Journal of Drug Delivery Science and Technology 60 (2020) 101986

Contents lists available at ScienceDirect

Journal of Drug Delivery Science and Technology

journal homepage: www.elsevier.com/locate/jddst

Research paper

Preparation of delayed-release multiparticulate formulations of diclofenac

sodium and evaluation of their dissolution characteristics using biorelevant
dissolution methods
Daniel Zakowiecki a, *, Maja Szczepanska b, Tobias Hess a, Krzysztof Cal b, Barbara Mikolaszek b,
Jadwiga Paszkowska c, Marcela Wiater c, Dagmara Hoc c, Grzegorz Garbacz c, d
a
Chemische Fabrik Budenheim KG, Rheinstrasse 27, 55257, Budenheim, Germany
b
Department of Pharmaceutical Technology, Faculty of Pharmacy, Medical University of Gdansk, Al. Gen. J. Hallera 107, 80-416, Gdansk, Poland
c
Physiolution Polska Sp. Z o.o., Skarbowcow 81/7, 53-025, Wroclaw, Poland
d
Physiolution GmbH, Walther-Rathenau-Strasse 49a, 17489, Greifswald, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: Diclofenac sodium was used as a model drug for preparation of delayed-release (DR) multiparticulates, which
Multiple unit pellets systems were further processed into solid oral dosage forms such as capsules and tablets. Multiple unit pellets systems
Calcium phosphate-based pellets (MUPS) were prepared from different types of starter pellets (inert cores) including microcrystalline cellulose
Diclofenac sodium
pellets, sugar spheres, isomalt pellets and novel calcium phosphate-based pellets. The study results showed that
Starter pellets
pHysio-grad
the material of the inert cores affected both mechanical properties of the drug-loaded pellets and the dissolution
characteristic of the model drug. Biorelevant dissolution method carried out with the help of a pHysio-grad
device allowed thorough examination of the developed formulations in the environment mimicking pH condi­
tions along gastrointestinal tract. This method revealed significant differences between the formulations and
their sensitivity to variable hydrodynamic conditions.

1. Introduction technology compared to extrusion and spheronisation process. More­

over, it enables the production of multiparticulates with a very uniform
Multiparticulate drug delivery systems (MDDS) are novel pharma­ particle size distribution as well as advantageous surface morphology [5,
ceutical dosage forms, which have been gaining increasing popularity in 6]. There is a wide range of various types of neutral starter pellets
recent years when compared to single unit dosage forms. These multiple commercially available on pharmaceutical market offering different
unit dosage forms (MUDF) are made of numerous independent subunits characteristics [7,8]. Sugar spheres (nonpareils), which were introduced
(microparticles), where each of them is an autonomous reservoir of a on the market and applied for the production of multiparticulates as
drug and releases the drug in a desirable way, independently of the other first, are still the most popular. Sugar spheres are made of sucrose
subunits. Multiparticulates are especially suitable for preparation of (usually up to 92%) and corn starch [9]. They are soluble in aqueous
modified-release solid oral dosage forms (delayed- or sustained-release), media and hygroscopic, which is often a challenge during coating [10].
which offer such benefits as less variable gastrointestinal transit or Absorbed water can be problematic and impact stability of moisture
reduced risk of dose dumping. At the present time, the most commonly sensitive drugs [8]. That is why alternative products have been intro­
used multiparticulates are coated pellets which are formulated into oral duced to the market and interest in their use is constantly growing. The
dosage forms either by filling them into hard gelatin capsules or com­ most commonly employed substitutes are cellulose pellets consisting of
pacting into tablets (Multiple Unit Pellet System, MUPS) [1–4]. 100% microcrystalline cellulose (MCC). They are insoluble in water and
One of the well-established ways of preparation of multiparticulates are characterized by high sphericity as well as good mechanical
is drug layering of spherical starter pellets (also called inert cores or strength. Due to these properties, the drug layering can be carried out
beads) followed by coating these drug-loaded pellets with a thin faster and with reduced formation of undesirable agglomerates, which
polymer-based film. Drug layering of inert cores offers relatively easier shortens the coating time and reduces cost of production. However,

* Corresponding author.
E-mail address: daniel.zakowiecki@budenheim.com (D. Zakowiecki).

https://doi.org/10.1016/j.jddst.2020.101986
Received 2 April 2020; Received in revised form 31 July 2020; Accepted 31 July 2020
Available online 18 August 2020
1773-2247/© 2020 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
D. Zakowiecki et al. Journal of Drug Delivery Science and Technology 60 (2020) 101986

