The purpose of this study was to develop a novel drug delivery system for challenging

drugs with potential for scale-up manufacturing and controlled release of incorporated
drug. Pre-liposomes powder containing metronidazole, lecithin and mannitol, prepared by
spray-drying, was mixed with different tableting excipients (microcrystalline cellulose,
lactose monohydrate, mannitol, dibasic calcium phosphate, pregelatinized starch, pectin or
chitosan) and compressed into tablets. The delivery system was characterized with respect
to (i) dry powder characteristics, (ii) mechanical tablet properties and drug release, and (iii)
liposomal characteristics. The pre-liposomes powder was free-flowing, and tablets of
similarly high qualities as tablets made of physical mixtures were prepared with all
excipients. Liposomes were formed in situ upon tablet disintegration, dissolution or
erosion depending on the type of tablet excipient used. The liposomal characteristics and
drug release were found to depend on the tablet excipient. The new delivery system offers a
unique synergy between the ability of liposomes to encapsulate and protect drugs and
increased stability provided by compressed formulations. It can be adjusted for drug
administration via various routes, e.g. oral, buccal and vaginal.

In spite of an excessive number of potential drug candidates, the major drawback of

many of those molecules remains to be their poor water solubility resulting in poor
bioavailability. Several promising formulation strategies have been proposed and
devel-oped to improve the delivery of poorly-soluble substances. Most of such
formulations can be categorized into those including crys-talline solid formulations,
amorphous formulations and lipid for-mulations [1].

Among the lipid formulations, we are particularly interested in liposome-based

formulations. The ability of liposomes to solubilize poorly-soluble drugs in the lipid
bilayer and incorporate water-sol-uble drug in the aqueous core is well-established [2].
Due to their biphasic characteristic and diversity in design, composition and construction,
liposomes offer a dynamic and adaptable technology for enhancing drug solubility as
well as offering protection to drugs that are easily degraded [1,2]. However, aqueous
dispersions of lip-osomes may be subjected to a variety of stability problems associ-ated
with aggregation, fusion and phospholipid hydrolysis, which may limit their shelf life.
The drying of liposomal dispersions by freeze-drying or spray-drying is an important step
resulting in more stable dry powder formulations that can be rehydrated in contact with
water with maintaining their structure and function-ality [3–5]. Spray-drying is a
well-established technique that enables the design of particles with the desired
physicochemical properties, e.g. particles size, density, shape and solid form [6].
Additionally, spray-drying has been demonstrated to be a suitable method for preparation
of proliposomes [7], where liposomes are formed upon hydration of the dry powder.
Preparation of prolipo-somes by spray-drying should be distinguished from the
spray-drying of preformed liposomes, even though both products will form liposomes in
contact with water. The drying, as well as hydra-tion, of preformed liposomes is critical
steps due to the risk of fusion and/or aggregation and the disruption of the bilayer
struc-ture of the membrane [4,8]. The stress that the preformed liposomes are subjected to
during the spray-drying can be circumvented by spray-drying the liposomal components,
or lipo-some precursors, since the liposome formation will only take place upon
hydration of the dry product. For a precise terminology, and in order to distinguish these
types of systems from proliposomes prepared by other methods [9], it is suggested that
these systems should be referred to as pre-liposomes.

Up to now, only a few studies have successfully demonstrated formulation of

lipid-based nano-/micro-particulate delivery sys-tems that allow compression into tablets
[10–14]. These systems were either formulated as emulsions stabilized by polymer [10],
sugar [11] and/or porous silica particles [12] and subjected to spray-drying, or the lipids
were simply loaded onto porous silica particles by adsorption [13,14]. However, only the
silica-based systems resulted in tablets of high mechanical quality.

We propose a novel approach for manufacturing of a stable, spray-dried pre-liposomes

(PreLipo) powder that compressed into tablets (PreLipo tablets), combines the
solubilizing and protection properties of liposomes with the stability and ease of
administra-tion of a solid tablet formulation, as a new delivery system for chal-lenging
drugs. The originality of this approach is related to the formation of liposomes in situ
during disintegration, dissolution or erosion of the tablet. The release behavior of
associated drug can be tailored by the optimization of the liposomal composition as well
as appropriate selection of the tablet matrix. The new deliv-ery system has high potential
for industrial manufacturing since both spray-drying and tableting are already
well-established industrial scale manufacturing processes.

The aim of the current paper is to describe and confirm the con-cept and elucidate the
potential of PreLipo tablets as a novel drug delivery system. Metronidazole was chosen
as a model drug to illustrate that the concept is applicable for drugs destined to be
administered by several routes, e.g. oral, buccal and vaginal, not because of its solubility.
Metronidazole is classified as BCS class I drug and is slightly to sparingly soluble in both
hydrophilic and lipophilic media (aqueous solubility 10 mg/ml at 20 LC; log P 0.1 in
octanol/buffer pH 7.4) [15]. It is a known drug that is still the drug of choice for treatment
of several bacterial infections and protozoa [16]. It is associated with numerous side
effects when delivered systemically, therefore we were aiming at local therapy, e.g. local
treatment of periodontal diseases [17,18], local treatment of bacterial vaginosis [19,20] or
local treatment of Helicobacter pylori associated peptic ulcer disease [21]. Liposomes
have been shown to improve the delivery if antimicrobial drugs to bacteria by allowing
better contact with the bacterial membrane [22,23]. Another reason for encapsulating
metronidazole in liposomes is to retain the drug at the site of action (class I drugs are
highly per-meable), avoiding permeation to circulation and consequently reducing
systemic side effects and possible toxicity. By using lipo-somes the localized activity
enhanced even at lower doses as com-pared to the free drug provided by commercially
available dosage forms (e.g. immediately release tablets, cream and gel). Soybean
lecithin was chosen as the lipid, since it is known that it produces liposomes that are
suitable for improved local drug delivery, it is cheap compared to hydrogenated and
synthetic lipids, and thus well suited for industrial production. It has been proven suitable
for spray-drying [7] and was therefore highly interesting for this concept. To check
whether the PreLipo formulations are pres-sure-sensitive or otherwise challenging in the
tablet compression, both conventional tablet excipients and excipient used in ‘‘soft
tableting’’ [24] were used as tablet fillers. In respect to buccal and vaginal route of
administration, the bioadhesion of the system would be the most advantageous; therefore,
mucoadhesive poly-mers were also studied as possible tablet fillers.

