREPARATIONS FOR ORAL ADMINISTRATION
361
that the inactivation kinetics follows an expo-
nential law.
For examining possible concentration changes
of sorbic acid during incubation, the experiments
with medium glycerol/propylene glycol concen-
trations (28%/12%) were repeated with 100 mL
instead of 30 mL sample volumes. Besides the
sorbic acid concentration the spore count was
also monitored. In this case six measurements
(three inoculated bottles, two samples drawn from
each vial) were averaged to obtain a value for the
spore count reduction. Table 4 shows the loga-
rithms of the spore count reduction of the 100 mL
experiments compared to the 30 mL samples. The
spore count reductions obtained from 100 mL
samples are within the same log steps as the
respective ones from 30 mL. There is no uniform
Figure 1.
Inactivation kinetics of the formulations
with 0.15% sorbic acid and pH 8.
Table 3. Reduction (Log
10
Steps) of Viable Spore Count of
A. niger in Test Formulations After 14 and 28 days
Incubation (SD, Standard Deviation)
c(Glycerol)
(%)
c(Propylene
Glycol) (%)
pH
c(Sorbic
acid) (%)
14 Days
28 Days
Log
10
SD Log
10
Log
10
SD Log
10
36.4
20.7
5
0
0.312
0.154
0.429
0.062
0.05
0.324
0.071
1.073
0.074
0.15
1.029
0.132
>2.096
7
0
0.227
0.097
0.851
0.027
0.05
0.311
0.125
1.191
0.129
0.15
0.603
0.046
1.494
0.078
8
0
0.395
0.091
1.788
0.215
0.05
0.923
0.059
2.099
0.189
0.15
0.988
0.060
1.816
0.061
28.0
12.4
5
0
0.304
0.050
0.778
0.046
0.05
0.449
0.058
1.548
0.077
0.15
>2.221
>4.221
7
0
0.344
0.074
0.866
0.047
0.05
0.206
0.043
0.245
0.040
0.15
0.111
0.046
0.264
0.032
8
0
0.441
0.165
0.921
0.056
0.05
0.372
0.061
0.841
0.050
0.15
0.297
0.086
0.807
0.041
18.7
4.2
5
0
0.047
0.045
0.125
0.065
0.05
0.180
0.049
0.523
0.048
0.15
>2.164
>7.646
7
0
0.093
0.052
0.142
0.050
0.05
0.037
0.045
0.124
0.052
0.15
0.018
0.052
0.186
0.050
8
0
0.102
0.044
0.116
0.043
0.05
0.037
0.029
0.103
0.044
0.15
0.043
0.032
0.128
0.035
At pH 5 the preservation effectiveness increases as the concentration of sorbic acid increases. In case of pH 7 and 8 the
concentration of sorbic acid has a mixed influence on the antimicrobial effect which is positively correlated at high concentrations of
glycerol/propylene glycol and negatively correlated at medium and low concentrations of the alcohols.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
DOI 10.1002/jps
362
SCHELER ET AL.
shift towards lower or higher values observable
when the sample volume is changed by a factor of
3. With respect to the high standard deviations
the sample volume has to be considered without
any influence to the test results within the
investigated range. As expected the data prove
that sorbic acid provides its best antimicrobial
activity within an acidic environment and is only
poorly active in neutral solutions. Surprisingly
the activity increases in alkaline solutions, which
can be attributed to the presence of glycerol and
propylene glycol as discussed later.
From the raw data, representing a three-level
full factorial experimental design, a model for
interpretation and prediction purposes was calcu-
lated. A quadratic interaction model was applied
and could be improved by waiving the quadratic
terms for cosolvents and sorbic acid leading to the
following equation:
log Red
¼ a þ b cðGPÞ þ c pH þ d cðSorÞ
þ e pH
2
þ f cðGPÞ pH
þ g cðGPÞ cðSorÞ þ h pH cðSorÞ
(1)
with log Red being the logarithm of the spore
count reduction,
c(GP) the concentration of
glycerol
þ propylene glycol, pH the pH value,
c(Sor) the concentration of sorbic acid, and a, b,
c, d, e, f, g, and h are coefficients.