problems such as the absorption of certain drugs on the surface of cel­ of the intestinal pH was employed [34]. The pHysio-grad can precisely
lulose fibers or swelling in contact with water, which can affect disso­ monitor and adjust the pH value of hydrogen carbonate buffer, which is
lution pattern have been reported [11–13]. Another solution available considered to be one of the most biorelevant medium for simulation of
on the market are starter pellets consisted of polyalcohols such as iso­ intestinal conditions [35,36]. The carbonate buffer is a complex system
malt, mannitol, xylitol. Similarly to sugar spheres, they are hygroscopic and its pH is an effect of dynamic interplay between all its constituents.
and soluble in aqueous media, however, they are promoted due to their The thermodynamic instability leads to spontaneous increase of pH due
low glycemic index and lack of cariogenic effect [5,14]. Within this to the loss of carbon dioxide (CO2). This effect might be minimized by
group, isomalt pellets offer especially interesting functionality of either preventing CO2 evaporation (by using appropriate sealing or
creating internal osmotic pressure, which modulates dissolution pattern covering a dissolution medium with an organic layer) or by compen­
of some drugs [15]. Recently, a new type of water-insoluble pellets sating the loss of CO2 with the help of an automated system. Moreover,
based on anhydrous dibasic calcium phosphate (DCPA) has been intro­ carbonate buffers give the opportunity for simulation of dynamic
duced to the market. These pellets consist of 80% w/w of DCPA and 20% intraluminal pH changes, however, continuous and dynamic adjustment
w/w of MCC. Due to their elevated density of above 1000 g/l, very low of pH value must be provided in order to use them in routine dissolution
water content of less than 1%, and reduced hygroscopicity, they repre­ testing.
sent an attractive alternative to the other commercial products offered The main objective of this work was to investigate the suitability of
so far. A detailed comparison of various functional properties of calcium different types of starter pellets in the development of delayed-release
phosphate-based pellets with other commercially available inert cores multiparticulate formulations containing diclofenac sodium as a model
can be found elsewhere [16]. drug. The planned scope of research included the use of commercial
Inert cores are made of excipients commonly used in the pharma­ inert cores made of microcrystalline cellulose, sugar, isomalt as well as
ceutical industry, which are neutral and should not exhibit any phar­ novel calcium phosphate-based pellets in preparation of both hard
macological activity nor interact with the drug substance in a way, gelatin capsules and compressed tablets. The additional objective was to
which may adversely affect its stability and/or effectiveness. Never­ examine and compare the effect of the core material on the in vitro drug
theless, characteristics of starter pellets may have a significant impact on release of the model drug using compendial as well as biorelevant
the course of the production process as well as can influence the rate of dissolution methods.
drug dissolution not less than the properties of coating polymer. Thus,
the drug release pattern can be determined by the thickness of the 2. Materials and methods
coating and its permeability, but also by the geometry of the pellets,
surface morphology or the osmotic pressure originating from the pellets Diclofenac sodium (Amoli Organics, Mumbai, India). Film coating
cores [5,8,15,17–20]. systems: Vivacoat® FM-1M 000 (JRS Pharma, Rosenberg, Germany) and
As has been already mentioned, drug-loaded pellets can be filled into Aquarius® Control ENA (Ashland, Covington, KY, USA). Inert cores:
hard gelatin capsule shells or compressed into tablets. Especially, the calcium phosphate-based (DCPA) pellets - PharSQ® Spheres CM M
latter technology of tabletting of microparticulates is very interesting (Chemische Fabric Budenheim, Budenheim, Germany), microcrystalline
and allows combining the advantages of both tablets and pellets-filled cellulose pellets - VIVAPUR® MCC Spheres 500 (JRS Pharma, Rosen­
capsules in one dosage unit [21–23]. In this case, the properties of berg, Germany), sugar spheres - pharm-a-spheres™ MESH 35-25 (Hanns
pellet cores together with the coating polymer will determine the G. Werner GmbH, Tornesch, Germany), isomalt starter pellets - gale­
durability of the drug-loaded pellets and protect them from damages nIQ™ 960 (BENEO-Palatinit GmbH, Mannheim, Germany). Microcrys­
during tabletting. It has been frequently reported that the size of pellets talline cellulose VIVAPUR® 200 (JRS Pharma, Rosenberg, Germany).
is a very important factor to be considered during compaction. Nor­ Low-substituted hydroxypropyl cellulose (L-HPC) LH-11 (ShinEtsu,
mally, small pellets are employed in preparation of tablets as they are Wiesbaden, Germany). Magnesium stearate (Peter Greven Fett-Chemi,
less affected by the compression process. This is mainly due to the higher Venlo, The Netherlands). Transparent hard gelatin capsule shells, size
mechanical strength of smaller beads relative to their size as well as “0” (Kapselwelt, Hude, Germany).
reduced contact force on each individual pellet resulting in significantly
reduced degree of transformation [24,25]. 2.1. Preparation of drug-loaded DR pellets
In this work the suitability of different types of starter pellets in the
development of delayed-release multiparticulate formulations contain­ The starter pellets used in this study were initially calibrated be­
ing diclofenac sodium as a model drug was evaluated. Diclofenac so­ tween two sieves, 500 μm and 710 μm, in order to obtain grains of
dium is a nonsteroidal anti-inflammatory drug (NSAID) having both similar dimensions and to avoid the effect of different particle sizes on
analgesic and antipyretic activities. The drug is commonly used in the the coating process as well as on analyses results. Sieved pellets were
treatment of rheumatic disorders. Diclofenac sodium is available in a drug-layered with diclofenac sodium in a ProCepT 4M8-Trix Fluid-bed
number of preparations of 25 mg, 50 mg or 75 mg strength [26–28]. The system (FBS) equipped with a Wurster column (ProCepT nv, Zelzate,
drug is a weak acid of BCS Class II and shows pH-dependent solubility in Belgium). Around 100 g of starter pellets were first coated with a
physiological pH range which increases with the increase of pH value aqueous suspension containing 5% w/w of the drug substance and 5%
[29,30]. Diclofenac sodium is often formulated as enteric-coated prep­ w/w of Vivacoat® system up to about 20% of weight gain. Subsequently,
aration to avoid direct contact of the drug with the mucosa resulting in without breaking the process, the pellets were sprayed with purified
local gastrointestinal toxicity. For this purpose pH-sensitive enteric water (intermediate coating) and finally coated with 20% w/w aqueous
polymers are commonly used. These polymers should limit the release of suspension of Aquarius® Control ENA (enteric-coating) until around
the drug in acidic conditions of stomach and allow rapid dissolution of 10% of weight gain was reached. The use of an intermediate coating step
the drug in duodenum and distal parts of the gastrointestinal tract (GIT) allowed maintaining the continuity of the entire process and adjusting
[31]. the process parameters before the following coating phase. Moreover, it
The development of drug products is a time-consuming process and reduced interactions between the two layers and the accumulation of
involves high costs. Thorough examination of drug formulations with static charges. During all phases the machine settings were maintained
the help of reliable in vitro biorelevant techniques facilitates screening at the following values: air speed of around 0.21 m3/min, inlet air
of the candidates and makes this process more efficient. Biorelevant temperature of around 63 ◦ C, product temperature of around 40 ◦ C,
dissolution methodologies allowing an adequate prediction of the in nozzle pressure of 1 bar.
vivo performance of drugs come here with help [32,33]. In the present
work a novel device, pHysio-grad, which enables biorelevant simulation