2. Materials and methods

2.1. Materials

Metronidazole was purchased from Fluka Analytical (Steinheim, Switzerland) and D( )

mannitol from Kemika (Zagreb, Croatia). Lecithin (Lipoid S 75; soybean
phosphatidylcholine 68–73%, phos-phatidylethanolamine 7–10%,
lysophosphatidylcholine <3%) was kindly provided by Lipoid (Ludwigshafen, Germany),
microcrystal-line cellulose (Avicel PH102) was from FMC biopolymer (Leeds, England),
partially pregelatinized maize starch (Starch 1500 ) from Colorcon (Dartford, England)
and the direct compressible mannitol (Pearlitol 100SD) from Roquette (Basel,
Switzerland). The pectins of different degrees of methoxylation (DM; 5%, 10% and 25%)
were donated by Herbstreith & Fox GmbH (Neuenbürg, Germany). Chitosans of different
degree of deacetylation (DD; 77%, 82% and 92%) were obtained from Sigma Aldrich
(Milwaukee, USA). All other chemicals and solvents used in experiments were of
analytical grade.

2.2. Test media

Na2HPO4–NaH2PO4 buffer solution, in the following referred to as phosphate buffer
pH 7.0, was made from 305 ml 0.2 M Na2HPO4 and 195 ml 0.2 M NaH2PO4 diluted to
1000 ml with distilled water. The pH was adjusted to 7.0 with either HCl or NaOH.
Vaginal fluid simulant (VFS) was prepared by modification of the composition
originally reported by Owen and Katz [25]. It con-tained 3.51 g/l NaCl, 1.40 g/l KOH,
0.222 g/l Ca(OH)2, 2 g/l lactic acid, 1 g/l acetic acid, 0.16 g/l glycerol, 0.4 g/l urea and 5
g/l glu-cose. The pH of the mixture was adjusted to 4.5 with either HCl or NaOH.

2.3. Methods
2.3.1. Preparation of spray-dried pre-liposomes
Dry PreLipo powder was prepared by the modified spray-drying procedure
previously described by Škalko-Basnet et al. [7]. Briefly, empty and
metronidazole-containing PreLipo powder was prepared by mixing the ethanolic solution
of lecithin and aqueous solution of mannitol and metronidazole (where applicable),
whereupon the mixture was subjected to spray-drying in a Büchi 190 Mini Spray Dryer
(Büchi Labortechnik AG, Flawil, Switzerland). The spray-drying conditions were as
follows: the flow rate was 0.25 l/h, while inlet and outlet temperatures were 120 and 80
LC, respectively. The mixing ratio of lecithin to mannitol was 1:2 (w/w) for the empty
PreLipo powder, and for the drug-containing PreLipo powder, the ratio of lecithin to
mannitol to metronidazole was 1:2:0.3 (w/w/ w). The ratio of lecithin to metronidazole
1:0.3 (w/w) was chosen based on the solubility of metronidazole and phospholipid to
yield maximal entrapment of the drug upon hydration. All spray-dried powders were
stored in the dry (over silica) and cool (4 LC) condi-tions until hydration or tablet
preparation.

2.3.2. Characterization of pre-liposomes powder

The particle size distribution of the spray-dried PreLipo powder was determined by
analytical sieving, using a mechanical sieve shaker (Retsch VE 1000, Haan, Germany)
and sieves (Retsch, Haan, Germany) of 63, 90, 125, 180, 250, 355, 500 and 710 lm.

The true particle density was determined with a helium gas pycnometer (AccuPyc
1330, Micromeritics Instrument Corporation, Norcross, GA, USA). Reported results
represent the mean of three independent experiments with ten repetitive purge cycles and
three runs for each experiment.

Powder flowability was estimated from the Hausner ratio according to Ph.Eur. 2.9.36.
The bulk and tapped densities (Erweka tapped volumeter, type SVM, Heusenstamm,
Germany) were determined according to Ph.Eur. 2.9.34, using a 10 ml cylinder (10 mm
in diameter) as described by Salbu et al. [26] to reduce the amount of required powder.

Particle shape and morphology of the PreLipo powders were assessed with scanning
electron microscopy (SEM; JSM-6300 SEM, Japan Electron Optics Laboratory Ltd.,
Tokyo, Japan). The sam-ples were mounted on aluminum stubs with double-sided
adhe-sive carbon tape and sputter-coated with gold/palladium (Polaron SC7640 sputter
coater, Quorum Technologies Ltd., Ringmer, East Sussex, UK).

Metronidazole content in the PreLipo powder (sum of free and liposomally-associated

drug) was determined by dissolving sam-ples (2.000 mg) of the powders in methanol.
Quantification was done by UV–VIS spectrophotometry (Agilent Technologies GmbH,
Waldbronn, Germany) at a wavelength of 311 nm for methanolic solution. Samples were
taken by systematic sampling to ensure uniformity of drug content in the whole batch (n
= 10). Two sepa-rate batch productions were evaluated.

The phospholipid content in the spray-dried powders (empty and

metronidazole-containing) was estimated based on determi-nation of the total amount of
phosphorous according to Bartlett [27]. Two separate batch productions of the
drug-containing and empty pre-liposomes powder were evaluated.

Differential Scanning Calorimetry (DSC) was carried out using a calorimeter DSC 1
(Mettler Toledo AG, Schwerzenbach, Switzer-land). DSC analyses were performed on
the raw materials sepa-rately (lipid, mannitol and metronidazole), the spray-dried PreLipo
powders (drug-containing and empty pre-liposomes) and the PreLipo tablets (tablets of
drug-containing PreLipo powder with filler, and of physical mixture of all components in
the same ratio as PreLipo with filler). The tablets were grinded before the analysis. All
samples were weighed in aluminum pans and scanned from 10 to 170 LC at a heating rate
of 10 LC/min under nitrogen gas (flow rate 50 ml/min). Indium was used for temperature
and enthalpy calibration.

2.3.3. Preparation of pre-liposome tablets

The PreLipo powders, both empty and drug-containing, were mixed with different
tableting excipients in the ratio of the PreLipo powder to filler 15:85 (w/w) and
compressed to tablets (flat-faced, 6 mm diameter) on a costume made compaction
simulator (ServoPress 450, Schmidt Technology GmbH, St. Georgen, Germany, with
compaction module, IBR, Waldkirch, Germany). Prior to start of tableting and further on
demand, the tips and die were lubricated using a magnesium stearate in acetone
suspen-sion. Tablets were prepared by the manual die-filling, and the com-paction speed
was 10 mm/s (saw-tooth profile). The applied compaction force was 2.8 ± 0.2 kN,
corresponding to pressures around 100 MPa. The tablet mass was approximately 75.0 ±
2.5 mg. For comparison, tablets were also made from the physical mixtures of the
components under the same conditions.

