ELLEN VERHOEVENVan den Mooter (Laboratorium voor Farmacotechnologie en Biofarmacie, Katholieke...

Ghent University

Faculty of Pharmaceutical Sciences

HOT-MELT EXTRUSION AS PROCESSING TECHNIQUE FOR MULTIPARTICULATE DOSAGE FORMS

CONTAINING LIPOPHILIC AND HYDROPHILIC POLYMERS

ELLEN VERHOEVEN Pharmacist

Thesis submitted to obtain the degree of Doctor in Pharmaceutical Sciences

2008

Promoters:

Prof. Dr. J.P. Remon

Prof. Dr. C. Vervaet

Laboratory of Pharmaceutical Technology – Ghent University

The author and the promoters give the authorization to consult and to copy parts of this

thesis for personal use only. Any other use is limited by the Laws of Copyright, especially

concerning the obligation to refer to the source whenever results are cited from this thesis.

Ghent, December 15th, 2008

The Promoters The Author

Prof. Dr. Jean Paul Remon Prof. Dr. Chris Vervaet Ellen Verhoeven

DANKJEWEL…

DANKJEWEL…

DANKJEWEL…

Mijn doctoraatsonderzoek is voltooid en mijn verhaal geschreven, maar dat zou nooit gelukt

zijn zonder de hulp, ervaring en steun van zovele boeiende mensen die ik heb mogen leren

kennen tijdens de afgelopen jaren, die elk hun steentje hebben bijgedragen op hun manier,

en een dankjewel meer dan verdienen.

Vooreerst gaat mijn oprechte dank uit naar mijn promotoren, Prof. Jean Paul Remon en Prof.

Chris Vervaet. Professor, bedankt dat u mij, na mijn specialisatie opleiding in de

Farmaceutische Biotechnologie, de kans bood om gedurende een jaar in samenwerking met

uw labo te werken als wetenschappelijk medewerker aan de Universiteit Gent voor de

farmaceutische industrie. Gedurende dat jaar heb ik veel geleerd over de ontwikkelingen die

in de farmaceutische industrie groeien, kon ik mijn kennis over farmaceutische technologie

verrijken, en kon ik de sfeer opsnuiven van wat doctoreren op uw labo inhoudt. Na dat

boeiend jaar, kwam de vraag ‘Wat verder nu?’ en u wist me over het onderzoek van ‘Hot-

melt extrusie voor het ontwikkelen van multiparticulaire geneesmiddelenvormen’ te vertellen.

Bedankt voor dat uitnodigend gesprek toen en me ervan te overtuigen dat ik van dit project

een interessante onderzoeksthesis zou kunnen maken. Ik wil u ook bedanken voor de

enthousiaste en stimulerende wijze waarop u dit onderzoek begeleidde, de leerrijke

discussies en uw positieve energie voor onderzoek en wetenschap. Chris, ook jouw

aanstekelijk enthousiasme en nauwe betrokkenheid bij het onderzoek heeft in belangrijke

mate bijgedragen tot deze doctoraatsthesis. Jou wil ik heel speciaal dankjewel zeggen voor

de vele uren kritisch nalezen van manuscripten en dit proefschrift, en de suggestie ‘to keep it

short and simple’. Dank aan jullie beiden om mij de opportuniteit te bieden tot presenteren

van dit werk, bijwonen van wetenschappelijke congressen en workshops in binnen- en

buitenland en mij te laten groeien in mijn werk. Mijn zin voor organisatie kon ik gedurende de

voorbije jaren ontplooien door het organiseren van de praktijksessies Artsenijbereidkunde en

Farmaceutische Technologie voor de studenten farmacie, een paar edities van de

wetenschapsweek, en enkele ‘all-round’ verantwoordelijkheden op het labo, iets waar ik mij

graag en met veel toewijding voor inzette. Dus dankjewel voor de unieke ervaring die jullie

mij hebben aangeboden en om in mij en het onderzoek te blijven geloven, zeker als de

onzekerheid weer eens ging toeslaan!

DANKJEWEL…

Een woordje van dank gaat ook uit naar een aantal mensen van andere diensten: Prof. Guy

Van den Mooter (Laboratorium voor Farmacotechnologie en Biofarmacie, Katholieke

Universiteit Leuven), bedankt voor uw inbreng bij de MTDSC experimenten en de

mogelijkheid te bieden analyses op uw labo uit te voeren, en Patrick Rombaut wil ik

bedanken voor het analyseren van de vele binaire mengsels (gelukkig was er de

autosampler waardoor de bezetting van jullie toestel voor mijn stalen tot een minimum kon

beperkt worden). Prof. Etienne Schacht (Vakgroep Organische Chemie – Polymer Chemistry

& Biomaterials Research Group, Universiteit Gent), hartelijk dank dat ik gedurende enkele

weken de infrastructuur van uw labo mocht gebruiken voor het uitvoeren van de

gelfiltratiechromatografie analyses, en uw verduidelijkingen wanneer het om het gedrag van

polymeren ging. Ook dank aan Olivier Janssens (Vakgroep Vaste-stofwetenschappen,

Universiteit Gent) voor de XRD opnames en een welgemeende dank aan Denis Van Loo

(Centrum voor X-stralen Tomografie - UGCT, Universiteit Gent) voor de beelden verkregen

via micro-CT.

A special thanks to Prof. Jürgen Siepmann and Florence Siepmann (College of Pharmacy,

Université de Lille 2), it was a great pleasure to collaborate with you. Thank you for all the

very interesting discussions in Ghent and Lille about the research work, for sharing your

knowledge about modelling with me and adding this dimension to my work, for careful

revising abstracts, for the follow-up of the project and your encouragement!

Ik wil hier ook het Laboratorium voor Farmaceutische Chemie en Geneesmiddelenanalyse

(Prof. Willy Baeyens) bedanken, en in het bijzonder Thomas De Beer voor zijn bijdrage bij

het uitvoeren van de Raman analyses. Thomas, bedankt voor de uitleg en verhelderende

discussies met betrekking tot dit werk, en voor jouw inspanningen om mijn stalen meermaals

overnacht te willen laten ‘runnen’. Ook dankjewel voor de leuke en ontspannende babbels!

Het Laboratorium voor Farmaceutische Microbiologie (Prof. Hans Nelis): Kristof De Prijck,

dankjewel dat ik met mijn ‘melk-kanneke’ geregeld achter een ‘soeplepel’ vloeibare stikstof

mocht komen, en voor de toffe babbels als we mekaar in ’t FFW eens tegenkwamen, over

ditjes en datjes en de vooruitgang van ons beider onderzoek.

Daniël Tensy, jou wil ik heel speciaal danken voor de deskundige uitvoering van de

dierenexperimenten en de vele staalnames, ook in de vroege en late uurtjes en weekends.

Het was fijn samenwerken om met jou de planningen hiervoor op te stellen en jouw

betrokkenheid en meeleven te voelen tijdens de vele uurtjes voorbereiding van de HPLC, het

extraheren en het afwachten tot ‘de goede piekjes’ op het scherm verschenen. Zelfs toen de

DANKJEWEL…

experimenten met de honden volbracht waren bleef je geïnteresseerd in het reilen en zeilen

van mijn project. Dankjewel ook voor de aangename gesprekjes tijdens de latere uurtjes in

het lab.

Voor hulp bij de statistische verwerking van de gegevens, wil ik Els Adriaens enorm

bedanken. Het betekende extra werk in je drukke schema, maar je hebt toch tijd gemaakt om

mij hiermee te helpen. En na onze intensieve SPSS-momenten, kon er natuurlijk wat

bijgekletst worden.

De thesisstudenten die ik begeleid heb in het kader van hun onderzoeksstage: Bob Van der

Jeugt, Joseph Nzalawahe, Inge Cnudde, Maïté Van Heuverswyn, Sérgio Manuel Campos da

Silva, Veronique Wauters en Kristof Vandenkerckhove, dankjewel voor jullie inzet en

interesse voor mijn doctoraat, jullie praktische assistentie en het plezier om samen te

werken!

Alle (ex)-collega’s voor de toffe werksfeer op het lab.

Special thanks to André Antunes for being there to solve computer problems and laughter in

the lab, I appreciated it a lot! Also your interest in ‘chauffage’ during my renovation project…

Katharine en Bruno voor al de SAP-, kopieer- en andere administratieve taken die jullie

hebben uitgevoerd, maar zeker ook voor de gezellige drukte op het secretariaat. Trientje,

dankjewel voor de opsmuk van enkele figuren, en Bruno om me te helpen als de computer

weer eens dwars lag.

Mijn eerste bureaugenoten Dieter en Jody, wil ik bedanken voor de leuke sfeer tijdens mijn

eerste jaar op het labo. Jody, het was boeiend om met jou samen te werken, en ook

dankjewel voor je steuntje in de rug om mijn doctoraat aan te vangen! Ook met Evy, Yves,

Joachim, Brenda, en later ook Evelien en Ana was het een aangename tijd op onze bureau,

dankjewel voor de leuke babbels en de boel op te vrolijken!

Delphine, dankjewel voor de leuke gesprekjes over allerlei, en nieuwtjes van in ons ‘rtc-

wereldje’.

Trientje, Bruno en Bruno (Germain en Gerard), Yves, Joachim, Els en Nathalie, bedankt voor

de ontspannende middagpauzes, de gevarieerde gespreksonderwerpen en de vele ‘jokes’,

de ene al wat bruiner dan de andere…

DANKJEWEL…

Mijn badmintonmaatjes: Katharine, Bruno en Bruno, Els en Nathalie, bedankt voor de vele

leuke enkeltjes en dubbeltjes! ’t Was soms nodig om de frustraties eruit te kloppen en er

werd ook heel wat gelachen op het badmintonplein!

Katharine, Els en Nathalie, jullie aanmoedigingen om het onderzoek naarstig verder te zetten

deden deugd, maar jullie wil ik ook heel erg bedanken omdat jullie er voor me waren voor de

babbels op persoonlijk vlak. Meer dan eens werd een lach en traan gedeeld, en deed het

deugd om ons hart te luchten! Dankjewel voor jullie steun en vriendschap! Els en Bruno, en

Nathalie, we hebben veel plezier gehad op onze skireisjes en ook de sfeer op etentjes en

andere avondjes uit was fun! Els en Bruno, we hebben ons kostelijk geamuseerd tijdens

onze ‘mojito-sortiekes’ in ‘Depot Central’ (vaak tot in de vroege uurtjes…) en hebben veel

leute gehad als we ‘een stapke gingen zetten’, maar ook dankjewel voor jullie hulp bij de

laatste ‘opsmukperikelen’ van de thesis.

Yves, fijn dat je mijn maatje was op het lab. We hadden veel plezier op onze bureau en

hebben heel wat afgelachen, dankjewel voor mijn assertiviteit aan te wakkeren, en voor nu

en dan eens een ‘vette sjief’ te draaien (we kunnen een mooi top 3-tje samenstellen van de

voorbije jaren: ons drummoment ‘In the Air Tonight’ van de Phil, een ‘Relax, Take It Easy’-

moment wisten we ook wel vaak te appreciëren en soms hadden we zin in een stevige

‘beat’). ’t Is dan wel niet mijnen favoriet, maar toch tof dat je mij hebt laten kennismaken met

het repertoire van d’ een André… Maar ook was er bij ons beiden vaak de frustratie dat het

tokkelen op de pc voor de publicaties niet altijd zo goed scheen te lukken, hoewel onze

taalvaardigheid en welbespraaktheid toch wel heel vlotjes zijn?! Dankjewel dat je een goede

vriend bent geweest en dat we ook lief en leed konden delen.

Al de vrienden en de ‘rtc-compagnie’, dankjewel voor jullie vriendschap de voorbije jaren, het

zorgen voor de nodige ontspanning op leuke uitjes en weekendjes, het enthousiasme om

samen met jullie allerlei activiteiten ineen te boksen en het plezier op vele feestjes en

avondjes uit! Ook dankjewel voor jullie interesse voor mijn werk en de aanmoedigingen de

laatste maanden! In goede of minder goede tijden, jullie zijn er voor me, dus dikke merci!

Tenslotte richt ik mijn belangrijkste dankwoordje aan Ina, mama en papa, bedankt voor alles,

voor de kansen die jullie me gegeven hebben en me te steunen in al mijn keuzes, om overal

en altijd voor me klaar te staan, mijn opvoeding, jullie bezorgdheid en betrokkenheid, om me

te leren altijd mezelf te blijven, nooit op te geven en hard te werken voor hetgeen ik wil

bereiken. Zonder jullie had ik de eindstreep nooit gehaald…

TABLE OF CONTENTS

TABLE OF CONTENTS INTRODUCTION Hot-melt extrusion and sustained-release

from multiparticulate polymeric dosage forms 3

OUTLINE AND AIMS 41

CHAPTER 1 Xanthan gum to tailor ibuprofen release:

in vitro and in vivo evaluation 47

CHAPTER 2 Xanthan gum to tailor metoprolol tartrate release:

influence of formulation and process parameters 95

CHAPTER 3 Polyethylene oxide to tailor metoprolol tartrate release:

in vitro and in vivo evaluation 139

CHAPTER 4 Mathematical modelling of drug and polymer release

from hot-melt extruded mini-matrices 179

SUMMARY AND GENERAL CONCLUSIONS 213

SAMENVATTING EN ALGEMEEN BESLUIT 221

CURRICULUM VITAE 229

1111

INTRODUCTION HOT-MELT EXTRUSION

AND SUSTAINED-RELEASE FROM MULTIPARTICULATE

POLYMERIC DOSAGE FORMS

3333

INTRODUCTION

4444

HOT-MELT EXTRUSION AND SUSTAINED-RELEASE FROM MULTIPARTICULATE POLYMERIC DOSAGE FORMS

INTRODUCTION HOT-MELT EXTRUSION

AND SUSTAINED-RELEASE FROM MULTIPARTICULATE

POLYMERIC DOSAGE FORMS

Hot-melt extrusion comprises the forcing of material through an opening (extrusion die) by

means of pressure. During the process, thermoplastic polymer(s), active ingredient(s) and other

processing aids are fed into a heated barrel and the polymer-drug blend is transferred through the

machine towards the die: the excipients melt or soften at the elevated temperature and the molten

mass is continuously kneaded and pressurized, and finally the homogeneous melt is pumped

through the die attached to the end of the barrel. As material passes through the die, the material

acquires the shape of the die opening and the material solidifies: the extruded product is referred

to as the extrudate. When the shape of the die is changed, the final product may take the form of

a film, cylinder or tube [1-4].

It is not the aim of this introduction to provide a complete literature review about hot-melt

extrusion, but the most important aspects about this processing technique as well as its use for

pharmaceutical applications will be discussed in the next paragraphs, as a motivation why this

production technique was selected to produce sustained-release mini-tablets.

Materials can be extruded in the molten state or in the solid state. It is called melt-fed

extrusion when the extruder is fed with molten polymer. Solid-state extrusion occurs when the

material is fed to the extruder as a solid; the material will melt or soften by the elevated

temperatures of the extruder screw as it is conveyed from the feed port to the die [2].

5555

INTRODUCTION

Extrusion process technology can be divided into two categories based on their mode of

operation: ram extrusion (discontinuous) and screw extrusion (continuous). Ram extrusion

operates with a positive displacement ram capable of generating high pressures to force material

through a shaping die, while screw extrusion consists of a rotating screw or set of screws inside a

heated barrel. During ram extrusion, materials are introduced into a heated cylinder. After an

induction period to soften the materials, a ram (or a piston) pressurizes the soft materials through

the die and transforms them into the desired shape (Figure 1). High pressure is the operating

principle of ram extrusion. The major drawback of ram extrusion is the limited melting capacity that

causes poor temperature uniformity in the extrudate. Also, extrudates prepared by ram extrusion

have lower homogeneity, in comparison with extrudates processed by screw extrusion [5].

Figure 1: Ram extrusion using a capillary die [5].

Unlike ram extrusion, a screw extruder provides more shear stress and intense mixing. A

distinction can be made based on the number of screws incorporated in the extruder, resulting in

single-screw, twin-screw and multi-screw extruders. As indicated by the name, the difference

between a single-screw and twin-screw extruder is whether one or two screws are utilized in the

machine. Also, they differ in the type of material transport and mixing abilities that take place in the

extruder: using a single-screw extruder, one screw rotates inside the barrel and is used for

feeding, melting and pumping the material in the direction of the die (a drag-induced type of

transport), whereas the transport in a twin-screw extruder is to some extent a positive

displacement type of transport, with a reduced degree of leakage flow since there is a possibility

6666


of exchanging material from one screw to the other. An important implication is that a stagnant

layer may remain at the screw surface: single-screw extruders are less efficient processors than

twin-screw extruders, and therefore have generally a longer equipment length. On the other hand,

single-screw extruders do have the advantage over twin-screw extruders in terms of their

mechanical simplicity, are less expensive and hence provide high productivity-to-cost ratios [6].

The twin-screw extruder has two agitator assemblies on parallel shafts and rotate together

with the same direction of rotation (co-rotating) or in the opposite sense (counter-rotating) (Figure

2). Co-rotating screws can rotate either clockwise or counterclockwise, and both directions are

equivalent from a processing point of view. If the two screws rotate in different directions, they can

either rotate towards the center, or rotate away from the center: using the counter-rotating screw

designs, the material is squeezed through the gap between the two screws as they come

together. Generally, counter-rotating screw extruders suffer from disadvantages of potential air

entrapment and low maximum screw speeds and output: these types of extruders are operated at

lower speeds because of the pressure which is developed from the outward pushing effect as the

screws come together in rotation. Co-rotating twin-screw extruders can be operated at higher

screw speeds and achieve higher outputs, while maintaining good mixing and conveying

characteristics. Both rotational designs with twin-screw extruders have the basic processing

advantages of positive conveying and effective mixing as compared with single-screw extruders

[6-9].

Figure 2: Twin-screw configurations: (top) intermeshing co-rotating twin-screw extruder, (bottom)

intermeshing counter-rotating twin-screw extruder [6].

7777

INTRODUCTION

These two types can be further separated into non-intermeshing and fully intermeshing.

For the latter, each agitator element wipes both the surface of the corresponding element on the

adjacent shaft and the internal surfaces of the mixing chamber, thereby preventing material from

rotating with the screw which is a significant advantage over single-screw extruders, where the

material is not being conveyed forward in an efficient manner. The agitators are said to be self-

wiping, an arrangement that eliminates stagnation areas within the mixing chamber and ensures a

narrow and well-defined residence time distribution. If the agitators are chosen not to intermesh,

the arrangement comes close to a single-screw set-up and have a lower degree of positive

conveying: the non-meshing types are also described as double-screws as they are not true twin-

screw extruders, they consist of essentially two single-screws side by side and work in a similar

way as single-screw extruders. In general, co-rotating shafts have better mixing capabilities as the

surfaces of the screws move towards each other. As the screws rotate, the flight of one screw

element wipes the flank of the adjacent screw, causing material to transfer from one screw to the

other. In this manner the material is transported along the extruder barrel [6-9]. Illustrations of the

two types (co-rotating and counter-rotating), both intermeshing, are shown in Figure 2.

Twin-screw extruders have several advantages over single-screw extruders, such as

easier material feeding, high kneading and dispersing capacities and less tendency to overheat

the material. In addition, twin-screw extruders have the advantage of a shorter residence time,

therefore, a smaller equipment size is required to achieve an equivalent output, and twin-screw

extruders are currently used to process larger quantities of material [6-10].

The extruder is typically composed of a feeding hopper, barrel, screw, die, screw driving

unit, and a heating/cooling device. Downstream equipment is used to collect the extrudates for

further processing. Monitoring devices on the equipment, used for performance and product

quality evaluation, include temperature gauges, a screw speed controller, an extrusion torque

monitor and pressure gauges. The extrusion channel is conventionally divided into three sections

along the length of the barrel: the feed zone (solid conveying zone), the transition zone (melting,

compression) and the metering zone, as shown in Figure 3.

8888


Figure 3: Schematic of an extruder illustrating various functional zones including the hopper, solid

conveying zone, melting zone, metering zone and die [11].

The starting material is fed from a hopper directly into the feed section, which has deeper

flights or flights of greater pitch (Figure 4).

Figure 4: Extrusion screw geometry [8].

This geometry enables the feed material to fall easily into the screw for conveying along

the barrel; pitch and helix angle determine the throughput. The channel depth is usually widest in

this section to facilitate mass flow and transfers the solid material forward as a solid plug into the

9999

INTRODUCTION

melting section where it is mixed, compressed, melted and plasticized. Compression is developed

by decreasing the pitch and channel depth, and entrapped air is removed by the increasing

pressure. The polymer gradually begins to soften and melts as it enters the melting section;

melting results from the heat transferred to the barrel from the heating devices as well as from

heat generated because of the shearing effect of the rotating screw(s). The melt moves by

circulation in a helical path when entering the metering section; the function of the metering zone

is to reduce the pulsating flow and to ensure a uniform delivery rate through the die cavity (a

variety of shapes is possible) [7-10].

Most extruders have a modular design, providing a choice of screw elements or

interchangeable sections which alter the configuration of the feed, transition and metering zones

(Figure 5). This makes it possible to modify the process to meet particular requirements and to

improve the melting process, from standard extrusion with low shear to high shear extrusion [7-9,

12, 13]. In an extrusion process, the dimensions of the screws are given in terms of L/D ratio,

which is the length of the screws divided by the diameter [14].

Figure 5: Screw and kneading elements [8].

In a hot-melt extrusion process, polymers are subjected to both thermal and shearing

stresses. The temperature can contribute to depolymerization of polymer chains, whereas

polymer chain scission may result from the shearing effects of the screw [4]. So one of the major

drawbacks of hot-stage extrusion is that it can only be applied when the products are thermally

stable. For a number of drugs, proteins and other excipients, hot-stage extrusion can not be used

unless a plasticizer is added in order to reduce the viscosity of the mixture in the extruder and

therefore to lower the process temperature settings [15]. But, as the residence time in the extruder

is rather short (up to 2-5 min, depending on the feed rate and screw speed) and the temperature

of all the barrels are independent and can be accurately controlled from low temperatures to high

temperatures, degradation by heat can be minimized [7].

10101010


For pharmaceutical applications, it can be concluded that this technique offers many

advantages over conventional pharmaceutical production methods. Neither solvents nor water are

used for processing since the molten polymer can function as a thermal binder; due to the

anhydrous nature of the process any potential drug degradation from hydrolysis can be avoided.

Therefore, fewer processing steps are needed and time-consuming drying steps are eliminated.

When used as a molding technique, there are no requirements on the compressibility of the

materials used in the formulation, and the entire process is simple, continuous and efficient. The

intense mixing and agitation during processing cause suspended drug particles to de-aggregate in

the molten polymer, resulting in a more uniform dispersion of fine particles. Furthermore, the

bioavailability of the drug substance may be improved when it is solubilized or dispersed at the

molecular level in hot-melt extruded dosage forms. Its simplicity on an industrial scale makes it

popular with pharmaceutical manufacturers [16, 17]. The popularity of continuous processes is

increasing in the pharmaceutical industry [18]. This way of processing offers several advantages,

such as improved process efficiency due to reduction of costs and time. Furthermore, scale-up is

avoided as the amount of material processed can be increased by simply prolonging the process

time [19].

Hot-melt extrusion originates from the polymer and food industry [7-9, 16], but several

research teams explored the possibilities of hot-melt extrusion as an alternative method to

develop pharmaceutical dosage forms such as granules, sustained-release pellets, matrix tablets,

matrix-in-cylinder systems, capsules and implants. Hot-melt extrusion also has applications in

topical drug delivery. In Table 1, an overview is given of hot-melt extrusion used as processing

technique for pharmaceutical applications: it lists the active drug substances, carriers, extruded

dosage forms and the references to literature sources. Crowley and co-authors also provided a

partial listing of some of the drug substances and carriers that have been formulated and

processed with hot-melt extrusion [9, 20].

11111111

INTRODUCTION

Ref

eren

ce

21-2

3

24

25

26, 2

7

28, 2

9, 3

0

13

30

31

32

11, 3

3

34, 3

5

36

37

Car

rier

carn

auba

wax

agar

, mic

rocr

ysta

lline

cellu

lose

, pol

yeth

ylen

e ox

ide,

Eud

ragi

t® L

100

wax

(Pre

ciro

l® A

TO 5

, Ste

rote

x® K

)

Eud

ragi

t® R

S P

O

α-la

ctos

e m

onoh

ydra

te

lact

ose,

cor

n st

arch

, hyd

roxy

prop

ylce

llulo

se, p

olyv

inyl

alc

ohol

α-la

ctos

e m

onoh

ydra

te

α-la

ctos

e m

onoh

ydra

te

poly

viny

lpyr

rolid

one,

hyd

roxy

prop

ylm

ethy

lcel

lulo

se, p

olox

amer

407

, po

lyet

hyle

ne o

xide

ethy

lcel

lulo

se, c

ellu

lose

ace

tate

but

yrat

e, E

udra

git®

RS

PM,

poly

(eth

ylen

e-co

-vin

yl a

ceta

te)

mic

rocr

ysta

lline

cellu

lose

, Eud

ragi

t® 4

135

F

Acry

l-EZE

®, C

arbo

pol®

974

P, M

etho

cel®

K4M

ethy

lcel

lulo

se, p

olye

thyl

ene

oxid

e

Dru

g

dicl

ofen

ac s

odiu

m

phen

ylpr

opan

olam

ine

HC

l

hydr

ochl

orot

hiaz

ide

sodi

um g

uaia

zule

ne s

ulfo

nate

ribof

lavi

n (s

odiu

m p

hosp

hate

)

dica

lciu

m p

hosp

hate

anh

ydra

te

itrac

onaz

ole

dilti

azem

HC

l

theo

phyl

lin

guai

fene

sin

Tabl

e 1:

Ove

rvie

w o

f diff

eren

t hot

-mel

t ext

rude

d do

sage

form

s.

Dos

age

form

Gra

nule

s

Pel

lets

12121212


38

39

4, 4

0

26, 2

7, 4

1-43

43

44

43

42

42

45

46, 4

7

48

49-5

1

39, 5

2

36

53, 5

4

glyc

eryl

fatty

aci

d es

ters

, pol

yeth

ylen

e gl

ycol

fatty

aci

d es

ters

poly

ethy

lene

, cel

lulo

se a

ceta

te b

utyr

ate,

pol

yvin

yl a

ceta

te, p

olyc

apro

lact

one

poly

ethy

lene

oxi

de

Eud

ragi

t® R

S P

O

Eud

ragi

t® E

PO

chito

san,

xan

than

gum

, pol

yeth

ylen

e ox

ide,

mic

rocr

ysta

lline

cel

lulo

se

Eud

ragi

t® R

S P

O, E

udra

git®

E P

O

Eud

ragi

t® R

S P

O

Eud

ragi

t® R

S P

O

α-la

ctos

e m

onoh

ydra

te

ethy

lcel

lulo

se, h

ydro

xypr

opyl

met

hyce

llulo

se, x

anth

an g

um

Eud

ragi

t® R

S PO

, mic

rocr

ysta

lline

cel

lulo

se

corn

sta

rch

poly

ethy

lene

, cel

lulo

se a

ceta

te b

utyr

ate,

pol

yvin

yl a

ceta

te, p

olyc

apro

lact

one

Acry

l-EZE

®, C

arbo

pol®

974

P, M

etho

cel®

K4M

glyc

eryl

pal

mito

stea

rate

, gly

cery

l trim

yris

tate

chlo

rphe

nira

min

e m

alea

te

acet

ohyd

roxa

mic

aci

d

dilti

azem

HC

l

indo

met

haci

n

hydr

ochl

orot

hiaz

ide

ibup

rofe

n

theo

phyl

line

(mon

ohyd

rate

)

Tabl

ets

13131313

INTRODUCTION

39

55

56

57

58

59

60

61

62

62

62

63

64

61

61, 6

5, 6

6

poly

ethy

lene

, cel

lulo

se a

ceta

te b

utyr

ate,

pol

yvin

yl a

ceta

te, p

olyc

apro

lact

one

ethy

lcel

lulo

se

Acry

l-EZE

®, E

udra

git®

L 1

00-5

5

®®

Eud

ragi

t S

100

, Eud

ragi

t L

100

poly

ethy

lene

oxi

de

Eud

ragi

t® L

100

corn

sta

rch

ethy

lcel

lulo

se, h

ydro

xypr

opyl

met

hyce

llulo

se, m

ethy

lcel

lulo

se,

G,

Sup

poci

re® A

P, S

uppo

cire

® B

P, W

iteps

ol® H

15, W

iteps

ol® E

85

eluc

ire® 4

4/14

, Gel

ucire

® 5

0/13

, Ovu

cire

® 3

260,

Ovu

cire

® 3

460

silic

one

silic

one

silic

one

colla

gen

silic

one

ethy

lcel

lulo

se, h

ydro

xypr

opyl

met

hyce

llulo

se, m

ethy

lcel

lulo

se,

eluc

ire® 4

4/14

, Gel

ucire

® 5

0/13

, Ovu

cire

® 3

260,

Ovu

cire

® 3

460

G,

Sup

poci

re® A

P, S

uppo

cire

® B

P, W

iteps

ol® H

15, W

iteps

ol® E

85

ethy

lcel

lulo

se, h

ydro

xypr

opyl

met

hyce

llulo

se, m

ethy

lcel

lulo

se,

G,

eluc

ire® 4

4/14

, Gel

ucire

® 5

0/13

, Ovu

cire

® 3

260,

Ovu

cire

® 3

460

Sup

poci

re® A

P, S

uppo

cire

® B

P, W

iteps

ol® H

15, W

iteps

ol® E

85

salic

ylic

aci

d

guai

fene

sin

5-am

inos

alic

ylic

aci

d

nife

dipi

ne

keto

prof

en

theo

phyl

line

(mon

ohyd

rate

)

indo

met

haci

n

antip

yrin

e

keto

prof

en

reco

mbi

nant

hum

an g

ranu

locy

te

colo

ny-s

timul

atin

g fa

ctor

iver

mec

tin

hydr

ochl

orot

hiaz

ide

prop

rano

lol H

Cl

Dou

ble

mat

rix

14141414


67

68, 6

9

70

70

71

72

72

73, 7

4

75

73

76

76, 7

7

77-8

1

82, 8

3

84

hydr

oxyp

ropy

lmet

hylc

ellu

lose

ace

tate

suc

cina

te, p

olyv

inyl

ace

tate

pht

hala

te

poly

(D,L

-lact

ide-

co-g

lyco

lide)

poly

(D,L

-lact

ide)

, pol

y(ε-

capr

olac

tone

)

poly

(D,L

-lact

ide)

, pol

y(ε-

capr

olac

tone

)

poly

(L-la

ctic

aci

d)

silic

one

silic

one

hydr

oxyp

ropy

lcel

lulo

se

hydr

oxyp

ropy

lcel

lulo

se, p

olye

thyl

ene

oxid

e

hydr

oxyp

ropy

lcel

lulo

se

poly

ethy

lene

oxi

de

poly

ethy

lene

oxi

de

poly

ethy

lene

oxi

de, h

ydro

xypr

opyl

cellu

lose

hydr

oxyp

ropy

lcel

lulo

se, p

olye

thyl

ene

oxid

e

hydr

oxyp

ropy

lcel

lulo

se

hydr

alaz

ine

mel

anot

an

ethi

dium

bro

mid

e

isom

etam

idiu

m c

hlor

ide

vapr

eotid

e pa

moa

te

hum

an s

erum

alb

umin

inte

rfero

n

chlo

rphe

nira

min

e m

alea

te

vita

min

E T

PGS

hydr

ocor

tison

e

guai

fene

sin

keto

prof

en

clot

rimaz

ole

keto

cona

zole

itrac

onaz

ole

Cap

sule

s

Impl

ants

Film

s

15151515

INTRODUCTION

85

3 3

86-8

8

89

86, 8

7

86

12, 9

0

86

86

91

1

5, 9

2

93

hydr

oxyp

ropy

lcel

lupy

lmet

hylc

ellu

lose

lo

se, h

ydro

xypr

o

Eud

ragi

t® E

100

Eud

ragi

t® E

100

poly

viny

lpyr

rolid

one,

pol

yvin

ylpy

rrol

idon

e-co

-vin

yl a

ceta

te

Eud

ra L

100

poly

viny

lpyr

roe-

co-v

inyl

ace

tate

git®

RL

PO

, Eud

ragi

t® R

D 1

00, E

udra

git®

S 1

00, E

udra

git®

lidon

e, p

olyv

inyl

pyrr

olid

on

poly

viny

lpyr

rolid

one

hydr

oxyp

ropy

lmet

hyl

late

/ ac

etat

e su

ccin

ate

cellu

lose

pht

ha

poly

viny

lpyr

rolid

one

poly

viny

lpyr

rolid

one

ethy

lcel

lulo

se

hydr

oxyp

ropy

lmet

hylc

ellu

lose

, pol

yvin

ylpy

rrol

idon

e-co

-vin

yl a

ceta

te, E

udra

git®

E 1

00

hydr

oxyp

ropy

lmet

hylc

ellu

lose

, pol

yvin

ylpy

rrol

idon

e-co

-vin

yl a

ceta

te, E

udra

git®

E P

O

lidoc

ai (H

Cl)

ne

diph

e

indo

met

haci

n

nhyd

ram

ine

HC

l

laci

dipi

ne

nife

dipi

ne

piro

xica

m

tolb

utam

ide

ibup

rofe

n

itrac

onaz

ole

nim

odip

ine

Sol

id d

ispe

rsio

ns

16161616


Oral solid dosage forms are available as single-unit (non-divided formulation) or

multiple-unit (divided formulation) forms which can be used for immediate- or sustained-

release drug delivery. Multiple-unit dosage forms consist of numerous sub-units which are

filled into capsules or sachets, or compressed into tablets which disintegrate instantaneously

[94, 95]. The sub-units of these multiparticulate dosage forms are granules, pellets or mini-

tablets. Mini-tablets are tablets with a diameter equal to or smaller than 2-3 mm [96].

Multiple-unit dosage forms offer a high degree of flexibility during the design and

development of oral dosage forms: they can be divided into desired doses without

formulation and process changes, and sub-units with different release profiles can be

administered simultaneously. Multiparticulates also have therapeutic advantages over single-

units, such as a more predictable gastric emptying which is less dependent on the nutritional

state: particles smaller than 2-3 mm are rapidly emptied from the stomach regardless of the

feeding state of the patient and the influence of gastric emptying rate on the upper gastro-

intestinal transit time of mini-tablets is minimized. Furthermore, mini-matrices have a high

degree of dispersion in the digestive tract (minimizing the risk of high local drug

concentrations which could damage the mucosa), less variance in transit time through the

gastro-intestinal tract, less inter- and intra-subject variability (owing to the improvement in

both transit reproducibility and dispersion in the digestive tract), and a lower risk of dose-

dumping [94, 95, 97-100].

Intensive research work is conducted on developing pellets as multiparticulate dosage

form by the classical way of extrusion and spheronization [101-108]. Extrusion/spheronization is

one of the most established techniques for the production of pellets with high quality, but due to

the fact that it is a multi-step process, a number of process parameters should be controlled.

Furthermore, formulation development is of special concern since the moistened mass needs to

possess specific characteristics in order to be successfully extruded and spheronized, dried and

coated. Therefore, for this thesis, mini-tablets (3 mm diameter, 2 mm height) will be produced

instead of pellets, using a hot-melt twin-screw extruder as continuous processing technique since

mixing, melting, homogenizing and shaping are carried out in a single step. The multiple-unit

dosage forms were obtained by cutting the extrudates manually into small cylinders using surgical

blades and are called mini-matrices. For larger scale batches, this cutting can also be performed

mechanically [109]: the strand discharged from the die is collected on a moving belt, eventually

cooled by air cooling, followed by cutting the extrudate using a sizing machine. If the extrudate

needs to be cut to a predetermined length, this can be accomplished with several downstream

equipments such as a guillotine cutter or rotating knife cutter (a blade is mounted to a disc which

is driven by a servomotor to cycle the blade in rotary motion) (Figure 6).

17171717

INTRODUCTION

(a) (b)

Knife

Nozzle

Extruder screw

Extruder

Polymer melt

(c) (d)

Figure 6: Cut-to-length devices: (a) guillotine cutter, (b) chopper to pelletize tablets [9], (c) rotating

knife cutter after extrudate is cooled on a moving belt, (d) rotating knife cutter immediately after

extrusion.

With conventional dosage forms, high peak blood concentrations may be reached soon

after administration with possible adverse effects related to the high concentration. Repeated

administration of a drug may cause serious side effects, which in many cases necessitates the

patient to stop taking medication. Sustained-release formulations are usually intended to optimize

a therapeutic regimen by providing slow and continuous drug delivery over the entire dosing

interval in combination with reduced side effects, whilst also providing greater patient compliance

and convenience by reducing the number of daily doses [110].

18181818


The classical way to design an oral sustained-release dosage form is by coating capsules,

tablets or pellets with insoluble, pH-independent polymers (reservoir type), whereby the

composition and thickness of the coat determine the drug release pattern. However, coating is a

time-consuming and expensive process with possible problems related to reproducibility of drug

release and dose-dumping. Therefore, another common sustained drug delivery system is the

matrix type, whereby the drug is uniformly dissolved or dispersed throughout the rate-controlling

polymer and drug release is determined by the excipients incorporated into the formulation and

not by an outer polymeric membrane. The matrix principle can be applied to single-unit as well as

to multiple-unit dosage forms, the latter having definite advantages, as described above.

Different strategies have been investigated in an attempt to sustain drug release from

matrix dosage forms. One strategy is to restrict the matrix surface area exposed to the release

medium, leading to a decreased matrix hydration and swelling rate and consequently a slower

drug release rate (e.g. multi-layered hydrophilic tablets [111-115], partial coating of the matrix with

an impermeable film to restrict matrix swelling [116], a core-in-cup tablet (an inert cup into which a

drug-containing matrix layer was compressed, thus leading to a matrix with only one side in

contact with the surrounding medium) [117], RingCap®, an oral controlled-release drug delivery

system consisting of a capsule-shaped matrix tablet to which bands of insoluble material are

applied circumferentially to the surface of the tablet (these bands modified drug release by

controlling the surface area, which changed over time as the area around the bands became

hydrated and eroded creating new surface area) [118-120], matrix-in-cylinder systems or covered-

rod-type formulations (a hydrophilic matrix tablet placed in an impermeable cylinder, reducing drug

release to the open ends of the cylinder) [61-66, 121], etc.). Varying the geometry is also a

possible means of altering release kinetics from matrix systems: matrix systems having spherical,

cylindrical, biconvex, donut shapes, etc. [122-128].

However, for this thesis, instead of manipulating the surface available for drug release in

order to have sustained drug release, mini-matrices will be produced. Therefore, the ultimate goal

of this project is to obtain sustained drug release using polymers which decrease drug dissolution

and diffusion/erosion since we deal with a large surface area from the multiparticulate dosage

form under investigation. Matrix tablets composed of drug and the release-retarding material

(polymer) offer a simple approach to design a sustained-release system. The materials that are

commonly used include: (a) hydrophobic materials, (b) insoluble, erodible materials or (c)

hydrophilic polymers [129].

The mechanism of drug release from the first two classes (a and b) of insoluble polymers

is considered to be diffusion of the drug from the matrix as a result of liquid penetration, therefore

19191919

INTRODUCTION

hydrophobic matrix carriers (e.g. ethylcellulose, glycerides, waxes) have been investigated as

drug carriers [21-23, 25, 53, 54, 130-132].

In contrast, especially hydrophilic polymer-based dosage forms are very popular and well-

described since drug release from these matrices is controlled by a combination of polymer

swelling, front movement, drug dissolution and diffusion through the hydrated gel, and matrix

erosion [133]. Despite their high water solubility, they are very successful in sustaining drug

release: upon contact with the surrounding medium, the hydrophilic polymer swells and develops

a highly viscous surface barrier which controls the liquid penetration into and the drug release

from the system, thus, retarding drug diffusion. Until now, hydroxypropylmethylcellulose has been

widely studied, reviewed by many authors and is the excipient chosen by most formulators for

hydrophilic matrix systems [134-143] as well as polyethylene oxides [40, 58, 76, 80, 144-150].

Also natural materials were extensively investigated as hydrophilic matrix carrier: xanthan gum

[44, 151-167], galactomannan [156, 159, 168], guar gum [169-173] and carrageenan [174-178].

However, drug release from a hydrophilic matrix is generally characterized by a burst

effect; the drug present at the surface is quickly released, then with time, as the diffusion path

length increases the release rate is progressively reduced. Therefore, it is often desirable to mix

different polymers in an optimum ratio to synchronize the movement of the swelling and erosion

fronts in the tablets [179-182]. Only few investigators performed research on the mechanism of

drug release from solid dispersions prepared with water insoluble and water soluble polymers,

combining ethylcellulose and hydroxypropylmethylcellulose [46, 47, 183]. No detailed

investigations were conducted on the drug release from matrix tablets formulated with

ethylcellulose as carrier material with xanthan gum or polyethylene oxides to tailor drug release.

Combining the advantages of hot-melt extrusion as processing technique and sustained-

release multiparticulate dosage forms – to achieve a plasma concentration-time profile within the

optimal therapeutic range in combination with reduced side effects – resulted in the production of

a sustained-release multiple-unit matrix dosage form manufactured by hot-melt extrusion. The

challenge of this doctoral thesis is to sustain drug release from the mini-matrices using a

combination of hydrophobic (ethylcellulose) and hydrophilic polymers (xanthan gum, polyethylene

oxides): hydrophilic polymers will be added to obtain zero-order release kinetics with complete

drug release within 24 h, by making the hydrophobic drug carrier more accessible to the release

medium, without resulting in burst release.

20202020


REFERENCES [1] D.H. Morton-Jones (ed.), Polymer processing, Chapman and Hall, London, UK (1989).

[2] C. Rauwendaal (ed.), Polymer extrusion, Hanser, München, Germany (1994).

[3] C. Aitken-Nichol, F. Zhang, J.W. McGinity, Hot-melt extrusion of acrylic films, Pharm.

Res. 13 (1996) 804-808.

[4] F. Zhang, J.W. McGinity, Properties of sustained-release tablets prepared by hot-melt

extrusion, Pharm. Dev. Technol. 4 (1999) 241-250.

[5] J. Perdikoulias, T. Dobbie, Die design, in: Pharmaceutical extrusion technology –

Drugs and the pharmaceutical sciences, Vol. 133 (I. Ghebre-Sellassie, C. Martin, ed.),

Marcel Dekker, New York, USA (2003) 99-109.

[6] M. Mollan, Historical overview, in: Pharmaceutical extrusion technology – Drugs and

the pharmaceutical sciences, Vol. 133 (I. Ghebre-Sellassie, C. Martin, ed.), Marcel

Dekker, New York, USA (2003) 1-18.

[7] J. Breitenbach, Melt extrusion: from process to drug delivery technology, Eur. J.

Pharm. Biopharm. 54 (2002) 107-117.

[8] R. Chokshi, H. Zia, Hot-melt extrusion technique: a review, Iran. J. Pharm. Res. 3

(2004) 3-16.

[9] M.M. Crowley, F. Zhang, M.A. Repka, S. Thumma, S.B. Upadhye, S.K. Battu, J.W.

McGinity, C. Martin, Pharmaceutical applications of hot-melt extrusion: part I, Drug

Dev. Ind. Pharm. 33 (2007) 909-926.

[10] J.W. McGinity, F. Zhang, Melt-extruded controlled-release dosage forms, in:

Pharmaceutical extrusion technology – Drugs and the pharmaceutical sciences, Vol.

133 (I. Ghebre-Sellassie, C. Martin, ed.), Marcel Dekker, New York, USA (2003) 183-

208.

21212121

INTRODUCTION

[11] N. Follonier, E. Doelker, E.T. Cole, Various ways of modulating the release of diltiazem

hydrochloride from hot-melt extruded sustained release pellets prepared using

polymeric materials, J. Control. Release 36 (1995) 243-250.

[12] K. Nakamichi, T. Nakano, H. Yasuura, S. Izumi, Y. Kawashima, The role of the

kneading paddle and the effects of screw revolution speed and water content on the

preparation of solid dispersions using a twin-screw extruder, Int. J. Pharm. 241 (2002)

203-211.