2
D. Zakowiecki et al. Journal of Drug Delivery Science and Technology 60 (2020) 101986

2.2. Preparation of multiparticulate dosage forms CO2 and compressed air in this case (Fig. 1).
Dissolution was carried out in a paddle apparatus (USP apparatus 2)
Enteric-coated pellets containing the model drug were used to at a 50 rpm paddle rotational speed. Dissolution test was composed of
formulate two types of solid oral dosage form - hard gelatin capsules and two stages. In the first stage tablets or pellets (taken out from capsule
compressed tablets. Developed MUPS formulations consisted of around: shells) were placed in a vessel filled with 250 mL of 0.01 M HCl for 30
50.0% w/w of the diclofenac sodium pellets (equivalent to 25 mg of min. After this time 750 mL of Hank’s buffer concentrate was added and
diclofenac sodium), 44.5% w/w of MCC type 200, 5.0% w/w of L-HPC the test continued for 60 min. After addition of buffer, the pH of
and 0.5% w/w of magnesium stearate. Hard gelatin capsules were pre­ hydrogen carbonate medium was controlled by pHysio-grad device and
pared by manual filling of the diclofenac sodium pellets into capsule set at 6.8. The amount of the dissolved drug was determined using an
shells of “0” size. Tablets were compressed on a Korsch EK0 eccentric Agilent 8453 UV/VIS Spectrophotometer (Agilent Inc. Santa Clara, CA,
tablet press (Korsch, Berlin, Germany) using concave punches (R 25 USA) at the detection wavelength of 276 nm using quartz cuvettes with
mm) of 12 mm diameter under a compaction forces of 18–20 kN. 1 cm path length.

2.3. Scanning electron microscopy 3. Results

To visualize surface properties of the drug-loaded DR pellets alone, 3.1. Mechanical strength
as well as compressed into tablets, a scanning electron microscope
Phenom Pro (Phenom World Thermo Fisher, Eindhoven, Netherlands) Mechanical strength of starter pellets and drug-loaded DR pellets was
was employed. Standard sample holder and a carbon tape were used to assessed with a texture analyzer. The maximum force at the point of
fix a sample and acceleration voltage of 5–10 kV was applied to record fracture (hardness) and the distance to break vs. pellets breaking
images at a magnification of 300x. strength [N] are shown in Figs. 2 and 3. The highest value of hardness
was recorded for MCC spheres, and the lowest for sugar-based inert
2.4. Mechanical strength cores. Generally, it was observed that for all types of starter pellets the
coating process significantly improved the mechanical strength of the
The mechanical strength of starter pellets as well as drug-loaded DR particles. The increase in robustness was the most pronounced in case of
pellets was assessed with a texture analyzer equipped with a 6 mm water-insoluble pellets (MCC and DCPA). It can be noted that for iso­
diameter cylindrical stainless steel probe in compression mode (TA.XT malt-, sugar- and MCC-based pellets values of the standard deviation are
Plus, Godalming, Surrey, England). The probe was moved vertically relatively high which could result from their not quite spherical shape.
downwards at a speed of 2.0 mm/s until the triggering force of 0.1 N was Drug-loaded enteric-coated pellets based on isomalt and sugar exhibited
detected and further with a speed of 0.05 mm/s until a pellet was very similar response to the compression. They deformed relatively fast
crushed. The maximum force at the point of fracture and the distance to when exposed to external force until reaching the point of fracture as
break were recorded. All measurements (for each type of sphere n = 12) shown in Fig. 3. Deformation of DCPA-based pellets was significantly
were performed at room temperature. slower. It can be noted that the maximum force at the breaking point is
at similar level as for isomalt pellets, but the distance to reach fracture
2.5. Dissolution test (modified compendial method) point was significantly extended. This can be the effect of the specific
composition of DCPA-based inert cores, which combine the features of
According to The United States Pharmacopeia monograph for brittle calcium hydrogen phosphate and a ductile material (microcrys­
diclofenac sodium delayed-release tablets, dissolution test is carried out talline cellulose).
in two steps, between which test samples must be transferred from the
one medium to the other. In the case of enteric-coated tablets a paddle 3.2. Scanning electron microscopy
apparatus is applicable because the transfer of such tablets is not a major
problem. However, in the case of microparticulates the use of a basket Surface structure and cross-sections of the drug-loaded DR pellets
apparatus is more convenient. In this study both hard gelatin capsules before and after compression into tablets were studied with a scanning
and tablets containing enteric-coated diclofenac sodium pellets were electron microscope. Fig. 4 shows SEM micrographs of drug-loaded
placed in a basket apparatus (USP apparatus 1) PTWS 820D (Phar­ diclofenac sodium pellets before compression (left column), com­
maTest AG, Hainburg, Germany) and the dissolution test was carried out pressed pellets inside tablets (middle column) and pellets extracted from
according to the monograph for Diclofenac Sodium Delayed-Release the surface of tablets (right column). SEM images of cross-sections show
Tablets given in USP42/NF37 using a rotational speed of 100 rpm.
During the first phase the capsules were incubated in 0.1 M hydrochloric
acid for 2 h. In acidic stage, according to both USP/NF and Ph.Eur.
gastro-resistant pellets should not release more than 10% of labeled
content of a drug. In the subsequent buffer stage, release of the drug
should be rapid and reach minimum 80% of the labeled amount of
diclofenac sodium within 45 min. The amount of the dissolved drug was
determined using a T70 UV/VIS Split-Beam Spectrophotometer (Phar­
maTest AG, Hainburg, Germany) at the detection wavelength of 276 nm
using flow-through quartz cuvettes with 1 cm path length.