2.3.4. Characterization of pre-liposome tablets

The mass of each tablet and its dimensions, i.e. thickness and diameter, were measured
using a micrometer screw (IP54 Wilson Wolpert, Maastricht, the Netherlands),
immediately after ejection, after storage for 24 h and 1 week, respectively. The crushing
force was determined (Erweka TBH 20, Heusenstamm, Germany) and the tensile strength
calculated according to Fell and Newton [28].

2.3.5. Characterization of in situ liposomes

The PreLipo powders (200 mg), both empty and drug-contain-ing, were hydrated with
1 ml of phosphate buffer pH 7.0 under continuous stirring (400 rpm) for 45 min, before 4
ml was added to dilute the liposomes. For the PreLipo tablets, each tablet was first
weighted, then disintegrated in 5 ml phosphate buffer, pH 7.0 (37 LC) under constant
stirring (400 rpm) for 1–3 h, depending on the filler. To remove the excipient from the
liposomes, the mix-tures were centrifuged (Megafuge 10 Heracus Sepatech, Osterode am
Harz, Germany) for 3 min at 4000 rpm (20 LC). The supernatant containing liposomes
and unentrapped drug was placed in a volu-metric flask and diluted with phosphate buffer
pH 7.0 up to 10 ml. This liposomal suspension was used for the particle size, entrap-ment
efficiency and zeta potential determinations.

The hydrodynamic diameter (dh) and polydispersity index (PI) of liposomes were
measured by dynamic light scattering (DLS) with a vertically polarized He–Ne laser
beam at a wavelength of 633 nm (Zetasizer 3000 HS, Malvern Instruments, Malvern,
UK). The scattering angle was fixed at 90L and the temperature was maintained at 25 LC.
Liposomes were diluted with 1 mM NaCl, pre-viously filtered through 450 nm Minisart
filters (Sartorius Stedim Biotech, Göttingen, Germany), to achieve a count rate between
100 and 300 Kcps. Prior to the measurements, all of the examined samples of liposomes
were sonicated for approximately 15 s in an ultrasonication bath. The hydrodynamic
diameter and PI were cal-culated from the autocorrelation function of the intensity of
light scattered from particles [29]. However, since the liposome size was expected to be
large, potentially in the upper end of the reli-able range of DLS, all samples were also
examined in an optical image analysis system (Olympus BH-2 microscope equipped with
a computer-controlled image analysis system Optomax V, Cambridge, UK). The
appearance of the sample was noted (vesicles, no-vesicles, aggregation) and the size of
the vesicles was estimated.

The entrapment efficiencies were determined in order to evalu-ate whether the

compression of PreLipo powder into tablets would influence the encapsulation efficiency
of drug in the in situ formed liposomes as compared to the non-compressed powder, and
also to check if the entrapment efficiencies were influenced by the type of tableting
excipient used as filler. The separation of free from lipos-omally-associated drug was
performed by two methods; dialysis and ultracentrifugation. Dialysis: Sample of
liposome suspension was placed in a dialysis bag (MWCO 12–14,000 Da) and
exten-sively dialyzed against the phosphate buffer pH 7.0 during con-stant stirring (100
rpm) for 4 h (23 LC). The volume of buffer was adjusted so that the concentration of the
drug was kept below the solubility of the drug. Ultracentrifugation method: The liposome
suspension (1 ml) was diluted with 2 ml of phosphate buffer pH 7.0 and ultracentrifuged
(Beckman Optima LE-80 K Ultracentri-fuge, Beckman Coulter Inc., Fullerton, USA) for
1 h at 100,000g (20 LC). Supernatant (unentrapped drug) was removed and the pel-let
was washed with 3 ml of phosphate buffer pH 7.0 under the same conditions. Drug
concentrations (both in liposomes and free) were examined spectrophotometrically. The
unentrapped metroni-dazole was determined at 320 nm in phosphate buffer pH 7.0, while
the liposomally-entrapped metronidazole was detected at 311 nm, after dissolving
liposomes in methanol. Blank tests were conducted with tablets containing the empty
PreLipo powder (i.e. tablets without metronidazole) under the same conditions. The
results were expressed as lg of metronidazole per mg of lipid, i.e. all values were
corrected for the loss of drug and lipid during the spray-drying process.

The influence of the type of tableting excipient as well as the vaginal environment on
the surface charge of liposomes was inves-tigated by measuring the zeta potential on
vesicle surface. Zeta potentials of liposomes, after disintegration of PreLipo tablets both
in the buffer pH 7.0 and VFS pH 4.5, respectively, were obtained by laser Doppler
velocimetry using a Zetasizer 3000 HS (Malvern Instruments, Malvern, UK). Samples
were placed in an electropho-retic cell, where a potential of 150 mV was established. To
ensure valid measurements, the instrument was calibrated through mea-surements of the
Malvern Zeta Potential Transfer Standard ( 50 ± 5 mV). All of the zeta potential
determinations were per-formed at 25 LC [29].

2.3.6. In vitro drug release

In order to monitor a possible differentiation between the release of unentrapped and
liposomally-entrapped drug from the PreLipo tablets, the release test was performed in a
dialysis bag. Each tablet was weighted and placed in a dialysis bag (MWCO 12–14,000
Da) together with 2.00 ml test medium (phosphate buf-fer pH 7.0 or VFS pH 4.5,
respectively). The dialysis bag containing a tablet and the buffer was placed in a beaker
with 100 ml test medium (receiver solution) at constant stirring (100 rpm). Aliquots (1.00
ml) were withdrawn from the receiver solution at predeter-mined time intervals. The
amount of drug released was quantified using a UV–VIS spectrophotometer (Agilent
Technologies GmbH, Waldbronn, Germany) at a wavelength of 320 nm (aqueous
solu-tions). After the last sampling, the remaining content from the dial-ysis bag was
dispersed in 10 ml of methanol, and the potential rest of metronidazole quantified
spectrophotometrically at 311 nm (methanolic solution). The recovery was calculated,
and the effect of the type of tableting excipient on the release behavior was evaluated.
The released metronidazole from PreLipo tablets was compared with that from the tablets
of physical mixtures of the same components.

2.3.7. Statistics
All data were expressed as mean ± SD. Significant differences between the values were
evaluated using Student’s unpaired t-test, and p < 0.05 was considered statically
significant.

