University of Rhode Island
DigitalCommons@URI
Open Access Dissertations
1991
OPTIMIZATION OF IBUPROFEN
FORMULATION: MOLECULAR
PHARMACEUTICS OF THE BIOLOGICALLY
ACTIVE STEREOISOMER
Alain Joseph Romero
University of Rhode Island
Follow this and additional works at: http://digitalcommons.uri.edu/oa_diss
Recommended Citation
Romero, Alain Joseph, "OPTIMIZATION OF IBUPROFEN FORMULATION: MOLECULAR PHARMACEUTICS OF THE
BIOLOGICALLY ACTIVE STEREOISOMER" (1991). Open Access Dissertations. Paper 185.
http://digitalcommons.uri.edu/oa_diss/185
This Dissertation is brought to you for free and open access by DigitalCommons@URI. It has been accepted for inclusion in Open Access Dissertations
by an authorized administrator of DigitalCommons@URI. For more information, please contact [email protected].
OPTIMIZATION OF IBUPROFEN FORMULATION:
MOLECULAR PHARMACEUTICS
OF THE BIOLOGICALLY ACTIVE STEREOISOMER
BY
ALAIN JOSEPH ROMERO
A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
IN
PHARMACEUTICAL SCIENCES
UNIVERSITY OF RHODE ISLAND
1991
(
DOCTOR OF PHILOSOPHY DISSERTATION
OF
ALAIN JOSEPH ROMERO
APPROVED:
Dissertation Committee
Major Professor
€:?_,,~
~' 7
~-~-
DEAN OF THE GRADUATE SCHOOL
UNIVERSITY OF RHODE ISLAND
1991
ABSTRACT
A pr ogr a m of e x peri me nt s comparing
rac - ibuprof en
to
t h at
of
t he
( - )- S- i bu pr ofe n
Earl y i n ve s tigat i on s re v ealed that al t h ough
t he
Uni t ed
Stat es
formulatio n
Pharmac opeia
was pe r f or med.
compl y ing
pr o perties.
and
with
c ompendial standards five
different sources of rac-ibuprofen had different
characteristics
of
pr ocessing
as a result variable biophar maceuti c al
Cr y stal habits critically influenced pr oces s ing
parameters.
It was possible t o identify l ow and h i gh liquid
requirement
powders
Further
analysis
of
for
the
wet
granulation
rac - ibupr ofen
during the different stages o f
end - point.
crystals was perf o rmed
tablet
manufacture .
Phase
diagrams c onfirmed that rac - ibupr o fen crystallizes preferentially in the
racemate .
monoclinic
space
P2
c
group as a true
1
It was found that crystal distortion translated
int o an hydrophobic network of ibuprofen causing a
8.5
be
KJ mole -
1
of
in the enthalpy of fusion. This is thought t o
resp onsible
tablets.
dr op
for
The
the
poor
performance
of
ibuprofen
exte nt of this network seemed to be dose de -
pendent as suggested by the dissolution profiles.
Using single crystal x-ray diffraction. the crystal lat tice of (+)- S- Ibuprofen was
pharmaceutics
of
isomer. although
molecules
elucidated
a nd
the
molecular
the S and racemate investigated .
crystallizing
with
the
same
The (+)
number
of
in the unit cell. exhibited a t otally independent
crys~al ~ith a mel~ing poin t of 54°C
-1
sion ( c.E ) of 17 . 9 KJ mo le less.
ibuprofec was more soluble
media .
However ,
a
than
study
a~d a~ enthalpy of fu ".'he
stereoisomer
rac - ibuprofen
of
in
aqueous
the solution thermodynamics
revea::.ed that standard free ener gy of s o luti on ( L>G o =
rac
and t>G o = 29.5
s
in KJ mole
and entropy of
solution
( L>H
rac
32. 2
=
and
) were comparable.
s
is omer
2
to the racemate (0. 34 m 1g )
pared
the
slower
biopharmaceutical
tions.
however ,
the
S (+)
at
)
( 2.8.10- 3
.
pH
1. 3
The small
2
m g)
com-
is pr obably responsible
intrinsic
dissolution
( IDRrac =
11.6
µg . sec - 1 .cm- 2 ) . The study of
properties
of
indicated
an
dissolution than the racemate .
of
30.3
whereas heats
different
= 51 . 5 in KJ ! mole
specific surf ace area of the
for
very
were
L>HS
of
stereoisomer
(+)-S- Ibuprofen
excellent
flow
f o rmula and better
Extensive eutectic
behavi or
might be of some concern to the
formulators.
In
order
to
formulate
the pure enantiomer. the phar -
macokinetic of rac - ibuprofen was
Stellatm
simulation
investigated.
Using
the
software it was determined that 1 / 3 of
the (+)- S- Ibuprofen was derived from the inversion of (-)- Ribuprofen
systemically
rather
than pre - systemically. Thus
150 mg of (+)-S-ibuprofen might be
alent to 200 mg of rac-ibuprofen .
therapeutically
equiv -
ACKNOWLEDGEMENTS
I
want to express my profound appreciation for not only
the guidance and assistance of C.T.
Rhodes.
but
also
his
support of all my endeavors . The project led me down avenues
that nearly exhausted the
thought
breadth
of
explorati o n
I
ever
possible. Without the help of my major professor my
growth over these last three years would have
been
und oub-
tedly stunted.
I want especially to thank
the
award
Ciba-Ceigy
Corp o rati o n
of a fellowship over the last three years. Under
the supervision of Dr. George Lukas, Executive
PPT .
I
for
had
guidance
and
A
beyond the call
handful
of
duty
of
advice to continue on the right
direction and also the freedom to
curiosities.
Director
of
alleviate
my
scientific
individuals at Ciba- Geigy went
to
assist
me .
among
them
Lou
Savastano. Russ Rackley and Franck Clarke: they made my stay
in Summit most productive, fruitful and enjoyable.
I
would like to extend my thanks to Dr . Bauer and Ethyl
Corp. for providing me with the (-)- S- ibuprofen. thus giving
me the opportunity to work on a very challenging topic.
Mrs Reba Whitford deserves
Time
particular
acknowledgement.
and again she has been the cohesion of the department.
taking under her wing all the
myself.
Over
graduate
students
including
my tenure at URI I have enjoyed and benefited
from her professionalism. care and devotion .
iv
?~~a::y.
anc
~c
a~tic~pacion
all
of
o:
~hose
gradua~e
wtc have shared in the trial
work. my fellow students. thank
you. The comple tion of this dissertation was achieved within
the heart of this group. To Alex and Mary Kay
have and will enjoy the greatest of friendship.
v
with
whom
I
PREFACE
I
elected t o write this dissertation following the for-
mat of the manuscript plan described in section 11-3 of
Graduate manual at the University of Rhode Island.
tion was most appropriate to present my results
in
the
This opseveral
sections.
Section I consists of
problem
with
tions. The five
Section
a
general
introduction
presentation
manuscripts.
chronologically
II. are the core of this study .
have been either
accepted
of
the
and objectives of my investiga-
or
submitted
numbered
in
Most of the papers
for
publication.
Section III. a published manuscript on the topic of clinical
research. was not directly related to the core of this work.
but
some of the analytical method was later employed in the
pharmacokinetic methodology used for the (•)- S- ibuprofen.
Section
IV
is
a
set
of appendices A to D giving ex -
perimental details on this work.
vi
TABLE OF CONTENTS
ABSTRACT.
ACKNOWLEDGEMENTS.
. .. iv
PREFACE .... .... ...... .
. .. vi
LIST OF TABLES .. .
. . ix
LIST OF FIGURES ... .
. . .. . . . . .. .... xii
LIST OF PUBLICATIONS AND PRESENTATION.
. . . ... . . . . . xviii
SECTION I
INTRODUCTION .......... . .
. . . . .. . . . . . . . 2
SECTION II
MANUSCRIPT I :
INFLUENCE OF DRUG SOURCES ON THE
BIOPHARMACEUTICAL PROPERTIES AND PROCESSING OF
HIGH DOSE IBUPROFEN FORMULATIONS.
MANUSCRIPT II:
MONITORING CRYSTAL MODIFICATIONS
IN SYSTEMS CONTAINING IBUPROFEN.
MANUSCRIPT III:
. .. 55
AN EVALUATION OF IBUPROFEN
BIOINVERSION BY SIMULATION ...
MANUSCRIPT IV:
. .... 10
. ..... 86
AN APPROACH TO STEREOSPECIFIC
PREFORMULATION.
. . ... . 102
vii
MANUSCRIPT V:
S'.:'EREOCiEMICA L ASPECTS OF THE
MOLECULAR PHARMACEUTICS OF IBUPROFEN.
MANUSCRIPT VI:
. .127
FORMULATION OF THE BIOLOGICALLY
ACTIVE STEREOISOMER OF IBUPROFEN.............
CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK.
.149
. ... 169
SECTION III
MANUSCRIPT VII:
USE AND LIMITATIONS OF THE
TRUNCATED AREA UNDER THE CURVE IN
. . 174
BIOEQUIVALENCE TESTING ...
SECTION IV
APPENDI X A.
.. . .... . . . . . ......... . . .. . .. . ..... 220
APPENDIX B.
. ..... . . . ............ . . .. . . .. ..... 223
APPEND I X C ... .. ...... . ... . . ..... . .......... ..... . ... 231
APPENDIX D . ... . . . . ... ... . . . .. . . . ... .. .... . ...... .... 253
BIBLIOGRAPHY ... . .. .. .... .. . .. . .. . .. .. ... . . ....... ... 258
viii
LIST OF TABLES
?age
Tables
Manuscript I
Table I
Source and Origin of Ibuprofen . .
.31
Table II
Starting Formulation ........ . . .
.32
Table III
Mean Particle Size.
Table IV
Surface Area. BET values ................. 34
Table
v
. .33
Melting Ranges. Differential Scanning
Calorimetry .
Table VI
. ....... ........ 35
Enthalpy of Fusion of Various Sources
( J i g ) Analysis of Variance.
Table VII
. .... 36
Liquid Requirements for Ibuprofen
Granulations End-Point.
.37
Table VIII: Effect of Different Sources on Liquid
Requirements.
. . . . . . . . . . . . . . . . . . . 38
Table IX
Moisture Content of Wet Granules.
Table X
Hardness Levels.
Table XI
Analysis of Variance for Tablet Hardness
. . 39
... ........... .......... 40
One Degree of Comparison.
. .. 41
Manuscript II
Table I
Formulation Used in This Study ............ 72
Table II
Experimental Design.
ix
. . . . . . . . . . . . . . . 73
Table III:
Thermal AnalyE i S of Ibuprofe=
Formulation.
Table I V
. . . . . . .. ..... . .. . ...
Ana lysis of Variance for Enthalpy
of Fusion .............. .
Table v
~{ ~
. . 75
Crystal Modifications and Initial
Dissolution Rates.
. .. . ... . . 76
Manuscript III
Table I
Simulation Parameters and Results for
Bi o inversion of Ibuprofen ....
Table II:
. . . . . . . . . . . 99
AUC Values in Man after administration of
Individual Ibuprofen Enantiomers or
Racemate .. ..... ....... . .
. . . . . . . . 100
Manuscript IV
Table I
Thermal Analysis of Ibuprofen ..... .... ..... 116
Table II: Heats of Solution for ibuprofen.
. ... 117
Manuscript v
Table I
: Crystal Data of (+)-S-ibuprofen.
. .... 140
Tabl e II: Thermodynamic Functions of Melting and
Solubility .......... .
. ................. 141
Manuscript VI
Table I
Experimental Design ..... . .. .
Table II
Thermal Analysis of (+)- S-ibuprofen
x
........ 161
24
H O U~E
of
M~xi ng
..
162
Tab : e III : Thermal Anal ys is of (-)- S- ibuprofen
One Week St o rage at Room Temperature
Table IV
Thermal Analysis of ( +)- S-ibuprofen
Tablets.
Table v
.. 163
. .... 164
Biopharmaceutical Properties of
( +
)- S- ibuprofen Tablets.
. 165
Manuscript VII
Table I
One Compartment Model Simulation
Parameters.
Table II
.. . . 198
Tw o Compartment Model Simulation
Parameters ... . ...... .
199
Table III: Pharmacokinetic Parameters of
the Reference Products ..... . ........ .
Table IV
Pharmacokinetic Relevance of
AUCtmax ........... . .. .
Table v
. .. 200
.. .. 201
One Compartment Model
AUG ratios for Bioequivalence ..
. ... 202
Table VI : Bioinequivalent Products-AUG Ratios ....... 203
Table VII: Fal se Inequi valence. .
. 204
Table VIII:Truncated AUG ratios of Bioavailability
Percentage and Corresponding AUCinf . ...... 205
Appendix D
Table I
:Pure Ibuprofen Compacted:
Thermal Analysis
xi
.... 257
LIST OF FIG URES
Figu r es
Pa ge
Manuscript I
Figure
Size Frequency distribution
Surface-Weighted Mean .............. .. . . .... 42
Figure 2
Apparent Bulk and Tapped Density ... .... .... 43
Figure
Scanning Electron Microscopy
3
Photographs of Ibuprofen Crystals
from I : Fld. II: Fhd, III: Boots
and IV:Ethyl.
Figure 4
. . . . . . . . . . . . . . . . 44
Scanning Electron Microscopy Photographs of
Ibuprofen Crystals from I and II: Cheminor,
II and IV: different batches of Ethyl.
46
Figure 5
Typical DSC endotherms of Ibuprofen ..... . .. 48
Figure 6
X-Ray Diffraction Patterns of Ibuprofen .... 49
Figure 7
Power Consumption for Wet Granulation I.
Figure 8
Power Consumption for Wet Granulation II ... 51
Figure 9
Loss on Drying Profiles at 40°c.
.. 52
Figure 10: Dissolution of Ibuprofen Cores ..
. . .. 53
.50
Figure 11: Dissolution of Ibuprofen Cores (CHEM ) ...... 54
Manuscript II
Figure
Unit Cell of Rae-ibuprofen ............ .
.77
Figure 2
Crystal Packing ......... .
.78
xii
Fibure 3
X- Ray Di ffraction Patcerns of A) pure
ibuprofen. E ) Granulati ons and C) Tablet s .... 79
Figure 4
Scanning Electron Mi c r osco py Ph otogr aphs . . . 80
Figu re 5
Effect of Compact i on on Melting .
Figure 6
Effe c t of Compacti on o n Enthalpy .
Figure 7
Dissolution Pr of iles of Ibupr o fen
.82
Cor es (0 and 1 Intragranular Ratio ) .
Figure 8
... 83
. .84
Dissolution Profiles of Ibuprofen
Cores ( 1 , 2 and 1 / 3 Intragranular Ratio ) .... 85
Manuscript III
Figure 1 : Chiral Bioinversion Model Pr oposed
. .. ... 101
by Jamali et al ( 5).
Manuscript IV
Figure
Cross Section of the Modified Woods
Rotati n g Apparatus ..... ...... ............ . 118
Figure 2 : Powder X- Ray Diffraction Patterns:
Effect of Grinding . . . . . . . . . . . . . . . . . . . .
Figure 3
119
Scanning Electron Microscopy of
. .. . .. . . .. 120
(+)- $ - ibuprofen.
Figure 4 : Flowability / Packing
Log V / V against the Number of Taps ... . . .. 122
0
Figure 5 : Intrinsic Dissolution of
(+)- S- ibuprofen . .. .
. . .. . . . .... . . . ... 123
Figure 6 : Effect of Aging . .. .
. .. . . . .. . . . . . . . . .... 124
xiii
[rac-ib~pr of en ) .
Figc:re 7
Vant' Hoff Plots
!'igc:re 8
Van t ' Hoff Pl ots [(-)- 5 - ibupr of en l.
.125
. 126
Manuscript v
Figure
Thermograms of ibuprofen A) Pure
rac - ibupr of en. B) pure (-)- 8 - ibupr of en ..... 142
Figure 2
Thermograms of ibuprofen
Different Enantiomeric Compositions.
. .. 144
Figure 3
Phase Diagram of Ibuprofen.
Figure 4
Test of the Prigofine -Defay Equati on.
Figure 5
Molecular Modeling
a ) Unit Cell of (+)-S-ibupr ofen.
C~)
.145
. . 146
. 147
b ) Crystal Lattice of (+)- S- isomer.
c ) Superposition of
.1 43
Stereoisomer
Molecules Involved in H2 Bonding .. ... .. 148
Manuscript VI
Figure
Crystal Distortion in Tablets ...... .... ... 166
Figure 2
Dissolution of Ibuprofen Cores ..... ... .. .. 167
Manuscript VII
... 206
Figure
One Compartment Model .... .. . . .
Figure 2
Two Compartment Model . .
Figure 3
Pharmacokinetic Relevance of AUCtmax a ) .. . 208
. ...... 207
Pharmacokinetic Relevance of AUCtmax b ) ... 209
xiv
Figure 4
Figure 5
Percentage of
AUC~n:
Reference
a ) One
Compartmen~
Kodel.
b) Two
Compartmen~
Model ..... ............. 211
. . 210
Percentage of AUCinf Required to show
Bioequivalence.
........
. .. . 212
Figure 6
Kinetics of AUC Rati os ( Slow Absorpti o n ) .. 213
Figure 7
Kinetics of AUC Rati os ( Fast Absorption ) .. 2 14
Figure 8
Percentage of AUCinf
when AUC
Figure 9
ratio
~
( F ) ...
. ..... 215
Per ce ntage of AUCinf at any time after
dosing
a ) One compartment Model ..
.216
b ) Tw o Compartment Model . ............ .
. . 217
... 218
Figure 10: Therapeutic Window.
Appendix A
Figure
Typical Power Consumpti on Trace.
.221
Figure 2
Calibration Curve for Ibuprofe n.
. . . 222
Appendix B
Figure
Molecular Packing in the Crystal
Lattice of Rae-ibuprofen . .
Figure 2
. ..... . ... . . . 225
Representation of the S-(+ ) -ibuprofen
Crystal Lattice from a different angle . ... 226
Figure 3
Arrays of (+) -S and (-)- R Molecules
Cryst al Lattice of Rae-ibuprofen .
xv
. . 227
Fi gu:::e 4
Details of the
~omochiral
Eydrogen Bond
in the S Crystal Lattice.
. .... 228
Figure 5
Identification of the Molecu lar Torsi on .. . 229
Figure 6
Theoreti cal X- ray Diffr act og rams .......... 230
Appendix C
Figure
Typical DSC thermogram of
Ethyl ibuprofen.
Figure 2
. . 235
DSC Scan of Polyvinyl Pirrolidone
( Povidone ) in the temperature ran ge of
interest . . .
Figure 3
..... 236
DSC Scan of Na Starch Glyc olate ( Expl otab)
in the temperature range of interest ...... 237
Figure 4
DSC Scan of Lact ose ( Fast Flow Lactose ) in
. . 238
the temperature range of interest .
Figure 5
Example of a typical endotherm of Ethyl
El3l;An 77% ibuprofen formulat i on with 1%
of disintegrant 113 intrgranular .
. ... 239
Figure 6 : Example of a typical endotherm of a ground
ibuprofen tablet obtained at regular
. ............... 240
compaction.
Figure 7
Example of a typical endotherm of a ground
physical mixture ( before wet granulation ) .241
Figure 8
Typical Endotherm of (+)- S-ibuprofen . . . . . 242
Figure 9
Endotherms of (+) -S-ibuprofen and
rac - ibuprofen recrystallized at 4°C from
xvi
methan ol liquors.
.24 3
F igur e 10: Therm og ram of a Sand racemate mix"ure
( 25 - 75 ) melted and
recrystallized at 4
°c .. . .... . .. ... ....... 244
Figure 11: Example of Sand racemate physical mixtures
[(75-25%)-2 4 h ou rs-Lab shaker-Room
Temperature l .
. . 245
Figure 12: Thermogram of a directly compressible
(+) S-ibuprofen formulation [24 hours of
mixing ].
. .246
Figure 13: Thermogram of a directly compressible
(+)- S- ibuprofen formulation [72 hours of
mixing l . .
...................
Figure 14: Thermogram of a ground
. ..... 247
C~)- S-ibuprofen
tablet obtained at 600 lbs ................ 248
Figure 15: Thermogram of a ground
(~)- S-ibuprofen
tablet obtained at 1200 lbs .......... . .. . . 249
Figure 16: High Purity Indium Melt ......... .
. . .. 250
Figure 17: (+)-S-ibuprofen Purity Run ............ . ... 251
Figure 18: (+)- S-Ibuprofen Purity Determination ...... 252
Appendix D
Figure 1
Disintegration Time of IBU Formulation .... 255
Figure 2
Effect of Compression on Hardness.
xvii
.. 256
LIST OF PUBLICATIONS AND PRESENTATIONS
Manuscript
I
has
been
published in the journal Pharm
Acta Helv 66(2 ) :34-43. ( 1991 ) and was presented. in part. at
the 1989 A.A.P.S. meeting in A"lanta.
Manuscript II was presented in part at the 1990 A.A.P.S.
nati o nal
meeting in Las Vegas and will be submitted "o Int .
J. Pharm.
Manuscript
III has been accepted for publicati on in the
journal Chirality 1991.
Manuscript
IV
has been accepted for publication in the
journal Drug Dev . and Ind . Pharm. 17(5 ) , 1991.
Manuscript
V
will
be submitted for publicati on in the
J ournal of Pharmacy and Pharmacology
Manuscript
VI
will
be
submitted for publication as a
techn ological note in the Journal de Pharmacie de Belgique.
Manuscript VII has been published in the journal
Clin . Res. Practices and Reg. Affairs 8 ( 2): 123 - 151.
and
was
presented.
in
part ,
( 1990 )
at the GRASP ' 90 meeting in
Chapell Hill, NC .
xviii
"The hard Things We Do Immediately.
The Impossible Takes A Little Longer "
Pete DeMaria
Ciba-Geigy Corporation
xix
To my Parents and to Teri.
To Colette.
xx
SECTION I
INTRODUCTION
The
production
o f pharmaceutical compressed tablets is
very common despite the fact that our understanding
pr ocess
is
by
no
means
than
based
on
an
intuit ive
a rati onal function . Thus the processing tech -
nology may or may n ot be opti mal .
problems
the
complete. In many instances. the
choice of formulati on variables is
rather
of
As
a
result
there
are
in fully validating the process as required by the
Food and Drug Administrati on ( FDA ).
Ibuprofen
is
currently
administered as a racemate
oral dosage forms are manufactured
using
wet
and~
granulation.
This technology improves the flow and compactibility of pow ders by increasing
effect
the
particle
size
and
cohesion .
The
of processing on crystal and granule characteristics
have been carefully discussed in the
tribution
of
particle
size
literature .
depends
The
dis-
substantially on the
binder solution (1). its volume ( 2 ) . the mixing time ( 3 ) and
many
other
factors ( 4 ). The drying stage may have critical
effects on the hardness ( 5 ) and other physical properties of
the
tion
granules ( 6 - 9 ). Several authors have correlated compac parameters
to
the
granule
characteristics
( 10 - 12 )
Similarly . the properties of pharmaceutical tablets such as
dissolution ( 13 - 16 ), disintegration time
( 14 ) ,
( 16 )
or
hardness
were related to the primary processing technology. To
date it is generally recognized that some characteristics of
2
the
r a~
materials are responsible f o r c ertain a spe cts o f t he
pr oc e ss in g
b ehavi o r
( 16 - 21 ) .
Alth ou gh
g ranu l e
g r owth
mech anisms have been studied rather successful l y ( 17.21 - 25 ),
the the o reti c al models proposed fail
biguities
of
of
describing
behavior
the
relationship
some
am -
betveen
the
of povdered drugs and tablet processing
( 26-29 ) . These studies addressed the
of
explain
the ibuprofen formulati o n. There are a number
articles
molecular
to
crystal
modifications
carbamazepine(28), sulfanilamide ( 29 ) . phen obarbital(30 ) .
aspirin ( 31 ) or many other drugs ( 32 ) but at this time. there
are no such publications for ibuprofen.
The
development
tiinflammatory
of
agent
ibuprofen ,
( NSAI ),
a
non
steroidal
an -
with several doses strengths
presents many challenges (33) to the formulators.
Yet.
new
challenges emerge from the recent possibility of manufactur ing the biologically active
stereoisomer
[( + )- S- ibuprofenl
using an economically viable chemical synthesis .
During the course of this study ( Spring 1990 ) .
able
to
the
time
we
vere
obtain a substantial amount of (+)- S-ibuprofen. At
several
prestigious
pharmaceutical
companies
( Johnson & Johnson. Merck Sharp & Dohme and McNeil ) were ac tively investigating possible synthetic routes to obtain the
(+)
isomer
in
a
large
scale
fashion
and presently the
benefits as well as possibilities of formulating
pound
are
under
heavy
this
com-
scrutiny . This general interest in
stereospecif ic drug development
3
meets
the
new
trends
in
regu l a to r y
bodies .
es pecially t he FDA under the leader s h ip
of Carl Peck. in pr omocing the pharmaceutical devel opmen t of
pure pharmaco l ogicall y active enanciomers.
With the exception o f
currently
used
as
Napr oxen
( Sy ntex ).
all
pr o fens
antiinflammat or y products in the United
States are marketed as racemates
( 34 ).
cases,
optical isomer seems to be
the
dextrorotary
responsible
for
stereospecific
the
che
S
therapeutic
most
activity
of
that
chese
is
the
inhibition of the cyclooxygenase and further
prostaglandin
reports
or
In
have
been
these aryl propionic
stereoisomer
could
synthetase.
Various
published.
acids
suggesting that for some
bioinversion
take
pharmacokinetic
place
in
of
the
vivo
by
of ~
inactive
enzymatic
mechanisms ( 35 ) .
While
it
was my intention to investigate the relations
between processing and ibuprofen
level,
(in
at
at
a
molecular
order to improve and optimize ics formulation ),
it would have
racemate
crystals
been
unreasonable
to
consider
sol ely
the
this stage. Therefore during the spring 1990 I
decided to redirect my research work
with
an
emphasis
comparing the rac-ibuprofen to ( +) -S - ibuprofen crystals.
a result, a combination of several "expertise "
some
on
As
ex -
clusively reserved to basic research (i.e. single crystal XRay diffraction ) were applied to the
study
isomer.
ibuprofen
The
hypothesis
modified during
were
formulation
that
and
4
in
turn
of
the
active
crystal was
influenced
the
pr ope r tie s
chat .
of
using
resultin g f o rmula c i o n s . We a l s o hyp othe s ized
similar
techniques
the
f o rmulati on
of
the
b i olog ically active scereoisomer was indeed po ssible.
My review of the
predicting
or
published
understanding
literature
the
indicated
molecular
behavior
ibuprofen under processing had never been reported.
chirality
issue.
most
of
the
reviews on the topic approached the
problem from a pharmacodynamic or drug metabolism
view
On
that
point
of
( 33 , 36). There were no published reports investigating
the possibility of developing the pure ibuprofen enantiomers
nor
addressing the issue of stereospecif ic drug development-
in terms of molecular pharmaceutics. It is believed that
this approach is a unique concept in the development of pure
enantiomers that can be used in many othe r comparable cases.
The specific obj ectives of this research work were
1 ) to demonstrate the effect of ibuprofen crystal on the
processing and biopharmaceuti cal properties of resulting
formulations
2)
to
assess
crystal
distortion
qualitatively
and
quantitatively during formulation
3 ) to use these crystal properties in the development of
the pharmacologically active (+)- S-ibuprofen
4)
and
to
compa re
the
molecular
racemate and the S enantiomer
5
pharmaceutics
of
REFERENCES
( 1)
J. W. Wallace. J.T. Capozzi. R.F. Shangraw.
Pharm.Tech. Sept 1983.
(2)
A. Mehata . K. Adam s. M.A. Zoglio. J.T.
Carstensen , J.Pharm.Sci.
(3)
66(10 ) :1462 -1 464 (197 7 )
H.M. Unvala . Joseph B. Schwartz and R . L . Schnaare
D.D.I.P. 14 ( 10). 1327 - 1349 ( 1988 )
(4)
N- 0 Lindberg , C . Jo n sso n and B . Holmquist
D. D.I.P.11( 4 ). 917-930 (1 985 )
(5)
Z.T. Chowan ,
J.Pharm.Pharmacol.
32:10 -1 4 (Jan )
1980
(6)
P . Arnaud et al .
Pharm.Acta . Helv .
58 (1 1 )
290 - 297 (1 983 )
( 7 ) L.Benkerrour. F. Puisieux. D. Duchene .
Pharm.Acta . Helv . 57 ( 10 ) :301-308 ( 1982 )
(8) H. Leuenberger.
Int.J.Pharm.
12:41-55 (1 982 )
( 9 ) Ian Kryce, David G. Pope , J ohn A. Hersey
Int.J.Pharm. 12:113-134 ( 1982 )
( 10 ) Z.T. Chowan and Y.T. Chow
Int . J.Pharm . Tech . &
Prod.Manuf. 2 ( 1 ) :29-34 ( 1981 )
( 11 ) F.A . Menard , M. G. Dehdiya , and C.T. Rhodes
D. D.I.P. 14(11 ), 1352 , 1988
( 12 ) R.N. Chilamkurti . J.B. Schwartz and C.T. Rhodes ,
Pharm.Acta .Helv.
58:253 ( 1983 )
6
( 13 ) Z . T . Ch owan and L Palagy i
J. Pharm.Sci.
67 ( 10 ) :1385 - 138 9 (1 978 )
(1 4 ) J. P. Rern on and J. B. Schwartz
D.D . I.P.
13 ( 1 ) .
1- 14 ( 1984 )
( 16 ) T . M. J ones J. Pharrn . Pharrnaco. 31, 17-23 ( 198 0 )
( 17 ) J . Carstensen et al.
J.Pharrn.Sci .
65:992 - 997
( 1977 )
(18 ) H. Vrornans. A.H. DeBoer. G.K. Bolhuis and C . F. Lerk
Acta Pharm. Suec. 22:163-172 ( 1985 )
(19 ) H. M. Unvala. J.B. Schwartz and R.L. Schnare
D.D.I.P. 11(14):1327-1350 1988
(20 ) Z . T. Chowan and A.A. Amaro
D . D.I.P . 8 ( 14 ) :1 079-1106 ( 1988 )
(21 ) zoglio M.A. et al ,
J.Pharm.Sci.