2.6. Biorelevant dissolution in hydrogen carbonate buffer using a pHysio-

grad device

The pHysio-grad device used in this study was composed of a

microcomputer, which controlled independently each measurement
channel and a valve island. pH probes assured constant pH measurement
in each vessel and triggered adjustment of pH when necessary. The
adjustment of pH values was obtained by addition of a titrant, which was Fig. 1. Schematic diagram of the pHysio-grad device.

3
D. Zakowiecki et al. Journal of Drug Delivery Science and Technology 60 (2020) 101986

Fig. 2. Comparison of breaking strength (hardness) of inert cores made of DCPA, MCC, sugar and isomalt with corresponding drug-loaded DR pellets (given are
means of min. 10 repetitions, SD is indicated by the error bars).

Fig. 3. The distance to break and the maximum force at the point of fracture of drug-loaded enteric-coated pellets based on DCPA, MCC, sugar and isomalt (given are
examples of single deformation profiles).

that the shape of sugar- and DCPA-based drug-loaded pellets is spher­ cracks are visible for pellets made of sugar, isomalt and MCC. Especially
ical, whilst isomalt and MCC show more ellipse-like character. In all sugar-based pellets were heavily deformed, which led to partial expo­
cases a clear distinction between a core and a coating layer can be sure of their cores.
spotted with no visible deformation of the coating layer. The coating
layer is uniform through the whole cross-sections with layer thickness of 3.3. Dissolution test (modified compendial method)
approximately 30 μm. Although the irregular shape of the some types of
examined inert cores could have caused a variation in the coating Four types of commercially available inert cores (DCPA pellets, MCC
thickness, no such observations were made. It can be seen that even after spheres, sugar spheres and isomalt pellets) were used to prepare
compression the original shape of the pellets located inside the tablets delayed-release multiparticulate formulations (hard gelatin capsules
generally remained unchanged, without visible deformations or dam­ and tablets) of diclofenac sodium at a dose of 25 mg. All of them were
ages of the coating layer. However, it should be noted here that many analyzed using modified compendial methods for diclofenac sodium
pellets positioned on the very surface of the tablets, which were in direct delayed-release tablets (USP42/NF37). Comparison of the amount of the
contact with the punches during compaction, displayed cracks in the drug released in the acidic medium is given in Fig. 5. During this phase
polymer layer and partial shape deformation. The most pronounced both hard gelatin capsules and tablets disintegrated completely and

4
D. Zakowiecki et al. Journal of Drug Delivery Science and Technology 60 (2020) 101986

Fig. 4. SEM micrographs of drug-loaded diclofenac sodium DR pellets (magnification of 300x).

Fig. 5. The release of diclofenac sodium from gastro-resistant diclofenac sodium preparations in acid phase (after 2 h incubation in 0.1 M HCl) (given are mans of n
= 6, SD is indicated by the error bars).

water-insoluble ingredients formed a cone at the bottom of dissolution amount of diclofenac sodium within 2 h of the test. However, in case of
vessels. In the baskets remained practically only pellets with very small preparations containing water-soluble starter pellets higher dissolution
residue of excipients. It was observed that the water-soluble inert cores of the drug substance can be detected. It can also be noticed that tablets
(made of sugar and isomalt) tended to float in the upper part of baskets, containing inert cores with MCC, sugar and isomalt released less amount
whereas the insoluble ones (DCPA and MCC) remained mostly at the of the drug than analogous capsules. The opposite situation occurred in
bottom of baskets. All examined formulations demonstrated their case of phosphate pellets, where the release of diclofenac sodium from
gastro-resistance and released much less than 10% of the labeled tablets is slightly higher than from capsules.