3. Results

3.1. Spray-dried pre-liposomes powder

The PreLipo powder, or granules, obtained by spray-drying the mixtures of lipid,

mannitol and metronidazole in the ratio 1:2:0.3 (w/w/w) was pale yellowish, non-sticking
and free-flowing. The spray-dried powder without metronidazole (lipid and mannitol in
the ratio 1:2 (w/w)) exhibited similar properties. These two types of the spray-dried
powders are in the following text referred to as metronidazole-containing PreLipo
powder and empty PreLipo powder, respectively. Fig. 1 shows the differences in the
morphol-ogy of the two powders, indicating a smoother surface of the empty PreLipo
powder as compared to that of the metronida-zole-containing PreLipo powder.

The loss of both the drug and the lipid in the spray-drying pro-cess was estimated in
both types of powders, based on the quanti-fication of drug in the
metronidazole-containing PreLipo powder and lecithin, as determined by phosphorous
assay. The recovery of metronidazole was approximately 82% (82.2 ± 2.0% and 82.3 ±
1.5%) in the PreLipo powder, indicating a loss of only around 18% of the drug during the
process. The phosphorous assay con-firmed that about 20% of the lipid was lost during
the process; slightly more for the drug-containing PreLipo powder (22.6 ± 5.7% and 22.5
± 7.6%) compared to the empty PreLipo pow-der (17.8 ± 5.2%). After correction for the
loss of the drug and the lipid in the preparation process, the PreLipo powder was found to
contain 320 lg metronidazole/mg lipid.

Before preparation of tablets, both types of PreLipo powders were fully characterized
with respect to their powder properties (Table 1). The particle size distribution was found
to be rather broad for both powders; with 90% of the particles smaller than approximately
515 lm for the drug-containing and 625 lm for empty PreLipo powder, respectively. The
true density was slightly lower for the drug-containing PreLipo powder as compared to
the product composed of only lipid and mannitol. To assess the powder flow properties,
which is an important parameter for tablet pro-duction, the Hausner ratio (HR) was
calculated from the fluffed and tapped powder density. The value of 1.13 for the
drug-containing PreLipo powder indicated good flow properties, whereas the empty
PreLipo powder showed fair powder flow with a HR of 1.25.

The DSC curves (Fig. 2) showed that both metronidazole and mannitol raw materials
were crystalline with melting peaks at 159.6 LC (Fig. 2b) and 165.8 LC (Fig. 2c),
respectively. The DSC curves of the two PreLipo powders (Fig. 2d and e) showed
absence of metronidazole melting peak. The shift of mannitol melting point to 160.7 LC
and 162.5 LC, respectively, for these samples, indicates drug excipient interaction in the
formulation.

3.2. Pre-liposomes tablets

The PreLipo powders were mixed with the different types of excipients as fillers in
the ratio 15:85 (w/w), and tablets (6 mm, 75 mg) were compressed (100 MPa). The
typical direct compres-sion excipients, such as Avicel PH102 (microcrystalline cellulose),
Pearlitol 100SD (mannitol), Spherolac 100 (a-lactose monohy-drate), Emcompress
(dibasic calcium phosphate) and Starch 1500 (pregelatinized starch), were selected to
investigate the application of the PreLipo powder in some well-characterized, frequently
used fillers. For comparison, the tablets consisting of physical mixtures of the same
components in the same ratio were prepared. The tensile strengths of the two sets of
tablets are sum-marized in Table 2. The strongest tablets were prepared using Avi-cel
PH102, and the tablets containing Starch 1500 were too weak to allow easy handling
under the applied conditions. The PreLipo tablets showed similar tensile strength to those
tablets containing the physical mixture for each of the respective fillers. Also, the ‘‘soft
tableting’’ materials, chitosan (three different DD) and pectin (three different DM),
produced tablets of similar strength for the PreLipo powder and physical mixtures (Table
2). For the pectin-based PreLipo tablets the tensile strength seemed to decrease with an
increasing DM; a similar relationship between DD and mechan-ical strength was not
observed for the chitosan-based PreLipo tablets.

In DSC similar curves were found for the tablets of drug-con-taining PreLipo powder
as for the PreLipo powder as such, as shown in Fig. 2 for tablets containing Avicel PH102
(Fig. 2f). The drug melting peak was detected in the tablets made of the physical mixture
(Fig. 2g), however mannitol melting point was decreased in both types of tablet samples
(161.1 LC and 161.4 LC), confirming interaction between these two components.

3.3. In situ liposomes

The tablets disintegrated, dissolved or eroded in aqueous med-ium depending on the

type of tableting excipient used as filler, and the test medium. It was confirmed by the
image analysis that lip-osomes were formed in situ from the PreLipo tablets in aqueous
medium (data not shown), which was in contrast to the tablets made of physical mixtures
of the same components in the same ratio; they did not form vesicles. The characteristics
of liposomes formed from the PreLipo tablets were compared to those of lipo-somes
formed by the direct hydration of the PreLipo powder (con-trol). All analyses were
performed in phosphate buffer pH 7.0. The influence of the different tablet fillers on
particle size, drug encap-sulation efficiency and zeta potential of liposomes was
evaluated (Table 3).

The mean particle size of liposomes formed by the direct hydra-tion of PreLipo
powder was approximately 550 nm for the empty liposomes and 600 nm for the
drug-containing liposomes (Table 3). The polydispersity indices (PI) indicated that the
size distributions were broad, but acceptable (PI 0.4–0.5). The most pronounced fea-ture
of liposomes formed through dissolution of the PreLipo tablets was an broadening of the
size distribution (PI > 0.7), meaning that liposomes of various sizes were formed ranging
from smaller to very large ones, possibly also in or even above the upper size range that
can be accurately measured by DLS. Very high polydispersity (PI > 0.7) is a warning that
the mean size is only a rough estimate, and that the numbers should not be taken as
representative. Nevertheless, it was confirmed by the image analysis that spherical
vesicles of varying sizes were formed from the PreLipo tablets. The mean size estimated
for empty and metronidazole-containing lip-osomes seemed to be in the same order of
magnitude for each type of tablet excipient used as fillers, confirming consistency of
lipo-some formation with the respective filler (Table 3), thus an inter-pretation of the
main trends was proposed. In general, the particle size of liposomes formed from the
PreLipo tablets was about 100–200 nm larger as compared to the not-compressed
con-trol. However, some of the fillers (e.g. Starch 1500 , pectin DM 5% and DM 10%)
lead to formation of liposomes of smaller sizes (approx. 400–500 nm) as compared to the
control, and one of the chitosan grades (DD 77%) seemed to produce liposomes of 2.5–3
times larger mean size (1700–1800 nm).