65:1205 - 1208
( 1978 )
(22 ) Newitt and Conway - Jones ,
Trans.Intn.Chem.Eng.
36: 42 2. (1958)
(23 ) H. Leuenberger , H-P. Bier. H.B. Sucker
Pharrn.Tech. 60-66 , June 1979
(24 ) A. Stamm and L. Paris
D.D.I.P .. 11(2&3). 333-360 ( 1985 )
(25 ) Rumpf H. Chem.Ing.Tech. , 30:144-158 (1 958 )
( 26 ) H.G. Brittain
D.D.I.P. 15(13), 2083-2103 (1989 )
(27 ) C . Lefebvre. A.M. Guyot-Hermann , M. DraguetBrughmans , R. Bouche and J.C. Guyot
7
D.D.I.P.
12 ( 11 - 13 ) . 1913 - 1927 ( 1986 )
( 28 ) Y. Nakai D. D. I. P. 12(7 ) . 1017 - 1039 (1986 )
( 29 ) o. Cruaud. D. Duchene. F.
Puisieux. A. Chau vet et
J. Masse J. Pharm.Belg. 36(1 ) . 15-20. 1981
(30)
s. Kopp. c. Beyer.
E.
Graf. F. Kubel and E . Doelker
J. Pharm.Pharmaco .. 41:79-82 ( 1989 )
( 31 ) M.P.
Summers. R.P. Enever. and J.E. Carless
J.Pharm.Sci (66)8 11723-1175 (1977 )
(32) A Chan and E. Doekler D.D.I.P.
11(2&3) 315-332
(1 985 )
(33) Robert M. Franz
U.S.
Patent
~
4.6 09.6 75. Sept.2.
1986
( 34) J.M. Mayer
Acta Pharm.Nord. 2(3). 196-216. 1990
( 35 ) A. J. Hutt and J. Caldwell
J.Pharm.Pharmaco. 35:693-704 , 1983
( 36 ) D.B. Campbell
Eur.J . Drug Met .Pharmacok . 15 ( 2 ): 109-125. 1990
8
SECTION II
9
MANUSCRIPT I
INFLUENCE OF DIFFERENT SOURCES
ON THE PROCESSING AND BIOPHARMACEUTICAL PROPERTIES
OF HIGH DOSE IBUPROFEN FORMULATIONS
10
ABSTRACT
It is known that depending on the manufacturing and
synthetic
processes. drugs may exist as different forms. As
a result. physico-chemical properties.
teristics.
intrinsic
dissolution
compression
and
charac-
bioavailability may
vary substantially . The purpose of this study was to
tigate
the
effect of different sources of ibuprofen on the
pr ocessing of tablets and on their properties .
phasis
of
this
Another
granulation
parameters
and
to
related
eluding
manufacturers
X- ray
calorimetry.
and
to-
the behavior of final
tablets. Commercially available ibuprofen was obtained
different
em -
work was to rati onal ize one or s everal key
characteristics of the raw material as directly
wet
inves -
from
a preformulation program , in -
crystallography.
differential
scanning
scanning electron microscopy , determinati on of
particle size distribution and fl owability , was performed to
characterize the raw material. Granules were prepared with a
planetary mixer and liquid requirements
were
for
the
end - point
obtained by monitoring power consumption. Tablets were
manufactured on Stokes rotary and single punch
presses .
Data
acquisition
interfaces produced compression
data for each formulation. Granules and final
analyzed
for
uniformity.
variance
and
instrumented
hardness ,
dissolution
profiles
Statistical
evaluations
using
tablets
were
and content
analysis
of
multiple comparison procedures were performed
11
on
the
re s u lts
vari abili t y
to
be tween
d ecermine
the
independent
s i gn if ic an c e
of
the
paramecer s. The ibuprofe n
tested was f ound c o be a unique p o l y mo rphic form
wi t h
so me
diffe r enc e s in the external c r y stal linity . The particl e s ize
characteristics of the material als o allowed a
tion
between
differenti a-
s ources and although there was no difference s
in dissolution patterns or content uniformity. particle size
was
f ound
to
account for 50% of the variability in tablet
hardness . Tw o sources of ibuprofen with l ower mean
size
sh owed
particle
significant variations in end - p oint liquid re -
quirements resulting in variable tablet crushing strength .
I NTRODUCTI ON
It is now widely recognized that grade
of
the
starting material can be responsible for major dif -
ferences when processing
dosage
variability
forms.
Inadequate
and
formulating
( 1)
oral
solid
control of the synthetic process
can lead to the production of different polymorphs or
tal
forms
having
variable
exhibiting differences in
behavior
(3)
or
wet
intrinsic
bioavailability
granulation
crys-
dissolution
(2) ,
and
compaction
parameters ( 4 ) .
Those
phenomena have been frequently addressed in the pharmaceuti cal
literature.
For
example.
the
pharmaceutics resulting from grinding ,
changes
in
compression
molecular
and
in
general processing, have been discussed extensively ( 5 ) . The
12
kn owl edge o f phy s ic o - c hemical charact eristi cs of the
s car t -
i n g materia l s i s critical f or the f ormulat o r (6 ) . es pe c iall y
when high dose drugs are formulated. where the nature of the
acti ve
itself can also influence subs tantially the pr ocess -
ing of the final products.
pharmaceutical
To
specifications
date .
materials used in a formulation .
small
changes
may
occur
although.
recommend
it
between
has
been
(7 ) .
As
a
result,
shown
thac
products from different
manufacturers and within products provided by
plier
industrial
tw o suppliers for
a
same
sup -
formulation problems arise when
processing the correspondin g formulations
(4 ) .
It
is
the-
role of the formulating pharmacist t o understand and monitor
the transfer of technology involved in switching sources
or
suppliers. in order to avoid nonideal or unexpected behavior
during large scale
good
quality
of
manufacturing.
theref o re
insuring
a drug product to guarantee the patient ' s
safety. Ibuprofen, our model compound. is a widely used
Non
Steroidal
the
OTC
Antiinflammatory Agent. Different polymorphs
of recrystallized ibuprofen have been shown to exhibit vari able
extent
and rate of biological absorption (2 ) and this
molecule could exist under different crystal forms depending
on
the
synthetic process. As many therapeutic applications
of ibuprofen may become available for children at
levels
(B ).
minor
changes
dose
in the crystal structure could
result in dramatic changes in the pharmacologic
of
low
disposition
this compound. The objectives of this investigation were
13
to determine if several sources of ibupr ofen could exist
dif:erent
forms
and
exhibit
as
variations in their physical
pharmacy p r of ile . A correlation between s ome characteristics
of the raw material. processing parameters and properties of
the final products was studied. In further studies . the
fect
of
processing
on
ibuprofen
is investigated in more
detail and at a molecular level. For example.
shown
that
particle
it
has
been
size (9). particle morph ology ( 4 ) and
surface area ( 10 ) of the
significantly
ef-
granule
starting
formation
material
and
can
binding
influence
pr operties
during compaction .
Fractional
factorial
designs were utilized to in-
vestigate the extent of variability
between
the
different
sources and multiple comparison procedures were performed on
the results for processing parameters and
characteristics.
This
study
is the first of a three paper
serial investigation leading to
aspects
of
the
to
optimization
of
some
ibuprofen formulation. The information obtained
in this work served to correlate key
tics
biopharmaceutical
product
properties ,
material
characteris -
isolate them from processing
parameters and support recommendations regarding the use
of
different validation procedures for various raw materials if
they are provided by different suppliers as it is
in most industrial pharmaceutical settings.
14
the
case
EXPERIMENTAL
MATERIALS
Ibuprofen was obtained
tributors
through
intermediate
dis -
and manufacturers. The identity and origin of the
five different sources analyzed in this study are
presented
in table I. Monobasic phosphate and sodium hydroxide were of
analytical grade and obtained through the Fisher
Company.
Ibuprofen
obtained from the
States
standards for calibration purposes were
Drug
Pharmacopeial
using
Standard
Division
Convention
granulations containing
Povidone
Scientific
Fast
flow
in
of
the
Rockville ,
Lactose
United_
MD.
Wet
(Schieffield).
( GAF Co. ). Explotab ( Edward Mendell ) were prepared
purified
water.
Lubrication
was
performed
with
Magnesium Stearate ( Fisher Scientific Co. ).
METHODS
GRANULATION PROCESS
Ibuprofen
strengths , in
amounts
of
was
order
active
formulated
to
study
the
at
three
effect
of
different
increasing
on the processing and biopharmaceutical
properties of the final
products.
tigated
may not be the most appropriate for
in
this
work
15
The
formulation
inves-
large scale manipulations. Nevertheless.
combinat ion
convenien~
variables i nvolved in
available.
the
wet
it
wa s
the
most
giv en the complexity of
granulation
process.
thus
unablin g the study of pure s ource effect.
Blends composed of the
disintegrant
and
active.
the
diluent.
the binder were dry mixed for ten minutes
in a Turbular mixer. The mixture was transferred on
strumented
the
planetary
an
in-
mixer ( Kitchen Aid Model K5 - A Hobart )
interfaced with an IBM personal computer. A data acquisiti on
software from Extech Co . all owed the recording of Power consumption . The pre-mixed powder was then
planetary
dry
mixed
using
a
peristaltic
ml / min. with five seconds
granulation
the-
mixer allowing the Watt - reading to stabilize t o a
baseline (11) . The granulating fluid was added to
ture
in
was
pump
at
the
mix-
a rate adjusted t o 10
interruption
every
minute.
Wet
proceeded until the end point. In order to
have a uniform distribution of liquid bridges the mixer
was
stopped three minutes after the end-point.The power data was
plotted against the granulation time
the
real
which
corresponds
to
wet mixing time and the volume of water added was
recorded
Although we fully understand
time
rate at which the liquid is added, it was not the
and
scope of this investigation to
granulating
process
study
Granules
the
this
were
12
aspect
then
screened through a number 8 mesh screen and
during
importance
of
gently
dried
at
of
the
hand
40°c
hours in a convection oven to reach a one to two
16
percent
final
screened
moisture
through
The
dry
granules
were
a • 16 mesh screen. mixed for ten minutes
with the lubricant and
tegrant.
content.
the
appropriate
amount
of
disin-
and compressed into tablets on an instrumented B-2
Stokes rotary press. An instrumented single punch press ( F3 Stokes)
was used to validate 350 mg tablets with acceptable
tensile strength ranging from 8 to
force
was
15
Sc.
The
recorded and the different formulations compared
using this parameter. Another experiment was
the
compaction
hardness
conducted
and
of tablets , made at the same level of compac-
tion. was recorded . Three different
levels
of
compression-
force were investigated as some biopharmaceutical properties
are known to be
proportional
to
the
compaction
behavior
( 12 ). Compression data were recorded as fingerprints of each
formulation. Table II shows the starting formulation.
PRKFORMUI.ATION AND PHYSICAL TESTING
Analytical
testing
was
stages of the study and during the
The
five
d if ferent
sources
performed
at different
formulation
process ( 8).
of raw material were sc reened
through a so l id state preformulation program to characterize
the
active . Thi s preliminary testing included the following
analysis:
-particle size analysis
-surface area determination
17
-differential scanning calorimetr y
- x - rays diffraction patterns
- scanning electron microscope
phot ographs
This
reduced physical pharmacy profile was under -
taken to detect any remarkable differences between
The
particle
suspensions
Analyzer
size
( 10
was
characterized
mg ml )
with
1
model
2334A
a
Brinckman
This
method
technique
Analyzer
from
ibuprofen water
Particle
a
gas
was
available
on
a
Quantasorb
Thermal
Analysis
microscopy
NY .
photographs
and
of
Ciba-Geigy
granules;
a
flow
powder
meter
on
density
variance
in
was
recorded
with
was
ibuprofen
flow linearity index was
an
Apparent
Erweka tap density
tester with 2000 taps. Statistical evaluation with
of
scanning
of the raw materials were
derived from the flow charts when applicable (13).
tapped
Perkin-
Flowability measurements were obtained with a
custom designed recording powder
powders
a
Series 7 interfaced with a Perkin-
performed by the Analytical R&D services
Ardsley.
Sorption
Quantachrome , NJ. The melting processes were
Elmer P7500 E computer. X-Ray crystallography and
electron
adsorption .~
and calculated using the B.E.T. equation.
measured by differential scanning calorimetry on
Elmer
Size
using a laser light scattering tech -
nique . Surface area was determined using
monolayer
on
sources.
analysis
used to differentiate between independent
18
variables and support interpretations .
was
measured
using
a
Final
granule
size
conventional sieve method. the size
distributions were compared by plotting the percentage over size
vs the amount of active on the screen .The median point
wa s used as the mid - point to compare the different
tio n s.
Mois ture
determined
Co.
NJ.
contents
formula -
and loss on drying profiles we re
on a Computrac Mo isture Analyzer from
Computrac
Tabl et c rushing s trength wa s measured on an Erweka
Aut omat ed Hardness tester. Dissolution testing was performed
with an Easylift model 63 - 734-100 from Hanson Co. The method
consisted of a rotating paddle at 50 RPM in
phosphate
buffer
dilutions.
the
nanometers.
working
This
wavelength
technique
was
was
l o ss
on
drying
curves.
granulation end point, f l owability
Ibupr ofen
cores
pH
7.2
USP ~
adjusted
to
26 4
applied successful ly in a
previ ous work (1 4 ). Gr anulati ons were
content,
a
at 37°c . In order to avoid time consuming
tested
for
moisture
liquid requirement s for
and
size
distribution.
were analyzed f o r hardness and dissolution
pr ofiles in correlation t o compaction forces.
RESULTS AND DISCUSSION
PREFORKULATION
The
preformulation profiles of ibuprofen showed
several differences in the solid
19
state
characteristics
of
the
various
sources.
Neveroheless .
it is underst ood that
anal y tical testing can present s o me variations and
to
be
extremly cautious with
one
has
interpretation. For example.
the ibuprofen - water suspensions used with the particle
size
analyzer may be a fraction not representative of the overall
sample
populations
and
each
analysis
was
performed
in
triplicate. The mean particle sizes and log-normal frequency
size distributions are given
respectively.
in
table
III
and
Both measuremen•s were performed on ibuprofen
particles suspended in an inert solvent and
ten
minutes
micronized
but
experiments
creasing
par-
The
characteristics.
including a study of the effect of in-
micronization
performed.
primary
did not generate sufficient energy to dislocate
the primary crystals related to processing
Further
for
to obtain uniform suspensions. we believe that-
the micronization divided the aggregates into
ticles
figure
time
on
particle
size
will
be
average particle size was obtained from the
surface weighed equation (15). It is appreciated
that
sam-
pling may also be subject to certain variation in this case.
The Francis High density had the largest mean particle
with
size
a right skewed tendency. The Cheminor source seemed to
exhibit a narrow distribution with an
intermediate
average
particle size. The Boots and Francis Low density sources exhibited the lowest mean particle si ze ,
distribution
of
almost
micrometers.
The
surface
100%
area
20
of
with
the
results
the
sizes
of
narrowest
below
25
unmicronized
ibuprofen
samples
are
presented
surface measurements of average
used
in table IV . The v olume -
parcicle
size
are
mostly
for pharmaceuticals because they are inversely propor -
tional to the specific surface. Combined with
measurements.
it
surface
allows an accurate evaluation of physical
properties of medicinal powders. Some BET resu lts
expected
as
Francis
were
n ot
low D. and Boots . which exhibited the
l owest mean particle size,
values.
area
had
intermediate
surface
area
On the other hand , Francis High D. which showed the
largest mean particle exhibited the highest surface area in dicating
a
very broad size distributi on in accordance
the frequency curve. The apparent tapped densities
in
figure
ibuprofen
2
powder
summarize
based
the
on
micromeritics
the
previous
with~
reported
behavior
of
experiments .
According to the packing theory , as a result of size characteristics and surface area, Franc is High
largest
density.
D.
exhibited
the
Boots and Francis Low D. showed the smal-
lest apparent density. probably indicating the unif ormity of
the
shape
distributions. Figures 3-4 are scanning electr on
microscopy photographs of ibuprofen raw material .
roscopic
observation
allowed rational
sources
as
of
The
mac-
X500 magnified ibuprofen crystals
conclusions
on
the
differences
between
sorted by particle size , BET values and density
results . There was no significant visible difference between
Ethyl
and Boots , which exhibited lamellar needle type crys -
tals . On the other hand , Francis low D. has a
21
very
uniform
distribu•ion
of the smallest needle type particles. Francis
high D . has the largest
rounded
macrocrystals
with
small
microcrystals and Cheminor exists as laminated square plates
of intermediate size. The surface of
macrocrystals
seems
to
evaluation
High
density
be very irregular. probably an ex-
plaination of the high surface
visual
Francis
area
value.
Clearly.
this
indicated that the final crystallisation
step of the synthetic process could be
very
different
for
the various sources leading to differences in crystal formE.
This observation could not be predicted from the BET results
which
did
not
show
any
significant
differ e nces
between~
Cheminor.Ethyl and Boots but appeared to vary from the
face
area
sur-
of the two Francis sources as indicated by an F-
test. The thermal
analysis
gave
more
information
on
the
crystal structure. All the DSC curves exhibited a unique endotherm in the
range
75-76°c
with
enthalpies
of
fusion
ranging from 113 J i g to 118 J i g. An example of a typical DSC
profile of ibuprofen is shown in figure 5
points
and
all
melting
are reported in table V. An analysis of variance did
not show any significant
differences
indicating
that
the
ibuprofen tested do not exist as different polymorphic forms
and the internal crystal structure
sources.
is
equivalent
for
Nevertheless. the enthalpies of fusion are statis-
tically different. The results of an F-test , shown in
VI.
all
suggest
that
ibuprofen
has
22
table
variable crystal surface
structure. as anticipated from the scanning electron micros copy
photographs
and
the
X-ray
figure 6. The gene ral shapes are
diffraction
similar
for
patcerns in
all
sources
however. for low angles of the spectra. the intensity of the
first deflection peak varies between materials indicating
difference
in
external crystallinity.
Except from Francis
high D. which exhibited a flow index of 18.5 ( a
of
a
flow
index
19 is representative of good flowability properties ) all
ibuprofen powder did not flow through
the
orifice
of
the
flowmeter.
The wet granulation process divided the sources into
groups:
raw
end-point
ments
materials with l ow liquid requirements for the
~f igure
( figure
two~
7) and crys tal s with high liquid
8).
It
require-
is appreciated that the power scale
does not represent means of differentiating between sources.
rather the general shape of the power consumption curves was
analyzed in detail with emphasi s
given
to
the
inflection
points where the torque required to rotate the paddle at the
same speed within the wet granules increased
suddenly.
The
increase in wattage was attributed to a change in the physi cal state of the wet mass, which we associated with the end point
of
the granulation also represented by the arrows on
the power consumption curves. The addition of water was
terrupted
in-
upon observat ion of this increase. The arithmetic
average of liquid requirements are reported
23
in
table
VII.
This
categ orizati o n
between
sources was not perf ormed ar -
bitrarily but using an analysis of
variance
mulciple
which
comparison
procedures
and
Duncan ' s
divided
the
raw
materials into two groups with a 95% confidence level ( table
VIII ) .
Ibuprofen
powders with high liquid requirements and
ibuprofen sources with low liquid requirements. As a
quence
of
end
the wet mass. measured ac 65°c before the
hibited
differences
recorded
in
drying
step
in
the
channels
and
pore
granules in which the moisture migrates to
evaporate.
Granules
indicating
tortuosity
the
of
surface
force
was
applied
comparison
and
reported
as
in
table
X.
particle
size
or
of active. Although, not using an interactive model.
some inferences could be made on the size
ANOVA
com-
procedures were applied to evaluate the
effect of independent variables such
amount
to
(10-15 KN ), the hardness of the
corresponding tablets measured
Several
the~
from different sources were mixed with
lubricant for ten minutes and tabletted. An acceptable
paction
ex -
table IX. Loss on drying
profiles ( figure 9 ) were also differ ent probably
variations
conse -
point requirements. the moisture content of
in
table
XI.
effect
with
The sum of square due to the particle
size of the raw material demonstrated that about 50% of
hardness
variability
the
among
granulations
was
the
due to dif -
ferences in particle size fraction and distribution
of
the
starting material. The differentiation and classification of
24
the sources by liquid requirements was confirmed on the compacti on versus hardness investigation. The crushing strength
of tablets made with Boots and Francis Low
higher
sources
did
powders.
not
Dissolution
profiles
of
ibuprofen
indicate any significant differences in the
release from the various formulations and based on the
vious
was
for the same level of compression forces as compared
to the other
cores
D.
preformulation
experiments,
we
do not forecast any
problems of biological availability with
ferent
sources.
Further
pre-
the
use
of
dif-
studies will include the analysis
for the enantiomeres
of
various
profiles and calibration curve for the
dissolution
ibuprofen .
Figure
ibuprofen cores. In o rder to preserve
figure.
the
percentage
the
10
shows
clarity
of
the.
the
dissolved after 70 minutes are not
represented as they did not
bring
further
information
on
possible differences between sources. Figure 11 shows a complete dissolution profile of ibuprofen cores made
source
(Cheminor ) .
with
position in the tablet formulation (100% intragranular)
responsible
for
the
slow
of
are
ibuprofen release rate. Further
studies include the optimisation of
position
one
The low disintegrant level ( 1% ) and its
the
concentration
and
the disintegrant when the active / diluent ratio
is increased.
l
25
CONCLUSIONS
The process of chemical synthesis
drug
substances
and
or
isola0ion
although designed to produce materials of reproducible
chemical
purity ,
of
excipiencs used in tablet formulation
may
high
not necessarily result in batches of
product with equivalent
physico-technical
properties .
The
nature of the solvents or the concentration of intermediates
present in the liquors used for crystallisation
particle
can
affect
morphology including crystal dislocations. surface
rugosity and surface area. Those
reflected
in
solubility
significant
or
crystal
properties.
differences
forms.
can
in
although
n ot.
melting points.
influence
compression
characteristics and possibly the amount of granulating fluid
required to produce a coherent mass. This conclusion
lines
the
importance
under-
of the preformulation and in-process
testing when using different
suppliers
and
possibl y
dif-
ferent batches of the same material.
As the different ibuprofens tested did
major
not
variations in physical-chemical properties and d o not
exist as different polymorphs, various sources of
tive
of
exhibit
this
ac-
could be used in oral solid dosage forms without risks
altering
during
ferences
liquid
the
the
biological
course
were
of
detected.
requirements
for
availability.
this
study
several important dif-
Possibilities
the
26
Nevertheless,
end-point.
of
variations
which
could
in
be
predicted from particle size anal ysis and
measurements.
have
tabl ets
are
in
a
is
of
higher
differences
end-p oint
a
crit ica l
sources.
As
slight
hardness
differences
a consequence. when consistent varia -
tions between two sources can be detected
state
preformulation
program ,
sity.
crystal
surface
size .
may
in
liquid requirements ) .
coatability functi on may be affected by
morphology
commercial
f ormulati on techn ology. Certainly , s ince dif -
result
between
all
hardness.
ferent sources led to substantial
( as
date
coated and the ease of coatability.
mostly relat ed t o friability and
parameter
density
to be kept in mind as they affected the
final hardness of ibuprofen cores. To
ibupr of en
apparent
through
a
solid,
key parameters such as den area
and
crystal
s urface
be used to predict pr oblems in the formula -
tion behavior. It is appreciated
that
the
conclusions
of
this work d o not advantage one s ou rce over another. since at
any
moment
parameters
of
the
can
formulati on
of
the
a
those
strict
and
observations
detailed
and
the
pharmacy
suggest
usefulness of two validation procedures or two standard
operating procedures specific to each one
thus
underline
physical
profile for materials from different suppliers
the
processing
be modified to obtain final products in ac-
cept able ranges , rather
importance
stage
avoiding
costly
unexpected
during large scale operations.
27
of
the
sources.
pharmaceutical behavior
Akn o wled geme nts
(
One of us ( AJR ) thanks Ciba-Geigy Co. for
of
a
summer
fellowship
and
laboratories .
award
the opportunicy of using the
equipment available in the Pharmaceutics and
Technology
the
Pharmaceutical
we also would like to acknowledge
the support and expert advice of many Ciba - Geigy scientists.
especially
Mr .
Louis
Savastano
made this project possible.
28
whose careful supervision
REFERENCES
1. Hakan Nyqvist
Dru g Dev .Ind . Phar m. 15 ( 16&17 ). 957 - 964 (1989)
2. N. Udupa
Dr u g Dev . Ind.Pharm . 13 ( 15 ). 2749-2769 (1987)
3. M. P . Summers. R.P. Enever and J.E. Carless
J.Pha rm. Sci . 66 (8), 1172-1175 ( 1977 )
4. C. Lefebvre. R. Bouge. J. Ringard and A.M. Guyot-Herman
"Modification de l ' aptitude a la granulation d ' un
principe actif imputable a l ' augmentation de la
solubilite de surface de ses particules.
Pre sented
in Agpi Paris (1 989 )
5. C. Lefebvre . H . M. Guyot-Herman. J.C. Guyot. R. Bouchet ,
and J . Ringard
Dr u g Dev . Ind . Pharm . 13(9-11 ) . 224-235 (1988)
6 . Harry G. Britain
Drug Dev . Ind . Pharm . 15(13 ), 2083-2103 ( 1989 )
7. T . M. Jones
J.Pharm . Pharmaco. 31, 17-23 ( 1980 )
8. P.D. Walds on , Gary Galleta. Nancy Jo Brade n and Laura
Alexander
Acta . Pharmaco.Toxico . 59 ( Supp . V).
155-
(1986)
9 . H. Vromans. A. M. DeBoer , G.K. Bolhuis and C.F . Lerk
Ac ta Pharm . Su ec . 22, 163-172 ( 1985 )
10 . Jue Chen Lin et al
29
Drug Dev .Ind.Pb arm. 13 ( 12 ) . 2087 - 211 0 ( 1987 )
11. A.
Stamm and L.
Paris
Drug Dev.Ind .Pbarm . 11 ( 2&3 ). 333 - 360 ( 1987 )
12. Kumar A. Khan and C.T. Rhodes
J . Pharm . Sci . 65 ( 12). 1835 - 1837 (1976 )
13. R . P . Jordan and C.T. Rhodes
Drug Dev . Ind . Pbarm .. 5. 151-
( 1979)
14 . A.J . Romer o. L.T. Grady and C.T. Rhodes
Drug Dev . Ind . Pbarm . 14 ( 11 ) . 15 49 -158 6 (1988 )
15. I.e. Edmonson
"Advances in Pharmaceutical Sciences· Vol.2
Edited by H. S . Bean. J . E. Careless and A.M. Beckett
Acade mic Pr ess. London 1967 pp-95
16. Z.T. Chowan and L.
Palagyi
J . Pbarm . Sci . 67 (10). 1385-1389 (1 978 )
30
Table I
SOURCE AND ORIGIN OF IBUPROFEN
Origin
Name
Italy
Italy
U.S.A.
U.S.A.
India
Francia Low Density
Francia High Density
Boots
Ethyl
Chemin or
31
Abbreviation
Franl.D
FranH.D.
Boots
Eth
Chem.
Table II :
STARTING FORMULATION
57
36
6
IBUPROFEN
FAST FLOW LACTOSE
PLASDONE
EXPLOTAB
LUBRICANT
GRANULATING FLUID (WATER)
32
"'
"'
"'"'
"'
q.s.
Table Ill
MEAN PARTICLE SIZE
SOURCE
<.. cRoHsl
lurfac•-Numb.,.
Mean
FRANCIS LOW D.
BOOTS
ETHYL
CHEMINOR
FRANCIS HIGH D.
3.17
3 . 58
5 . 23
7.94
10.54
SRINCKMAN PART ICLE SIZE ANALYZER
M1cron1 sed Water Suspensions
33
.....
Surface-Wtlgtlttd
5.03
6.22
18.18
31.07
38.25
Table IV
SURFACE AREA
B.E.T. VALUES
2
(SD) in m /gram
FRANCIS LOW D.
"'
ii>
FRANCIS HIGH D.
ETHYL
BOOTS
CHEMINOR
0.76
0.86
0.34
0.41
0.36
(•) UNMICRONIZED SAMPLES
(0.03)
(0.09)
(0.01)
(0.03)
(0.04)
Table V
MEL TING RANGES
DIFFERENTIAL SCANNING CALORIMETRY
SOURCE
"'
(JI
CHEMINOR
ETHYL
BOOTS
F.LOW D.
F.HIGH D.
ONSET
73.0
73.2
73.8
72.4
73.4
(SD)
(0.2)
( 1.4)
(0.2)
(0.1)
(0.1)
All valuas ara reported in degrees Celsius
MEL TING POINT (SD)
75.4
75.6
76.1
7 5.1
75.3
(0.05)
(0.10)
(0.10)
(0.10)
(0.10)
Table VI
ENTHALPY OF FUSION OF VARIOUS SOURCES (Jig)
ANALYSIS OF VARIANCE TABLE
SOURCES
ERROR
TOTAL
df
SS
MS
F
4
16
20
67.63
57.96
125.60
. ;.91
3.62
4 .67
p
0.01
At the 991' confidence l•vel tt'•«• Is • • lgnUlcent difference In AH'•
o t 3 to 5 J/ g
36
Table VII
LIQUID REQUIREMENTS FOR IBUPROFEN
GRANULATIONS END-POINT
MEAN VOLUME (SD)
SOURCE
VI
-'I
ETHYL
F.HIGH D.
CHEMINOR
BOOTS
F.LOW D.
Liquid requir emen t s for
27.8
31.2
31.5
44.0
58.2
IN ML
(4.2)
(3.4)
(5. 1)
(6.7)
(2.2)
1 7 5 gm batches
The average re por ted was obtained trom ell granulations
(
Table VIII
EFFECT OF DIFFERENT SOUR - cS :
LIQUID REQUIREMENTS FOR GRANULATION END-POINT
DUNCAN'S MULTIPLE RANGE TEST
SIGNIFICANCE LEVEL ex • 0 . 05
27.8
ETHYL
31.2
FHIGHD.