5
D. Zakowiecki et al. Journal of Drug Delivery Science and Technology 60 (2020) 101986

In the buffer stage all tested formulations demonstrated rapid contrary, MCC- and DCPA-based pellets exhibited more stationary
dissolution and released the whole amount of the drug within required behavior. Especially pellets containing DCPA, due to their elevated
45 min (Fig. 6). It can be noted that multiparticulates containing water- density, were lying still at the bottom of the vessels even during addition
soluble substances showed identical, measurably faster dissolution rate of gaseous titrants. The limited movement resulted in much lower
in comparison to insoluble MCC and DCPA cores. In the case of phos­ dissolution rate, particularly in comparison with sugar and isomalt
phate pellets, the release of diclofenac sodium from tablets is slightly cores. DCPA-based pellets did not release the whole of the labeled
slower than from hard gelatin capsules, while in the other examined amount of diclofenac sodium within 60 min of the test. Additional
formulations these differences are not so visible. After completing this intensive mixing was needed to release all the substance (results not
phase in the baskets containing DCPA and MCC microparticulates still included in the diagram).
the whole quite spherical cores could be observed while for water- Significant differences in dissolution pattern between tablets and
soluble sugar and isomalt beads the baskets were practically empty. In free pellets could be observed, especially at the very beginning of the
the case of MCC spheres, a large amount of retained red dye (which is a second stage of the dissolution test. First samples were withdrawn in 1
component of the enteric-coating) can be seen. (Fig. 7). min after addition of Hank’s buffer (Fig. 9). All tested tablets have
released more than 20% of label amount of diclofenac sodium within
3.4. Biorelevant dissolution in hydrogen carbonate buffer using a pHysio- this time, which could be caused by the presence of excipients that
grad device promoted movement of freed pellets and increased their exposure to the
dissolution medium. For comparison, at the same time the free pellets
Biorelevant dissolution test has shown the high discriminatory released a much smaller amount of the drug substance (about 12–13%
power and revealed significant differences between the tested formu­ for sugar and isomalt pellets, 11% for MCC cores and only 4% for DCPA
lations in terms of the impact of the dosage form as well as the core cores). In case of pellets based on water-insoluble material, contrary to
material on the dissolution rate. Moreover, the use of a paddle apparatus tablet formulations, free pellets showed less tendency to move and were
enabled visual observation of the tested samples directly in dissolution lying at the bottom of the dissolution vessels. It could be also observed
vessels. that dosing of the gaseous titrant had a greater effect on movement of
The results presented in Fig. 8 show that drug-loaded pellets con­ isomalt and sugar cores and facilitated their floating in the entire volume
taining sugar and isomalt, both alone (without capsule shells) and of the dissolution medium.
compressed into tablets, released diclofenac sodium faster than corre­
sponding MCC and DPCA cores. In addition, the release of the model 4. Discussion
drug from these tablets and free pellets was virtually identical and the
dosage form had no effect on its dissolution rate. These findings are in The aim of this work was to compare the suitability of different types
line with the results obtained in the dissolution test based on the com­ of starter pellets in the development of delayed-release multiparticulate
pendial method (Fig. 6). A different situation can be observed in the case formulations containing the model drug - diclofenac sodium. Four types
of pellets containing water-insoluble MCC and DCPA. Here, the differ­ of commercial inert cores of pharmaceutical grade were selected for the
ences between the two dosage forms are clearly visible. The dissolution research. Two of them were soluble in aqueous media – sugar spheres
rate from pellets is much slower than from corresponding tablets. This and isomalt pellets. Other two, microcrystalline cellulose spheres and
difference is noticeably larger in the case of DCPA cores. calcium phosphate-based pellets, were insoluble in water. Starter pellets
Examination of free pellets allowed observation of their different available on the market differ significantly in size which can signifi­
dissolution behavior which might be related to the core material. Sugar- cantly impact properties of multiparticulate formulations including their
and isomalt-based pellets were mobile and tended to float in the entire dissolution characteristics [37–39]. In order to eliminate impact of
volume of dissolution medium. Dosing of a gaseous titrants in the second particle diameters on results of analyses the cores used in this study were
stage of the test additionally stimulated the movement of pellets in the normalized and only the 500–710 μm fraction was utilized. A chemical
vessel which could promote dissolution of diclofenac sodium. On the nature of the core material determined not only the solubility but also

Fig. 6. Dissolution rate of diclofenac sodium from various multiparticulates in the buffer stage (given are mans of n = 6, SD is indicated by the error bars).

6
D. Zakowiecki et al. Journal of Drug Delivery Science and Technology 60 (2020) 101986

Fig. 7. Residues of diclofenac sodium DR pellets (DCPA, MCC, sugar, isomalt) at the bottom of basket after completing the buffer stage.

Fig. 8. Dissolution rate of diclofenac sodium from various multiparticulates in the hydrogen carbonate buffer pH 6.8 (biorelevant test condition) (given are mans of
n = 3 SD is indicated by the error bars).

the functional properties of starter pellets that were influencing the inert cores. The coating of inert cores – in this case with a water sus­
process of drug layering of the cores. For instance, different tendencies pension of hydroxypropyl methylcellulose (HPMC) mixed with the equal
to accumulate electrostatic charges have been observed. Fig. 10 shows amount of diclofenac sodium – significantly improved mechanical
pictures of FBS chamber taken about 5 min after the start of the process. strength of the pellets. The greatest improvement was observed for
In case of MCC, sugar and isomalt cores, high adherence of the particles DCPA pellets for which hardness value increased by over 2.25 times. In
to the glass walls of the granulator, caused by static electricity, is clearly the course of research, no clear relationship between pellets hardness
visible. Fluidization of DCPA beads is smooth and undisturbed. More­ and resistance to crush during tabletting could be observed. Considering
over, in the case of water-soluble inert cores, especially made of isomalt, that the coating thickness for all pellets was identical (around 30 μm),
initial extensive formation of dust has been noticed. This observation the differences in their robustness had to be due to the properties of the
corresponds to the data published earlier and showing higher friability core material. Drug-loaded enteric-coated sugar spheres with the lowest
of these inert cores [16]. value of breaking strength were affected the most during compression
Starter pellets differed in terms of hardness expressed here as (see Fig. 4). Nevertheless, when comparing coated pellets of similar
breaking strength (see Fig. 2) The most durable were MCC spheres, the hardness, i.e. based on isomalt and DCPA, one can notice much less
lowest mechanical strength was observed for sugar- and DCPA-based compression damages of these latter. This difference is probably due to

7
D. Zakowiecki et al. Journal of Drug Delivery Science and Technology 60 (2020) 101986

Fig. 9. The release of diclofenac sodium after 30 min incubation in 0.01 M HCl and 1 min after addition of hydrogen carbonate buffer (given are mans of n = 3 SD is
indicated by the error bars).