The encapsulation efficiency of metronidazole in liposomes was determined after

separation of the drug encapsulated in liposomes from the unencapsulated (free) drug by
two separation methods (dialysis and ultracentrifugation). The results obtained by the two
methods were in an agreement, therefore only the results from the dialysis method were
displayed (Table 3). Liposomally-encapsulated metronidazole content in the control
(not-compressed PreLipo powder) was approximately 30 lg drug/mg lipid. Some of the
tableting excipients did not seem to affect the encapsulation efficiency to a greater extent,
e.g. Spherolac 100 and Emcompress , whereas other excipients, such as Avicel PH102
and Starch 1500 , strongly reduced encapsulation as com-pared to the control.
Interestingly, some of the highly swelling tab-leting excipients (chitosan DD 77% and
DD 92%) seemed to assist to an increased drug encapsulation compared to the control.

The zeta potential of liposomes formed by the direct hydration of the PreLipo
powder (control) was strongly dependent on the medium in which the tablets were
disintegrated (Table 3). In phos-phate buffer pH 7.0 liposomes, both empty and
metronidazole-containing exhibited highly negative zeta potentials. When the same
PreLipo powder was hydrated in VFS pH 4.5, a zeta potential close to zero ( 2.2 mV) was
detected. Furthermore, the various tablet fillers exhibited different effects on the zeta
potential of in situ formed liposomes in the two test media. The conventional tableting
excipients seemed to shield the negative surface charge of liposomes to different extent in
the phosphate buffer, as the zeta potential moved closer to zero for most of the excipients
as com-pared to the control. The zeta potentials of the metronidazole-containing
liposomes tended to be slightly more negative than the corresponding empty liposomes
from tablets of the same type of filler.

For liposomes formed from chitosan- and pectin-based PreLipo tablets, respectively,
the zeta potential was found to be strongly depended on the medium; in phosphate buffer
pH 7.0 the zeta potential of liposomes formed from chitosan-based PreLipo tablets was
negative, whereby with increasing the degree of chitosan deacetylation the zeta potentials
adopted more negative values. On the contrary, in VFS pH 4.5, similarly positive zeta
potentials (+37 mV) were observed. Liposomes from pectin-based PreLipo tablets were
of more negative zeta potentials between 40 and 50 mV at neutral conditions, and around
20 mV in VFS pH 4.5.

3.4. In vitro drug release

The dissolution rate of metronidazole from the PreLipo tablets prepared with the
conventional tableting excipients in phosphate buffer pH 7.0 is depicted in Fig. 3. In
addition, the drug release pro-file from the tablets composed as physical mixture of
lecithin, mannitol, metronidazole and one of the fillers (Avicel PH102) was included in
the figure and served as a comparison. The physi-cal mixtures of all formulations were
tested and showed similar release pattern as the one shown in Fig. 3. The dissolution rate
of the drug was sustained from all of the PreLipo tablets as compared to tablets of
physical mixtures. The release profiles from tablets made of all the conventional fillers
were more or less similar; how-ever, the fastest release was seen from the tablets
containing Pearlitol 100SD (65.7 ± 6.8% after 60 min) and the slowest for Emcompress
(49.6 ± 6.8% after 60 min). Since Starch 1500 did not produce coherent tablets under the
applied conditions, the fil-ler was disregarded in this test.

The release from PreLipo tablets containing chitosan or pectin was followed in two
media; in phosphate buffer pH 7.0 and VFS pH 4.5 (Fig. 4A–D). The release was faster in
the phosphate buffer pH 7.0 from both types of polysaccharide matrices. Larger
differ-ences were observed in the dissolution profiles of the PreLipo tablets containing
different chitosan grades (Fig. 4A) as compared to the pectins (Fig. 4C). In VFS pH 4.5 a
lag-time of about 15 min was observed for both types of swelling polymers.

4. Discussion

Although poorly-soluble drugs can be solubilized in the bilayers of liposomes,

liposomal suspensions, liquefied in nature, have limited storage stability and require
refrigeration. Spray-drying is, therefore, an attractive process that may transform the
liposo-mal suspension into a dry powder with increased stability [2–4]. However, a major
draw-back of such a process is the stressful con-dition that the liposomes are subjected to
during spray-drying, which results in a risk of fusion and/or aggregation and the
disrup-tion of the bilayer structure of the liposomes [4,8]. These problems may be
controlled or circumvented by the spray-drying of liposo-mal components or precursors,
here described as pre-liposomes (PreLipo), instead of the preformed liposomes. The
working hypothesis was that liposomes will be formed by the hydration of the PreLipo
powder. If the PreLipo powder can be compressed into tablets without the loss of
functionality, i.e. ability to form in situ liposomes, the new drug delivery system would
be highly attractive for industrial production. The system offers a unique synergy
between the ability of liposomes to solubilize and protect drugs with the increased
stability offered by compressed formula-tions. Moreover, the system can be adjusted to
manufacture tablets for application through various routes of drug administration, e.g. the
oral, buccal and vaginal route. Therefore, in order to illustrate the features and highlights
the potential, we focused on metroni-dazole as a model drug and full characterization of
the dry (powder and tablet properties) and wet (liposome properties) phases of the
delivery system, and different ways of tailoring the drug release.

4.1. Spray-dried pre-liposomes powder

In respect to tablet production, the powder flowability is a key parameter to obtain

mass and content uniformity of tablets, and was the selection criteria when producing
powders with the desired powder qualities. A dry, non-sticking powder was prepared by
spray-drying the mixtures of lipid, mannitol and metronidazole. Mannitol plays an
important role in preparation of a dry free-flow-ing powder, and in enabling hydration of
the spray-dried PreLipo powder. Lecithin was confirmed as a suitable lipid, and the
optimal ratio of lipid to mannitol was found to be 1:2 (w/w) using soybean lecithin
containing 70% of pure phosphatidylcholine. The recover-ies of both the drug and lipid
after the spray-drying were high; only about 20% of the materials were lost in the
process, which were in the same order of magnitude as in earlier reports on spray-dried
proliposomes [7].