[13] K. Nakamichi, T. Nakano, H. Yasuura, S. Izumi, Y. Kawashima, Stabilization of sodium

guaiazulene sulfonate in granules for tableting prepared using a twin-screw extruder,

Eur. J. Pharm. Biopharm. 56 (2003) 347-354.

[14] R. Steiner, Extruder design, in: Pharmaceutical extrusion technology – Drugs and the

pharmaceutical sciences, Vol. 133 (I. Ghebre-Sellassie, C. Martin, ed.), Marcel Dekker,

New York, USA (2003) 19-37.

[15] G. Verreck, A. Decorte, K. Heymans, J. Adriaensen, D. Cleeren, A. Jacobs, D. Liu, D.

Tomasko, A. Arien, J. Peeters, P. Rombaut, G. Van den Mooter, M.E. Brewster, The

effect of pressurized carbon dioxide as a temporary plasticizer and foaming agent on

the hot stage extrusion process and extrudate properties of solid dispersions of

itraconazole with PVP-VA 64, Eur. J. Pharm. Sci. 26 (2005) 349-358.

[16] J.W. McGinity, F. Zhang, M.A. Repka, J.J. Koleng, Hot-melt extrusion as a

pharmaceutical process, Am. Pharm. Rev. 4 (2001) 25-36.

[17] J.W. McGinity, J.J. Koleng, M.A. Repka, F. Zhang, Hot-melt extrusion technology, in:

Encyclopedia of pharmaceutical technology, Vol. 19 (J. Swarbrick and J.C. Boylan,

ed.), Marcel Dekker, New York, USA (2001) 203-226.

[18] H. Leuenberger, New trends in the production of pharmaceutical granules: batch

versus continuous processing, Eur. J. Pharm. Biopharm. 52 (2001) 289-296.

[19] C. Vervaet, J.P. Remon, Continuous granulation in the pharmaceutical industry, Chem.

Eng. Sci. 60 (2005) 3949-3957.

22222222


[20] M.A. Repka, S.K. Battu, S.B. Upadhye, S. Thumma, M.M. Crowley, F. Zhang, C.

Martin, J.W. McGinity, Pharmaceutical applications of hot-melt extrusion: part II, Drug

Dev. Ind. Pharm. 33 (2007) 1043-1057.

[21] Y. Miyagawa, T. Okabe, Y. Yamaguchi, M. Miyajima, H. Sato, H. Sunada, Controlled

release of diclofenac sodium from wax matrix granule, Int. J. Pharm. 138 (1996) 215-

224.

[22] H. Sato, Y. Miyagawa, Dissolution mechanism of diclofenac sodium wax matrix

granules, J. Pharm. Sci. 86 (1997) 929-934.

[23] Y. Miyagawa, H. Sato, T. Okabe, T. Nishiyama, M. Miyajima, H. Sunada, In vivo

performance of wax matrix granules prepared by a twin-screw compounding extruder,

Drug Dev. Ind. Pharm. 25 (1999) 429-435.

[24] J.G. Lyons, D.M. Devine, J.E. Kennedy, L.M. Geever, P. O’Sullivan, C.L.

Higginbotham, The use of agar as a novel filler for monolithic matrices produced using

hot-melt extrusion, Eur. J. Pharm. Biopharm. 64 (2006) 75-81.

[25] J. Liu, F. Zhang, J.W. McGinity, Properties of lipophilic matrix tablets containing

phenylpropanolamine hydrochloride prepared by hot-melt extrusion, Eur. J. Pharm.

Biopharm. 52 (2001) 181-190.

[26] Y. Zhu, N.H. Shah, A.W. Malick, M.H. Infeld, J.W. McGinity, Influence of thermal

processing on the properties of chlorpheniramine maleate tablets containing an acrylic

polymer, Pharm. Dev. Technol. 7 (2002) 481-489.

[27] Y. Zhu, N.H. Shah, A.W. Malick, M.H. Infeld, J.W. McGinity, Solid-state plasticization of

an acrylic polymer with chlorpheniramine maleate and triethyl citrate, Int. J. Pharm. 241

(2002) 301-310.

[28] E.I. Keleb, A. Vermeire, C. Vervaet, J.P. Remon, Continuous twin screw extrusion for

the wet granulation of lactose, Int. J. Pharm. 239 (2002) 69-80.

[29] E.I. Keleb, A. Vermeire, C. Vervaet, J.P. Remon, Twin screw granulation as a simple

and efficient tool for continuous wet granulation, Int. J. Pharm. 273 (2004) 183-194.

23232323

INTRODUCTION

[30] B. Van Melkebeke, C. Vervaet, J.P. Remon, Validation of a continuous granulation

process using a twin-screw extruder, Int. J. Pharm. 356 (2008) 224-230.

[31] D. Djuric, B. Van Melkebeke, P. Kleinebudde, J.P. Remon, C. Vervaet, Comparison of

two twin-screw extruders for continuous granulation, Eur. J. Pharm. Biopharm., doi:

10.1016/j.ejpb.2008.06.033.

[32] D.A. Miller, J.T. McConville, W. Yang, R.O. Williams, J.W. McGinity, Hot-melt extrusion

for enhanced delivery of drug particles, J. Pharm. Sci. 96 (2007) 361-376.

[33] N. Follonier, E. Doelker, E.T. Cole, Evaluation of hot-melt extrusion as a new technique

for the production of polymer-based pellets for sustained-release capsules containing

high loadings of freely soluble drugs, Drug Dev. Ind. Pharm. 20 (1994) 1323-1339.

[34] C.R. Young, J.J. Koleng, J.W. McGinity, Production of spherical pellets by a hot-melt

extrusion and spheronization process, Int. J. Pharm. 242 (2002) 87-92.

[35] C.R. Young, J.J. Koleng, J.W. McGinity, Properties of drug-containing spherical pellets

produced by a hot-melt extrusion and spheronization process, J. Microencapsul. 20

(2003) 613-625.

[36] C.R. Young, C. Dietzsch, M. Cerea, T. Farrell, K.A. Fegely, A. Rajabi-Siahboomi, J.W.

McGinity, Physicochemical characterization and mechanisms of release of theophylline

from melt-extruded dosage forms based on a methacrylic acid copolymer, Int. J.

Pharm. 301 (2005) 112-120.

[37] C.R. Young, M. Crowley, C. Dietzsch, J.W. McGinity, Physicochemical properties of

film-coated melt-extruded pellets, J. Microencapsul. 24 (2007) 57-71.

[38] W. Prapaitrakul, O.L. Sprockel, P. Shivanand, Release of chlorpheniramine maleate

from fatty acid ester matrix disks prepared by melt extrusion, J. Pharm. Pharmacol. 43

(1991) 377-381.

[39] O.L. Sprockel, M. Sen, P. Shivanand, W. Prapaitrakul, A melt extrusion process for

manufacturing matrix drug delivery systems, Int. J. Pharm. 155 (1997) 191-199.

24242424


[40] M.M. Crowley, F. Zhang, J.J. Koleng, J.W. McGinity, Stability of polyethylene oxide in

matrix tablets prepared by hot-melt extrusion, Biomaterials 23 (2002) 4241-4248.

[41] Y. Zhu, N.H. Shah, A.W. Malick, M.H. Infeld, J.W. McGinity, Influence of a lipophilic

thermal lubricant on the processing conditions and drug release properties of

chlorpheniramine maleate tablets prepared by hot-melt extrusion, J. Drug Del. Sci.

Tech. 14 (2004) 313-318.

[42] Y. Zhu, K.A. Mehta, J.W. McGinity, Inlfuence of plasticizer level on the drug release

from sustained release film coated and hot-melt extruded dosage forms, Pharm. Dev.

Technol. 11 (2006) 285-294.

[43] M. Fukuda, N.A. Peppas, J.W. McGinity, Floating hot-melt extruded tablets for

gastroretentive controlled drug release system, J. Control. Release 115 (2006) 121-

129.

[44] M. Fukuda, N.A. Peppas, J.W. McGinity, Properties of sustained release hot-melt

extruded tablets containing chitosan and xanthan gum, Int. J. Pharm. 310 (2006) 90-

100.

[45] E.I. Keleb, A. Vermeire, C. Vervaet, J.P. Remon, Cold extrusion as a continuous

single-step granulation and tabletting process, Eur. J. Pharm. Biopharm. 52 (2001)

359-368.

[46] C. De Brabander, C. Vervaet, J.P. Remon, Development and evaluation of sustained

release mini-matrices prepared by hot melt extrusion, J. Control. Release 89 (2003)

235-247.

[47] C. De Brabander, C. Vervaet, J.P. Remon, Bioavailability of ibuprofen from hot-melt

extruded mini-matrices, Int. J. Pharm. 271 (2004) 77-84.

[48] M. Kidokoro, N.H. Shah, A.W. Malick, M.H. Infeld, J.W. McGinity, Properties of tablets

containing granulations of ibuprofen and an acrylic copolymer prepared by thermal

processes, Pharm. Dev. Technol. 6 (2001) 263-275.

[49] D. Henrist, R.A. Lefevre, J.P. Remon, Bioavailability of starch based hot stage

extrusion formulations, Int. J. Pharm. 187 (1999) 185-191.

25252525

INTRODUCTION

[50] D. Henrist, J.P. Remon, Influence of the formulation composition on the in vitro

characteristics of hot stage extrudates, Int. J. Pharm. 188 (1999) 111-119.

[51] D. Henrist, J.P. Remon, Influence of the process parameters on the characteristics of

starch based hot stage extrudates, Int. J. Pharm. 189 (1999) 7-17.

[52] F. Zhang, J.W. McGinity, Properties of hot-melt extruded theophylline tablets

containing poly(vinyl acetate), Drug Dev. Ind. Pharm. 26 (2000) 931-942.

[53] C. Reitz, P. Kleinebudde, Solid lipid extrusion of sustained release dosage forms, Eur.

J. Pharm. Biopharm. 67 (2007) 440-448.

[54] C. Reitz, C. Strachan, P. Kleinebudde, Solid lipid extrudates as sustained release

matrices: the effect of surface structure on drug release properties, Eur. J. Pharm. Sci.,

doi: 10.1016/j/ejps.2008.08.002.

[55] M.M. Crowley, B. Schroeder, A. Fredersdorf, S. Obara, M. Talarico, S. Kucera, J.W.

McGinity, Physicochemical properties and mechanism of drug release from

ethylcellulose matrix tablets prepared by direct compression and hot-melt extrusion, Int.

J. Pharm. 269 (2004) 509-522.

[56] C. Bruce, K.A. Fegely, A.R. Rajabi-Siahboomi, J.W. McGinity, Crystal growth formation

in melt extrudates, Int. J. Pharm. 341 (2007) 162-172.

[57] L.D. Bruce, N.H. Shah, A.W. Malick, M.H. Infeld, J.W. McGinity, Properties of hot-melt

extruded tablet formulations for the colonic delivery of 5-aminosalicylic acid, Eur. J.

Pharm. Biopharm. 59 (2005) 85-97.

[58] L. Li, O. AbuBaker, Z.J. Shao, Characterization of poly(ethylene oxide) as a drug

carrier in hot-melt extrusion, Drug Dev. Ind. Pharm. 32 (2006) 991-1002.

[59] R. Yang, Y. Wang, X. Zheng, J. Meng, X. Tang, X. Zhang, Preparation and evaluation

of ketoprofen hot-melt extruded enteric and sustained-release tablets, Drug Dev. Ind.

Pharm. 34 (2008) 83-89.

26262626


[60] D. Henrist, L. Van Bortel, R.A. Lefebvre, J.P. Remon, In vitro and in vivo evaluation of

starch-based hot stage extruded double matrix systems, J. Control. Release 75 (2001)

391-400.

[61] E. Mehuys, C. Vervaet, J.P. Remon, Hot-melt extruded ethylcellulose cylinders

containing a HPMC-Gelucire® core for sustained drug delivery, J. Control. Release 94

(2004) 273-280.

[62] H. Maeda, E. Ohashi, A. Sano, H. Kawasaki, Y. Kurosaki, Investigation of the release

behavior of a covered-rod-type formulation using silicone, J. Control. Release 90

(2003) 59-70.

[63] H. Maeda, T. Nakagawa, N. Adachi, Y. Sakai, T. Yamamoto, A. Matsuoka, A. Sano, Y.

Satoh, T. Miyata, K. Fujioka, Design of long-acting formulation of protein drugs with a

double-layer structure and its application to rhG-CSF, J. Control. Release 91 (2003)

281-297.

[64] H. Maeda, M. Brandon, A. Sano, Design of controlled-release formulation for

ivermectin using silicone, Int. J. Pharm. 261 (2003) 9-19.

[65] E. Mehuys, C. Vervaet, I. Gielen, H. Van Bree, J.P. Remon, In vitro and in vivo

evaluation of a matrix-in-cylinder system for sustained drug delivery, J. Control.

Release 96 (2004) 261-271.

[66] E. Mehuys, J.P. Remon, A. Korst, L. Van Bortel, R. Mols, P. Augustijns, C. Porter, C.

Vervaet, Human bioavailability of propranolol from a matrix-in-cylinder system with a

HPMC-Gelucire® core, J. Control. Release 107 (2005) 523-536.

[67] E. Mehuys, J.P. Remon, C. Vervaet, Production of enteric capsules by means of hot-

melt extrusion, Eur. J. Pharm. Sci. 24 (2005) 207-212.

[68] R. Bhardwaj, J. Blanchard, In vitro evaluation of poly(D,L-lactide-co-glycolide) polymer-

based implants containing the α-melanocyte stimulating hormone analog, Melanotan-I,

J. Control. Release 45 (1997) 49-55.

27272727

INTRODUCTION

[69] R. Bhardway, J. Blanchard, In vitro characterization and in vivo release profile of a

poly(D,L-lactide-co-glycolide)-based implant delivery system for the α-MSH analog,

Melanotan-I, Int. J. Pharm. 170 (1998) 109-117.

[70] Y. Lemmouchi, E. Schacht, P. Kageruka, R. De Deken, B. Diarra, O. Diall, S. Geerts,

Biodegradable polyesters for controlled release of trypanocidal drugs: in vitro and in

vivo studies, Biomaterials 19 (1998) 1827-1837.

[71] A. Rothen-Weinhold, K. Besseghir, E. Vuaridel, E. Sublet, N. Oudry, F. Kubel, R.

Gurny, Injection-molding versus extrusion as manufacturing technique for the

preparation of biodegradable implants, Eur. J. Pharm. Biopharm. 48 (1999) 113-121.

[72] M. Kajihara, T. Sugie, T. Hojo, H. Maeda, A. Sano, K. Fujioka, S. Sugawara, Y. Urabe,

Development of a new drug delivery system for protein drugs using silicone (II), J.

Control. Release 73 (2001) 279-291.

[73] M.A. Repka, T.G. Gerding, S.L. Repka, J.W. McGinity, Influence of plasticizers and

drugs on the physical-mechanical properties of hydroxypropylcellulose films prepared

by hot-melt extrusion, Drug Dev. Ind. Pharm. 25 (1999) 625-633.

[74] M.A. Repka, J.W. McGinity, Influence of chlorpheniramine maleate on topical

hydroxypropylcellulose films produced by hot-melt extrusion, Pharm. Dev. Technol. 6

(2001) 297-304.

[75] M.A. Repka, J.W. McGinity, Influence of vitamin E TPGS on the properties of

hydrophilic films produced by hot-melt extrusion, Int. J. Pharm. 202 (2000) 63-70.

[76] M.M. Crowley, A. Fredersdorf, B. Schroeder, S. Kucera, S. Prodduturi, M.A. Repka,

J.W. McGinity, The influence of guaifenesin and ketoprofen on the properties of hot-

melt extruded polyethylene oxide films, Eur. J. Pharm. Sci. 22 (2004) 409-418.

[77] V.S. Tumuluri, M.S. Kemper, I.R. Lewis, S. Prodduturi, S. Majumdar, B.A. Avery, M.A.

Repka, Off-line and on-line measurements of drug-loaded hot-melt extruded films using

Raman spectroscopy, Int. J. Pharm. 357 (2008) 77-84.

[78] M.A. Repka, S. Prodduturi, S.P. Stodghill, Production and characterization of hot-melt

extruded films containing clotrimazole, Drug Dev. Ind. Pharm. 29 (2003) 757-765.

28282828


[79] V.S. Tumuluri, S. Prodduturi, M.M. Crowley, S.P. Stodghill, J.W. McGinity, M.A.Repka,

B.A. Avery, The use of near-infrared spectroscopy for the quantitation of a drug in hot-

melt extruded films, Drug Dev. Ind. Pharm. 30 (2004) 505-511.

[80] S. Prodduturi, R.V. Manek, W.M. Kolling, S.P. Stodghill, M.A. Repka, Solid-state

stability and characterization of hot-melt extruded poly(ethylene oxide) films, J. Pharm.

Sci. 94 (2005) 2232-2245.

[81] S. Prodduturi, K.L. Urman, J.U. Otaigbe, M.A. Repka, Stabilization of hot-melt extrusion

formulations containing solid solutions using polymer blends, AAPS Pharm. Sci. Tech.

8 (2007) Article 50.

[82] P.K. Mididoddi, S. Prodduturi, M.A. Repka, Influence of tartaric acid on the bioadhesion

and mechanical properties of hot-melt extruded hydroxypropyl cellulose films for the

human nail, Drug Dev. Ind. Pharm. 32 (2006) 1059-1066.

[83] P.K. Mididoddi, M.A. Repka, Characterization of hot-melt extruded drug delivery

systems for onychomycosis, Eur. J. Pharm. Biopharm. 66 (2007) 95-105.

[84] S.M. Trey, D.A. Wicks, P.K. Mididoddi, M.A. Repka, Delivery of itraconazole from

extruded HPC films, Drug Dev. Ind. Pharm. 33 (2007) 727-735.

[85] M.A. Repka, K. Gutta, S. Prodduturi, M. Munjal, S.P. Stodghill, Characterization of

cellulosic hot-melt extruded films containing lidocaine, Eur. J. Pharm. Biopharm. 59

(2005) 189-196.

[86] A. Forster, J. Hempenstall, I. Tucker, T. Rades, The potential of small-scale fusion

experiments and the Gordon-Taylor equation to predict the suitability of drug/polymer

blends for melt extrusion, Drug Dev. Ind. Pharm. 27 (2001) 549-560.

[87] A. Forster, J. Hempenstall, I. Tucker, T. Rades, Selection of excipients for melt

extrusion with two poorly water-soluble drugs by solubility parameter calculation and

thermal analysis, Int. J. Pharm. 226 (2001) 147-161.

[88] R.J. Chokshi, H.K. Sandhu, R.M. Iyer, N.H. Shah, A.W. Malick, H. Zia, Characterization

of physico-mechanical properties of indomethacin and polymers to assess their

29292929

INTRODUCTION

suitability for hot-melt extrusion process as a means to manufacture solid

dispersion/solution, J. Pharm. Sci. 94 (2005) 2463-2474.

[89] Y. Zhu, N.H. Shah, A.W. Malick, M.H. Infeld, J.W. McGinity, Controlled release of a

poorly water-soluble drug from hot-melt extrudates containing acrylic polymers, Drug

Dev. Ind. Pharm. 32 (2006) 569-583.

[90] K. Nakamichi, T. Nakano, S. Izumi, H. Yasuura, Y. Kawashima, The preparation of

enteric solid dispersions with hydroxypropylmethylcellulose acetate succinate using a

twin-screw extruder, J. Drug Del. Sci. Tech. 14 (2004) 193-198.

[91] C. De Brabander, G. Van den Mooter, C. Vervaet, J.P. Remon, Characterization of

ibuprofen as a non-traditional plasticizer of ethyl cellulose, J. Pharm. Sci. 91 (2002)

1678-1685.

[92] K. Six, T. Daems, J. de Hoon, A. Van Hecken, M. Depre, M. Bouche, P. Prinsen, G.

Verreck, J. Peeters, M.E. Brewster, G. Van den Mooter, Clinical study of solid

dispersions of itraconazole prepared by hot-stage extrusion, Eur. J. Pharm. Sci. 24

(2005) 179-186.

[93] X. Zheng, R. Yang, X. Tang, L. Zheng, Part I: characterization of solid dispersions of

nimodipine prepared by hot-melt extrusion, Drug Dev. Ind. Pharm. 33 (2007) 791-802.

[94] A.C. Vial-Bernasconi, E. Doelker, P. Buri, Gélules à liberation prolongée: formes

divisées et monolithiques, STP Pharma Sci. 4 (1988) 397-409.

[95] N. Follonier, E. Doelker, Biopharmaceutical comparison of oral multiple-unit and single-

unit sustained-release dosage forms, STP Pharma Sci. 2 (1992) 141-155.

[96] P. Lennartz, J.B. Mielck, Mini-tabletting: improving the compactability of paracetamol

powder mixtures, Int. J. Pharm. 173 (1998) 75-85.

[97] H. Bechgaard, G.H. Nielsen, Controlled-release multiple-unit and single-unit doses,


30303030


[98] B. Edgar, C. Bogentoft, P.O. Lagerstrom, Comparison of 2 enteric-coated

acetylsalicylic-acid preparations by monitoring steady-state levels of salicylic-acid and

its metabolites in plasma and urine, Biopharm. Drug Dispos. 5 (1984) 251-260.

[99] C. Graffner, K. Josefsson, O. Stockman, Intra- and intersubject variation of

erythromycin absorption from single-unit and multiple-unit enteric-coated products,

Biopharm. Drug Dispos. 7 (1986) 163-171.

[100] J. Krämer, H. Blume, Biopharmaceutical aspects of multiparticulates, in:

Multiparticulate oral drug delivery (I. Ghebre-Sellassie, ed.), Marcel Dekker, New York,

USA (1994) 307-332.

[101] J.W. Conine, H.R. Hadley, Preparation of small solid pharmaceutical spheres, Drug

Cosmet. Ind. 106 (1970) 38-41.

[102] A.D. Reynolds, A new technique for the production of spherical particles, Manuf.

Chemist 41 (1970) 40-43.

[103] R.C. Rowe, Spheronization: a novel pill-making process, Pharm. Int. 6 (1985) 119-123.

[104] D.S. Hicks, H.L. Freese, Extrusion and spheronizing equipment, in: Pharmaceutical

pelletization technology (I. Ghebre-Sellassie, ed.), Marcel Dekker, New York, USA

(1989) 71-100.

[105] C. Vervaet, L. Baert, J.P. Remon, Extrusion-spheronisation: a literature-review, Int. J.

Pharm. 116 (1995) 131-146.

[106] J.M. Newton, Extrusion and extruders, in: Encyclopedia of pharmaceutical technology

(J. Swarbrick, J.C. Boylan, ed.), Marcel Dekker, New York, USA (2002) 1220-1236.

[107] D.F. Erkoboni, Extrusion/spheronization, in: Pharmaceutical extrusion technology –

Drugs and the pharmaceutical sciences, Vol. 133 (I. Ghebre-Sellassie, C. Martin, ed.),


[108] N.R. Trivedi, M.G. Rajan, J.R. Johnson, A.J. Shukla, Pharmaceutical approaches to

preparing pelletized dosage forms using the extrusion-spheronization process, Crit.

Rev. Ther. Drug 24 (2007) 1-40.

31313131

INTRODUCTION

[109] B. Bessemer, Shape extrusion, in: Pharmaceutical extrusion technology – Drugs and

the pharmaceutical sciences, Vol. 133 (I. Ghebre-Sellassie, C. Martin, ed.), Marcel


[110] Y.W. Chien, Fundamentals of controlled-release of drug administration, in: Novel drug

delivery system (J. Swarbrick, ed.), Marcel Dekker, New York, USA (1982) 465-574.

[111] U. Conte, L. Maggi, P. Colombo, A. La Manna, Multilayered hydrophilic matrices as

constant release devices (GeomatrixTM systems), J. Control. Release 26 (1993) 39-47.

[112] U. Conte, L. Maggi, Modulation of the dissolution profiles from Geomatrix® multi-layer

matrix tablets containing drugs of different solubility, Biomaterials 17 (1996) 889-896.

[113] U. Conte, L. Maggi, Multilayer tablets as drug delivery devices, Pharm. Tech. Int. 10

(1998) 174-182.

[114] U. Conte, L. Maggi, A flexible technology for the linear, pulsatile release of drugs,

allowing for easy accommodation of difficult in vitro targets, J. Control. Release 64

(2000) 263-268.

[115] H.G. Zerbe, M. Krumme, Smartrix system: design characteristics and release

properties of a novel erosion-controlled oral delivery system, in: Modified-Release Drug

Delivery Technology (M.J. Rathbone, J. Hadgraft, M.S. Roberts, ed.), Marcel Dekker,

New York, USA (2003) 35-48.

[116] P. Colombo, U. Conte, A. Gazzaniga, L. Maggi, M.E. Sangalli, N.A. Peppas, A. La

Manna, Drug release modulation by physical restrictions of matrix swelling, Int. J.

Pharm. 63 (1990) 43-48.

[117] M.P. Danckwerts, Development of a zero-order release oral compressed tablet with

potential for commercial tabletting production, Int. J. Pharm. 112 (1994) 37-45.

[118] P.S.L. Wong, D.E. Edgren, L.C. Dong, V.J. Ferrari, Active agent dosage form

comprising a matrix and at least two insoluble bands, US Patent 5 534 263, ALZA

Corporation (1996).

32323232


[119] P.S.L. Wong, D.E. Edgren, L.C. Dong, V.J. Ferrari, Banded prolonged release active

agent dosage form, US Patent 5 667 804, ALZA Corporation (1997).

[120] D.A. Dickason, G.P. Grandolfi, RingCap Technology, in: Modified-Release Drug

Delivery Technology (M.J. Rathbone, J. Hadgraft, M.S. Roberts, ed.), Marcel Dekker,

New York, USA (2003) 49-57.

[121] I. Krögel, R. Bodmeier, Development of a multifunctional matrix drug delivery system

surrounded by an impermeable cylinder, J. Control. Release 61 (1999) 43-50.

[122] J. Cobby, M. Mayersohn, G.C. Walker, Influence of shape factors on kinetics of drug

release from matrix tablets: I. Theoretical, J. Pharm. Sci. 63 (1974) 725-731.

[123] J. Cobby, M. Mayersohn, G.C. Walker, Influence of shape factors on kinetics of drug

release from matrix tablets: II. Experimental, J. Pharm. Sci. 63 (1974) 734-737.

[124] E.Z. El-Din, M.H. El-Shaboury, H.A. El-Aleem, Effect of tablet shape on the in vitro and

in vivo availability of directly compressed non-disintegrating tablets, Pharm. Ind. 51

(1989) 694-696.

[125] A.C. Shah, N.J. Britten, Novel divisible design for sustained-release formulations, J.


[126] A.Y. Benkorah, J.N. McMullen, Biconcave coated, centrally perforated tablets for oral

controlled drug delivery, J. Control. Release 32 (1994) 155-160.

[127] C. Kim, Compressed donut-shaped tablets with zero-order release kinetics, Pharm.

Res. 12 (1994) 1045-1048.

[128] S.K. Chopra, Drug delivery systems based on geometric configuration, in: Modified-

Release Drug Delivery Technology (M.J. Rathbone, J. Hadgraft, M.S. Roberts, ed.),


[129] N. Lordi, Sustained release dosage forms, in: The theory and practice of industrial

pharmacy (L. Lachman, H. Lieberman, J. Kanig, ed.), Lea and Febiger, Philadelphia,

USA (1986) 430-456.

33333333

INTRODUCTION

[130] W. Sutananta, D.Q.M. Craig, J.M. Newton, An evaluation of the mechanisms of drug

release from glyceride bases, J. Pharm. Pharmacol. 47 (1995) 182-187.

[131] V. Ratsimbazafy, E. Bourret, C. Brossard, Drug release from matrix tablets and

minitablets containing glycerides, Pharm. Ind. 58 (1996) 442-446.

[132] J. Hamdani, A.J. Moës, K. Amighi, Physical and thermal characterisation of Precirol®

and Compritol® as lipophilic glycerides used for the preparation of controlled-release

matrix pellets, Int. J. Pharm. 260 (2003) 47-57.

[133] G. Di Colo, S. Burgalassi, P. Chetoni, M.P. Fiaschi, Y. Zambito, M.F. Saettone, Gel-

forming erodible inserts for ocular controlled delivery of ofloxacin, Int. J. Pharm. 215

(2001) 101-111.

[134] P. Gao, J.W. Skoug, P.R. Nixon, T.R. Ju, N.L. Stemm, K.C. Sung, Swelling of

hydroxypropyl methylcellulose matrix tablets. II. Mechanistic study of the influence of

formulation variables on matrix performance and drug release, J. Pharm. Sci. 85 (1996)

732-739.

[135] L.W.S. Cheong, P.W.S. Heng, L.F. Wong, Relationship between polymer viscosity and

drug release from a matrix system, Pharm. Res. 9 (1992) 1510-1514.

[136] L.S.C. Wan, P.W.S. Heng, L.F. Wong, Relationship between swelling and drug release

in a hydrophilic matrix, Drug Dev. Ind. Pharm. 19 (1993) 1201-1210.

[137] N. Shah, G. Zhang, V. Apelian, F. Zheng, M.H. Infeld, A.W. Malick, Prediction of drug

release from hydroxypropyl methylcellulose (HPMC) matrices: effect of polymer

concentration, Pharm. Res. 10 (1993) 1693-1695.

[138] E. Papadimitriou, G. Buckton, M Efentakis, Probing the mechanisms of swelling of

hydroxypropyl methylcellulose matrices, Int. J. Pharm. 98 (1993) 57-62.

[139] H. Khurahashi, H. Kami, H. Sunada, Influence of physicochemical properties on drug

release rate from hydroxypropylmethylcellulose matrices, Chem. Pharm. Bull. 44

(1996) 829-832.

34343434


[140] T.D. Reynolds, S.H. Gehrke, A.S. Hussain, L.S. Shenouda, Polymer erosion and drug

release characterization of hydroxypropylmethylcellulose matrices, J. Pharm. Sci. 87

(1998) 1115-1123.

[141] I. Katzhendler, K. Mäder, M. Friedman, Structure and hydration properties of

hydroxypropyl methyl cellulose matrices containing naproxen and naproxen sodium,

Int. J. Pharm. 200 (2000) 161-179.

[142] V. Mahaguna, R.L. Talbert, J.I. Peters, S. Adams, T.D. Reynolds, F.Y.W. Lam, R.O.

Williams, Influence of hydroxypropyl methylcellulose polymer on in vitro and in vivo

performance of controlled release tablets containing alprazolam, Eur. J. Pharm.

Biopharm. 56 (2003) 101-108.

[143] M.L. Vueba, L.A.E. Batista de Carvalho, F. Veiga, J.J. Sousa, M.E. Pina, Role of

cellulose ether polymers on ibuprofen release from matrix tablets, Drug Dev. Ind.

Pharm. 31 (2005) 653-665.

[144] L. Yang, G. Venkatesh, R. Fassihi, Characterization of compressibility and compatibility

of poly(ethylene oxide) polymers for modified release application by compaction

simulator, J. Pharm. Sci. 85 (1996) 1085-1090.

[145] C.J. Kim, Drug release from compressed hydrophilic Polyox®-WSR tablets, J. Pharm.

Sci. 84 (1995) 303-306.

[146] C.J. Kim, Effects of drug solubility, drug loading, and polymer molecular weight on drug

release from Polyox® tablets, Drug Dev. Ind. Pharm. 24 (1998) 645-651.

[147] L. Maggi, R. Bruni, U. Conte, High molecular weight polyethylene oxides (PEOs) as an

alternative to HPMC in controlled release dosage forms, Int. J. Pharm. 195 (2000) 229-

238.

[148] L. Maggi, L. Segale, M.L. Torre, E. Ochoa Machiste, U. Conte, Dissolution behaviour of

hydrophilic matrix tablets containing two different polyethylene oxides (PEOs) for the

controlled release of a water-soluble drug. Dimensionality study, Biomaterials 23

(2002) 1113-1119.

35353535

INTRODUCTION

[149] N. Wu, L. Wang, D.C. Tan, S.M. Moochhala, Y. Yang, Mathematical modeling and in

vitro study of controlled drug release via a highly swellable and dissoluble polymer

matrix: polyethylene oxide with high molecular weights, J. Control. Release 102 (2005)

569-581.

[150] A. Apicella, B. Cappello, M.A. Del Nobile, M.I. La Rotonda, G. Mensitieri, L. Nicolais,

Poly(ethylene oxide) (PEO) and different molecular weight PEO blends monolithic

devices for drug release, Biomaterials 14 (1993) 83-90.

[151] V. Dhopeshwarkar, J.L. Zatz, Evaluation of xanthan gum in the preparation of

sustained release matrix tablets, Drug Dev. Ind. Pharm. 19 (1993) 999-1017.

[152] M.M. Talukdur, A. Michoel, P. Rombaut, R. Kinget, Comparative study on xanthan gum

and hydroxypropyl methylcellulose as matrices for controlled drug delivery. I.

Compaction and in vitro drug release behaviour, Int. J. Pharm. 129 (1996) 233-241.

[153] M.M. Talukdur, R. Kinget, Comparative study on xanthan gum and hydroxypropyl

methylcellulose as matrices for controlled-release drug delivery. II. Drug diffusion in

hydrated matrices, Int. J. Pharm. 151 (1997) 99-107.

[154] M.M. Talukdur, I. Vinckier, P. Moldenaers, R. Kinget, Rheological characterisation of

xanthan gum and hydroxypropyl methylcellulose with respect to controlled-release drug

delivery, J. Pharm. Sci. 85 (1996) 537-540.

[155] M.M. Talukdar, G. Van den Mooter, P. Augustijns, T. Tjandra-Maga, N. Verbeke, R.

Kinget, In vivo evaluation of xanthan gum as a potential excipient for oral controlled-

release matrix tablet formulation, Int. J. Pharm. 169 (1998) 105-133.

[156] F. Ughini, I.F. Andreazza, J.L.M.S. Ganter, T.M.B. Bresolin, Evaluation of xanthan and

highly substituted galactomannan from M. scabrella as a sustained release matrix, Int.

J. Pharm. 271 (2004) 197-205.

[157] H. Santos, F. Veiga, M.E. Pina, J.J. Sousa, Compaction, compression and drug

release characteristics of xanthan gum pellets of different compositions, Eur. J. Pharm.

Sci. 21 (2004) 271-281.

36363636



release properties of diclofenac sodium and ibuprofen pellets comprising xanthan gum

as a sustained release agent, Int. J. Pharm. 295 (2005) 15-27.

[159] C.W. Vendruscolo, I.F. Andreazza, J.L.M.S. Ganter, C. Ferrero, T.M.B. Bresolin,

Xanthan and galactomannan (from M. scabrella) matrix tablets for oral controlled

delivery of theophylline, Int. J. Pharm. 296 (2005) 1-11.

[160] M.M. Talukdur, R. Kinget, Swelling and drug release behaviour of xanthan gum matrix

tablets, Int. J. Pharm. 120 (1995) 63-72.

[161] M.M. Talukdar, J. Plaizier-Vercammen, Evaluation of xanthan gum as a hydrophilic

matrix for controlled-release dosage form preparations, Drug Dev. Ind. Pharm. 19

(1993) 1037-1046.

[162] M.F. Lu, L. Woodward, S. Borodkin, Xanthan gum and alginate based controlled

release theophylline formulations, Drug Dev. Ind. Pharm. 17 (1991) 1987-2004.

[163] P.J. Cox, K.A. Khan, D.L. Munday, J. Sujja-areevath, Development and evaluation of a

multiple-unit oral sustained release dosage form for S(+)-ibuprofen: preparation and

release kinetics, Int. J. Pharm. 193 (1999) 73-84.

[164] N. Billa, K. Yuen, M.A.A. Khader, A. Omar, Gamma-scintigraphic study of the

gastrointestinal transit and in vivo dissolution of a controlled release diclofenac sodium

formulation in xanthan gum matrices, Int. J. Pharm. 201 (2000) 109-120.

[165] H.M. Ingani, A.J. Moes, Utilisation de la gomme xanthane dans la formation des

matrices hydrophiles, STP Pharm. 4 (1988) 188-195.

[166] X. Mu, M.J. Tobyn, J.N. Staniforth, Influence of physiological variables on the in vitro

drug-release behaviour of a polysaccharide matrix controlled-release system, Drug

Dev. Ind. Pharm. 29 (2003) 19-29.

[167] S. Baumgartner, M. Pavli, J. Kristl, Effect of calcium ions on the gelling and drug

release characteristics of xanthan matrix tablets, Eur. J. Pharm. Biopharm. 69 (2008)

698-707.

37373737

INTRODUCTION

[168] S.K. Baveja, K.V.R. Rao, J.Arora, N.K. Mathur, V.K. Vinayak, Chemical investigations

of some galactomannan gums as matrix tablets for sustained drug delivery, Indian J.

Chem. B 30 (1991) 133-137.

[169] S.A. Altaf, K. Yu, J. Parasrampuria, D.R. Friend, Guar gum-based sustained release

diltiazem, Pharm. Res. 15 (1998) 1196-1201.

[170] P. Khullar, R.K. Khar, S.P. Agarwal, Evaluation of guar gum in the preparation of

sustained-release matrix tablets, Drug Dev. Ind. Pharm. 24 (1998) 1095-1099.

[171] Y.S.R. Krishnaiah, R.S. Karthikeyan, V. Satyanarayana, A three-layer guar gum matrix

tablet for oral controlled delivery of highly soluble metoprolol tartrate, Int. J. Pharm. 241

(2002) 353-366.

[172] S.M. Al-Saidan, Y.S.R. Krishnaiah, V. Satyanarayana, P. Bhaskar, R.S. Karthikeyan,

Pharmacokinetic evaluation of guar gum-based three-layer matrix tablets for oral

controlled delivery of highly soluble metoprolol tartrate as a model drug, Eur. J. Pharm.

Biopharm. 58 (2004) 697-703.

[173] F. Tugcu-Demiroz, F. Acarturk, S. Takka, O. Konus-Boyunaga, In-vitro and in-vivo

evaluation of mesalazine-guar gum matrix tablets for colonic drug delivery, J. Drug

Target. 12 (2004) 105-112.

[174] M. Hariharan, T.A. Wheatley, J.C. Price, Controlled-release tablet matrices from

carrageenans: compression and dissolution studies, Pharm. Dev. Technol. 2 (1997)

383-393.

[175] K.M. Picker, Matrix tablets of carrageenans. I. A compaction study, Drug Dev. Ind.

Pharm. 25 (1999) 329-337.

[176] K.M. Picker, Matrix tablets of carrageenans. II. Release behavior and effect of added

cations, Drug Dev. Ind. Pharm. 25 (1999) 339-346.

[177] V.K. Gupta, M. Hariharan, T.A. Wheatley, J.C. Price, Controlled-release tablets from

carrageenans: effect of formulation, storage and dissolution factors, Eur. J. Pharm.

Biopharm. 51 (2001) 241-248.

38383838


[178] J. Nerurkar, H.W. Jun, J.C. Price, M.O. Park, Controlled-release matrix tablets of

ibuprofen using cellulose ethers and carrageenans: effect of formulation factors on

dissolution rates, Eur. J. Pharm. Biopharm. 61 (2005) 56-68.

[179] P. Colombo, A. Gazzaniga, C. Caramella, U. Conte, A. La Manna, In vitro

programmable zero-order release drug delivery system, Acta Pharm. Technol. 33

(1987) 15-20.

[180] U. Conte, P. Colombo, A. Gazzaniga, M.E. Sangalli, A. La Manna, Swelling-activated

drug delivery systems, Biomaterials 9 (1988) 489-493.

[181] R.S. Harland, A. Gazzaniga, M.E. Sangalli, P. Colombo, N.A. Peppas, Drug/polymer

matrix swelling and dissolution, Pharm. Res. 5 (1988) 488-494.

[182] K.V. Rao Rango, K.P. Devi, P. Buri, Influence of molecular size and water solubility of

the solute on its release from swelling and erosion controlled polymeric matrices, J.


[183] T. Ohara, S. Kitamura, T. Kitagawa, K. Terada, Dissolution mechanism of poorly water-

soluble drug from extended release solid dispersion system with ethylcellulose and

hydroxypropylmethylcellulose, Int. J. Pharm. 302 (2005) 95-102.

39393939

40404040

OUTLINE AND AIMS

41414141

OUTLINE AND AIMS

42424242

OUTLINE AND AIMS

OUTLINE AND AIMS

Previous research work showed that sustained-release mini-matrices formulated with a

combination of ethylcellulose and a hydrophilic polymer (hydroxypropylmethylcellulose, xanthan

gum) could be manufactured using the hot-melt extrusion technique. The release of ibuprofen,

selected as model drug, was influenced by the concentration, viscosity grade and substitution type

of hydroxypropylmethylcellulose, and the investigators only briefly discussed a formulation

consisting of ethylcellulose, ibuprofen and xanthan gum [1]. No detailed study was performed to

evaluate the influence of xanthan gum concentration and particle size on ibuprofen release.

Therefore, CHAPTER 1 further explores the possibilities of xanthan gum as hydrophilic polymer to

tailor ibuprofen release from ethylcellulose/xanthan gum mini-matrices. The influence of xanthan

gum concentration and particle size on the in vitro behaviour of the mini-matrices (drug release

and mechanism, liquid uptake/swelling/erosion, porosity/friability/hardness) was evaluated and a

stability study over 12 months at different storage conditions was conducted. In addition, the

ibuprofen bioavailability from these hot-melt extruded mini-matrices was evaluated in dogs.

Since ibuprofen acts as a plasticizer for ethylcellulose, ibuprofen was substituted in

CHAPTER 2 by metoprolol tartrate, a non-plasticizing drug. We studied the influence of different

plasticizers and plasticizer concentrations on the performance of the hot-melt extrusion process,

the quality of the extrudate and mini-matrices, and the in vitro release properties of metoprolol

tartrate. The influence of xanthan gum concentration was investigated in order to obtain mini-

matrices with zero-order release kinetics. In addition, the effect of extrusion process parameters

(screw design, powder feed rate and screw speed) on the extrusion process, extrudate and mini-

matrices quality, drug and polymer crystallinity, drug homogeneity and in vitro release

characteristics is described.

CHAPTER 3 deals with the addition of polyethylene oxides as hydrophilic polymer to tailor

metoprolol tartrate release from the lipophilic ethylcellulose matrix-cores: the influence of

polyethylene oxide concentration as well as its molecular weight was investigated on drug and

hydrophilic polymer release, drug distribution, and drug and polymer crystallinity. An in vivo study

in dogs was performed and the bioavailability of metoprolol tartrate from

ethylcellulose/polyethylene oxide and ethylcellulose/xanthan gum mini-matrices was compared.

CHAPTER 4 reports the modelling of metoprolol tartrate and polyethylene oxide release

from hot-melt extruded mini-matrices. Based on time-independent diffusion coefficients, drug

43434343

OUTLINE AND AIMS

release was predicted from mini-matrices with different dimensions. In case the drug release was

not only diffusion-controlled, a new mathematical model was developed (taking the porosity

changes during dissolution into account) to better describe drug release and predict drug and

polymer distribution in the mini-tablets in function of dissolution testing.

44444444

OUTLINE AND AIMS

REFERENCES


release mini-matrices prepared via hot melt extrusion, J. Control. Release 89 (2003)

235-247.

45454545

46464646

CHAPTER 1 XANTHAN GUM TO TAILOR

IBUPROFEN RELEASE: IN VITRO AND IN VIVO EVALUATION

Parts of this chapter were published in:

Xanthan gum to tailor drug release of sustained-release ethylcellulose mini-matrices prepared via hot-

melt extrusion: in vitro and in vivo evaluation.

E. Verhoeven, C. Vervaet, J.P. Remon, Eur. J. Pharm. Biopharm. 63 (2006) 320-330.

Laboratory of Pharmaceutical Technology, Ghent University, Ghent, Belgium.

474747

CHAPTER 1

ABSTRACT

Mini-matrices (multiple-unit dosage form) with release-sustaining properties were

developed by means of hot-melt extrusion using ibuprofen as model drug and ethylcellulose

as sustained-release agent. Xanthan gum, a hydrophilic polymer, was added to the

formulation to increase drug release since ibuprofen release from the

ibuprofen/ethylcellulose matrices (60/40, w/w) was too slow (20% in 24 h). Changing the

xanthan gum concentration as well as its particle size modified the in vitro drug release.