Fig. 10. Chamber of fluid bed system after around 5 min of the coating process containing 1) DCPA pellets, 2) MCC spheres, 3) sugar spheres, 4) isomalt pellet.

the longer time needed to reach the fracture point for the phosphate- allow quick dissolution in the duodenum and distal GIT. In the present
based pellets, as shown in Fig. 3. Furthermore, the inert cores made of studies, the quality of the coating layer has been evaluated after pro­
ductile MCC with the highest value of breaking strength were damaged longed storage under acidic conditions, reflecting those prevailing in the
even more than previous two types. It should be mentioned here that the stomach, as it is recommended in Ph.Eur. (chapter 2.9.3.) or USP/NF
pellets of the sizes used in this study (500–710 μm) are normally not (chapter 711). After 2 h of maceration in 0.1 M HCl, it was found that all
employed for preparation of MUPS tablet. However, a similar approach tested microparticulates met the compendial requirements and did not
has been successfully applied for research purposes and reported by release more than 10% of the labeled content of the active substance (see
Hiew et al. [40] In this case, however, such particle dimensions allowed Fig. 5) proving their gastro-resistance. It can be noted, however, that the
precise observation of the coating layer and its deformation or damages drug-loaded pellets based on water-insoluble cores released less diclo­
arose during compression into tablets. It is also important that the fenac sodium than water-soluble ones. When comparing MUPS tablets
damages were limited to the pellets located on the very surface of the and capsules, one can also observe slight differences in the amount of the
tablets, which had direct contact with steel parts of tableting tooling. It drug substance dissolved. For tablets containing sugar, isomalt or MCC
can therefore be assumed that many of the observed cracks were caused cores, the release of diclofenac sodium under acidic conditions was
mechanically and the use of larger amounts of cushioning excipients lower than from corresponding capsules. The smallest change was
could significantly reduce the damages. observed in the case of microparticulates based on MCC. This observa­
Diclofenac sodium is known to cause adverse GI side effect after oral tion is unexpected as the analysis of SEM micrographs (see Fig. 4)
administration [29]. Thus its preparations are commonly enteric-coated revealed damages to the polymer layer of many pellets after compres­
in order to reduce gastric exposure. Enteric-coating should limit release sion. Probably during prolonged maceration in aqueous solution of hy­
of the active substance in the stomach as much as possible and then drochloric acid, the polymer coating was hydrated and formed hydrogel