ANT TWO AVERAGE NOT
SIG. .ICANTLT DFFERENT
31.5
44. 0
CHEM.
BOOTS
~ERLINED
58.2
FLOWD.
BY THE SAllE SEGMENT ARE
38
(ML)
Table IX
WET GRANULES: MOISTURE CONTENT
ETHYL
FRANCIS HIGH D.
CHEMINOR
BOOTS
FRANCIS LOW D.
11 %
13 %
18 %
23 %
32 %
Loss on cty1no at 65 degrees C
39
Table X
HARDNESS LEVELS
COMPACTION RANGE (KN)
SOURCE
ETHYL
BOOTS
CHEMIN OR
FRAN.LOW
FRAN.HIGH
10-15:
FORCE(KN)
HARDNESS(Sc)
10 - 15
19-22
22-29
12-14
12-14
11-13
10-12
12-24
19-28
18-25
40
Table XI
AHlyal1 of Variance for tablet hardneaa
Ona degree of lreedoom compariaons
"'....
Source
di
SS
MS
Among granulations
11
3651.1
331 . 9
1
1
127 . 1
390.1
891 .6
427 .5
Ethyl
Size1
Slze2
Size1
Va.
Va.
Va.
Va.
Boote
Size3
Size3
Size2
Within Granulations
1
1
66
852 . 1
-
-
-
F
25 .7
9.8
71 .0
69 . 1
33 . 1
12. 9
S1ze1 : Sieve Frac tion 20/40 ; S1ze3: Sieve Frac ti on BO/ Pan
S1ze2: ibuprofen as 1s
p - value
0 .001
0 .025
0.000
0 . 000
0.000
.
,_
0
~
c:
.
0
:;::
:::s
:e
.!::
VI
Q
I
c
••
:IE
0
0
l""
;""
~
1"
r
<l,,
~
~
+
+
~
+
'~--
,'<l'', q
'\.
'
=
+
'ea~~&. ..
--:~:'~
+
'Q:i
~ .
+
- .. ... . . . ___
..
' l'l
+
- -..............._
0
+
.....,_
\~
+ O<i
'~
\
+"\
\
"
I
It)
0
0
..,0
42
00
figure 2
App.rent Butk •nd T•pped Den11tle1
-
fiml
Bul k
Tapped
100
75
!
i
50
25
0
ETHYL
BOOTS
CHEM.
43
FRANLO.
FRANHO.
Figure 3:
Scanning Electron Microscopy Photographs of
Ibuprofen crystals
I)
FLO:
Francis Low Density
II) FHD: Francis High Density
III) Boots
I V) Ethyl
44
45
Figure 4:
Scanning Electron Microscopy Photographs of
Ibuprofen crystals
I) and II) Cherninor
II) and IV) Ethyl: different batches
46
47
Figure 5: Typical DSC Endotherm of Ibuprofen
Sample: Ibuprofen Boots 187327
Size:
6 . 3000 11g
Method: 5•c1a1n. 60" to 95•c
Co•••nt: 100 •L/a11n N2 I Al 1gned Al pan
DSC
73.eJ•c
119 . 5J/ g
or---~~~~~~-+-.:+-~~~-,,-+~~~~~~~~~-
II>
Cl>
7& . 13•c
60
65
70
75
80
Te•pereture c•c1
85
90
95
F,~ure
6: X- Ray Diffraction Patterns of Ibuprofen
Sources
A. Francis High o.
c.
Cheminor
B. Francis Low o.
D. Boots
E. Ethyl
A
B
c
D
E
49
z
0
j::
<(
.....
::>
z<(
a:
"
I-
"'
.
.....
i=
... :ta:
E
.. 0
c
"
.~z
.2
... 0
j::
ii
:;
.
Q.
c
2
::>
<;
Ill
z
0
(,)
a:
"'
:t
0
Q.
.
"
\
'<l
=
•n•M
50
=
~
•0
~~
~II
I
I
...
.
~
0
~
311
•
I
I
"'
z
0
j::
«(
...J
:::>
z«(
"'
0:
~
co
.......3:
•
"· ...
0:
O>
c
.2
-----...--·--.-~
• ,, <=i
...j::
:I
:::>
...
Cl)
z
0
<..>
0:
.•
.
I
~'
"'
ii
..
(;
' --.,,
,,.
I
0
0
0
51
•E
j::
···
.~z
... 0
...3:
...
,.'~
···
0
0
~
• •, . . ..
'-.._
.. .
...
::s
~
z
..'
';
c
•
(
0
:t
a:
_,
>-
....w
"'....0
<iI
•
:z:
u
0
f
:;:
z
iw
.
:t
0
;i:
"z z0_,
...<a: ...<a:
0
I
I
+
I
t
r
,,
:
I
I
"'......
I
I
I
I
en u:::
0
I
a::
CL
/I
I
,....
I
.!!!'CJ
+
'/
LL Z
1,
>
a:
I
I/
Q
z
I
0
I
"'"'
I
0
...I
I
I
I
I
!
...
0
~
I
I
I
I
I
I
I
I
I
./]
:I
c
,
r
"'
w
...ti
·/
<I
I
/ ,'
,,
...I
0
I
1
I
,
I
<
.
•:+-
1.·
I
I
~
0
;:i
0
-"'
co
ti
E
j::
..
)
/'
I
0
I
I
./]
I
+
I
· 1
I
I
:;
I
I
/ :
I
I
I
.
0
~
0
(M/M)
0.59 J8 iuaiuoo aJnJS!OW
52
0
Q
.:>
...
Q
0
0
...
~)Ci
h•
f~
"'~p'
'2
0 I .
0
I
~Q
z fI
w
I
LL :
o•
a:
0
~
:I
~
u...
~· ·
~~
LL tb'
0 '
'
z•
c
0
0
28
:::>'
-0
"'
'
w
:I
j:::
<I,
I- •
...I
!
O •
"'
cE=
T
0
0
0
0
"'
0311\10SSIO .l.N31::>113d
53
0
0
"'
::E
w
::t:
0
fl)
-----<
0
-1
w
a:
"'
M
0
zw
c
u..
0
.,, !
a:
0
~ Cl.
""'
ii:
:::>
i=
u..
0
z
•"'
0
j::
_,
:IE
:::> w
J:
0
()
2i
f
fl)
fl)
.
E
~
0
0
....
..
0
P•AIOHIO
54
,.
.."'
0
0
MANUSCRIPT II
MONITORING CRYSTAL MODIFICATIONS IN SYSTEMS CONTAINING
IBUPROFEN
55
Key words: Ibuprofen: Crystal analysis: Hydrophobic network:
Intermolecular interactions: Formulation effects.
SUMMARY
Qualitative and quantitative crystal analysis . including
differential
scanning calorimetry, X-ray powder diffracti on
and scanning electron microscopy were performed at different
stages
of
ibuprofen
tablet
manufacture obtained at three ._
levels of compaction. Melting points and enthalpy of
were
carefully
monitored
techniques ( ANOVA and one
Drug - disintegrant
and
compared
degree
interactions
of
were
fusi on
using statistical
freedom
procedures ).
investigated
using a
fractional factorial design. wetting and compaction affected
the
crystal
surface
as
measured
by a 0.2 to 8 .6 KJ mole
decrease in the heat of fusion.and a shift of 2-3
°c
in
the
melting point. The differences were too small to suggest the
existence of
enantiotropically
polymorphs.
The
results.
modification of ibuprofen
dissolution
rates
or
however.
during
appeared
monotropically
related
indicated
a
lattice
The
initial
processing .
to be inversely related to the
amounts of ibuprofen in the formulation and the fastest drug
release was obtained for a 1 / 3 intragranular ratio .
56
INTRODUCTION
Fundamenoal
investigations.
especially in the field of
compaction and weo granulation. have long
established
that
pharmaceutical processing can modify some characteristics of
raw materials in such a way that can be detrimental
overall
performance
al 1986, Chan et al, 1985 ). Monitoring crystal
become
essential
in
order
to
changes
was
altered
as
forces or exposure to
physical
liquids
interactions
The
patibility
crystal
( Cruaud
et
al.
1981 )
and
between ibuprofen and excipients can
Mura
et
al.
latter does not necessarily mean adverse incom but
pharmaceutical
may
explain
handling
manipulations
difficulties.
Many
will affect the crystal habit
of drug substances and these modifications may have
consequences
has
a function of increased compression -
induce eutectic behavior (Gordon et al. 1984;
1987 ) .
the
optimize many formulations
(Haleblian et al, 1975). For example, sulfanilamide
habit
to
of the final drug product ( Lefevbre et
adverse
on the formulation ( Cruaud et al , 1981 ) or the
drug bioavailability (Aguiar et al, 1967 ).
Crystal
properties
of ibuprofen are known to influence
the processing behavior ( Romero et al. 1991; Hiestand et al .
1981 ) .
This
mented
and
aspect
it
is
of ibuprofen formulation is well docu generally
recognized
that
the
drug
undergoes changes due to processing ( Franz et al. 1986). For
57
example. eutectic behavior has been proposed with some phar maceutical
excipients
( Gordon
et
al.
1984 ) and although
never experimentally proven . surface sintering has been sug gested
as
a
theory
for
rearrangement of crystal lattice
during compression ( Alhe c et al. 1990 ).
Nevertheless.
very little has been published to support
evidence of the crystal modifications
of
ibupr ofen.
Thus .
the mechanisms and consequences of such alterations have yet
to be identified for this particular compound.
The
objectives
mechanisms of
modified
fects
of
cryst al
this
work
were
distortion
by
to
which
elucidate the
ibuprofen
is -
during its processing and to investigate these ef -
on ·the
biopharmaceutical
properties
of
a
model
formulation.
EXPERIMENTAL
Materials
Ibup r ofen USP grade was
obtained
from
the
Ethyl
Co .
(Lo ttLH-6-72 ) . Wet granulations containing Fast Flow Lactose
( Sheffield lot *59009). Povidone
and
Explotab
purified
Magnesium
(P.V. P.-GAF
lotiG-30223A)
( Edward Mendell lot #l 336 ) were prepared using
water.
Granule
stearate
lubrication
( Fisher
was
achieved
using
Scientific Co. ). The potassium
58
monobasi c phosphate and sodium hydroxide used
tion
medium
Scientific.
and
All
Ibupr o fen
buffe rs
chemicals
standards
were
were
for
obtained
of
dissolu-
from
Fisher
grade.
analytical
for spectrophotometry and differential
scanning calorimetry ( DSC ) were
provided
by
the
Standard
Divisi on of USP Rockville. MD .
Methods
The experimental design consisted of analyzing ibuprofen
after dry mixing with excipients , wet massing and tableting. The model formulation . defined in table I was prepared using
five process steps presented in table II . The percentage
active were
57, 67 , and 77 percent .
the diluent , the binder ( 6
amount
Aid.
model
a
into
appropriate
liquid
hours
350
mixer
The
( Kitchen
K5 - A. Hobart ) until the end - point , monitored by
remained
Flow rate o f
constant
mg
the
throughout
a
tray
at
granula -
the
40
entire
°c
for
and later mixed with lubricant in a V-blender
for ten minutes. Lubricated granules
press.
the
planetary
experiment. Granules were dried on
twelve
and
of disintegrant were dry mixed for ten minutes.
power co nsumption , was reached.
tion
Mixtures of ibupr ofen.
percent ) .
powder was wet granulated in
of
were
then
compressed
tablets using an instrumented F3 single punch
Three compaction pressures were investigated :
59
low.
int er med i ate
high
and
( aver ag in g
1.1 0
and
3C
KN
respe c oivel y ) .
At
the
end
of
the manufacturing steps I , III , I V and v
( table II ) . s amples were withdrawn
powder
diffraction
and
and
anal y zed
by
X-Ray
scanning electron microsc opy ( SEM )
photographs. Thermal analysis ( DSC )
was
performed
on
all
samples ( Kim et al , 1985 ) using a Perkin Elmer. series 7 instrumented unit, calibrated with i ndium and interfaced
with
a P 7500 E computer. For whole tablets , the electron microscopy photographs were shot at 35 and 80
and
° angles
on
pressed
side surfaces, on ho r izontal and vertical cross section _
of tabl e ts embe dded
a nd
prepared
according
to
a
method
describe d by He ss , ( 1978 ) .
Cry stal Packing
The
unit
cel l
of ibuprofen crystal was analyzed using
the mo l ecular mode l ing software. The coordinates
X- ray
reflection
data
was
obtained
from
of
single
the literature
(McCon ne ll , 1974 ) and t he molecular arrangement of a crys t al
lattice ,
simulated on thi s program.
Comparative Analysis
In
an
effort
to
ibuprofen and physical
mimic
mixtures
60
the
of
effect
the
of
processing,
formulation
were
gr ound tho r oughly for t en minute s in a mo rtar o r melted at a
°c
temperat u re above 80
and recry stallized u p o n
coo ling
at
r oom t e mperature ( RT ) . Di fferential scannin g cal o rimetry wa s
further perf o rmed on the samples and their thermal
compared
those
to
of
pure
and
Additionally. ibuprofen hygroscopicity
storage at 35
°c
formulated
was
pr o files
ibupr ofen .
mea s ured
after
and 85% relative humidity ( RH ) . Karl Fisher
analysis was performed at regular time intervals on
100
mg
samples exposed to humidity.
Biopharmaceutical Properties
The
dissolution
apparatus used a paddle rotating at 50
RPM in a USP phosphate buffer at pH 7 . 4 and a temperature of
37
°c.
This method using a six vessel dissolution apparatus
(Vankel ) had been
shown
to
discriminate
between
various
ibuprofen formulations ( Romero et al , 1988 ) . An ultra-violet
spectrophotometer wa s used to determine the concentration of
ibuprofen
at
264
nm
in
the
dissolution
fluid. For low
ibuprofen con centrations. the percentage dissolved was
calculated f r om measurements obtained at 220 nm.
Statistic al Analysi s
61
also
All
results
were
analyzed
statistically
analys is of variance at the 99% confidence level
mine
differences
between
enthalpy
of
to
of
freedom
comparisons
using
an
deter -
fusion and melting
ranges . Sums of squares were calculated to perform
gree
using
one
de -
orthogonal contrasts.
These tests allowed the investigation of pure compaction ef fect
on
the
factorial
thermal
design
was
parameters .
used
to
drug disintegrant
interactions.
were
active
amounts
tragranular
of
disintegrant.
and
A restricted fracti onal
test
The
the
the
effects
of
independent variables
concentration
of
ex -
All factors had thr e e levels . In _
the interpretation of the data greatest weight was placed on
a ny effects on dissolution.
RESULTS AND DISCUSSION
Crystal Packing
The single crystal unit cell for the
four
molecul es:
hydrogen
bonds
dextrorotary
racemate
included
two R(-) and two S(+) isomers. two central
between
the
carboxy l ic
f unctions
of
and levorotary molecul es ( Fig . 1). In addition
figure 2 shows
the
juxtaposition
of
eight
crystal
unit
cells. The hydrogen bonds between cells could be identified.
Each inte r molecular
other
cells
as
interaction
was
shared
betwe en
fou r
favored by the preferential positioning of
62
S ( ~)
R(-) and
molecules .
Except
for
the
top - left
cell
retained f o r baseline comparison. each unit has been cleared
of the molecules not involved in the intercellular
tions.
interac -
The resulting effect is the delimitation of a plane.
on which intermolecular distances are most likely to be
af -
fected during tangential stress. Thus this eight cell system
may explain the observed lattice weakness.
The
mass
fraction of water obtained by the Karl - Fisher
technic , averaged ( 0.063% + 1- 0.001 ) and (0.55%
before
and
after
exposure
+
0.004 )
-
to humidity respectively. This
analysis confirmed that although the moisture increased
fold
after
ten _
exposure to 85% relative humidity for 76 hours .
it did not.exceed 0.55% possibly concentrating at
the
sur -
face since ibuprofen does not include crystallization water.
This amount of moisture. was defined by
Alhec
and
Zografi
(Alhec et al, 1990 ) as plasticization or molecular mobility.
Thermal Analys is
Thermal
analysis
of ibuprofen indicated that only com-
paction or grinding of the
melting
point
parameters
Furthermore.
and
were
the
compared
physical
heat
at
mixture
of
fusion
the
99%
affected
the
(table III ) . All
confidence
level.
the enthalpy of fusion decreased progressively
to as low as 18.l KJ / mole during the tablet manufacture.
If
the assumption that a pure equilibrium exists at the melting
63
point ( Tm ) is valid. then the changes are
of
tive
enthalpic
modifications
interactions ) at the
crystal
( weaker
surface
~ndica -
probably
in termolecular
before
compression.
then :
t.G
-
at
t.H
Tm
- Tm . t.S
the
equal zero. with
t.t.Hf
equation 1
melting
O and t.H
Mixing
with
t.Tm
the
Tm . t.S
equation 2
excipients and process ing are the combina-
statistical
enthalpy
analysis
parameters . All enthalpy of fusion were
ferent
as
determined
by
the
F
drop.
of
orthogonal
thermodynamic
significantly
dif -
processing.
the
enthalpy
of
and ibuprofen in lower strength formulation appeared
to be less sensitive than in higher
heat
IV
contrasts Ll , L2.L3 were found statistically
significant. Compression had an effect on
fusion
Table
test ( see table IV ) . The
magnitude of the shift depended on the stage of
The
sample,
the melting decrea se.
tion factors responsible f or the
summarizes
sh ould
t.G
t.Hf the enthalpy of fusion of the
the enthalpy loss and
t.G
t he free energy
po int.
of
strength
tablets .
The
fusion for ibupr of en was less affected in granules
than in tablets.
64
X- Ray Crystallography
In figure 3
ibupr o fen
tal
the
X-Ray
diffraction
changes
of
ibuprofen
dilution
a
slight
during
pure
crys-
pharmaceutical
with excipients only decreased the
intensity of the diffractogram.
Wet
granulation ,
h owever.
rearrangement of the cr··stal lattice . At
low angles of the spectrum , the 12.2° defle ct i on
pletely
of
dry granules and ground tablets indicated
manipulations:
induced
patterns
peak
com -
d isappeared leaving an amorphous region. No further
changes were visible on the X-ray diffraction pattern
after -
compaction.
Scanning Electron Microscopy
In
order to complement results of the thermal analysis ,
qualitative
observations
were
performed
by
SEM.
Morph ological changes from pure to formulated ibuprofen were
visible at the XlOOO magnification as shown on figure 4.
the
tablets.
the
crystals
are
visible
boundaries are not detectable (indicating
fusion
to result from sintering of the
( dark).
but the particle
cold
bonding
or
at the surface). Thus , an ibuprofen network appeared
processing.
( light
On
Cross
membrane )
Some
ibuprofen
crystals
during
sections of tablets showed a film of PVP
covering
starch
packs
of
ibuprofen
crystals
glycolate particles are also visible.
65
The macroscopic observation of SEM phocographs confirmed the
crystal
disorder
suggested
by
granulation process indu c ed the
up on
chermal
lattice
analysis : the wet
fragilization
and
co mpactio n . further crystal disrupti on occur red with a
p oss ible consolidation of the hydrophobic matrix .
integration
and
dissolution
analysis
The
confirmed
dis -
that the
ibuprofen network was indeed hydrophobic. When two ibuprofen
particles
are
in
contact.
pr operties have already
enough
within
been
a formulation . thermal
disturbed;
upon
compaction.
mechanical ene r gy is provided to induce cold welding
or sintering of the crystalline envelopes
of
ibupr o fen
as _
visualized on these S.E.M photographs.
Effect of Proces s ing
The
2-3
°c
decrease in Tm averaged a statistically significant
upon compaction. ATm was not proportional to the com -
pression
as
indicated
by
paired
t
tests
(f ig.5 ) .
reported values of compaction are arithmetic means of
The
upper
punch compression forces recorded on the instrumented press.
The e n thalpy decrease AAHf was defined as t he difference between the heat of fusion of pure ibuprofen and the energy of
fusion of the formulated ibuprofen . AAHf was
the
extent
of
some
have
of
crystal modifications. The effects of
compaction and low ibuprofen strength on
parameter
indicative
the
thermodynamic
been found statistically significant ( table
66
I V). I n f i gure 6 f or all Sorengths studied t he enthalpy l oss
in c re as e d
a nd stabilized at a plateau value . Low co n c ent ra -
oi o n s of ibupr o fen seemed t o be less sensitive t o me c hanical
stre s s .
The
latter.
however. had the highest shif t before
tableting. These results are conclusive of tw o opposite
fects
of
a
low
active excipient
rati o
during
ef -
tablet
manufacturing. Similar effects had been previously sug ge s ted
by a study on the pharmaceutical processing of sulfanilamide
( Cruaud et al. 1981 ) .
Dissolution Studies
The dissolution profiles of
hibited
similar
trends ,
as
ibuprofen
seen
in
ibuprofen content led to cores with fast
as
compared
to
cores
also
ex -
figures
7- 8 .
Low
dissolution
rates
higher strength dosage forms. Although the
data suggested that initial drug release rates might be
lated
re -
to the extent of the ibuprofen network ( table V) . any
correlation was insignificant because of the limited
number
of points. The crystal modi fication of ibuprofen created the
surface hydrophobic network within the tablet
quantified
by
AAHf.
The
that
for
the
lowest
be
latter seemed to be the limiting
factor of drug release . The dissolution efficiency
timal
may
was
op -
amount of ibuprofen and the highest
concentration of disintegrant (3% ) in the 1 / 3
67
intragranular
rati o.
oegrant
Fo rmulati on
had
formulati on
with 100% ex,ra or intragranular disin -
l ower
dissolution
c ombining
efficiencies
extra ' intragranular
than
did
disintegrant at
the same c o nc entrati on.This might be an other evidence of the
existence o f the hydrophobic network.
CONCLUSIONS
The crystal packing of ibuprofen occurs with a preferen tial orientation in which a weak plane has
been
identified
as probably responsible for the crystal modifications during _
compaction.
Pharmaceutical
processing
does alter the crystal habit
of ibuprofen ( not its internal structure ) in a stepwise man ner. We identified three progressive mechanisms:
Mixing with excipients: destabilization and fragilization of
intermolecular
interactions,
wet granulation: distribution
of water in the amo r phous regions. Both
for
a
drop
of
4.1
mechanisms
account
KJ / mo l e in the enthalpy of fusion and
predispose the ibuprofen to co l d welding. Compaction: brings
the
enthal py
dec r ease
to
another
6.2 to 8.6 KJ / mole and
provides enough energy to catalyze sintering as observed
the
scanning
electron
rearrangement ,
hydrophobic
and
the
could
microscope.
resulting
on
In addition to lattice
ibuprofen
network
is
be the limiting factor of drug dis-
solution. As a consequence
formulators
68
must
be
extremely
cautious
vhen
increasing ibuprofen concent ration in t ablet
formulation.
REFERENCES
A . J . Aguiar , J . Krc, A . W. Kint e l and J .C . Samy n
"Effect of Polymorphism on the Absorption of
Chl o ramphenicol from Chl o ramph en icol Palmitate "
J.Pharm.Sci . , 56 (7), pp 847 - 853. 1967
C . Alhec and G. Zografi
"The Molecular Basis of Moist ure Effects on Drugs
in the Solid State."
Accepted Int . .r.Pharm . , 1991
J . F . McConnell
" 2 -( 4-Isobutyl Pheny l ) Propionic Acid.
Cr~:;it.S:truQ:!; .
CQmm . 3 , pp 73 - 75. 1974
H. K . Chan and E . K. Doelker
"Polymo r phic Transf o rmatio n o f Some Drugs under
Comp r ession"
I.llll.E . 1 1 ( 2&3 )' pp 315 -332. 1 985
O . Cruaud , D. Duchene, F. Puisieux, A . Chauvet, J . Masse
"Etude des Transformations Poly morphiques du
Sulfanilamide au cours de la Fabrication des
Comprimes"
.r . Pham . Belg. , 36 ( 1 ). pp 15 -20, 1981
Robert . J . Franz ,
69
" Stable High Dose. Hi gh Bulk Density Ibupr ofen
Granulati o ns for Tablet and Capsule Manufacturing"
U.S . Patent 4 , 609.675 Sept. 2.
1986
P.E. Gordon , C . L . Van.Koevering , and D.J. Reits
"U tilizati o n of DSC in the Compatibilty Screening
of Ibupr o fen with Stearate Lubricants and
Constructi on of Phase Diagrams "
Int.J . Pharm. 21 : 99 -105. 1984
John K . Haleblian
" Characterization of Habits and Crystalline
Modifications of Solids and their Pharmaceutical
Applicati o ns "
J.Pharm.Sci
64 ( 8 ), pp 1269-1288. 1975
H . Hess
"Tablets Under the Microscope. "
Pharm.Tech., pp 36 - 50 , Sept.
1978
E . N. Hiestand , G.E . Amidon , D. P. Smith , and B.D . Tiffany
"Mechanical Property Changes of Compacts from
Various Rates of Crystallization of the Solid.
Proc.Techn.Prog.Int.Powd.Bulk Solids Handling
Proc. , Rosemont. IL 1981
H. Kwon Kim , M. J. Franck , and N. L . Henderson
"Application of DSC to the Study of Solid Drug
Dispersion " J.Pharm.Sci .. 74 ( 3 ), pp 283-289 . March
1985
C . Lefevbre, A . M. Guyot - Hermann , M. Draguet , R . Bonde ,
70
and J.C. Guyot
" Polymorphic Transition of Carbamazepine during
Grinding and Compression·
rilll.E.... 12(11&13) , pp 1913-1927. 1986
P. Mura , A. Liguori, G. Bramanti , and L. Poggi
"Phase Equilibria , Crystallinity and Dissolution
Rates of Ibuprofen-PEG 20.000 Solid Dispersions ·
Il Farrnaco-Ed . Pr., Vol. 42. Fasc. 6. pp 157-164 ,
1987
A. J. Romero , L.T . Grady and C. T . Rhodes
"Diss olut ion Testing of Ibuprofen Tablets "
D. D.I.P. , 14 ( 11 ). pp 1429-1457 , 1988
A. J . Romero , G. Lukas, C. T . Rhodes
" Influence of Drug Sources on the Processing and
the Biopharmaceutical Properties of High Dose
Ibuprofen Formulati ons.·
Pharm.Acta Rely ..
66(2 ), 34-43 , 1991
Aknowledgements
This work was supported by Ciba-Geigy Corporation
We aknowledge, with gratitude the help of Dr . Franck Clarke ,
Basic
Research
at Ciba- Geigy who made invaluable contribu-
tions to the success of this project
71
Teble I: Formuletlona UHd In thla Study
Ingredient
Ibuprofen
Feat Flow LectoH
P.Y.P.
Ilg Steerete
% Ne Sterch Glycolete • 1
Amount In %
57
36
67
11
26
16
6
6
1
6
1
2
3
1
2
3
1
2
-'1
"'
•D11tails on th11
111• lntragr11nul11r ratio 1 1/2 1/3
Purified Weter
1 1/2 1/3
q.1.
1 1/2 1/3
q.1.
All ay1tema contelned 1 % of lntregrenuler dl1lntegrant
.'
3
q.1.
Table I: Experimental llethodology
ProceHlng
-<I
Vl
Dry ••Int (V-Blender) 1 O mlnutea
(Ibuprofen, LactoH, Explotab, PVP)
Wet Granulatlon
(llonltored by power conaumptlon)
12 how tray drying at 40
C
•xllllg (V-Blender) 10 mlnutH
(QranulH, Explotab, Lubricant)
Compaction
Stepa
Analyala
DSC, SEii, X
II
DSC, SEii,
DSC, SEii,
DSC, SEii,
DSC, SEii,
DIHolutlon
Hardneaa
HI
IV
v
DSC: clfferentlal acannlng calorlmetry, SEii: acannlng electron
.icroacopy, X: X-ray cryatallography,
'
.
X
X
X
X
T8ble •: Thermal Analyal• of Ibuprofen
II.Pt.( SD)
AH (SD)
[oC)
[KJ/mole)
AS [J.G - . 'K
IBUPROFEN
~
ti>
77.7(0.6)
76.8(0.2)
76.9(0.,)
77.2(0.3)
Pwe
Ground
Melted(•)
RH(••)
Phyalcal
Mixture
Ground
77.,
75.3
Granule•
Tablet•
77.3(0.2)
75.2(0.9)
-
-
25.7(0.53)
23.7(0.93)
23.0(0.91)
25.6(0.06)
0.35
0.33
0.11
0.35
21.0 18.1
0.29
0.25
21.9(1.29):
18.2(1.71)
0.30
0.25
-
(•) recryatalllzed at RT, (••) 76 hour• at 85 "
(., atellotlNllr alllllf!CMI ...,.,....,. _ . , . . . lo - · . _....
'
'
RH
]
Table IV: ANALYSIS OF VARIANCE (ENTHALPY OF FUSION)
One degrH of Freedom comparisons
SOURCE
DF
SS
llS
F
F
o .0 1. 1,4t
-.;i
(JI
Among
Formulations
9
•365.7
•B5.1
L 1 KN Vs. NoKN
1
-
-
-
-
1
-
-
33
1157.3
L.Z Low KN VS
High/Reg KN
£3 A-57% Vs.
C-67% or
E-77%
Within
Formulations
1
13.62
948 •
•
429
19.8
7.3
•
35.6
(•) indicates a significant difference (99% confidence level)
'
'
'al
c
•
•c0
:::
.~Cit
: E
..
-• -••
•
••
;:: •
a:
• •••
~
Q
0
a:
• a:
0
•C! •.,
0
...0
0
CID
-..• -• .. ....... ...... •.,
;f
1- - -.... i....
ii •
Cit
4'!
0
,.. a:
'#.
.. I
!. c u•
ID
...
""
GI!
i....
....
iii
76
..••
Iu
..•c
•
i
a
0
;
=
:!
.....