Contour plots which were calculated from this
equation illustrate that at pH 5 the antimicrobial
activity depends nearly exclusively on the con-
centration of sorbic acid whereas at pH 8 it is
correlated mainly to the cosolvents concentration
(Fig. 2). The contour plots show that at pH 8 the
preservation effectiveness of sorbic acid increases
as the concentration of the alcohols increases,
while at pH 5 there is a slight trend towards
a lower effectiveness with increasing alcohol
concentrations.
Although
the
main
correlations
can
be
derived from this model it has to be pointed out
that the extraordinarily high values obtained
at pH 5 and 0.15% sorbic acid could not be
fitted properly by this equation even if the
c(Sor)
2
term is retained. This means inclusion
of these values in a fitting procedure over- and
understates some other values to an unacceptable
degree.
For this reason a modified Shepard’s method
was applied as a second model which acts as an
exact interpolator.
13
This method is based on an
inverse distance-weighted least-squares inter-
polation following the principle that each data
point within a certain radius has some local
influence on the value at the prediction location
that diminishes with its distance. The results
were visualized by displaying the spore count
reduction versus all three different factors of
influence. This was achieved by using a three-
dimensional diagram with the concentration of
glycerol
þ propylene glycol, the concentration of
sorbic acid, and the pH value assigned to the
X,
Y, and Z axis, respectively. In the diagram of
Figure 3a all positions representing those
experimental settings which lead to a spore
count reduction equal or larger than 2 log
10
steps
within 14 days are enclosed by a dark gray
colored isosurface. This spacial region of strong
antimicrobial effectiveness is surrounded by
another light gray colored isosurface in a onion
skin-like manner which delimits the region
where the spore count reduction is only at least
1 log
10
step. The latter mentioned region is also
shown in Figure 3b from a front perspective. The
interpolated density distribution was calculated
Table 4. Spore Count Reduction in 30 mL Versus 100 mL Sample Volume (Sorbic Acid:
0.05, Glycerol: 27%, Propylene Glycol: 12%, Initial Spore Count: About 1
10
6
mL
1
)
(SD, Standard Deviation)
Sample
Volume (mL)
pH
14 Days
28 Days
Log
10
SD Log
10
Log
10
SD Log
10
30
5
0.449
0.058
1.548
0.077
100
0.647
0.079
2.164
0.226
30
7
0.206
0.043
0.245
0.040
100
0.070
0.044
0.133
0.044
30
8
0.372
0.061
0.841
0.050
100
0.094
0.046
0.560
0.081
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
PRESERVATION OF LIQUID DRUG PREPARATIONS FOR ORAL ADMINISTRATION
363
by data gridding and subsequent interval
volume tetrahedrization using the software
package IDL 6.0 (Research Systems, Inc.,
Boulder, CO).
The diagram shows that a tenfold or larger
decrease of the spore count within 14 days cannot
be achieved without any preservatives. If the
sorbic acid concentration exceeds 0.05% two
different areas of activity occur. One of them
is at low pH values and low concentrations of
glycerol and propylene glycol and the other is at
high pH values and high concentrations of the
alcohols. At low glycerol/propylene glycol concen-
trations sorbic acid develops a maximum anti-
microbial activity below pH 6. If the glycerol/
propylene glycol fraction of the solution exceeds
40% a second antimicrobial optimum appears
within the alkaline range (pH > 7) whereas the
antimicrobial effect at low pH values declines.
The data also reveal an antimicrobial activity of
sorbic acid-free preparations which increases with
rising cosolvent concentration. There is a clear
pH dependence with higher pH values enhancing
the antimicrobial effect. This correlation becomes
superimposed by an inversed influence of pH
in formulations with increased concentration of
sorbic acid due to its maximum activity in acidic
media.
In neutral and alkaline media, the effect of
sorbic acid is increased by glycerol/propylene
glycol. Such an enhancement of the antimicrobial
activity of various preservatives and antimicro-
Figure 3. (a,b) Isosurface plots of spore count reduc-
tion. (a) All positions representing experimental set-
tings causing a spore count reduction equal or larger
than two log
10
steps within 14 days enclosed by a dark
gray colored isosurface (surface of constant data value).
The light gray colored isosurface delimits the region
where the spore count reduction is only at least 1 log
10
step. (b) The latter mentioned region from a front
perspective.
Figure 2.
(a) RSM contour plot illustrating the viable
spore count reduction of
A. niger after 14 days (log
10
steps) in dependence of the concentrations of glycer-
ol
þ propylene glycol (alcohols) and sorbic acid at pH 5.