8
D. Zakowiecki et al. Journal of Drug Delivery Science and Technology 60 (2020) 101986

which sealed microparticulates interior. Interestingly, however, this acid phase. This may result in lower concentration of diclofenac sodium
phenomenon has a larger magnitude in the case of pellets compressed in the gastric lumen and reduce stomach irritation.
into tablets which might result from the very close contact of The tableting of drug-loaded enteric-coated pellets revealed slightly
drug-loaded pellets surface with excipients caused by compaction different resistance to compression and susceptibility to mechanical
forces. DCPA-based DR pellets behaved differently and after compres­ damages of investigated inner cores. It can be assumed, the use of starter
sion into tablets released more active substance. This could be explained pellets of smaller size would result in obtaining more rugged multi­
by a less effective sealing of the enteric-coating during maceration. It particulates, more suitable for formulation into tablets as reported
should be noted that the difference between tablets and capsules elsewhere [37,39,41].
formulation was not big and the amount of dissolved diclofenac sodium The dissolution rate of diclofenac sodium from developed formula­
in this case was at the same level as for MCC cores and furthermore, tions in neutral conditions mimicking these in the distal GIT was clearly
much lower than for both sugar- and isomalt-based multiparticulates. related to the solubility of the core material in water – very rapid for
The assumption about the sealing of the coating layer is to a certain soluble cores made of isomalt and sugar, slower for insoluble ones (MCC-
degree justifiable, taking into account the results shown in Fig. 9 where and DCPA-based).
four times shorter maceration in much smaller volume of diluted acid
did not cause such sealing effect and allowed much higher release of References
active substance from tablets when compared to the corresponding free
pellets (taken out from capsule shells). The highest difference was [1] N. Al-Hashimi, N. Begg, R.G. Alany, H. Hassanin, A. Elshaer, Oral modified release
multiple-unit particulate systems: compressed pellets, microparticles and
observed for DCPA-based microparticulates, nevertheless, it should be nanoparticles, Pharmaceutics (2018), https://doi.org/10.3390/
noted that the release of diclofenac sodium from the tablets containing pharmaceutics10040176.
calcium phosphate cores is lower that from other tested MUPS tablets. [2] I. Ghebre-Selassie, Pellets: a general overview I, in: Pharmaceutical Pelletization
Technology, Marcel Dekker, 1989.
For formulation containing MCC- and isomalt-based DR pellets, big error [3] S. Muley, T. Nandgude, S. Poddar, Extrusion–spheronization a promising
values can be spotted. In the first case, this may be due to the plastic pelletization technique: in-depth review, Asian J. Pharm. Sci. (2016), https://doi.
nature of the cores and their significant deformation during compression org/10.1016/j.ajps.2016.08.001.
[4] M. Zaman, S. Saeed-Ul-Hassan, R.M. Sarfraz, N. Batool, M.J. Qureshi, M.A. Akram,
resulting in a different coating thickness. For isomalt-based drug-loaded S. Munir, Z. Danish, Pellets and pelletization: emerging trends in the pharma
pellets this phenomenon is related to the high solubility of the core industry, Acta Pol. Pharm. - Drug Res. 73 (3) (2016) 1415–1425.
material in water. From the very beginning of the dissolution test, both [5] O. Luhn, N. Kállai, Z.K. Nagy, K. Kovács, B. Fritzsching, I. Klebovich, I. Antal,
Dissolution profile of novel composite pellet cores based on different ratios of
uncompressed and compressed multiparticulates were partly floating on
microcrystalline cellulose and isomalt, J. Pharm. Sci. (2012), https://doi.org/
the surface of the dissolution medium, where parts of the pellets were 10.1002/jps.23186.
not immersed in the medium. [6] R. Deb, A.B. Ahmed, Pellets and pelletization techniques: a critical review, Int. Res.
Comparing results of the dissolution carried out in buffers of neutral J. Pharm. (2016), https://doi.org/10.7897/2230-8407.04414.
[7] P. Tschopp, Starting pellets: materials, manufacturing methods, and applications,
pH using modified compendial method in a basket apparatus (see Fig. 6) Tablets & Capsules Magazine 13 (5) (2015) 10–13.
and the biorelevant method in a paddle apparatus (see Fig. 8), a very [8] R. Sidwell, J. Hansell, M. Rane, A.R. Rajabi-Siahboomi, Characterization of Inert
similar relationship between the drug release rate and the chemical Cores for Multiparticulate Dosage Forms, 2017, https://doi.org/10.1007/978-1-
4939-7012-4_2.
nature of inert cores was found: dissolution rate from water-soluble [9] D. Werner, Sugar spheres: a versatile excipient for oral pellet medications with
sugar- and isomalt-based multiparticulates was much faster than from modified release kinetics, Pharmaceut. Technol. Eur. 18 (4) (2006) 35–41.
insoluble ones. Furthermore, in the case of drug-loaded pellets based on [10] E. Gryczová, M. Rabišková, D. Vetchý, K. Krejčová, Pellet starters in layering
technique using concentrated drug solution, Drug Dev. Ind. Pharm. (2008),
water-soluble inert cores the release of diclofenac sodium from both https://doi.org/10.1080/03639040802130046.
tablets and hard gelatin capsules was insensitive to differences in hy­ [11] S.S. Al-Nimry, S.M. Assaf, I.M. Jalal, N.M. Najib, Adsorption of ketotifen onto some
drodynamic conditions generated in a paddle or a basket apparatus, pharmaceutical excipients, Int. J. Pharm. (1997), https://doi.org/10.1016/S0378-
5173(96)04857-0.
showing similarly fast dissolution rate, normally exceeding 85% of the [12] A.S. Narang, D. Desai, S. Badawy, Impact of excipient interactions on solid dosage
amount of the drug substance within first 15 min of the test. The form stability, Pharm. Res. (N. Y.) 29 (2012) 2660–2683, https://doi.org/10.1007/
dissolution rate from tablets containing MCC- and DCPA-based multi­ s11095-012-0782-9.
[13] S.L. Rivera, S. Ghodbane, In vitro adsorption-desorption of famotidine on
particulates was slightly slower and their release profiles had a similar
microcrystalline cellulose, Int. J. Pharm. (1994), https://doi.org/10.1016/0378-
shape. In the case of uncompressed DR pellets filled in capsules, a very 5173(94)90413-8.
high susceptibility of the drug release to the hydrodynamic conditions in [14] G. Livesey, Health potential of polyols as sugar replacers, with emphasis on low
an apparatus was found. Comparing the results obtained using the glycaemic properties, Nutr. Res. Rev. (2003), https://doi.org/10.1079/nrr200371.
[15] N. Kállai, O. Luhn, J. Dredán, K. Kovács, M. Lengyel, I. Antal, Evaluation of drug
basket and paddle methods, one can observe much slower dissolution release from coated pellets based on isomalt, sugar, and microcrystalline cellulose
rate for the latter method. This is due to the fact that the water-insoluble inert cores, AAPS PharmSciTech (2010), https://doi.org/10.1208/s12249-010-
pellets cores, especially DCPA-based ones, had a higher density and 9396-x.
[16] D. Zakowiecki, M. Lachmann, V. Schaum, T. Hess, Novel starter pellets based on
tended to lay on the bottom of dissolution vessel directly under the dibasic calcium phosphate anhydrous: properties and application, Express Pharma
paddles, where mixing intensity is normally the lowest. Interestingly 15 (1) (2019) 34–36.
though, under these conditions the release profile observed for calcium [17] S. Abdul, A.V. Chandewar, S.B. Jaiswal, A flexible technology for modified-release
drugs: multiple-unit pellet system (MUPS), J. Contr. Release (2010), https://doi.
phosphate pellets was very close to 0-order kinetics. org/10.1016/j.jconrel.2010.05.014.
[18] A. Dashevsky, K. Kolter, R. Bodmeier, Compression of pellets coated with various
5. Conclusions aqueous polymer dispersions, Int. J. Pharm. (2004), https://doi.org/10.1016/j.
ijpharm.2004.03.019.
[19] A.B. Bashaiwoldu, F. Podczeck, J.M. Newton, The application of non-contact laser
In the course of this study, delayed-release multiparticulate formu­ profilometry to the determination of permanent structural change induced by
lations of diclofenac sodium in the form of tablets and hard gelatin compaction of pellets: II. Pellets dried by different techniques, Eur. J. Pharmaceut.
Sci. (2004), https://doi.org/10.1016/j.ejps.2004.02.005.
capsules were developed. As a basis for the preparation of the multi­ [20] C. Seiler, Oral controlled release technology and development strategy, in:
particulates, the starter pellets containing sucrose, isomalt, microcrys­ Pharmaceutical Formulation: the Science and Technology of Dosage Forms,
talline cellulose and anhydrous dibasic calcium phosphate were used. It Croydon, Cambridge RSC, 2018, https://doi.org/10.1039/9781782620402.
[21] F. Zeeshan, K.K. Peh, Y.T.F. Tan, Exploring the potential of a highly compressible
has been shown that the core material significantly influenced the
microcrystalline cellulose as novel tabletting excipient in the compaction of
coating process, the properties of drug-loaded pellets as well as their extended-release coated pellets containing an extremely water-soluble model drug,
dissolution behavior. For diclofenac sodium, it seems more advanta­ AAPS PharmSciTech (2009), https://doi.org/10.1208/s12249-009-9278-2.
geous to use starter pellets based on water-insoluble substances such as
MCC or DCPA, which showed the lowest degree of the drug release in the