Increasing xanthan gum concentrations yielded a faster drug release due to a higher liquid

uptake, swelling and erosion rate. Regarding the effect of the xanthan gum particle size, no

difference was observed for formulations containing 10 and 20% xanthan gum. However,

using 30% xanthan gum, drug release was influenced by the particle size of the hydrophilic

polymer due to the susceptibility of the coarser xanthan gum particles to erosion. Drug

release from the mini-matrices was mainly diffusion-controlled, but swelling played an

important role to obtain complete drug release within 24 h. Drug release was influenced by

the ionic strength of the medium as the conformation of xanthan gum molecules is

determined by the salt concentration. An oral dose of 300 mg ibuprofen was administered to

dogs (n = 6) either as an immediate-release preparation (Junifen®), as a sustained-release

formulation (Ibu-Slow® 600 mg (½ tablet)) or as the experimental mini-matrices (varying in

xanthan gum concentration). Administration of the experimental formulations sustained the

ibuprofen release. Although a significant difference in dissolution rate of the 20 and 30%

xanthan gum mini-matrices was detected in vitro, the difference in relative bioavailability

(using Ibu-Slow® 600 mg as reference) was limited (70.6 and 73.8%, respectively).

484848

XANTHAN GUM TO TAILOR IBUPROFEN RELEASE: IN VITRO AND IN VIVO EVALUATION


IBUPROFEN RELEASE: IN VITRO AND IN VIVO EVALUATION

1. INTRODUCTION

Hot-melt extrusion is becoming a widely used technology in the pharmaceutical industry to

produce matrix formulations into which a drug is homogeneously embedded. Its major advantage

over conventional techniques (e.g. direct compression) for manufacturing matrices is the

continuity of the production process as the different steps (mixing, melting, homogenizing and

shaping) are carried out on a single machine. This implicates a decrease in investment costs and

offers more opportunities for automation of the production. Furthermore, this technique has a high

throughput and limited material loss, yielding material having excellent homogeneity [1-3].

The excellent feasibility of ethylcellulose, a polymer with thermoplastic properties, for

hot-stage extrusion has been established in a variety of applications [4-7]. Previous work [8-

10] has shown that hot-melt extrusion can be used as an appropriate technique to develop

mini-matrices using ethylcellulose to sustain the release of ibuprofen, selected as the model

drug. The combination of ethylcellulose and a hydrophilic component such as

hydroxypropylmethylcellulose (HPMC) offered a flexible system to tailor the drug release by

changing the viscosity, substitution type and concentration of HPMC. Substituting HPMC for

xanthan gum yielded formulations having a nearly zero-order drug release without burst

effect and complete drug release within 24 h. In addition, the incorporation of xanthan gum

resulted in a longer sustained-release effect, allowing to use a lower concentration of

hydrophilic polymer [11]. Rheological and drug diffusion studies in hydrated HPMC and

xanthan gum compacts elucidated the difference in the release-controlling ability of both

polymers [12-14]. The higher ability of xanthan gum to control drug release in comparison

with HPMC most probably originates from their different hydrophilicity, hydration properties

494949

CHAPTER 1

and swelling behaviour. Whereas a gradual increase in liquid uptake and swelling prevailed

for the HPMC systems, the maximum liquid uptake and swelling was reached within 2 h for

the xanthan gum formulation, quickly forming a viscous gel around the matrix core.

In the present study, sustained-release mini-matrices were developed by hot-melt

extrusion of an ibuprofen/ethylcellulose mixture with the addition of xanthan gum to tailor

drug release, whereby the influence of the concentration as well as the particle size of

xanthan gum on the in vitro characteristics of the mini-matrices was investigated. The in vivo

performance of these experimental formulations was evaluated in dogs and compared with

an equivalent dose of a sustained-release ibuprofen matrix tablet (Ibu-Slow® 600 mg).

505050


2. MATERIALS AND METHODS

2.1. Materials

2.1.1. Ibuprofen

Ibuprofen (IBP) (average diameter: 25 µm) was received from Knoll Pharmaceuticals

(Nottingham, UK) and selected as model drug (Figure 1).

CH3

CH3

CH3

O

OH

Figure 1: Chemical structure of ibuprofen.

It is a chiral non-steroidal anti-inflammatory drug (NSAID), which is widely used for

the treatment of acute and chronic musculoskeletal inflammations and the control of mild to

moderate pain due to its anti-inflammatory, analgesic and antipyretic properties. Ibuprofen is

administered as a racemic mixture of R-(-)-ibuprofen and S-(+)-ibuprofen. S-(+)-ibuprofen is

the pharmacological active compound. However, R-(-)-ibuprofen is also related with the

therapeutic activity of the drug since it is in vivo partially converted to S-(+)-ibuprofen via a

unidirectional mechanism [15-18], and investigators performed research work on the

stereoselective disposition of ibuprofen enantiomers [19, 20]. Maximum ibuprofen plasma

concentrations are achieved within 1-2 h after ingestion of the drug, but due to its short half

life of approximately 2 h therapeutic blood concentrations can only be maintained if the drug

is administered frequently (multiple daily dosing). Slowing down the ibuprofen release from

the formulation could reduce the dose regimen to a once or twice a day scheme, enhancing

patient compliance and therefore making it a suitable candidate for sustained drug delivery.

Besides, a slower ibuprofen release could also decrease the occurrence of gastro-intestinal

(GI) side effects since the main adverse effects are GI disturbances [21]. Hot-melt extrusion

is a viable method to develop solid dispersions and solid solutions to improve the

bioavailability of poorly soluble compounds [3, 22-26]. The low aqueous solubility of

ibuprofen provides a scientific rationale for its incorporation in the dosage form under

investigation.

515151

CHAPTER 1

2.1.2. Ethylcellulose

The matrix consisted of ethylcellulose (EC) (Ethocel® Std 10 FP Premium; particle

size: 3-15 µm; ethoxyl content: 48-49.5% (w/w); viscosity: 9-11 mPa.s (5% solution in

toluene/EtOH 80/20, 25°C)), kindly donated by Dow Chemical Company (Midland, USA).

Ethylcellulose, an ethylether of cellulose, is a long chain polymer of β-anhydroglucose

units, where the number of monomers (n) can vary to provide a wide variety of molecular

weights (Figure 2; with partially ethoxyl substitution, degree of substitution = 2).

Ethylcellulose ethers are a family of inert hydrophobic polymers (organosoluble

thermoplastics) that are essentially tasteless, odourless, colourless, non-caloric and

physiologically inert. Ethylcellulose is useful in a variety of pharmaceutical applications: as

rate-controlling polymer in matrix tablets, as tablet binder and cushioning agent, as viscosity-

increasing agent, as coating polymer, as granulation binder and as matrix for solid

dispersions [27]. In our study, ethylcellulose was applied as sustained-release agent.

O

OHO

O

O C2H5

C2H5

n

Figure 2: Chemical structure of ethylcellulose.

2.1.3. Xanthan gum

Xanthan gum (XG), available in different particle sizes (Xantural® 75 (XG75),

Xantural® 180 (XG180) and Xantural® 11K (XG11K) having a mean particle size of 75, 180

and 1180 µm, respectively) was supplied by CP Kelco (Liverpool, UK).

Xanthan gum (Figure 3) is a high molecular weight extracellular polysaccharide

produced by the gram-negative bacterium Xanthomonas campestris through a fermentation

process. The primary structure of this naturally produced cellulose derivative consists of a

cellulosic backbone (β-D-glucose residues) and a trisaccharide side chain containing two

525252


mannose units and one glucuronic acid residue. The terminal D-mannose may carry a

pyruvate function, the non-terminal D-mannose unit (closest to the backbone) in the side

chain contains an acetyl function. The degree of acetylation and pyruvylation varies,

depending on the fermentation conditions. The anionic character of the polymer is due to the

presence of both glucuronic acid and pyruvic acid groups in the side chain. Next to its use as

thickening agent in the food and cosmetics industry, it has also applications in

pharmaceutical formulations to thicken, suspend, stabilize and emulsify. It has also been

reported by many authors that xanthan gum can be used as an effective excipient for

sustained-release formulations [27-32]. For our purposes, xanthan gum was incorporated in

the dosage form to tailor drug release from the lipophilic carrier, ethylcellulose.

OH OH

OH

OCH2OH

O

O

OCH2OH

O

O n

CCOO M

CH3

O

OHO

OH

OCH2

O

OH

OH

OCOO M O

OH

OCH2OCCH3

OH

- +

- +

Figure 3: Chemical structure of xanthan gum (where M+ = Na+, K+ or ½ Ca2+).

All other chemicals were at least of analytical grade (VWR, Leuven, Belgium).

535353

CHAPTER 1

2.2. Methods

2.2.1. Composition of the mixtures

The ibuprofen content was kept constant at 60% (w/w) and the remaining part of the

formulation consisted of ethylcellulose and xanthan gum. The xanthan gum concentration

varied between 10 and 30% (w/w). All formulations are listed in Table 1.

Table 1: Composition (%, w/w) of the hot-melt extruded formulations.

10% XG 20% XG 30% XG

Ibuprofen 60 60 60

Xanthan gum (XG75, XG180 or XG11K) 10 20 30

Ethylcellulose 30 20 10

2.2.2. Production of mini-matrices

Prior to hot-melt extrusion the formulations were blended in a planetary mixer (15 min, 90

rpm) (Kenwood Major Classic, Hampshire, UK). Hot-melt extrusion was performed using a lab-

scale intermeshing co-rotating twin-screw extruder (MP19TC-25, APV Baker, Newcastle-under-

Lyme, UK) with a length-to-diameter ratio of 25/1 (Figure 4).

Figure 4: Schematic diagram of a twin-screw extruder with (a) control panel, (b) driving unit,

(c) feeding hopper, (d) barrel with screws and (e) die.

The machine was equipped with a Brabender twin-screw powder feeder, a Technodrives

DC motor and a Tricool cooling unit. The barrel, containing the screws, is divided into 5 zones of

which the temperature can be set separately. A control panel is used for setting the temperature,

the screw speed and the powder feed rate (the process parameters). Some process variables

such as the pressure, the material temperature and the torque can be recorded from the control

a b

e c d

545454


panel. A screw with two mixing sections (Figure 5) and a cylindrical die of 3 mm was used for the

production of the mini-matrices. For all formulations, the following extrusion conditions were used:

a screw speed of 30 rpm, a powder feed rate of 6 g/min and a temperature of 50°C for the five

heating zones along the barrel. After cooling down to room temperature, the extrudates (∅ = 3

mm) were manually cut into mini-matrices of 2 mm length.

Figure 5: Configuration of the intermeshing co-rotating screws. Screw configuration with 2

kneading blocks: (1) transport zone, (2) mixing zone, (3) transport zone, (4) mixing zone, (5)

transport zone, (6) densification zone.

2.2.3. In vitro evaluation

2.2.3.1. In vitro drug release

The mini-matrices (approximately 60 mg) were introduced in a basket (USP 27,

dissolution apparatus 1 [33]). The dissolution was performed in a VK 7010 dissolution system

combined with a VK 8000 automatic sampling station (VanKel, New Jersey, USA).

Phosphate buffer KH2PO4 (pH 7.2, µ = 0.11) was used as the dissolution medium. The

temperature of the medium (900 ml) was kept at 37 ± 0.5°C, while the rotational speed of the

baskets was set at 100 rpm. Samples of 5 ml were withdrawn at 0.5, 1, 2, 4, 6, 8, 12, 16, 20

and 24 h and spectrophotometrically analyzed for ibuprofen at 221 nm by means of a Perkin-

Elmer Lambda 12 UV-VIS double beam spectrophotometer (Zaventem, Belgium). The

ibuprofen concentrations were calculated from a calibration curve between 0 and 50 µg/ml.

The dissolution was simultaneously performed in 6 dissolution vessels, each vessel

containing 4 mini-matrices. As the solubility of ibuprofen is pH-dependent, one could think to

investigate the impact of different pHs on the formulations tested. Phosphate buffer pH 7.2

was selected for dissolution testing (as recommended by USP for ibuprofen dosage forms

[33]) since it is most challenging at this pH to develop a sustained-release platform as the

solubility of ibuprofen is the highest at this pH.

1 2 3 4 5 6

555555

CHAPTER 1

Additional rotational speeds of 50 and 200 rpm were used during dissolution testing in

order to evaluate the influence of matrix erosion on drug release.

The susceptibility of the mini-matrices to ionic strength was verified by dissolution

testing in diluted phosphate buffer and phosphate buffer with increasing NaCl concentrations:

µ = 0.011 (10-fold dilution), µ = 0.022 (5-fold dilution), µ = 0.055 (2-fold dilution), µ = 0.11, µ =

0.15 and µ = 0.2. Sodium chloride was added to the medium, since these are the main

electrolytes in the gastro-intestinal fluid [34].

To determine the in vitro release properties of the formulations used for in vivo

evaluation, a dissolution was also performed on a hard-gelatin capsule (n° 00) filled with 500

mg mini-matrices.

To understand the drug release mechanism from swellable matrices, the data (Mt/M∞)

were fitted to the following exponential equation (equation 1) [35-38]:

nt ktMM

=∞

(1)

This equation generally holds for the initial phase of the release profile (Mt/M∞ ≤ 60%). Mt/M∞

is the percentage drug released at time t, k denotes a constant incorporating the structural

and geometric characteristics of the release device and the exponent n is a release constant

used to characterize the transport mechanism. The values of n were obtained by logarithmic

calculation of equation 1.

Images of the mini-matrices, after immersion in the dissolution medium, were made

with a digital camera (C3030 Olympus) attached to an image analysis system (analySIS®,

Soft Imaging system, Münster, Germany).

2.2.3.2. Liquid uptake, erosion and swelling measurements

The mini-matrices were introduced into USP dissolution apparatus 1 and submitted to

a dissolution test (n = 3) under the conditions described above (100 rpm, µ = 0.11), each

vessel containing one mini-matrix. At predetermined time intervals (dissolution sampling

points), they were withdrawn from the medium and weighed after removing excess surface

water.

565656


Liquid uptake (expressed as % weight gain of the polymer content (ethylcellulose and

xanthan gum)) was determined from the weight of the mini-matrices, taking into account the

drug released (equation 2):

( ) ( )( )0

0%DRW

DRWDRWuptakeliquidi

itw

−−−−

= x 100 (2)

where Ww = weight of the mini-matrix at time ‘t’ after immersion in the dissolution medium

Wi = initial weight of the mini-matrix at time ‘0’

DR0 = amount of drug in the mini-matrix at time ‘0’

DRt = amount of drug in the mini-matrix at time ‘t’

The degree of erosion (expressed as % erosion of the polymer content) was determined

based on the weight difference between the dried mini-matrix and its initial weight, considering the

amount of drug released (equation 3):

( ) ( )( )0

0%DRW

DRWDRWerosioni

tdi

−−−−

= x 100 (3)

where Wd = dry weight of the mini-matrix at time ‘t’ after immersion in the dissolution medium

Wi = initial weight of the mini-matrix at time ‘0’

DR0 = amount of drug in the mini-matrix at time ‘0’

DRt = amount of drug in the mini-matrix at time ‘t’

The height and diameter of the mini-matrices were measured using a digital calliper

(Bodson, Luik, Belgium) to determine the axial and radial swelling, respectively.

2.2.3.3. Porosity

The porosity of the mini-matrices was calculated, based on the difference between the

apparent and true volume of the extrudates. After measuring the exact height and diameter of the

extrudate (∅ = 3 mm, length = 30 mm), the total apparent volume was calculated (n = 10). The

true volume of 10 extrudates was measured using a helium pycnometer (AccuPyc 1330,

Micromeritics, Norcross, USA) (one analysis is the mean of 10 measurements, n = 10). The

porosity was expressed as a percentage of the apparent volume.

575757

CHAPTER 1

2.2.3.4. Friability

10 g of mini-matrices together with 200 glass beads (∅ = 4 mm) (to increase the

mechanical stress) were placed in a friabilator, equipped with an abrasion wheel (Pharma

Test Type PTF E, Hainburg, Germany). After rotating for 10 min (25 rpm), the glass beads

were removed. The size fraction > 250 µm was recovered by shaking for 5 min at an

amplitude of 2 mm using a Retsch VE 1000 Shaker (Retsch, Haan, Germany), and the

weight loss was expressed as a percentage of the initial weight of the mini-matrices. A

weight loss of not more than 1% was considered acceptable [33].

2.2.3.5. Crushing strength

The hardness of the mini-matrices was analyzed using a hardness tester (Pharma

Test Type PTB 311, Hainburg, Germany). 10 tablets were analyzed and the force needed to

break each tablet was registered.

2.2.3.6. Stability study

The XG75 formulations were selected for stability testing according to the USP

guidelines [33]. Mini-matrices were stored at 25 ± 2°C/60 ± 5% relative humidity (RH) and at

40 ± 2°C/75 ± 5% RH.

The in vitro drug release, moisture content and hardness of the mini-matrices were

determined as a function of storage time (12 months), storage conditions and xanthan gum

concentration.

The moisture content of the mini-matrices was determined using a Mettler Toledo

DL35 Karl Fisher titrator (Mettler Toledo, Beersel, Belgium) in combination with a Mettler

DO337 oven operated at 170°C. The samples (approximately 100 mg) were placed in the

oven and the moisture evaporated from the sample during 10 min was carried to the titration

vessel by a nitrogen stream (300 ml/min), after which the titration was started. Hydroquant

Uniquant 2® (Biosolve LTD, Valkenswaard, The Netherlands) with a theoretical titre of 2 mg

H2O/ml was used as the titrant solution. The analysis was performed in triplicate.

585858


Texture profile analysis was conducted on the mini-matrices, stored at the

aforementioned storage conditions, using a Texture Analyser Model TA.XT.plus (Stable

Micro Systems, Goldalming, UK). The force of resistance (N) against penetration (maximum

penetration depth of 1.0 mm) of a flat-tipped cylindrical steel probe (2 mm diameter) under

increasing load (test speed of 0.8 mm/s) was measured. The maximum recorded force is a

measure for the hardness of the mini-matrices. Data acquisition and analysis was performed

using the software Texture Exponent 32® (version 1.0.0.91). The analysis was performed in

tenfold.

2.2.3.7. Statistical analysis

The effect of xanthan gum concentration and particle size on the release mechanism and

drug release (using the area under the curve of the release profiles) was assessed by a two-way

ANOVA. To further compare the main effect of the different treatments a multiple comparison

among pairs of means was performed using a Bonferroni post-hoc test with P < 0.05 as

significance level. If a significant interaction between both factors was present, the simple effects

were investigated with a Bonferroni post-hoc test. The normality of the residuals was tested with a

Kolmogorov-Smirnov test. The homogeneity of variances was tested with Levene’s test. SPSS

version 12.0 was used to perform the statistical analysis.

2.2.4. In vivo evaluation

Although dogs are more sensitive to ibuprofen compared to humans and the

ulcerogenic activity of ibuprofen after chronic use is related to the plasma concentration of

the drug [16, 39-41], no gastro-intestinal side effects were expected after the single-dose

treatment used in this study.

All procedures were performed in accordance with the guidelines and approval of the

local Institutional Animal Experimentation Ethics Committee.

2.2.4.1. Subjects and study design

A group of 6 male mixed-breed dogs (weight 26.0-45.5 kg) was used in this study. To

investigate the influence of the xanthan gum concentration in the experimental mini-tablets on the

bioavailability of ibuprofen, the following drug formulations were administered:

595959

CHAPTER 1

F-1: hot-melt extruded mini-matrices, consisting of 60% IBP, 30% EC and 10% XG75 (10%

XG75)


XG75)


XG75)

F-4: Ibu-Slow® 600 mg (commercially available hydrophilic matrix tablets containing 600 mg

ibuprofen (Therabel Pharma, Brussels, Belgium))

F-5: Junifen® (100 mg/5 ml) (commercially available sugar-free syrup of ibuprofen (Boots

Healthcare, Wemmel, Belgium))

The mini-matrices of the experimental formulations were filled in hard-gelatin capsules n°

00, each capsule containing 300 mg ibuprofen. Ibu-Slow® 600 mg (½ tablet) and Junifen® (15 ml)

were administered as sustained- and immediate-release reference formulations, respectively. The

formulations were administered in randomized order with a wash-out period of at least 8 days

between consecutive sessions. On the experimental days the dogs were fasted for 12 h prior to

the study period, although water was available ad libitum. Before administration of the

formulations, an intravenous cannula was placed in the lateral saphenous and a blank blood

sample was obtained. The formulations were orally administered with 10 ml water. The blood

samples (2 ml at each sampling) were collected in dry heparinized tubes 1, 2, 4, 6, 8, 12, 24 and

36 h after intake of F-1, F-2, F-3, F-4, and at 0.25, 0.5, 1, 1.5, 2, 4, 6, 8, 12, 24 and 36 h after

administration of F-5. No food was administered to the dogs during the initial 24 h of the test,

afterwards they resumed their usual diet. Water could be taken freely. Within 1 h after collection,

the blood was centrifuged for 10 min at 1450g and the plasma was immediately assayed for

ibuprofen.

2.2.4.2. Ibuprofen assay

The plasma ibuprofen concentrations were determined by a validated HPLC-UV

method. All chemicals were of analytical or HPLC grade.

Fifty microlitres of an internal standard solution (30 µg/ml indomethacin in ethanol), 50 µl

ethanol and 500 µl plasma were transferred into a borosilicate glass tube. After 1 min of vortexing,

100 µl HCl (2 N) was added and homogenized by 1 min of vortexing. Consecutively, 4 ml

hexane/ether mixture (4/1, v/v) was added. After 3 min of vortexing and 5 min of centrifuging at

2500g, the upper organic layer was transferred into a new glass tube and evaporated to dryness

606060


under a nitrogen stream. The residue was dissolved in 200 µl mobile phase and 50 µl of this

solution was injected onto the column. The ibuprofen plasma concentrations were determined via

a calibration curve. These standards for the calibration curve were extracted using the same

procedure as described above: 500 µl blank plasma was spiked with 50 µl of internal standard

solution and 50 µl of a standard solution with a known concentration of ibuprofen in ethanol (0, 3,

6, 12, 30, 60, 90, 180 and 240 µg/ml).

The HPLC equipment (Merck-Hitachi, Darmstadt, Germany) consisted of a solvent

pump (L-7100) set at a constant flow rate of 1.5 ml/min, a variable wavelength UV-detector

(L-7400) set at 220 nm, a reversed-phase column and precolumn (LiChroCART® 125-4 and

4-4, LiChrospher® 100 RP-18 5 µm), an auto-sampler injection system (L-7200) with a 50 µl

loop (Valco Instruments Corporation, Houston, Texas, USA) equipped with an automatic

integration system (software D-7000 Multi-Manager). The mobile phase consisted of 0.1 M

KH2PO4 (adjusted to pH 7.0 with 2 M NaOH)/acetonitrile (12/3, v/v).

2.2.4.3. Method validation

Based on the guidelines of the International Conference on Harmonisation (ICH)

about the validation of analytical procedures, the following parameters were evaluated [42].

2.2.4.3.1. Specificity

The specificity can be defined as the ability to measure the analyte accurately in the

presence of interfering substances such as impurities and plasma components present in the

sample.

In our experiment, specificity was determined at 220 nm by comparing the

chromatograms after extraction of blank plasma, plasma spiked with internal standard

indomethacin, and plasma spiked with ibuprofen and indomethacin.

2.2.4.3.2. Linearity

The linearity of an analytical procedure is its ability to obtain test results which are

directly proportional to the concentration of the analyte in the samples.

616161

CHAPTER 1

During validation of the HPLC method, different calibration curves (n = 10) were

prepared. For each calibration curve, blank plasma of dogs was spiked with different

ibuprofen solutions to obtain plasma concentrations from 0 to 20 µg/ml, and internal standard

(indomethacin, 2.5 µg/ml) to cover the concentration range of the unknown samples.

Linearity was expressed by the coefficient of determination (R2) and the mean linear

regression equation (equation 4):

Y = aX + b (4)

where a = slope

b = intercept

X = ibuprofen concentration (µg/ml)

Y = absorption of ibuprofen at 220 nm

2.2.4.3.3. Precision

The precision of an analytical method expresses the closeness of agreement

between a series of measurements obtained when the method is applied repeatedly to

multiple samples of the same homogeneous sample. Precision may be considered at three

levels: repeatability (intra-assay precision), intermediate precision (within-laboratory

precision) and reproducibility (between-laboratory precision). Due to the fact that all samples

were analyzed at the same laboratory by the same person using the same equipment,

reproducibility was not considered.

The within-day repeatability was determined using the variation of the peak area of

samples injected at different times (n = 4) during the same day. The analyses performed over

five different days can be considered as the intermediate precision or within-laboratory

precision (between-day repeatability).

The precision was expressed as the coefficient of variation of a series of

measurements and was calculated for different ibuprofen concentrations (0.25, 0.5, 1, 2.5, 5,

7.5, 15 and 20 µg/ml). To have an acceptable precision the coefficient of variation of the

mean value should not exceed 15% [43].

626262


2.2.4.3.4. Accuracy

To determine the accuracy (or trueness) of an analytical procedure, the closeness of

agreement (expressed as %) between the value accepted as the conventional true value

(known spiked amount of analyte in the sample) and the mean value (n = 10) obtained using

the test procedure different times was calculated.

Accuracy was determined for different ibuprofen concentrations (0.25, 0.5, 1, 2.5, 5,

7.5, 15 and 20 µg/ml). The mean value should be within 15% of the actual value [43].

2.2.4.3.5. Recovery

The recovery is defined as the extent of agreement (expressed as %) between the

peak areas of an extracted sample (n = 10) versus the peak areas of a non-extracted sample

(n = 3).

The recovery experiments were performed at ibuprofen and indomethacin

concentrations of 0.25, 2.5, 7.5 and 20 µg/ml and 2.5 µg/ml, respectively.

2.2.4.3.6. Limit of detection and limit of quantification

The limit of detection is the lowest amount of analyte in a sample which can be

detected but not necessarily quantitated as an exact value. The limit of quantification is the

lowest amount of analyte in a sample which can be quantitatively determined with suitable

precision and accuracy.

The limit of detection (LOD) and limit of quantification (LOQ) were calculated based

on the standard deviation of the intercept and slope of the mean calibration curve (equation 5

and equation 6):

LOD = 3.3 x σ/S (5)

LOQ = 10 x σ/S (6)

where σ = standard deviation on the intercept (b) of the mean calibration curve

S = slope (a) of the mean calibration curve

636363

CHAPTER 1

2.2.4.4. Data analysis

The peak plasma concentration (Cmax), the time to reach Cmax (Tmax) and the extent of

absorption (area under the curve, AUC0-36h) were calculated using the MW-Pharm Program

version 3.0 (Mediware 1987-1991, Utrecht, The Netherlands). The AUC0-36h was calculated

using logarithmic and linear trapezoidal rules. The relative bioavailability (Frel, expressed in

%) was calculated as the ratio of AUC0-36h between a test formulation and the sustained-

release reference formulation (Ibu-Slow® 600 mg). The sustained-release characteristics of a

formulation were evaluated by the time span during which the plasma concentrations were at

least 50% of the Cmax value (HVDt50%Cmax, the width of the plasma concentration profile at

50% of Cmax) [44, 45]. The HVDt50%Cmax values were determined from the individual plasma

concentration-time profiles. The ratio between the HVDt50%Cmax values of a test formulation

and the immediate-release reference formulation (Junifen®) (expressed as RD) is indicative of

its sustained-release effect: a ratio of 1.5, 2 and > 3 indicating a low, intermediate and strong

sustained-release effect, respectively [44].


The effect of ibuprofen formulation on the bioavailability was assessed by repeated

measures ANOVA (univariate analysis). To further compare the effects of the different treatments,

a multiple comparison among pairs of means was performed using a Bonferroni post-hoc test with

P < 0.05 as significance level. The normality of the residuals was tested with a Kolmogorov-

Smirnov test. The sphericity of covariances was tested with Mauchly’s test. If the assumption of

sphericity was not fulfilled, the Huynh-Feldt correction was performed. SPSS version 12.0 was

used to perform the statistical analysis.

646464


3. RESULTS AND DISCUSSION

3.1. In vitro drug release: influence of xanthan gum concentration and particle size

Hot-melt extrusion was performed at 50°C and produced smooth extrudates. This

rather low extrusion temperature, far below the glass transition temperature of EC (130°C),

can be explained by the plasticizing effect of IBP on EC, favourably affecting the stability of

both the drug and the polymer [46]. Previous work at the Laboratory of Pharmaceutical

Technology confirmed the chemical stability of drug and polymers when hot-melt extruded at

this processing temperature using similar shearing forces [47]. The pressure and torque

during processing varied between 0 and 3 bar, and 30 and 44%, respectively. Figure 6

presents an example of an extrudate processed via hot-melt extrusion and of mini-matrices

obtained after manually cutting the extrudate.

Figure 6: Extrudate (left) and mini-matrices (right).

The drug release from the IBP/EC matrix was too slow as only 20% of IBP was

released in 24 h (Figure 7). The same slow drug release pattern was seen for theophylline

(40%, w/w) embedded in ethylcellulose tablets [48]. This sustained-release effect was

attributed to the integrity of the matrix structure which was maintained during the dissolution

experiment. Hydrophilic polymers such as HPMC [8] and XG had to be added to enhance the

drug release rate. Changing the XG/EC ratio allowed to modify the drug release rate as

increasing concentrations of XG (type XG75) enhanced drug release (Figure 7): after 24 h,

only 50% IBP was released from formulations containing 10% XG, whereas the total drug

load was released within this time from mini-matrices containing 20 and 30% XG.

Conventionally, in case of XG as a hydrophilic matrix, the drug release rate from matrices

decreased at higher XG concentrations, due to an increase in viscosity and thickness of the

hydrated gel layer, instantly formed around the dosage form upon contact with the dissolution

medium [11, 49]. This gel barrier restricted water penetration into and drug diffusion from the

tablet, delaying drug release [12-14, 34, 50-52]. However, as the system under investigation

is not a pure hydrophilic matrix but a combination of a hydrophobic matrix and a hydrophilic

656565

CHAPTER 1

polymer, the higher degree of swelling did not impede drug release as the swelling of XG

opened the structure of the mini-matrices, creating pores in the lipophilic EC matrix through

which IBP was released. The XG particle size had no effect on the drug release from

formulations containing 10 and 20% XG since the XG180 and XG11K formulations (results

not shown) yielded similar release profiles as presented in Figure 7 for the XG75 mini-

matrices. However, at 30% XG, the drug release was influenced by the particle size of the

hydrophilic polymer: a faster drug release from the mini-matrices formulated with coarser XG.

Those data are in agreement with the results reported by Dhopeshwarkar and Zatz [11] on

XG hydrophilic matrix tablets: a fine particle size of XG produced the slowest and most

reproducible sustained-release profiles, whereas the coarser XG fractions did not hydrate

fast enough to form a protective layer and the matrix tablet disintegrated during dissolution

testing. To evaluate the susceptibility of the coarser XG fractions to erosion, the erosion

behaviour (including liquid uptake and swelling) of the hydrated mini-matrices was

determined since these parameters affect the mechanism and kinetics of drug release.

0

20

40

60

80

100

120

0 4 8 12 16 20 24

Time (h)

IBP

rele

ased

(%)

Figure 7: Influence of xanthan gum concentration and particle size on the dissolution profiles

(mean ± standard deviation (S.D.), n = 6) of mini-matrices containing 60% IBP, EC and XG:

(△) 0% XG, (○) 10% XG75, (□) 20% XG75, (▲) 30% XG75, (■) 30% XG180, (●) 30%

XG11K.

666666


The mean porosity data of the mini-matrices (Table 2) indicated that the initial

porosity is independent of the composition of the mini-matrices and therefore not responsible

for any difference in drug release. Other researchers already reported that melt-extruded

dosage forms had a low porosity since compression and intense mixing of molten materials

during processing resulted in a product with low free volume [53].

Table 2: Porosity (%, mean ± S.D., n = 10) of the hot-melt extruded mini-matrices.

10% 20% 30%

XG75 1.4 ± 0.1 1.6 ± 0.1 3.2 ± 0.5

XG180 2.4 ± 0.7 4.2 ± 0.6 4.6 ± 0.1

XG11K 3.2 ± 0.1 5.3 ± 0.4 5.5 ± 0.1

3.2. Liquid uptake, erosion and swelling

Since the rate of swelling and erosion determines the mechanism and kinetics of drug

release, matrix swelling, liquid penetration and polymer (hydrophilic/lipophilic) erosion of the

hydrated mini-matrices were determined.

The influence of XG concentration and particle size on the liquid uptake, swelling and

erosion is presented in Figure 8. A higher liquid uptake (Figure 8a) was observed at

increasing XG180 concentrations: as expected from the hydrophilic nature of the polymer,

the mini-matrices became more accessible to the dissolution medium with increasing XG

concentration, resulting in a higher weight gain of the mini-matrices after immersion in the

medium (equation 2). Radial (Figure 8b) as well as axial (data not shown) swelling also

depended on the concentration of XG and was in all cases directly proportional to the liquid

uptake. Anisotropic swelling (more swelling in the longitudinal direction than in the radial

direction) was seen for the 10 and 20% XG formulations, independent of XG particle size.

However, anisotropic swelling for 30% XG mini-matrices could only be seen using the finest

XG particle size. When mini-matrices were formulated with 30% XG180 and XG11K, no rigid

structure could be recovered due to the higher impact of erosion, and therefore the swelling

in both directions was comparable. Similar phenomena during the swelling of HPMC

compacts were observed in literature [54, 55]. They related the axial relaxation of the HPMC

compacts to the relief of stress induced during compaction and the unidirectional swelling to

the orientation of molecules during compression. Since the hot-melt extrusion process also

676767

CHAPTER 1

induces stress on the extrudates, the asymmetric swelling behaviour could also be due to

this phenomenon. Formulations containing 10 and 20% XG (XG180 grade) did not swell to a

great extent, in contrast to the swelling capacity of the 30% XG formulation: for the latter, the

hydrated polymer layer was less cohesive, and therefore more sensitive to erosion (Figure

8c). After 4 h immersion in the medium, these mini-matrices became brittle and as a

consequence a higher erosion rate was seen, the matrix was completely eroded after 6 h.

The liquid uptake, swelling and erosion profiles for the 10 and 20% XG75 and XG11K

formulations (data not shown) coincided with those presented for the XG180 mini-matrices,

so no influence of the particle size was observed at these concentrations. However, when

comparing formulations containing 30% XG, the erosion rate was affected by the XG grade

(Figure 8c): using the coarser XG types, the mini-matrices were more susceptible to erosion,

whereas the gelled surface of the mini-matrices formulated with the XG75 grade seemed to

be stronger and therefore these mini-matrices were more resistant to erosion. The lower

liquid uptake and swelling of the XG180 and XG11K matrices at 4 h are also due to the

higher erosion rate of these formulations as a smaller amount of material was recovered.

686868


(a)

(b)

(c)

Figure 8: Influence of xanthan gum concentration and particle size on (a) liquid uptake, (b)

radial swelling and (c) erosion (mean ± S.D., n = 3) of mini-matrices containing 60% IBP, EC

and XG: (○) 10% XG180, (□) 20% XG180, (▲) 30% XG75, (■) 30% XG180, (●) 30% XG11K.

0

500

1000

1500

2000

2500

3000

0 2 4 6

Time (h)

Liqu

id u

ptak

e (%

)

8

-1

0

1

2

3

4

5

6

0 2 4 6 8

Time (h)

Rad

ial s

wel

ling

(mm

)

-20

0

20

40

60

80

100

120

0 2 4 6

Time (h)

Eros

ion

(%)

8

696969

CHAPTER 1

The difference in gel strength of XG when using different grades was confirmed by

visual inspection of the mini-matrices after 4 h immersion in the dissolution medium (Figure

9): the mini-matrices containing XG75 were embedded in a gel layer, and the inner core was

not yet fully hydrated. In contrast, the entire matrix was hydrated when using the coarsest XG

grade (XG11K), only a few non-hydrated individual particles remained. These pictures also

confirmed a decrease in diffusion path length with increasing XG particle size, resulting in a

faster drug release from the 30% XG11K mini-matrices (Figure 7).

(a) (b) (c)

Figure 9: Mini-matrices containing 60% IBP, 10% EC and 30% XG after 4 h immersion in the

dissolution medium: (a) XG75, (b) XG180, (c) XG11K.

3.3. Analysis of drug release mechanism

To investigate if the XG concentration and particle size had an effect on the IBP

release mechanism, the exponent characterizing drug release (n) and the area under the

curve (AUC) were compared. As the ranking of n (Table 3) and AUC (Table 4) in function of

concentration was independent of particle size, the main effect of XG concentration could be

analyzed (comparison of the overall mean of each column). However, as the ranking in

function of particle size depended on XG concentration only the simple effect could be

analyzed (at each specific XG concentration, i.e. within each column).

When increasing the XG concentration in the mini-matrices (regardless of the XG

particle size), the drug release tended towards transport by Case II diffusion (i.e., erosion) as

indicated by a significantly increasing exponent n (Table 3) (coefficient of determination of

0.9943 ± 0.0076 (n = 162)). This observation was confirmed by the erosion study (Figure 8c)

and the dissolution profiles (Figure 7) and correlated with the AUC values (Table 4), since a

higher contribution of erosion resulted in a faster drug release.

707070


Table 3: Influence of xanthan gum concentration and particle size on the calculated n values

(mean ± S.D., n = 6) of the dissolution profiles of mini-matrices (100 rpm).

For a cylinder, Fickian or Case I diffusion is defined by n = 0.45, zero-order drug release due to erosion or

relaxation (Case II diffusion) by n = 0.89 and anomalous behaviour or non-Fickian transport is indicated by

values between 0.45 and 0.89. Super Case II transport is defined by n values > 1.0 [35, 37].

10% 20% 30%

XG75 0.58 ± 0.08a 0.77 ± 0.03a 0.86 ± 0.07a

XG180 0.64 ± 0.03a

0.85 ± 0.09a 1.00 ± 0.07b

XG11K 0.55 ± 0.05a 0.85 ± 0.10a 1.06 ± 0.05b

Main effect concentration 0.59 ± 0.07A 0.82 ± 0.08B 0.97 ± 0.11C

a, b: means in the same column with different superscript are different at the 0.05 level of significance A, B, C: means in the same row with different superscript are different at the 0.05 level of significance

Table 4: Influence of xanthan gum concentration and particle size on the calculated AUC

values (mean ± S.D., n = 6) of the dissolution profiles of mini-matrices (100 rpm).

10% 20% 30%

XG75 794 ± 32a 1637 ± 27a 1993 ± 36a

XG180 792 ± 31a

1412 ± 28b 2136 ± 33b

XG11K 981 ± 23b 1469 ± 12c 2236 ± 25c

Main effect concentration 856 ± 95A 1506 ± 101B 2122 ± 107C

a, b, c: means in the same column with different superscript are different at the 0.05 level of significance A, B, C: means in the same row with different superscript are different at the 0.05 level of significance

Regarding the effect of XG particle size, the IBP drug release mechanism did not

differ significantly for the 10 and 20% XG formulations, although a significant, but not

relevant, higher AUC value was obtained for the XG11K and XG75 formulations,

respectively. Both formulations followed a similar release pattern: in all cases anomalous

transport with n values from 0.55 up to 0.85. As the core of these mini-matrices could be

recovered after 24 h, there is a strong indication that the drug release mechanism is not

primarily erosion-controlled. However, for the 30% XG formulation, anomalous transport was

seen for the XG75 type (n = 0.86), whereas the release mechanism was mainly erosion-

717171

CHAPTER 1

controlled for the mini-matrices with a coarser XG particle size (n = 1.06). This observation

was expected based on the erosion study (Figure 8c and 9) since a higher erosion rate was

seen for the XG180 and XG11K mini-matrices.

3.4. In vitro drug release: influence of hydrodynamics

The influence of the hydrodynamic stress on the release rate is negligible for a

diffusion-controlled system, but has a great influence on an erosion-based system [56].

Therefore, dissolution experiments were carried out at different stirring rates (50, 100 and

200 rpm). The n and AUC values for IBP release (50 and 200 rpm data not shown) clearly

indicated that for all formulations, regardless of the XG concentration and particle size, the

drug release mechanism was not affected by the rotational speed. For the 30% XG

formulations, this was surprising in view of the high erosion rates observed for the XG11K

formulations (Figure 8c and 9). As an example, the mean dissolution profiles of 30% XG11K

mini-matrices at different rotation speeds are shown (Figure 10).

0

20

40

60

80

100

120

0 4 8 12 16 20 2Time (h)

IBP

rele

ased

(%)

4

Figure 10: Dissolution profiles (mean ± S.D., n = 6) of mini-matrices containing 60% IBP,

10% EC and 30% XG11K at different stirring speeds: (▲) 50 rpm, (■) 100 rpm, (○) 200 rpm.

727272


Since drug release was unaffected by the hydrodynamic conditions, we concluded

that, although erosion of the mini-matrix is contributing to faster drug release from the 30%

XG formulations, the drug release mechanism from the mini-matrices was mainly diffusion-

controlled. The faster drug release observed using a coarser particle size (XG11K type) for

the 30% XG formulations was more related to the faster hydration of the polymer particles.

3.5. In vitro drug release: influence of ionic strength of the dissolution medium

The susceptibility of the 20% XG75 containing matrices to the ionic strength of the

dissolution medium is shown in Figure 11. The slowest drug release was seen at the lowest

ionic strength: after 24 h in a medium of µ = 0.011, only 50% IBP was released, whereas the

total drug load was released in a medium of µ = 0.055 or higher. Above an ionic strength of

0.11 no differences in drug release profiles were observed, which was also reported by

Talukdar and Kinget [34], since under these conditions XG formed a helical structure and

further addition of salts had no influence on its rheological properties [57].

0

20

40

60

80

100

120

0 4 8 12 16 20

Time (h)

IBP

rele

ased

(%

24

)

Figure 11: Dissolution profiles (mean ± S.D., n = 6) of mini-matrices containing 60% IBP,

20% EC and 20% XG75 at different ionic strengths: (●) µ = 0.011, (■) µ = 0.022, (▲) µ =

0.055, (○) µ = 0.11, (□) µ = 0.15, (△) µ = 0.2.

737373

CHAPTER 1

A similar drug release dependency on the ionic strength was noticed for the 10 and

30% XG75 mini-matrices (data not shown). Visual inspection of the XG matrices (Figure 12)

after 24 h of dissolution testing showed enhanced swelling of the mini-matrices in a

dissolution medium of high ionic strength, which correlated with a higher IBP release. This

behaviour might be attributed to the conformation of XG molecules which depended on the

salt concentration. In aqueous solutions, a transition of XG occurs from a random coil at low

ionic strength to an ordered elongated double helix with increasing salt concentrations. The

ordered elongated conformation is able to form a gel, unlike the random coil [34]. Hence, at

the lower ionic strengths, the disordered XG molecules resulted in less gel formation, a

reduced swelling of the mini-matrices and a slower drug release rate. Based on these

findings it is likely that within the physical range of gastric and intestinal ionic strengths (µ =

0.01-0.166 according to Johnson et al. [58]) the drug release of XG containing formulations

would be affected. The higher drug release in media of increasing salt concentrations is

consistent with data reported by other investigators [12, 13, 34, 49, 59-62].

(a) (b)

Figure 12: Mini-matrices containing 60% IBP, 20% EC and 20% XG75 after 24 h immersion

in the dissolution medium at different ionic strengths: (a) µ = 0.011, (b) µ = 0.2.

3.6. Friability and hardness

The physical properties of the mini-matrices extruded with different XG concentrations

and particle sizes are summarized in Table 5. Increasing the XG concentration resulted in an

increase of friability, which was correlated with a decrease of crushing strength. This can be

due to the weaker binding forces between EC and XG in comparison with EC-EC bonds. The

influence of XG particle size was negligible. The low friability (0.01-0.32%) of the mini-

matrices gives an indication of the resistance to abrasion, and the results were conform to

the requirements set forth in section 2.2.3.4.

747474


Table 5: Friability (%, n = 1) and hardness (N, mean ± S.D., n = 10) of the hot-melt extruded

mini-matrices.