•
I.
it
c
77
!i
a
c
:;
Q
:. :
•~=-•
:i
•
0• ';
Q
! >/
I
/ '
I )'(\ '
.;.
..-e •""
o!
• ti
c
Q.
iJ
-
••
as
78
N
N
N
...OU
•
e:
.! •
.
;
a. •
c '!
.! :
u&
~ c
0
N
"'
"'~====~S:~~=
~~
•
0
,.. i
~.
I
)( iii
;;
•
I
ii:
79
,,-..
Figure 4: Identification of the Ibuprofen network
Electron microscopy photographs of
A) Ibuprofen, Bl grHulea, C) tablets:
C1,C2: preaHd surface, C3: aide walls
C4: vertlcal croaa section
O>
0
81
w
w
i=
i=
>
u<l:
~
r---
l!l
>
u<l:
~
r--r---
D~
.
..
0
0
0
...
.!
("")
••
•
1111
,\~~"~
0 ..
u
-z
~
A.
2i
; :!•
IL
!
.. 'i
¥•
=.
0 • ii
1111
~I
1115
,;;
•
u
I
0
IL
("")
0
(\J
(~)
'ld'lll .. J..-tS
82
Figure 6: Effect of Compression Forces
on the Enthalpy Parameter
-i!t- 77% DRUG
. u:
--'Y-- 67% DRUG
- -..-·- 57% DRUG
50
.,...
"'"'
•-z
l
_,,,-'"'l-----------------------------------------'V
,,
rv-·<.,'. .---·-1· ·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·--- ...T.
25
~
•
I
o
I
I
1
I
--~~~~~~~~~--~~~~~~~~__,
0
15
COMPRESSION LEVEL (KN)
30
f'ltur• 7 : DIHolutlM
of IMlpfofetl CorH
Dlalntegronl (0 MCI 1 lntHgronulor r•tlol
-0-
57% QR.JG
110 ...------------------~
55
Dlalol ...Mt l•••I t ...
1001'11t!nw1N11ar
o"---------'-----------l100
0
50
110 .------------------~
55
o.-.--------'-----------'
0
50
T IPllE (M1N.JTE)
84
100
Figure 8 : DluolutlH of Ibuprofen C.rH
DlalntegrHt ( 112 Hd 1 II lntr1gr1nul1r retlol
--.- 77% CA...G
-:>--
57%
~L.G
-+- ·
57% OR\...G
a lt:.~~~~~~~~~~~~~~~~~
0
~
~
77% Of:l.UG
~
g
55
i5
Dlollt-.rant LeHI I'll
1/S mt:r•1r.w•
o ~~~~~~~~~......~~~~~~~--'
0
50
T IME (MINUTE )
85
100
(
MANUSCRIPT III
AN EVALUATION OF IBUPROFEN BIOINVERSION BY SIMULATION
86
INTRODUCTION
In an e ffort to formulate the
enantiomer
examined
of
ibuprofen
the
pharmacologically
active
into a solid or a l dosage form. we
literature
regarding
investigations
of
stereoselective pharmacokinetics of this arylpr opionic acid .
Early rep orts (1) indicated that urinary
lites
were
essentiall y
following
all
administration
Sensitivity
of
dextro-rotary
of
the
stereoselective
separati ons have improved and
that
ibupr ofen
it
is
(S enantiomer )
racemate
assays
metabo-
and
in
man.
enanti omeric
generall y
recognized .
there is an enzyme catalyzed inversion of the inactive
R enantiomer into the therapeutically
(2.3) .
Bioinversion
active
S
enantiomer
of the R to the S enanti omer ha s been
suggested to occur either presystemically ( 4 , 5) and or
temically
(6 ) .
Als o.
differences
in
the
inverted int o Shave been reported (7. 8 ).
sys-
fraction of R
To date only four
studies have involved the administration of a pure ibuprofen
enanti omer t o man (6, 9 ).
Using
the
pharmacokineti c
model
proposed by Jamali et al . ( 5 ), we simulated (-)-R- and
(+)- S - ibuprofen plasma
ministration
racemate .
of
concentrations
(+)- S- ibuprofen,
following
oral
(-)- R-ibuprofen ,
Simulated and literature values
for
ad -
or the
area
under
the plasma concentration - time curves ( AUC ) were used to compare rati os of SI R for different cases of the model and
87
for
comparison
of
different
methods calculating fraccion of R
in verced to S.
METHODS
Simulations
The
one-compartment model proposed by Jamali et al . ( 5 )
was used for all simulati o ns of ibuprofen enantiomer
concentrations.
This
model
(Fig.
plasma
1) assumes first-order
processes for absorption and elimination for both
R
and
S
enantiomers , as well as for the bioinversion of the R to the .
S enantiomer.
that
total
This model appears to have been designed such
elimination
of
R.
by
bioinversion
bioinversion routes , should be equal to
elimination
and non of
S.
The literature , however, supports a faster apparent elimina tion for the R isomer ( 6-8 ). The racemate dose
was
to
mg of (+)- S-
be
200
mg
of
ibupr ofen enantiomer.
to
those
used
by
(-)-R-ibuprofen
Pharmacokinetic
Jamali
et
al.
and
200
parameters
rate
constants
( kar.kas )
similiar
( 5) were incorporated.
Volumes of distribution (Vd) were 10 L for each
Absorption
assumed
enantiomer.
and elimination rate
constants (kr , ks) for both enantiomers were 1.0 and
hr - 1 ,
0.34
respectively .
The
four
cases of presystemic
(kip) and / or systemic ( kic ) bioinversion reported by
et
Jamali
al. ( 5 ) were reproduced as follows: 1 ) kip=l.5 and kic=O
2) kip =l.O and kic =0.08 hr- 1 , 3)
and
kip =0.5
hr - 1 ,
88
kip-0 and kic -0.225 hr - 1 .
kic-0.15hr- 1 .
and
4)
simulation of
the
model
simulation
software
step of 0.01 hr.
constants
was
performed
with
Numerical
the
Stella
( 10 ), using Euler ' s method with a time
The
apparent
terminal
elimination
rate
for the R and S ibuprofen enantiomers ( App kr and
App ks) were estimated from simulat ed plasma concentrations.
The
area
under
the
simulated
plasma
concentration-time
profile from time zero to 24 hr after start of dose.
24) ,
was
calculated for both enantiomers.
AUC's were used to calculate an S I R AUC
from
these
AUC (0-
These simulated
ratio.
The
AUC 's
simulations were also used to compare different .
methods of calculation the fraction of the R enantiomer
verted
to
the
in-
S enantiomer following oral administration.
In additon. literature values for AUC of Rand S
(4.6-9,11)
were used to calculate an S I R AUC ratio and assess different
methods for estimating fractions inverted.
Calculation of Fraction Inverted
For calculations of fraction of
ibup orfen
inverted
in
the body. it was assumed that both enantiomer s have similiar
absorption and disposition parameters.
Also , it was assumed
that oral administration of racemate was similar to oral administration
enantiomers.
of
equal
These
amounts
of
S
and
R
ibupr of en
assuptions are recognized as potential
limitiations of the calcualtion method but have been used by
several authors (5- 9, 12 , 13 ) .
89
The relati o nship assumed was :
( AUC
of
s after racemate ) - ( AUC of s after S) - ( AUC of S
after R)
Equation 1
where dose of racemate - dose of s - dose of R. and the dose
of R and S are equal.
The
fraction of R inverted to S ( Fr
s ) may be approximated
by :
Fr- >S
=
( AUC of
s
after R) / ( AUC of
s
after S )
Equation 2
for equal doses of R and S.
Substitution
Equation
tions for Fr
s may be derived:
Fr - >s
into Equation 2 , additional equa-
- (( AUC of S after racemate )-( AUC of S after S )) i ( AUC
of S after S )
Equati on 3
where dose of recemate
Fr
2•dose of S:and ,
s = ( AUC of S after R) / (( AUC of S after
Equation 4
of S after R))
where dose of racemate
Based
on
the
Racemate )-( AUC
2*dose of R.
previous assumptions of identical absorption
and disposition for both enantiomers , it may be possible
90
to
approximate fraction inverted in the body following oral ad ministration of the R and S enantiomers. the racemate and
S
enantiomer, or the racemate and R enantiomer.
In addition one can estimate
chiral
R
contribution
of
Rather than calculating
the
fraction
converted to S ( Equation 2-4). an alternate method of
looking at bioinversion is by estimation of the fraction
S
the
bioinversion to the plasma levels of the therapeuti -
cally active isomer.
of
the
which
of
is inverted from R ( Fs- >r) for a dose of racemate.
This may be approximated by the following equation:
FS- >r
( AUG of S after R) / (AUC of S after racemate )
Equati on 5
where dose of racemate
2* dose of R(-).
Additional variations of Equation 5 are also possible if one
assumes the relationship in Equation 1 to be valid.
RESULTS AND DISCUSSION
Simulated
( -)-R-
and (+)-S-ibuprofen enantiomer plasma
concentration-time profiles and t he corresponding S I R plasma
concentration
ratios
appeared
to
be
reported Jamali et al. (5) for the four
91
identical
different
to
cases
that
of
presystemic
and or
systemic
bioinversion
following
ad -
ministration of 400 mg of racemate as 200 mg each of R and S
enantiomer s
Simulations
of the S R AUC rati os. following
administrati on of the racemate. ranged from a value
for
a
presystemic-only
1 .66 for
systemic -only
bioinversion
of
4.0
( Sim #1. Tabl e I) to
bioinversion
( Sim
*4.
Table
I).
Simulati ons of the S R AUC ratios. for administrati on of 200
mg of the R enantiomer o nl y, ranged from a value of 1. 5
a
presystemic -o nly
for
systemic-only
bi oinversion
f or
( Sim #5 . Table I ) to 0.66
bioinversi o n
~s.
( Sim
Table
I ).
Calculati ons S I R AUC rati os from literature reports of AUC ' s .
of R and
S
enanti omers
racemate
averaged
following
1.53±0.20
( average±sd ) .
ratios following administration of
found
to
administration
be 0.50±0.09 ( Table II ) .
the
R
while
appear
data.
on
to
be
in
of
were
Thus. these simulations
man
s ince
good agreement with the literature
Different conclusions by Jamali et al .
use
the
S / R AUC
enantiomer
support a conclusion of systemic bioinversion in
they
of
(4,5) .
based
a plasma concentration S I R ratio . may be due to
increased variability in plasma concentrations R and S enan tiomers .
Since the emphasis of that ratio was placed toward
the end of the sampling period
when
plasma
concentrations
were relatively low ( 4,5 ), a greater error in calculation of
the plasma concentration SI R rati o could
assay variability.
have
arisen
from
It is speculated that calculations S and
92
R AUG'S. for th e S R AUG ratio. are not
as
susceptible
to
the same degree of such err o r .
Agreement of simul ated S R AUG
values
does
not
necessarily
ratios
support
with
the
literature
hypothesis
sytemic bioinversion , since the mode prop os ed by
al.
(5)
may
be
questioned.
Jamali
of
et
The mode ( Fig . 1 ) appears to
have been designed to demonstrate similiar eliminati on for R
and S.
The sum total of R is ( kr - ki c)T kic. which equals ks.
This creates some confusion since elimination of R
bioinversion
routes
comparable.
than
and
elimination
However , elimination of R
by
of
S
by
could
processes
defined
by
was
Jamali
et
al.
derived
to
demonstrate
elimination
other
bioinversion
(5).
It was assumed that this
similiar
case
holds
true
for
the
' apparent '
presystemic-only
(Sim # l and Sim *5. Table. I ) .
as the model is admusted toward systemic-only
apparent
model
rate constants for both R and S ( App kr and App
ks , Table I ), which
the
be
bioinversion was adjusted by the systemic bioinversion
rate constant ( i.e . kr-kic, Fig. 1 ), according to the
model
non -
eliminat i on
rate
However.
bioinversion ,
constant for S , determined
from simulated plasma conce n trations
(App
ks ) ,
decreases.
ne suggested modification of the model is not to adjust non b i oinversion elimination of R by the changes in kic .
model
modifications
are
a l so
possible , such as suggested
from the studies by Ahn et al. in the dog ( 12 ) .
93
Other
from
which
it
may
be speculated that the rate and extent of bioinver -
sion decreases with increasing amounts
Thus.
a
model
fact
that
(+)-S-ibuprofen.
incorporating saturable enzymatic inversion
may prove to be a more correct model.
the
of
gastro-intestinal
We
also
membranes
appreciate
of
different
animal species contain the enzymatic system re sponsible
the stereospecific inve rsion.
for
Therefore. presystemic bioin-
version. if any. could be formulation dependent.
An assessment of the fraction of R inverted to S (Fr- >s)
was based on c alculations using Equati ons
ti o ns
and
2-4
on
racemate.
the same model ( Sim #9, Table I ).
2. and the results for simulation
tiomers,
simula-
literature values for AUC ' s of the S enanti omer . •
following administration of S and R. S and
based
for
the
fraction
of
the
inverted
0.66
and
from
for
S
enan -
0.6
for
pre systemic-only
simulation
simulation.
must be n ot ed that Jamali et al. ( 5 ) chose
It
to
R
Usin g Equati on
R
ranged
or
parameters for pres ys temi c and systemic
would
syst emi c-only
bioinversion
which
approximate a fraction inverted of 0 . 6 , which i s con-
sistent with the data of Lee
Geisslinger
et
et
al.
(8).
Assessment
of
al . ' s data ( 7 ), using Equation 2. indicates
the fraction of R inverted to S was
0.48
( Table
II ),
al -
though these autors who report ed a fraction inverted of 0.33
did
not
indicate
Calculations
of
their
exact
mathematical
operati on .
Fr - >S using Equation 2 ranged from 0.36 to
94
0.64 f or all literature
ministration
of
the
values
R
and
of
S
AUC
after
S
enantiomers
for
( Table
Fractions inverted calculated for Equations 3 using
ture
values
ranged
form
0.25
to
ad II ).
litera -
0.74. while Equation 4
yielded values of Fr- >s ranging from 0.51 to
0.96 .
It
is
important to note that when AUC data was not available for a
particular enantiomer, data from Lee et al.
(6 )
were
used
after
ferences were found
inverted
was
normalization
for
literature
calculated
using
( 8 ) or Cox et al
for dose.
values
Since dif -
when
fraction
Equations 2-4. it might be
concluded that the relationship described in Equation
not
a
rigorous as originally assumed.
In the four studies
involving administration of (- ) -R - ibuprofen. the
and
R
is.
AUC
averaged 1 / 2 of the AUC of Rafter R (49 . 7%).
of
S
Using
Equation 5 , we calculated that after oral administration
of
the racemic mixture. (+)-S-ibuprofen bioinverted from ( - ) -R ibuprofen averaged 38 . 5% of the total ( T) -S-ibuprofen in the
systemic circulation.
CONCLUS I ONS
The
model presented by Jamali et al.
(5 ) , for bioinver-
sion of ibuprofen after administration of a racemic mixture ,
was
reviewed
and
found
to be more indicative of systemic
bioinversion when the AUC S I R ratio was taken
and
compared
to literature values.
95
into
account
Results of simulations
with this model demonstrate n o diffen ece
fo r
Fr - •s.
between
equaoions
based on AUC of S following administration of R
and S. racemate and S. or racemate and R. for ohe four cases
of
presystemic
However.
and or
comparison
syst emi c bi oinversi on inv e stigated.
of
results
of
Equati ons
2-4
using
literature data differ such that it does n ot appear th at the
relationships assumed f o r the
dosing situati ons .
model
h old
true
under
all
Calculations Fr - s ( Equati on s 2-4 ) using
literature data averaged 0.52 overall.
We
estimated
that
after racemate oral administrati on.
1 / 3 of the total (+)- S-ibuprofen
from
the
dose of
inversion
~n
the
plasma
of (-)- R- ibuprofen.
(~)-S-ibuprofen
will be bioequivalent to
dose of rac-ibuprofen after oral administration .
not
the
formulation
therapeutic
of
the
active
is
derived .
Possibly a 15 0 mg
a
200
mg
Whether or
stereoisomer
is
a
improvement can be still argued and ot her ph ar -
maceuticial characteristics of the drug must be considered.
96
REFERENCES
l.
Adams SS . Cliffe EE. Lessel B. Nicholson JS
J
2.
Pharm Sci. 56:1686, 1967 .
Adams SS. Bresloff P, Mason CG. J Pharm Pharmac. 28:256 257 .1976.
3.
Mayer JM, Acta Pharm Nord. 2 (3) : 197-216, 1990.
4.
Jamali
F.
Sing
NN ,
Pasutto FM. Russell AS. Coutt RT .
Pharm Res, 5 ( 1 ) :40 - 43 , 1988 .
5.
Mehvar R , Jamali F , Pharm Res , 5:76 - 79 , 1988 .
6.
Cox SR . February
1988.
Clin
Pharm
Therapeu,
43:146 .
1988.
7.
Gesslinger G. Stock KP. Back GL. Loew D, Brune K. Agents
and Actions , 27 ( 3 / 4 ) :455 - 457, 1989.
8.
Lee EJD, Williams K, Day R, Graham G , Champion D.
Br
J
Clin Pharmac, 19:669-674, 1985 .
9.
Baillie TA , Adams WJ , Kaiser DG , Olanoff LS. Hastead GW.
Harp oot lian H. Van Giesen GJ , J Pharmaco Exp Thera.
249 ( 2):517 - 523, 1989.
10. Richm ond
B,
Ve scuso
P,
Peterson S , A Business user ' s
Guide to STELLA. High Performance
Systems ,
inc.,
Lyme
New Hampshire , 1987 .
11. Cox SR, Brown MA , Squires DJ, Murrill
EA.
Lednicer
D,
Knuth DW , Biopharm Dru g Disp. , 9:539-549, 1988.
12 . Hae - Young Ahn. G. L . Amidon and D.E. Smith , presented
the A. A.P.S meeting in Las Vegas , Pharm Res ,
97
at
7 . ( 9 ) :234.Sept . 1990.
13 .
Kni~inicki
RD.
Day
Ro.
Graham
Chirality 2:134 - 140. ( 1990 ) .
98
GG
and
Williams KM.
Table I
S1mulat10n Parame lers and Rosull s lof 81omve1s1on ol Ibupr ofen,
Based on lhe Model ol Jamah el al. (5)
w11h Cak:ulalions ol AUC SIA Aalio and Fr action ol R lnvef1ed lo S (fr->s )
Dosing ·
Race male
Racemale
Race male
Race male
Ronly
A only
A only
A only
1.5
1.0
0.5
0 .0
0.0
0.08
0.15
0.225
0.341
0 .341
0.341
0.341
0.341
0.317
0.305
0.296
23.52
29.40
39.20
58.80
35.28
36.30
36.84
38.78
1.501
1.23
0.94
0.66
<D
<D
• Dose of A .. 200 mg
Dose ol S • 200 mg
Dose of Racemale • 200 mg of A and S each
Addilional Parameters used tor all simulations
A
S
ka.
1.0
1.0
hr... · 1
k•
Vd •
0.34
10
0.34
10
hr• -1
L
l .iblu II
Al.JC Values in Man alttH Adm1rns1ia1ion of lnd1vldual lbuprnhm l:nan homurs or A aca ma1a
wllh Cak:ulalions ol Al.JC SIA A allo, F1ac11on ol A Inverted 10 S (Fr ·>S), and Frac11on ol S lnve11ed F1om A (fs <- r)
,_,
0
0
-·
Geissl"Cox
Cox
Cox
Cox
Jama'
Jamai
Jamai
Jamai
7
6
6
11
11
4
4
4
4
PO
PO
IV
PO
PO
PO
PO
PO
PO
600
600
600
400
400
600
600
600
600
89.8
91 .0
89.0
58 0
580
106.9
121 .7
90.8
100.2
57.0
61 .0
59.0
42.0
50.0
74.2
73.9
56.9
51.5
1.58
1.49
1.5 1
1.38
1.16
1.44
1.65
1.60
1.95
Average (OveraU) •
SO (Overall\ Average (True"") sp tlrue .. ) •
Avg ol Eqns 2-4.
Ava of Eans 2 ·4 llrue""I •
•compared lo lee dala (8) tor admimslralion ol S, normalized for dose
6 compared Jo Cox data (6) for adm1nistralion ol A. normalized tor dose
•compared 10 lee data (8) tor adrmnis11auon ol A. normalized lor dose
•• al parameters used in 1he equation were obtained from individual stud10s
0 ·I '
0.34
• 0.30
• 0.27
• 0.25
• 0.25
' 0.53
• 0.74
• 0.30
• 0.44
0.56
0 .63
0.5 1
"0.67
• 0.7 1
• 0.58
• 0 96
• 0.80
0.38
0 70
.6 0 .67
0.36
0.38
0 34
" 0.40
"0.40
• 0.42
• 0 .37
• 0.49
• 0.44
0 41
0 10
0 16
0 IL
0.05
0.56
0.36
0.64
0.39
0 11
0 03
0 15
0 05
0.52
0.55
(.)
:::;::
Cii
E
,co
'
Vl
~
::.::
>.
(.)
.0
-0
Q)
Vl
:::;::
Cf
+
~
"(])
0
a.
0
a:
Qi
-0
0
Vl
ca
co
::.::
::.::
:::?
c
0
·u;
a.
'
~c
Cf
al
1
"Ci
~
:::;::
+
Q)
::;
ci
--,
I
I
.c
(.)
....:
"(])
I
l::l %OS pue S %OS =
uaJOJdnq1 a1ewaoel::j
Cl
u:
101
-
-'
MANUSCRIPT IV
APPROACHES TO STEREOSPECIFIC PREFOR.MULATION OF IBUPROFEN
102
Abstract
In an effort to formulate the
ibuprofen
pharmacologically
isomer ((- )- S- ibupro fen ) into a solid oral dosage
form , preformulation studies
racemate
were
performed
on
both
predict
physical
the
pharmacy
pharmaceutical
profiles
were
compared
aqueous
the
very
small
specific .
area. This characterisitic could be a limiting step
in the formulation
was
of
the
optical
isomer
although
less
required for the solution of S (+) ibuprofen . On
this crystal , there was ten times more moisture
the
rac emate
media but exhibited l ower intrinsic dissolution
rates. as would be expected from
surface
to
behavior of the S ( - ) ibuprofen
compound. The enantiomer was more soluble than the
energy
the
and this stereoisomer of ibuprofen . Results of the
respective
in
active
layered
at
surface and comparative thermal analysis indicated that
for both compounds a
grinding .
l oss
Properties
in
in
crystallinity
occured
upon
the solid state of S (+) ibuprofen
included higher density and excellent
flowability
as
com -
pared to the racemate.
INTRODUCTION
Recently
extensive
emphasis
has
been
given
to
stereospecific nature of drug disposition in the context
103
the
of
dr ug developmen t. Re gu l ato r y bodies in Eur ope alre ady r e commend che use of s tere os pecific assays
fo r mu lace d
( 1 ).
a
a
r a c emate
is
In this case. when the chi r al s y nthe s i s i s
p oss ible. the f o rmulati on of pure
becomes
when
active
enanti omer s
als o
priority for research and devel opment. Thu s . f or
drug candidates in the development
considerations
are
critical .
pipeline
For
drug
marketed as racemates. the study of
stere o specific
products
al r eady
optical
i somers
their
might present a different set of problems ( 2 ) .
The use of an enantiomer as part of a solution for
mon
pharmaceutical
problems
is
not
new .
In
the
c omearly .
seventies. formulators at Wyeth had described the advantages
of
using
overcome
the
a
pure
optical isome r of a cytotoxic agent to
solubility
physico - chemical
problem
properties
crystal
The
structures
long
the
effect
of
time ;
s olid
chiral
been
biopharmaceutical
state
investigated.
asymmetry
properties was reviewed and it was concluded that
compounds.
in
and the effect o f is omeric
purity on phase solubilities ( 4-6 ) have
Recently
differences
between racemic and enantiomer
have been known and studied for a
properties ,
(3).
on
crystal
for
such
characteristics must be care-
fully monitored ( 7 ).
Ibuprofen is administered as a racemate and is one of 50
compounds for which detailed stereospecif ic pharmacokinetics
have
been
reported.
Properties
104
in solution and the solid
state have been documented in
tional
preformulation
study
of
this
the
programs
literature
have
antiinflammatory
been
agent
and
conven -
applied to the
( 8).
It
is
now
generally recognized that formulating ibuprofen is difficult
and relies mainly on the expertise of
low
solubility
in
the
formulator .
aqueous media at low pHs as well as its
poor handling properties contribute to its tedious
ing.
S(+)
ibuprofen ,
ibuprofen,
through
is
now
the
biologically
available
economically
in
large
process -
active
scale
isomer of
quantities
viable chemical synthesis and the ob-
jectives of this study are to draw a preformulation
for
this
Its
profile .
enantiomer. investigate the feasibility of a con-
ventional formulation and compare this stereoisomer
to
the
racemate currently used.
EXPERIMENTAL
MATERIALS
Rae-ibuprofen (lot# LH6 - 72 )
AC-lR)
were
obtained
and
(+)-S-ibuprofen
(lot¥
from the Ethyl Co., Baton Rouge , LA.
Monobasic potassium phosphate and sodium hydroxide from
Fisher Scientific Company, were of analytical grade.
METHODS
105
the
Analvtical:
were
all quantitative determinations in solut i on
performed
using
an
ultra
vi olet
spectrophotometer
(Hewlett Packard 8450) at 264 and 220 nanometer wavelengths.
Intrinsic Dissolution Rate ( IDR )
The
procedure
desribed
by Woods
~
( 9 ) was used to
determine the intrinsic dissolution rates. A modified
apparatus ,
designed
used. One gram of sample powder was weighted in the die
precompressed
After
at
cleaning
recompressed
500
the
lbs
with
exposed
lated
surface
the
compact
The
sampling
regimen .
was 4 cm . withdrawn with micro-syringes and
volume
of
of
SIF
with
at 37°c. All samples were passed
analysis.
drug dissolved were plotted versus time and the
slope of the straight line portion was divided by
of
included
replaced
through a 0.45 um cellulose acetate filter before
Amounts
simu-
and 35 minute time points. Sample volume
3
same
USP
fluid ( SIF ) at 37 + / - 0.8 °c and rotated
rpm.
2.5 , 10.15.20 , 25 . 30
the
was
at 1000 lbs for a dwell time of 5 seconds. The
intestinal
100
and
a Carver hydraulic press.
rotating disk assembly was immersed in 500 ml of
at
Woods
by Ciba-Geigy researchers ( fig. 1 ) was
the
area
the pellet ( A-1.281 cm 2 ) to yield the intrinsic dissolu -
tion rates in mg.sec-l cm- 2 .
106
Solubility and Heat of Solution
The method us ed wa s the procedure described
et
by
Higuchi
al ( 10 ). Ibupr of en was added in large excess t o buffered
solut i on s in screw-capped vials at various
were
pHs.
The
tubes
r otated on a labquake shaker in dry ove n s for 24 hours
to reach saturation ( 11 ) . The vials were then centrifuged at
2000g
for 10 minutes, supernatants withdrawn. filtered. and
further analyzed for pH and drug concentrations.
Four
pH ' s
were studied at temperatures ranging from 25 to 51 ° c .
Solid State Properties
Particle
size was measured using a laser light scatter -
ing technique ( Brinckman Particle
areas
were
performed
on
equation.
et
a
instrument.
scanning
proposed
by
were
performed
on
an
Moisture contents were estimated with a
calorimetry
to
conducted
with
107
included
monitor
enthalpies of fusion on a Perkin-Elmer
analysis
experiment
al (12). About 100 mg of powder were weighted in a
Karl-Fisher technique. Crystal analysis
ti al
This
procedure
v o lumetric cylinde r . Up to 2000 taps were
Erweka
Surface
a Quantasorb instr ument. Compressibilit y
and density were evaluated using
Rees
Analy zer ) .
determined with a Nitr ogen ads o rption technique
and calculated using the B . E . T .
was
Size
a
differen-
temperatures
P7500 ,
all
and
thermal
heat fl ow of 5°C / minute;
sample s were als o analyzed thr ough scanning electron micros c opy
and
X- Ray
powder
diffraction
at the R&D analy tical
ser v ices o f Ciba-Geigy in Ardsley , NY.
RESULTS AND DI SCUSSIONS
Solid State Properties
Under similar storage conditions ( 22°c + / - 2 and 35% RH )
the mass fraction of moisture for the rac - ibuprofen averaged
0 . 065
+1-
0 . 013%
where ( + ) -S-ibuprofen exhibited 0 . 34 + ' -
0.24% of water corresponding to ten times more moisture
for
this enantiomer. After careful analysis of X-Ray diffraction
patterns of dried and humidity
there
was
no
change
exposed
ibuprofen
upon removal of water. It was speculated that
essentially
sticks
to
the
moisture
the crystal surface . Both compounds
had a very low water content but according to
( 13 )
samples.
in the crystallinity of both powders
Ahelc
et
al
they still require special attention because the water
distributes only in amorphous regions. X- ray powder diffrac tion
patterns indicated different crystal structure for the
racemate and its enantiomer with very
peaks
different
deflection
especially at low angles of the spectrum. Ten minutes
grinding resulted in a loss of crystallinity for
both
pow -
ders with a decrease in deflection peaks intensity (fig. 2).
For ground S(+ ) ibuprofen , some peaks were more
intense
at
low angles of the spectrum ( 11-15). This difference might be
108
due to
the
decrease
modi f i c ati o n
of
in
the
particle
c rystal
p a rameters. presented in
size
but
nature.
table
Th e
confirmed
ind i cate d
a
therm odynam i c
these
co n c l u -
sions . The endotherm temperature was 20 ° c l ower f o r the S (-)
enantiomer and the enthalpy of fusion averaged 35
Jtg
less
than the racemate. Upon grinding there was a decrease in the
enthalpy
entr opy
of
fusion
changes.
for
the
ibuprofens
proporti o nal
to
There was no modifications in the melting
points in either case confirming no rearrangement of the in ternal
lattice ( s ) .
Nevertheless .
although
the
internal
structure might not have been modified , upon grinding
was
a
there
decrease in the particle size of ( + ) -S-ibuprofen and
the compound became difficult
to
handle
as
a
result
of
flowabili t y los s .