(b) RSM contour plot illustrating the viable spore count
reduction of
A. niger after 14 days (log
10
steps) in depen-
dence of the concentrations of glycerol
þ propylene
glycol (alcohols) and sorbic acid at pH 8.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
DOI 10.1002/jps
364
SCHELER ET AL.
bials by cosolvents is described in literature.
14–16
In case of sorbic acid this correlation vanishes
at pH 5.
It is well known that the lipophilic, uncharged,
and undissociated molecules of sorbic acid which
predominate at lower pH values (p
K
a
¼ 4.75)
permeate the lipid layers of the cell membranes
much better than negatively charged sorbate ions.
In the environment of the cell cytosol which has
a pH of about 7.6 in case of
A. niger dissociation
takes place and the charged anions are not able to
diffuse back across the membrane.
17
The result
of this ion capturing process is an intracellular
acidification which inhibits key metabolic activ-
ities and activates protecting mechanisms like the
membrane H
þ
-ATPase. This mechanism leads to
an increased ATP consumption and results in an
energy depletion of the cells.
11,18,19
According to Eklund
20
the relative inhibitory
efficiency between undissociated and dissociated
sorbic acid can be calculated by fitting the
following mathematical model to the experimen-
tal MIC values at different pH:
MIC
t
¼
1
ð1 aÞ=k
1
þ a=k
2
(2)
with MIC
t
being the MIC for total sorbic acid
(dissociated
þ undissociated) and a is the ratio
between undissociated and total acid
a
¼
1
10
pH
pK
a
þ 1
(3)
where
k
1
is the MIC for dissociated sorbic acid,
and
k
2
the MIC for undissociated sorbic acid.
Fitting the model to the MIC values cited above
obtains
k
1
¼ 41.2 mmol/L and k
2
¼ 2.0 mmol/L
indicating that undissociated sorbic acid has an
about 20 times higher antimicrobial activity
against
A. niger than the sorbate ion. This value
is in the same range as those reported for yeasts
which are
k
1
/
k
2
¼ 10 and 30 for C. albicans and
Saccharomyces cerevisiae, respectively.
20,21
It is reported that the dissociation of sorbic acid
depends on the amount of cosolvents present in
the solution. According to Pethybridge in aqueous
glycerol solutions the p
K
a
value of sorbic acid can
be estimated by the following equation
22
:
p
K
a
¼ 4:680 þ 0:0123M 2:66 10
4
M
2
(4)
with
M being % (w/w) glycerol within the mixture.
On the basis of this equation p
K
a
values of
4.82 and 4.78 were calculated for the lowest and
highest glycerol concentration of the tested
solutions (18% and 36%). This means that by
varying the glycerol concentration between the
lower and the higher levels the fraction of
undissociated sorbic acid changes only by 2% or
less, depending on the pH. It is unlikely that
propylene glycol has a stronger impact. Consider-
ing this calculation it can be concluded that the
cosolvents do not change the antimicrobial activ-
ity of sorbic acid by influencing their dissociation.
Nevertheless, the experimental results show an
interaction between the concentration of sorbic
acid and the cosolvents, which can be concluded
also from the finding that in the factorial analysis
the coefficient for this two-factor interaction is not
insignificant. At pH 7 and 8 a change of the sorbic
acid concentration has an opposite effect on the
spore count reduction depending on whether the
cosolvent concentration is high or low. At 36%/
25% glycerol/propylene glycol the 14 days spore
count reduction is enhanced by rising amounts of
sorbic acid, but surprisingly the antimicrobial
effect of the cosolvents is reduced by sorbic acid
at glycerol/propylene glycol concentrations equal
or lower than 28%/12%. In contrast the 28 days
value is mainly unaffected. This indicates that
the antimicrobial effect of glycerol (
28%) or
propylene glycol (
12%) is decelerated by sorbic
acid.
At pH 5 the effect of sorbic acid is enhanced by
cosolvent concentrations up to medium levels
(28%/12%) but attenuated if the concentrations
are further increased to high levels (36%/25%). A
potentiating effect of an alcohol to the antimicro-
bial action of a weak acid has been also been
reported for ethanol and acetic acid with respect to
metabolic inhibition and internal acidification in
S. cerevisiae.
23
Since it has been shown that the
cosolvents do not alter the dissociation of sorbic
acid significantly, the attenuation observed at
higher concentrations might be explained by an
increasing lipophilicity of the incubation medium
which shifts the distribution equilibrium of
undissociated sorbic acid towards the incubation
medium thus impeding its permeation into the
spores.