Friability (%) Hardness (N)

XG75 10% 0.01 39.1 ± 8.3

20% 0.03 32.5 ± 5.5

30% 0.19 28.6 ± 3.5

XG180 10% 0.01 41.0 ± 7.4

20% 0.07 32.0 ± 4.7

30% 0.23 25.9 ± 3.9

XG11K 10% 0.03 40.6 ± 3.2

20% 0.13 29.5 ± 3.1

30% 0.32 24.5 ± 3.8

3.7. Stability study

Due to the incorporation of the hydrophilic polymer, these formulations might be

susceptible to storage conditions. Therefore, the stability of these mini-matrices was

evaluated under specific conditions of relative humidity and temperature.

The release profiles indicated that all experimental mini-matrices were stable for at

least 12 months during storage at 25°C/60% RH. Storage at 40°C/75% RH did not affect

ibuprofen release from 10% XG mini-matrices. However, the drug release from the 20 and

30% XG formulations was increased: after 12 months, the amount of drug released after 4 h

increased from 30 and 42% for the 20 and 30% XG mini-matrices, respectively, to 51 and

75%, respectively.

Immediately after production of the mini-matrices, the water content was similar (0.04,

0.08 and 0.09% for the 10, 20 and 30% XG formulation, respectively) (Figure 13). Storage during

12 months at both conditions resulted in an increase of the water uptake which depended on XG

concentration and storage conditions: 3.6 and 4.6% for the 20 and 30% XG formulations,

respectively, when stored at 25°C/60% RH; 4.3 and 6.8%, respectively, when stored at 40°C/75%

RH. No difference was seen in water uptake of the 10% XG formulations between both storage

conditions (2.3 and 2.0% for 25°C/60% RH and 40°C/75% RH, respectively).

757575

CHAPTER 1

0

1

2

3

4

5

6

7

8

10% XG75 20% XG75 30% XG75

Wat

er c

onte

nt (%

)

0 months 12 months 25°C/60% RH 12 months 40°C/75% RH

Figure 13: Water content (%, mean ± S.D., n = 3) of mini-matrices in function of xanthan gum

concentration, and storage time and conditions.

Although the mini-matrices had the same water content after the extrusion process, they

were softer with increasing XG concentration using the indentation test (116, 95 and 73 N for the

10, 20 and 30% XG mini-matrices, respectively) (Figure 14).

0

20

40

60

80

100

120

140

10% XG75 20% XG75 30% XG75

Har

dnes

s (N

)

0 months 12 months 25°C/60% RH 12 months 40°C/75% RH

Figure 14: Hardness (N, mean ± S.D., n = 10) of mini-matrices in function of xanthan gum

concentration, and storage time and conditions.

767676


From the results of the texture analysis it is clear that the mini-matrices became softer

during 12 months storage: at 25°C/60% RH 103 and 80 N for the 10 and 20% XG mini-matrices,

respectively, and at 40°C/75% RH 83 and 71 N for the 10 and 20% XG mini-matrices,

respectively. No difference in tablet hardness was observed for the 30% XG mini-matrices when

stored at the different conditions (51 and 52 N at 25°C/60% RH and 40°C/75% RH, respectively).

The stability data revealed that drug release, water uptake and hardness of the XG

formulations changed when stored for 12 months at 25°C/60% RH and 40°C/75% RH, and

therefore these mini-matrices should be appropriately packaged.

3.8. In vivo evaluation 3.8.1. Method validation

3.8.1.1. Specificity

Chromatograms after extraction of blank plasma, plasma spiked with indomethacin

(internal standard) and plasma spiked with ibuprofen and indomethacin are shown in Figure

15. The chromatograms indicate that there was no interference of drug and internal standard

with endogenous compounds, and separation between ibuprofen and indomethacin was

complete. As shown in Figure 15, a retention time of ± 12 min and ± 32 min was obtained for

ibuprofen and indomethacin, respectively. It was concluded that the method was selective to

determine ibuprofen in dog plasma.

777777

CHAPTER 1

Figure 15: Chromatograms showing ibuprofen and indomethacin after extraction of (top)

blank dog plasma, (middle) plasma spiked with indomethacin and (bottom) plasma spiked

with ibuprofen and indomethacin.

Absorption

Time (min)

787878


3.8.1.2. Linearity

Linearity was expressed by the mean linear regression equation (slope and intercept)

and its coefficient of determination (n = 10).

The mean calibration curve (Y = 0.2527 (± 0.0436) X – 0.0101 (± 0.0235)) was linear

between 0 and 20 µg/ml ibuprofen, with a coefficient of determination of 0.9997 ± 0.0003.

3.8.1.3. Precision

The coefficients of variation (C.V.) (%) of the repeatability (intra-assay precision) and

the intermediate precision (within-laboratory precision) are shown in Table 6.

Table 6: Coefficients of variation (%) of the repeatability (n = 4) and the intermediate

precision (n = 5).

Coefficient of variation (%) Concentration (µg/ml)

Repeatability Intermediate precision

0.25 8.0 12.5

0.50 6.3 7.9

1.00 4.3 7.1

2.50 3.8 4.9

5.00 5.8 4.8

7.50 2.9 6.2

15.00 3.9 4.4

20.00 3.7 7.6

For the repeatability as well as the intermediate precision, the coefficients of variation

at all concentrations were lower than 15%.

797979

CHAPTER 1

3.8.1.4. Accuracy

The accuracy was determined for all ibuprofen concentrations and the mean results

are presented in Table 7.

Table 7: Accuracy (%, mean ± S.D., n = 10).

Concentration (µg/ml) Calculated concentration (µg/ml) Accuracy (%) C.V. (%)

0.25 0.22 ± 0.03 86.8 ± 12.3 14.1

0.50 0.47 ± 0.04 94.3 ± 8.3 8.8

1.00 0.95 ± 0.03 95.4 ± 3.2 3.3

2.50 2.54 ± 0.11 101.4 ± 4.3 4.2

5.00 4.97 ± 0.13 99.4 ± 2.5 2.5

7.50 7.50 ± 0.29 100.0 ± 3.9 3.9

15.00 15.51 ± 0.28 103.4 ± 1.8 1.8

20.00 19.63 ± 0.15 98.1 ± 0.7 0.7

The proposed method can be considered as accurate as an accuracy of at least

86.8% was obtained and the highest coefficient of variation was 14.1%.

3.8.1.5. Recovery

The recovery of ibuprofen and indomethacin from plasma are shown in Table 8.

Table 8: Recovery (%, mean ± S.D., n = 10) of ibuprofen and indomethacin from plasma.

Concentration (µg/ml) Ibuprofen Indomethacin

0.25 83.3 ± 10.9 -

2.50 84.4 ± 5.6 73.5 ± 14.7

7.50 82.4 ± 5.1 -

20.00 88.9 ± 9.1 -

808080


3.8.1.6. Limit of detection and limit of quantification

Based on the mean linear regression equation of ibuprofen the limit of detection and

the limit of quantification for ibuprofen are 0.31 and 0.93 µg/ml, respectively. Both

parameters were higher than the lowest concentration of the calibration curve but as

accuracy and precision for the concentration of 0.25 µg/ml were within accepted limits [43],

the concentration of 0.25 µg/ml was considered as limit of quantification.

3.8.2. In vivo performance

In the previous paragraphs of this chapter, the in vitro behaviour of the mini-matrices

formulated with ethylcellulose and xanthan gum was described. It showed that the drug release

from these mini-tables could be modified by appropriate selection of the type and concentration of

the hydrophilic component (xanthan gum). To investigate the biopharmaceutical properties of

these sustained-release formulations an in vivo study using dogs was performed to determine the

bioavailability of ibuprofen from a sustained-release hot-melt extruded mini-tablet formulation. The

bioavailability of the experimental formulations was compared to the in vivo behaviour of a

commercial sustained- and immediate-release dosage form.

An in vitro burst release was observed at 30% XG when using the coarsest particle size

XG11K (Figure 7). Therefore, to prevent the burst release in vivo, the formulations containing

XG75 (finest XG particle size) were selected for in vivo analysis.

818181

CHAPTER 1

Figure 16 shows the mean plasma concentration-time profiles after oral administration of

300 mg ibuprofen to 6 dogs as Junifen®, Ibu-Slow® 600 mg (½ tablet) and the experimental mini-

matrices. The pharmacokinetic parameters are reported in Table 9.

0

10

20

30

40

0 10 20 30

Time (h)

Con

c. IB

P (µ

g/m

l)

40

Figure 16: Mean plasma concentration-time profiles (± S.D., n = 6) after oral administration of

300 mg ibuprofen to dogs: (●) 10% XG75 mini-matrices, (▲) 20% XG75 mini-matrices, (■)

30% XG75 mini-matrices, (○) Ibu-Slow® 600 (½ tablet), (△) Junifen® (15 ml).

With regard to the extent of absorption, the mean AUC0-36h of the formulations F-1, F-

2, F-3 and Ibu-Slow® 600 mg (½ tablet), were 34.8, 129.7, 134.9 and 186.2 µg.h/ml,

respectively; yielding relative bioavailabilities of the mini-matrices versus Ibu-Slow® 600 mg

(½ tablet) of 19.3, 70.6 and 73.8%, respectively. The similar in vivo behaviour of the 20 and

30% XG formulations was not reflected in their in vitro dissolution profiles (Figure 7) since a

significant difference in drug release was noticed for these formulations. Therefore, it was

investigated if the drug release properties were altered when a larger sample size (500 mg)

of mini-matrices was filled into a capsule (similar to the dosage form that was administered to

the dogs).

828282


Table 9: Mean pharmacokinetic parameters (± S.D., n = 6) after oral administration of 300

mg ibuprofen to dogs as 10% XG75 mini-matrices, 20% XG75 mini-matrices, 30% XG75

mini-matrices, Ibu-Slow® 600 (½ tablet) and Junifen® (15 ml).

Cmax

(µg/ml)

Tmax

(h)

AUC0-36h

(µg.h/ml)

HVDt50%Cmax

(h)

Frel*

(%)

RD

10% XG75 3.0 ± 0.5a 5.7 ± 3.9a,b 34.8 ± 8.8a 10.6 ± 3.8a,b 19.3 ± 5.5a 3.6 ± 1.4a

20% XG75 9.8 ± 2.3b

6.3 ± 2.9a,b 129.7 ± 40.0b 9.5 ± 4.0a,b 70.6 ± 22.9b 3.1 ± 1.2a

30% XG75 15.8 ± 4.6b,c 3.7 ± 0.8b 134.9 ± 32.6b 6.9 ± 2.1b 73.8 ± 19.5b 2.2 ± 0.6a

Ibu-Slow® 600 (½ tablet)

®

18.4 ± 5.7c 4.3 ± 2.0a,b 186.2 ± 39.8b 8.5 ± 1.8b - 2.8 ± 0.7a

Junifen(15 ml) 32.1 ± 3.7d 1.0 ± 0.0a 175.4 ± 46.9b 3.1 ± 0.4a - -

-: not applicable *: using Ibu-Slow® 600 (½ tablet) as reference a, b, c, d: means in the same column with different superscript are different at the 0.05 level of

significance

Similar dissolution profiles (Figure 17) and swelling (Figure 18) were obtained for the

10% XG75 formulation, irrespective of the amount of mini-matrices. However, the release

rate for the mini-matrices containing 20 and 30% XG was reduced (Figure 17) by increasing

the sample size in the dissolution basket, although sink conditions were maintained (solubility

of ibuprofen: 4.48 ± 0.08 mg/ml at 37°C [8]). Due to the rapid swelling of XG at 20 and 30%

after hydration by the dissolution medium, the mini-matrices stick together with increasing

sample size (Figure 18): the increased diffusion path length reduced the drug release rate

and similar drug release profiles were obtained at both XG concentrations. The formulations

F-2 and F-3 had a pronounced sustained-release profile with RD values of 3.1 and 2.2,

respectively. In comparison with the 30% XG mini-tablets, the 20% XG formulation is

characterized by a higher Tmax (3.7 and 6.3 h, respectively) and a lower Cmax (15.8 and 9.8

µg/ml, respectively), which is typical for sustained-release formulations and which is reflected

in its higher HVDt50%Cmax and RD. It was interesting to observe that during the initial phase of

the plasma concentration-time profiles a slower absorption was seen for the mini-matrices

containing 20% XG. This delayed onset of action can be seen as an advantage since it could

favourably affect the clinical effect of an ibuprofen dose administered in the evening by

preventing the morning stiffness typically associated with rheumatoid arthritis. Examining the

plasma concentration-time profiles revealed a constant drug absorption pattern over 36 h for

838383

CHAPTER 1

the 20% XG mini-matrices compared with the 30% XG mini-matrices and the reference

formulation. From the data in Table 9, it can be concluded that Cmax and Tmax are similar for

the 30% XG mini-tablets and Ibu-Slow® 600 mg (½ tablet). The similar pharmacokinetic

parameters of both formulations supported the hypothesis that the 30% XG mini-matrices

behaved in vivo as a single-unit dosage form instead of multiparticulates due to the

immediate swelling of the 30% XG mini-matrices upon contact with the gastro-intestinal fluids

and the formation of a plug. Although swelling was also noticed for the 20% XG mini-tablets,

they performed in vivo as a multiparticulate dosage form since they differ in Cmax and Tmax

values from Ibu-Slow® 600 mg (½ tablet). This bioavailability study demonstrated that the

matrix mini-tablets formulated with EC and XG can be used to prepare sustained-release

dosage forms.

0

20

40

60

80

100

120

0 4 8 12 16 20 2

Time (h)

IBP

rele

ased

(%)

4

Figure 17: Influence of xanthan gum concentration and sample size introduced in the

dissolution medium on the drug release profiles (mean ± S.D., n = 6) of mini-matrices

containing 60% IBP, EC and XG: (●) 10% XG75 60 mg, (○) 10% XG75 500 mg, (▲) 20%

XG75 60 mg, (△) 20% XG75 500 mg, (■) 30% XG75 60 mg, (□) 30% XG75 500 mg.

848484


(a) (b) (c)

(d) (e) (f)

Figure 18: Mini-matrices containing 60% IBP, EC and XG after 6 h immersion in the

dissolution medium: (a) 10% XG75 60 mg, (b) 20% XG75 60 mg, (c) 30% XG75 60 mg, (d)

10% XG75 500 mg, (e) 20% XG75 500 mg, (f) 30% XG75 500 mg.

858585

CHAPTER 1

4. CONCLUSIONS

The present work was conducted in order to assess the influence of xanthan gum

(concentration and particle size) on the in vitro release of ibuprofen from ethylcellulose mini-

matrices manufactured by hot-stage extrusion. Since the drug release was modified by

changing the xanthan gum concentration as well as the xanthan gum particle size, the mini-

matrices based on a combination of ethylcellulose and this hydrophilic additive offered a

flexible system to tailor the drug release to the required specifications. Increasing xanthan

gum concentrations yielded a faster drug release, higher liquid uptake, swelling and erosion

rate. The low susceptibility to the agitation speed suggested that diffusion is the predominant

mechanism of drug release. However, diffusion was not the only mechanism controlling drug

release from these systems as the rate and extent of drug release was also dependent on

the swelling (relaxation) of the hydrated polymer mass.

From the validation, it could be concluded that a specific, linear, precise and accurate

HPLC method was obtained. In vivo data showed that oral administration of a xanthan

gum/ethylcellulose formulation was able to sustain plasma levels in dogs.

868686


REFERENCES

[1] W. Mueller, R. Spengler, S. Grabowski, A. Sanner, Feste pharmazeutische

Retardforme, European Patent 0 544 144 B1, BASF Aktiengesellschaft (1992).

[2] J. Breitenbach, S. Grabowski, J. Rosenberg, Extrusion von Polymer-Wirkstoff-

Gemischen zur Herstellung von Arzneiformen, Spektrum der Wissenschaft, July, 18-20

(1995).


Pharm. Biopharm. 54 (2002) 107-117.





for the production of polymer based pellets for sustained release capsules containing






McGinity, Physicochemical properties and mechanism of drug release from ethyl

cellulose matrix tablets prepared by direct compression and hot-melt extrusion, Int. J.

Pharm. 269 (2004) 509-522.



235-247.

[9] W.M. Van Den Abeele, C. De Brabander, C. Vervaet, J.P. Remon, L.M. Van Bortel,

Bioavailability of ibuprofen from hot-melt extruded mini-matrices, Brit. J. Clin.

Pharmacol. 57 (2004) 366-367.

878787

CHAPTER 1

[10] C. De Brabander, C. Vervaet, L. Van Bortel, J.P. Remon, Bioavailability of ibuprofen

from hot-melt extruded mini-matrices, Int. J. Pharm. 271 (2004) 77-84.

[11] V. Dhopeshwarkar, J.L. Zatz, Evaluation of xanthan gum in the preparation of

sustained release matrix tablets, Drug Dev. Ind. Pharm. 19 (1993) 999-1017.

[12] M.M. Talukdur, A. Michoel, P. Rombaut, R. Kinget, Comparative study on xanthan gum

and hydroxypropyl methylcellulose as matrices for controlled drug delivery. I.

Compaction and in vitro drug release behaviour, Int. J. Pharm. 129 (1996) 233-241.

[13] M.M. Talukdur, R. Kinget, Comparative study on xanthan gum and hydroxypropyl

methylcellulose as matrices for controlled-release drug delivery. II. Drug diffusion in

hydrated matrices, Int. J. Pharm. 151 (1997) 99-107.

[14] M.M. Talukdur, I. Vinckier, P. Moldenaers, R. Kinget, Rheological characterisation of

xanthan gum and hydroxypropyl methylcellulose with respect to controlled-release drug

delivery, J. Pharm. Sci. 85 (1996) 537-540.

[15] S.S. Adams, P. Bresloff, C.G. Mason, Pharmacological differences between the optical

isomers of ibuprofen: evidence for metabolic inversion of the (-) isomer, J. Pharm.

Pharmacol. 28 (1976) 256-257.

[16] N.M. Davies, Clinical pharmacokinetics of ibuprofen – The first 30 years, Clin.

Pharmacokinet. 34 (1998) 101-154.

[17] K.J. Nichol, The medicinal chemistry of ibuprofen, in: Ibuprofen. A critical bibliographic

review (K.D. Rainsford, ed.), Taylor and Francis, Londen, UK (1999) 25-51.

[18] D.R. Brocks, F. Jamali, The pharmacokinetics of ibuprofen in humans and animals, in:

Ibuprofen. A critical bibliographic review (K.D. Rainsford, ed.), Taylor and Francis,

Londen, UK (1999) 87-142.

[19] E.J.D. Lee, K. Williams, R. Day, G. Graham, D. Champion, Stereoselective disposition

of ibuprofen enantiomers in man, Brit. J. Clin. Pharmacol. 19 (1985) 669-674.

[20] H.Y. Ahn, G.L. Amidon, D.E. Smith, Stereoselective systemic disposition of ibuprofen

enantiomers in the dog, Pharm. Res. 8 (1991) 1186-1190.

888888


[21] F. Highton, The pharmaceutics of ibuprofen, in: Ibuprofen. A critical bibliographic

review (K.D. Rainsford, ed.), Taylor and Francis, Londen, UK (1999) 53-86.

[22] S. Hülsmann, T. Backensfeld, S. Keitel, R. Bodmeier, Melt extrusion – An alternative

method for enhancing the dissolution rate of 17 β-estradiol hemihydrate, Eur. J. Pharm.

Biopharm. 49 (2000) 237-242.

[23] C. Leuner, J. Dressman, Improving drug solubility for oral delivery using solid

dispersions, Eur. J. Pharm. Biopharm. 50 (2000) 47-60.

[24] A. Forster, J. Hempenstall, T. Rades, Characterisation of glass solutions of poorly

water-soluble drugs produced by melt extrusion with hydrophilic amorphous polymers,

J. Pharm. Pharmacol. 53 (2001) 303-315.

[25] A. Forster, J. Hempenstall, I. Tucker, T. Rades, The potential of small-scale fusion

experiments and the Gordon-Taylor equation to predict the suitability of drug/polymer

blends for melt extrusion, Drug Dev. Ind. Pharm. 27 (2001) 549-560.

[26] A. Forster, J. Hempenstall, I. Tucker, T. Rades, Selection of excipients for melt

extrusion with two poorly water-soluble drugs by solubility parameter calculation and

thermal analysis, Int. J. Pharm. 226 (2001) 147-161.

[27] R.C. Rowe, P.J. Sheskey, P.J. Weller (ed.), Handbook of pharmaceutical excipients

(Fourth Edition), Pharmaceutical Press and American Pharmaceutical Association,

Washington D.C., USA (2003).

[28] M.M. Talukdar, G. Van den Mooter, P. Augustijns, T. Tjandra-Maga, N. Verbeke, R.

Kinget, In vivo evaluation of xanthan gum as a potential excipient for oral controlled-

release matrix tablet formulation, Int. J. Pharm. 169 (1998) 105-133.

[29] F. Ughini, I.F. Andreazza, J.L.M.S. Ganter, T.M.B. Bresolin, Evaluation of xanthan and

highly substituted galactomannan from M. scabrella as a sustained release matrix, Int.

J. Pharm. 271 (2004) 197-205.


release characteristics of xanthan gum pellets of different compositions, Eur. J. Pharm.

Sci. 21 (2004) 271-281.

898989

CHAPTER 1


release properties of diclofenac sodium and ibuprofen pellets comprising xanthan gum

as a sustained release agent, Int. J. Pharm. 295 (2005) 15-27.

[32] C.W. Vendruscolo, I.F. Andreazza, J.L.M.S. Ganter, C. Ferrero, T.M.B. Bresolin,

Xanthan and galactomannan (from M. scabrella) matrix tablets for oral controlled

delivery of theophylline, Int. J. Pharm. 296 (2005) 1-11.

[33] U.S. Pharmacopeia 27 & National Formulary 22, The United States Pharmacopeial

Convention, Rockville, MD, USA (2004).

[34] M.M. Talukdur, R. Kinget, Swelling and drug release behaviour of xanthan gum matrix

tablets, Int. J. Pharm. 120 (1995) 63-72.

[35] N.A. Peppas, Analysis of Fickian and non-Fickian drug release from polymers, Pharm.

Acta Helv. 60 (1985) 110-111.

[36] P.L. Ritger, N.A. Peppas, A simple equation for description of solute release: I. Fickian

and non-Fickian release from non-swellable devices in the form of slabs, spheres,

cylinders or discs, J. Control. Release 5 (1987) 23-26.

[37] P.L. Ritger, N.A. Peppas, A simple equation for description of solute release: II. Fickian

and anomalous release from swellable devices, J. Control. Release 5 (1987) 37-42.

[38] N.A. Peppas, J.J. Sahlin, A simple equation for description of solute release: III.

Coupling of diffusion and relaxation, Int. J. Pharm. 57 (1989) 169-172.

[39] R. Scherkl, H.H. Frey, Pharmacokinetics of ibuprofen in the dog, J. Vet. Pharmacol.

Ther. 10 (1987) 261-265.

[40] S.S. Adams, R.G. Bough, E.E. Cliffe, W. Dickinson, B. Lessel, K.F. McCullough, R.F.N.

Mills, J.S. Nicholson, G.A.H. Williams, Some aspects of the pharmacology, metabolism

and toxicology of ibuprofen, Rheumatol. Phys. Med. 9 (1970) 9-26.

[41] R. Sapierzyaski, M. Galanty, Gastric ulcers – Five cases in dogs, Med. Weter. 60

(2004) 392-395.

909090


[42] International Conference on Harmonisation, ICH Harmonised Tripartite Guideline –

Validation of Analytical Procedures: Text and Methodology Q2 (R1) (Parent Guideline,

1994 and Complementary Guideline on Methodology, 1996) (2005).

[43] V.P. Shah, K.K. Midha, S. Dighe, I.J. McGilveray, J.P. Skelly, A. Yacobi, T. Layloff,

C.T. Viswanathan, C.E. Cook, R.D. McDowall, K.A. Pittman, S. Spector, Analytical

methods validation: bioavailability, bioequivalence and pharmacokinetic studies, J.

Pharm. Sci. 81 (1992) 309-312.

[44] J. Meier, E. Nuesch, R. Schmidt, Pharmacokinetic criteria for the evaluation of retard

formulations, Eur. J. Clin. Pharmacol. 7 (1974) 429-432.

[45] V.W. Steinijans, Pharmacokinetic characterization of controlled-release formulations,

Eur. J. Drug Metab. Pharmacokinet. 15 (1990) 173-181.

[46] C. De Brabander, G. Van den Mooter, C. Vervaet, J.P. Remon, Characterisation of


1678-1685.

[47] C. De Brabander, Development of sustained release multiple unit dosage forms using

hot-melt extrusion, Thesis presented at the Laboratory of Pharmaceutical Technology,

Ghent University, Ghent, Belgium (2002).

[48] O. Chambin, D. Champion, C. Debray, M.H. Rochat-Gonthier, M. Le Meste, Y.

Pourcelot, Effects of different cellulose derivatives on drug release mechanism studied

at a preformulation stage, J. Control. Release 95 (2004) 101-108.

[49] M.M. Talukdar, J. Plaizier-Vercammen, Evaluation of xanthan gum as a hydrophilic

matrix for controlled-release dosage form preparations, Drug Dev. Ind. Pharm. 19

(1993) 1037-1046.

[50] M.F. Lu, L. Woodward, S. Borodkin, Xanthan gum and alginate based controlled

release theophylline formulations, Drug Dev. Ind. Pharm. 17 (1991) 1987-2004.

[51] P.J. Cox, K.A. Khan, D.L. Munday, J. Sujja-areevath, Development and evaluation of a

multiple-unit oral sustained release dosage form for S(+)-ibuprofen: preparation and

release kinetics, Int. J. Pharm. 193 (1999) 73-84.

919191

CHAPTER 1

[52] N. Billa, K. Yuen, M.A.A. Khader, A. Omar, Gamma-scintigraphic study of the

gastrointestinal transit and in vivo dissolution of a controlled release diclofenac sodium

formulation in xanthan gum matrices, Int. J. Pharm. 201 (2000) 109-120.


containing granulations of ibuprofen and acrylic copolymer prepared by thermal


[54] E. Papadimitriou, G. Buckton, M. Efentakis, Probing the mechanisms of swelling of

hydroxypropylmethylcellulose matrices, Int. J. Pharm. 98 (1993) 57-62.

[55] P. Gao, J.W. Skoug, P.R. Nixon, T.R. Ju, N.L. Stemm, K.C. Sung, Swelling of

hydroxypropyl methylcellulose matrix tablets. 2. Mechanistic study of the influence of

formulation variables on matrix performance and drug release, J. Pharm. Sci. 85 (1996)

732-739.

[56] W.D. Lindner, B.C. Lippold, Drug release from hydrocolloid embeddings with high or

low susceptibility to hydrodynamic stress, Pharm. Res. 12 (1995) 1781-1785.

[57] F. Callet, M. Milas, M. Rinaudo, Influence of acetyl and pyruvate contents on

rheological properties of xanthan in dilute solution, Int. J. Biol. Macromol. 9 (1987) 291-

293.

[58] J.L. Johnson, J. Holiney, M.D. Williams, Influence of ionic strength on matrix integrity

and drug release from hydroxypropyl cellulose compacts, Int. J. Pharm. 90 (1993) 151-

159.

[59] H.M. Ingani, A.J. Moes, Utilisation de la gomme xanthane dans la formation des

matrices hydrophiles, STP Pharm. 4 (1988) 188-195.

[60] X. Mu, M.J. Tobyn, J.N. Staniforth, Influence of physiological variables on the in vitro

drug-release behaviour of a polysaccharide matrix controlled-release system, Drug

Dev. Ind. Pharm. 29 (2003) 19-29.

[61] M. Fukuda, N.A. Peppas, J.W. McGinity, Properties of sustained release hot-melt

extruded tablets containing chitosan and xanthan gum, Int. J. Pharm. 310 (2006) 90-

100.

929292


[62] S. Baumgartner, M. Pavli, J. Kristl, Effect of calcium ions on the gelling and drug

release characteristics of xanthan matrix tablets, Eur. J. Pharm. Biopharm. 69 (2008)

698-707.

939393

94949494


METOPROLOL TARTRATE RELEASE: INFLUENCE OF FORMULATION AND

PROCESS PARAMETERS

Parts of this chapter were published in:

Influence of formulation and process parameters on the release characteristics of ethylcellulose

sustained-release mini-matrices produced by hot-melt extrusion.

E. Verhoeven1, T.R.M. De Beer2, G. Van den Mooter3, J.P. Remon1, C. Vervaet1, Eur. J. Pharm.

Biopharm. 69 (2008) 312-319.

1 Laboratory of Pharmaceutical Technology, Ghent University, Ghent, Belgium. 2 Department of Pharmaceutical Analysis, Ghent University, Ghent, Belgium. 3 Laboratory of Pharmacotechnology and Biopharmacy, University of Leuven, Leuven, Belgium.

95

CHAPTER 2

ABSTRACT

Mini-matrices (multiple-unit dosage form) with release-sustaining properties were

developed by hot-melt extrusion (cylindrical die: 3 mm) using metoprolol tartrate as model

drug and ethylcellulose as sustained-release agent. A burst release of metoprolol tartrate

was seen from mini-matrices formulated with triethyl citrate or triacetin (hydrophilic

plasticizers), whereas drug release from matrices plasticized with dibutyl sebacate or diethyl

phthalate (hydrophobic) was slow and followed zero-order kinetics. Despite the different

hydrophilic character of the plasticizers, all had the same plasticizing effect during

processing. Dibutyl sebacate was selected as plasticizer for further development and its

concentration was optimized to 50% (w/w) of the ethylcellulose concentration. Xanthan gum,

a hydrophilic polymer, was added to the formulation to increase drug release. Changing the

xanthan gum concentration modified the in vitro drug release: increasing xanthan gum

concentrations (1, 2.5, 5, 10 and 20%, w/w) yielded a faster drug release. Zero-order drug

release was obtained at 5% (w/w) xanthan gum. The production of the extrudates was

reproducible, and after an equilibration time of 5 min extrudates with a constant quality were

produced. Incorporation of kneading paddles in the screw configuration allowed to

manufacture smooth extrudates when processing the formulation at 60°C. The mixing

efficacy and drug release were not affected by the number of mixing zones or their position

along the extruder barrel. Using a screw configuration only consisting of transport zones

resulted at 60°C in extrudates with an irregular surface structure and powder spots (identified

as metoprolol tartrate via Raman spectroscopy) were seen on the extrudates surface. Using

a higher processing temperature (70°C) improved the surface structure but metoprolol

tartrate spots were still visible on the surface. Raman analysis revealed that metoprolol

tartrate was homogeneously distributed in the mini-matrices, independent of screw design

and processing conditions. Simultaneously changing the powder feed rate (6-50 g/min) and

screw speed (30-200 rpm) did not alter extrudate quality or dissolution properties.

96

XANTHAN GUM TO TAILOR METOPROLOL TARTRATE RELEASE: INFLUENCE OF FORMULATION AND PROCESS PARAMETERS


METOPROLOL TARTRATE RELEASE: INFLUENCE OF FORMULATION AND

PROCESS PARAMETERS

1. INTRODUCTION

Hot-melt extrusion is a technology used in the pharmaceutical industry to produce

matrix formulations into which a drug is homogeneously embedded. Its major advantage over

conventional techniques (e.g. direct compression) for manufacturing sustained-release

matrices is the continuity of the production process as the different steps (mixing, melting,

homogenizing and shaping) are carried out on a single machine [1-3].

The excellent feasibility of ethylcellulose, a polymer with thermoplastic properties, for

hot-stage extrusion has been established in a variety of applications [4-7]. Previous work has

shown that hot-melt extrusion is an appropriate technique to develop mini-matrices using

ethylcellulose to sustain the release of ibuprofen: the combination of ethylcellulose and a

hydrophilic component (hydroxypropylmethylcellulose [8-10], xanthan gum [8-11]) offered a

flexible system to tailor the in vitro as well as in vivo drug release. This technique to

manufacture sustained-release mini-matrices using a combination of ethylcellulose, xanthan

gum and ibuprofen is discussed in detail in Chapter 1.

Due to the specific drug-matrix interaction, the low-melting ibuprofen (melting point

76°C) was identified as a plasticizer for ethylcellulose [12]. Consequently, the characteristics

of ethylcellulose/hydrophilic polymer mini-matrices containing ibuprofen are not predictive of

the extrusion and dissolution properties of ethylcellulose mini-matrices containing non-

plasticizing drugs. Therefore, ibuprofen was substituted by a drug with a higher melting point

(metoprolol tartrate, 123°C) and a conventional plasticizer was added to the formulation.

97

CHAPTER 2

In the present study sustained-release mini-matrices were developed by hot-melt

extrusion of a metoprolol tartrate/ethylcellulose mixture with the addition of xanthan gum to

tailor drug release. The aim was to examine the effect of different plasticizers on the

extrusion behaviour and extrudate quality. After optimization of the formulation (type and

concentration of plasticizer, xanthan gum concentration) to obtain zero-order drug release,

the effect of extrusion process parameters (screw design, powder feed rate and screw

speed) on the quality and drug release properties of the mini-matrices was evaluated.

98



2.1. Materials

The matrix consisted of ethylcellulose (EC) (Ethocel® Std 10 FP Premium; particle

size: 3-15 µm; ethoxyl content: 48-49.5% (w/w); viscosity: 9-11 mPa.s (5% solution in

toluene/EtOH 80/20, 25°C)), kindly donated by Dow Chemical Company (Midland, USA), and

a hydrophilic component: xanthan gum (XG) (Xantural® 75, mean particle size of 75 µm)

supplied by CP Kelco (Liverpool, UK).

2.1.1. Metoprolol tartrate

Metoprolol tartrate (MPT) (average diameter: 10 µm) was received from Esteve

Quimica (Barcelona, Spain) and selected as model drug (Figure 1).

Figure 1: Chemical structure of metoprolol tartrate.

Metoprolol tartrate is a β1-selective adrenergic blocking agent and widely used as a

drug of choice in the management of hypertension, angina pectoris, arrhythmias and heart

failure [13]. It is a drug with high solubility and permeability, is well-absorbed over a large part

of the gastro-intestinal tract and a well-defined relationship between the beta-blocking effect

and plasma drug concentration is known. Its relatively short half life (3-4 h) makes it a

suitable candidate for an extended-release formulation [14-19]. Metoprolol tartrate is a

racemic drug, and the S-enantiomer is pharmacologically active. The influence of

stereoselective pharmacokinetics and first pass metabolism on the development of an in

vitro/in vivo correlation for its enantiomers has been described [20, 21]. Metoprolol tartrate is

characterized by rapid systemic clearance (hepatic first pass metabolism) and thus

99

CHAPTER 2

necessitates repeated daily dosing. Therefore, the enhanced therapeutic efficacy of this drug

through the provision of constant drug input and maintenance of steady-state blood levels is

well-documented. Although it is challenging to deliver highly water soluble drugs via

sustained-release formulations – due to their susceptibility to dose-dumping – several

approaches have been used to formulate different types of controlled-release metoprolol

tartrate formulations (matrix tablets [22-26], multi-layer matrix tablets [27-29], pulsatile drug

delivery capsules [30], multiple-unit dosage forms [31-36], solid dispersions [37]) in order to

improve the clinical efficacy of metoprolol tartrate.

2.1.2. Plasticizers

Different plasticizers (dibutyl sebacate (DBS), diethyl phthalate (DEP), triethyl citrate

(TEC) and triacetin (TA) (Figure 2)) were tested as potential plasticizers for ethylcellulose,

and were obtained from Sigma-Aldrich (Steinheim, Germany).

Plasticizers are typically low molecular weight compounds capable of softening

polymers by weakening the intermolecular forces that hold the polymer chains together and,

in doing so, improve the movement of polymer chains with respect to each other. These

agents are able to decrease the glass transition temperature (Tg) and the melt viscosity of

the polymer by increasing the free volume between polymer chains. With the addition of a

plasticizer, a hot-melt extrusion process can be conducted at a lower temperature and with

less torque (shear forces), thereby improving the processing stability of both the polymer and

drug. Plasticizers may also improve the physico-mechanical properties of hot-melt extruded

dosage forms: there is a decrease in both the tensile strength and Young’s modulus, while

the polymer flexibility is increased, resulting in elongation of the final product when dealing

with extruded films [38-45]. Conventional plasticizers are used in a concentration range of 5-

10% by weight of the extrudable mass [46, 47], and the efficiency of a plasticizer is related to

its chemical structure and the interaction between its functional groups with those of the

polymer(s) [48]; because polyethylene glycol (PEG) and polyethylene oxide (PEO) are

composed of the same structural repeating unit, PEG and PEO low molecular weight can be

used as processing aid for PEO high molecular weight to reduce the melt viscosity, friction

and chain entanglements between the PEO molecules [49-51].

100


(a) (b)

(c) (d)

Figure 2: Chemical structure of (a) dibutyl sebacate, (b) dietyl phthalate, (c) triethyl citrate

and (d) triacetin.

Whereas the behaviour and properties of traditional plasticizers such as phthalates

have been well-documented, some drugs were found to have a plasticizing effect. It is

reported that ibuprofen acted as a plasticizer for ethylcellulose [12, 41, 52] and Eudragit® RS

PO [53], whereas the Tg of the latter could also be lowered by chlorpheniramine maleate and

indomethacin [44], and methylparaben [54]. Eudragit® E 100 was plasticized by lidocaine HCl

[39], whereas Eudragit® RS 30 D was plasticized by methylparaben, chlorpheniramine

maleate and ibuprofen [41]. Hydrocortisone and chlorpheniramine maleate softened the

matrix structure of hydroxypropylcellulose films [40], whereas vitamin E TPGS could be used

as processing aid when extruding hydroxylpropylcellulose/polyethylene oxide films by

decreasing the barrel pressure and torque of the extruder equipment [43].

The utility of various surfactants as plasticizers for hot-melt extrusion was also

investigated [55].


101

CHAPTER 2

2.2. Methods

2.2.1. Preparation of co-evaporates and hot-melt extruded samples

The glass transition temperature (Tg) of EC in EC/plasticizer co-evaporates (containing 0,

10, 20 and 30% (w/w) DBS, DEP, TEC or TA), in hot-melt extruded EC/DBS mixtures (ratio: 95/5,

80/20 and 66/33 (w/w)) and in hot-melt extruded MPT/EC/plasticizer (30/50/20, w/w/w) mixtures

was determined via differential scanning calorimetry (DSC).

2.2.1.1. Co-evaporates

Co-evaporates were prepared by dissolving the plasticizer and EC in ethanol. The

polymer solution (total solid content: 0.5 g in 150 ml) was transferred into a Teflon flask and a

rotavapor (Büchi Rotavapor R-200, Flawil, Switzerland) was used to remove the solvent under

reduced pressure at 70°C. The films were peeled from the perfluoroalkoxy surface of the flask and

stored in an oven at 70°C for 2 days to ensure complete removal of ethanol. Subsequently the

films were pulverized in a mortar using liquid nitrogen, further dried and stored at 40°C before

subjecting the samples to thermal analysis.

2.2.1.2. Hot-melt extruded samples

Hot-melt extruded EC/DBS samples were prepared at a powder feed rate of 6 g/min and a

screw speed of 30 rpm, using the standard screw configuration and different extrusion

temperatures (range: 50-200°C with 25°C intervals). Hot-melt extruded MPT/EC/plasticizer

mixtures were processed at 65°C.

Additionally, the influence of MPT on the Tg of EC is measured in an extruded sample

having an EC/MPT ratio of 33.3/30 (w/w) (i.e. the lowest EC/MPT ratio in the formulations tested

thus offering the highest probability for interaction).

The thermal profile (via modulated temperature differential scanning calorimetry) of MPT

before and after processing (extrusion of MPT powder as well as the production of mini-matrices,

both at a processing temperature of 65°C) was compared to obtain information about the thermal

stability during processing.

102


2.2.2. Optimization of plasticizer

To select a suitable plasticizer EC was combined with different plasticizers (DBS,

DEP, TEC, TA) in a ratio of 2.5/1 (w/w). The MPT content of these formulations was 30%

(w/w). The components were blended in a planetary mixer (15 min, 90 rpm) (Kenwood Major

Classic, Hampshire, UK) and incubated overnight at room temperature to achieve sufficient

interaction between EC and plasticizer. The mixture was passed through the screws of the

powder feeder of the extruder and recycled into the powder reservoir to grind the

MPT/EC/plasticizer mixture prior to hot-melt extrusion. Hot-melt extrusion was performed

using a laboratory-scale intermeshing co-rotating twin-screw extruder (MP19TC-25, APV

Baker, Newcastle-under-Lyme, UK) having a length-to-diameter ratio of 25/1. The machine

was equipped with a Brabender twin-screw powder feeder and an extrusion screw having

two mixing sections and a densification zone (this screw configuration is further referred to as

the standard screw profile). The die block (2.6 cm thickness) was fixed to the extruder barrel,

and additionally, an axially mounted die plate (1.9 cm thickness) was attached to the die

block, with a cylindrical hole of 3 mm diameter for shaping the extrudates. The following

extrusion conditions were used: a screw speed of 30 rpm, a powder feed rate of 6 g/min and

a temperature of 65°C for the five heating zones along the barrel. After cooling down to room

temperature, the extruded rods (∅ = 3 mm) were manually cut, using surgical blades, into

mini-matrices of 2 mm length. The influence of the plasticizer type on drug release was

evaluated via in vitro dissolution testing.

To optimize the extrudate quality and mini-matrix properties, formulations with variable

DBS concentrations were processed (EC/DBS ratio: 5/1, 3/1, 2/1 and 1.4/1, w/w). The MPT and

XG content were 30 and 10% (w/w), respectively. The mini-matrices were manufactured using the

same process as described above, except for the temperature: the initial extrusion temperature

was set at 80°C and was lowered in steps of 10°C until shark skinning of the extrudate occurred.

The surface properties of the extrudates were visually inspected for any defects and evaluated for

their suitability to be cut into mini-matrices (deformation due to cutting, smoothness of the cutting

surfaces and edges) using a digital camera (C3030 Olympus) linked to an image analysis system

(analySIS®, Soft Imaging system, Münster, Germany) (magnification 9.5x).

2.2.3. Influence of xanthan gum

To modify the drug release and obtain zero-order release kinetics from the EC mini-

matrices, XG was added to the formulation in concentrations of 0, 1, 2.5, 5, 10 and 20%

103

CHAPTER 2

(w/w). The MPT content was 30% (w/w), the EC/DBS ratio 2/1 (w/w). The formulations were

processed via hot-melt extrusion: screw speed of 30 rpm, powder feed rate of 6 g/min,

extrusion temperature of 60°C, using the standard screw profile described above. In vitro

dissolution testing and swelling properties of the mini-matrices were evaluated via

microscopic analysis (digital camera (C3030 Olympus) linked to an image analysis system

(analySIS®, Soft Imaging system, Münster, Germany) (magnification 9.5x)) after 4 h

immersion in the dissolution medium.

2.2.4. Influence of the extrusion process parameters

The influence of process parameters (screw design, powder feed rate and screw speed)

on the extrusion process, extrudate quality and drug release properties of an optimized

formulation (MPT/XG/EC/DBS ratio: 30/5/43.3/21.7, w/w/w/w) was evaluated.

The standard configuration (Figure 3a) of the screw (length: 47.5 cm, diameter: 1.9

cm) consisted of transport elements (forwarding elements who serve as drivers to transport

the material through the extruder), 2 kneading blocks and a discharge element (densification

zone). The conveying sections are built from transport elements having a double helix and a

pitch of 9.5 mm. The kneading blocks consist of disks (4.75 mm thickness, mixing elements)

which ensure blending of the powder mixture. The first kneading block (at 26.6 cm from the

discharge end) consists of 10 disks with increasing staggering angle (4 disks at 30°, 4 disks

at 60° and 2 disks at 90°). The second kneading block (at 15.2 cm from the discharge end)

has 6 disks with a staggering angle of 60°. The discharge element is placed at the end of the

screw and has a length of 2.8 cm and a pitch of 5.0 mm. Next to the standard screw design,

experiments were also performed with a screw having only one kneading block (the first or

second mixing zone, the kneading paddles of the other mixing zone were replaced by

transport elements) and without kneading block (the screw consisted of only transport

elements and the discharge zone) (Figure 3b). The extrusion was carried out at a powder

feed rate of 6 g/min, screw speed of 30 rpm and a processing temperature of 60°C.