Based on the monolayer gas adsorption theory
lated
the
we
calcu-
true surface area of the two compounds using the
B.E.T. equation . The B . E . T. values for racemate and the S( T)
were respectively 3 . 4 10-l ( 0.01 ) and 2.8 10 - 3
( 1.9
m2 1gram indicating a specific surface area more
than
isomer
10- 4 )
100
found
times
smaller for the enantiomer. Although it has been
that
BET
values
vary
between
sources
( 14 ).
the
amplitude of this difference might be a potential problem of
the
formulation
specifically
in
the
dissolution / bioavailability behavior. The particle size with
a mean rangi ng from 83 to 149 um was very large compared
5 - 38
um
for
to
the racemate ( 14 ) and was responsible for the
109
excellent flowability of the
ele ctro n
microscopy
bulk
material.
The
scanning
obs ervati ons in fig.3 indeed confirmed
the descriptiv e a nalysis and the unusual nature of the c r ystal
surface
o f (+)-S-ibuprof en . The optical isomer existed
as large b ox y needle s consisting of the
crystal
un i t .
All
particles had a very sm ooth surface.
Bulk and tapped densities averaged 34 and 56
tively
and
were
similar
to
compressibility / flowability as
Vo / V
against
the
number
initial volume and v the
above
the
0.1
the
racemate
evaluated
by
%
respec -
analyzed. The
pl otti ng
Log
of t aps ( fig .4 ) . where Vo is the
tapped
assymptote
volume.
was
consist entl y .
for the racemate and considered
poor ( 12 ) as compared to the S(T ) enantiomer. Thus the opti cal is omer might be a good candidate for direct compression .
Dissolution Kinetics
Because
of
the
unique dissolution characterisitics of
each compound. the rate plots presented in
ferent
slopes ,
and
consequently
fig.5
different
had
dif -
intrinsic
dissolution rates. Surprinsingly the IPR for S (+) ibuprofen
averaging 8.
ug.sec -1 .cm -2 was smaller than the IPR of the
racemate with a mean at 11.6 ug.sec - 1 .cm- 2 . It appears that
under
our
experimental conditions , the dissolution rate of
the enantiomer was limited by its very
( rather
small
surface
area
than enhanced by its high solubility ) . In addition.
110
ag e d c ompac t s at Room Temperatu re f or
analyzed
3
d ays
a nd
further
u nd er the same co ndi ti o n s had lowe r in t rin sic dis -
solution rat e s I DR than ori g inal disk s immed i ately
ana ly zed
a ft e r ma nuf a ct ure ( f ig. 6 ).
Heat of Soluti o n
The
s olubilit y
of the tw o ibupr o fens wa s determined at
different temperat ures in v ari ous buffers .
equat i on
relates
the
s olubilit y
per c ent to the inverse of the
The
van ' t
Ho ff
in mole fraction or mole
absolute
temperature
of
an .
ideal solution:
-Log Xi
(
~Hf / RT )
+ constant
where Xi is the ibupr ofen c on c entration
ti o n ,
~Hf
in
mole
frac -
the heat of s ol u ti on , R the perfect gas constant
and T the absolute temperature in degree Ke l vin .
The
Van ' t
Hoff
plots for rac - ibuprofen ( fig.7 ) and (+)- S- ibuprofen
(fig . 8 )
y ield heats of solution at different pH ' s presented in table
2. The slopes varied with pH indicating that
sorbed
by
the
systems
the
heat
ab-
during dissolution varied with the
extent of ionization . The energy of solubilization decreased
with ionized species and at pH 7.7 , the heat of s o lution be came slightly ex othermic for both
111
ibupr ofens .
This
result
confirmed
that
when i o nizati o n was the principal f a c t or in
di ssolut i on ( at high pHs ) this phen omena
ener gy .
At
ohis
level
ibupro fen co n s istently exhibited
than
its
actuall y
rele ase d
of l ower [ H+) c o ncentrat i on s . S ( - )
l ower
heat
of
s olu t i ons
racemate form showing that the enantiomer i s more
soluble in aqueou s media . For example the aqueous solubility
of ( - )- S- ibuprofen in a pH 7 . 7 ph o sphate buffer at 37
°c
6.0 mg / ml compared to 5.0 mg t ml f o r the racemate . Under
was
our
experimental settings . at pH 4.5 ( pKa of ibuprofen ) the con centrations were most
variable
and
yield
close
to
zer o
slopes for both ibuprofens.
CONCLUS I ONS
The physical pharmacy profiles of S(+) ibuprofen and its
racemate form were compared . Thermal analysis indicated that
the
optical
isomer existed as a different crystal form ex -
hibiting different solid state properties which could be
concern
for the formulator. Particle size was increased and
the flowability was
large
boxy
crystals
improved.
The
enantiomer
as
solid
formulation.
fact the intrinsic dissolution rate of this compound was
found smaller than the IDR of the
not
existed
with unusually low surface area. This
might be a major limitation for an oral
In
of
predicted
documented
in
from
the
rac-ibuprofen
which
was
the solubulity data and not previously
literature.
112
Solubility
determinations
revealed
that
( • )- S-ibuprofen
media at pHs higher than 4.5
ticipated
f r om
a
was mo re soluble in aqueous
but
both
ibuprofens
were
indicating
from
a
the
at
It
pH
the
has
been
pharmacokinetic / pharmacodynamic stand po int
( 9 ).
however,
considering
therapeutic
these
im -
elements
physical pharmacy special attention should be given
formulation ,
an -
exothermic
this
energy.
that formulating S (+) ibuprofen might be a
provement
extent
slightly
that
solubilization pr ocess released some
argued
to
review of the stereochemical literature.
Al so at pH 7.7 heats of solution
for
not
to
of
the
in order to overcome the potential problems of .
dissolution . low melting
levels
and
compatibility .
Thus.
S (+) ibuprofen might be readily absorbed through the j ejunum
but its dissolution characteristics could be a rate limiting
step
of
bioavailability
lower doses
are
for an oral solid dosage form. If
required ,
it
is
anticipated
that
S(+)
ibuprofen could be a good candidate for direct compression .
In summary,
ibuprofen
we
believe
is
that
cer t ainly
the
formulation
of
(+)- S-
provided
the
unique
achievable
characterist i cs of t his drug are kept
significant
racemate
and
diffe rences
the
between
potential
in
this
mind.
There
enantiomer
advantages
are
and t he
(therapeutic
and
biopharmaceutical ) of the S(+) drug might be considerable.
Aknowledgements :
This
Corporation.
greatly
We
work
was
supported
appreciate
113
by
Ciba- Geigy
the input from Dr. G.
Lukas . Director of PPT and we thank Dr. D. Bauer from
Ethyl
Co . who provided us with the ibuprofen.
REFERENCES
l .
A.C. Cartwrig ht, rLI....A.. Vo l 24. pp 115-116, 1990
2.
PMA Ad Hoc Committee on Racemic Mixtur e s.
Pharm.Tech. pp 46 -52. May 1990
3 .
A. Repta , M.J. Baltez or and PC Raussal , J. Pharm .Sci.
Vol 65.2, pp 238-242 , Feb. 1976
4 .
A. Formi. I . Moretti and G. Torre .
J.Chem.Soc.Perkin.Trans. 2 . pp 791-797. 1984
5.
G.P . Bettinetti. F . Gioardano , A. Italia .
R. Pellegata and P . Ventura, A. G.P .I.
, 3.
pp 232-240.1988
6.
S-Tsuen Liu and A. Hurvitz , J . Pharm.Sci . . Vol 67.
7.
Harry R. Britain . Pharm Res . . Vol 7. 7 , 683-690.
5 , pp 636-638 , 1978
1990
8.
C.D. Herzfeldt and R. Kummel , DD I.P., 9(5),
767-793 , 1983
9.
J.H. Wood, J . E . Syarto and H. letterman .
10.
D.J.W . Grant , M. Medhizadeh , A. H. Chow and
J.Pharm.Sci. 54 , 1068 -10 75 , 1965
J.E.Fairbrother Int .J. Pharm. 18 , 25-38, 1984
11.
C.D. Herzfeldt and R Kummel , D.D . I .P .. 9(5 ),
114
767-793. 1983
12.
J.E. Rees. Bull.Chem . Pharm .. 112. 216 - 220, 1973
13.
C. Alhec and G. zografi
In press Int.J.Pharm.
14.
A.J.
1991
Romero. G. Lukas and C.T . Rhodes
In press Pharm . Acta Helv .. 1991
115
Table I
: THERMAL ANALYSIS OF IBUPROFEN
"As Received"
MELTING
RANGE
ENTHALPY
OF FUSION
.....
(j)
Racemate
S( + )
75 - 77 oC
53 - 55
R(-)
0
13 5 .1 J/G
0
c
8 6. 8 J/ G
"Ground"
MELTING
POINT
ENTHALPY
OF FUSION
75 . 3
11 5 . 2
c
55. 1
J/G
c
8 2 . 3 J /G
53-55
c
87.0 J/G
TABLE II: HEAT OF SOLUTION FOR
(+)-S AND RAC-IBUPROFEN
pH
~H
(KJ.lolOLE
RAC-IBUPROFEN
....
....
-;z
1.3
4.5
6.0
7.7
( + )-S-IBUPROFEN
1.3
4.5
6.0
7.7
32.2
-0.3
38.8
-5.2
51.5
-0.04
29.9
-16.4
_,
_,
• • K)
Figure l
cross-section ot the Modified Woods Rotating Apparatus
JJ
disk holder
i-------i--die hole
ie
118
Figure 2
f Grinding
"as
Diffraction
db.;
" p r( Bof
)Effect
"asibupr
re cofen
Patterns:
. en
S (+)
ei v ed"
X-R~:, profen
". ga rco.217
( C) :
R
(A Powder
) : S(+)
l"
r eceiv ed
0
,
( A)
(8 )
I
UVI 11,11~~
'
119
1
\ 1
I
(CJ
Figure 3
Scanning Electron Microscopy Photographs of S(+) Ibuprofen
so , 250, soo and lOOOX Magnification
1 20
121
0
~
\[)
N
'
•a.
•
I-
0
0
N
- "'
0
0
z
"'. •
••
:I
<1>
E
<1>
u
ro
0
([
\[)
~
a~
>
>
....
~
qi
0
0
c
\[)
0
>
a
•
0
-'
'
us+
f
•
;
:I
u:
•Cl.
I-
'
--~. .. .•... ,.
0
l
.... ...... . . ....
0
("')
N
0
0
0
ci
ci
ci
0
0
A/OA 1101
122
q
..•
J
0
(')
...
•
.
••
c
0
:;::
:I
•
:I
0
..,
•
Ill
-ro Ill
c
E-
Ill 0
.!
A
QA
c
~~
a: -
u
·;
c
·;:
,,• •
~
31
•...
oa
•o•
·~
•
-·
:
I
.:: I
-.
-.
": lt
~.
I II:
-.,, ,,
•. 8
j
I II:
...
0
0
0
(\J
i•
I-
0
.!
:a
:I 0
.!l~
u.
II.
~
0
I~
If)
-
a
0
0
<O
0
(')
(1111/ln) UOA•4UHUO:)
123
0
0
l()
-•..
c:
0
a.
::J
:e
-- ~
•~
•• •
+
Ci) Cl
c:
·a.
0
~
c:
0
:::
0
::J
0
i5
•
l()
(\J
::J
Cl
u::
I
j:
u
w
co
•.
i
~I
D
8
0
l()
(1111/ln) UOA1411HUO:)
1 24
.""
0 ;;I
!!!
cti
...0...
a.
=i
.Cl
....
:z:Q,
•
·r
Q
• ci
•... .!5
=i
~
."
0
0
•
a- i
·0
u. -A.
... ...
.•
0
:z:
c ....
> ..:
i
•
0
..
I
"
I
uxJloi
125
.
I
..
0
.,,c>
I
I
~
0
......
.
., c:;
0
I
!!!
c
-.
ti
0
a.
:::IJ
"!
0
:z:A
0
:9
I
Cl)
co
..!.
• !.
....,
...... •
0
:::IJ
0
0
a. i
u:: 0
A:
-..
<l
0
::c
~
•
>
...,.;
i
.
0
0
...
I
(IX)Ool
126
0
.. 0
I
';-
-
._
Ille
......
(
MANUSCRIPT V
STEREOCHEMICAL ASPECTS OF THE MOLECULAR PHARMACEUTICS
OF IBUPROFEN
127
ABSTRACT
Thermal
analysis.
thermody namic s
of
s oluti on
and
molecular modeling of (+)- S- ibuprofen and rac - ibuprofen gave
information on how heterochiral o r
would
affect
the
processing
homochiral
interactions
of ibuprofen. The study con -
firmed that rac-ibuprofen exists as a true racemate
10%
with
a
eutectic pure enantiomer composition. Both the racemate
and the ( +) isomer crystal unit cells include four molecules
and
crystallize
tively.
Thus
between
the
the
in
the P2 / c and P2 space groups respec 1
1
forces were
different _
intermolecular
crystals.
As a consequence the (+) enantiomer
lattice was more fragile but only slightly more soluble than
the
racemate
in
aqueous
media. The solid - state structure
contributions to solubility were very different between
two
crystals
the
(AH~( +)- 51 . l and AH~ac- 32.2 in KJ mole ) but
the standard free energy of solution were found to
be
com -
parable for both compounds.
INTRODUCTION
It
is
generally recognized that (+)- S-ibuprofen is the
enantiomer of ibuprofen inhibiting
thetase
(1 - 2 ) .
Stereospecif ic
the
prostaglandin
and
conformational
characteristics of this isomer are required for
128
syn-
interaction
with
the cell re ceptors responsible for the therapeutic an -
tiinflammatory activity. Another
could
be
the
consequence
of
chi ralit y
differences between the crystal habit of the
tw o s e parate isomers and the racemate. We now report a
con-
tinuation of ou r studies in this area ( 3).
We have previously
melting
point
and
reported
higher
solubility ,
smaller intrinsic diss olution rate s f or
the (+) isomer compared to the racemate form ( 3 ).
bination
of
tests
lower
used
in
this
study
is ,
The
com-
we believe ,
essential when one is investigating a chiral compound or
optically
an
pure isomer. For example , the thermal behavior of _
sobrerol and diastereoisomers was reported ( 4 ) and the crystal
structures
of
cytostatic
agents
(s tere oisome rs and
mixtures ) were elucidated (5), both in support
tion
efforts.
of solution of non steroidal
cluding
ibuprofen
formula -
in
antiinflammat or y
agents ,
in -
have been thoroughly studied by Pe cc i
al (6) to determine the
structures
of
In another study the thermodynamic functions
contribution
promoting
solubility.
teristics can be related to the
of
the
solid
e_t_
state
The physical charac -
molecular
packing
of
the
crystals under study ( 7 ) .
The combination of all these tests and
analysis
is
ibuprofen
enantiomers
before
a
unique
approach
that
has
to
been
the
the
comparative
formulation
largely
of
overlooked
the stereospecific synthesis became economically vi -
able .
129
EXPERIMENTAL
Materials
Racemic
Corp.
Co.
and
S(+)
ibupr of en were supplied b y the Ethyl
( Baton Rouge . LA ). Methan ol from the Fisher Scientifi c
was
used
for
ibuprofen
recrystallization and was of
analytical grade. Potassium ph osphat e monobasic
hydroxide were obtained
and
sodium
from the Malinckrodt and J . T. Baker
chemical companies , respectivel y.
Me thods
Thermal Analysis: Ibuprofen mixtures containing
enantiomeric
proportions
were
various
prepared by slow recrystal -
lization at 42 .6 °F from methan ol and after melting. Thermal
analysis
were
performed
on (-)- S- ibuprofen [S (+)J. (-)- R-
ibuprofen [R(-)] , rac-ibuprofen [Rael and mixtures
differential
scanning
using
a
calorimeter [DSC] from Perkin Elmer,
series 7. The heating rate was
set
at
5
°c 1minute
under
nitrogen flushing . Thermal endotherms were integrated t o obtain thermodynamic functions used
for
the
phase-diagrams.
The oretical solid-liquid equilibr ium curves were drawn using
the Prigofine - Defay equation ( eq.l ) for
pletely
dissolved
in
the
the
racemate
com-
melt and the Schroeder Van - Laar
130
(
equati on ( eq.2 ) for che simple eutectic formation (9). Thus.
eute ct i c temperature and compositions could be determined:
Ln4x ( l - x )
equati on 1
Lnx
equati on 2
where x is the mole fraction of the more abundant
tiomer
in the mixture. whose melting ends at Tf (°K)
and AHrac are the enthalpy of fusion of the
the
enan -
r ac emic
form
pure
AHs
S(+)
and
respectively: Tms and Tmrac are the cor-
responding melting points and R is the gas constant at 1.987
cal.mol- 1 . _ °K- 1 .
Crystal Analysis:
performed
on
small
Single crystal X-ray diffraction
crystals
o f (+)-S- ibuprofen from the
bulk compound. Reflection data was obtained from
and
processed
on
was
Ethyl
Co .
a molecular modeling program (Shelxtlr ).
Crystal data is given in Table I . The structure
was
solved
in the space group P2 / c
1
Analysis of the unit cell allowed
the identification of the
molecular
bonds
packing
and
hydrogen
network within the monoclinic crystal . Similar infor-
mation on the racemate was obtained from the literature ( 10)
and compared to the newly obtained molecular packing data of
the (+) enantiomer .
Solubility:
Excess
amounts
of
both
compounds
suspended in 0.0 5M aqueous buffered solutions at pH 1.3.
131
were
At
this
pH.
the
drug
is
essentially
unionized
since
ibuprofen pKa is believed to be in the range of 4.6 to
Screw
cap vials were rotated on a Labshaker for 24 hours at
temperatures ranging from 25
°
Quantitative
performed
analysis
was
to 52
°c
in
using
walk-in
ovens.
ultra
violet
spectrophotometry at selected wavelengths (220 and 264
The
the
5 .2 .
thermodynamic
nm ) .
functions of solubilities were evaluated
as follows:
The
chemical
potential
of a solute (s) in equilibrium
with its pure form (s) may be written as:
µ
µ
+
R'T *Ln x
The variation
equation 3
of the solubility expressed in xw (mo le frac -
tion in the solution ) can be integrated to
Cstant + ( - ~Ho
and
I
R)
>
(l /T )
equation 4
experimental data can be analyzed by plotting Ln of the
solubility versus l / T. The free energy of
solubility
at
a
given temperature can be obtained from
-R *T*Ln xw
equation 5
and the entropy of solution is derived from the third law of
thermodynamics:
132
e quat i on 6
Al l thermody namic parameter s were
the
racemate
ibuprofen
solubility ,
and
ibupr o fens.
S (+)
stereochemistry
was
( solid
investigated
analyzed
The
state
In
to
compare
c ontribution
of
structures )
to
a recent study, ther -
modynamic functions were used to predict
and
separate
the
roles of solid- state stru c tures from s o lute - s olvent interac tions in promoting the solubility of a solid
nonelectrolyte
(6) .
RESULTS AND DISCUSSION
Thermal Behavior
Thermodynamic functions f o r both compounds are
in
Table
II.
reported
Melting parameters ( Tm and 6H ) obtained from
the thermograms in figures 1 - 2 , we re used in equations l and
2 to obtain the phase - diagram in figure 3 (10 ) . Experimental
data were in good agreement with the theoretical
lines
in -
dicating the fusion of the Eutectic at about Teu_ 321 °K and
a eutectic composition of 10.0 % on each side. As previously
anticipated
( 3 , 10 )
the
thermal
analysis
confirmed
ibuprofen naturally occurs as a unique racemic
133
that
compound.
A
Peter son
"i"
ratio
of 1.77 was calculated from equati on 7
(9) :
"i "
equation 7
where Tmeu is the eutectic temperature as determined ex perimentally and from the phase diagram.
Although somewhat arbitrary
clearly
indicated
that
in
character.
optical isomers was 20 to 22
°c
enantioselective
impossible . The test of
perf ormed
a
posteriori
point
of
both
lower than the racemate f orm -
and eutectics were very close to the edges
any
rati o
ibuprofen has a strong tendency t o
crystallize as a true racemate. The melting
making
this
of
the
diagram
resolution by crystallizat io n
the
Prigofine - Defay
and
equation
was
the straight line in figure 4
confirmed the model.
Crys tal Packing
Perspective drawings of the
molecular
packing
in
the
crystals of (+)- S-ibuprofen are given in figures 5a-5c.
(+)-S-ibuprofen is more water soluble than the racemate
and
it
is
solubilitie s or thermal behavior in terms of
attractions
(3)
of interest to seek the basis for the differing
forces.
Although
having
the
intermolecular
same
number of
molecules , S (+) crystals have a totally different unit
134
cell
than
the
volved
racemace
in
( Fig.5a ).
homochir al
mechanical
The array of S mole cul e s in -
interactions
stability / strength
probably
spreads
the
of the crystal ( Fig.5b ) . The
preferential molecular arrangement in the P2 1 plan exhibit
some of the acid groups " face-up " and others "face down· so
that all the layers of
pairs
of
hydrogen
molecules
are
interconnected
with
bonds to carboxyl groups. As suming that
the top layer of the c r ystal is one. crystal surface is dif ferent
for
ibuprofen
the
two
compounds. Thus in the case of
there
are
more
hydrophobic
layer s.
There
exposed
carboxyls
C~)- S -
and
less
are several points of interest: _
first the gr eater number of crystallographically independant
molecules
in
the S crystals ; secondly there are no obvious
relati o nships between molecular packing in
the
racemate
structural
and
enantiomers .
Finally.
the
lattice
reflects different intermolecular environments. In order
pack
together,
molecules
of
requirements
to
of
lattice. A qualitative measure was the superposition of
two (+)-S- ibuprofen molecules involved in the same
bond
data
the same chirality had to be
somewhat flexed in order to meet the space
the
of
which
clearly
demonstrate
hypothesize that further
crystal
would result from the packing.
135
the
hydrogen
torsion (Fig.5c). We
elasticity
or
fragility
Solubility
The s o lubility of crystalline solid is determined b y c he
free energy changes from the solid state to a
indicated
in
table II. ~Go at 25
S C~)
the racemate than for its
the
differing
°c
solution .
is slightly higher f o r
isomer which may account
solubilities.
As
for
Similar conclusions could be
drawn from the analysis of fusion parameters.
The
enthalpy
and entropy contributions to water solubility as revealed by
the thermodynamic functions in table II. are very
suggesting
that
different
solid state structures are responsible for
these differences.
CONCLUS I ONS
It was confirmed that rac-ibuprofen naturally occurs
a
racemic
compound
(10)
with
as
a eutectic temperature ap -
proaching 320 °K. This behavior although quite conventional .
has
some
serious
implications
in
the formulation of the
biologically active stereoisomer. Heterochiral
formed
preferentially.
Thus,
titatively, the intermolecular network
crystal
unit
cells
of
reasonably
account
of
interactions
in
the racemate significantly exceeds
that existing within cells of the pure
also
interactions
both qualitatively and quan -
for
the
enantiomer
and
can
differing solubilities.
thermal behavior and further processing characteristics. The
136
literature
ind i cates chat in some in s tances. when the melt -
ing p o inc of
pure
ste re oisomer s
is
substantially
l ower.
these optical isomers are several fold mo re s oluble than the
correspondin g racemate ( 5,8 ) . In this case, the S isomer was
only
slightly more water soluble than r ac -ibuprofen . We at -
tribute this phenomena to the mole cular arrangement
lattice
tions
isomer
to s olubilit y were
and
the
(+)
the racemate. At pH 1 .3 the entr opy effect
( ~S )
equivalent
d if ferent
between
free
energy
intrinsic
rate of ( +)- S- ibupr of en. In addition , the crys-
tal lattice exhibi ted potentials for mechanical
perturbed
were
for the two c rystals. These results con -
firmed the l ow specific surface area and the slow
diss o lution
if
the
of ( +)-S- ibupr of en . Solid - state structu re contribu( ~H )
counterbalanced this effect and standard
almost
in
by
components
of
instabilit y
high hydr og en bonding af -
finities.
Acknowledgements :
We are indebted to Ciba- Geigy Corpo rati on
for the award of a fellowship and to Dr. Lukas , Director
P.P.T.
fo r
his
superv i sion
during
the
course
of
project. We thank Dr D. Bauer. Ethyl Corp. for providing
with single crystal X-ray diffraction data.
REFERENCES
137
of
this
us
l.
Pharmac o l og ical
of
~buprofen:
differences between the optical is omer s
evidence for metabolic inversion of the
( - )-R-isom er .
SS Adams , P. Bresloff
and
CG
Ma son
J.Pharm.Pharmac o.
28:256-257. 1976
2.
Stereoselectivity in Clinical Pharmacokinetics and
Drug
Development
D.B. Campbell. Europeen J. Drug Metab . &
Pharmacokinetics , 5(2 ) pp 109-125. 1990
3.
Approaches to Stereospecific Preformulation of Ibuprofen
A.J .
Romero and C.T. Rhodes accepted for publicati on in
D.D . I.P ., 17 ( 4 ). 1991
4.
Thermal
Behavior
and
Phase
Diagrams
of
Sobrerol
Enantiomers and Racemates.
G. P . Bettinetti . F . Giordano , A. Italia. R . Bellegata
and P.Ventura , A. G.P.I. Paris 1989 , pp 232 - 242
5.
Stereochemistry
Pr opanediyl bis
Molecular
of
the
Antitumor
Agent
(4 - Piperazine -2, 6,Dione ) :
Structures
of
4.4 '-( l , 2
Crystal
and
the Racemate ( ICRF - 159 ) and a
Soluble Enantiomer.
A.Hempel.
N. Camerman and A. Cammerman
2, 3453-56. 1982
138
J.Ame Chem.Soc .
6.
Solubility
and
Par titioning
I:
Solubility
of
Nonelectrolytes in Wat er .
S.H. Yalkowsky and S.C . Valvan i J.Pharm.Sci. 69 ( 8 ) .
9 1 2 - 921. August 1980
7.
Relationships
between
Solid-State
Enantiomers and the Co rresp onding Racemic
Structures
Compound s
of
in
Small derivatives
A. Forni . I. Moretti , G. Torre. S. Bruckner. L. Malpezzi
and G. DiSilvestro. J . Chem.Soc . Perkin Trans II
2:791-797 , 1984
8.
Enantiomers. Racemates and Resolutions
J.Jacques . A. Collet and s.w. Wilen J. Wiley 1981
a ) pp 32 - 43: b ) pp 88 -104
9.
2- ( 4 - Isobutyl Phenyl ) Pr opionic Acid
J.F. McConnell , Cryst . Struct.Comm. 3 : 73 - 75 , 1974
139
Table I: Crystal Data for (+)-S-Ibuprofen and Rae-ibuprofen
(+)-S-ibuprofen
,_,
""0
Cl3H1802
206.3 grams
Monoclinic
P21/c
12.46
8.03
13.53
Formula
Molecul.Weight
Crystal System
Space Group
a( i\)
b(i\)
c( i\l
()l(i\)
(
~(
112. 95
e
)
Rae-ibuprofen
C13H1802
206.3 grams
Monoclini c
P21
14.67
7.88
10.73
99.36
# of molecules
in the cell
Density (g. cm
CuK GI Radiation
4
)
1. 098
4
Table II: Thermodynamic Functions of Melting and Solubility
Ibuprofen
Melting
Tm
Rae
(~K)
Melting Point
349
(+)-S
327
(-)-R
327
·1
>--'
ti>
oH
Enthalpy
oS
Entropy
(KJ.mole
·1
(J.mole
)
25.5
17.9
17.9
73.2
54.8
54.8
32.2
51. 5
6.7
73.4
30.2
29.6
·1
·~ K
)
>--'
Solution
·1
oH
( KJ.mole
oS
( J.mole
oG,
< KJ.mole
)
·1
.o K
)
·1
)
l'lts• t
Typlcal DIC ladetllenM
Al ,... rH-llluprot .. M4I I) ,... (+)-1-..._,ot•
"'-C
71L 111
TZ
• 5. c
•c
11173..QZI aJ
117. -
(A )
J/ 1
4 1• .S ••
35. 0
""121
•c
30. 0
l 5. 0
20.0
ts.:
10. 0 ·
5.0
,__-----+--)4 - + - - -
~
n
35.0
'1nc •c
'5!. HD •c
''·o"' •c
l
......_
!le l t.a N
'"/
10.C
... 0
.'.! D.
~z
;a._.
I
i 1Zl.S7'&1
r.'."8 1:9
l~'.___ -- ~~:: ~~ -
Q
I \
:s.o
I
\~-------­
10.0
~
5.0
.v
.....
.... o
....
1111.D
70.D
....
T~<"O
142
ID.Q
IDD.0
,..... 2
DIC TMrmo1rame of lbuprof..,
Dfft•rut Enantlom- Compoaltlon
I
40
I
I
50
I
60
70
143
I
80
0
0
0
C!
0
.I
i
f
e
. !J
I
_u
.2
iiJ
.a•
~J
ii!
&
~
~
41
41
E
0
!!
-
-•
0
-
Cl
I
ii'
s:.
E
0
!!
... · D ····
+
c;;
Cl Cl
-41
Cl
0
0
IO
IO
0
0
s:.
0
-•
c:
0
;
c:
.2
u
u
•
~
~
....................
&I.
&I.
•
•
0
2
0
2
0
C!
."
.
'""o
0
•
{)I.) •Unl8J9dW9,l
144
0
0
oo
0
"
-
FICllll'•'
Teet of the Pritoflne-Det•r Equ•Uon
- 1.JO ~---AH--.-l-1.-1-K-.a..ol----,.-.K-, 1
0
0
-...