24
CONCLUSION
Liquid preparations for oral use typically contain
excipients like sorbitol, glycerol, and propylene
glycol acting as sweeteners, cosolvents, or viscosity
enhancers but generating also a certain antimi-
crobial activity. Our investigation reveals that
combinations of these excipients with concentra-
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
PRESERVATION OF LIQUID DRUG PREPARATIONS FOR ORAL ADMINISTRATION
365
tions up to 15% sorbitol, 36% glycerol, and 21%
propylene glycol are not able to reduce the count of
A. niger spores by the factor 10 within 14 days.
A supplementary preservative is necessary to
protect those preparations against microbial
contamination if they are supplied in multi-dose
bottles. Sorbic acid, one of the most commonly
used preservatives for oral preparations, exhibits
maximum activity in acidic media. It was found
that all acidic formulations (pH 5) with 0.15% but
no preparation with 0.05% sorbic acid decreased
the spore count by the factor 10 or more within
14 days. By mathematical interpolation it was
found that the concentration limit necessary for a
tenfold reduction of viable spores ranges between
0.08% and 0.10% with only a slight dependence
from the cosolvents concentration. This corre-
sponds very well to the minimal inhibitory
concentration reported for this pH. High concen-
trations of glycerol and propylene glycol were
found to increase the antimicrobial activity of
sorbic acid in alkaline media, reaching a nearly
tenfold decrease of the initial spore count at pH 8
with 36% glycerol, 21% propylene glycol, and
0.15% sorbic acid. In all cases in which the spore
count reduction exceeds 1 log step after 14 days a
further reduction of more than 1 log step was
achieved after 28 days. All tested preparations,
even those free of sorbic acid, met the criteria for
oral products with aqueous bases according to
USP27 and JP XV which claim no increase from
the initial count at 14 and 28 days. In distinction,
due to EP 6. ed. 2008 specifications not less than
1 log
10
reduction from the initial count has to
be achieved at 14 days and no increase from the
14 days count must occur at 28 days. This criterion
is only met by preparations with pH 5 and not less
than 0.15% sorbic acid. The study shows that
current USP and JP requirements for oral
preparations can easily be fulfilled by formula-
tions with a specific amount of cosolvents like 18%
glycerol and 4% propylene glycol whereas the
more strict specifications of the EP are only met by
application of a supplementary preservative.
ACKNOWLEDGMENTS
This work was funded by the German Federal
Institute for Drugs and Medical Devices, Bonn,
Germany (independent higher federal authority
within the portfolio of the Federal Ministry of
Health).
REFERENCES
1. United States Pharmacopeia USP 32-NF 27, 2008,
General Chapters: <51> Antimicrobial effective-
ness testing. U.S. Pharmacopeia, Rockville, Mary-
land, USA.
2. Pharmacopoeia European, 6th edition, 2008 (6.4),
5.1.3 Efficacy of antimicrobial preservation, EDQM—
European Directorate for the quality of medicines,
Strasbourg, France.
3. The Japanese Pharmacopeia, General Information:
19. Preservatives-Effectiveness Tests. 15th edition.
Society of Japanese Pharmacopoeia, Tokyo, Japan.
4. Na’was TE, Salem MS, Alkaysi HN. 1990. Microbial
contamination and preservation efficacy of cough
preparations. J Clin Pharm Ther 15:365–369.
5. Van Doorne H, Scheffers WA, Hadiutomo M,
Van den Bosch E. 1984. Microbial contamination
of a vitamin A formulation, prepared in local
pharmacies, and its preservation against yeasts
and moulds. Antonie Van Leeuwenhoek 50:405–
416.
6. Shihab FA, Ezzedeen FW, Stohs SJ. 1988. Effect of
some syrup constituents on the solubility of sorbic
acid. J Pharm Sci 77:455–457.
7. Radus TP, Gyr G. 1983. Determination of antimi-
crobial preservatives in pharmaceutical formula-
tions using reverse-phase liquid chromatography.
J Pharm Sci 72:221–224.
8. Barr M, Tice LF. 1956. A study of the inhibitory
concentrations of glycerine-sorbitol and propylene
glycol-sorbitol combinations on the growth of micro-
organisms. J Am Pharm Assoc 46:217–218.