The effect of production rate (depending on powder feed rate and screw speed) was

determined by manufacturing mini-matrices at the following conditions: standard screw

configuration, processing temperature of 60°C and variable powder feed rate and screw

speed: 6 g/min in combination with 30 rpm, 25 g/min with 100 rpm, and 50 g/min with 200

rpm.

104


(a)

(b)

Figure 3: Configuration of the intermeshing co-rotating screws. (a) Standard screw

configuration with 2 kneading blocks: (1) transport zone, (2) mixing zone, (3) transport zone,

(4) mixing zone, (5) transport zone, (6) densification zone. (b) Screw configuration without

kneading paddle elements: (7) transport zone, (8) densification zone.

Extrudate quality and mini-matrix properties in function of process parameters (screw

design and production rate) were visually evaluated (microscopic analysis, using a digital

camera (C3030 Olympus) linked to an image analysis system (analySIS®, Soft Imaging

system, Münster, Germany) (magnification 9.5x)) and via in vitro dissolution testing.

Using the standard extrusion settings (standard screw configuration, processing

temperature of 60°C, powder feed rate of 6 g/min and screw speed of 30 rpm), the

reproducibility of the extrusion process was evaluated by processing 3 batches on 3

consecutive days, and the equilibration time of the production process was determined by

collecting extrudate samples at several time points (5, 10, 15, 20 and 25 min after start-up).

In vitro dissolution testing was used as evaluation method to differentiate between samples

collected at different days and at different sampling points.

MPT distribution in the mini-matrices prepared via different extrusion processes

(using 0, 1 or 2 kneading blocks) was evaluated by Raman spectroscopic mapping. Three

1 2 3 4 5 6

7 8

105

CHAPTER 2

areas (2150 x 1150 µm2) (two at the edges and one in the middle of the mini-matrix) of each

mini-matrix (n = 3) were scanned using a 10x long working distance objective lens (laser spot

size: 50 µm) in point-by-point mapping mode with a step size of 100 µm in both the x and y

directions. The resulting map provides an overview of the MPT distribution in the mapped

area. The mapping system used for this study was a RamanRxn 1 Analyzer and Microprobe

(Kaiser Optical Systems, Ann Arbor, USA), equipped with an air cooled CCD detector (back-

illuminated deep depletion design). The laser wavelength during the experiments was the

785 nm line from a 785 nm Invictus NIR diode laser. All spectra were recorded at a resolution

of 4 cm-1 using a laser power of 400 mW and a laser light exposure time of 20 s per collected

spectrum. Before data analysis, spectra were baseline corrected. Data collection and data

analysis was done using the HoloGRAMSTM data collection software package, the

HoloMAPTM data analysis software and Matlab® software package (version 6.5.).

2.2.5. Thermal analysis

The thermal behaviour of powders, co-evaporates and hot-melt extruded samples

was evaluated using a 2920 Modulated DSC (TA Instruments, Leatherhead, UK) equipped

with a refrigerated cooling system (RCS). Dry helium at a flow rate of 40 ml/min was used as

the purge gas through the DSC cell and 150 ml/min of nitrogen through the RCS unit.

Samples (± 10 mg) were run in closed aluminium pans supplied by TA Instruments; the mass

of each empty sample pan was matched with the mass of the empty reference pan to ± 0.10

mg. The experimental method consisted of an initial 5 min isothermal equilibration period at

0°C. During the subsequent heating run the following experimental parameters were used:

an underlying heating rate of 2°C/min from 0 to 200°C, a modulation amplitude of 0.212°C

and a period of 40 s. Temperature and enthalpic calibration was performed with an indium

standard, whereas calibration of the heat capacity was performed with a sapphire standard.

The results were analyzed using the TA Instruments Universal Analysis Software.

Measurements were performed in duplicate and the Tg values (midpoint half height) are

reported.

2.2.6. In vitro drug release

The mini-matrices (approximately 60 mg) were introduced in a basket (USP 27,

dissolution apparatus 1 [56]). The dissolution was performed in a VK 7010 dissolution system

combined with a VK 8000 automatic sampling station (VanKel, New Jersey, USA). Demineralized

106


water was used as the dissolution medium. The temperature of the medium (900 ml) was kept at

37 ± 0.5°C, while the rotational speed of the baskets was set at 100 rpm. Samples of 5 ml were

withdrawn at 0.5, 1, 2, 4, 6, 8, 12, 16, 20 and 24 h and spectrophotometrically analyzed for MPT at

222 nm by means of a Perkin-Elmer Lambda 12 UV-VIS double beam spectrophotometer

(Zaventem, Belgium).

Due to interference of DEP in the absorption spectrum of MPT, dissolution samples of

mini-matrices containing DEP were analyzed using a validated HPLC-UV method. All chemicals

were of analytical or HPLC grade. The HPLC equipment (Merck-Hitachi, Darmstadt, Germany)

consisted of a solvent pump (L-7100) set at a constant flow rate of 1.0 ml/min, a variable

wavelength UV-detector (L-7400) set at 222 nm, a reversed-phase column and precolumn

(LiChroCART® 250-4 and 4-4, LiChrospher® 100 RP-18 5 µm), an auto-sampler injection system

(L-7200) with a 100 µl loop (Valco Instruments Corporation, Houston, Texas, USA) equipped with

an automatic integration system (software D-7000 Multi-Manager). The mobile phase consisted of

0.01 M KH2PO4/acetonitrile/methanol (55/22.5/22.5, v/v/v).

For both spectrophotometric methods, the MPT concentrations were calculated from a

calibration curve between 0 and 50 µg/ml. The dissolution was simultaneously performed in 6

dissolution vessels, each vessel containing 4 mini-matrices.


For the general considerations about method validation of an analytical procedure, we

refer to Chapter 1 and the guidelines of the International Conference on Harmonisation (ICH) [57].

The following parameters were evaluated: specificity, linearity, precision, accuracy and the limit of

detection as well as of quantification.


Specificity of the HPLC method was assessed at 222 nm by comparing the

chromatograms of dissolution medium spiked with MPT, dissolution medium spiked with DEP

and a dissolution sample (containing MPT and DEP).

107

CHAPTER 2


Different calibration curves (n = 10) were prepared. For each calibration curve,

different metoprolol tartrate concentrations were investigated (2.5, 5, 10, 20, 30, 40 and 50

µg/ml in water) to cover the range of the unknown samples.





precision (between-day repeatability). The precision was calculated for different metoprolol

tartrate concentrations (2.5, 5, 10, 20, 30, 40 and 50 µg/ml).

2.2.6.1.4. Accuracy

Accuracy was determined for different metoprolol tartrate concentrations (2.5, 5, 10,

20, 30, 40 and 50 µg/ml) (n = 10).


The limit of detection (LOD) and limit of quantification (LOQ) were calculated based on the

standard deviation of the intercept and slope of the mean calibration curve.

2.2.7. X-ray diffraction

To investigate the crystallinity of the components in the mini-matrices, X-ray

diffraction was performed (on pulverized mini-matrices). The X-ray patterns of MPT, EC, XG

and mini-matrices were determined using a D5000 Cu Kα Diffractor (λ = 0.154 nm)

(Siemens, Karlsruhe, Germany) with a voltage of 40 mA in the angular range of 10° < 2θ <

60° using a step scan mode (step width = 0.02°, counting time = 1 s/step).

108



3.1. Optimization of plasticizer

Previous work [12] identified ibuprofen (melting point 76°C) as a plasticizer for EC,

and therefore the characteristics of ibuprofen/EC mini-matrices are not predictive of the

extrusion and dissolution properties of EC mini-matrices containing non-plasticizing drugs.

The aim of this work was to substitute ibuprofen by MPT: due to its higher melting point

(123°C), a specific drug-matrix interaction (due to melting of the drug) is unlikely in case

extrusion is performed at a temperature below the drug melting point. Therefore, the effect of

MPT on the polymer Tg was examined. Analyzing the thermal behaviour of an EC/MPT

(33.3/30, w/w) hot-melt extruded sample (extrusion temperature: 65°C) only revealed the

melting endotherm of MPT (Tm 122.6 ± 0.4°C), whereas no Tg of EC was detected. Since the

Tg of EC powder was determined 127.9 ± 0.2°C, one can assume that the Tg of EC in the

melt-extruded samples is masked by the melting endotherm of MPT (peak ranging from

108.5 to 126.0°C). Although a slight decrease in Tg of EC could be overwhelmed by the

melting signal of MPT, any interaction occurring during extrusion between MPT and EC is

limited (since no Tg of EC was detected at a temperature below the melting endotherm of

MPT) and the incorporation of a conventional plasticizer is required to lower the extrusion

temperature in order to reduce thermal degradation and improve extrudate quality.

The Tm of MPT prior to extrusion and when processed at 65°C (as powder as well as

in the MPT/EC/plasticizer mixtures) was similar, proving its thermal stability during

production. Previous work at the Laboratory of Pharmaceutical Technology confirmed the

chemical stability of ethylcellulose and xanthan gum after hot-melt extrusion at this

processing temperature using similar shearing forces [58]. The thermochemical stability and

volatility of the plasticizer during processing must also be taken into consideration. Since

other researchers found that plasticizer concentration only decreased when processed in

films at 180°C [59], the amount of plasticizer in our formulations will remain constant since

extrusion was performed at 60°C.

109

CHAPTER 2

3.1.1. Plasticizer type

3.1.1.1. Co-evaporates

The effect of plasticizer type and concentration is shown in Figure 4 (n = 3). Tg of EC

decreased at higher plasticizer concentrations, although an increase from 20 to 30% (w/w)

did not induce a further drop of Tg. Despite the different polarity of the plasticizers

(hydrophilic: TEC, TA / lipophilic: DBS, DEP) the efficiency of all plasticizers was similar.

0

20

40

60

80

100

120

140

0 10 20 3

Concentration plasticizer (%)

Tg E

C (°

C)

0

Figure 4: Influence of plasticizer concentration and type on the Tg of EC (mean ± S.D., n = 3)

processed as co-evaporates: (□) DBS, (△) DEP, (○) TEC, (x) TA.


Using co-evaporates it was not possible to select a specific plasticizer. However, co-

evaporates could overestimate the plasticizing efficacy since ideal solid dispersions were

manufactured, maximizing plasticizer/polymer interactions. Therefore, hot-melt extruded

samples (MPT/EC/plasticizer in a ratio of 30/50/20 (w/w/w)) were processed to select the

optimal plasticizer. Neither the extrusion process nor the extrudate quality was influenced by

110


the plasticizer type since the pressure (4, 3, 8 and 3 bar for DBS, DEP, TEC and TA

formulations, respectively) and torque (37, 37, 37 and 34% for DBS, DEP, TEC and TA

formulations, respectively) were similar when extruding with the different plasticizers

(extrusion temperature: 65°C). For all hot-melt extruded samples, the Tg of EC was detected

in the temperature range of 55-60°C. Comparing these results with the Tg shifts of the co-

evaporates showed that the plasticizing efficiency of the 4 plasticizer types was similar in

both methods, indicating that intense intermolecular mixing also occurred during the hot-melt

extrusion process, ensuring complete miscibility of plasticizers and EC. In contrast, Aitken-

Nichol and co-workers [39] suggested that intermolecular mixing was better in a solution-cast

film compared to a highly viscous melt: lidocaine HCl being a more effective plasticizer in

solution cast films than in the extruded films.



The chromatograms (Figure 5) showed that MPT and DEP were completely

separated, and therefore the method was selective to determine MPT in dissolution samples:

a retention time of ± 7 min and ± 15 min was obtained for MPT and DEP, respectively.

111

CHAPTER 2

Figure 5: Chromatograms showing metoprolol tartrate and diethyl phthalate in (top) water

spiked with MPT, (middle) water spiked with DEP and (bottom) dissolution sample.

Absorption

Time (min)

112



The mean calibration curve (Y = 82741 (± 1564) X - 30938 (± 8556)) was linear

between 0 and 50 µg/ml metoprolol tartrate, with a coefficient of determination of 0.9999 (±

0.0001).





precision (n = 5).



2.5 1.9 7.9

5.0 0.7 3.4

10.0 0.7 2.7

20.0 0.8 2.0

30.0 0.4 2.9

40.0 0.8 1.4

50.0 0.6 1.7



3.1.1.3.4. Accuracy

The accuracy was determined for all metoprolol tartrate concentrations and the mean

results are presented in Table 2.

113

CHAPTER 2



2.5 2.84 ± 0.40 113.5 ± 16.0 14.1

5.0 5.63 ± 0.63 112.6 ± 12.5 11.1

10.0 9.59 ± 0.14 95.9 ± 1.4 1.4

20.0 18.98 ± 0.13 94.9 ± 0.6 0.7

30.0 31.30 ± 1.22 104.3 ± 4.1 3.9

40.0 40.10 ± 0.29 100.2 ± 0.7 0.7

50.0 48.29 ± 0.69 96.6 ± 1.4 1.4




Based on the mean linear regression equation of metoprolol tartrate the limit of

detection and the limit of quantification for metoprolol tartrate are 0.34 and 1.03 µg/ml,

respectively.

3.1.1.4. Influence of plasticizer type on drug release

Drug release in function of plasticizer type is shown in Figure 6: the release rate was

affected, indicating that the chemical nature of the plasticizer plays a major role for the

underlying drug release mechanism. DBS and DEP provided a linear release of MPT,

however, not all drug was released within 24 h. On the other hand, complete release was

noticed for TEC- and TA-containing mini-matrices, but the burst effect was significant

probably due to the hydrophilicity of these plasticizers. Other investigators [60-63] already

described a similar effect of water solubility of the plasticizer on the drug release mechanism.

Lecomte and co-workers [64] compared a lipophilic (dibutyl sebacate) and hydrophilic

(triethyl citrate) plasticizer and reported that triethyl citrate rapidly leached from a polymer

coat affecting film permeability and drug diffusion. In contrast dibutyl sebacate remained in

114


the polymer layer, reducing the drug release rate. Furthermore, the plasticizer concentration

can also affect the drug release profile, as reported by other investigators [58, 63, 65, 66].

Due to the burst effect of the hydrophilic plasticizers, DBS was selected for MPT/EC

formulations, since this combination seemed able to provide zero-order drug release from the

mini-matrices.

0

20

40

60

80

100

120

0 4 8 12 16 20

Time (h)

MPT

rele

ased

(%

24

)

Figure 6: Influence of plasticizer type on the dissolution profiles (mean ± S.D., n = 6) of mini-

matrices containing 30% (w/w) MPT and EC/plasticizer (2.5/1, w/w): (■) DBS, (□) DEP, (▲)

TEC, (△) TA.

3.1.2. Plasticizer concentration

Depending on the DBS concentration in the formulation and the processing

temperature, the pressure and torque during production varied (Table 3): both parameters

increased with decreasing processing temperatures (irrespective of the DBS content) due to

the higher viscosity of the material at these production conditions. Processing was also

easier at higher DBS levels as Tg of EC was lowered to about 100.1, 43.8 and 38.8°C after

extrusion (at 125°C) of binary EC/DBS mixtures containing 5, 20 and 33% (w/w) DBS,

115

CHAPTER 2

respectively (Table 4). The minimum extrusion temperature allowing processing also

confirmed the efficiency of DBS during extrusion of EC: a 66/33 (w/w) and 80/20 (w/w)

EC/DBS mixture could be processed at 50 and 75°C, respectively, whereas a 95/5 (w/w)

mixture already blocked the extruder at an extrusion temperature of 100°C.

Table 3: Pressure (bar) and torque (%) during production of mini-matrices using different

EC/DBS ratios (w/w) at different processing temperatures. The MPT and XG content were 30

and 10% (w/w), respectively.

EC/DBS ratio

(w/w)

Processing

temperature (°C)

Pressure

(bar)

Torque

(%)

Extrudate

Mini-matrices

5/1 80 0 8 sticky -

70 4 12 smooth cracks

60 66 93 shark skinning twisted -

3/1 80 0 16 sticky -



50 26 43 shark skinning -

2/1 80 0 18 sticky -

70 1 26 smooth smooth


50 10 45 rough -

40 39 78 shark skinning -

1.4/1 80 0 19 sticky -




40 9 59 rough -

-: not applicable

116


For all formulations, the extrudates were sticky and had droplets (mainly DBS) on

their surface when extruded at 80°C (Table 3). At higher DBS concentration it was possible

to process the extrudates at lower temperature before surface defects were observed: at

40°C the extrudate with an EC/DBS ratio of 2/1 (w/w) showed shark skinning (Figure 7a),

whereas at the same processing temperature the extrudate showed only a rough surface

(Figure 7b) when an EC/DBS ratio of 1.4/1 (w/w) was used. Since mini-matrices with cracks

(Figure 7c) are excluded from further optimization and evaluation (drug release from these

mini-matrices would be difficult to control), an EC/DBS ratio of at least 2/1 (w/w) is required

to produce mini-matrices with good quality (without cracks) (Figure 7d). As it is preferential to

use the lowest plasticizer concentration possible, an EC/DBS ratio of 2/1 (w/w) is selected for

further experiments.

(a) (b)

(c) (d)

Figure 7: Extrudates and mini-matrices in function of EC/DBS ratio and processing

temperature: (a) MPT/XG/EC/DBS (30/10/40/20, w/w/w/w) produced at 40°C showing shark

skinning, (b) MPT/XG/EC/DBS (30/10/35/25, w/w/w/w) produced at 40°C showing rough

surface, (c) MPT/XG/EC/DBS (30/10/50/10, w/w/w/w) produced at 70°C showing mini-

matrices with cracks, (d) MPT/XG/EC/DBS (30/10/40/20, w/w/w/w) produced at 60°C.

117

CHAPTER 2

Shark skinning can be described as a regular ridged surface distortion, with the ridges

running perpendicular to the extrusion direction. Shark skinning is thought to be formed in the

die channel or at the exit. The melt, as it proceeds along the die channel, has a velocity

profile, with maximum velocity at the centre and zero velocity at the wall. The material at the

wall has to accelerate to the velocity at which the extrudate is leaving the die, generating

tensile stress. If the stress exceeds the tensile strength, the surface ruptures causing the

surface roughness. The shark skinning problem can generally be reduced by increasing the

die temperature, reducing the screw speed or the addition of a lubricant [67, 68].

MPT/XG/EC/DBS (30/10/50/10, w/w/w/w) extrudates processed at 60°C also showed

a twisting distortion: a polymer can remember its history of turning along the spiral screw,

even after it has been passed through the die. This phenomenon is called turning memory.

During passage down the screw, the melt receives a prolonged mechanical treatment during

which alignment of the polymer chains can occur. Since passage through the die is relatively

short, there is no time to replace the spiralled configuration by a new one. The turning

memory of the melt can be removed by installing a breaker plate-screen pack assembly in

the die zone [67].

As can be seen from Table 4, the Tg of EC decreased with increasing DBS content:

110, 80 and 45°C for an EC/DBS ratio of 95/5, 80/20 and 66/33 (w/w), respectively (before

hot-melt extrusion). After processing the Tg was decreased, irrespective of the processing

temperature: about 100, 50 and 40°C for an EC/DBS ratio of 95/5, 80/20 and 66/33 (w/w),

respectively. Due to the mixing efficiency of the kneading elements of the extrusion screw the

better plasticizer-polymer interaction during hot-melt extrusion explains the lower Tg after

hot-melt extrusion.

118


Table 4: Glass transition temperature (Tg, °C) of EC for different EC/DBS ratios (w/w) at

different processing temperatures (mean ± S.D., n = 3).

EC/DBS ratio (w/w) Processing temperature (°C) Tg (°C)

95/5 before extrusion 111.3

200 101.5

175 107.5

150 104.2

125 100.1


200 46.6

175 43.0

150 45.5

125 43.8

100 48.5

75 54.7


200 38.9

175 38.6

150 42.8

125 38.8

100 41.4

75 38.5

50 37.2

119

CHAPTER 2

3.2. Influence of xanthan gum

The drug release rate from hot-melt extruded dosage forms is highly dependent upon

the characteristics of the carrier material. Most of the materials used in hot-melt extruded

dosage forms are water insoluble [5, 6, 69, 70] or have slow hydration or gelation rates [40,

49]. To improve or modulate drug release from these systems, various excipients with

different physicochemical properties can be added. In this study XG was incorporated in the

formulation to increase the drug release rate by increasing the hydrophilicity and porosity of

the dosage form during dissolution [11]: drug release from MPT/EC/DBS (30/50/20, w/w/w)

mini-matrices was limited to 45% in 24 h.

The different formulations were easily processed, independent of XG concentration:

the pressure and torque varied between 1 and 5 bar, and 30 and 35%, respectively. All

formulations were produced at 60°C except for the 20% (w/w) XG formulation when a barrel

temperature of 65°C was required to produce smooth extrudates. Increasing XG

concentrations yielded a faster drug release and a zero-order drug release was only obtained

at ≤ 5% (w/w) XG (Figure 8).

0

20

40

60

80

100

120

0 4 8 12 16 20 2

Time (h)

MP

T re

leas

ed (%

)

4

Figure 8: Influence of xanthan gum concentration on the dissolution profiles (mean ± S.D., n

= 6) of mini-matrices containing 30% (w/w) MPT, XG, EC and DBS (EC/DBS 2/1, w/w). XG

concentration (w/w): (■) 0%, (▲) 1%, (●) 2.5%, (□) 5%, (△) 10%, (○) 20%.

120


The 5% XG formulation was selected as reference, since incomplete drug release

and burst effect were seen at lower and higher XG concentrations, respectively. Similar to

EC/XG matrices containing ibuprofen [11], swelling of the mini-matrices containing 10 and

20% (w/w) XG was responsible for their faster MPT release (Figure 9).

(a) (b)

(c) (d)

Figure 9: Mini-matrices containing 30% (w/w) MPT, XG, EC and DBS (EC/DBS 2/1, w/w) after 4 h

immersion in the dissolution medium. XG concentration (w/w): (a) 0%, (b) 5%, (c) 10%, (d) 20%.

121

CHAPTER 2

Figure 10 shows the X-ray diffraction spectra of MPT, EC, XG and mini-tablets containing

5, 10 and 20% XG. As can be seen from the figure, the drug was crystalline, while EC and XG

were amorphous. The drug remained crystalline in the hot-melt extruded mini-tablets as sharp

peaks of MPT could be observed in the diffraction pattern of the mixtures. Furthermore, no

difference in drug crystallinity could be seen in mini-matrices with different XG content, since the

diffraction spectra of the different hot-melt extruded mixtures (varying in XG concentration) were

overlapping. However, the diffraction peaks were smaller in comparison with the pure material,

illustrating the presence of smaller crystals.

0

500

1000

1500

2000

10 20 30 40 50 60

2-teta (°)

Lin

(Cou

nts)

Figure 10: X-ray diffraction pattern of (red) XG, (purple) MPT, (green) EC, (blue) 5% XG

mini-matrices, (pink) 10% XG mini-matrices, (yellow) 20% XG mini-matrices.

3.3. Reproducibility and equilibration of the production process

The extrusion process did not vary between the production of 3 batches on

consecutive days: the pressure and torque varied between 3 and 5 bar, and 30 and 32%,

respectively. The production of the 5% (w/w) XG extrudates was reproducible since the drug

release profiles coincided, revealing the robustness of the production method. After an

equilibration time of 5 minutes extrudates of constant quality were obtained since no

influence on the extrudate quality and release characteristics was seen for different extrusion

times.

122


3.4. Influence of the extrusion process parameters

Modifications of the screw configuration allow to alter the production method as the

different screw elements (feed, mixing, discharge) can be optimized to suit a particular application

[3, 71]. Literature reports described a significant influence of the screw design on the efficiency of

a hot-melt extrusion process [4], extrudate quality and the physico-chemical properties

(crystallinity and dissolution properties) of extruded material [72, 73].

In our study, the pressure and torque were neither influenced by the number nor the

position of the mixing zones: 3, 7 and 3 bar, and 30, 36 and 33% when using both mixing zones,

only the first mixing zone and only the second mixing zone, respectively.

Using a screw design with two mixing sections, smooth extrudates were obtained, yielding

mini-matrices of good quality (Figure 11a). In case of extrusion (at 60°C) without kneading

paddles (pressure: 2 bar, torque: 14%) the extrudates had a rough surface with powder spots

(which Raman identified as MPT) (Figure 11b) confirming that kneading paddle elements are

required for intensive mixing which was also reported by others [72, 73]. Using this screw

configuration, smooth extrudates could be obtained when the barrel temperature was increased

with 10°C (pressure: 0 bar, torque: 10%). However, powder spots were still detected (Figure 11c)

since the kneading elements have an effect on the residence time and mixing intensity in the

extruder and facilitate dispersion of the drug in the polymer. Mini-matrices from extruded rods

produced with only transport elements did not have good quality (cracks) (Figure 11b and 11c)

and were therefore not submitted to dissolution testing.

(a) (b) (c)

Figure 11: Extrudates and mini-matrices consisting of MPT/XG/EC/DBS (30/5/43.3/21.7, w/w/w/w)

produced at (a) 60°C with 2 kneading blocks, (b) 60°C without kneading blocks, (c) 70°C without

kneading blocks.

123

CHAPTER 2

Although it was reported that a screw configuration with less kneading paddles induced

less shear and decreased the residence time of material in the extruder [72-74], in our study the

mixing efficacy was not influenced if the number of mixing zones was reduced, provided that at

least one mixing zone was present. Drug release was also not influenced by the number or the

position of the mixing zones (Figure 12), what was also noticed by other investigators [75].

0

20

40

60

80

100

120

0 4 8 12 16 20

Time (h)

MPT

rele

ased

(%

24

)

Figure 12: Influence of screw design on the drug release profiles (mean ± S.D., n = 6) of

mini-matrices containing MPT/XG/EC/DBS (30/5/43.3/21.7, w/w/w/w): (■) kneading block 1

and 2, (▲) kneading block 1, (●) kneading block 2.

124


The distribution of MPT in the mini-matrices was monitored via Raman using the 627-653

cm-1 spectral band as no overlap with other ingredients occurred at these wavenumbers (Figure

13).

c:\hologram\data\thomas\microprobe\ellen verhoeven\stefanie\zuivere stoffen\000001 metapold tartrate.spc

600 650 700 750 800 850 900

0.5

1

1.5

2

2.5

3

3.5

x 104

Raman Shift (1/cm)

Cou

nts

600 650 700 750 800 850 900

Raman shift (cm-1)

Cou

nts

(AU

)

627- 653 cm-1

DBS

EC

XG

MPT

c:\hologram\data\thomas\microprobe\ellen verhoeven\stefanie\zuivere stoffen\000001 metapold tartrate.spc

600 650 700 750 800 850 900

0.5

1

1.5

2

2.5

3

3.5

x 104

Raman Shift (1/cm)

Cou

nts

600 650 700 750 800 850 900

Raman shift (cm-1)

Cou

nts

(AU

)

627- 653 cm-1

DBS

EC

XG

MPT

Figure 13: Raman spectra from the compounds in the mini-matrices.

Figure 14a is a picture from a mini-matrix tablet produced via extrusion at 60°C with 2

kneading blocks. The blue dots in the picture indicate the spots on the mini-matrix surface where

Raman spectra were collected. Based on the intensity of the Raman signal across the scanned

sections of the mini-matrices it was confirmed that MPT is homogeneously distributed in the mini-

matrix, since the same MPT intensities were measured in the centre and at the edges of the mini-

matrices (Figure 14b, red indicating an area with high MPT concentration at the tablet surface,

blue an area without MPT).

(a) (b)

Figure 14: (a) Mapped area of mini-matrix (border) – extruded at 60°C with 2 kneading blocks, (b)

interpolated map from a mapped mini-matrix (border) – extruded at 60°C with 2 kneading blocks,

representing uniform MPT distribution.

125

CHAPTER 2

ata\thomas\microprobe\ellen verhoeven\dringende stalen\60 graden\spots op extrudaat\0

0 500 1000 1500 2000 2500 30000

1

2

3

4

5

6

x 104

Raman Shift (1/cm)

Cou

nts

pure MPT

a b

c d Raman shift (cm-1)

Cou

nts

(AU

)


0 500 1000 1500 2000 2500 30000

1

2

3

4

5

6

x 104

Raman Shift (1/cm)

Cou

nts

pure MPT


0 500 1000 1500 2000 2500 30000

1

2

3

4

5

6

x 104

Raman Shift (1/cm)

Cou

nts

pure MPT

a b

c d Raman shift (cm-1)

Cou

nts

(AU

)

Similar results were obtained for all mapped areas of the analyzed mini-matrices,

irrespective of the production process, provided that kneading paddles were present in the screw.

In case of tablets prepared at 70°C without kneading blocks MPT distribution on the tablet surface

was not homogeneous (Figure 15b). This area of high MPT intensity in the Raman map

corresponded to a white particle that visually could be seen (Figure 15a and 15c at a

magnification of 10 and 100, respectively). The Raman spectrum (Figure 15d) identified this

particle as an API cluster as only the spectrum of MPT was detected.

Figure 15: Mini-matrices extruded at 70°C without kneading elements have local high MPT

amounts: (a) mapped area of mini-matrix (border), (b) interpolated map from a mapped mini-

matrix (border) representing inhomogeneous drug distribution, (c) white particle on the tablet

surface, (d) Raman spectrum of white particle on the tablet surface.

To evaluate the effect of production rate on the extrusion process and extrudate quality,

the powder feed rate and screw speed were simultaneously adjusted to ensure a constant fill level

of the extrusion chamber. By increasing both process parameters, the residence time in the

extruder was influenced [76]. Smooth extrudates were obtained at all processing conditions. A

pressure of 3, 2 and 2 bar, and a torque of 30, 46 and 42% were recorded at a powder feed rate

and screw speed of 6 g/min and 30 rpm, 25 g/min and 100 rpm, 50 g/min and 200 rpm,

respectively. Using a high powder feed rate (25 and 50 g/min) in combination with a high screw

speed (100 and 200 rpm), a lower barrel temperature was set (50°C) to compensate for the heat

126


generated due to the friction [4, 71]. The extrudate quality was independent of the powder feed

rate and screw speed, whereas Henrist et al. observed an increase in radius and porosity and a

decrease in mechanical strength of the extrudate when processed at a higher screw speed and

powder feed rate [77]. The production rate did not alter the drug release characteristics of the

mini-matrices (Figure 16), although other investigators reported a lower drug release when

increasing the powder feed rate and screw speed [77]. Nakamichi and co-workers reported that

the extrudate quality and properties were not influenced by the screw speed, in case kneading

elements were used [72].

0

20

40

60

80

100

120

0 4 8 12 16 20 2

Time (h)

MPT

rele

ased

(%

4

)

Figure 16: Influence of powder feed rate and screw speed on the drug release profiles (mean ±

S.D., n = 6) of mini-matrices containing MPT/XG/EC/DBS (30/5/43.3/21.7, w/w/w/w): (■) 6 g/min -

30 rpm, (▲) 25 g/min - 100 rpm, (●) 50 g/min - 200 rpm.

127

CHAPTER 2

4. CONCLUSIONS

This chapter showed that hot-melt extrusion can be used to produce

ethylcellulose/xanthan gum mini-matrices with controlled drug delivery. Dibutyl sebacate was

found to be a suitable plasticizer to obtain smooth extrudates and mini-matrices. Using

xanthan gum, zero-order release kinetics with almost complete drug release could be

obtained. Modifications of process parameters (screw design, powder feed rate and screw

speed) had no influence on the extrusion process or on the extrudate quality, drug

homogeneity and release characteristics, confirming the robustness of the process.

However, a drawback of this robustness is the fact that it offers limited flexibility for

optimization of drug release.

Therefore, further experiments (discussed in the next chapter) were performed using

polyethylene oxides as an alternative for xanthan gum.

128


REFERENCES




Gemischen zur Herstellung von Arzneiformen, Spektrum der Wissenschaft, July, 18-

20, 1995.


Pharm. Biopharm. 54 (2002) 107-117.

[4] J.W. McGinity, J.J. Koleng, M.A. Repka, F. Zhang, Hot-melt extrusion technology, in: J.

Swarbrick and J.C. Boylan (ed.), Encyclopedia of pharmaceutical technology, Vol. 19,











Pharm. 269 (2004) 509-522.



235-247.



Pharmacol. 57 (2004) 366-367.

129

CHAPTER 2



[11] E. Verhoeven, C. Vervaet, J.P. Remon, Xanthan gum to tailor drug release of

sustained-release ethylcellulose mini-matrices prepared via hot-melt extrusion: in vitro

and in vivo evaluation, Eur. J. Pharm. Biopharm. 63 (2006) 320-330.



1678-1685.

[13] B.B. Hoffman, Catecholamines, sympathomimetic drugs, and adrenergic receptor

antagonists, in: Goodman and Gilman’s - The pharmacological basis of therapeutics

(J.G. Hardman, L.E. Limbird, A.G. Gilman, ed.), McGraw-Hill, New York, USA (2001)

255-256.

[14] C.G. Regardh, K.O. Borg, R. Johansson, G. Johnsson, L. Palmer, Pharmacokinetic

studies on the selective β1-receptor antagonist metoprolol in man, J. Pharmacokinet.

Biopharm. 2 (1974) 347-364.

[15] G. Jobin, A. Cortot, J. Godbillon, M. Duval, J.P. Schoeller, J. Hirtz, J.J. Bernier,

Investigation of drug absorption from the gastrointestinal tract of man. I. Metoprolol in

the stomach, duodenum and jejunum, Brit. J. Clin. Pharmacol. 19 (1985) 97S-105S.

[16] N. Vidon, D. Evard, J. Godbillon, M. Rongier, M. Duval, J.P. Schoeller, J.J. Bernier, J.

Hirtz, Investigation of drug absorption from the gastrointestinal tract of man. II.

Metoprolol in the jejunum and ileum, Brit. J. Clin. Pharmacol. 19 (1985) 107S-112S.

[17] J. Godbillon, D. Evard, N. Vidon, M. Duval, J.P. Schoeller, J.J. Bernier, J. Hirtz,

Investigation of drug absorption from the gastrointestinal tract of man. III. Metoprolol in

the colon, Brit. J. Clin. Pharmacol. 19 (1985) 113S-118S.

[18] B. Mistry, J. Leslie, N.E. Eddington, A sensitive assay of metoprolol and its major

metabolite α-hydroxy metoprolol in human plasma and determination of

dextromethorphan and its metabolite dextrorphan in urine with high performance liquid

chromatography and fluorometric detection, J. Pharm. Biomed. Anal. 16 (1998) 1041-

1049.

130


[19] M.J. Kendall, S.R. Maxwell, S.A. Westergren, Controlled release metoprolol. Clinical

pharmacokinetic and therapeutic implications, Clin. Pharmacokinet. 5 (1991) 319-330.

[20] N. Sirisuth, N.D. Eddington, Influence of stereoselective pharmacokinetics in the

development and predictability of an IVIVC for the enantiomers of metoprolol tartrate,

Pharm. Res. 17 (2000) 1019-1025.

[21] N. Sirisuth, N.D. Eddington, The influence of first pass metabolism on the development

and validation of an IVIVC for metoprolol extended release tablets, Eur. J. Pharm.

Biopharm. 53 (2002) 301-309.

[22] N.D. Eddington, P. Marroum, R. Uppoor, A. Hussain, L. Augsburger, Development and

internal validation of an in vitro – in vivo correlation for a hydrophilic metoprolol tartrate

extended release tablet formulation, Pharm. Res. 15 (1998) 466-473.

[23] R.V. Nellore, G.S. Rekhi, A.S. Hussain, L.G. Tillman, L.L. Augsburger, Development of

metoprolol tartrate extended-release matrix tablet formulations for regulatory policy

consideration, J. Control. Release 50 (1998) 247-256.

[24] G.S. Rekhi, R.V. Nellore, A.S. Hussain, L.G. Tillman, H.J. Malinowski, L.L. Augsburger,

Identification of critical formulation and processing variables for metoprolol tartrate

extended-release (ER) matrix tablets, J. Control. Release 59 (1999) 327-342.

[25] V. Pillay, R. Fassihi, A novel approach for constant rate delivery of highly soluble

bioactives from a simple monolithic system, J. Control. Release 67 (2000) 67-78.

[26] J. Wikstrand, B. Andersson, M.J. Kendall, H. Stanbrook, M. Klibaner, Pharmacokinetic

considerations of formulation: extended-release metoprolol succinate in the treatment

of heart failure, J. Cardiovasc. Pharmacol. 41 (2003) 151-157.

[27] Y.S.R. Krishnaiah, R.S. Karthikeyan, V. Satyanarayana, A three-layer guar gum matrix

tablet for oral controlled delivery of highly soluble metoprolol tartrate, Int. J. Pharm. 241

(2002) 353-366.

[28] S.M. Al-Saidan, Y.S.R. Krishnaiah, V. Satyanarayana, P. Bhaskar, R.S. Karthikeyan,

Pharmacokinetic evaluation of guar gum-based three-layer matrix tablets for oral

131

CHAPTER 2

controlled delivery of highly soluble metoprolol tartrate as a model drug, Eur. J. Pharm.

Biopharm. 58 (2004) 697-703.

[29] C. Narendra, M.S. Srinath, B.P. Rao, Development of three layered buccal compact

containing metoprolol tartrate by statistical optimization technique, Int. J. Pharm. 304

(2005) 102-114.

[30] R. Löbenberg, J.S. Kim, G.L. Amidon, Pharmacokinetics of an immediate release, a

controlled release and a two pulse dosage form in dogs, Eur. J. Pharm. Biopharm. 60

(2005) 17-23.

[31] M.J. Kendall, V.A. John, C.P. Quarterman, P.G. Welling, A single and multiple dose

pharmacokinetic and pharmacodynamic comparison of conventional and slow-release

metoprolol, Eur. J. Clin. Pharmacol. 17 (1980) 87-92.

[32] A. Sandberg, G. Ragnarsson, U.E. Jonsson, J. Sjörgren, Design of a new multiple-unit

controlled-release formulation of metoprolol-metoprolol CR, Eur. J. Clin. Pharmacol. 33

(1988) S3-S7.

[33] A. Sandberg, I. Blomquist, U.E. Jonsson, P. Lundborg, Pharmacokinetic and

pharmacodynamic properties of a new controlled-release formulation of metoprolol: a

comparison with conventional tablets, Eur. J. Clin. Pharmacol. 33 (1988) S9-S14.

[34] A. Sandberg, B. Abrahamsson, A. Svenheden, B. Olofsson, R. Bergstrand, Steady-

state bioavailability and day-to-day variability of a multiple-unit (CR/ZOK) and a single-

unit (OROS) delivery system of metoprolol after once-daily dosing, Pharm. Res. 10

(1993) 28-34.

[35] B. Abrahamsson, M. Alpsten, U.E. Jonsson, P.J. Lundberg, A. Sandberg, M. Sundgren,

A. Svenheden, J. Tölli, Gastro-intestinal transit of a multiple-unit formulation

(metoprolol CR/ZOK) and a non-disintegrating tablet with the emphasis on colon, Int. J.

Pharm. 140 (1996) 229-235.

[36] Z. Ye, P. Rombout, J.P. Remon, C. Vervaet, G. Van den Mooter, Correlation between

the permeability of metoprolol tartrate through plasticized isolated

ethylcellulose/hydroxypropyl methylcellulose films and drug release from reservoir

pellets, Eur. J. Pharm. Biopharm. 67 (2007) 485-490.

132


[37] J. Varshosaz, H. Faghihian, K. Rastgoo, Preparation and characterization of metoprolol

controlled-release solid dispersions, Drug Deliv. 13 (2006) 295-302.

[38] J.C. Gutierrez-Rocca, J.W. McGinity, Influence of aging on the physical-mechanical

properties of acrylic resin films cast from aqueous dispersions and organic solutions,


[39] C. Aitken-Nichol, F. Zhang, J.W. McGinity, Hot-melt extrusion of acrylic films, Pharm.

Res. 13 (1996) 804-808.




[41] C. Wu, J.W. McGinity, Non-traditional plasticization of polymeric films, Int. J. Pharm.

177 (1999) 15-27.

[42] A.D. Godwin, Plasticizers, in: Applied polymer science: 21st Century (C.D. Craver,

J.C.E. Carraher, ed.), Elsevier, New York, USA (2000) 157-175.



[44] Y. Zhu, N.H. Shah, W. Malick, M.H. Infeld, J.W. McGinity, Solid-state plasticization of

an acrylic polymer with chlorpheniramine maleate and triethyl citrate, Int. J. Pharm. 241

(2002) 301-310.

[45] M.M. Crowley, F. Zhang, M.A. Repka, S. Thumma, S.B. Upadhye, S.K. Battu, J.W.

McGinity, C. Martin, Pharmaceutical applications of hot-melt extrusion: part I, Drug

Dev. Ind. Pharm. 33 (2007) 909-926.

[46] J.W. McGinity, F. Zhang, Melt-extruded controlled-release dosage forms, in:

Pharmaceutical extrusion technology (I. Ghebre-Sellassie, C. Martin, ed.), Marcel


133

CHAPTER 2




[48] J.C. Gutierrez-Rocca, J.W. McGinity, Influence of water soluble and insoluble

plasticizers on the physical and mechanical properties of acrylic resin copolymers, Int.

J. Pharm. 103 (1994) 293-301.







[52] R. Bodmeier, O. Paeratakul, The effect of curing on drug release and morphological

properties of ethylcellulose pseudolatex-coated beads, Drug Dev. Ind. Pharm. 20

(1994) 1517-1533.


containing granulations of ibuprofen and an acrylic copolymer prepared by thermal


[54] C. Wu, J.W. McGinity, Influence of methylparaben as a solid-state plasticizer on the

physicochemical properties of Eudragit® RS PO hot-melt extrudates, Eur. J. Pharm.

Biopharm. 56 (2003) 95-100.

[55] A.N. Ghebremeskel, C. Vemavarapu, M. Lodaya, Use of surfactants as plasticizers in

preparing solid dispersions of poorly soluble API: selection of polymer-surfactant

combinations using solubility parameters and testing the processability, Int. J. Pharm.

328 (2007) 119-129.



134





[58] C. De Brabander, Development of sustained release multiple unit dosage forms using

hot-melt extrusion, Thesis presented at the Laboratory of Pharmaceutical Technology,

Ghent University, Ghent, Belgium (2002).

[59] M.A. Repka, Physical – mechanical and chemical properties of topical films produced

by hot-melt extrusion, The University of Texas at Austin, Austin, Texas, USA (2000).

[60] M.A. Frohoff-Hülsmann, A. Schmitz, B.C. Lippold, Aqueous ethyl cellulose dispersions

containing plasticizers of different water solubility and hydroxypropyl methylcellulose as

coating material for diffusion pellets I. Drug release rates from coated pellets, Int. J.

Pharm. 177 (1999) 69-82.

[61] M.A. Frohoff-Hülsmann, A. Schmitz, B.C. Lippold, J.W. McGinity, Aqueous ethyl

cellulose dispersion containing plasticizers of different water solubility and

hydroxypropyl methylcellulose as coating material for diffusion pellets II. Properties of

sprayed films, Eur. J. Pharm. Biopharm. 48 (1999) 67-75.

[62] F. Sadeghi, A. Sadr, The effect of plasticizer type and concentration on characteristics

of ethylcellulose matrices prepared by solid dispersion technique, J. Pharm.

Pharmocol. 54 (2002) S92.

[63] W.J. Lin, H.K. Lee, D.M. Wang, The influence of plasticizers on the release of

theophylline from microporous-controlled tablets, J. Control. Release 99 (2004) 415-

421.

[64] F. Lecomte, J. Siepmann, M. Walther, R.J. MacRae, R. Bodmeier, Polymer blends

used for the aqueous coating of solid dosage forms: importance of the type of

plasticizer, J. Control. Release 99 (2004) 1-13.

[65] D. Henrist, J.P. Remon, Influence of the formulation composition on the in vitro

characteristics of hot stage extrudates, Int. J. Pharm. 188 (1999) 111-119.

135

CHAPTER 2

[66] Y. Zhu, K.A. Mehta, J.W. McGinity, Influence of plasticizer level on the drug release

from sustained release film coated and hot-melt extruded dosage forms, Pharm. Dev.