The PreLipo powder showed a broad particle size distribution exhibiting good powder
flow properties (Table 1); moreover, it did not exhibit the cohesive nature often reported
for small spray-dried particles [30]. In the spray-drying process the droplets (size <5 lm)
were deposited on top of each other resulting in aggregates with good powder flow
properties. The drug-containing powder showed better flowability (lower Hausner ratio)
than the powder prepared without the drug. SEM micrograph (Fig. 1) indi-cated that the
drug-containing powder has a different morphology (containing some solid/crystalline
material) compared to the one without drug, which may explain the difference in flow
character-istic as well as the small deviation observed in powder density.
The raw materials, both metronidazole and mannitol, were of crystalline nature, as
confirmed by DSC. Also, after the spray-dry-ing, a melting peak indicating crystalline
material was observed in the DSC curves of both the drug-containing and the empty
Pre-Lipo powder, which could (partly) be explained by the presence of crystalline
mannitol in both samples. From a tableting perspective, the crystalline mannitol in the
spray-dried PreLipo powder may contribute positively to the powder flow of the
PreLipos. One lim-itation of the DSC analyses was that the melting point of mannitol is
very close to the melting point of metronidazole, which made a separation of the two
rather difficult. The proportion of metronida-zole to mannitol in the PreLipo powder is
low, which may contrib-ute to the explanation of why it was challenging to clearly
separate the metronidazole peak from the mannitol peak. Low heating rate (1 LC/min)
was used for this purpose, however, the full separation of the two peaks remained to be
challenging for the spray-dried PreLipo powders; in the case of a physical mixture the
two compo-nents could be detected (Fig. 2g). The shift observed in the melting points
compared to the points for the raw materials may be a result of the decrease in particle
size, or possibly, an interaction between the two substances.

4.2. Pre-liposomes tablets

In order to display the potential of the new concept, we focused on the combination of
the PreLipo powder with the different types of tableting excipients at a constant mixing
ratio (PreLipo powder to filler 15:85 (w/w)). This ratio was identified as suitable in
preli-minary experiments, but the ratio was not optimized in the current study. For several
of the excipients it would be possible to further increase the PreLipo-proportion. The
tablets were manufactured without experiencing any problems of sticking for all the
investi-gated excipients; except for the expected need of frequent lubrica-tion of punches
and dies (magnesium stearate in acetone suspension) for mannitol (Pearlitol 100SD), as
there was no lubri-cant in the powder mixtures.
Tablets of suitable mechanical strength were prepared from all excipients, except from
the pregelatinized starch (Starch 1500 ). This finding is not surprising since starch is an
elastically deforming material, and is therefore not expected to form strong tablets
with-out co-mixing with other tableting excipients. The other so-called conventional
fillers evaluated in the study were typical constituents of tablet formulations, with
different deformation behavior (micro-crystalline cellulose: plastic, lactose and dibasic
calcium phosphate: fragmenting). They all seemed to be suitable for the PreLipo tablet
manufacturing, and may therefore be applied in future formulation development of
PreLipo tablets to achieve the right balance of plas-ticity/fragmenting properties. Most
importantly, the PreLipo tablets showed similar tensile strength to the tablets containing a
physical mixture of the same components in the same ratios. This indicated that the
PreLipo powder did not interfere or disturb the ability to form tablets under the applied
conditions.
Also the PreLipo tablets prepared with chitosan and pectin showed suitable mechanical
strength to allow easy handling. The mechanical strength of the pectin-based tablets was
found to be dependent on the degree of methoxylation [26]. Both biopolymers are
suitable for soft tableting [24,31] as well as exhibiting mucoad-hesive properties [32,33].
The PreLipo tablets made of such matri-ces provide the potential to increase the retention
time on mucosal surfaces, e.g. the vaginal mucosa, targeting local therapy [34].

4.3. In situ liposomes

Liposomes were formed by hydration of the spray-dried PreLipo powder. The hydration
step is always critical for the in situ formation of liposomes. Spray drying of liposome
suspensions is more stressful for liposomes than spray drying of liposomal components
[4,8]. Spray drying of liposomal components is also a simpler and cheaper procedure
with respect to an industrial pro-duction. The high content of freely water-soluble
mannitol will facilitate the hydration of the PreLipo powder and hence the lipo-somes
formation.

Liposomes formed by the hydration of the PreLipo powder were large with a broad
size distribution (high PI), and assumed to be heterogeneous. The image analysis
technique did not provide information to the morphology of vesicles, but it might be
specu-lated that most of the large liposomes formed were oligolamellar or multilamellar
vesicles (MLV), as they were formed by passive hydration, whereas the smallest
liposomes were probably unila-mellar. It should be noted that the continuous stirring
applied in the in vitro studies, does not correlate directly to the in vivo condi-tions, even
though some motility is also expected in vivo. Continu-ous stirring might promote vesicle
formation more readily that what will occur in vivo. Liposomes of comparable size to
those formed from PreLipo tablets were recently reported for metronida-zole-containing
EPC/EPG-Na liposomes prepared by the classical film hydration method, prior to the
extrusion [29]. However, smal-ler metronidazole-containing liposomes have also been
prepared by other methods [35,36]. Since MLV could encapsulate more of the drug than
smaller liposomes, they are considered to be suitable for topical drug delivery, e.g. in
vaginal therapy [29,37]. Liposomes composed of soybean lecithin are known to be
suitable for improved local delivery, but their uses have been restricted due to poor
storage stability. Since the bilayers of lecithin liposomes are not rigid, storage in the form
of liposome suspensions could result in leakage of the entrapped hydrophilic drug.
Therefore, the concept of preparing pre-liposomes and storing the liposomal components
in ‘‘dry’’ condition is a way to improving physical sta-bility of liposomes and put them to
use. As a matter of fact, most of the commercially available liposomal products for
parenteral administration are in the solid form (lyophilizates, e.g. AmBisome, Abelcet,
Myocet [38], even though they are made of more physi-cally stable hydrogenated and
synthetic phospholipids. In terms of industrial production, soybean lecithin has the
benefit over the above mentioned lipids of being more economically rational and easily
accessible.

It is known that the particle size of liposomes formed from the spray-dried liposome
powder is also influenced by the volume of fluid available for the hydration [7]. In the
current study the hydra-tion volume was kept constant at 5 ml; however liposomes were
formed upon disintegration, dissolution or erosion of the tablet, and the tablet excipient
influenced the liposome characteristics such as the size, entrapment efficiency, zeta
potential as well as drug release behavior. Tablet fillers of very different characteris-tics,
spanning from freely water-soluble (mannitol) to highly swel-lable (starch, chitosan and
pectin) and poorly water-soluble (microcrystalline cellulose), were investigated.
Generally, the poly-dispersity of the liposome size increased when liposomes were
formed from PreLipo tablets as compared to the direct hydration of PreLipo powder
(control). Since the polydispersity was very high (PI > 0.7), the mean sizes should only
be taken as crude estimates. It was confirmed by image analysis that liposomes varied
greatly in size. Aggregation is less likely considering the high surface charges. Since the
mean size estimated for empty and metronidazole-con-taining liposomes seemed to be in
the same order of magnitude for each type of tablet excipient used as fillers, certain
consistency of the liposome formation with the respective filler was confirmed.
Therefore, a crude interpretation of the trends found for the filler on the liposome sizes
was proposed: For tablets prepared with fillers that were either waters-soluble (mannitol,
lactose, dibasic calcium phosphate) or poorly water-soluble (microcrystalline cellulose)
the size increased by approximately 100–200 nm for both the drug-containing and empty
liposomes. The water-soluble fillers were found not to interfere much with the formation
of lip-osomes and the observed broadening of the size distribution may be a result of the
hydration rate of the PreLipo powder as it is hom-ogenously distributed in the tablet. The
PreLipo powder on the sur-face of the tablet has free access to aqueous media during the
hydration, whereas it takes longer time until the core of the tablet will be fully hydrated,
which may be reflected in the size of the lip-osomes formed.