9. Karabit MS, Juneskans OT, Lundgren P. 1989.
Studies on the evaluation of preservative effi-
cacy—IV: The determination of antimicrobial char-
acteristics of some pharmaceutical compounds in
aqueous solutions. Int J Pharm 54:51–56.
10. Herman EB, Haas GJ, Crosby WH, Cante CJ. 1980.
Antimicrobial action of short chain alcohols and
glycols. J Food Safety 2:131–139.
11. Plumridge A, Hesse SJA, Watson AJ, Lowe KC,
Stratford M, Archer DB. 2004. The weak acid pre-
servative sorbic acid inhibits conidial germination
and mycelial growth of Aspergillus niger through
intracellular acidification. Appl Environ Microbiol
70:3506–3511.
12. Razavi-Rohani SM, Griffiths MW. 1999. Antifungal
effects of sorbic acid and propionic acid at different
pH and NaCl conditions. J Food Safety 19:109–
120.
13. Renka RJ. 1988. Multivariate interpolation of large
sets of scattered data. ACM Trans Math Softw 14:
139–148.
14. Prickett PS, Murray HL, Mercer NH. 1961. Poten-
tiation of preservatives (parabens) in pharma-
ceutical formulations by low concentrations of
propylene glycol. J Pharm Sci 50:316–320.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
DOI 10.1002/jps
366
SCHELER ET AL.
15. Darwish RM, Bloomfield SF. 1995. The effect of
co-solvents on the antibacterial activity of paraben
preservatives. Int J Pharm 119:183–192.
16. De Spiegeleer B, Wattyn E, Slegers G, Van der
Meeren P, Vlaminck K, Van Vooren L. 2006. The
importance of the cosolvent propylene glycol on the
antimicrobial preservative efficacy of a pharmaceu-
tical formulation by DOE-rudgedness testing.
Pharm Dev Technol 11:275–284.
17. Hesse SJA, Ruijter GJG, Dijkema C, Visser J. 2002.
Intracellular pH homeostasis in the filamentous
fungus Aspergillus niger. Eur J Biochem 269:
3485–3494.
18. Bracey D, Holyoak CD, Coote PJ. 1998. Comparison
of the inhibitory effect of sorbic acid and amphoter-
icin B on Saccharomyces cerevisiae: Is growth
inhibition dependent on reduced intracellular pH?
J Appl Microbiol 85:1056–1066.
19. Przybylski KS, Bullerman LB. 1980. Influence of
sorbic acid on viability and ATP content of conidia of
Aspergillus parasiticus. J Food Sci 45:375–376.
20. Eklund T. 1983. The antimicrobial effect of disso-
ciated and undissociated sorbic acid at different
pH levels. J Appl Bacteriol 54:383–389.
21. Rehm HJ, Lukas EM. 1963. Zur Kenntnis der anti-
mikrobiellen Wirkung der Sorbinsa¨ure—Die Wir-
kung der undissoziierten und dissoziierten Anteile
der Sorbinsa¨ure auf Mikroorganismen. Zentral-
blatt Bakteriol Parasitenkd Infekt Hyg 117:306–
318.
22. Pethybridge AD, Ison RW, Harrigan WF. 1983.
Dissociation constant of sorbic acid in water and
water-glycerol mixtures at 258C from conductance
measurements. Int J Food Sci Technol 18:789–
796.
23. Pampulha ME, Loureiro-Dias MC. 1989. Combined
effect of acetic acid, pH and ethanol on intracellular
pH of fermenting yeast. Appl Microbiol Biotechnol
31:547–550.
24. Liewen MB, Marth EH. 1985. Growth and inhibi-
tion of microorganisms in the presence of sorbic
acid: A review. J Food Prot 48:364–375.
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 1, JANUARY 2010
PRESERVATION OF LIQUID DRUG PREPARATIONS FOR ORAL ADMINISTRATION
367
Wyszukiwarka
Podobne podstrony:
jps 21681
jps 21579
jps 22023
jps 22081
jps 22139
Viofor JPS jest aparatem do magnetostymulacji, FIZJOTERAPIA
jps 21484
jps 21792
jps 21755
jps 22267
jps, kryminologia
jps 22075
jps 21904
jps 21667
jps 21681
jps 22219
jps 21704
jps 22346
jps 21947
więcej podobnych podstron