Technol. 11 (2006) 285-294.

[67] D.H. Morton-Jones (ed.), Polymer processing, Chapman and Hall, London, UK (1989).


[69] Y. Miyagawa, T. Okabe, Y. Yamaguchi, M. Miyajima, H. Sunada, Controlled-release of

diclofenac sodium from wax matrix granule, Int. J. Pharm. 138 (1996) 215-254.

[70] H. Sato, Y. Miyagawa, T. Okabe, Dissolution mechanism of diclofenac sodium from

wax matrix granules, J. Pharm. Sci. 86 (1997) 929-934.

[71] R. Chokshi, H. Zia, Hot-melt extrusion technique: a review, Iran. J. Pharm. Res. 3

(2004) 3-16.

[72] K. Nakamichi, T. Nakano, H. Yasuura, S. Izumi, Y. Kawashima, The role of the

kneading paddle and the effects of screw revolution speed and water content on the

preparation of solid dispersions uing a twin-screw extruder, Int. J. Pharm. 241 (2002)

203-211.

[73] K. Nakamichi, T. Nakano, H. Yasuura, S. Izumi, Y. Kawashima, Stabilization of sodium

guaiazulene sulfonate in granules for tableting prepared using a twin-screw extruder,

Eur. J. Pharm. Biopharm. 56 (2003) 347-354.

[74] M. Yoshinaga, S. Katsuki, M. Miyazaki, L. Liu, S. Kihara, K. Funatsu, Mixing

mechanism of three-tip kneading block in twin screw extruders, Polym. Eng. Sci. 40

(2000) 168-178.

[75] D. Henrist, Hot stage extrusion for the production of starch based matrices for oral

controlled drug delivery, Thesis presented at the Laboratory of Pharmaceutical

Technology, Ghent University, Ghent, Belgium (2001).

[76] H. De Ruyck, Modelling of the residence time distribution in a twin screw extruder, J.

Food Eng. 32 (1997) 375-390.

136


[77] D. Henrist, J.P. Remon, Influence of the process parameters on the characteristics of

starch based hot stage extrudates, Int. J. Pharm. 189 (1999) 7-17.

137

138138138138

CHAPTER 3 POLYETHYLENE OXIDE TO TAILOR

METOPROLOL TARTRATE RELEASE: IN VITRO AND IN VIVO EVALUATION

Parts of this chapter were submitted for publication in:

Influence of polyethylene glycol/polyethylene oxide on the release characteristics of sustained-release

ethylcellulose mini-matrices produced by hot-melt extrusion: in vitro and in vivo evaluation.

E. Verhoeven1, T.R.M. De Beer2, E. Schacht3, G. Van den Mooter4, J.P. Remon1, C. Vervaet1, Eur. J.

Pharm. Biopharm.

1 Laboratory of Pharmaceutical Technology, Ghent University, Ghent, Belgium. 2 Department of Pharmaceutical Analysis, Ghent University, Ghent, Belgium. 3 Department of Organic Chemistry, Ghent University, Ghent, Belgium. 4 Laboratory of Pharmacotechnology and Biopharmacy, University of Leuven, Leuven, Belgium.

139139

CHAPTER 3

ABSTRACT

Mini-matrices with release-sustaining properties were developed by hot-melt

extrusion (diameter 3 mm, height 2 mm) using metoprolol tartrate as model drug (30%, w/w)

and ethylcellulose as sustained-release agent. Polyethylene glycol or polyethylene oxide was

added to the formulation to increase drug release. Changing the hydrophilic polymer

concentration (0, 1, 2.5, 5, 10, 20 and 70%, w/w) and molecular weight (6000, 100.000,

1.000.000 and 7.000.000) modified the in vitro drug release: increasing concentrations

yielded faster drug release (irrespective of molecular weight), whereas the influence of

molecular weight depended on concentration. Smooth extrudates were obtained when

processed at 40 and 70°C for polyethylene glycol and polyethylene oxide formulations,

respectively. Raman analysis revealed that metoprolol tartrate was homogeneously

distributed in the mini-matrices, independent of hydrophilic polymer concentration and

molecular weight. Also drug and polymer crystallinity were independent of both parameters.

An oral dose of 200 mg metoprolol tartrate was administered to dogs either as immediate-

release preparation (Lopresor® 100), as sustained-release formulation (Slow-Lopresor® 200

Divitabs®), or as experimental mini-matrices (varying in hydrophilic polymer concentration).

The sustained-release effect of the experimental formulations was limited and relative

bioavailabilities (using Slow-Lopresor® 200 Divitabs® as reference) of 66.2 and 148.2% were

obtained for 5 and 20% PEO 1.000.000 mini-matrices, respectively.

140140

POLYETHYLENE OXIDE TO TAILOR METOPROLOL TARTRATE RELEASE: IN VITRO AND IN VIVO EVALUATION

CHAPTER 3 POLYETHYLENE OXIDE TO TAILOR

METOPROLOL TARTRATE RELEASE: IN VITRO AND IN VIVO EVALUATION

1. INTRODUCTION

Hot-melt extrusion is a technology used in the pharmaceutical industry to produce

matrix formulations where a drug is homogeneously embedded in a release-controlling

polymer matrix. Since the different steps (mixing, melting, homogenizing and shaping) are

continuously performed as a one-step process on a single machine, it has advantages over

direct compression and granulation as conventional technique to manufacture controlled-

release matrices [1-3].

Ethylcellulose, a polymer with thermoplastic properties, has been intensively

investigated as drug carrier in hot-stage extrusion [4-7]. Mini-matrices could be developed by

hot-melt extrusion using ethylcellulose to sustain the release of ibuprofen: the combination of

ethylcellulose and a hydrophilic component (hydroxypropylmethylcellulose [8-10], xanthan

gum [8-11]) offered a flexible system to tailor the in vitro as well as in vivo drug release.

Due to the specific drug-matrix interaction, the low-melting ibuprofen (melting point

76°C) was identified as a plasticizer for ethylcellulose [12]. As a consequence, the

characteristics of ethylcellulose/hydrophilic polymer mini-matrices containing ibuprofen were

not predictive of the extrusion and dissolution properties of ethylcellulose mini-matrices

containing non-plasticizing drugs. Therefore, ibuprofen was substituted by a drug with a

higher melting point (metoprolol tartrate, 123°C) and a conventional plasticizer was added to

the formulation [13].

In the present study sustained-release mini-matrices were developed by hot-melt

extrusion of a metoprolol tartrate/ethylcellulose/dibutyl sebacate mixture with the addition of

polyethylene glycol or polyethylene oxide as hydrophilic polymer to tailor drug release. The aim

141141

CHAPTER 3

was to examine the effect of different concentrations (0-70%, w/w) and molecular weights

(polyethylene glycol 6000 and polyethylene oxide 100.000-7.000.000) of these hydrophilic

polymers on drug and polyethylene glycol/polyethylene oxide release, drug homogeneity and drug

and polymer crystallinity. The influence of polyethylene oxide concentration on bioavailability in

dogs was also evaluated and compared with xanthan gum mini-matrices [13] and an equivalent

dose of a sustained-release metoprolol tartrate matrix tablet (Slow-Lopresor® 200 Divitabs®).

142142



2.1. Materials

Metoprolol tartrate (MPT) (10 µm) (Esteve Quimica, Barcelona, Spain) was selected

as model drug. The matrix consisted of ethylcellulose (EC) (Ethocel® Std 10 FP Premium;

particle size: 3-15 µm; ethoxyl content: 48-49.5% (w/w); viscosity: 9-11 mPa.s (5% solution in


a hydrophilic component: xanthan gum (XG) (Xantural® 75, mean particle size: 75 µm) (CP

Kelco, Liverpool, UK) or polyethylene glycol or polyethylene oxide. Dibutyl sebacate (DBS)

(Sigma-Aldrich, Steinheim, Germany) was used as plasticizer for ethylcellulose.

2.1.1. Polyethylene glycol, polyethylene oxide

Polyethylene glycol (PEG) having a molecular weight (MW) of 6000 was purchased

from Fagron (Waregem, Belgium). Different types of polyethylene oxide (PEO, SentryTM

PolyoxTM WSR N10, N12K and 303 having a MW of 100.000, 1.000.000 and 7.000.000,

respectively) were supplied from Dow Chemical Company (Midland, USA).

Polyethylene glycol (Figure 1c) and polyethylene oxide (Figure 1d) are non-ionic,

water soluble, linear, semi-crystalline, high molecular weight homo-polymers manufactured

by polymerization of ethylene glycol (Figure 1a) and ethylene oxide (Figure 1b), respectively.

The polymer backbone of both is identical, but the major difference between the two

materials is the terminal end-groups: PEG has hydroxyl groups at both ends of the chain,

whereas PEO has one hydroxyl end-group. They are available in a broad range of molecular

weights, i.e. PEG up to 20.000 g/mol (daltons, Da), PEO from 100.000 up to 8.000.000 Da

[14-16]. Polyethylene oxides can be used as tablet binder in direct compression [14, 17-27],

in hot-melt produced dosage forms (e.g. tablets, films, beads, solid dispersions, solid

solutions) [16, 28-41] or in other drug delivery systems (e.g. nanoparticles, hydrogels) [42-

46]. High molecular weight PEOs have been successfully applied in controlled-release

dosage forms, because the rate of swelling, mucoadhesive properties and erosion of the

polymer allow to sustain the release of active ingredients. As the swelling properties (thus

matrix forming capacity) depend on the molecular weight, modification of the release kinetics

is possible by choosing the appropriate type of PEO grade.

143143

CHAPTER 3

(a) (b)

(c) (d)

Figure 1: Chemical structure of (a) ethylene glycol, (b) ethylene oxide, (c) polyethylene

glycol, (d) polyethylene oxide.


2.2. Methods

2.2.1. Hot-melt extrusion

2.2.1.1. Production of mini-matrices

The MPT content was kept constant at 30% (w/w). The concentration of hydrophilic

polymer varied between 0 and 70% (w/w) for all PEG/PEO types under investigation. The

remaining part of the formulation consisted of EC/DBS, in a ratio of 2/1 (w/w) [13]. The

concentrations of PEG, PEO and EC/DBS were varied according to Table 1.

Table 1: Composition (%, w/w) of the mini-matrices.

MPT 30 30 30 30 30 30 30

PEG/PEO - 1 2.5 5 10 20 70

EC 46.7 46 45 43.3 40 33.3 -

DBS 23.3 23 22.5 21.7 20 16.7 -

144144


The components were blended in a planetary mixer (15 min, 90 rpm) (Kenwood Major


interaction between EC and plasticizer. The mixture was passed through the screws of the

powder feeder of the extruder and recycled into the powder reservoir to homogeneously distribute

DBS in the powder prior to hot-melt extrusion. Hot-melt extrusion was performed using a

laboratory-scale intermeshing co-rotating twin-screw extruder (MP19TC-25, APV Baker,

Newcastle-under-Lyme, UK) having a length-to-diameter ratio of 25/1. The machine was

equipped with a Brabender twin-screw powder feeder and an extrusion screw having two mixing

sections and a densification zone. The die block (2.6 cm thickness) was fixed to the extruder

barrel and the axially mounted die plate (1.9 cm thickness, with a cylindrical hole of 3 mm

diameter for shaping the extrudates) was attached to the die block. The following extrusion

conditions were used: a screw speed of 30 rpm, a powder feed rate of 6 g/min and a temperature

of 40 and 70°C for the five heating zones along the barrel for extrusion of PEG and PEO

formulations, respectively. After cooling down to room temperature, the extruded rods (∅ = 3 mm)

were manually cut, using surgical blades, into mini-matrices of 2 mm length.

The surface properties of the extrudates were visually inspected for defects (shark

skinning) and evaluated for their suitability to be cut into mini-matrices (deformation due to cutting,

smoothness of the cutting surfaces and edges) using a digital camera (C3030 Olympus) linked to

an image analysis system (analySIS®, Soft Imaging system, Münster, Germany) (magnification

9.5x). Drug and PEG/PEO release was evaluated by dissolution testing. Drug (and polymer)

crystallinity and homogeneity was evaluated by X-ray diffraction and Raman analysis,

respectively.


To investigate the thermal stability of the polymers (EC, PEG and PEO) during the

hot-melt extrusion process, the polymers were extruded at 100 and 200°C, using the process

parameters described above. The samples (powder before extrusion, extruded samples at

100 and 200°C) were analyzed by means of gel filtration chromatography with the

appropriate polymer standards.

Binary mixtures (EC/DBS, EC/PEG(PEO), EC/MPT, DBS/PEG(PEO) and

MPT/PEG(PEO)) were extruded under the same processing conditions as described above,

and evaluated for changes in glass transition temperature, melting point, crystallinity and

possible interactions by modulated temperature differential scanning calorimetry (MTDSC)

145145

CHAPTER 3

(by comparing the results with the thermograms of the powder blends prior to extrusion). All

types of PEG/PEO were investigated. The ratios of both components in these blends were

identical to their ratio in the mini-matrices (containing 2.5, 10, 20 and 70% (w/w) hydrophilic

polymer).

2.2.2. In vitro drug and PEG/PEO release

Drug release from the mini-matrices (approximately 60 mg) was determined using

USP apparatus 1 (baskets) [47], in a VK 7010 dissolution system combined with a VK 8000

automatic sampling station (VanKel, New Jersey, USA). Demineralized water was used as

dissolution medium. The temperature of the medium (900 ml) was kept at 37 ± 0.5°C, while

the rotational speed of the baskets was set at 100 rpm. Samples of 5 ml were withdrawn at

0.5, 1, 2, 4, 6, 8, 12, 16, 20 and 24 h and spectrophotometrically analyzed for MPT at 222 nm

by means of a Perkin Elmer Lambda 12 UV-VIS double beam spectrophotometer (Zaventem,

Belgium). The dissolution was performed in 6 dissolution vessels, each vessel containing 4

mini-matrices.

PEG/PEO release from the mini-matrices was determined using the USP paddle

method under the same conditions as described above. PEG/PEO concentrations in the

samples were measured by gel filtration chromatography (GFC) (detection via refractive

index). Due to the sensitivity of the GFC analysis, the sample size for dissolution testing was

increased (36, 9, 5 and 2 g mini-matrices for the 2.5, 10, 20 and 70% PEG/PEO

formulations, respectively) and the paddle method was used. The PEG/PEO release was

performed in twofold. The stability of PEG/PEO following hot-melt extrusion was also

determined via GFC. For PEG 6000 and PEO 100.000 analysis, a Waters 600

Controller/Waters 610 Fluid Unit pump (flow rate: 1.0 ml/min) with a Waters 410 Differential

Refractometer was used. 20 µl sample (using a 100 µl Hamilton syringe) was injected on the

column (TSK-gel® G 4000 PW (TosoHaas; 7.5 mm ID x 30 cm – particle size 17 µm) and

TSK-gel® G 3000 PW (7.5 mm ID x 30 cm – particle size 10 µm) coupled in series). PEO

1.000.000 and 7.000.000 samples were analyzed using a Waters 515 HPLC pump (flow: 1.0

ml/min) combined with a Waters Differential Refractometer R401. 20 µl sample was injected

on the column (TSK-gel® G 6000 PW (7.5 mm ID x 30 cm – particle size 17 µm) and TSK-

gel® G 5000 PW (7.5 mm ID x 30 cm – particle size 10 µm) coupled in series). The

chromatograms were integrated using the Millenium 2010 Chromatography Manager

(Version 2.15.01). The mobile phase consisted of water (Milli-Q® Water Purification System),

filtered using Durapore® membrane filters (0.45 µm – HVLP type) and degassed with helium.

146146


The PEG/PEO concentrations were calculated from a calibration curve between 0.10 and

5.00 mg/ml, using polymer standards (analytical or HPLC grade).

To determine the in vitro release properties of the formulations used during the in vivo

experiments, a dissolution experiment was performed with hard-gelatin capsules (n° 000)

containing 667 mg mini-matrices. Their swelling behaviour (after 4 and 24 h immersion in the

dissolution medium) was also evaluated by microscopic analysis using a digital camera

(C3030 Olympus) linked to an image analysis system (analySIS®, Soft Imaging system,

Münster, Germany) (magnification 9.5x).

2.2.3. Raman analysis

MPT distribution in the mini-matrices containing different concentrations and types of

PEG/PEO was evaluated by Raman spectroscopic mapping. Three areas (2150 x 1150 µm2)

(two at the edges and one in the middle of the mini-matrix) of each mini-matrix (n = 3) were

scanned using a 10x long working distance objective lens (laser spot size: 50 µm) in point-

by-point mapping mode with a step size of 100 µm in both the x and y directions. The

resulting map provides an overview of the MPT distribution in the mapped area. The

mapping system used for this study was a RamanRxn 1 Analyzer and Microprobe (Kaiser

Optical Systems, Ann Arbor, USA), equipped with an air cooled CCD detector (back-







2.2.4. Thermal analysis

The glass transition temperature (Tg) and/or melting point (Tm) of the pure components

(EC, MPT, DBS, PEG and PEO) and in binary mixtures of these materials (powder blends prior to

extrusion and hot-melt extruded samples) were determined via modulated temperature differential

scanning calorimetry (MTDSC).

147147

CHAPTER 3

The thermal behaviour of powders and hot-melt extruded samples was evaluated

using a 2920 Modulated DSC (TA Instruments, Leatherhead, UK) equipped with a

refrigerated cooling system (RCS). Dry helium at a flow rate of 40 ml/min was used as purge

gas through the DSC cell and 150 ml/min nitrogen through the RCS unit. Samples (± 10 mg)

were run in closed aluminium pans supplied by TA Instruments; the mass of each empty

sample pan was matched with the mass of the empty reference pan to ± 0.10 mg. The

experimental method consisted of an initial 5 min isothermal equilibration period at 0°C.

During the subsequent heating run the following experimental parameters were used: an

underlying heating rate of 2°C/min from -70 to 200°C, a modulation amplitude of 0.212°C and

a period of 40 s. Temperature and enthalpic calibration were performed with an indium

standard, whereas calibration of the heat capacity was performed with a sapphire standard.

The results were analyzed using the TA Instruments Universal Analysis Software.

Measurements were performed in duplicate and the Tg values (midpoint half height) are

reported.

2.2.5. X-ray diffraction

To investigate the crystallinity of the components in the mini-matrices, X-ray

diffraction was performed (on pulverized mini-matrices). The X-ray patterns of MPT, EC,

PEG, PEO and mini-matrices were determined using a D5000 Cu Kα Diffractor (λ = 0.154

nm) (Siemens, Karlsruhe, Germany) with a voltage of 40 mA in the angular range of 10° < 2θ

< 60° using a step scan mode (step width = 0.02°, counting time = 1 s/step).

2.2.6. In vivo evaluation

All procedures were performed in accordance with the guidelines and approval of the

local Institutional Animal Experimentation Ethics Committee.

2.2.6.1. Subjects and study design

A group of 6 male mixed-breed dogs (weight 22.0-38.0 kg) was used in this study. To

investigate the influence of the concentration of the hydrophilic component (PEO or XG) on

the bioavailability of metoprolol tartrate, the following drug formulations were administered:

148148


F-1: hot-melt extruded mini-matrices, consisting of 30% MPT, 5% XG, 43.3% EC and

21.7% DBS (5% XG)

F-2: hot-melt extruded mini-matrices, consisting of 30% MPT, 10% XG, 40% EC and 20%

DBS (10% XG)

F-3: hot-melt extruded mini-matrices, consisting of 30% MPT, 20% XG, 33.3% EC and

16.7% DBS (20% XG)

F-4: hot-melt extruded mini-matrices, consisting of 30% MPT, 5% PEO 1.000.000, 43.3%

EC and 21.7% DBS (5% PEO)

F-5: hot-melt extruded mini-matrices, consisting of 30% MPT, 20% PEO 1.000.000, 33.3%

EC and 16.7% DBS (20% PEO)

F-6: Lopresor® 100 (commercially available immediate-release tablets containing 100 mg

metoprolol tartrate (Sankyo Pharma, Louvain-la-Neuve, Belgium))

F-7: Slow-Lopresor® 200 Divitabs® (commercially available hydrophilic matrix tablets

containing 200 mg metoprolol tartrate (Sankyo Pharma, Louvain-la-Neuve, Belgium))

The mini-matrices of the experimental formulations were filled in hard-gelatin

capsules n° 000, each capsule containing 200 mg metoprolol tartrate. Lopresor® 100 (2

tablets) and Slow-Lopresor® 200 Divitabs® (1 tablet) were used as immediate- and sustained-

release reference formulations, respectively. The formulations were administered in

randomized order with a wash-out period of at least 8 days between different sessions. On

the experimental days the dogs were fasted for 12 h prior to the study period, although water

was available ad libitum. Before administration of the formulations, an intravenous cannula

was placed in the lateral saphenous and a blank blood sample was obtained. The

formulations were orally administered with 10 ml water. The blood samples (2 ml at each

sampling) were collected in dry heparinized tubes 0.5, 1, 2, 3, 4, 6, 8, 12, 24 and 36 h after

intake of the formulations. No food was administered to the dogs during the initial 24 h of the

test, afterwards they resumed their usual diet. Water could be taken freely. Within 1 h after

collection, blood was centrifuged for 10 min at 1450g and the plasma was immediately

assayed for metoprolol tartrate.

2.2.6.2. Metoprolol tartrate assay

The metoprolol tartrate plasma concentrations were determined by a validated HPLC-

fluorescence method. All chemicals were of analytical or HPLC grade.

149149

CHAPTER 3

A solid phase extraction (SPE) procedure was used to extract metoprolol tartrate. The

SPE equipment consisted of Oasis® HLB (1 cc 30 mg) cartridges (Waters, Brussels,

Belgium) and a 16-port vacuum manifold (Alltech Europe, Laarne, Belgium). Extraction

cartridges were first conditioned by rinsing consecutively with methanol (1 ml), water (1 ml)

and phosphate-buffered saline (PBS) (1 ml). Plasma samples (300 µl) were mixed with 20 µl

alprenolol (5.625 µg/ml alprenolol in water, as internal standard) and 680 µl PBS,

homogenized by 30 s vortexing, and then loaded on the column. The columns were washed

with water (1 ml) and the analytes were then eluted with methanol (1 ml). The eluates were

evaporated to dryness under N2, reconstituted in 150 µl water (30 s of vortexing) and 20 µl

solution was injected in the HPLC system. The metoprolol tartrate plasma concentrations

were determined via a calibration curve. For the standard curves, spiked samples were

prepared by mixing 280 µl plasma with 20 µl of a solution containing different concentrations

of metoprolol tartrate in water (0.375, 0.5625, 0.75, 2.25, 5.625, 7.5, 15.0 and 22.5 µg/ml). To

these spiked samples 20 µl alprenolol (internal standard solution) and 680 µl PBS was

added. These mixtures were extracted as described above.

The HPLC equipment (Merck-Hitachi, Darmstadt, Germany) consisted of a solvent

pump (L-7110) set at a constant flow rate of 0.8 ml/min, a variable wavelength fluorescence

detector (L-7480) set at an excitation and emission wavelength of 275 and 300 nm,

respectively, a LiChrospher® 100 CN 5 µm column (250 mm x 4 mm) and precolumn (4 mm x

4 mm), an auto-sampler (Gilson 234 autoinjector, Middleton, Wisconsin, USA) with a 50 µl

loop (Valco Instruments, Houston, Texas, USA) and an automatic integration system

(software D-7000 Multi-Manager) to collect and process the signals. The mobile phase

consisted of acetonitrile/sodium dihydrogen orthophosphate buffer (2 M)/water (5/0.5/94.5,

v/v/v). The mixture was adjusted to pH 3.0 with phosphoric acid.


For the general considerations about method validation of an analytical procedure, we

refer to Chapter 1 and the guidelines of the International Conference on Harmonisation (ICH) [48].


In our experiment, specificity was determined at an excitation and emission wavelength of

275 and 300 nm, respectively, by comparing the chromatograms after extraction of blank plasma,

150150


plasma spiked with internal standard alprenolol, and plasma spiked with metoprolol tartrate and

alprenolol.


Different calibration curves (n = 10) were prepared. For each calibration curve, blank

plasma of dogs was spiked with different metoprolol tartrate solutions to obtain plasma

concentrations from 0 to 3 µg/ml, and internal standard (alprenolol, 0.75 µg/ml) to cover the

concentration range of the unknown samples.





precision (between-day repeatability). The precision was calculated for different metoprolol

tartrate concentrations (0.05, 0.075, 0.1, 0.3, 0.75, 1, 2 and 3 µg/ml).

2.2.6.3.4. Accuracy

Accuracy was determined for different metoprolol tartrate concentrations (0.05, 0.075,

0.1, 0.3, 0.75, 1, 2 and 3 µg/ml) (n = 10).

2.2.6.3.5. Recovery

The recovery experiments (extracted samples: n = 10, non-extracted samples: n = 3)

were performed at metoprolol tartrate and alprenolol concentrations of 0.05, 0.3, 1 and 3

µg/ml, and 0.75 µg/ml, respectively.

151151

CHAPTER 3


The limit of detection (LOD) and limit of quantification (LOQ) were calculated based on the

standard deviation of the intercept and slope of the mean calibration curve.

2.2.6.4. Data analysis

The peak plasma concentration (Cmax), the time to reach Cmax (Tmax) and the extent of

absorption (area under the curve, AUC0-36h) were calculated using the MW-Pharm Program

version 3.0 (Mediware 1987-1991, Utrecht, The Netherlands). The AUC0-36h was calculated

using logarithmic and linear trapezoidal rules. The relative bioavailability (Frel, expressed in

%) was calculated as the ratio of AUC0-36h between a test formulation and the sustained-

release reference formulation (Slow-Lopresor® 200 Divitabs®). The sustained-release

characteristics of a formulation were evaluated by the time span during which the plasma

concentrations were at least 50% of the Cmax value (HVDt50%Cmax, the width of the plasma

concentration profile at 50% of Cmax) [49, 50]. The HVDt50%Cmax values were determined from

the individual plasma concentration-time profiles. The ratio between the HVDt50%Cmax values

of a test formulation and the immediate-release reference formulation (Lopresor® 100)

(expressed as RD) is indicative of the sustained-release effect: a ratio of 1.5, 2 and > 3

indicating a low, intermediate and strong sustained-release effect, respectively [49].


The effect of metoprolol tartrate formulation on the bioavailability was assessed by

repeated-measures ANOVA (univariate analysis). To further compare the effects of the

different treatments, a multiple comparison among pairs of means was performed using a

Bonferroni post-hoc test with P < 0.05 as significance level. The normality of the residuals

was tested with a Kolmogorov-Smirnov test. SPSS version 16.0 was used to perform the

statistical analysis.

152152



3.1. In vitro evaluation

The drug release rate from hot-melt extruded dosage forms is mainly determined by

the ratio of hydrophilic and hydrophobic polymers included in the matrix formulations [5, 6,

16, 51-53]. Based on this principle polymeric hot-melt extruded mini-matrices consisting of

EC (hydrophobic matrix former) and XG (hydrophilic additive to modify drug release) yielded

zero-order release kinetics of freely water soluble drugs. Drug release depended on XG

concentration and grade [11, 13] as these factors determined the swelling behaviour and

water uptake of the mini-matrices. As an alternative to XG, PEG and PEO could be used as

these polymers have already been used for hot-melt extrusion. Since it is known that the

swelling properties of PEO depend on their molecular weight [16-24, 28], variable

concentrations and grades of PEG/PEO were processed in combination with EC to modify

drug release from these matrices.

The different formulations (Table 1) were easily processed at an extrusion

temperature of 40 and 70°C for PEG and PEO formulations, respectively, independent of

PEG/PEO concentration and MW: no surface defects were detected and the extrudates

having a smooth surface could be cut into high quality mini-tablets (without cracks or other

irregularities at the cutting edges).

153153

CHAPTER 3

Whereas previous work by the authors [13] already confirmed that MPT is stable at

these processing temperatures, the thermal stability of EC, PEG and PEO after hot-melt

extrusion at 100 and 200°C was analyzed using GFC since polymers are subjected to

mechanical, thermal and oxidative degradation during hot-melt extrusion. For all polymers,

the MW profile of the samples prior to and after extrusion overlapped. As an example, the

GFC chromatograms of PEO 1.000.000 are shown (Figure 2). The retention time of PEG

6000, PEO 100.000, 1.000.000 and 7.000.000 during GFC analysis were 15, 11, 13 and 10

min, respectively. Other investigators reported that PEO stability in matrix tablets prepared

by hot-melt extrusion depended on storage and processing temperature, screw speed and

the MW of the polymer [16, 29].

0

2

4

6

8

10

0 5 10 15 20 25

Time (min)

Sign

al (m

V)

Figure 2: Gelfiltration chromatograms of PEG 6000: (□) prior to extrusion, hot-melt extruded

sample at (△) 100°C and (○) 200°C.

154154


An increase of PEG/PEO content significantly increased the drug release rate,

irrespective of their MW due to the higher hydrophilicity of the matrix. Figure 3 shows

exemplarily the results obtained for PEO 100.000-containing mini-matrices. For all mini-

matrices, the release from a formulation containing 1 and 2.5% PEG/PEO was similar,

whereas the data of the 5% PEG/PEO mini-matrices coincided with the 10% PEG/PEO

formulations. Interestingly the mini-matrices retained their cylindrical shape during dissolution

testing and did not significantly swell, except for the 70% PEG/PEO formulations which had

dissolved after approximately 1 h. Although the swelling of purely hydrophilic PEG/PEO

matrices and films has been reported during dissolution [16-24, 28], the cohesive nature of

the EC matrix structure probably restricted swelling of the mini-matrices upon wetting.

Formulations with 0, 1, 2.5, 5 and 10% showed zero-order, but incomplete release kinetics.

Complete drug release was observed at PEG/PEO concentrations of 20% and above,

however the burst release from these formulations was significant: 43.6 and 81.6% released

after 30 min (PEO 100.000).

0

20

40

60

80

100

120

0 4 8 12 16 20 24

Time (h)

MPT

rele

ased

(%)

Figure 3: Influence of PEO 100.000 concentration on the dissolution profiles (mean ± S.D., n

= 6) of mini-matrices containing 30% (w/w) MPT, EC/plasticizer (2/1, w/w) and PEO 100.000:

(■) 0%, (▲) 2.5%, (●) 10%, (□) 20% and (△) 70%.

155155

CHAPTER 3

At low PEG/PEO concentration (1-2.5%) no effect of PEG/PEO MW on drug release

was observed. Although a slower drug release is described from hydrophilic matrices

containing higher MW PEO due to the increasing viscosity of the polymer [16-24, 28, 30] (a

higher MW increases gel strength, which tends to decrease the diffusion of the drug), drug

release from matrices containing 5 and 10% PEG 6000 and PEO 100.000 was significantly

slower in comparison to the PEO 1.000.000 and 7.000.000 mini-matrices (Figure 4). This

might be due to the faster release of low MW PEG/PEO from the matrix (in comparison with

the release of high MW PEO), resulting in molecular rearrangement of EC, thus modifying

the porous network inside the matrix and hindering drug release. The dissolution conditions

provided sufficient mobility for the EC polymer chains as MTDSC analysis showed that Tg of

EC in combination with DBS (EC/DBS ratio: 2/1) was 40°C.

0

20

40

60

80

100

120

0 4 8 12 16 20 24

Time (h)

MPT

rele

ased

(%)

Figure 4: Influence of PEG/PEO MW on the dissolution profiles (mean ± S.D., n = 6) of mini-

matrices containing 30% (w/w) MPT, EC/plasticizer (2/1, w/w) and 10% (w/w) hydrophilic

polymer: (■) PEG 6000, (▲) PEO 100.000, (●) PEO 1.000.000, (□) PEO 7.000.000.

156156


At 20 and 70% PEG/PEO the drug release rate was slower using high MW PEO

(Figure 5), which is consistent with data reported by other investigators [16-24, 28, 30], due

to the increasing viscosity at higher MW since the properties of PEG/PEO became more

predominant for drug release.

0

20

40

60

80

100

120

0 4 8 12 16 20 2

Time (h)

MPT

rele

ased

(%

4

)

Figure 5: Influence of PEG/PEO MW on the dissolution profiles (mean ± S.D., n = 6) of mini-

matrices containing 30% (w/w) MPT and 70% (w/w) hydrophilic polymer: (■) PEG 6000, (▲)

PEO 100.000, (●) PEO 1.000.000, (□) PEO 7.000.000.

PEG/PEO release increased at higher PEG/PEO concentration (irrespective of MW)

as less matrix forming polymer was present and the mini-matrices became less hydrophobic.

In addition, PEG/PEO release from mini-matrices depended on the MW of the hydrophilic

polymer as the viscous nature of high MW PEO reduced its mobility and consequently

release kinetics. MPT release was not in all cases correlated with PEG/PEO leaching,

especially in relation to the effect of PEG/PEO MW: although a slower PEG/PEO release

was observed for all formulations formulated with high MW PEO, drug release was not

affected by PEO MW in case of mini-matrices containing low PEG/PEO concentrations (1-

2.5%) and drug release from mini-matrices containing 5-10% PEG/PEO increased when

using higher MW PEG/PEO.

157157

CHAPTER 3

The MPT distribution in the mini-matrices was monitored off-line via Raman using the

627-653 cm-1 spectral band as no overlap with other ingredients occurred at these

wavenumbers (Figure 6). No difference was seen between the different PEG/PEO types.

Based on the intensity of the Raman signal across the scanned sections of the mini-matrices

it was confirmed that MPT is homogeneously distributed in the mini-matrix irrespective of the

PEG/PEO concentration and MW.

Figure 6: Raman spectra from the compounds in the mini-matrices.

158158


Figure 7 shows the X-ray diffraction spectra of MPT, EC, PEO 1.000.000 and 20%

PEO 1.000.000 mini-tablet. The drug and hydrophilic polymer were crystalline, while EC was

amorphous. The X-ray pattern of the hot-melt extruded 20% PEO 1.000.000 mini-tablet

showed only diffraction peaks corresponding to MPT and PEO 1.000.000, indicating that

these products remained crystalline during extrusion. However, the diffraction peaks were

smaller in comparison with the pure materials, illustrating the presence of smaller crystals.

Other investigators also noticed a reduction of PEO peak intensity (decrease of crystallinity)

following thermal processing in combination with peak broadening (due to a wider

distribution of crystal size) [31-34].

0

500

1000

1500

2000

10 20 30 40 50 60

2-teta (°)

Lin

(Cou

nts)

Figure 7: X-ray diffraction pattern of (red) PEO 1.000.000, (purple) MPT, (green) EC and

(blue) 20% PEO 1.000.000 mini-matrix.

To check for changes in the thermal behaviour of the components or to identify

interactions between components, MTDSC analysis was performed: the values of Tg of EC

and Tm of MPT were 127.9 ± 0.2°C and 122.6 ± 0.4°C, respectively. The Tm of PEG 6000,

PEO 100.000, PEO 1.000.000 and PEO 7.000.000 is 60.3 ± 0.0°C, 65.4 ± 0.4°C, 67.7 ±

0.1°C and 69.5 ± 0.3°C, respectively. After extrusion of PEG and PEO samples at 40 and

70°C, respectively, (blends without PEG or PEO were processed at both temperatures) no

shifts in Tg and Tm were detected in case of binary mixtures and the individual thermal

159159

CHAPTER 3

signals were detected, indicating that no interactions between components occurred during

extrusion under the processing conditions described above. However, others reported

changes in drug and polymer thermal transitions when drug and PEOs were co-processed,

associated with their (partially) miscibility and formation of solid dispersions and solid

solutions [24, 28-36]. All binary blends of EC/DBS, EC/PEG(PEO), EC/MPT and

MPT/PEG(PEO) could be processed via extrusion and were subsequently analyzed via

MTDSC. In contrast, only a binary mixture of PEG(PEO)/DBS in a ratio of 20/16.7

(corresponding to a formulation with 20% hydrophilic polymer) could be processed via

extrusion for PEO 1.000.000 and 7.000.000 as the other PEG(PEO)/DBS ratios (Table 1)

and PEG/PEO grades resulted in phase separation (due to the high liquid fraction).

3.2. In vivo evaluation

3.2.1. Method validation

3.2.1.1. Specificity

Chromatograms after extraction of blank plasma, plasma spiked with alprenolol

(internal standard) and plasma spiked with metoprolol tartrate and alprenolol are shown in

Figure 8. The chromatograms indicate that there was no interference of drug and internal

standard with endogenous compounds, and separation between metoprolol tartrate and

alprenolol was complete. As shown in Figure 8, a retention time of ± 19 min and ± 56 min

was obtained for metoprolol tartrate and alprenolol, respectively. It was concluded that the

method was selective to determine metoprolol tartrate in dog plasma.

160160


Absorption

Time (min)Figure 8: Chromatograms showing metoprolol tartrate and alprenolol after extraction of (top)

blank dog plasma, (middle) plasma spiked with alprenolol and (bottom) plasma spiked with

metoprolol tartrate and alprenolol.

161161

CHAPTER 3

3.2.1.2. Linearity

The mean calibration curve (Y = 2.6975 (± 0.3274) X + 0.0435 (± 0.0258), n = 10)

was linear between 0 and 3 µg/ml metoprolol tartrate, with a coefficient of determination of

0.9976 ± 0.0015.

3.2.1.3. Precision




precision (n = 5).



0.05 6.3 10.5

0.075 9.5 7.8

0.1 6.1 11.4

0.3 8.6 1.7

0.75 7.9 11.6

1.0 6.8 6.1

2.0 10.5 5.5

3.0 3.7 4.1



3.2.1.4. Accuracy

The accuracy was determined for all metoprolol tartrate concentrations and the mean

results are presented in Table 3.

162162




0.05 0.06 ± 0.01 125.7 ± 14.8 11.8

0.075 0.08 ± 0.01 109.8 ± 9.5 8.7

0.1 0.11 ± 0.01 107.9 ± 9.0 8.3

0.3 0.31 ± 0.03 102.8 ± 11.2 10.9

0.75 0.69 ± 0.07 91.7 ± 8.8 9.6

1.0 1.02 ± 0.11 101.9 ± 10.9 10.7

2.0 1.96 ± 0.18 98.1 ± 8.9 9.1

3.0 3.08 ± 0.13 102.6 ± 4.4 4.3



3.2.1.5. Recovery

The recovery of metoprolol tartrate and alprenolol from plasma are shown in Table 4.

Table 4: Recovery (%, mean ± S.D., n = 10) of metoprolol tartrate and alprenolol from

plasma.

Concentration (µg/ml) Metoprolol tartrate Alprenolol

0.05 79.3 ± 13.4 -

0.3 85.0 ± 14.1 -

0.75 - 85.1 ± 5.3

1.0 83.7 ± 9.9 -

3.0 80.4 ± 5.4 -

163163

CHAPTER 3

3.2.1.6. Limit of detection and limit of quantification

Based on the mean linear regression equation of metoprolol tartrate the limit of

detection and the limit of quantification for metoprolol tartrate are 0.03 and 0.10 µg/ml,

respectively. The limit of quantification was higher than the lowest concentration of the

calibration curve but as accuracy and precision for the concentration of 0.05 µg/ml were

within accepted limits [54], the concentration of 0.05 µg/ml was considered as limit of

quantification.

3.2.2. In vivo performance

MPT, a drug with high solubility and permeability is well-absorbed over a large part of

the gastro-intestinal tract. Its relatively short half life (3-4 h) makes MPT a suitable candidate

for an extended-release formulation [55-60]. The enhanced therapeutic efficacy of this drug

through the provision of constant drug input and maintenance of steady-state blood levels is

well-documented and different types of controlled-release formulations improved the clinical

efficacy of MPT. Its administration in dogs is considered to be safe [61, 62].

Prior to the investigation of the influence of PEG/PEO concentrations on MPT

release, the authors investigated in Chapter 2 the effect of XG concentration on in vitro drug

release [13]. Based on the different swelling behaviour of the mini-matrices formulated with

both hydrophilic polymers upon contact with the dissolution medium (swelling in case of XG

versus no swelling for PEG/PEO mini-matrices), it is of interest to compare the bioavailability

of both types of mini-tablets. Drug release from formulations without hydrophilic polymer and

with low concentrations (1 and 2.5%, w/w) of XG [13] and PEG/PEO (Figure 3) was not

complete after 24 h, and these mini-matrices were therefore not selected for in vivo analysis.

The MPT release kinetics from 5, 10 and 20% XG mini-matrices (all resulting in complete

drug release after 24 h) were significantly different [13], and these formulations were

selected for the in vivo study. Only formulations with 5 and 20% PEO were administered to

the dogs as 10% PEO mini-tablets yielded similar dissolution profiles as 5% PEO mini-

matrices. Mini-matrices with 70% PEG/PEO gave a burst release (Figure 3) and were

consequently not of interest as sustained-release dosage form. PEO 1.000.000 (intermediate

MW) was selected as hydrophilic polymer for the in vivo analysis.

164164


Figure 9 shows the mean plasma concentration-time profiles after oral administration

of 200 mg MPT to 6 dogs as Lopresor® 100 (2 tablets), Slow-Lopresor® 200 Divitabs® (1

tablet) and the experimental mini-matrices. The pharmacokinetic parameters are reported in

Table 5.

0

10

20

30

40

0 10 20 30

Time (h)

40

Con

c. M

PT (µ

g/m

l)

Figure 9: Mean plasma concentration-time profiles (± S.D., n = 6) after oral administration of

200 mg metoprolol tartrate to dogs: (■) Lopresor® 100 (2 tablets), (□) Slow-Lopresor® 200

Divitabs® (1 tablet), (●) 5% XG mini-matrices, (○) 10% XG mini-matrices, (x) 20% XG mini-

matrices, (▲) 5% PEO 1.000.000 mini-matrices and (△) 20% PEO 1.000.000 mini-matrices.

0

10

20

30

40

0 1Time (h)

Con

c. M

PT (µ

g/m

l

0

)

165165

CHAPTER 3

Table 5: Mean pharmacokinetic parameters (± S.D., n = 6) after oral administration of 200

mg metoprolol tartrate to dogs as 5% XG mini-matrices, 10% XG mini-matrices, 20% XG

mini-matrices, 5% PEO 1.000.000 mini-matrices, 20% PEO 1.000.000 mini-matrices,

Lopresor® 100 (2 tablets) and Slow-Lopresor® 200 Divitabs® (1 tablet).

Cmax

(µg/ml)

Tmax

(h)

AUC0-36h

(µg.h/ml)

HVDt50%Cmax

(h)

Frel*

(%)

RD

5% XG 3.9 ± 0.7 3.3 ± 1.4 39.2 ± 16.2 4.6 ± 1.8 51.5 ± 27.8 1.6 ± 0.4

10% XG 6.6 ± 2.3 3.5 ± 1.4 72.1 ± 22.5 5.7 ± 2.3 94.4 ± 43.2 2.2 ± 1.1

20% XG 13.9 ± 3.5 2.7 ± 0.8 100.6 ± 39.9 4.0 ± 1.6 131.5 ± 70.0 1.4 ± 0.2

5% PEO 1.000.000 5.9 ± 2.1 2.2 ± 0.4 51.5 ± 22.6 4.3 ± 1.2 66.2 ± 31.0 1.6 ± 0.5

20% PEO 1.000.000 21.3 ± 7.9 2.7 ± 0.5 113.5 ± 54.9 3.7 ± 1.0 148.2 ± 87.7 1.4 ± 0.6

Lopresor® 100 (2 tablets) 31.8 ± 4.6 1.3 ± 0.5 151.3 ± 51.4 2.9 ± 1.2 - -

Slow-Lopresor® 200 Divitabs® (1 tablet) 13.3 ± 2.8 3.2 ± 1.0 87.8 ± 35.8 4.9 ± 1.2 - 2.0 ± 1.0

-: not applicable *: using Slow-Lopresor® 200 Divitabs® (1 tablet) as reference

166166


The in vivo behaviour of the experimental mini-matrices was reflected in their in vitro

dissolution profiles (Figure 10).

0

20

40

60

80

100

0 4 8 12 16 20 24

Time (h)

MPT

rele

ased

(%)

Figure 10: Influence of XG and PEO 100.000 concentration on the dissolution profiles (mean

± S.D., n = 6) of mini-matrices containing 30% (w/w) MPT, EC/plasticizer (2/1, w/w) and

hydrophilic polymer: (●) 5% XG, (○) 10% XG, (x) 20% XG, (▲) 5% PEO 1.000.000 and (△)

20% PEO 1.000.000.