With some of the swelling polymers used as fillers (starch, pec-tin DM 5% and DM
10%) in situ formed liposomes were of smaller size, whereas for the most highly swelling
polymers (pectin DM 25% and all chitosans) bigger sizes were determined. It may be
postulated that the swelling of the polymer will limit the space available for liposome
formation, i.e. the liposome formation takes place within the three dimensional polymer
networks. Consequently, the vesicle size will be reduced. In the case of the highly
swelling polymers, three scenarios may explain the bigger sizes observed: The increased
viscosity of the dispersion may result in an aggregation of liposomes, so that the bigger
sizes are in real-ity clusters of liposomes. This may be the case for the chitosan (DD
77%), as the vesicles size of in situ formed liposomes was measured to be 2.5–3 times
bigger as compared to the control. This type of chitosan was also the one with the highest
molecular weight among the investigated grades. For the other two chitosan grades (DD
82% and DD 92%) and the most highly swelling pectin (DM 25%), it may be postulated
that the in situ formed vesicles were large due to extensive hydration and swelling of the
polymer chains, which reduced the volume of available liquid for liposome formation [7].
At the same time, the polymer might be also inside the in situ formed liposomes, which
will have an osmotic effect pulling water inside the liposomes hence the liposomes
become larger. It should also be noted that the sample undergoes several steps from the
dissolution of the tablet to the sample is measured on the PCS instrument, several of
which involved dilution, and this might influence the measured vesicle size.

Metronidazole is a relatively small molecule (171.15 Da), and exhibits, to a certain

level, both hydrophilic and hydrophobic char-acter. Based on its solubility (10 mg/ml)
and the log P ( 0.1), met-ronidazole may be expected to be entrapped in the aqueous part
of the vesicle, located at the interface between the lipid bilayer and aqueous interior of
the liposome or between the bilayers of MLV. However, the exact position of the drug in
the vesicle remains unproven. The entrapment was efficiency approximately 30 lg/drug
mg lipid for the prelipo containing some tablets of the conventional tablet excipients (e.g.
Spherolac 100, Pearlitol 100SD), which constitutes approximately 10% of the drug
content in the PreLipo powder (320 lg drug/mg lipid). Compared to Škalko-Basnet et al.
[7] where 104 lg metronidazole/mg lipid was liposomally-associated after the hydration
of spray-dried prolipo-somes, the current liposomes were bigger and entrapped less
met-ronidazole. However, an advantage of the PreLipo tablet system is that all drug
associated to the spray-dried powder will be present in the delivery system, and the
amount of drug available for local effect will therefore be higher than the liposomally
entrapped drug. Most importantly, liposomes will contribute to the drug retention at the
site of action, assuring local effect and avoiding unwanted absorption [29]. The impact of
the free drug remains to be investigated, but it is possible that some excipients, for
instance the mucoadhesive polymers, can contribute positively to enhanced local activity
also of the non-liposomally associated drug by maintaining it in close proximity to where
it is released.

Some of the tested excipients (e.g. Avicel PH102 and Starch 1500 ) seemed to have a
negative influence on the metronida-zole-entrapment in liposomes and the soft tableting
materials may therefore seem like an interesting alternative [24,31]. Other fillers
produced liposomes with a higher content of metronidazole, as compared to the control,
but some of these formulations also showed a larger particle size. It is therefore not a
strait forward task to evaluate whether the soft tableting materials actually pro-vided
better protection of the PreLipo powder during compression. However, there are several
factors suggesting that successful for-mulation of the PreLipo tablets is not restricted to
materials used in soft tableting. A range of conventional tableting excipients have shown
promising behavior. It can, therefore, be concluded that the PreLipo powder is not
particularly pressure-sensitive, at least not in 15% (w/w) dilutions with other materials.
This is of great impor-tance for industrial manufacturing.

The zeta potentials of liposomes were in accordance with the properties of materials
used for preparation of PreLipo tablets (lipid, drug, tablet excipient) as well as the
properties of the med-ium in which the tablets were disintegrated or dissolved
(phos-phate buffer pH 7.0; VFS pH 4.5). The negative charge of the controls can be
attributed to the low purity of the lecithin used for preparation of the PreLipo powder
(70% phosphatidylcholine), whereas for the metronidazole-containing vesicles the even
more negative values are probably a consequence of the negative charge of the
surface-associated drug (pKa 2.5), additionally contributing to the already negative zeta
potential of vesicles. On the other side, the presence of the fillers in the PreLipo tablets
seemed to shield some of the surface charge for the in situ formed liposomes.
 In the case of chitosan and pectin PreLipo tablets, the polymer worked as a hydrogel
matrix at high concentrations, i.e. little avail-able fluid, forming a liposome-in-hydrogel
like system [39], whereas in excess fluid a surface coating of polymers on the lipo-somes
can be assumed [40,41]. Positively charged chitosan has been used to coat negatively
charged liposomes by electrostatic deposition [40,42]. The amino group in chitosan has a
pKa value of about 6.5, which leads to protonation in acidic to neutral solu-tion with a
charge density dependent on pH and the DD. The values of zeta potential in phosphate
buffer pH 7.0 and VFS pH 4.5 were significantly different for these in situ liposomes,
reflecting that one medium was above and one below the pKa of the polymer. Similarly,
negatively charged pectin has been used in coating of positively charged liposomes
[33,41]. Even though the liposomes in the current study have negative surface, pectin
may still be sur-face oriented though not electrostatically bound. Pectin exhibited a
negative charge in both media investigated (pKa 3.4–4.1). The shift in zeta potential
toward zero (but still on the negative side), as seen for pectins, can be explained by the
shielding of the lipid sur-face by less negatively charged polymer chains. In VFS pH 4.5
this effect is more pronounced as the control liposomes exhibit a zeta potential closer to
zero without the polymer present, and when formed in situ from pectin-containing tablets
a zeta potential around 20 mV was determined.