)i
I
;c
-1.70
3
-2.10 L---~-~--~-~
2.84
2.91
2.98
145
Figure Ila
Cryatal Unit Cel of (+1-S-llluprofen
146
Flfllre lb
Crrttal Lattice of (+ J-1-lbuprolen
147
FIOur• le
luperpo1illon of (+)-1-lbuprolen moleculH lnwolved
In the Hm• hydrogen bond
';,....:· _
,
,'
..:'-
>-\
\
148
MANUSCRIPT VI
FORMULATION OF THE BIOLOGICALLY ACTIVE
STEREOISOMER OF IBUPROFEN
149
ABSTRACT
As part of an effort to formulate the
tive
stereoisomer
pharmaceutical
ibuprofen
of
processing
ibuprofen.
on
biologically
the
effect s
rac-ibuprofen
and
not
crystals were compared. This comparative anal ys i s
found
Tablets
so
compression
excipients
decreased
fully
the
by
DSC
acceptable.
enthalpy
ibuprofen and compaction induced low
indicated
appeared
most
formulated showed rapid diss olution
and other tablet properties were
with
It
possible - - formulate the (+ ) isomer using
wet granulation , however direct
promising.
of
( + )- S-
is a unique concept used in stereoselective formulati on.
was
ac-
of
temperature
Mixing .
fusion
of
eutectics
endotherms. Stress storage of the (+)-S-
ibuprofen seriously affected the handling properties of
the
dry formulations.
INTRODUCTION
With
the
exception
of Naproxen, all profens currently
used as non-steroidal antiinflammatory agents
as
racemates
( 1).
For
most
are
marketed
of these drug substances the
dextro-rotary or S (+) optical isomer seems to be responsible
for the therapeutic activity , that is the stereospecific inhibition
of
pharmacokinetic
cyclooxygenase.
In
addition ,
various
reports have been published suggesting that
150
for some of these aryl p r opio ni c acids. bioinve rsion of
inactive
enanti ome r
the
int o the active is omer S (-) take place
in vivo. by enzymatic mechanisms ( 2 ) . For ibupr ofen. a s much
as
33% of the S (+) fo rm could result fr om this biotransfor-
mation. Thus a racemate (contain ing 50% of R( -)) could yield
2 / 3 of active ibupr of en in the systemic ci rculati on ( 3 ).
Unfortunatel y, this pharmac okinetic
rationale
is
only
one of the factors which must be considered for the success ful
f ormulation of the pure enantiomer in a
system .
Previous
investigations
shown that the presentl y
relatively
what
of
available
delivery
by Romero et al ( 4 ) have
(+)- S-ibuprofen
has
a
small specific surface area which might be somean
solubility
impediment
was
found
to
bioavailability
(although
higher than the racemate). Some han -
dling properties such as flowability were
while
drug
greatly
impr oved
compatibility screening indicated substantial crystal
distortion under processing ( low temperature eutectic ).
molecular
packing
structure
in
the
crystal
The
lattice was
elucidated ( 5 ) and it was concluded that (+)- S- ibupr of en exists
as a totally different crystal than the racemate form.
Analysis of homochiral interactions also revealed
enantiomer
that
the
crystal might be fragile and more susceptible to
distortion.
The
study of a pure enantiomer as an answer for a phar -
maceutical problem is not new in process development ( 6 ) but
the
analysis
comparing the relationships between molecular
151
aspects of crystal
performances
stereochemistry
and
biopharmaceutical
appears to be a new approach co stereospecif ic
drug formulacion. Recommendations on the processing of
(+) -
S-ibuprofen were f o rmulated based on this study and previous
reports.
EXPERI MENTAL
Materials
Ibuprofens
(racemate
and
S isomer) were obtained from _
the Ethyl Corporation ( Baton Rouge. LA ). Fast
flow
lactose
(Sheffield). polyvinylpirrolidone ( GAF) and Explotab ( Edward
Mendell) were selected as diluent. binder
respectively .
and
disintegrant
Simulated intestinal fluid was prepared using
potassium phosphate monobasic ( Fisher Scientific ), distilled
water
and
the
pH
adjusted
to
7.4 with sodium hydroxide
(Malinkrodt) .
Methods
A formulation program was undertaken, including
patibility
screening,
tablet
manufacture
a
com-
a nd analysis of
biopharmaceutical properties.
As
in previous investigation (7), emphasis was given to
the study of crystal
modifications
152
during
processing
and
their
effects
on
the perf o rman ce of the final product for
both ibuprofens .
Thermal a nd Comp a t i b i lity St udies
Calorimetry was the analytical tool selected
studies.
A
Perkin
and
these
Elmer Series 7 thermal unit was used. A
P7500E computer. interfaced with the system.
aquisition
for
integration
of
the
all owing
thermal
data
endotherms .
Systems o f increasing concentrations of (+)-S- ibuprofen
excipients
and
were mixed at room temperature for 24 hours on a
Labshaker rotating at 40 RPM. Table
design .
perimental
Samples
were
I
summarizes
withdrawn ,
the
ex -
tested
for
differential scanning calorimetry ( DSC ) profiles on a Perkin
Elmer
differential scanning calorimeter and stability after
exposure to the
chambers ;
37
followi n g
stress
conditions
in
humidity
°c at 85 % relative humidity ( RH) , and 50°c at
75% RH for one and seven days. Macroscopic observations were
also
recorded.
Tablets made of 67% (T)- S- ibuprofen and ex -
cipients were ground and analyzed for endotherms
to
assess
crystal mod if i cation.
Attempts made to formulate the (+)- S-isomer into tablets
by
wet
granulation
we r e unsuccessful . It was found impos -
sible to dry the gran ul e s at
40°c.
A
temperature s
betwee n
30
a nd
direct compression formulation was deve l oped using
the same excipients used for the racemate
153
and
compared
to
tablets havi n g t he same rac - i buprofen co n centration. Tablets
of rac-ibup r ofen were prepared according
design
already
ibuprofen.
validated
23%
disintegrant
Fast
( 6).
flow
( Exp lotab )
6%
and
with
five
stearate
form ulation
binder
prepared
r otat ing at 30 RPM for ten minutes
1% magnesium
a
Dry mixtures of 67% (-)-S-
lact os e .
were
to
for
using
and
further
minutes.
hydraulic press was used to produce 350 mg tablets
lbs
compressi on
force
for
a
3%
a V- b le nder
lubricated
A Carver
at
2500
15 second dwell t ime. Three
regions of the compaction spectrum
(low.
intermediate
and
high ) were als o investigated f o r crystal distortion . The ex perimental protocol included measurement
of
disintegration
time , hardness . dissolution profiles and thermal analysis.
When possible statistical anal ys is was performed at
95%
confidence
the
level either using a paired t test or ANOVA
to compare (+)- S-ibuprofen and Rae-ibupr ofen powders.
RESULTS AND DISCUSSION
DSC profiles of ibuprofen mixtures were
integrated
for
melting points , enthalpy of f u sion and heat capacity. Mixing
with
excipients
parameters
of
had
fusion
an
influence
on
the
thermodynamic
largely due to increasing amounts of
impurities (tabl e II ) and the extent of surface crystal dis tortion
appeared
concentration.
inversely
Similar
proportional
observations
154
were
to the ibuprofen
made
f or
the
racemate.
The mixtures became more difficult to handle with
increasing conce ntrations of (+)- S- ibupr o fen. Sample s of the
formulations.
kept
at
different temperatures and relative
humidities ranging from 35° t o 50°c and 35
tivel y
were
withdrawn
at
to
85%
re s pec-
dif ferent time intervals . After
only 24 hours under the above storage co ndition s it was
possible
to
handle
any
50°C , 37 °C and 75% RH . All
had
im-
mixtures which had been sto red at
high
strength
minif or mul ations
"melted ". After one week of storage at r oom temperature
the thermal analysis revealed that eutectic f ormation lower ing
the
melting
point
and
heat
of
fusion
(tabl e III ) _
apparently made the ibuprofen mechanically unstable.
It
appeared that S(+) had a greater tendency t o form eutec -
tics than the
racemate
with
the
selected
pharmaceutical
excipients. The crystal distortion is further facilitated by
stressful mixing ( e.g. planetary mixer ). This observation i s
in
agreement
rangements
with
in
localization
the
the
of
previous analysis of molecular ar -
crystal
unit
and
the
peripherical
the intermolecular "Hydrogen" bond network
of the (+) enantiomer (7).
Biopharmaceutical
analysis
was performed on the 350 mg
tablets. Conventional rac-ibuprofen formulations ( using
Granulation )
Compressible )
8 '( 4 ')
were
compared
formulation.
to
(+)- S-ibuprofen
Disintegration
times
Wet
(Directly
averaged
for the S (+). These tablets did not erode as did the
corresponding racemate tablets
155
averaging
53 '( 6 ') .
instead
they
disintegrated rather quickly and became a paste stick -
ing to the grid. The mean hardness
30 (1 9 )
for
these
tablets
the
extent
forms
(7 )
was
N.
Using a procedure testing
modifications
in
solid
for
dosage
of
crystal
thermodynamic
parameters of ground tablets were analyzed to compare (+)-8ibuprofen
to
the
racemate.
Results
of
the calorimetric
analysis are presented in table IV and figure l. Thermal en d othe rms
of
mixtures
after compaction indicated a "cl ear "
eutectic ( first endotherm) at lower temperature. The
tic
also
appeared
proportional to the
found
fo~
the
after
three days of mixing and was not .
compaction
racemate
level,
(7).
of
compaction
between
different
as
was
previously
As predicted from molecular
modeling the manifestation
are
eutec-
lattice
rearrangement
upon
(+)- 8-ibuprofen and the
racemate.
Although
higher
intrinsic
dissolution
found for the racemate (partly due
to
size),
(+)-8-ibuprofen from the
the
dissolution
for
67%
its
rates had been
directly compressible formulation was faster
racemate
(Table
small
than
58.2(12)%
for
the
V). The time for 50% dissolution of (+)-8-
ibuprofen was three times smaller (figure 2 ) and
minutes
particle
( average ( 8D ))
of
at
twenty
(+)-8 isomer had dis -
solved whereas only 22.8(4) % of rac-ibuprofen was released.
It is speculated that the presence of excipients and crystal
156
distortion might be b ene ficial in enhan cing the
dissolution
of the S isomer.
CONCLUSIONS
The
results
of
this
study
reporting
preformulati on profile ( 6 )
crystal
particles
are
were
in a gr eement with a
(~ J -S-i bupr ofen
that
very "fragile" and loose their han-
dling properties when stressed or ground. Therefore
processing
will
further
require a gentle mixing. Crystal arrays of
(+) and (-) molecules in racemic crystals are
totally
dif -
ferent than molecules in the lattice of (T)-S- ibuprofen ( 7 ).
As a consequence. crystal distortion was significantly
ferent
between
the
two
compounds.
In
dif -
additi on
to
modification of the crystal habit. the S isomer exhibited
a
eutectic upon compact ion . The dissolution of (+)- S-ibuprofen
from the dosage
forms
racemate .
is
It
was
faster
hypothesized
decreased upon comp ression and / or
the
when
that
the
compared
the
to
particle
excipie nts
the
size
improved
dissolution . Another result of the thermal analysis in -
dicated
protected
that
the
as
observed
ibuprofen
for
crystal
the
racemate.
excipients
from further distortion.
Lower strength formulations had decreased heat of fusion but
improved mechanical properties .
In conclusion a directly compressible formulation of the
S (T)
isomer
is a
feasible alternative to the conventional
157
wet
granulation
ibuprofen.
formulation
currenoly
used
for
molecular level , suggestions for future work can
lated.
Research
for
neutral
excipients
properties and the effect of storage
tions
could
bring
techniques ) appear
preserve
rac -
Based on the analysis of the stereoisomer at the
the
S
practical
to
be
critical .
It
formu -
including
under
answers.
be
various
Mixing
is
flow
condi -
(time
and
important
to
crystal intact before tabletting. Finally.
the effect of storage on the biopharmaceutical properties of
the
tablets should be studied. There was eutectic formation
resulting from compaction.
Acknowledgement s: We greatly appreciate the support provided
by Ciba-Geigy Corporation and the
input
of
Dr.
G.
Lukas
Director of Pharmaceutics and Pharmaceutical technology.
REFERENCES .
1.
Stereoselective Metabolism of Antiinflammatory
2-Aryl Propioniates.
J.M. Mayer
Acta Pharm.Nord. 2(3 ), 196 - 216 , 1990
2.
Metabolic Chiral Inversion of 2 Aryl Propionic
Acids: A Novel Route with Pharmacological
158
Consequenses.
A.J. Hu"t and J . Caldwell
J . Pharm . Pharmac ol. 35:693 - 704. 1983
3.
An
Evaluation
of
Ibuprofen
Bioinversion
Simulation.
A.J. Romero. R. Rackley and C.T. Rhodes
Chirality. III ( 5 ) . 1991
4.
An Approach to stereospecific Preformulation of
Ibuprofen.
Accepted for publication D.D.I.P. 17(4 )
5
Molecular Biopharmaceutics and Stereochemistry of
Ibuprofen. Unpublished report
A.J. Romero, F. Clarke and C.T. Rhodes
6.
Utilization of an Enantiomer as a Solution to a
Pharmaceutical Problem: Application to
Solubilization of l,2-Di ( 4 - piperazine -2.6 Dione)
Propane.
A.J. Repta.M.J. Baltezor and P . C. Bansal
J.Pharm.Sci , 65(2):238-242, 1976
7.
Monitoring Crystal Modification in Systems
Containing Ibuprofen.
159
by
r
A.J. Romero. L. Savastano and C.T. Rhodes
submitted for publication
160
Table I: Experimental Design
a
Formulation
,_,
(+)-S-ibuprofen
Binder (PVP)
Diluent (F.F.Lactose)
Disintegrant (Explotab)
Lubricant (Mg.St.)
A
17
6
73
3
-
a
B
37
6
53
3
-
O>
*· Tabletted at several compression forces
(ranging from 600 to 15000 lbs)
a. subjected to stress storage
,,
c
57
6
33
3
-
*
a
D
67
6
23
3
l
a
E
77
6
13
3
-
%
%
%
%
%
Table II: Thermal Analysis of (+)-S Formulation (24 hr mixing)
Formulations:
Pure
ibuprofen
Thermodynamic
Functions
A
c
B
E
I-'
m
a
0
(\)
Melting Point (Tm in
K)
Enthalpy of Fusion (KJ/mole)
average +/- (SD)
327
18.3
(0 . 2)
a
325.5
11. 6
( 1.1)
a
a
325.8
14.6
(1.6)
326.4
16.0
(1.1)
a. Relative standard deviation (RSD) less or equal than 0 . 05
a
326.4
16.5
(1. 0)
Table III: Thermal Analysis of (+)-S Formulation (One week storage at RT)
Formulations:
Thermodynamic
Functions
Pure
ibuprofen
Melting Point (Tm in
B
324.2
325.0
c
E
a
0
,...,
A
K)
327
325.3
325.5
Ol
c;I
Enthalpy of Fusion (KJ/mole)
18.3 +/0.2
10.5
12.7
16.2
a. Relative standard deviation (RSD) less or equal than 0 . 05
15.4
'
Table IV: Thermal Analysis of (+)-S-ibuprofen Tablets (67%)
Effect of Compaction
1
Compression Force
1
.6H
Tm
0
(KN)
( K )
2
2
Tm
AH
0
(KJ/mole)
.( K )
(KJ/mole)
a
,_,
Ol
ol>
0.0
o.o
o.o
2.7
5.3
10.7
22.2
26.7
35.6
44.5
b
-
-
327.8
18.3
-
-
325.7
17.3
c
314.5
1.
79
325.4
16.8
313. 9
318.1
314.3
314 .1
313. 8
314 . 5
314 .0
0.66
0.92
0.53
0.53
0.68
0.35
0.41
325.1
324.7
325.7
325.9
325.4
326.2
325.7
16.9
16.2
17.7
18. 6
17. 5
19.0
18.5
a. Pure (+)-S-ibuprofen "as is"
b. 67% (+)-S-ibuprofen formulation after ten minutes of mixing
c. 67% (+)-S-ibuprofen formulation after three days of mixing
Table V: Biopharmaceutical Properties of Ibuprofen Tablets
Ave rage +/- (SD)
Time for 50%
(Minutes)
(+)-S-ibuprofen
17
% Dissolved
in 20 Min .
Hardness
(KN)
5 8.2(12)
30.0(19)
Disintegration
Time (Minutes)
8 ( 4)
..
'-'
Ol
(JI
Ra e-ibuprofen
53
22.8 (4)
2 2. 2 (5)
53 ( 6 )
.
•
0
OI
a:
.:!
:a• •
• e
... li
...
0
s50
c,,c
.
!~
~
0
A
:ao
:. g . .i;·.:::::::·::·:::; .•....
...
... f
0 -!
•
'()
•
••..
0
; Ill
i5
.
....
: !'.'.~:::.~..
e
...•
_j)
. -~· ·· ·· ..
··~-0
0
'°
0
!; u
a
ii:
·-
166
.J
.
0
a
~
..
:II
.-.
,,
::
•
E
...•
ii
:§
•
a:
I
•...
•I
ti
0
ti
c
ti
E
ti
-0
u
Ill
I
0
I
+
I
0
o-;
<IJ
~a:
..•
:I "Cl
:e
0
!
...
a. •
0 a:
I
c
Ill
-;
.....
c
+
g u;
s
0:I
-
....s
•
;;
0
•• 0c
•
..:
Q
.
C"1
l
~
!; 0
a
u::
167
CONCLUSIONS
168
CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK
In the early stages of this study. I defined
methods
that
made
possible
tests
properties that affect the processing of ibuprofen.
shown
that
crystal
characteristics.
standards ,
was
had
very
These difference s had a
dramatic effect upon the selection of processing
and
It
different sources of ibuprofen , all meeting the
United States Pharmacopeia compendial
variable
and
the identification of crystal
parameters
not all the sources of ibuprofen can be regarded as in -
terchangeable . This report is one of the first promoting the use
of
different standard operating procedures (SOP's) for
sources from different suppliers of ibuprofen.
Also .
given
the five sources of this NSAID agent , it is conceivable that
it might be advantagous to make USP standards more rigorous.
I
have no reason to doubt that other drugs may also require
this adjustment of the formulation process if they
tained
from
are
ob -
different manufacturers or different synthetic
routes.
This
study
underlines
gineering and strict
development.
the
physical
importance
pharmacy
of crystal en -
profiles
drug
I was able to demonstrate , with relative con-
fidence that sintering is the molecular mechanism
ibuprofen
in
by
which
transformation is achieved during formulation.
resulting hydrophobic network within tablets was
169
the
A
cause
of
many
formulation
problems with chis compound.
analytical tools usually re served to other fields
maceutical
proved
research
extremely
regular
( e.g.
useful
SEM.
and
DSC.
should
formulation / preformulation
Several
of
phar -
molecular modeling )
be
incorporated
activities
in
especially
when single isomers are being considered as replacement
for
racemic drugs.
The same methodologies were applied to the
the
(-)-S- ibuprofen .
there
and
inactive
might
racemate
be
with
of
It is well established that drug sub -
stances with a chiral center may
active
analysis
exhibit
pharmacologically -
stereoisomers. When this is the case.
indeed
a
powerful
argument
to
replace
the single pure enantiomer. The rational for
this change has to be carefully reviewed.
It
is
believed
that
this
thesis is one of the first
report to demonstrate that in addition
ferences
(previously
to
biological
publicized ). rac - ibuprofen and (+)- S-
ibuprofen are essentially different in terms
and
of
a
certain
degree
of
properties and solubility .
is
processing
formulation. Both quantitatively and qualitatively dif-
ferences in homo and heterochiral interactions
for
dif-
the
differences
For example, direct
can
account
in mechanical
compression
impossible and wet granulation is the only way to formu -
late the racemate. whereas the reverse may be true for the S
isomer .
Some
characteristics of the racemate must be kept
170
in mind b ut I c an assess with
i bupr ofen
is
a
co nfidenc e
that
the
(-)- S-
totally independent cry stal that s h ould be
for mul a ted a s a new drug subs tan c e .
co mparative
The
analysis o f mole c ular pharmaceutic s of
the racemate and stereoisomer s revealed a
u s eful
tool
f or
stereospecific drug development.
Enantiomer s should be co n -
sidered
as
as
early
formulation.
It
as
possible
new
c andidates
may well be that conclusi ons for ibupr ofen
could be applied to others chiral compounds and in
the
above
factors.
I
from racemate to pure
molecular
fo r
scrutiny
view
of
strongly recommend that all changes
isomers
in
be
subjected
to
extensive _
early preformulation program . using
the approach that was designed.
My
results
suggest,
however. that more studies on the
effect of stress storage of ( +)- S- ibuprofen and its formula tions
should
be
conducted. This compound is very fragile.
Formulators must proceed with extreme caution as it is
likely
that
( +) - S-ibuprofen will sublime even at room tem -
perature. Other studies are currently under
the
model
of
way
the
to
refine
stereospecific bioinversion in man . adding a
feed back inhibition from the S isomer. Such
involve
more
administration
position in a rat
model
and
studies
could
of different enantiomeric com the
monitoring
of
systemic
plasma concentrations of S and R ibuprofen. A faster absorption of (+)- 8-ibuprofen has been observed when
171
this
isomer
is
administered
within
the
racemate
form
Optimizing i buprofen therapy could lead to
than
alone.
administering
a
small amount of (-) -R-ibuprofen. In this thesis it was found
that chiral bioinversion occur essentially systemically. The
results
suggested
that
tremely
powerful
in
enantiomeric
studying
AUC
ratios were ex-
realistic
ibuprofen
pharmacokinetics.
The work reported in this dissertation has
demonstrated
that careful attention has to be directed to crystal characteristics and physico-chemical properties of raw
Consequently
USP
rigorous. The
results
change
of
the
compendial
clearly
approach
standards
indicate
materials. -
should
that
a
more
profound
to formulating drug substances is
necessary using combinations of new testing methods
available.
be
readily
This concept is in my opinion essential when the
drug substance includes a chiral center and developing
isomers might be a possibility.
172
pure
SECTION III
173
MANUSCRIPT VII
USE AND LIMITATIONS OF THE TRUNCATED
AREA UNDER THE CURVE IN BIOEQUIVALENCE TESTING
17 4
ABSTRACT
Comput er simulat i ons were used t o evaluate the truncated
area under the plasma level - time curves ( AUCt ) as indicators
of
the
bioequivalence between test and reference products.
Plasma concent rations were simulated from one and
partrnent
open
models
using
first
order
respectively
45 to 200% and 60 to 140% of the reference values. The
pharrnac okinetic parameters were selected
range
of
to
cover
a
trapezoidal
was
regimen.
calculated ,
The
using
extent
of
integrals
to
AUCinfwere
determined.
absorpti on
of the general
blood equations. The ratios of AUCt: test to
AUCt
using
rule at each time point ( t ) according to a
conventional sampling
( AUCinf )
wide
disp osit ion rate constants (0.0 1 -0 .7 9 hr . - 1 ) . The-
area under the blood level - time curves was calculated
the
corn-
abs or p tion rate
constants ( Ka ) and bioavailability ( F ) ranging
from
two
reference
and
For most simulations . the
ratios changed very little between the end of the absorpti on
period.
AUCtrnax
the
was
last
not
bioavailability
ferent
of
mathematical
bioequivalence
when using a
however.
a
time
point
good
the
parameter
time
to
infinit y .
compare
the
two drug products. Three groups of dif behavior
determinati on
single
and
AUG.
were
might
Several
identified ,
present
truncated
in
which
some problems
AUC
ratios,
could provide meaningful information on absorption
175
rates for bioequivalence testing. In ou r st udy the Auc
80
was
con sistently a good indicator of bioequivalence.
I NTRODUCTION
Since the seventies. the rate of discovery and
ment
of
develop -
new therapeutic entities has been decrea si ng while
the number of major drugs going off patent has been increasing
(1) .
As
a result there has been a steady expansion o f
the generic market (2). This phenomena was enhanced
Drug
by
Price Competition and Patent Restoration Act which ex - -
pedited the approval of generic drugs.Under
therapeutic
sumption
this
act,
being
that
once
fundamental
independent
of
the
dosage
forms .
to
metabo-
Therefore,
products showing "similar " bioavailability
according
as -
in the general circulation, the
same active drug undergoes the same disposition and
bioequivalent
the
equivalency of generic products may be assessed
on the basis of a bioequivalence test. The
lism
the
could be
drug
termed
specifications provided by the
Food and Drug Administration ( FDA ).
Bioavailability
has been defined as the rate and extent
at which a drug ingredient is absorbed from
mulation
and
becomes
the
drug
for-
available to the site of action (3).
Details on the experimental
design
and
statistical
tech -
niques for these bioavailability studies have been described
extensively elsewhere ( 4 - 5 ) and for some drugs the
176
Divisio n
of
Bioequivalence
at
the FDA provides
guide line s for the
required in-vivo bioequivalence study ( 6).
the
In
our
reporc .
staciscical issues. although int i mately related to bie -
quivalence inferences , will not be dis cussed. The rat ios
area
und er
blood
level
profiles
for
of
reference and test
products will be analy zed.
To
date ,
although no cases of bioinequivalence b e tween
approved drug products have been documented . the re are
reports
showing
concerns
some
on the therapeutic efficiency of
generic products (7- 11 ). In a recent anal ysis. however ,
FDA
found
that
in
difference of area under the plasma
the
tests
to
the
80% of 224 bioequivalence studies , the _
level -t ime
curves
reference products averaged 3.5%
For such products , the natural intersub j ect
!
5%
for
(1 2 ).
variabilit y
is
more likely to affect the pharmacodynami c response of drugs.
than differences in
absorption ). While
the
ultimate
bioavailability
AUCinf
extent
concentration)
of
provides
( extent
and
rate
of
complete information on
absorption.
Cmax
( maximum
peak
and Tmax ( time to the peak ) are dependant on
the sampling regimen . Therefore , the confidence interval for
those
parameters
is
often
wide and the regulatory agency
gives less weight to the variation of these parameters (12).
Thus ,
there is a need for alternative parameters to provide
more reliable information on the rate
Two
recent
reports
of
absorpt i on
( 15 ) .
suggested that more emphasis should be
placed o n truncated area under the blood profiles
177
( AUCt
or
AUCxx· where t is the sampling time corresponding to XX per cent of the AUCinf ) in the determination
of
( 16.23 ) .
compared
Many
bioequivalence
studies
bioequivalence
AUCtlast
(tlast =t ime of last measurable concentration ) (17 - 19), but
only
recently. few used the truncated area AUCt ( 22 ) or ad -
dressed the
pharmacokinetic
(20,2 1 , 23).
In
Japan.
relevance
statistical
of
this
testing
instead of AUCinf is required to assess
on
parameter
AUCtlast
bioequivalence
be -
tween two products (19). The issue of the last sampling time
remains a question mark. The approach required by the FDA is
that
AUC's
should
be
calculated from plasma levels which _
have fallen to at least 10% of the peak concentration.
method
is
useful
simple rule of
but
thumb
empirical.
governing
This
In essence. there is no
the
principles
of
bioe -
qui valence testing.
The objectives of this computer simulation are to inves tigate
the
use
and
limitations
bioequivalence studies and
relevance
of
this
to
of
stress
parameter.
We
truncated
the
areas
in
pharmacokinetic
hope to define rational
limitations and boundarie s in the practical use of incremental
AUC 's.
Our
goals
are
to
demonstrate that ratios of
truncated AUC's can be used with good reliability
quivalence
testing
appropriateness
and
in
bioe-
as they conform to both the statistical
the
pharmacokinetic / pharmacodynamic
relevance
178
METHODS
Plasma
levels
ment open model
were simulated from one and two compart -
equati o n s
using
first
or der
elimination
( KE ) rangin g from 0 . 04 to 0.2 hr - 1 and sets of al -
constants
pha ( a ). beta(~)
hr - 1 respectivel y .
macokinetic
fr om
0. 17
to
0.79
and
0.01
to
0. 17
For each KE and each set of a ,~ the phar-
parameters
characterizing
the
dosage
forms.
bioavailability ( F ) and rate of absorption ( Ka ), were ranged
from 60% to 140% and 45% to 200% of
values .
All
the
reference
product
conditions are reported in table I and II. All
necessary nomenclature is given in the appendix. It is
well _
known that AUCinf remains constant with varying rates of ab sorption and we were interested in observing the behavior of
truncated
areas
in
the various combinations . The one com-
partment open model ( figure
compartment
open
model
1,
( 24 )
equation
with
1)
and
elimination
central compartment , ( figure 2. equation
2)
were
the
two
from
the
used
these simulations.
Ct
( F.D.Ka / V) . ( e - KEt _ e - Kat )
Ct
-v~c fa = a ;~c~= a ;-
eq.l
Ka.F.D. ( K21 - a )
[ e - atl
+
Ka.F.D.(K21- ~)
[ e- ~ tl
+
-v~cfa=~;~ca=~;-
179
in
eq . 2
The disp o siti o n parameters for a drug. were
be
assumed
to
independent o f the d o sage f o rm and for each KE or set of
C a .~) .
and
the pharmacokinetic parameters of the dosage fo rms. F
Ka
were
varied.
The
reference
products that we ar -
bitrarily designated for comparison purposes had the same F.
Ka over the KE or
Simulation
TurboPascal
a. ~
ranges.
algorithms
were
coded
in
Pascal
and compiled on IBM Personal Computers. Mos t o f
the equations we used were derived from mono , bi and
ponential
and
traditional
time.
depended
relationships
Calculations
on
the
of
calculations
AUC
triex -
between concentration
ratios
(AUCtest AUCref )
time available for absorption T ( 20. 25 ).
where T was the end of the
and
and
for
one
absorption
and
period.
Derivat ions
two model compartments are
given in Appendix 2.
T.
the
time for 99% of the absorption to occur. can be
estimated from the first order
process
using
the
wagner -
Nelson method (22) where the fraction unabsorbed is obtained
with :
T
4.606 / Ka
eq . 3
T ranged from 1 . 6 to 5.8 hours. It
that
AUC
should
be
calculated
180
has
been
suggested
at least through 3.5 half
lives f o r accuracy ( 24 ) . However. it appeared that
AUCt
at
t =2T is usually within a few percent of the AUCinf ( 17 , 24 ) .