An increasing XG concentration of 5 to 20% enhanced drug release, which was

reflected in the higher AUC0-36h and Frel. The same observation was valid for the PEO mini-

matrices as the AUC0-36h values correlated with the in vitro release profiles. Concerning the

AUC values, no hot-melt extruded formulation was significantly different when compared with

the sustained-release reference formulation Slow-Lopresor® 200 Divitabs® (P > 0.05),

although the AUC values tended to increase at higher hydrophilic polymer concentration.

None of the formulations showed a strong sustained-release effect since RD values ranged

from 1.4 to 1.6, except for the 10% XG mini-matrices which showed an intermediate

sustained-release effect (RD 2.2). In comparison with the 20% XG mini-tablets, the 10% XG

formulation is characterized by a higher Tmax (2.7 and 3.5 h, respectively) and lower Cmax

(13.9 and 6.6 µg/ml, respectively), which is typical for sustained-release formulations and

167167

CHAPTER 3

which is reflected in its higher HVDt50%Cmax (4.0 and 5.7 h, respectively) and RD (1.4 and 2.2,

respectively). Only the Cmax values of 5% XG (mean difference: 9.4 (95% C.I.: 1.9-16.8; P =

0.018)) and 5% PEO (mean difference: 7.4 (95% C.I.: 0.4-14.4; P = 0.039)) formulations

were significantly different from Slow-Lopresor® 200 Divitabs® at the 0.05 level of

significance. For all experimental formulations, Tmax and HVDt50%Cmax values showed no

significant difference in comparison with the sustained-release reference formulation (P >

0.05). Frel of the 5 and 10% XG formulations were significantly different (mean difference:

42.9 (95% C.I.: 3.6-82.3; P = 0.035)), whereas no significant difference between the PEO

formulations was seen (P > 0.05). Concerning the sustained-release effect, the RD values of

the experimental formulations were not significantly different from Slow-Lopresor® 200

Divitabs®. Based on the pharmacokinetic data (Table 5) Cmax and Tmax of 20% XG mini-

tablets (multiparticulate dosage form) and Slow-Lopresor® 200 Divitabs® (single-unit dosage

form) were similar which supported the hypothesis that 20% XG mini-matrices behaved in

vivo as a single-unit dosage form instead of multiparticulates. This was confirmed by the

swelling behaviour of the 20% XG mini-matrices during dissolution (using a larger sample

size of mini-matrices filled into a capsule, similar to the dosage form administered to the

dogs): due to the rapid swelling of XG, the mini-matrices stick together and the increased

diffusion path length from this plug reduced drug release, AUC0-36h and Frel [11] (Figure 11a

and 11b). In contrast, PEO mini-matrices did not swell during dissolution testing (Figure 11c

and 11d). Previous work [11] has also shown that ibuprofen mini-matrices with 20 and 30%

XG behaved similar in vivo, caused by their similar swelling behaviour upon immersion with

the dissolution medium. However, in this study a significant difference in pharmacokinetic

parameters and plasma concentration-time profiles between 10 and 20% XG formulations

was observed. This can be due to the fact that in this study a drug with a high water solubility

was used (in comparison with ibuprofen, which is a BSC class II drug) which is less affected

by the rapid swelling of XG.

168168


(a) (b)

(c) (d)

Figure 11: Swelling behaviour of mini-matrices after immersion in the dissolution medium: (a)

20% XG – 4 h, (b) 20% XG – 24 h, (c) 20% PEO 1.000.000 – 4 h, (d) 20% PEO 1.000.000 –

24 h.

169169

CHAPTER 3

4. CONCLUSIONS

This chapter showed that hot-melt extrusion can be used to produce mini-matrices

with controlled drug delivery: the presented ethylcellulose-polyethylene glycol/polyethylene

oxide mini-matrices allow to manufacture sustained-release dosage forms with a wide range

of drug release patterns. Importantly, not only the slope, but also the shape of the resulting

release curves can be adjusted by varying the polyethylene glycol/polyethylene oxide content

and molecular weight. All components remained stable under the processing conditions.

Neither drug and polymer crystallinity, nor drug homogeneity were influenced by the

concentration and type of hydrophilic polymer.

From the validation, it could be concluded that a specific, linear, precise and accurate

HPLC method was obtained. In vivo data showed that oral administration of ethylcellulose-

xanthan gum matrices was able to sustain metoprolol tartrate plasma levels in dogs. Xanthan

gum and polyethylene oxide behaved differently upon contact with the swelling medium.

In the next chapter, further work was conducted in order to determine the underlying

mass transport mechanism of drug and hydrophilic polymer during dissolution, and

experiments were performed to predict the effects of the mini-matrix geometry on the

resulting drug release kinetics and to evaluate the validity of these theoretical predictions.

170170


REFERENCES




Gemischen zur Herstellung von Arzneiformen, Spektrum der Wissenschaft, July, 18-

20, 1995.


Pharm. Biopharm. 54 (2002) 107-117.

[4] J.W. McGinity, J.J. Koleng, M.A. Repka, F. Zhang, Hot-melt extrusion technology, in: J.

Swarbrick and J.C. Boylan (Eds.), Encyclopedia of pharmaceutical technology, Vol. 19,







polymeric materials, J. Contr. Release 36 (1995) 243-250.




Pharm. 269 (2004) 509-522.



235-247.



Pharmacol. 57 (2004) 366-367.

171171

CHAPTER 3








1678-1685.

[13] E. Verhoeven, T.R.M. De Beer, G. Van den Mooter, J.P. Remon, C. Vervaet, Influence

of formulation and process parameters on the release characteristics of ethylcellulose

sustained-release mini-matrices produced by hot-melt extrusion, Eur. J. Pharm.

Biopharm. 69 (2008) 312-319.




[15] www.polymerlabs.com



[17] C.J. Kim, Drug release from compressed hydrophilic Polyox®-WSR tablets, J. Pharm.

Sci. 84 (1995) 303-306.

[18] C.J. Kim, Effects of drug solubility, drug loading, and polymer molecular weight on drug

release from Polyox® tablets, Drug Dev. Ind. Pharm. 24 (1998) 645-651.

[19] L. Maggi, R. Bruni, U. Conte, High molecular weight polyethylene oxides (PEOs) as an

alternative to HPMC in controlled release dosage forms, Int. J. Pharm. 195 (2000) 229-

238.

[20] L. Maggi, L. Segale, M.L. Torre, E. Ochoa Machiste, U. Conte, Dissolution behaviour of

hydrophilic matrix tablets containing two different polyethylene oxides (PEOs) for the

172172

http://www.polymerlabs.com/


controlled release of a water-soluble drug. Dimensionality study, Biomaterials 23

(2002) 1113-1119.

[21] N. Wu, L. Wang, D.C. Tan, S.M. Moochhala, Y. Yang, Mathematical modeling and in

vitro study of controlled drug release via a highly swellable and dissoluble polymer

matrix: polyethylene oxide with high molecular weights, J. Control. Release 102 (2005)

569-581.

[22] M. Efentakis, S. Politis, Comparative evaluation of various structures in polymer

controlled drug delivery systems and the effect of their morphology and characteristics

on drug release, Eur. Polym. J. 42 (2006) 1183-1195.

[23] Y. El-Malah, S. Nazzal, Hydrophilic matrices: application of Placket-Burman screening

design to model the effect of POLYOX-carbopol blends on drug release, Int. J. Pharm.

309 (2006) 163-170.

[24] S.I. Hong, S.Y. Oh, Dissolution kinetics and physical characterization of three-layered

tablet with poly(ethylene oxide) core matrix capped by Carbopol, Int. J. Pharm. 356

(2008) 121-129.

[25] A.G. Thombre, L.E. Appel, M.B. Chidlaw, P.D. Daugherity, F. Dumont, L.A.F. Evans,

S.C. Sutton, Osmotic drug delivery using swellable-core technology, J. Control.

Release 94 (2004) 75-89.

[26] H. Kojima, K. Yoshihara, T. Sawada, H. Kondo, K. Sako, Extended release of a large

amount of highly water-soluble diltiazem hydrochloride by utilizing counter polymer in

polyethylene oxides (PEO)/polyethylene glycol (PEG) matrix tablets, Eur. J. Pharm.

Biopharm., doi: 10.1016/j.ejpb.2008.05.032.

[27] V. Pillay, R. Fassihi, A novel approach for constant rate delivery of highly soluble

bioactives from a simple monolithic system, J. Control. Release 67 (2000) 67-78.

[28] A. Apicella, B. Cappello, M.A. Del Nobile, M.I. La Rotonda, G. Mensitieri, L. Nicolais,

Poly(ethylene oxide) (PEO) and different molecular weight PEO blends monolithic

devices for drug release, Biomaterials 14 (1993) 83-90.

173173

CHAPTER 3



[30] S. Prodduturi, R.V. Manek, W.M. Kolling, S.P. Stodghill, M.A. Repka, Solid-state

stability and characterization of hot-melt extruded poly(ethylene oxide) films, J. Pharm.

Sci. 94 (2005) 2232-2245.

[31] M.M. Crowley, A. Fredersdorf, B. Schroeder, S. Kucera, S. Prodduturi, M.A. Repka,

J.W. McGinity, The influence of guaifenesin and ketoprofen on the properties of hot-

melt extruded polyethylene oxide films, Eur. J. Pharm. Sci. 22 (2004) 409-418.



[33] L. Li, O. AbuBaker, Z.J. Shao, Characterization of poly(ethylene oxide) as a drug

carrier in hot-melt extrusion, Drug Dev. Ind. Pharm. 32 (2006) 991-1002.

[34] T. Ozeki, H. Yuasa, Y. Kanaya, Application of the solid dispersion method to the

controlled release of medicine. IX. Difference in the release of flurbiprofen from solid

dispersions with poly(ethylene oxide) and hydroxypropylcellulose and the interaction

between medicine and polymers, Int. J. Pharm. 155 (1997) 209-217.

[35] D.M. Schachter, J. Xiong, G.C. Tirol, Solid state NMR perspective of drug-polymer

solid solutions: a model system based on poly(ethylene oxide), Int. J. Pharm. 281

(2004) 89-101.

[36] P.K. Mididoddi, M.A. Repka, Characterization of hot-melt extruded drug delivery

systems for onychomycosis, Eur. J. Pharm. Biopharm. 66 (2007) 95-105.

[37] V.S. Tumuluri, M.S. Kemper, I.R. Lewis, S. Prodduturi, S. Majumdar, B.A. Avery, M.A.

Repka, Off-line and on-line measurements of drug-loaded hot-melt extruded films using

Raman spectroscopy, Int. J. Pharm. 357 (2008) 77-84.



174174


[39] S. Thumma, M.A. ElSohly, S. Zhang, W. Gul, M.A. Repka, Influence of plasticizers on

the stability and release of a prodrug of ∆9-tetrahydrocannabinol incorporated in poly

(ethylene oxide) matrices, Eur. J. Pharm. Biopharm., doi: 10.1016/j.ejpb.2008.06.009.



[41] S. Prodduturi, K.L. Urman, J.U. Otaigbe, M.A. Repka, Stabilization of hot-melt extrusion

formulations containing solid solutions using polymer blends, AAPS Pharm. Sci. Tech.

8 (2007) Article 50.

[42] E. Alexandre, Poly(ethylene oxide)-based hydrogels designed for artificial organs,

Polym. Mater. Sci. Eng. 89 (2003) 240-241.

[43] M.L.T. Zweers, G.H.M. Engbers, D.W. Grijpma, J. Feijen, In vitro degradation of

nanoparticles prepared from polymers bases on DL-lactide, glycolide and poly(ethylene

oxide), J. Control. Release 100 (2004) 347-356.

[44] S. Dhawan, M. Varma, V.R. Sinha, High molecular weight poly(ethylene oxide)-based

drug delivery systems. Part I. Hydrogels and hydrophilic matrix systems, Pharm. Tech.

29 (2005) 72-79.

[45] S. Dhawan, K. Dhawan, V.R. Sinha, Application of poly(ethylene oxide) in drug delivery

systems. Part II, Pharm. Tech. 29 (2005) 82-96.

[46] A. Kelner, E.H. Schacht, Tailor-made polymers for local drug delivery: release of

macromolecular model drugs from biodegradable hydrogels based on poly(ethylene

oxide), J. Control. Release 101 (2005) 13-20.






175175

CHAPTER 3

[49] J. Meier, E. Nuesch, R. Schmidt, Pharmacokinetic criteria for the evaluation of retard

formulations, Eur. J. Clin. Pharmacol. 7 (1974) 429-432.

[50] V.W. Steinijans, Pharmacokinetic characterization of controlled-release formulations,

Eur. J. Drug Metab. Pharmacokinet. 15 (1990) 173-181.

[51] Y. Miyagawa, T. Okabe, Y. Yamaguchi, M. Miyajima, H. Sunada, Controlled-release of

diclofenac sodium from wax matrix granule, Int. J. Pharm. 138 (1996) 215-254.

[52] H. Sato, Y. Miyagawa, T. Okabe, Dissolution mechanism of diclofenac sodium from

wax matrix granules, J. Pharm. Sci. 86 (1997) 929-934.




[54] V.P. Shah, K.K. Midha, S. Dighe, I.J. McGilveray, J.P. Skelly, A. Yacobi, T. Layloff,

C.T. Viswanathan, C.E. Cook, R.D. McDowall, K.A. Pittman, S. Spector, Analytical

methods validation: bioavailability, bioequivalence and pharmacokinetic studies, J.

Pharm. Sci. 81 (1992) 309-312.

[55] C.G. Regardh, K.O. Borg, R. Johansson, G. Johnsson, L. Palmer, Pharmacokinetic

studies on the selective β1-receptor antagonist metoprolol in man, J. Pharmacokinet.

Biopharm. 2 (1974) 347-364.

[56] G. Jobin, A. Cortot, J. Godbillon, M. Duval, J.P. Schoeller, J. Hirtz, J.J. Bernier,

Investigation of drug absorption from the gastrointestinal tract of man. I. Metoprolol in

the stomach, duodenum and jejunum, Brit. J. Clin. Pharmacol. 19 (1985) 97S-105S.

[57] N. Vidon, D. Evard, J. Godbillon, M. Rongier, M. Duval, J.P. Schoeller, J.J. Bernier, J.

Hirtz, Investigation of drug absorption from the gastrointestinal tract of man. II.

Metoprolol in the jejunum and ileum, Brit. J. Clin. Pharmacol. 19 (1985) 107S-112S.

[58] J. Godbillon, D. Evard, N. Vidon, M. Duval, J.P. Schoeller, J.J. Bernier, J. Hirtz,

Investigation of drug absorption from the gastrointestinal tract of man. III. Metoprolol in

the colon, Brit. J. Clin. Pharmacol. 19 (1985) 113S-118S.

176176


[59] B. Mistry, J. Leslie, N.E. Eddington, A sensitive assay of metoprolol and its major

metabolite α-hydroxy metoprolol in human plasma and determination of

dextromethorphan and its metabolite dextrorphan in urine with high performance liquid

chromatography and fluorometric detection, J. Pharm. Biomed. Anal. 16 (1998) 1041-

1049.

[60] M.J. Kendall, S.R. Maxwell, S.A. Westergren, Controlled release metoprolol. Clinical

pharmacokinetic and therapeutic implications, Clin. Pharmacokinet. 5 (1991) 319-330.

[61] R. Löbenberg, J.S. Kim, G.L. Amidon, Pharmacokinetics of an immediate release, a

controlled release and a two pulse dosage form in dogs, Eur. J. Pharm. Biopharm. 60

(2005) 17-23.

[62] J. Fang, H.A. Semple, J. Song, Determination of metoprolol, and its four metabolites in

dog plasma, J. Chromatogr. B 809 (2004) 9-14.

177177

178178178178

CHAPTER 4 MATHEMATICAL MODELLING

OF DRUG AND POLYMER RELEASE FROM HOT-MELT EXTRUDED MINI-MATRICES

179179179179

CHAPTER 4

180180180180

MATHEMATICAL MODELLING OF DRUG AND POLYMER RELEASE FROM HOT-MELT EXTRUDED MINI-MATRICES

CHAPTER 4 MATHEMATICAL MODELLING

OF DRUG AND POLYMER RELEASE FROM HOT-MELT EXTRUDED MINI-MATRICES

1. INTRODUCTION

Incorporation of a drug in a polymeric matrix is often used to sustain drug release. To

produce these sustained-release matrices, hot-melt extrusion is becoming a widely used

technology in the pharmaceutical industry using several thermoplastic polymers to control

drug release [1-7]. Cylindrical-shaped mini-matrices manufactured via hot-melt extrusion

provided zero-order release kinetics of freely water soluble drugs using ethylcellulose as

matrix former and hydroxypropylmethylcellulose [8-10], xanthan gum [8-13, Chapter 1,

Chapter 2] or polyethylene glycol/polyethylene oxide [13, Chapter 3] as hydrophilic polymers

to modify the release rate. Increasing xanthan gum concentrations caused swelling of the

mini-matrices (creating pores in the matrix through which the drug was released), resulting in

accelerated drug release [11, 12]. In the previous chapter, it has been shown that a wide

range of metoprolol tartrate release patterns can be obtained by adding different amounts

and types (molecular weight) of polyethylene glycol/polyethylene oxide to the formulation

[13].

The present study evaluates the release mechanism of a model drug (metoprolol

tartrate) and the hydrophilic polymer (polyethylene glycol/polyethylene oxide) from hot-melt

extruded mini-matrices (3 mm diameter, 2 mm height) formulated with polyethylene

glycol/polyethylene oxide and ethylcellulose. The drug release kinetics from these diffusion-

controlled mini-matrices was quantitatively predicted in function of the polyethylene

glycol/polyethylene oxide concentration (0-70%, w/w) and grade (polyethylene glycol 6000

and polyethylene oxide 100.000-7.000.000) and in function of the device geometry (diameter

varied between 1 and 12 mm for constant height, height varied between 1 and 32 mm for

181181181181

CHAPTER 4

constant diameter). The validity of these theoretical predictions was evaluated by

independent experiments. In addition, a comprehensive mathematical theory was developed

allowing a quantitative description of drug and polyethylene glycol/polyethylene oxide

transport in ethylcellulose-based mini-matrices considering time-dependent diffusivities.

182182182182



2.1. Materials

Metoprolol tartrate (MPT) (10 µm) (Esteve Quimica, Barcelona, Spain) was selected

as model drug. The matrix consisted of ethylcellulose (EC) (Ethocel® Std 10 FP Premium;

particle size: 3-15 µm; ethoxyl content: 48-49.5% (w/w); viscosity: 9-11 mPa.s (5% solution in


a hydrophilic component: polyethylene glycol (PEG) (PEG 6000, Fagron, Waregem,

Belgium) or polyethylene oxide (PEO) (SentryTM PolyoxTM WSR N10, N12K and 303,

molecular weight (MW): 100.000, 1.000.000 and 7.000.000, respectively) (Dow Chemical

Company, Midland, USA). Dibutyl sebacate (DBS) (Sigma-Aldrich, Steinheim, Germany) was

used as plasticizer for ethylcellulose. All other chemicals were of analytical grade (VWR,

Leuven, Belgium).

2.2. Methods

2.2.1. Preparation of mini-matrices via hot-melt extrusion

The composition of the investigated mini-matrices is shown in Table 1. For all

formulations the MPT content was kept constant at 30% (w/w), while the concentration of

hydrophilic polymer (PEG or PEO) was varied between 0 and 70% (w/w). The remaining part

of the formulation consisted of EC/DBS, blended at a ratio of 2/1 (w/w) [12].

Table 1: Composition (%, w/w) of the hot-melt extruded mini-matrices.

MPT 30 30 30 30 30 30 30

PEG/PEO - 1 2.5 5 10 20 70

EC 46.7 46 45 43.3 40 33.3 -

DBS 23.3 23 22.5 21.7 20 16.7 -

The components were blended in a planetary mixer (15 min, 90 rpm) (Kenwood Major


interaction between ethylcellulose and the plasticizer. The mixture was passed through the screws

of the powder feeder of the extruder at room temperature and recycled into the powder reservoir

183183183183

CHAPTER 4

in order to homogeneously distribute DBS in the powder prior to hot-melt extrusion. Hot-melt

extrusion was performed using a laboratory-scale intermeshing co-rotating twin-screw extruder

(MP19TC-25, APV Baker, Newcastle-under-Lyme, UK) with a length-to-diameter ratio of 25/1.

The machine was equipped with a Brabender twin-screw powder feeder, a screw with two mixing

sections and a densification zone [11-13]. The die block (2.6 cm thickness) was fixed to the

extruder barrel and the axially mounted die plate (1.9 cm thickness, with a cylindrical hole of 3 mm

diameter for shaping the extrudates) was attached to the die block. The following extrusion

conditions were used: a screw speed of 30 rpm, a powder feed rate of 6 g/min and a temperature

of 40 and 70°C for the five heating zones along the barrel for extrusion of PEG and PEO

formulations, respectively. After cooling down to room temperature, the extruded rods were

manually cut, using surgical blades, into mini-matrices of the desired length. Dies with different

dimensions (varying diameters between 1 and 12 mm) were used in order to produce extrudates

with different diameters.

2.2.2. In vitro drug and PEG/PEO release

Drug release from the mini-matrices (approximately 60 mg) was determined using the

USP apparatus 1 (rotating baskets, 100 rpm, 4 mini-matrices per vessel) (VK 7010,

combined with a VK 8000 automatic sampling station, VanKel, New Jersey, USA).

Demineralized water (900 ml, 37 ± 0.5°C) was used as release medium. At predetermined

time points, 5 ml samples were withdrawn and analyzed for the drug content by UV

spectroscopy (λ = 222 nm, Lambda 12 UV-VIS double beam spectrophotometer, Perkin

Elmer, Zaventem, Belgium). Each experiment was conducted 6 times.

PEG/PEO release from the mini-matrices was determined using the USP paddle

method under the same conditions as described above. PEG/PEO concentrations in the

samples were measured by gel filtration chromatography (GFC) (detection via the refractive

index) [13, Chapter 3]. Due to the sensitivity of the GFC analysis, the sample size was

increased (36, 9, 5 and 2 g mini-matrices for the 2.5, 10, 20 and 70% PEG/PEO

formulations, respectively) and the paddle method was used. All experiments were

conducted twice.

In case drug or PEG/PEO release levelled off below 100%, the experimentally

determined plateau value was considered as the 100% reference value for the modelling of

drug and PEG/PEO release (= amount of mobile species).

184184184184


2.2.3. Porosity of the mini-matrices

To determine the changes in the porosity of the mini-matrices upon exposure to the

release medium, 40 mini-matrices were subjected to the same conditions as described

above for the in vitro release studies. At predetermined time points, the mini-matrices were

removed, frozen using liquid nitrogen and freeze-dried (Amsco-Finn Aqua GT4 freeze-dryer,

Amsco, Darmstadt, Germany) during 25 h (freezing to -45°C within 175 min at 1000 mbar,

primary drying was performed at -25°C and at a pressure varying between 0.8 and 1 mbar

during 13 h, followed by secondary drying at elevated temperature (10°C) and reduced

pressure (0.1-0.2 mbar) for 7 h). The samples were only introduced in the freeze-drier when

the shelf temperature was below -40°C. After freeze-drying, the mini-matrices were stored

over silica until true volume measurements with a helium pycnometer (AccuPyc 1330,

Micromeritics, Norcross, USA). The porosity of the mini-matrices was calculated based on

these values and on the mini-matrices’ apparent volume, which was determined using a

digital calliper (Bodson, Luik, Belgium).

2.2.4. X-ray tomography

The mini-matrices (before exposure to the release medium, and after 24 h exposure to the

release medium) were scanned with the in-house developed high resolution micro-CT scanner of

the Ghent University Centre for X-ray Tomography (www.UGCT.ugent.be). The system is

composed of a Feinfocus X-ray tube with transmission target and beryllium exit window, a Micos

high precision air-bearing rotation stage and a Varian Paxscan 2520 a-Si flat panel detector. The

tube was operated at 60 kV tube voltage in micro-focus mode at 2.8 W which results in a 3 µm

spot size. During the scans the sample was rotated over 360° in 0.45° steps where four shadow

images were recorded in every step. The detector consists of a CsI screen with 2000 x 1600

pixels of 127 µm and it was operated in non-binning mode with an exposure time of 2 s. The

source-to-object distance was 20 mm and the source-to-detector distance was 860 mm in order to

have a factor 42 magnification with 3 µm resolution.

The 800 shadow images were processed and reconstructed to 1500 cross-sectional

images of 1100 x 1100 pixels with 3 µm pixel size with Octopus software. From each mini-matrix a

central region of 300 slices was used for statistical analysis.

To process the images two selections were made: one with the material of the mini-matrix

and another one with the pores. This was achieved by segmentation based on the greyscale, and

subsequent some filtering to remove the noise (these results should be sceptically analyzed as

limiting factors such as noise and resolution make the obtaining of these results slightly operator

185185185185

http://www.ugct.ugent.be/

CHAPTER 4

dependent). The voids were subsequently separated and classified for equivalent diameters.

Separating the pores makes it possible to have a better analysis of the porosity but it removes the

pore interconnectivity, however the pores are mostly connected through ducts with a diameter of

about 1 µm and in order to analyze this the resolution of the scan is too low. Equivalent diameters

means that for each pore the volume was calculated in µm3 and that volume is then considered as

a perfect sphere, the diameter of that sphere is the equivalent diameter.

The set of 300 cross-sections that were statistically analyzed were subsequently loaded in

a 3D rendering software (VGstudio). In this way, it is made possible to represent the porosity and

the size distribution of the pores.

2.2.5. Raman spectroscopy

MPT distribution in the mini-matrices, before exposure to the dissolution medium and

24 h after exposure to the release medium, was evaluated by Raman spectroscopic

mapping. Prior to Raman analysis, the tablets obtained after 24 h of dissolution were frozen

using liquid nitrogen and freeze-dried over 25 h, as described above to determine the

porosity. After freeze-drying the samples were cut along their radial (Figure 1a) and axial axis

(Figure 1b). At the exterior surface as well as at the radial cutting surface three areas (1000 x

600 µm2, Figure 1a) (n = 3) were scanned using a 10x long working distance objective lens

(laser spot size: 50 µm) in point-by-point mapping mode with a step size of 50 µm in both the

x and y directions. In addition, 24 h after exposure to the release medium, 6 areas (1000 x

600 µm2, Figure 1b) (n = 3) of the axial cutting surface were mapped via Raman

spectroscopy.

The resulting map provides an overview of the MPT distribution in the mapped area.

The mapping system used for this study was a RamanRxn 1 Analyzer and Microprobe

(Kaiser Optical Systems, Ann Arbor, USA), equipped with an air cooled CCD detector (back-







186186186186


tabl

et h

eigh

t: 2

mm

tabl

et h

eigh

t: 2

mm

(a) (b)

Figure 1: Schematic drawing of the areas (presented as blue rectangles) mapped with

Raman spectroscopy: (a) surface and half height (radial cutting surface) are mapped along

the diameter of the mini-matrix, (b) 6 areas of the axial cutting surface are mapped.

tablet diameter: 3 mm

½ height

surface1000 x 600 µm2

tablet diameter: 3 mm

½ height

surface

1000

x 6

00 µ

m 2

187187187187

CHAPTER 4


Previous investigation in Chapter 3 showed that the release rates of MPT (used as

model drug) and PEG/PEO from hot-melt extruded mini-matrices strongly depended on the

concentration as well as on the type of PEG/PEO incorporated in the EC-based cylinders

[13]. The aim of this study was to better understand these phenomena and to be able to

quantitatively predict the effects of the device design and changes in the composition on the

resulting drug release kinetics.

3.1. “Simple” mathematical theory

As diffusion is likely to play a major role in the underlying drug and PEG/PEO release

mechanisms, Fick’s second law was used as a basis for the mathematical analysis. As the

investigated mini-matrices are cylindrical in shape, it is important to consider drug and

PEG/PEO transport in radial as well as in axial direction [14]:

⎭⎬⎫

⎩⎨⎧

⎟⎠⎞

⎜⎝⎛

∂∂

∂∂

+⎟⎠⎞

⎜⎝⎛

∂∂

∂∂

+⎟⎠⎞

⎜⎝⎛

∂∂

∂∂

=∂∂

zcrD

zθc

rD

θrcrD

rr1

tc k

kkkk

kk (1)

Where, ck and Dk are the concentration and diffusion coefficient of the diffusing species (k =

1: drug; k = 2: PEG/PEO), respectively; t represents time; and r, z and θ denote the radial,

axial and angular coordinate, respectively. This partial differential equation was solved

considering the following initial and boundary conditions:

• drug and PEG/PEO are initially homogeneously distributed throughout the mini-

matrices (before exposure to the release medium)

• perfect sink conditions are maintained through the experiments for all diffusing

species

• the cylinders do not significantly swell or shrink during drug release

• the diffusion coefficients of MPT and PEG/PEO are constant during the entire drug

release period

leading to:

( )( )

⎟⎟⎠

⎞⎜⎜⎝

⎛⋅⋅

⋅+−⋅

+⋅⎟⎟⎠

⎞⎜⎜⎝

⎛⋅⋅−⋅⋅−= ∑∑

∞

=

∞

=∞

tDH

π1p2exp1p2

1tDRq

expq1

π321

MM

k2

22

0p2k2

2n

1n2n

2,k

t,k (2)

188188188188


Where Mk,t and Mk,∞ are the amounts of drug or PEG/PEO released at time t, and at infinite

time (k = 1: drug; k = 2: PEG/PEO); qn are the roots of the Bessel function of the first kind of

order zero [J0(qn) = 0]; R and H represent the radius and length of the cylinder; Dk denotes

the time-independent diffusion coefficient of the drug or PEG/PEO.

Equation 2 was fitted to the various experimentally measured drug and PEG/PEO

release profiles from EC-based mini-matrices containing 0-70% (w/w) PEG/PEO of different

molecular weight: 6000-7.000.000 Da. Examples for such fittings are shown in Figure 2.

189189189189

CHAPTER 4

(a)

0

25

50

75

100

0 6 12 18 24

Time (h)

MPT

rele

ased

(%)

experiment

theory

(b)

0

25

50

75

100

0 6 12 18 24

Time (h)

PEO

rele

ased

(%)

experiment

theory

Figure 2: Experiment (symbols) and theory (curves, simple model with constant diffusion

coefficients, equation 2): (a) MPT release from mini-matrices containing 20% PEO

1.000.000, (b) PEO release from mini-matrices containing 10% PEO 100.000.

190190190190


As it can be seen, good agreement between theory and experiment was obtained in

the case of MPT and PEO release from cylinders initially containing 20% PEO 1.000.000 and

10% PEO 100.000, respectively. This indicates that the mass transport of both species is

likely to be dominated by diffusion with constant diffusion coefficients. Based on these

fittings, the apparent diffusivities of MPT and PEO could be determined to be equal to 1.56 .

10-7 (± 0.80) cm²/s and 6.74 . 10-8 (± 0.74) cm²/s, respectively. The higher diffusivity of the

drug compared to the PEO can be explained by the much smaller molecular weight.

However, also significant deviations were observed between the theoretically calculated MPT

or PEG/PEO release profile and the experimentally measured values. Figure 3 shows two

examples: drug and PEO release from mini-matrices initially free of PEG/PEO (i.e. containing

only 30% MPT and 70% EC) and containing 10% PEO 7.000.000, respectively. Thus, in

these cases the underlying mass transport mechanisms are more complex.

191191191191

CHAPTER 4

(a)

0

25

50

75

100

0 24 48

Time (h)

MPT

rele

ased

(%

72

)

experiment

theory

(b)

0

25

50

75

100

0 6 12 18 24

Time (h)

PEO

rele

ased

(%)

experiment

theory

Figure 3: Experiment (symbols) and theory (curves, simple model with constant diffusion

coefficients, equation 2): (a) MPT release from mini-matrices free of PEG/PEO, (b) PEO

release from mini-matrices containing 10% PEO 7.000.000.

192192192192


Table 2 gives an overview on the various systems which were studied and whether or

not the “simple” mathematical model considering constant diffusion coefficients (equation 2)

was applicable or not (either D values are indicated, or “non-constant” marked). Importantly,

there is a clear tendency: drug and PEG/PEO release follows pure diffusion control with

constant diffusion coefficients: (i) at high initial PEG/PEO loadings (≥ 20%), irrespective of

the molecular weight of PEG/PEO, and (ii) at low/intermediate PEG/PEO loadings (1-10%) in

the case of low/intermediate molecular weight PEG/PEOs (MW ≤ 100.000 Da). In the first

case, the mini-matrices become highly porous upon PEG/PEO leaching. In the second case

the PEG/PEO molecules are rather small, thus, time-dependent changes of the matrix

porosity during drug release can be expected to be only of minor importance for drug and

PEG/PEO mobility. In contrast, at low initial PEG/PEO loadings the mobility of large

molecules, such as high molecular weight PEG/PEOs is likely to be affected by changes in

the matrix porosity: with increasing pore size the mobility of these species significantly

increases. This hypothesis is in good agreement with the type of systematic deviations

between the fitted theory (“simple” model) and the experimentally determined amounts of

released MPT and PEG/PEO (Figure 3): at early time points diffusional mass transport is

theoretically overestimated, whereas at late time points it is underestimated (higher mobility

in reality).

An increase in the PEG/PEO content (irrespective of the PEG/PEO molecular weight)

resulted in increased drug and polymer diffusion coefficients, which is in good agreement

with the experimentally determined increasing drug and PEG/PEO release rates (Chapter 3).

This can be explained by the increased porosity of the cylinders upon drug and PEG/PEO

leaching (resulting in increased drug and polymer mobility) and to an increased hydrophilicity

of the systems (resulting in faster water penetration rates during dissolution). Furthermore,

the apparent diffusion coefficients of the drug and PEG/PEO generally increase with

decreasing PEG/PEO molecular weight in the systems (irrespective of the PEG/PEO

content), which can be described using equation 2 (Table 2). With decreasing molecular

weight, the mobility of PEG/PEO increases (due to the reduced size of the molecules and

lower viscosity of the solution), resulting in facilitated leaching into the bulk fluid and, thus,

increased system porosity and increased drug mobility.

193193193193

CHAPTER 4

Table 2: Effects of the initial PEG/PEO concentrations and molecular weights (MW) on the

apparent diffusion coefficient of metoprolol tartrate (Ddrug) and PEG/PEO (DPEG/PEO) in the

investigated mini-matrices (mean ± S.D., n = 6).

PEG/PEO concentration

(%, w/w)

MW PEG/PEO

(Da)

Ddrug

(cm2/s)

DPEG/PEO

(cm2/s)

0 - non-constant -

1 6000 - -

100.000 - -

1.000.000 - -

7.000.000 non-constant -

2.5 6000 - 4.21 . 10-8 ± 0.37

100.000 1.92 . 10-8 ± 0.01 non-constant

1.000.000 - non-constant

7.000.000 non-constant non-constant

5 6000 2.29 . 10-8 ± 0.27 -

100.000 2.22 . 10-8 ± 0.17 -



10 6000 2.60 . 10-8 ± 0.19 1.21 . 10-7 ± 0.02

100.000 2.50 . 10-8 ± 0.19 6.74 . 10-8 ± 0.74

1.000.000 non-constant 6.16 . 10-8 ± 1.51

7.000.000 non-constant non-constant

20 100.000 4.47 . 10-7 ± 0.86 1.36 . 10-7 ± 0.15

1.000.000 1.56 . 10-7 ± 0.80 7.12 . 10-8 ± 1.30

7.000.000 2.02 . 10-7 ± 0.15 7.32 . 10-8 ± 0.10

70 6000 1.13 . 10-5 ± 0.38 2.68 . 10-6 ± 0.62

100.000 1.38 . 10-6 ± 0.55 1.45 . 10-6 ± 0.60

1.000.000 2.98 . 10-7 ± 0.34 1.11 . 10-7 ± 0.15

7.000.000 2.36 . 10-7 ± 0.40 9.72 . 10-8 ± 0.99

-: not determined

194194194194


In addition to this deeper insight into the underlying mass transport mechanisms, the

mathematical modelling of the mini-matrices, which can be described by equation 2 (“simple”

theory) also offers an interesting practical benefit: the effects of the initial radius and height of

the cylinders on the resulting drug and PEG/PEO release kinetics can be predicted in a

quantitative way. Figure 4 shows some examples for such predictions. Knowing the diffusion

coefficients of MPT and PEG/PEO from the fittings of the “simple” model to the release

kinetics determined with mini-matrices initially containing 10% PEO 100.000 or 20% PEO

1.000.000 and with a height of 2 mm and a radius of 1.5 mm, drug and PEG/PEO leaching

into the release medium was predicted from cylinders with the same height, but varying

thickness: 0.5, 1.0, 3.0 and 6.0 mm, respectively (curves in Figure 4a and 4b). As it can be

seen, a significant decrease in the mass transport rates was predicted with increasing radius

of the cylinders. This is due to the decrease in the relative surface area available for diffusion

(decreasing “surface area/volume” ratio). Importantly, these theoretical predictions could be

confirmed by independent experimental MPT and PEG/PEO leaching measurements

(symbols in Figure 4a and 4b). For extrudates with a radius of 0.5 mm, there was an

important deviation between the experimental data and theoretical predictions. However, if

predictions were made taking die swelling [15, 16] of these matrices into account (the actual

radius of these mini-matrices was 0.7 mm) the correlation between theory and experiment

was good. Die swelling was less pronounced using other die geometries and die swelling

was not taken into account for predicting drug release. This clearly indicates that drug and

PEG/PEO release from these hot-melt extruded mini-matrices is predominantly controlled by

pure diffusion with constant diffusion coefficients. The theoretically predicted effects of the

height of mini-matrices initially containing 10% PEO 100.000 or 20% PEO 1.000.000 on the

resulting drug release kinetics from cylinders with a constant radius of 1.5 mm are illustrated

in Figure 4c and 4d. Clearly, also the variation of the cylinder height from 1 to 32 mm strongly

affects the resulting drug release patterns. Again, an increase in the relative surface area

available for diffusion (due to a decreasing matrix height) led to increased MPT release rates.

Compared to the effects of the cylinder radius, the consequences of the relative changes in

the matrix height are less pronounced. This can be attributed to the fact that the surface area

of a cylinder depends on the radius to the power of two and to the height to the power of one

only. Importantly, also in these cases, independent MPT and PEG/PEO release

measurements confirmed the theoretical predictions (symbols in Figure 4c and 4d). The

observed impact of the device geometry on the resulting drug release kinetics is in good

agreement with reports in the literature, including swellable tablets [17-20].

195195195195

CHAPTER 4

(a)

(b)

0

25

50

75

100

0 4 8

Time (h)

MPT

rele

ased

(%

12

) theory, R = 0.5 mm

theory, R = 1.0 mm

theory, R = 1.5 mm

theory, R = 3.0 mm

theory, R = 6.0 mm

experiment, R = 0.5 mm


experiment, R = 1.5 mm, reference



0

25

50

75

100

0 50 100 150 200

Time (h)

MPT

rele

ased

(%)

theory, R = 0.7 mm

theory, R = 1.0 mm

theory, R = 1.5 mm

theory, R = 3.0 mm

theory, R = 6.0 mm



experiment, R = 1.5 mm, reference



196196196196


(c)

0

25

50

75

100

0 50 100 150 200

Time (h)

MPT

rele

ased

(%) theory, H = 1 mm

theory, H = 2 mm

theory, H = 4 mm

theory, H = 32 mm

experiment, H = 1 mm

experiment, H = 2 mm, reference



(d)

0

25

50

75

100

0 4 8

Time (h)

MPT

rele

ased

(%

12

) theory, H = 1 mm

theory, H = 2 mm

theory, H = 4 mm

theory, H = 32 mm


experiment, H = 2 mm, reference



Figure 4: Theoretical predictions and independent experimental verification (except for the

reference formulations) of the effects of the cylinders’ radii and heights (indicated in the

figures) on the resulting MPT release kinetics from mini-matrices containing (a) 10% PEO

100.000 (constant height = 2 mm), (b) 20% PEO 1.000.000 (constant height = 2 mm), (c)

10% PEO 100.000 (constant radius = 1.5 mm), (d) 20% PEO 1.000.000 (constant radius =

1.5 mm).

197197197197

CHAPTER 4

3.2. “Complex” mathematical theory

To better understand the observed significant deviations between the “simple”

mathematical theory and the experimentally determined drug and PEG/PEO release kinetics

from mini-matrices with low/intermediate initial PEG/PEO loadings (1-10%) and high

molecular weight PEG/PEOs (MW ≥ 1.000.000 Da), the porosity of the mini-matrices was

determined before and upon exposure to the release medium using X-ray tomography as

well as helium pycnometry. As it can be seen in Figure 5a and 5b, the matrices showed

already an initial porosity, which was higher at the edges (Figure 5b: significant number of

smaller pores at the edges, whereas a few larger pores are located at the centre).

Importantly, this porosity significantly increased upon exposure to the release medium, at

least partially due to drug and PEG/PEO leaching into the bulk fluid (Figure 5c). After 24 h

exposure, a highly porous matrix was observed, which partially also showed crack formation

(Figure 5d and 5e).

198198198198


(a) (b)

(c)

0

10

20

30

40

50

60

0 6 12 18 24

Time (h)

Poro

sity

(%)

(d) (e)

Figure 5: Porosity of mini-matrices containing 10% PEO 7.000.000 before and after exposure

to the release medium: (a) cross-section at t = 0 h (before exposure), (b) visualization of the

pore distribution and sizes at t = 0 h (before exposure), (c) experimentally determined, time-

dependent increase in porosity, (d) cross-section after 24 h exposure, (e) visualization of the

pore distribution and sizes after 24 h exposure.

199199199199

CHAPTER 4

This significant increase in the matrix porosity is not taken into account by the “simple”

mathematical theory. Importantly, the time-dependent change in system porosity is likely to

result in significant changes in the mobility of the drug and PEG/PEO, especially at

low/intermediate PEG/PEO loadings. The following, more complex mathematical theory

takes such effects into account. It is also based on Fick’s law of diffusion for cylindrical

geometry (equation 1) and considers radial and axial mass transfer. The underlying model

assumptions are as follows:

• drug and PEG/PEO are initially homogeneously distributed throughout the mini-

matrices (before exposure to the release medium)

• perfect sink conditions are maintained through the experiments for all diffusing

species

• the cylinders do not significantly swell or shrink during drug release

As time-dependent drug and PEG/PEO diffusion coefficients are taken into account, there is

no analytical solution for this initial value problem. Thus, a numerical method, based on finite

differences was applied. The major principles of this theory are briefly described in the

following, the reader is referred to the literature for more details [21, 22]. As there is no

concentration gradient of any component with respect to the angle θ (Figure 6), equation 1

can be transformed into:

⎟⎠⎞

⎜⎝⎛

∂∂

∂∂

+∂∂

+⎟⎠⎞

⎜⎝⎛

∂∂

∂∂

=∂∂

zcD

zrc

rD

rcD

rtc k

kkkk

kk (3)

To minimize the computation time, also the symmetry plane at z = 0 (Figure 6) is taken into

account. Thus, it is sufficient to describe the changes in the MPT and PEG/PEO

concentrations within the 2-dimensional rectangle highlighted in Figure 6 to be able to

calculate the mass transport phenomena in the entire cylinder: upon rotation around the z-

axis the upper half of the mini-matrix is described, and due to the symmetry at the z = 0

plane, the whole cylinder can be considered.

200200200200


z

θ

R

Figure 6: Schematic presentation of a cylinder mini-matrix for the mathematical analysis. Two

symmetry planes are illustrated: for r = 0 and z = 0 as well as the rotational symmetry of the

system with respect to the angle θ. Furthermore, the locations of the two cross-sections

referred to in Figure 9 are illustrated.