4.4. Tailoring of drug release from the pre-liposomes tablets

The spray-dried PreLipo powder was highly porous, and easily hydrated in the
contact with aqueous medium. The release of drug was faster from the tablets made of
physical mixtures as compared to PreLipo tablets containing the same filler as illustrated
for microcrystalline cellulose in Fig. 3. The release from the physical mixture tablets
could be expected to more or less resemble the release from an immediate release tablet.
Our findings suggest that the drug was released from liposomes from the PreLipo tablets
in contrast to the tablets of physical mixture of the same components. The difference
proves that spray drying of the liposome compo-nents affects the formulation behavior,
and is necessary for the encapsulation of the drug into liposomes. The liposomes provide
a barrier that sustains the release of metronidazole, a slightly to sparingly water soluble
drug, to a certain degree. However, bilayer of soybean lecithin liposomes is not rigid, and
hydrophilic entrapped drug will be released faster than for lipids that produce more rigid
bilayers. However, it is likely to assume that the release behavior can be manipulated by
chancing the composition of the lipid phase.

Drug release from the PreLipo tablets was not only governed by the liposomes, but
also by the type of excipient used as filler for tablet production. This offers several
possibilities for tailoring the drug release profiles. For a fast disintegrating tablet, the
conven-tional tablet excipients Avicel PH102, Spherolac 100 and Emcompress may be
suitable. None of the materials were found to have a pronounced effect on the dissolution
rate of metronida-zole from PreLipo tablets. Such a system might in general terms be
interesting for oral administration [43,44]. For the metronida-zole-containing
formulations in current study, fast disintegrating PreLipo tablets could potentially be
interesting for local treatment of H. pylori, given that the liposomes would withstand the
harsh conditions of the stomach. Fast disintegrating PreLipo tablets could also be
interesting for vaginal therapy, however there might be other excipient that is even more
suitable for this purpose.

Incorporation of drug into liposomes is expected to enhance its activity at lower doses
as compared to free drug [45]. For metroni-dazole in the current study, it is assumed that
its localized effect will be enhanced firstly by liposomes providing better contact with
mucosa and bacteria, but also by liposomes retaining the drug at the site of action
preventing permeation and absorption. Hence fewer side effects can be expected and
lower doses can be applied as compared to immediate release tablets (systemic therapy)
or vaginal gels (local therapy). Minimal inhibitory concentration (MIC) for
metronidazole is reported to be 0.1–10 lg/ml for metro-nidazole [18,20,21]. Therefore,
obtained concentrations of metro-nidazole (200 lg/ml; PreLipo tablet dispersed in 5 ml)
should be well above the MIC to provide antimicrobial effect, even if the vol-ume is
further increased. For vaginal application, it is known that the daily production of vaginal
fluid is around 6 g with approxi-mately 0.5–0.75 g present in the vagina at any time [25],
which ensures concentrations well above the metronidazole MIC. In the case of local
buccal and oral therapy, the fluid volumes vary largely depending on fasted and fed state,
and also with respect to type and amount of food/drinks, therefore it is more difficult to
predict relevant metronidazole concentrations. Mucoadhesive polymers may be used to
provide required retention of the delivery systems on the mucosal surface to improve
therapy outcome, e.g. in vaginal therapy [34,35,46]. The PreLipo tablets were found to
form sort of in situ coated liposomes or liposomes-in-hydrogel formulations by swelling
and erosion of the tablet matrix in aqueous medium, depending on the access of fluid.
Both types of systems would be suitable for vaginal drug delivery. In this context, either
chitosan-or pectin-based tablets would be appropriate. The advantage of using chitosan, a
positively charged polymer, lies in its ability to form electrostatic interactions with
negatively charged mucin; thus enabling prolonged retention time of the formed
liposomes at the site of administration. However, the presence of chitosan in PreLipo
tablets was more affected by pH of the medium, both with respect to zeta potential and
dissolution rate, which would imply that the dosage form could potentially behave
differently in women of different age since the vaginal pH is known to increase from
around 4.2 in fertile age to around 6–7 during men-opause [25,47]. Pectin, on the other
side, was found to exhibit less variation in this pH range. Moreover, it has been shown
that also pectin interacts with mucin [33], providing the necessary mucoad-hesion to
increase the residence time on the mucosal surface.

Mucoadhesive liposomes are also promising drug delivery systems for therapy of
diseases in the oral cavity [48]. The PreLipo tablets offer the synergy of the two systems
and are therefore expected to be superior to conventional buccal tablets. Mannitol is
frequently used as matrix for orally disintegrating tablets, and the in vitro release studies
indicated that mannitol (Pearlitol 100SD) dissolved quickly, and would not interfere to a
greater extent with the release of the drug from the delivery system. By combining
mannitol with a mucoadhesive polymer, e.g. chitosan, the zeta potential results suggest
that the surface of the liposomes would be coated, which may offer mucoadhesive
properties beneficial for increasing the residence time on the buccal mucosa.
Mucoad-hesive liposomes would be formed in situ and released as the man-nitol matrix
dissolves in the saliva, leaving an ‘‘invisible’’ liposomal delivery system on the buccal
mucosa.

4.5. Perspectives

The concept of PreLipo tablets targets both local and systemic therapy, and the desired
drug release properties depend on the therapy goal, route of administration and the type
of drug in question. The objective of the current paper was to introduce the concept of the
PreLipo tablets by using a model drug to illustrate some of the features that the new
system may offer as an addi-tional tool for the formulation scientists. Further research is
required to fully characterize and understand the system and the factors affecting its
optimization with respect to specific sites of drug administration. Such studies would
cover among others stor-age stability, physicochemical characterization of the interaction
between the drug, lipid and mannitol in the dry as well as the wet state, the interaction
between liposomes and excipient, and applicability of the system for other drugs
including peptides and proteins and other biopharmaceuticals.
In order to move new delivery system closer to industrial man-ufacturing, simple
systems that allow easy scale-up and can be produced within the PAT-framework are
required, such as for instance the recently reported one-pot methods for preparation of
mucoadhesive liposomes [36]. The PreLipo tablets fulfill such requirements.

5. Conclusions

The PreLipo tablets, which are forming liposomes in situ during their disintegration,
dissolution or erosion, represent a new drug delivery system for the formulation of drugs
in advanced delivery systems with high potential for industrial manufacturing. The
sys-tem offers a unique synergy between the ability of liposomes to encapsulate and
protect drugs, and the increased stability offered by dry compressed formulations.
Moreover, the system can be adjusted to manufacture tablets destined for various routes
of drug administration, e.g. the oral, buccal and vaginal route. In addition, by using the
mucoadhesive polymers, it can be applied in the design and development of
mucoadhesive delivery systems. Further research is required to fully characterize and
optimize the new delivery system in specific routes of administration.