Fo r the two compartment study. the procedure was similar
except
the
classic
triexponential
function
( equation 2 )
describing an open two compartment model. If we maintain our
assumption that v. K21. D, a and
~
remain constant and inde -
pendent of time. we can express the extent of absorption as
At =
f
eq.4
Ctdt
Using the ranges of pharmacokinetic parameters reported
reference
and
values from table III. area under the curves were_
calculated using the trapezoidal rule (AUCt )
and
equations
previously defined (AUCinf) : the area at t = infinity was
A= F.D / V. KE
eq . 5
for one compartme nt a nd
A = AT + [ CT / a +
CT I ~
eq.6
for two compartme nts
AUC rat i os ( test / ref) were than evaluated.
such
Time
points
as t= T, 3 / 2T and 2T. 3T were given special attention.
This study is not exhaustive in any means because we had
use
a
to
limited set of parameters. Nevertheless , this report
presents an
approach in the potential use of truncated area
in bioequivalence determination.
181
RESULTS AND DISCUSSION
Area
under
the
curve
ratios
have
been discussed by
Wagner ( 24 ) and Loo and Riegelman (2 5 ) who proposed
mathematical
relations
pendent of the model. In complete availability can
the
absorption
pr ocess
cannot be calculated
levels
ing
several
for percent of drug absorbed. inde influence
and true absorption rate constants
with
complete
accuracy
from
plasma
( 27 ). Furthermore. when the absorption is the limit -
factor.
flip-flop
AUCinfcannot
be
stances . when
phenomena
correctly
both
zer o
may
estimated
and
first
occur
(28) .
order
( Ka <KE )
In
and
other in--
kinetic
model
describe the absorption process ( 29 ) . the rate of absorption
is very difficult
qui valence
to
testing,
estimate
however.
with
the
accuracy.
extent
and
absorption , as measured from blood concentrations ,
paring
using
AUCs ,
Tmax
and
rate
constants
and
bioe -
rate
by
of
com-
Cmax do not fall within the
previous limitations. In this study, the Ka ' s
absorption
For
represent
are
also
apparent
immediate
release characteristics of the dosage forms being tested . We
fully
understand
the limitations of the simulations since
the pharmacokinetic models apply
formulation
to
situations
where
the
behavior is the limiting factor of drug absorp -
tion .
182
In
man y bioe qu i v a lence stud ies. an anal ys i s of va r ia nce
is conducted on pharmacoki ne tic p ar ame ters
such
as
Ct
or
AUCt a nd t he class i c al hy p o these s a re te s ted:
H : Te st = Reference o r ~ = O
0
Ha . Te st # Referen c e o r ~ # 0 . where
~
is a measure of
the difference in bi oa v ailabilit y . Th i s t e st i s n ot
cons i s -
t ent wi th the definiti on o f bi oequi v alenc e b e c a use the p owe r
o f the test is aimed at testing Ha not the
H
0
null
hypothes i s
The current acceptance criteria f o r bi oequ iv alenc e i s
that the true AUG or AUCinf of the t est
f o rmulation
s h ould
Therefore the AUG ratios are c ompatible with the
nature
be within 20% of the reference mean :
AUCtest - AUCref ~
10. 20 1 ( AUCref )
which can be written as:
I 0. 2 I
of bioequivalence testing .
In a previous investigation ( 16 ) three
quivalence
testing
were simulated. In all
cases
of
bioe -
simulations the
truncated area under the curve ( AUCt ), accounting for 6 0% o f
183
the
AUCinf '
resulted in making the tests more sensitive to
bioavailability differences. It was
the
ratios
of
also
anti cipated
that
AUCtmax co uld be of critical importance fo r
the determination of bioequivalence as it i s intimately
lated
to
the
rate
of absorption . Unfortunately. we fou nd
that f or high bioavailability ( F ) . and low rate
tion
( Ka )
to
75%
of
the
reference
of
absorp -
value ).
the
AUCtmax still falls within the acceptable range 80- 120%
In
other
( 50
re -
cases. the AUCtmax ratio did not indicate equivalence
for bioequivalent products . The AUCtmax ratios are not reli able
indicators
of
the
indicate equivalence
bioavailability
the reference
absorption rate as they failed to_
in
were
instances
when
Ka
and
relative
actually within the acceptable 20% of
value
( Table
IV ).
In
the
bioavailability
profiles illustrated in figure 3. drug product B having a Ka
of 2.0
( 143%
of
bioavailability
0.8
( 50%
of
the
of
the
bioavailability
of
reference
70%,
value )
with
a
relative
and drug product C having a Ka of
value )
reference
130%,
were
with
a
relative
compared to the reference
formulation A .
All
truncated
AUC
ratios identified the inequivalence
except when approaching tmax for which the AUC ratios
cated
false
bioequivalence
indi -
as marked by the bold vertical
line . Furthermore , the calculation of this parameter is very
sensitive
to
the
sampling regimen and we do not recommend
its use in bioequivalence comparison .
184
In
(
this
proj e ct.
mo re
than 700 s imulat ions were per -
formed and analyzed. although only significant
presented
in
this
report.
l ) the
accep tabl e
are
The discussion is divided into
three scenarios whi ch we found
t est ing:
figures
critical
in
bioequivalenc e
re lat i ve bioavailability (F) is outside the
range
of
80
to
120%
2)
relative
the
bioavailability is within th e allowed int erval with Ka rang ing from 50 t o 200% of the reference v alue . and
boundaries o f bioequivalence F=80% and F =l20%.
3)
at
the
In each case
the effects of unacceptable performances for the rate of absorption were c arefull y monit o red.
Case 1: when Ka and F of the test product were both out side the range o f 80 to 120% of the reference values. 94% of
the time the truncated AUG detected the bioinequivalence
at
all time points (tables V and VI ) . In particular situati ons ,
when the bioavailability is above 120%
rate
and
the
absorption
i s below 80%. the effect of the truncated estimate for
the extent of absorption is
parameters
counterbalanced
those
in
the
more
absorption
relative
4,
the
bioavailability ( F ) compensates for the slow
rate of bioavailability ( Ka =5 7% )
false
is
truncated AUG ( 30 ) and larger portions of
AUCinf are sens itive to absorption rates. In figure
high
two
and the ratio of truncated areas indicated false
bioequivalence. As KE or a. C increase ,
included
by
bioequivalence.
This
and
AUC
ratios
indicate
situation shows that a product
185
wit h un acc e pt a b l y hi gh ex t ent of absorption and u n acce ptably
lo w rate of ab so rpti o n c an be f ound bioequi valent usi n g c er ta in truncated e s timates .
Cas e
2 : when the relative bioavailabili ty is within a c -
c eptable ranges and the Ka varies from 50
reference
product ,
74%
bioequivalence for only
to
200%
of
t he
o f the AUCt rati os ( RAT .XX) s h owed
30%
of
truly
bioequivalen t
c as e
products. indi c ating that truncated areas are mo re s en s it i ve
to the extent than the rate of absorpti on as shown i n
v.
Nevertheless .
rate
of
table
the sensitivity o f the AUCt rati os to t he
abs o rption
is
proportional
to
the
disp osit ion-
phases. It is known that changes in elimination affect phar macokinetic parameters of one
( 31 )
and
this
and
two
compartment
model s
study confirmed that as a consequence . suc h
changes also influence bioequivalence determination .
Case
AUC
20
3:
is
at
the
boundaries ,
when F =80% . the truncated
only discriminating when absorption rates are low .
When F =l20% and Ka is more than 20% of the acceptable limit.
AUc
20
cases ,
is sensitive to the high rate of absorption. In
if
you
increase
the
disposition
both
rate constants ,
larger portions of the the AUCinf become discriminating
show inequivalence. The percentage of the AUCinf
indicative
of true bioinequivalence due to the absorption rate .
with
the
elimination
and
varies
phases. Figure 5 illustrates two ex -
amples of minimum AUC ratios required
186
to
demonstrate
true
bioinequivalence. for formulations exhibiting l ow ab sorpcion
rate.
Particular
cases
where
the
AUC ratios showed bioine -
quivalence for products with extent and rates of
abs orptio n
within the acceptable range were critical. For all cases except a slow elimination rate constants of 0.01
F=l20%
and
the
absorption
rate
constants
within the acceptable 100 to 120% of the
the
truncated
AUC ' s
hrs
- 1
when
were high buc
reference
values.
for 20 to 50% of the AUCinf indicated
consistently fal se inequivalence ( Table VII). In our
the
.
study.
truncated racio of AUC accounting for 80% of the AUCinf-
did indicate bioequivalence
consistently
throughout
these
disposition situations.
Another limitation
when
relative
bioavialability
(F)
ranged from F=lOO to 120% in which AUC and Ka ranged from 67
to 196.4% the truncated area ratios did not indicate bioinequivalence
at
any time (t) o r for any proportional section
of the AUCinf. The same limitations
AUCinf
apply
to
the
use
of
in the bioequivalence comparison since AUCinf is in -
dependent of the absorption rate. Within this interval ,
truncated
AUC
was
completely
dependent
the
on the extent of
bioavailability.
In
all
cases ,
when
F was 80 to 120% of the reference
value , 98% of the time AUC
confirming
what
had
was also within the same range
80
been previously reported (21) on the
similarity of truncated and infinity AUC
187
ratios.
Truncated
ratios
for
early
p or tions of the AUCin f were smaller than
the relative bioavailabi lit y ( fi gure 6 ) when the
rate
constant
was
absorption
below the reference valu e 1 .4 hrs- 1 for
the test products. Conversel y. when Ka was greater than
reference
value,
earl y
truncated
the
area ratios were lar ger
than the relative bioavailability ( figure
7 ).
The
cut-off
point or time where the truncated AUG rati os became equal t o
the relati ve extent of abs orpt i on was related to the absorp tion rate const ants and exhibited a u-shaped curve in figure
8 , which indicated that these critical times or
of
AUCinf
were
high
for
low
absorpt ion
percentages
rate products.-
decreased as Ka increased and increased again when the
of
absorption
was
above
rate
100% of the re f erence value. The
slope of AUG ratio Vs. time, in early porti ons of the plasma
profile .
was proportional to rate and extent of absorption.
This parameter could be an additional
absorption
vestigations are needed on this topic.
cut-off
compare
In
our
study,
the
point ranged from 35 to 85% for acceptable formula -
tions ( table VIII). For
this
est imator t o
kinetics in bioequivalence study and further in -
perfectly
bioequivalent
products ,
point could theoretically be 0. The time corresponding
to these percentages depends on the elimination characteristics ,
(figure
9 ).
Furthermore , if the MEG or MIC [Minimum
Efficient Concentration or Minimum Inhibi tory Concentration ]
can
be
estimated
accurately ,
comparison of truncated AUC 's
it has been argued that the
for
188
bioequivalence
purposes
has
also
a pharmacodynamic relevance ( 31 ) . s ince the AUCxx
ratios compare portions of the plasma levels signif i cant for
the therapeutic window (figure 10 ) .
CONCLUSIONS
A
first
criticism
of
the emphasis given t o AUCinf in
bioequivalence t e st ing was the fact t hat
aimed
at
comparing
the
bioequivalence
is
absorption characteristics of two
dosage f orms. It is well kn own that AUCinf is not
dependent
on the rate of absorption Ka , thus on l y extent of absorption_
could be c ompared with accuracy .
We
feel
that extrapolation of the AUCtlast to infinity
could bring unnecessary
realistic
differences
variations
or
large ,
the
representative
of
equivalences between a reference
and a test produ c t. On one hand,
is
not
if the
extrapolated
area
true difference in extent of absorption may
become proportionally insignificant. On the other hand,
the
extrapolation is calculated from the last p oints , usually in
the most variable analytical region and this might introduce
differences
between
truly
bieoquiv alent products. In most
cases the AUCinf ratios and Auc
80
ratios for our simulations
did not vary significantly indicating that 80% of the AUCinf
was a good
AUCtmax
was
indicator
not
of
the
consistently
extent
of
bioavailability .
representative of AUCinf or
rate o f absorption . We do not rec ommend the use
189
of
AUCtmax
alone
to
compare bioavalailability because extent and rate
of absorption outside acceptable ranges can
in
some
cases
balance each other.
The proportion of the AUCinf required for sensit i vity in
bioequi valence
testing
depends
on
the
combination
of
e limination rate constants and the absorption rate co n stant .
In
all
cases.
the AUC ratios at earl y times overestimated
relative bioavailability for low absorption
products ,
and
underestimated
before reaching the
true
for
relative
high
rates
of
test
absorption
rate s
bioavailability
val ue.
Important failures of the truncated areas occurred where the.
relative bioavailability was high and the rate o f abs orpti on
was
l ower
than
the reference value as well a s cases where
acceptable l ow F and l ow Ka led t o false inequivalence as
result
of
a
over 0 . 0 1 hr.
counterbalancing effect . As
-1
a
or KE increased
. larger portions of the AUCinf
were
neces-
sary to assess bioinequivalence . These simulations suggested
that 80% of the AUCinf was
range
of
our
substantially
rate
ditional
the
bioequivalence
than
of absorption. Although truncated AUC ratios have
to be further
rates ),
in
study for determining the relative extent of
absorption which had more influence on
the
reliable
investigated
( as
indicators
of
absorption
we feel that examining truncated areas provides ad and
qualitative
information
bioequivalence testing.
190
that
strengthen
APPENDIX 1: Nomenclature
D
Dose
v
Vo lume of di str ibuti on
Ka
hr.
First
order
absorption
constant
in
- 1
KE
in hr.
First
order
elimination rate constant
-1
Distribution rate constant
Elimination rate constant
At
Area under the curve at time t
Ct
Blood concentration at time t
AT
Area
under the curve at the end of the
absorption
period
CT
Plasma
AUCi.
Area Under the Curve
AUCref
concentration
at
Area under the Curve for the reference product
AUCtest' Area under the curve for the test product
Ka%:
Ratio of Ka test to reference
in
per -
cent age
F% :
Ratio
of
bioavailability
reference in
RATXX
to
percentage
Ratio of
truncated
responding to XX%
i- inf
test
AUC ' s
at
t
cor -
of AUCinf reference
at time infinity
i - tmax : at time of the peak
191
i -
at time t
t
APPENDIX 2 : Calculations
For the one compartment model if t
~
T. blood levels are
given by eq.l and the extent of absorption can be integrated
and written as ( 21. 22 ) :
eq.
At
where
At
is the area under the blood level-time curves •
from 0 to time t. The following ratios can be
derived
(21)
where Ka' and F' indicate the pharmacokinetic parameters for
the reference product:
At
F
eq. 2
If t
, T the plasma levels can be described (17) by :
Ct
CT . e -KE(t-T)
where
eq . 3
CT is the blood concentration at time T . thus the
AUC at time t can be written (2 1 ) as :
192
AT - Ct. ( l - e-KE( t - T )) .l / KE
At;
eq . 4
with AT. the are a u nde r the curve from time O to T.
The
rat ios can th e r ef o re b e calcul ated ( 2 1 ) :
At
e q. 5
AT '
At "
+
( 1- e - KE ( t - T •) ) . Ct / KE
Fo r the two compartment mode l if t
T,
theref o re . At can be derived
At
+
a
1
+
-~-
F . D. ( K21 - Ka )
eq .6
- v ~r a = Ka )~\~= Ka )-
The area under the curve at t infinity
from
equation
9
and calculated using the
parameters independent of time :
A
a
193
+
can
be
derived
pharmac okineti c
1
-~ -
+
eq .7
The
ratio
At / At *
can also be estimated easil y . On the
other hand. if t >T , the blood levels can be expressed as:
Ct
~
CT
eq.8
therefore , At the area under t he curve at time t can be in tegrated and written as:
At
AT + (1-e - a ( t - T ) ].CT / a + (1 -e-~(t - T ) ].CT / ~
eq.9
The area at t - infinity is derived in equation 10
A
AT + [ CT / a +
CT I ~
]
eq. 10
and the ratios can be calculated from At / At • :
AT+ (1-e- a (t - T)].CT / a + (1-e - ~(t-T)) .CT I ~
At
eq . 11
At
0
AT "+ (1-e-a(t - T ) ) . CT 1a + (1-e-~(t-T)) .CT · / ~
0
194
with T
case.
4.606 Ka
and T
4 .606 Ka.
In
this
parcicular
AT and AT· are calculated from equacion 9 and CT. CT·
from
equati on 2 for
c~T.
REFERENCES
l.
Generic Drug Scandal Pharm . Te ch . (12). 10. 23-27. 1988
2.
Peter P. Lamy J . Clin . Pharmmacol ( 26 ) . 309 - 316. 1986
3.
"New Prescription Drugs-Pr op osed Bi oavailability
Requirements " FDA Washingt on, CFR Part 130. 1973
4.
Carl M. Metzler J.Clin.Pharmaco ( 29 ) . 289-292. 1989
5.
Robert
w.
Mackuch. Mary F. Johnson Arch.Intern . Med
( 146 ), 986 - 989 , 1986
6.
Di vision of Bioequivalence . "Guidance for In Vivo
Bioequivalence and In Vitro Dissolution Studies f or
Indomethacin " FDA 21 CFR 10.90 ( 1988 )
7.
8.
Stephen Levine Hippocrates March . April 1989
Elaine Wyllie , C. E. Pippenger , A. David Rother
s!AHA ( 258 ) , 9 , 1216-1217 , 1987
9.
Rudl Ansbache r slll.H.fl. ( 259 ), 2, 220. 1988
10.
Robert Schulma n ..rAM1I (261 ) , 17. 2499 , 1989
11 .
Sheldon S . Stoffer. Walter E. Szpuner <LAMA ( 244) , 15 .
1704 - 1705 , 1980
12 .
Stuart L . Nightingale. James C. Morrison .rAMA (258). 9.
1200-1204 , 1987
195
13 .
Joh n D. Hay nes J . Pha r m.Sci ( 70 ) . 6. 673 - 675. 1981
14 .
Stephen C. Ol so n. Michael A. Eldo n . Roge r D. Tooth aker .
J ames J. Ferr y a nd Wayne A. Co l bu rn J . Cli n . Ph ar macol .
( 27 ) . 342 - 34 5. 198 7
15.
David J . Greenblatt . Ri c hard I. Shader J .Clin . Ph armacol.
( 27 ) , 8 5-86. 1987
16.
S.E. Rosenbaum. C. Bon. C.T. Rh odes Drug Dev . Ind . Pharm
( 16 ) . 1. 157 - 163 . 1990
17.
H.P. Ringhand , W.A. Ritschel , M. C . Meyer. A. B. Straughn
and B.E. Cabana J.Pharm.Sci. (72 ). 12. 198 3
18.
W. A. Mah on. J . S . Leeder, M.M. Bril l- Edwards . J . Correia
and S.M . MacLeod Clin . Pharmacokinetics ( 13 ) . 118 - 124.
1987
19.
Mark L. Powell, Miryam Weisberger , Richard Gural. Menger
Chung , James A. Patrick . Elaine Radwanski and Samson S.
Symchowicz J.Pharm.Sci . ( 74 ) . 2 . 217 - 219, 1985
20.
E.G. Lovering. I.J. McGilveray. I.McMillan ,
W. Tostowaryk. T. Matula and G. Marier Can . J.Pharm . Sci.
( 10 ). 2 . 36 - 39 , 1975
21.
E.G. Lovering, I.J. McGilveray , I.McMillan and
w.
Tostowaryk J.Pharm . Sci. ( 64 ) . 9. 1521-1524. 1975
22.
K.K . Midha. B.S. Chakraborty , R . Schewde , E.M . Hawes. G.
McKay, J. W. Hubbard and M. Moore
J . Pharm . Sci . ( 79 ) . 1, 3-8 , 1990
196
23.
Nahoko Kaniwa. Hiroyasu Ogata. Nobuo Aoyagi. Yasushi
Takeda and Mitsuru Uchimaya J.Pharm.Sci.
(78). 12. 1020-
1024 . 1989
24.
John G. Wagner J.Pharm . Sci . ( 72 ) . 7. 838-884. 1983
25.
John G. Wagner and Eino Nelson
J.Pharm.Sci.
26.
(52), 6. 610-611.
1963
John G. Wagner "Pharmacokinetics: Notes· pp 132 - 138.
16 1-166 Published by J.M. Richards Laboratory 1969
27.
Donald Perrier and Mil o Gibaldi
J.Pharm.Sci.
28.
( 62 ) . 2. 225-228. 1973
Peter R. Byron and Robert E. Notari J.Pha rm. Sci. (65l.
8, 1140-1144. 1976
29 .
Lilian E. Riad. Keith K.Chan. William E. Wagner and
30.
William J. Jusko and Milo Gibaldi J.Pharm.Sci. ( 61). 8.
Ronald J. Sawshuk J .Pharm.Sci.
1270 -1273.
31.
( 75 ). 9. 897-900. 1986
1972
Jack N. Moss and C.T. Rhodes Can.J.Pharm . Sci. (9). 1.
30-32. 1974
197
Table I
: One compartment model simul a tions
Parameters
Variable Parameters
Elimination rate constant KE
l/hr
0.04,
Absorption rate constant Ka
l/hr
0. 8'
1. 4'
Absorption Period T
hrs.
,_,
tO
():>
Relative bioavailability
5 . 7 4'
J. JO'
0.08,
0.12,
0. 1 6 ,
0.2
o. 9 ' 1. 0' 1.1, 1. 2' 1.)'
1. 5 , 1. 6 ' 1. 7 ' 1. 8' 1. 9
5.12 ,
). 06,
4.60,
2.88,
0. 7' 0.8, 0.9,
4' 18' ) . 84'
2.70, 2.56,
1. 0,
1.1, 1. 2'
) .4
2. 4
1.)
Constant Parameters
Volume of distribution
Vd
Dose D
Sampling Regimen in hours:
5.
o,
1
10
mg
1000
0. )) ' 0.67, 1. 0' 1. 5' 2.0, 2 . 5 ' ) . 0' 4 . 0
6.0, 8. 6' 10, 12, 16, 48, 72, 96 , 1 20
Table II : Two Compartment Model Simulations
Parameters
Variable Parameters
Disposition Rate Constants
p
l/hr
l/hr
0. 17, 0.32, .0.4 7, 0.63, 0.79
0 .0 1, 0.09, 0. 17
Absorption Rate Constant
Ka
l/hr
0.65, 0 . 95, 1. 4.
Absorption Period
T
hrs
Relative Bioavailability
F
>-'
QI
2. 0. 2.7 5
7.08, 4.84, 2 . 30, 2.30,
0.6, 0. 8. 1. o, 1. 2.
1. 68
1. 4
Constant Parameters
<D
<D
Volume distribution V
1
10
D
mg
1000
Dose
Sampling Regimen:
1. 0, 1. 5, 2. 0. 2 . 5. 3. 0. 4. o, 5.0, 6. 0, 7.0, 8.0
10, 12, 15, 24, 48, 72, 96, 120, 144, 168, 192,
216 (hours)
Table III : Pharmacok inetic Parameters of the Refer e nce Produ c ts
One compartment model
First Order Absorpt i on Rate
Bioavailability,
Ka :
F :
1. 4
1. 0
l / hr.
Ka :
K21:
F :
1. 4
l/h r.
l/hr .
Two compartment model
ti)
0
0
First Order Absorption Rate
Disposition
Bioava ilabi 1 i ty
0. 2
1. 0
~
Table IV: Pharmacokinetic
Relevance of AUCtmax
Ka%
F%
AUCtmax
Ratio
AUC48
Ratio
AUC120
ratio
One Compartment Model
ti)
0
94.0
87.5
0.8
0.8
0.78
0.76
0.80
0.80
0.80
0.80
50.0
75.0
1. 3
1. 3
0.94
1.16
1. 30
1. 30
1. 30
1. 30
Two Compartment Model
142.8
50.0
78.5
85.7
0.7
1. 3
1. 3
0.8
0.81
0.87
1. 15
0.74
-
-
0.69
1. 30
1. 30
0.80
Table V : One Compartment model, AUC ratios for bioinequivalence
KE
0.04
0.20
ro
0
ro
F%
KA%
RAT.20
RAT.30
RAT.40
RAT . 50
RAT. 60
RAT. 80
70
70
130
130
57
143
57
143
0.63
0.73
1.17
1. 35
0.66
0. 71
1. 23
1. 32
0.67
0. 71
1. 25
1. 32
0.68
0.71
1. 29
1. 31
0.68
0.70
1. 39
1. 31
0.68
0.70
1. 30
1. 30
70
70
130
130
57
143
57
143
0.50
0.81
0.94
1. 51
0.56
0. 77
1. 03
1. 43
0.58
0.75
1. 07
1. 40
0.61
0.74
1.13
1. 37
0.63
0.72
1.17
1. 34
0.68
0.71
1. 27
1. 31
RAT.XX
AUC ratios (test to reference) at XX% of the AUC infinity
Table VI : Bioinequivalent Rroducts-AUC Ratios
Twocompartment Model
F%
KA%
ti)
0
60
46
140
196
46
196
RAT20
0.585
0.600
1. 35
1. 40
RAT30
0.590
0.600
1. 37
1. 40
RAT40
0.593
0.600
1. 38
1. 40
RAT50
0.596
0.600
1. 39
1. 40
RAT60
0.597
0.598
1. 39
1. 40
RAT70
0.598
0. 5 99
1. 40
1. 40
RAT80
0.599
0 . 600
1. 40
1. 40
VJ
Table VII : False Inequivalence
(for F = 120 % and Ka
"'
0
"'
=
114%)
RAT20
RAT40
RAT60
RA TSO
Ke
0 . 04
0.12
0. 20
1. 22
1. 24
1. 27
1. 22
1. 24
1. 25
1. 2 1
1. 22
1. 2 4
1. 20
1. 20
1. 21
0
N
'0.01
0 . 03
0.09
0.17
1. 20
1. 27
1. 28
1. 30
1. 20
1. 1 9
1. 19
1.19
1. 19
1.1 9
T
--
1. 21
1. 27
-1. 1 9
1. 2 2
MODEL
E
w
0
Table VIII: Truncated AUC Ratios equating Relative Bioavailability
Time and Corresponding \ of AUCinf
One Compartment
KE (l/hr)
Ka
(l/hr)
0.04
0. 1 2
0.20
1. 2
1. 7
1. 2
1. 7
1. 2
1. 7
16
46
12
36
24
61
24
61
24
61
24
61
14
74
14
74
15
84
15
84
15
84
15
84
13
13
84
84
14
89
84
0.95
2.00
0. 95
2.00
0.95
2. 00
24
32
10
13
12
70
7
49
7
82
6
75
36
39
12
16
15
78
8
54
15
87
9
65
48
52
24
32
15
78
9
59
F=0.7
Time (hr.)
\ AUCinf
F=l.O
Time (hr.)
\ AUCinf
ti)
0
(Jl
F=l. 3
Time (hr.)
\ AUCinf
Two Compartment
(l/hr)
'
Ka (l/hr)
13
14
89
0.17
0.09
0.01
14
89
F=0.6
Time (hr.)
\ AUCi nf
F=l.O
Time (hr.)
\ AUCinf
F=l. 4
Time (hr.)
\ AUCinf
9
90
5
65
206
Figwe 2 : Two Compartment Model
l\)
0
-'I
K12
Ka•
..
Centra l
Compartment
Peripherical
Compartment
..._
...K21
KE
~~
.
..
•
I
I
. .... 1.
I
I
I
.. '
)(
•E
()
()
;
::::>
-•
c
0
.... . .. .. .. ..
I
I
.. .... . .•I ..
I
I
A.
I
'
·•
'
'
•.,
chI
I
.,
0
()
.,• •> •
J,• ..• ;'
... a:
...
.2 .ci'
•
c
..
c
..•
:;:
A.
''
0
()
•.E
••
<lj
.I:
A.
•
II:
f
0
0
0
•
0
•
.
0
0
...
0
111111•-1 NOl.LYll.LNJONOO
208
0
~
::::>
0
e
w
:II
i=
(\J
'
''
/'
''
u
*'i5
l
ii::
'
I
'
I
.II
Ji
qi
I
~
.•
0
'
''
0..
<D
I
I
I
<I
··········· !fl
aJ
!0
I
I
*'i5
0..
I
I
I
'
diI
<i
ie
....I~
•
.!J
:
I
I
(j)
I
I-
<t
;.
IC
<l:
(\J
Qi
a:
T
8
0
<D
0
0
(j)
<t
0
(\J
(1111/l:IUI) NOIJ. YIU.NJONOO
209
=
,,"i
0
~
'
'
,''
''
!
~
J
I
,,
0
..
.
..
(\J
0
'ii
I
I
i ,
i :
i :
ii rf
i ,
i :
i . i '
I '
i : : ,'
I
(\J
0
•
<ll
1 + ~~
i / , ,'
I f : ,'
:
I!': ,,
•.
0
0
~'
<t
0
0
:I
a
iL
0
(\J
t
0
'<I:
0
0
q
"!
101i•Y
onv
210
8
oo
<ll
0
0
N
0
0
,,
"'
:I
I
"i
!
~
0
0
'Cl
<I>
I-
~
..
• CJ :
.Ci ~
•
0
(')
'
<i
'
=i
a
ii:
0
N
t
Q
II
I
I
i
i
i
i
i
i
i
i
i
i
I
I
:
I
I
.
'
I
I
I
I
I
I
I
1 ..
i
"'0
'
I
I
I
I
"'
;
I
;
;
I :
'tl
I
I
I'
<t
(//
0
0
///,'
//,:·
"h
0
C'!
•on•Y :>nv
211
8
oo
a:i
0
~
~
0
-
N
ci
<O
II
'#. :,(
-
<{
lol
._
<O
.t::
Cot
:J -I
er
• II.
ci
c•
~
•c
0
N
u
ci
J
~
I
q
Cl)
0
...•
<t
0
ci
:J 0
~I:.
II. •
0
0
0
If)
1uionv 10 •1181u•~ac1
212
0
Ill
0
..."!
c,g
-•..
...
I
...":
...
I
l{)
...0
...
I
D.
0
0
<t
.tea
c
•
0
0
co ~
(")
.
j
•~ •E
- ....
Cl·II.
•
...
......
•
0
N
'•
>
¥ J!
•
! ..