The initial condition of this theory takes into account the fact that MPT as well as PEG/PEO

are uniformly distributed throughout the mini-matrices before exposure to the release

medium (at t = 0). Thus, the drug and PEG/PEO concentrations at any position are equal to

the respective initial concentrations (c1 initial, c2 initial):

t = 0 c1 = c1 initial 0 ≤ r ≤ R 0 ≤ z ≤ Z (4)

t = 0 c2 = c2 initial 0 ≤ r ≤ R 0 ≤ z ≤ Z (5)

where R denotes the radius and Z the half height of the mini-matrix. As perfect sink

conditions are provided for all diffusing species throughout the experiments, the MPT and

r

Z z=0

201201201201

CHAPTER 4

PEG/PEO concentrations at the surface of the cylinders are always equal to zero upon

exposure to the release medium:

t > 0 c1 = 0 0 ≤ r ≤ R z = Z (6)

t > 0 c1 = 0 0 ≤ z ≤ Z r = R (7)

t > 0 c2 = 0 0 ≤ r ≤ R z = Z (8)

t > 0 c2 = 0 0 ≤ z ≤ Z r = R (9)

Due to the symmetries at z = 0 and r = 0 (Figure 6), there are no concentration gradients at z = 0

for 0 r ≤ R and at r = 0 for 0 ≤ z ≤ Z for any of the diffusing components: ≤

t > 0 0zc1 =∂∂

0 ≤ r ≤ R z = 0 (10)

t > 0 0rc1 =∂∂

0 ≤ z ≤ Z r = 0 (11)

t > 0 0zc2 =∂∂

0 ≤ r ≤ R z = 0 (12)

t > 0 0r

c2 =∂∂

0 ≤ z ≤ Z r = 0 (13)

Furthermore, the time-dependent porosity of the matrices is taken into account (Figure 5).

The increase in MPT and PEG/PEO mobility within the mini-matrices is considered as follows

as a function of the system porosity:

100)t(εD

)t(D critkk

⋅= (14)

where Dk crit represents a critical diffusion coefficient (as diffusion takes place in confined

regions, this value is not necessarily equal to the diffusivity in pure release medium), being

characteristic for the diffusing species, and ε denotes the porosity of the cylinder in percent.

Thus, the consequences of the dynamic changes in the structure of the mini-matrices during

drug release on the conditions for the multi-component diffusion are quantitatively taken into

account.

As the matrix porosity increases upon MPT and PEG/PEO leaching, there is no

analytical solution for the described set of partial differential equations (equation 3-14). To be

able to calculate the changes in the MPT and PEG/PEO concentrations within the 2-

202202202202


dimensional rectangle highlighted in Figure 6, the radius of the mini-matrix, R, and half

height, Z, are divided into I and J space intervals, ∆r and ∆z, respectively (Figure 7).

r

z

1 2 i-1 i i+1 I-2 I-1 I

12

j-1j

j+1

J-2J-1

J

00

t

t=t0+Δt

t=t0

b

Figure 7: Schematic presentation of the basis principle of the applied numerical analysis for

the complex mathematical model considering time-dependent diffusion coefficients.

This generates a grid of (I+1)x(J+1) grid points. The time is divided into g time intervals ∆t

(for most of the simulations I = J = 50 and g = 500.000 were chosen). At t = 0 (before

exposure to the release medium), the drug and PEG/PEO concentrations are known from the

initial conditions (equation 4, 5). Importantly, using equation 3 and 6-14, the concentration

profiles of the diffusing species at a new time step (t = t0+∆t) can be calculated, when the

concentration profiles are known at the previous time step (t = t0): the concentration of a

particular species at a certain inner grid point (ix∆r, jx∆z) at the new time step (t = t0+∆t) is

calculated from its concentrations at the same grid point (ix∆r, jx∆z) and the four direct

neighbours [(i-1)x∆r, jx∆z; ix∆r, (j-1)x∆z; ix∆r, (j+1)x∆z; (i+1)x∆r, jx∆z] at the previous time

step (t = t0). The respective concentrations at the outer grid points (i = 0 v i = I v j = 0 v j = J)

at a new time step (t = t0+∆t) are calculated using the boundary conditions (equation 6-13).

203203203203

CHAPTER 4

As all initial concentrations are known, the concentration profiles at t = 0+∆t, t = 0+2∆t, t =

0+3∆t,..., t = 0+g∆t can be calculated sequentially. Furthermore, the total amounts of MPT

and PEG/PEO within the cylinders are calculated at each time step (by integrating the

respective concentrations with respect to r, z and θ). Based on experimentally determined

porosity of the mini-matrices, the drug and PEG/PEO diffusivities can be calculated at any

time point using equation 14. For the implementation of the mathematical model the

programming language C++ was used (Borland C++ Builder V.6.0).

As it can be seen in Figure 8, much better agreement between the complex theory

and the experimentally measured MPT and PEG/PEO release kinetics from the mini-

matrices was obtained in the case of drug release from mini-matrices initially containing 10%

PEO 7.000.000 as well as in the case of PEO release from mini-matrices initially containing

10% PEO 7.000.000 or 2.5% PEO 100.000 than with the simple mathematical model. This is

due to the more realistic description of the diffusional mass transport: due to the increase in

device porosity, drug and PEG/PEO release are facilitated. Importantly, in the case of MPT

release from mini-matrices initially containing 2.5% PEO 100.000 the agreements between

the complex and the simple mathematical model with the experimentally measured drug

release kinetics were similar. This confirms the above described hypothesis that in these

cases, the changes in the conditions for drug and PEG/PEO release are not of significance

and that time-independent diffusion coefficients can be considered. Figure 8 only shows

some examples, the before mentioned is true for all the investigated mini-matrices.

It has to be pointed out that the agreement between theory and experiment is not yet

perfect, even in the case of the complex mathematical theory. For example, in Figure 8a and

8c still some systematic (but less important) deviations can be seen. This might at least

partially be attributed to the fact that crack formation sets on in these systems after a certain

lag-time, as it can be seen in Figure 5d. This leads to increased surface areas available for

diffusion at late time points and, thus, increased drug and PEG/PEO release rates. The

proposed mathematical theories do not take this dynamic process of crack formation into

account. Thus, the models (slightly) underestimate drug release at late time points. The

observed overestimation of drug release at early time points is a consequence of the fitting

procedure. It was beyond the scope of this study to include the onset and degree of crack

formation in the mathematical theories. This would require the knowledge of additional,

system-specific parameters, which are difficult to determine.

204204204204


(a)

0

25

50

75

100

0 4 8 12

Time (h)

MPT

rele

ased

(%

16

)

experiment

simple model

complex model

(b)

0

25

50

75

100

0 20 40 60 8

Time (h)

MPT

rele

ased

(%

0

)

experiment

simple model

complex model

205205205205

CHAPTER 4

(c)

0

25

50

75

100

0 4 8 12

Time (h)

PEO

rele

ased

(%

16

)

experiment

simple model

complex model

(d)

0

25

50

75

100

0 10 20

Time (h)

PEO

rele

ased

(%

30

)

experiment

simple model

complex model

Figure 8: Experiment (symbols) and theory (dashed curves: simple model with constant

diffusion coefficients, solid curves: complex model with time-dependent diffusion

coefficients): (a) MPT release from mini-matrices containing 10% PEO 7.000.000, (b) MPT

release from mini-matrices containing 2.5% PEO 100.000, (c) PEO release from mini-

matrices containing 10% PEO 7.000.000, (d) PEO release from mini-matrices containing

2.5% PEO 100.000.

206206206206


The distribution of MPT in the mini-matrices was monitored via Raman using the 627-

653 cm-1 spectral band as no overlap with Raman signals from the other tablet ingredients

occurred at this spectral range (see Chapter 3, Figure 6). Based on the intensity of the

Raman signal across the scanned sections of the mini-matrices it was confirmed that MPT is

homogeneously distributed in the mini-matrix before exposure to the release medium, since

the MPT intensities were similar in the centre and at the edges of the mini-matrices, for z = Z

(exterior surface) and z = 0 (half height). After exposure to the release medium (24 h),

surprisingly a lower drug concentration was only detected at the tablet edges, whereas even

at the exterior surface the centre still contained high MPT concentrations (Figure 9a).

However, mapping of the axial cutting surface (Figure 9b) showed that MPT preferentially

diffused in axial direction during dissolution. This can be explained by the fact that during hot-

melt extrusion the polymers are orientated parallel to the extrusion flow, resulting in a

preferential diffusion route for drug release. This observation is in agreement with the

anisotropic swelling due to axial relaxation of the polymers (Chapter 1). In addition, a drug

particle will select the shortest diffusion path length to the release medium: since the height

of our tablets is 2 mm in comparison with a diameter of 3 mm, the axial route is followed

(especially of great importance for particles at the centre of the tablet).

tablet diameter: 3 mm tablet diameter: 3 mm

surface

tabl

et h

eigh

t: 2

mm

(a) (b)

Figure 9: Schematic drawing of the experimentally determined MPT concentration with

Raman spectroscopy in mini-matrices containing 2.5% PEO 100.00): (a) after 24 h exposure

to the release medium – exterior surface and radial cutting surface (see also Figure 1a), (b)

after 24 h exposure to the release medium – axial cutting surface (see also Figure 1b). The

MPT distribution in the mini-matrices is schematically presented via a binary colour code

(blue indicating an area with low MPT concentration, red indicating an area with high MPT

concentration). A more detailed presentation of the MPT concentration was not possible as

the Raman signal could not be normalized over the different scanned areas.

tabl

et h

eigh

t: 2

mm

½ height

surface

½ height

207207207207

CHAPTER 4

4. CONCLUSIONS

The presented novel mathematical theories offer two major advantages: (i) deeper

insight into the underlying mass transport processes governing drug and excipient release

from mini-matrices can be gained, and (ii) the optimization of existing devices and

development of novel devices can be facilitated, because the effects of formulation

parameters, such as the tablet height and radius on the resulting drug release kinetics can

be predicted in a quantitative way.

208208208208


REFERENCES




Gemischen zur Herstellung von Arzneiformen, Spektrum der Wissenschaft, July, 18-20

(1995).


Pharm. Biopharm. 54 (2002) 107-117.













Pharm. 269 (2004) 509-522.



235-247.



Pharmacol. 57 (2004) 366-367.

209209209209

CHAPTER 4






[12] E. Verhoeven, T.R.M. De Beer, G. Van den Mooter, J.P. Remon, C. Vervaet, Influence

of formulation and process parameters on the release characteristics of ethylcellulose

sustained-release mini-matrices produced by hot-melt extrusion, Eur. J. Pharm.

Biopharm. 69 (2008) 312-319.

[13] E. Verhoeven, T.R.M. De Beer, E. Schacht, G. Van den Mooter, J.P. Remon, C.

Vervaet, Influence of polyethylene glycol/polyethylene oxide on the release

characteristics of sustained-release ethylcellulose mini-matrices produced by hot-melt

extrusion: in vitro and in vivo evaluation, submitted for publication in Eur. J. Pharm.

Biopharm.

[14] J. Crank (ed.), The mathematics of diffusion, 2nd ed., Clarendon Press, Oxford, UK

(1975).




[17] N. Wu, L.S. Wang, D.C.W. Tan, S.M. Moochhala, Y.Y. Yang, Mathematical modelling

and in vitro study of controlled drug release via a highly swellable and dissoluble

polymer matrix: polyethylene oxide with high molecular weights, J. Control. Release

102 (2005) 569-581.



[19] I. Katzhendler, A. Hoffman, A. Goldberger, M. Friedman, Modeling of drug release from

erodible tablets, J. Pharm. Sci. 86 (1997) 110-115.

210210210210


[20] J. Siepmann, N.A. Peppas, Modeling of drug release from delivery systems based on

hydroxypropyl methylcellulose (HPMC), Adv. Drug Deliver. Rev. 48 (2001) 139-157.

[21] J.M. Vergnaud (ed.), Liquid transport processes in polymeric materials. Prentice-Hall,

Englewood Cliffs, UK (1991).

[22] J.M. Vergnaud (ed.), Controlled drug release of oral dosage forms. Ellis Horwood,

Chichester, UK (1993).

211211211211

212212212212

SUMMARY AND GENERAL CONCLUSIONS

213213213213


214214214214



Hot-melt extrusion is a well-established technique in the plastics industry. In the

pharmaceutical field it is investigated as an application to develop pharmaceutical dosage forms:

granules, sustained-release pellets and matrix tablets, transdermal drug delivery systems and

drug/polymer solid dispersions or solid solutions. A detailed description of the extrusion process,

the different types of equipment, and an overview of drugs and polymers processed with hot-melt

extrusion are given in the INTRODUCTION. This chapter also describes the advantages of

multiparticulate systems compared with conventional dosage forms and some of the production

methods used to manufacture multiple-unit drug delivery systems. Matrix systems are commonly

used as oral sustained-release dosage forms, however their drug release profile is generally

characterized by a burst release effect. Since constant rate delivery is the primary goal of

sustained-release systems, considerable efforts have been made in the development of new drug

delivery concepts in order to avoid the burst effect and to obtain zero-order drug release. Different

strategies proposed to limit burst release from matrix systems are presented in the introduction.

The OBJECTIVE of this doctoral thesis was to develop a polymeric multiparticulate dosage

form by means of a simple, flexible and economically feasible manufacturing method. Therefore,

the properties of mini-matrices consisting of lipophilic as well as hydrophilic polymers to tailor drug

release and processed using hot-melt extrusion were evaluated.

CHAPTER 1 describes the formulation study of mini-matrices with release-sustaining

properties, developed by means of hot-melt extrusion, using ibuprofen as model drug and

ethylcellulose as sustained-release agent. Xanthan gum, a hydrophilic polymer, was added

to the formulation to increase drug release since ibuprofen release from the

ibuprofen/ethylcellulose matrices (60/40, w/w) was too slow (20% in 24 h). In the first part of

chapter 1, it was revealed that changing the xanthan gum concentration as well as its particle

size modified the in vitro drug release. Increasing xanthan gum concentrations yielded a

faster drug release due to a higher liquid uptake, swelling and erosion rate. Regarding the

215215215215


effect of the xanthan gum particle size, no difference was observed for formulations

containing 10 and 20% xanthan gum. However, using 30% xanthan gum, drug release was

influenced by the particle size of the hydrophilic polymer due to the susceptibility of the

coarser xanthan gum particles to erosion. The mean porosity data of the mini-matrices

indicated that the initial porosity (≤ 5.5%) is independent of the composition (xanthan gum

concentration, xanthan gum particle size) of the mini-matrices and therefore not responsible

for any difference in drug release. Drug release from the mini-matrices was mainly diffusion-

controlled, but swelling played an important role to obtain complete drug release within 24 h,

and ibuprofen release was unaffected by the hydrodynamic conditions of the in vitro

dissolution test. Drug release was influenced by the ionic strength of the medium as the

conformation of xanthan gum molecules is determined by the salt concentration. Increasing

the xanthan gum concentration resulted in a higher friability (≤ 0.32%), which was correlated

with a decrease of crushing strength; the influence of xanthan gum particle size was

negligible. The stability data revealed that drug release, water uptake and hardness of the

xanthan gum formulations changed when stored for 12 months at 25°C/60% relative humidity

and 40°C/75% relative humidity, and therefore these mini-matrices should be appropriately

packaged. In the second part of this chapter, an oral dose of 300 mg ibuprofen was

administered to 6 dogs either as an immediate-release preparation (Junifen®), as a

sustained-release formulation (Ibu-Slow® 600 mg (½ tablet)) or as the experimental mini-

matrices (varying in xanthan gum concentration). An HPLC-UV method, validated according

to the ICH-guidelines, was used for determination of drug levels in dog plasma. No

interference with endogenous compounds was detected. Calibration curves were linear in the

whole concentration range (R2 = 0.9997 ± 0.0003; n = 10). The recovery of ibuprofen (0.25-

20 µg/ml range) after extraction varied between 82.4 and 88.9%, while 73.5% of internal

standard (indomethacin) was recovered. The method was precise for the same concentration

range, since the coefficients of variation for repeatability and intermediate precision ranged

between 2.9 and 8.0% and between 4.4 and 12.5%, respectively. Ibuprofen could be

determined accurately as an accuracy between 86.8 and 103.4% was obtained. The limits of

detection and quantification were both set at 0.25 µg/ml. Administration of the experimental

formulations sustained the ibuprofen release. Although a significant difference in dissolution

rate of the 20 and 30% xanthan gum mini-matrices was detected in vitro, the difference in

relative bioavailability (using Ibu-Slow® 600 mg as reference) was limited (70.6 and 73.8%,

respectively).

In CHAPTER 2, mini-matrices (multiple-unit dosage form) with release-sustaining

properties were developed by hot-melt extrusion (cylindrical die: 3 mm) using ethylcellulose

as sustained-release agent and metoprolol tartrate as a model drug which has no plasticizing

216216216216


effect on the polymer. A burst release of metoprolol tartrate was seen from mini-matrices

formulated with triethyl citrate or triacetin (hydrophilic plasticizers), whereas drug release

from matrices plasticized with dibutyl sebacate or diethyl phthalate (hydrophobic) was slow

and followed zero-order kinetics. Despite the different hydrophilic character of the

plasticizers, all had the same plasticizing effect during processing. Dibutyl sebacate was

selected as plasticizer for further development and its concentration was optimized to 50%

(w/w) of the ethylcellulose concentration. Xanthan gum, a hydrophilic polymer, was added to

the formulation to increase drug release. Changing the xanthan gum concentration modified

the in vitro drug release: increasing xanthan gum concentrations (1, 2.5, 5, 10 and 20%, w/w)

yielded a faster drug release. Zero-order drug release was obtained at 5% (w/w) xanthan

gum. Based on X-ray diffraction analysis, it was revealed that metoprolol tartrate remained

crystalline in the hot-melt extruded mini-tablets, independent of the xanthan gum

concentration. The production of the extrudates was reproducible, and after an equilibration

time of 5 min extrudates with a constant quality were produced. Incorporation of kneading

paddles in the screw configuration allowed to manufacture smooth extrudates when

processing the formulation at 60°C. The mixing efficacy and drug release were not affected

by the number of mixing zones or their position along the extruder barrel. Using a screw

configuration only consisting of transport zones resulted at 60°C in extrudates with an

irregular surface structure, and powder spots (identified as metoprolol tartrate via Raman

spectroscopy) were seen on the extrudates surface. Using a higher processing temperature

(70°C) improved the surface structure but metoprolol tartrate spots were still visible on the

surface. Raman analysis revealed that metoprolol tartrate was homogeneously distributed in

the mini-matrices, independent of screw design and processing conditions. Simultaneously

changing the powder feed rate (6-50 g/min) and screw speed (30-200 rpm) did not alter

extrudate quality or dissolution properties.

In CHAPTER 3, it was proposed to substitute xanthan gum for polyethylene glycol or

polyethylene oxides, because it is known that the swelling properties of these polymers

depend on their molecular weight. Mini-matrices with release-sustaining properties were

developed by hot-melt extrusion (diameter 3 mm, height 2 mm) using metoprolol tartrate as

model drug (30%, w/w) and ethylcellulose as sustained-release agent. Polyethylene glycol or

polyethylene oxide was added to the formulation to increase drug release. Changing the

hydrophilic polymer concentration (0, 1, 2.5, 5, 10, 20 and 70%, w/w) and molecular weight

(6000, 100.000, 1.000.000 and 7.000.000) modified the in vitro drug release: increasing

concentrations yielded faster drug release (irrespective of molecular weight), whereas the

influence of molecular weight depended on concentration. Polyethylene glycol/polyethylene

oxide release increased at higher polyethylene glycol/polyethylene oxide concentration

217217217217


(irrespective of molecular weight), and a slower polyethylene glycol/polyethylene oxide

release was observed for all formulations formulated with high molecular weight polyethylene

oxide. Smooth extrudates were obtained when processed at 40 and 70°C for polyethylene

glycol and polyethylene oxide formulations, respectively. Raman analysis revealed that

metoprolol tartrate was homogeneously distributed in the mini-matrices, independent of

hydrophilic polymer concentration and molecular weight. Also drug and polymer crystallinity

were independent of both parameters. An oral dose of 200 mg metoprolol tartrate was

administered to 6 dogs either as immediate-release preparation (Lopresor® 100 (2 tablets)),

as sustained-release formulation (Slow-Lopresor® 200 Divitabs®), or as experimental mini-

matrices (varying in hydrophilic polymer concentration). An HPLC-fluorescence method,

validated according to the ICH-guidelines, was used for determination of drug levels in dog

plasma. Interference of metoprolol tartrate and alprenolol (internal standard) with

endogenous compounds was not detected. The calibration curves were linear (R2 = 0.9976 ±

0.0015; n = 10) in the 0.05 to 3.0 µg/ml concentration range. The recovery of metoprolol

tartrate after extraction varied between 79.3 and 85.0% depending on the concentration,

while 85.1% of internal standard was recovered. The method was precise: the coefficients of

variation for repeatability and intermediate precision ranged from 3.7 to 10.5% and from 1.7

to 11.6%, respectively. Metoprolol tartrate could be determined with an accuracy between

91.7 and 125.7%. The limits of detections and quantification were 0.05 µg/ml. The sustained-

release effect of the experimental formulations was limited and relative bioavailabilities (using

Slow-Lopresor® 200 Divitabs® as reference) of 51.5, 94.4, 131.5, 66.2 and 148.2% were

obtained for mini-matrices containing 5, 10 and 20% XG, and 5 and 20% PEO 1.000.000,

respectively.

Chapter 3 demonstrated that drug and polymer release was affected by the concentration

as well as the molecular weight of polyethylene glycol/polyethylene oxide. Therefore drug and

polymer release from hot-melt extruded mini-matrices were mathematically modelled in CHAPTER

4. Drug and polyethylene glycol/polyethylene oxide release mechanism depended on polymer

content and molecular weight. An increase of the hydrophilic polymer content resulted in higher

drug and polymer diffusion coefficients; an increase in molecular weight of the hydrophilic polymer

lowered the diffusion coefficient of drug and polymer. At high polyethylene glycol/polyethylene

oxide content (≥ 20%) (irrespective of the polymer molecular weight), and at low/intermediate

polymer content in the case of low/intermediate polymer molecular weight (≤ 100.000 Da), drug

and polymer release was controlled by diffusion with constant diffusivities (Fick’s second law of

diffusion). This resulted in a good correlation between theoretical modelling and experimental data

for these mini-matrices (3 mm diameter, 2 mm height). Based on the drug diffusion coefficients of

these mini-matrices, the metoprolol tartrate release patterns from cylinders with arbitrary

218218218218


dimensions (variable radius (0.5-6 mm) for constant height (2 mm), variable height (1-32 mm) for

constant radius (1.5 mm)) could be quantitatively predicted: a smaller radius and shorter length

increased drug release, whereas a higher radius and longer length decreased drug release. The

agreement between theoretical calculations and experimental data could be verified in all cases,

confirming the diffusion-controlled drug release mechanism and proving the validity of the model.

In contrast, at low/intermediate polyethylene glycol/polyethylene oxide content (1-10%) in case of

high polymer molecular weight (≥ 1.000.000 Da), drug and polymer release was not only

controlled by diffusion, resulting in significant deviations between the applied theory (considering

constant diffusivities) and experimental release profiles for mini-matrices with 3 mm diameter and

2 mm height. Thus, in these cases the underlying drug and polymer release mechanisms were

more complex and a new theory was required: a mathematical model was developed, taking into

account the experimentally determined increase in porosity of the cylinders upon drug and

polymer release as well as time-dependent drug and polymer mobilities. This model (with more

realistic conditions for drug and polymer diffusion) could more adequately describe the

experimentally measured release kinetics, resulting in a better fit between theoretical predictions

and experimental data. An interesting application of the model is the possibility to quantitatively

predict the effects of different formulation parameters on the resulting drug and polymer release

patterns.

The objective of this doctoral thesis was to develop by means of a simple, flexible and

continuous processing technique (i.e. hot-melt extrusion) a multiparticulate dosage form with zero-

order sustained-release properties. This study demonstrated that the in vitro and in vivo drug

release kinetics could be varied by selecting a formulation with an appropriate ratio of lipophilic

and hydrophilic polymers, and by changing the particle size and/or molecular weight of the

hydrophilic polymer (i.e. xanthan gum and polyethylene oxide). Nevertheless, a few topics can be

identified for future investigation:

- Since ethylcellulose was used as the main excipient, it would be of interest to investigate if hot-

melt extrusion of other lipophilic polymers (in combination with hydrophilic polymers) offers distinct

advantages compared to the ethylcellulose-based mini-matrices.

- All formulations used in this study were processed using a specific type of laboratory-scale

extruder. It is known that the equipment type and batch size (due to a difference in energy input

and heat generation during longer processing times) can affect the material behaviour, thus

influencing extrudate and matrix quality. Therefore, scale-up would be another challenging step

when evaluating hot-melt extrusion as processing technique for a multiparticulate dosage form.

- Future work can also focus on co-extrusion (multiple layers of material are simultaneously

extruded): incorporation of polymers with different hydration/swelling/release properties in different

layers could result in a dosage form with zero-order drug release kinetics.

219219219219

220220220220

SAMENVATTING EN ALGEMEEN BESLUIT

221221221221


222222222222



‘Hot-melt’ extrusie is een goed ingeburgerde techniek in de polymeer industrie. In de

farmaceutische sector vindt het zijn toepassing in de ontwikkeling van farmaceutische

geneesmiddelvormen zoals granules, pellets en tabletten voor gecontroleerde vrijstelling,

transdermale toedieningsvormen en vaste geneesmiddel/polymeer dispersies of oplossingen. In

de INLEIDING worden een gedetailleerde beschrijving van het extrusieproces en de verschillende

types van apparatuur weergegeven, alsook een overzicht van geneesmiddelen en polymeren die

reeds met behulp van ‘hot-melt’ extrusie werden verwerkt. Dit hoofdstuk vermeld eveneens de

voordelen van multiparticulaire vormen in vergelijking met ‘single-unit’ systemen, gevolgd door

een beschrijving van de mogelijke productiemethoden voor deze multiparticulaire vormen.

Matrices worden vaak aangewend voor de orale, vertraagde vrijgave van geneesmiddelen, maar

het voornaamste nadeel is dat het vrijstellingsprofiel gekarakteriseerd wordt door een ‘burst

release’ fenomeen. Aangezien een constante vrijstellingssnelheid van primordiaal belang is voor

‘sustained-release’ formulaties, is reeds heel wat onderzoek verricht naar nieuwe

geneesmiddelvormen die gekenmerkt worden door een nulde orde vrijstellingsprofiel. De inleiding

beschrijft de in de literatuur vermelde strategieën om ‘burst release’ van matrices te beperken of

zelfs te elimineren.

De DOELSTELLING van dit onderzoekswerk was om een multiparticulaire

geneesmiddelvorm te ontwikkelen op basis van polymeren, gebruik makende van een

eenvoudige, flexibele en economisch haalbare productietechniek. De eigenschappen van mini-

matrices, bestaande uit lipofiele en hydrofiele polymeren om de geneesmiddelvrijstelling te sturen

en geproduceerd via ‘hot-melt’ extrusie, werden geëvalueerd.

HOOFDSTUK 1 beschrijft het formulatie-onderzoek van mini-matrices voor vertraagde

geneesmiddelvrijstelling. Deze mini-matrices werden geproduceerd via ‘hot-melt’ extrusie, met

ibuprofen als modelgeneesmiddel en ethylcellulose als vrijstellingsvertragende drager.

Xanthaangom, een hydrofiel polymeer, werd aan de formulatie toegevoegd om de

223223223223


geneesmiddelvrijstelling te versnellen, vermits de vrijstelling van ibuprofen uit

ibuprofen/ethylcellulose matrices (60/40, g/g) te traag verliep (20% in 24 u). In het eerste deel van

hoofdstuk 1 werd aangetoond dat zowel de xanthaangom-concentratie als de deeltjesgrootte van

xanthaangom het in vitro vrijstellingsprofiel beïnvloedden. Hogere xanthaangom-concentraties

zorgden voor een snellere geneesmiddelvrijstelling, dit omwille van een hogere wateropname,

zwelling en erosiesnelheid. Wat betreft de invloed van de deeltjesgrootte van xanthaangom, werd

er geen verschil in vrijstelling vastgesteld tussen formulaties op basis van 10 en 20%

xanthaangom. Echter, voor de formulaties op basis van 30% xanthaangom werd de

geneesmiddelvrijstelling wel beïnvloed door de deeltjesgrootte van het hydrofiele polymeer, omdat

de grotere xanthaangomdeeltjes gevoeliger zijn voor erosie. Onderzoek naar de porositeit van de

mini-matrices gaf aan dat de initiële porositeit (≤ 5.5%) onafhankelijk is van de samenstelling

(xanthaangom-concentratie, xanthaangom-deeltjesgrootte) van de mini-matrices, en aldus niet

verantwoordelijk kan geacht worden voor verschillen in de vrijstellingsprofielen.

Geneesmiddelvrijstelling uit de mini-matrices was vooral diffusie-gecontroleerd, nochtans bleek

ook zwelling een belangrijke rol te spelen in het bekomen van volledige geneesmiddelvrijstelling

binnen de 24 u. De ibuprofen-vrijstelling werd niet beïnvloed door de hydrodynamische condities

tijdens de in vitro dissolutietest. De ionensterkte van het dissolutiemedium bleek wel een invloed

te hebben op de geneesmiddelvrijstelling, vermits de conformatie van xanthaangom afhangt van

de zoutconcentratie. Het verhogen van de xanthaangom-concentratie leidde tot een hogere

friabiliteit (≤ 0.32%), dewelke gecorreleerd was met een daling van de hardheid; de invloed van de

xanthaangom-deeltjesgrootte bleek verwaarloosbaar te zijn. Een stabiliteitsstudie toonde aan dat

de geneesmiddelvrijstelling, wateropname en hardheid van de xanthaangom-formulaties

wijzigden gedurende 12 maanden bewaring bij 25°C/60% relatieve vochtigheid en bij 40°C/75%

relatieve vochtigheid, wat inhoudt dat deze formulaties op een geschikte manier zullen moeten

verpakt worden. In het tweede deel van dit hoofdstuk werd een orale dosis van 300 mg ibuprofen

toegediend aan 6 honden, dit onder de vorm van een ‘immediate-release’ preparaat (Junifen®),

een ‘sustained-release’ preparaat (Ibu-Slow® 600 mg (½ tablet)) of de experimentele formulaties

(mini-matrices met variërende xanthaangom-concentratie). Via een HPLC-UV methode,

gevalideerd volgens de ICH-richtlijnen, werden de concentraties aan ibuprofen in hondenplasma

bepaald. Er werd geen interferentie vastgesteld tussen ibuprofen, de interne standaard en

endogene plasmacomponenten. De calibratiecurves waren lineair binnen het gehele

concentratiegebied (R2 = 0.9997 ± 0.0003; n = 10). De recovery van ibuprofen

(concentratiegebied: 0.25-20 µg/ml) na extractie varieerde tussen 82.4 en 88.9%, terwijl 73.5%

van de interne standaard (indomethacine) werd teruggevonden. De methode was precies binnen

dezelfde concentratierange, aangezien de variatiecoëfficiënten voor de herhaalbaarheid en

intermediaire precisie respectievelijk tussen 2.9 en 8.0% en tussen 4.4 en 12.5% lagen. Ibuprofen

kon nauwkeurig bepaald worden, vermits een accuraatheid tussen 86.8 en 103.4% verkregen

224224224224


werd. De detectie- en kwantificatielimiet werden vastgelegd op 0.25 µg/ml. Toediening van de

experimentele formulaties leidde tot een vertraagde ibuprofen-vrijstelling. Echter, het significante

verschil in in vitro dissolutiesnelheid tussen de 20 en 30% xanthaangom-formulaties werd niet

teruggevonden in vivo: het verschil in relatieve biologische beschikbaarheid (met Ibu-Slow® 600

mg als referentie) bleek immers beperkt (respectievelijk 70.6 en 73.8%).

In HOOFDSTUK 2 werden via ‘hot-melt’ extrusie (cylindrische ‘die’: 3 mm) mini-matrices

met vrijstellingsvertragende eigenschappen geproduceerd, op basis van ethylcellulose als

vrijstellingsvertragende drager en metoprololtartraat als modelgeneesmiddel dat geen

weekmakend effect heeft op het dragerpolymeer. Bij mini-matrices geformuleerd op basis

van triethylcitraat of triacetine (hydrofiele weekmakers) werd een ‘burst release’ vastgesteld,

terwijl geneesmiddelvrijstelling uit matrices met dibutylsebacaat of diethylftalaat (hydrofobe

weekmakers) een traag, nulde orde profiel vertoonde. Ondanks het verschil in hydrofiliciteit

van de weekmakers hadden ze allen hetzelfde weekmakend effect tijdens verwerking.

Dibutylsebacaat werd geselecteerd als weekmaker voor de verdere formulatie-ontwikkeling

en de concentratie werd geoptimaliseerd tot 50% (g/g) van de ethylcellulose-concentratie.

Xanthaangom, een hydrofiel polymeer, werd toegevoegd aan de formulatie om de

geneesmiddelvrijstelling te versnellen. Wijziging van de xanthaangom-concentratie leidde tot

een wijziging van de in vitro geneesmiddelvrijstelling: hogere xanthaangom-concentraties (1,

2.5, 5, 10 and 20%, g/g) gaven een snellere geneesmiddelvrijstelling. Een nulde orde

vrijstellingskinetiek werd bereikt met 5% (g/g) xanthaangom. Met behulp van X-

straaldiffractie werd vastgesteld dat metoprololtartraat in kristallijne vorm aanwezig was in de

geëxtrudeerde matrices, onafhankelijk van de xanthaangom-concentratie. De productie van

de extrudaten was reproduceerbaar, en na een equilibratietijd van 5 min konden extrudaten

van constante kwaliteit geproduceerd worden. Incorporatie van kneedelementen in de

schroefconfiguratie liet toe gladde extrudaten te produceren bij een procestemperatuur van

60°C. Het aantal mengzones en hun lokalisatie in het extrusietoestel hadden geen invloed op

de mengefficiëntie en de geneesmiddelvrijstelling. Een schroefconfiguratie van louter

transportzones leidde (bij 60°C) tot extrudaten met een onregelmatig oppervlak en met

poederzones op het oppervlak (geïdentificeerd als metoprololtartraat via Raman

spectroscopie). Verhogen van de extrusietemperatuur (70°C) verbeterde de

oppervlaktestructuur van het extrudaat, doch metoprololtartraat-zones bleven zichtbaar op

het oppervlak. Raman analyse toonde aan dat metoprololtartraat homogeen verdeeld was in

de mini-matrices, onafhankelijk van de schroefconfiguratie en procesomstandigheden.

Simultane wijziging van de poedertoevoersnelheid (6-50 g/min) en de schroefsnelheid (30-

200 tpm) had geen invloed op de extrudaatkwaliteit of de dissolutie-eigenschappen.

225225225225


In HOOFDSTUK 3 werd geopteerd om xanthaangom te vervangen door polyethyleenglycol

of polyethyleenoxides, omdat van deze polymeren gekend is dat hun zweleigenschappen

afhangen van hun moleculair gewicht. Er werden mini-matrices met vrijstellingsvertragende

eigenschappen geproduceerd via ‘hot-melt’ extrusie (diameter 3 mm, hoogte 2 mm), met

metoprololtartraat als modelgeneesmiddel (30%, g/g) en ethylcellulose als vrijstellingsvertragende

drager. Polyethyleenglycol of polyethyleenoxide werd toegevoegd aan de formulatie om de

geneesmiddelvrijstelling te versnellen. Wijziging van de concentratie van het hydrofiele polymeer

(0, 1, 2.5, 5, 10, 20 en 70%, g/g) en van het moleculair gewicht (6000, 100.000, 1.000.000 en

7.000.000) wijzigde het in vitro vrijstellingsprofiel: toenemende concentraties gaven een snellere

geneesmiddelvrijstelling (onafhankelijk van het moleculair gewicht), terwijl de invloed van het

moleculair gewicht afhing van de concentratie. Polyethyleenglycol/polyethyleenoxide-vrijstelling

nam toe bij hogere polyethyleenglycol/polyethyleenoxide-concentratie (onafhankelijk van het

moleculair gewicht), en een tragere polyethyleenglycol/polyethyleenoxide-vrijstelling werd gezien

voor alle formulaties met polyethyleenoxide met hoog moleculair gewicht. Er werden gladde

extrudaten bekomen bij een extrusietemperatuur van respectievelijk 40 en 70°C, voor

polyethyleenglycol- en polyethyleenoxide-formulaties. Raman spectroscopie toonde aan dat

metoprololtartraat homogeen verdeeld was in de mini-matrices, onafhankelijk van de concentratie

en het moleculair gewicht van het hydrofiele polymeer. Ook de geneesmiddelen

polymeerkristalliniteit bleken onafhankelijk te zijn van deze twee parameters. De biologische

beschikbaarheid werd bepaald na orale toediening van 200 mg metoprololtartraat aan 6 honden,

dit onder de vorm van een ‘immediate-release’ preparaat (Lopresor® 100 (2 tabletten)), een

‘sustained-release’ preparaat (Slow-Lopresor® 200 Divitabs®) of de experimentele formulaties

(mini-matrices met variërende concentraties van het hydrofiele polymeer). Voor de bepaling van

het geneesmiddelgehalte in hondenplasma werd gebruik gemaakt van een HPLC-fluorescentie

methode die gevalideerd was in overeenstemming met de ICH-richtlijnen. Er werd geen

interferentie waargenomen van metoprololtartraat en alprenolol (interne standaard) met endogene

componenten. De calibratiecurves waren lineair (R2 = 0.9976 ± 0.0015; n = 10) in een

concentratiegebied van 0.05 tot 3.0 µg/ml. De recovery van metoprololtartraat na extractie

varieerde afhankelijk van de concentratie tussen 79.3 en 85.0%, terwijl 85.1% van de interne

standaard werd gerecupereerd. De methode was exact: de variatiecoëfficiënten voor de

herhaalbaarheid en intermediaire precisie varieerden respectievelijk van 3.7 tot 10.5% en van 1.7

tot 11.6%. Metoprololtartraat kon accuraat bepaald worden, aangezien een nauwkeurigheid

tussen 91.7 en 125.7% behaald werd. De detectie- en kwantificatielimiet bedroegen 0.05 µg/ml.

Het vrijstellingsvertragend effect van de experimentele formulaties bleek echter beperkt. Een

relatieve biologische beschikbaarheid (met Slow-Lopresor® 200 Divitabs® als referentie) van

respectievelijk 51.5, 94.4, 131.5, 66.2 en 148.2% werd bekomen met mini-matrices op basis van

5, 10 en 20% xanthaangom, en 5 en 20% PEO 1.000.000.

226226226226


Hoofdstuk 3 toonde aan dat geneesmiddelen polymeervrijstelling beïnvloed werden door

zowel de concentratie als het moleculair gewicht van polyethyleenglycol/polyethyleenoxide. Om

deze reden werd in HOOFDSTUK 4 de geneesmiddelen polymeervrijstelling uit de geëxtrudeerde

mini-matrices wiskundig gemodelleerd. Het geneesmiddelen

polyethyleenglycol/polyethyleenoxide-vrijstellingsmechanisme was afhankelijk van de

polymeerconcentratie en het moleculair gewicht. Het verhogen van de concentratie hydrofiel

polymeer resulteerde in hogere geneesmiddelen polymeer-diffusiecoëfficiënten. Een stijging van

het moleculair gewicht leidde tot een daling van de diffusiecoëfficiënt van geneesmiddel en

polymeer. Bij hoge polyethyleenglycol/polyethyleenoxide-concentratie (≥ 20%) (onafhankelijk van

het moleculair gewicht van het polymeer), en bij lage/intermediaire polymeerconcentratie in geval

van laag/intermediair moleculair gewicht (≤ 100.000 Da), werd de geneesmiddelen de

polymeervrijstelling gecontroleerd door diffusie (tweede wet van Fick). Voor deze mini-matrices (3

mm diameter, 2 mm hoogte) resulteerde dit in een goede correlatie tussen theoretische

modellering en experimentele data. Op basis van de geneesmiddel-diffusiecoëfficiënten van deze

mini-matrices, kon de vrijstelling van metoprololtartraat uit cilinders met willekeurige afmetingen

(variabele diameter (0.5-6 mm) met constante hoogte (2 mm), variabele hoogte (1-32 mm) met

constante diameter (1.5 mm)) kwantitatief voorspeld worden: een kleine diameter en een kleine

lengte versnellen de geneesmiddelvrijstelling, terwijl een grote diameter en een grote lengte de

geneesmiddelvrijstelling vertragen. Er werd in alle gevallen overeenstemming tussen de

theoretische berekeningen en de experimentele data vastgesteld, waardoor het diffusie-

gecontroleerde vrijstellingsmechanisme en de validiteit van het model kon worden bevestigd.

Daarentegen, bij lage/intermediaire polyethyleenglycol/polyethyleenoxide-concentratie (1-10%) in

geval van hoog moleculair gewicht (≥ 1.000.000 Da) werd de geneesmiddelen

polymeervrijstelling niet enkel en alleen door diffusie gecontroleerd, waardoor er significante

afwijkingen waren tussen theoretische (uitgaande van constante diffusie) en experimentele

vrijstellingsprofielen voor mini-matrices van 3 mm diameter en 2 mm hoogte. Dus, in deze

gevallen bleek het geneesmiddelen polymeervrijstellingsmechanisme complexer te zijn en was

een nieuwe theorie noodzakelijk: er werd een mathematisch model ontwikkeld dat rekening hield

met de experimenteel vastgestelde toename in porositeit van de cilinders tijdens geneesmiddel-

en polymeervrijstelling, en met de tijdsafhankelijke geneesmiddelen polymeermobiliteit. Dit

model (met meer realistische condities voor geneesmiddelen polymeerdiffusie) kon de

experimenteel vastgestelde vrijstellingskinetieken adequater beschrijven, waardoor een betere

overeenkomst tussen de theoretische voorspellingen en de experimentele data werd bekomen.

Een interessante toepassing van dit model is de mogelijkheid om het effect van verschillende

formulatieparameters op de geneesmiddelen polymeervrijstellingspatronen kwantitatief te gaan

voorspellen.

227227227227


Het objectief van deze doctoraatsthesis was het ontwikkelen van een multiparticulaire

geneesmiddelvorm met een vertraagd, nulde orde vrijstellingsprofiel, met behulp van een

eenvoudige, flexibele en continue productietechniek, nl. ‘hot-melt’ extrusie. Dit onderzoek toonde

aan dat de in vitro en in vivo vrijstellingskinetiek kan aangepast worden door selectie van een

formulatie met een geschikte ratio van lipofiele en hydrofiele polymeren, en door wijziging van de

deeltjesgrootte en/of moleculair gewicht van het hydrofiel polymeer (xanthaangom en

polyethyleenoxide). Verder onderzoek zou zich kunnen richten op volgende onderwerpen:

- Vermits dit onderzoek steeds werkte met ethylcellulose als dragerpolymeer zou het interessant

kunnen zijn om na te gaan of ‘hot-melt’ extrusie van andere lipofiele polymeren (in combinatie van

hydrofiele polymeren) duidelijke voordelen biedt ten opzichte van mini-matrices op basis van

ethylcellulose.

- Alle formulaties onderzocht tijdens dit doctoraatswerk werden geproduceerd met een specifiek

type ‘lab-scale’ extruder. Het is echter bekend dat toesteltype en batchgrootte (omwille van een

verschil in energietoevoer en warmte-ontwikkeling tijdens de langere procestijden) het gedrag van

het materiaal kan beïnvloeden, en aldus ook de extrudaat- en matrixkwaliteit. Daardoor zou

procesopschaling een uitdagende stap kunnen vormen in de evaluatie van ‘hot-melt’ extrusie als

productietechniek voor multiparticulaire geneesmiddelvormen.

- Toekomstig werk kan ook focussen op co-extrusie (meerdere lagen van materiaal worden

simultaan geëxtrudeerd): incorporatie van polymeren met verschillende hydratatie-, zwelling- en

vrijstellingseigenschappen in verschillende lagen zou kunnen resulteren in een

geneesmiddelvorm met nulde orde kinetiek.

228228228228

ELLEN VERHOEVENVan den Mooter (Laboratorium voor Farmacotechnologie en Biofarmacie, Katholieke...

Documents

Transcript of ELLEN VERHOEVENVan den Mooter (Laboratorium voor Farmacotechnologie en Biofarmacie, Katholieke...