9
D. Zakowiecki et al. Journal of Drug Delivery Science and Technology 60 (2020) 101986

[22] V.S.N. Murthy Dwibhashyam, J. Ratna, Key formulation variables in tableting of [32] J. Van Den Abeele, J. Brouwers, R. Mattheus, J. Tack, P. Augustijns,
coated pellets, Indian J. Pharmaceut. Sci. (2008), https://doi.org/10.4103/0250- Gastrointestinal behavior of weakly acidic BCS Class II drugs in man - case study of
474X.45391. diclofenac potassium, J. Pharm. Sci. (2016), https://doi.org/10.1002/jps.24647.
[23] T. Chen, J. Li, T. Chen, C.C. Sun, Y. Zheng, Tablets of multi-unit pellet system for [33] Q. Wang, N. Fotaki, Y. Mao, Biorelevant dissolution: methodology and application
controlled drug delivery, J. Contr. Release (2017), https://doi.org/10.1016/j. in drug development, Dissolution Technol. (2009), https://doi.org/10.14227/
jconrel.2017.07.043. DT160309P6.
[24] J.L. Haslam, A.E. Forbes, G.S. Rork, T.L. Pipkin, D.A. Slade, D. Khossravi, Tableting [34] G. Garbacz, B. Kołodziej, M. Koziolek, W. Weitschies, S. Klein, A dynamic system
of controlled release multiparticulates, the effect of millisphere size and protective for the simulation of fasting luminal pH-gradients using hydrogen carbonate
overcoating, Int. J. Pharm. (1998), https://doi.org/10.1016/S0378-5173(98) buffers for dissolution testing of ionisable compounds, Eur. J. Pharmaceut. Sci. 51
00240-3. (2014), https://doi.org/10.1016/j.ejps.2013.09.020.
[25] B. Johansson, F. Nicklasson, G. Alderborn, Effect of pellet size on degree of [35] H.M. Fadda, H.A. Merchant, B.T. Arafat, A.W. Basit, Physiological bicarbonate
deformation and densification during compression and on compactability of buffers: stabilisation and use as dissolution media for modified release systems, Int.
microcrystalline cellulose pellets, Int. J. Pharm. (1998), https://doi.org/10.1016/ J. Pharm. (2009), https://doi.org/10.1016/j.ijpharm.2009.08.003.
S0378-5173(97)00355-4. [36] F. Liu, H.A. Merchant, R.P. Kulkarni, M. Alkademi, A.W. Basit, Evolution of a
[26] R.E. Small, Diclofenac sodium, Clin. Pharm. (1989), https://doi.org/10.2165/ physiological pH 6.8 bicarbonate buffer system: application to the dissolution
00003495-198835030-00004. testing of enteric coated products, Eur. J. Pharm. Biopharm. 78 (2011) 151–157,
[27] P.A. Todd, E.M. Sorkin, Diclofenac sodium: a reappraisal of its pharmacodynamic https://doi.org/10.1016/j.ejpb.2011.01.001.
and pharmacokinetic properties, and therapeutic efficacy, Drugs (1988), https:// [37] M. Babayeva, A.A. Marfo, R. Wolfe, Y.C.S. Hlaing, Z.G. Loewy, B.A. Selvi, Studying
doi.org/10.2165/00003495-198835030-00004. the factors that impact the dissolution characteristic of complex drug product,
[28] R.N. Brogden, R.C. Heel, G.E. Pakes, T.M. Speight, G.S. Avery, Diclofenac sodium: a Pharmaceut. Dev. Technol. 24 (2019), https://doi.org/10.1080/
review of its pharmacological properties and therapeutic use in rheumatic diseases 10837450.2019.1647544.
and pain of varying origin, Drugs (1980), https://doi.org/10.2165/00003495- [38] I. Partheniadis, P. Gkogkou, N. Kantiranis, I. Nikolakakis, Modulation of the release
198020010-00002. of a non-interacting low solubility drug from chitosan pellets using different pellet
[29] R. Altman, B. Bosch, K. Brune, P. Patrignani, C. Young, Advances in NSAID size, composition and numerical optimization, Pharmaceutics 11 (2019), https://
development: evolution of diclofenac products using pharmaceutical technology, doi.org/10.3390/pharmaceutics11040175.
Drugs (2015), https://doi.org/10.1007/s40265-015-0392-z. [39] S. Abdul, A.V. Chandewar, S.B. Jaiswal, A flexible technology for modified-release
[30] M. Tubic-Grozdanis, M.B. Bolger, P. Langguth, Application of gastrointestinal drugs: multiple-unit pellet system (MUPS), J. Contr. Release 147 (2010), https://
simulation for extensions for biowaivers of highly permeable compounds, AAPS J. doi.org/10.1016/j.jconrel.2010.05.014.
(2008), https://doi.org/10.1208/s12248-008-9023-x. [40] T.N. Hiew, Y.H. Tian, H.M. Tan, P.W.S. Heng, A mechanistic understanding of
[31] B. Jeganathan, V. Prakya, A. Deshmukh, Preparation and evaluation of diclofenac compression damage to the dissolubility of coated pellets in tablets, Eur. J. Pharm.
sodium tablet coated with polyelectrolyte multilayer film using hypromellose Biopharm. 146 (2020), https://doi.org/10.1016/j.ejpb.2019.11.006.
acetate succinate and polymethacrylates for pH-dependent, modified release drug [41] S.R. Béchard, J.C. Leroux, Coated pelletized dosage form: effect of compaction on
delivery, AAPS PharmSciTech (2016), https://doi.org/10.1208/s12249-015-0385- drug release, Drug Dev. Ind. Pharm. 18 (1992), https://doi.org/10.3109/
y. 03639049209052410.

10