0
:;:::
e
•
I
a:: •
a:
<>
:)
0
lili:
c
'
0
0
<t
0
<t:
0
213
!
<t
...
~
c0
:;
...a.0
...
I
...•
...
I
<Xl
• c:
•
•'"E•
I""•
0 0
Ill:~ co
•
~
• • C!
c.>
...
N
~ I •
••111:
9111:
•
II-
a:
>
.,.
=
•
0
·C'4
C: .,. I
:II·-
.~
I
. ..,
....
<(
I&.
I
'•
:i .i
~··
-
•
.c:a
.....
•c:i
...
...
...
N
...c:i
•
0
:;
<O
•
a:
(J
::::>
<(
0
0
<q
0
0
"':
IOllU
:>nv
214
0
l()
0 ..
c:
()
~
O
iI
\ ,,'\
0 ..
..
.!.
......
!!:
0
"1
CJ ....
! l
i
Cl
0
I
&.!!
~ii
c: ... o.
~
...
(,)
;::::)
:.
~
CIO
c:
e
.c:
•... •
:I
a
ii:
...
I
<J
..
''
0
•/I!/
o ..
..
•
...
...
I
I
0
I'-
!J
I
0
8
0
l()
215
o'°
0
(J\
>c:
.,,• -
,,'i
_ w
:..::
0
0
N
:I
-•
.••t
•
0
ci
c:
E
-•.
·u
.. u
.Q
0 II.I
E
Cl N
0
()
ie
·;
0
•c:
I-
Q.
c:
:::> c: '"-:
•0 ,,
~
0
..•. T
0
-•
••• .§•
.-
<t
ci
ii:
1
.I•
0
(')
0
:I
Cl
0
'°
0
8
I[)
(~)
IUIOflV •Ill JO IUOAJOcl
216
.. ,,
c
..
C>
0
•E
•""'
c:
..
-= -..
• <n
c:
()
~
al 0
a. :I
E
0
0
r
0
ClJ
i:
--~
i•
0
ci
~
Ill
0
'I"
.Q
O>
• °'ci
a
~
:I
~
0
u
c.
0
t-
~
ci
:;
.! 'ii
0
0
ii:
T
8
0
I[)
<~> 1111onv •111 10 lllOflJOcl
217
...
..1.
...
•
0
<t
''"' ""''' "'" "'"''''"'"""""''""""'••
.
..........
,,,,, ,,,,,,:
•
0
C'J
..
,,,,,,,,,,,,,, , ,,,,,,/~ ··
.f'
0
•
.
N
!
'
''
···········:·. ······· ······~...~ .. ··· ~:~······· w ······ "''"''""
.
.
.~·
.:"·
0
0
0
N
0
<O
(IW/15n) .:JN<)'.)
218
0
I
SECTION IV
219
APPENDIX A
In early stages of the wet granulation experiment. a
power
meter
or
wattmeter , was connected between the power
and our planetary mixer in order to validate
determination.
A
plotter
the variation in the
within
the
power
the
end - point
was wired to the meter to record
needed
to
rotate
the
paddle
granulation. Thus if the courant can be assumed
to be constant. one has a fair approximation of
the
torque
or resistance to movement of the mass being granulated.
The apparatus was improved
and
the
wattmeter
was
connected with a interface and an IBM PC. A data acquisition
software averaged and reduced the noise
recording
thus
allowing
the
of clear variation wattage. Figure 1 is a typical
power consumption t race for ibuprofen
formulations
without
the computerized interface.
Figure 2 is a calibration curve used in the dissolution
experiment.
Ibuprofen
concentration were measured by
ultra violet spectrophotometry at 264 nm.
220
..ct== ..,
0
....
....
~u
.-
~=
e::
:I •
g ~
u ;s
.. II.
•
c
1~
ii
,.. ..
A
.... 0
....
•
t
221
0
0
•
I
-·
••...
...c
Ill
>
.. !
... (J
·1
I~
c
(J
I
I
--
...
:i
a
0
0
... :::•
~
...•g
(J
0
0
...
~
0
0
0
•0
0
•H9Cl'O•CIY
222
.!
g
APPENDIX B
X- RAY CRYSTALLOGRAPHY AND MOLECULAR MODELING
223
Single X- ray diffraction data o f rac - ibupr o fen and
(-)- S- ibuprofen were used with the "macromodel " software
examine
the
3D
molecular
packing
to
in the crystal lattice
( figures 1 to 5 ) .
From single crystal coordinates it was a simple matter to generate powder X- ray
possible
to
predict
and
diffractograms.
Thus
it
was
validate the single powder X-ray
crystallography (figure 6 ) . however.
the
intensities
were
dependant on how randomly the crystal is oriented in the
x-ray beam. Therefore. only the position of the peaks
be predicted not the optical purity.
224
could
l'lour• 1: llolecul•r P•cklno in th•
cryat•I l.8ttlc• of Ru-ibuprofen
225
,...,. 2: llepr•-t•tloll et tH l+l-1M1prefe11
orretal lattioe ,,_ • 41fferut .....
226
l'lfUt• S: Arroyo of <•l-1 <-1-11 11...,010•
llolecllloo Ill Ill• oryotal of rac-lt...,ofoa
(Doo• - · oro hydr- bolMlol
227
l'igur• ' ' Detella of Ille ~--• h1•01.., l>oi. llMI <-1-1-. . .01.. ..,.... l•ttlo•
228
...... I: kle•t"IO•tlo• of tll• lllOloCuelr tor•lo•
l11porpHHIOft of • molectllH A 811411 I
IHOIH411 In tlle HIM hJdrogen bolHI
229
........ I: TheoreUc81 X-rey Dilfrectogre,.e
X-RAY POIJOER DIAGRAMS
Ibuprofen !optic ally active)
Ibuprofen (racernicl
l\)
VJ
0
10
20
30
40
so
APPENDI X C
THERMAL ANALYSIS : VALIDATION OF THE EXPERIMENTAL DESIGN
231
(
Differ e ntia l Scanni n g Calorimetry Procedures
DSC scans of all ingredients of the formulations in eluding
ibuprofen.
Qualitative
were
assessment
performed
) .
thermal
behavior
of
excipient s
affect the fusion of the pure active drug in the range
of 70 to 79
tested
to
of all melting characteristics were
evaluated to check if the
will
( Figures
for
°c
and 45 to 60
°c.
Physical mixtures were
als o
endotherms and compared to pure ibuprofen. From
all the thermograms analyzed. there was no evidence
of
ad -
verse interference in the range of our study.
When necessary. the fusion parameters ( Tm
and~
H)
were recalculated from the indium standard using the foll owing equations:
K
Mc
---------Ac
~Hf
equation 1
K . M _____
As _
-----Ms
with
K:
the
equation 2
calibration constant; Mc: the mass of
I ndium; Ac: the area under the thermogram for the calibrant;
As : the peak area of the sample: M: the MW of the sample
232
Purity Determination:
Using a differential scanning calorimetry method. it
was p o ssible to validate the enantiomeric purity of the (+)S-ibuprofen. This thermal approach has been described in the
Thermal Analysis Newsletter #3.
edited
by
and
¥5
by
A.P
Gray
and
Perkin-Elmer . Norwalk Conn. The method consisted
of calculating the amount of
impurities
from
the
melting
point depression
To - Tm
6T
equation 3
and using the Van ' t Hoff equation:
To-Tm
with To
Tm
RTo
2
equation 4
Absolute melting temperature of the sample
the experimental fusion temperature
the enthalpy of fusion in cal / mole
R
the molar gas constant:l.987cal.mole
X2
the mole frcation of impurity
-1 0 -1
. K
The fraction melted can be calcuated from
F
( To - Tm) / (To-Ts)
233
equation 5
Ts
and
~
To - ( To -Tm).
l F
equation 6
with Ts: the melting point of the sample for a gi ven
fracti on melted.
F
Fraction melted
Fr om a DSC scan of the pure indium, the angle a
determined
was
( figure 16 ) and used in figure 17 t o obtain true
values of melting. The cor resp onding
fracti on
mel ted
were
calculated.
Ts was plotted against l / F in figure 18 . The
cept
To
and
and used in equation 4 to produce the amount of
In
(+)- 8-ibuprofen ,
X2
the
mole
fraction
impurities .
of impur it ie s
ranged from 0.0081 to 0.0 115. Thus the purity of
pound
ranged
inter -
the slope were derived from the straight line
this
com -
from 98.85 to 99.18 %. It is not known if the
impurities responsible for Lattice deffects might be in part
( - )- R- ibuprofen
molecules. Enanti omeric purity was also in -
vestigated with phase diagrams in manuscript V.
234
Piqure 1: Typical DSC thermogram of Ethyl ibuprofen
DSC Data F t la. t.oot
Sa.ple Wai ght•
13. 300 .g
Sun Hor 11141501061990
ETHVL !BU
,---
70.0
ti)
VJ
(J1
.
~
~
~
ll
l
i
···1i
so. 0
•~o
30.0
- - -
Tl
7 3. 40o · c
T2
ao. '.i66 ·c
Poak
·1e. ·110 · c
llrno
159.(1. "/34 rnJ
Daito II
i
119. 1S3 Ji g
,..w
llutqht
~iS .
nn~:mt
.'Ii. "l l 7 "C
17fi
i
I
20. 0
...v
jt
10. 0
I -
0.0
25.
u
~o. o
- - - - - T - - - -- - ·· - ,--- --- --- - -, -·'/'J. o
100. 0
12s.n
Tc111po1 ·ature c• c)
2 et S0111f>h
f~
l:
--·-- -
,~:8 ~
u..; ,.
0.0 •In ftJllTl!: 1.
S .D C/•I "'
ALAlN
--··--
Figure 2: DSC scan of Polyvinyl Pirrolidone (Povidone) in
the temperature range of interest
DSC Doto Ft I• pov2
PERKIN-ELIER
So.pl• Wal!flt..
U. 400 "9
Tue 1Lr1 20 12' 241 30 :999
POV I DONE <2nd Rt.ti>
7 Series Ther110l Analysis Syste111
3. s
3.0
ti)
u
CJ)
..
.
2.
s
.!I
0
;;:
2. 0
j
1.5
I. 0
0.
s
o.o ~--
1 ··--
30.0
JIS 1:
g:n H:I
~:
•I"
1 -·
"1.0
O. O
RATf:
0.0 •In
. I -!50.0
T..,_.-otlM"'•
t.
- -i 70.0
I
llO.O
c•o
AA.
J. Ro.aro
111.0
Figure 3: DSC
~=an
of Na Starch Glycolate (Explotab) in the
temperature range of interest
PERI< IN-ELMER
DSC Data Ft 1.. e.cp12
S011f>le Weig-.t1 27. 700 1t9
Tue Jun 20 13. 31.00 1989
7 Serles Thermal Analysis System
EXPLOTAB Uat Rt.ti)
- - - - - - - - - ··----- ----- --
3. 0
11.10
Tl
2. s ~II
[\)
V>
-J
•
!
--
•c
S7.en •c
T2
111.•11
•c
1.maJ
D elta H
O.O!IO J/9
2.0il \ .......
O.D!!ll •
~
u:
"D
1. 5
:!
1.0
0. 5
0.0
ms: t
ii:&~
- , -·30. 0
;u: 1:
---
1--·
4D. 0
..
,-
!iO.O
1·
ea. o
r...,,.,.ot ...... c•o
8: ~ ::~
ltA ll! I+
-
- -·AA .
,- - - ·
7D. o
J. RolftQro
811.0
Figure 4: DSC Scan of Lactose (Fast Flow Lactose) in the
temperature range of interest
nsc
Doto Fl
l oct~
1111
So..,.p:o Wc lght1
21. If.it) 119
Tue ; , n-. 20 13157111 l999
:... ALTOSE {!st RUIH
I
O.lto H
0. 054 Jig
Hol!l>t
0. 054 ...
, ' -i\
2. 2
2. 0
Vl
Ol
.
~
Tl
IM. 183
T2
118. ll50
1. 1.. 11 •J
.. 8
:
I
"-
l. <
~
:. 2 .
\
0
0
u. 8
o. 6
tl. 4
1111.415
A r...
1.6 .
~
•c
•c
·c
-
2. 6
t\)
7 Series Thermal Analysis System
r-----~J\J
I
I
I
I
I
o. 2
0. 0
ao
I
I
75. 0
so. 0
11: ~~
lo
f£- ~ o
1E- 'loo
.l~·'' L
• 5 . 0 I..
15U.U C.
tl-.
f/-
lo
.l o
hi! "'' ' '
1. 1
• I••
" ' !~ :,
•<'\ 1 1. ..! o
0,, 1 ' · ' '" ' ''
11. , lJ 1;·..,, .,
I
I
100. 0
Tmwperature <• C>
12'5. 0
A. A.
J.
fi o • "l~r t •
riqur• 5: Example of a typical endotherm of Ethyl ElJl;
An 77\ ibuprofen formulation with l\ of
disintegrant 1/3 intrgranular
JS\.. ilat.o Fl
} Q1
00130
<:i••unplo 'io!qht1
!.~. 5:rr. "'9
Sc:.. ;..;; 29 ::. : ~ . 16 1999
:· - 1 1i.
:... : Ji
40.
c
I
35. c .
e5.1183 ·c
Tl
T2
78.1183 •c
Peak
11 . 518
[')
C>I
'°
30. G
;
25. 0 -
.~
"·..
~
·c
933. 91 2 .1
Ar-eo
Datte H
7• .Zll7 J ·9
Hotght
32. l :'I .. ~
Oneel
75. 3Je
•c
2G. 0
15. 0 .
IO. n
5. 0
1
0. u
30.0
Tk.loOI'
lt!MI"
nu• co: l !Jrllth.. 1
JI-. l o
al . D c
lo
z.
1.in.o 1:
I
.a.a
I
50. 0
I
80. 0
I
70. 0
I
80.0
T•mparoture
1-D • I "
"IA"tc.
lo
I
!ID. 0
c•c>
I
100. 0
I
110. 0
A. J . ROMl::'.RC
I
120. 0
I
130. 0
Figure 6: Example o f a typ i cal endotherm of a ground
ibuprofen tablet obtained at regular compaction
DSC Oato I· i l 01
c.ht;i111
S0111plo rlo1 g~t 1
13. 550 ing
VGd ~r l .4 151'j'3107 ) gqo
Cl I TRi o\L I I
1·
35.
o
30. 0
"'""
0
j
I
01
§
~5.
~
20. 0
.
I
l"
-,
f3 ·l. 916 •r.
:.•
: :· . ' ~· :11 • l.
..,,....
'· ..,
• 11.
: •;.
c;•· f.l.
~,.
j ~.• 1
')~ifl
: 1:. >:1;
11
l l1dq'1';
.
: · n .,1 •~
• 1•
n:·11 ·11;
f,'l j
.:r:. .'~1
..,·,;
."3.fl'.H
•r:
I
15. 0
J
l:::Jr0.J
.:u.i:
;:::;
.
·~ 1 ; .
u
ThQ For111u l o tton we ight 1•
l: 1 ~g ~ n.a lo
o. o •I n
t;t:. G
llilll.•·•
•4TW 1•
.!.! :
1
•1.l
T uw1p, wt..:t~11·,!
l "t.>
'. 11 1.l J
ALAIN
~ I
;n. ~
: :i:. r
Fiqure 7: Example of a typical endotherm of a ground
physical mixture (before wet granulation)
DSC Doto F 11 "' grr
So.pl" 'lig19ht1
? . 169 •g
'liq\.! Mor 14 12153130 \990
GROUND PHYSIC'L MIXT:JR£
r- --~- -
::J '
/ 6 . ".'611 • ·.
6JO. •1''11
.'\ nlu
>I>
.. ~.c j
.
~
.
~
20. 0
l
15.0
j
l
'·
°.'9.'.iiH "l.
P1.:°"
!\)
;·u. ·: &J
111 1
l '.,l :.• J II
Hfi. l.J .;
t!1.Jl •Jl 1...
.!H.
'i~ h
1t1 4
I ' -;
;.·1::1.J T•
·. · ; . ~ .u
· 1.
I
I
I
!
I
I
10. a ·
I
I
5. 0
_J
l
o. 0
i ! •. I,
'tr. II
h ll . • I
'.,,
:1
Th1m w'°'l9ht. of tl'\GI phys.11i x l o '10.7 mg
l~: ~:
,sg: ~ ~
fJ14:. l,
0 . 0 •l t1 MArli
I•
S . O C · , , ,..
tl ll.11
'
•1:J. !I
~ '" · l ' ' ' "t l h t'
l ' l.i
'
\l lU.11
ALAIN
; II• .• 1
II. ,,
'
I lt•. LI
Figure 81 Typical Endotherm of (+)-S-ibuprofen
JS: Jc :.:o t= t l a. •
Smpla 1auj"l':.1
14.CCC "9
~bl ;u : C2 GS. .:.e. 23 lG'ie
S<+> :aJ P'...'RE
PERK I N- ELM<'. R
- SC!riC?s -;-hC?rmcl .1~C~f'3is SyGtC!m
~·
7C..G
Ro~u~ :.1 ?oc:k
!s(•> :o:; P:;a..:
I
85. 0 "1
~
I
80. 0 -
i
55. 0
"'ii>
"'
i
50.0
.
!
"u
o!'"
~
45.0
40.0 ·'
I
35. c
30.
~
r:
' TZ
i p~
Arc:!o
: l)glt.c;; H
Uo:c;h t
Cnect
'.'/ 2 ' 90
49. 1sc:
-
-1
g, 53 ·~
I
•c
!
57. 233 °L
s-t. 6G8 •c
i~25.
~
670 .;
e:·. ~'8
;:g
.&6. 987 •i
52. 474 .,.
c
25.0
20.0 -.
15. 0
:a. 0
"
5. 0 -:
0. 0
·i
I
50.0
25. 0
I
75. 0
I
100. u
Tc111parotura c•C>
lll:t:i: ,itg
~
fllllllrO
lo
c. 0
. , ... "" ~"' 1.
'
~25.
0
I
150. ~
Figure 9: Endotherms of (+)-S-ibuprofen and rac-ibuprofen
recrystallized at 4°c from methanol liquors.
osc o. ... r n ...111r
PERKIN··i:l.M'..'.R
'laopl• •·1~.. 1a. 000 .,
Thu Jul 05 1a.21o'8 lllQll
IETH I · II RIDl19T
<....- ! ocol
7 Series Thermal
Analysis System
·1
2•. o
Z2.0
20.0 .
1a.o
ro
"'
(,)
~
1a.o
;;:
g
1'. 0
j
12. D
10. 0
I
I
2.0 I
o.o-l ... - 1
30.D
I
40. D
I
I
50. 0
60. 0
I
.. ll. O
lOlllj)Qroturo c• C)
ll:U: ,11:8£
Tl• h
0.0 •1" llA1• I t
L 0 C/•I"
A~. 0
I
90.0
I
I
100. 0
-!
110. D
~iqure 10:
Thermogram of a Sand racemate mixture (25- 75)
melted and recrystallized at 4 °c
. '6. r.
..?~.LI
"'""
~
."
""
..
~
~.!.
r,
.""r..
c
:a. G
:o. r.
;t,r.
::~. c
:;..r
u.r.
J. : .
.....
... u
l:. n
:!S. 0
:::;:;
~:
..
.:· : .
. tlil..•
5ll. Q
"".i.CJ
~ t.~~ ~ ..:1rp 1 :r ~ .'.. :1• , 1
:.
\ •L)
: . ~J. e
ISO. u
Fiqure 111 Example of
s and racemate physical mixtures
[(75-25\)-24 hours-Labshaker-Room Temperature]
,,
.! ';. I :
.oC. :J
'i'::.
l\)
"'
(]J
.
11
1r:. :.;
.'."~ .
¥
c .
.~ c
I:
I,
c
I\
"
l'i.G ~
.r
'"·
'·
'
5. G
c. 0 ~
25, 0
'..i0. 1,;
·: :i.:1
:r.t:. u
';'" l!ir.
~~~:: ~
:~:t
t
u.:irut J:~u
1
~ • L)
::!:.i. 0
150.0
riqure 12: Thermogram of a directly compressible
(+)-S-
ibuprofen formulation [24 hours of mixing]
DSC Data Ft l• eMKI ~
Saooplo Wotghto
9. !lOO '"9
Mon ju t C2 10. 2t1 52 1990
SC ..> lBU
PLRK IN··E'..Ml"R
7
s~rieG
Therma l Ana:/G!G System
r ·R~su : t • Puck
!
45. 0 .
SC +)
40. 0
!\)
"'
(l)
..
...
"
~
j
Tl
35.0 -.
25. 0 -
101 30 AM
•1. 2:n ·c
I T2
55 . "93
•c
; Pook
53. BU
•t
i Aruo
30. 0
11 2/ 90
IBU RECRYST METHANOL 0 C
8 48. 381 .1
j 0a 1to H
Hotght
65. 493 Jig
29. 538 • ..,
I Gnsot
si. 2e2 ·c
20. 0
15. 0
10. 0 5. 0 -
o. 0 -i
I
25.o
n: ~:
~::l.
0
'.'5.
c
:r.c.o
:"umpnro tur c <"C>
,118
~
•ow
1o
I
1 ~5.
G
I
150. 0
riqure 13: Thermogram of a directly compressible
(+)-S- ibuprofen formulation [72 hours of mixing]
i)SC Dato i: 11 .. c30kn
So.pl• Wotqhto
10. 4QQ "S
T..... Jul 12 IS. 5S. 14 1990
FOR>!. C- 3 ro ~N J days
30. a~
RMult.1 Peok
I
25. a
N
"'
-.;z
s:
!J
!...
J
31
5Q PM
doy•
i
37. 583
Jl. ns
· Ar.a
311. 250 •J
3. 742 J i g
1. 171 .w
311.200 •c
I
•c
•t. 3511 ·c
Dolt.a H
H.l!#'t
0.-t
I
I
•c
: Tl
· T2
iPoak
20. a~
11
J
j
;
10. a
iI
"
I
!
i
a.al
I
30.0
JO. O
'.JC. ::
6~.r.
:·n. o
T{',,1 1~~rat:.Jrn
¥A: ,11:8 E
Tl•h
a..o
•t"
IVIT• h
I
8! .. :i
( • !: >
uci. a
'
·1
PEAK l
I
1s. a
7/ 12/ QO
FORM. C··3 rm KN 3
lOC. 0
110. 0
--·--,
.,
Q)
.,..
.c
""
c
....0
Q)
...c.
::l
.c
.....
I
"'
+
I
'O
c
::l
...0
..
Ill
.c
...:
~=
.."' ""
.,"'
.....a ..
....0
0
0
'O
g c
.....
...a .,
Q)
Q)
.c
E-<
.
.c
0
......
•k
-g"
...&
""
0
,;!
-, - '
0
0
.,
0
0
~
1'1
!! !!
~
N
- , •• · -
t -
0
248
-"'T ·- l·· · - ··r - · : ·
0
0
0
..
!::!
~
al
'
0
ol
.
0
...
~
0
0
<\!
d
Piqure 15: Thermogram of a ground (+)-S-ibuprofen tablet
obtained at 1200 lbs
DSC Dato F' 1• c3k2.4
!looplo Welcf>to
ID. 150 "'l
Thu Jul 12 lllo 52. 211 1111111
FIJl!N C- 3 IZIJO Lbe
CllNPAtTI~
i
25.01
i
l\)
II>
<O
~
!...
j
20.0 1
!
15. 0 -
10. 0 ~
o.o ...
30. 0
,40.G
I
30.G
eu. e
!
70.
7oinpQroturo
lllS 1: .11: 8 ~
" .. h
D.O •l" ""Tl: lo
R: :. D
[J
c•o
I
9G.G
100. 0
i
: 10.0
Fiqure 16: High Purity I ndium Melt
DSC Data F I }Q1
PERKIN-ELMER
t a9 t
Sor.pl a 'ial 9ht 1
9. 390 ing
Tua Fab OS 17104 1 38 1991
7 Series Thermal Analysi s System
Indium
50.0 {
45. 0
ro
.
WO
35. 0
0
30. 0
.
i
1
.!>
0"'
'L
"
u
!/
25. 0
::::~ :~
Tl
T2
Pooh
1si . 111 ·c
Ar oo
252. 957 .J
[)QJto H
2ti. 968 Jig
Holght
Oneot
30.302·•
1se.002
l
I
20 . 11 } - ••
15.
·c /
\I
I
I
I\
I
J. \,
---~
I
oI
I
10.
I
01
I
5. 0 .
0. 0 ·\
150. 0
I 152. 0
I
15 4. 0
I
I
156.0
158.0
1
160. 0
I
I
· 1
162. 0
164. 0
166. 0
h,lllJ>Oratura ( • C>
TE .... l o
TE-
z,
•O .O C
:J .&O. 0 C
I
168. 0
~
170. 0
Fiqure 17: DSC Sca n of Ibuprofen
( + )- S-i buprofen Purity Run
DSC Doto F- I I a, gpura
PERKIN-ELMER
Sa111pla Wul g h t 1
7 Ser ies Thermal
Thu Ju l 12 121 481 IS 1990
Analys i s Sys tem
CONTROL S< •l 42. 6 I-
55. o
II
Tl
•e. 503 ·c
T2
56. s16 · c
50. 01
Paok
45 . 0
Arao
s•. 693 ·c
12 14. 27 4
m)
I
ti)
(JJ
....
.
".!)
40.0
35.
o
0
~
30. 0
~
l
2r,. o
?O. 0
~
I
l
f"
I
'
I
'
I
"'t-
,'
'
!",
'
''
'
\
'
I ,' /
'
I
/
I
' '
I
-~
/ ' 'I ' .
~ _/__ ,· _ ~ _}_ L. ---- ·- -- I
10. 0
,
I
I
•
5. 0 \
0. 0
I
50.0
I
40. 0
46. 0
I
52.0
I
54.0
Tc mparotura <•c>
TE"IP
u: ...
lo
2.
l~~: g ~
TIMf"
lo
0 . U "' ' " Jt ... TF.:
lo
I
56.0
I
58. 0
j
I
Figure 18
(+)-8-lbuprofen Purity Determlnetlon
53
-....•
•..
~
•~
....•
B
U1
!\)
-
.e1ee
•
0.0 11
r:;.~
;,;
Reqr •
• : To-Tm
4H 1
(x2) • RTq 1 I
.
xi •
!'urKr LH
·•··.......... .
l'
!\)
$lope •
•
ai • ::
0. 9
98 .8~"
0~•1<;:~:.>r;.:.;: 1AF
.
52
51
0.25
1.00
1.75
1/F
2.50
APPENDIX D
EXPERIMENTAL COMPLEMENT OF MANUSCRIPT II
253
In this work other
complement
the
results.
experiments
were
performed
to
However I found some of it incom-
plete or noc statistically powerful. Therefore. in agreement
with
the
authors
it
was dec i ded not to incorporate these
data in the manuscript for publ ication. Some of the data
is
graphically presented in figures 1 and 2 or tabulated (table
I ) .
(
254
c:
0
;;
f'f'-
•
-;
E
~
0
c:
.....
... ..
•...
-0
•...
::I
e!
c:
0
f'-
<O
:I.
,gt E
u.. ·-
•
c:
u
0
()
....
-•...
-•••
111=
c:
'"l
~.
a
..•...
~i
:I
c:
;. ~
0
Q.
..
!i5 .....
c:
i5
0
N
f'l{)
0
co
0
""
<•inu111> 8W!.1
255
0
:e
0
.i-
'#
....
....
I
c
0
...
·;;
.
••a.
I
I
I
.
-·.. '°
...
I
E
'#
0 .. ....
•c
Ill
~
:: 0
c-.
•.
::I
a
u::
0
N
:i
u
cQ,
I
I i
~-r
I
::.
0
u
1,
t-~~Q~
\
I
c
t
0
t--\~
....'#
IO
............
' ',
0
.i-
0
f'l
z
0
j::
! \I
I
...u
...0
II:
n
I
u
z
~
~t
•
U%
f'l
I
I
•
:i
0
()
0
<j> ~
I
~
0
N
I'S) SSJNOllYH
256
0
''
',,
''
'
0
0
Table I: Pure ibuprofen Compacted: Thermal Analysis
Compaction
Level in KN
N
(Jl
-'1
Mean (SD)
0
1(1)
8(3)
30(3)
1
4H
(J/g)
Tm
( 'C)
127.1
127.8
124.5
125.0
77.5
75.3
74.7
75.5
BIBLIOGRAPHY
Adams SS. Bresloff P , Mason CG,
Journal of Pharmacology & Pharmacology 28:256-257. 1976.
Adams SS. Cliffe EE , Lessel B, Nicholson JS .
Journal of Pharmaceutical Sciences 56:1686, 1967.
Aguiar AJ . Krc J, Kintel AW and Samy n JC
Journal of Pharmaceutical Sciences 56 , pp 847-853, 1967
Ahn Hae-Young, Amidon GL and Smith DE
presented
at the A. A.P.S . meeting in Las Vegas
Pharmaceutical Research
7. (9): 234. Sept . 1990 .
Akbuga J. and Gursoy A.
Drug Development & Industrial Pharmacy 13 ( 14 ) :2541-1552 (1 987 )
Alhec C . and Zografi G.
Ac c epted International Journal of Pharmaceutics . , 1991
Ansbacher Rudl Journal of the American Medical Association
(2 59 ), 2 , 220,
(1988 )
Arnaud P. Pharmaceutica Acta Helvetica
258
58 ( 11 ), 290 - 297 ( 1983 )
Baille TA. Adams WJ. Kaiser DG. Olanoff LS. Halstead GW
Har pootlian H. VanGie sen GJ. Journal of Pharmaco l ogy &
Experime ntal Therapeutics 249 ( 2 ) :517 - 523. (1989).
Benkerrour L. Puisieux F. Duchene. D.
Pharmaceutica Acta Helvetica 57 ( 10 ) :301 - 308 ( 1982 )
Bettinetti GP. Gioardano F. Italia A, Pellegata R
and Ventu ra P , AGPI , 3 , pp 232 - 240. ( 1988)
Britain, Harry G. Drug Developement & Industrial Pharmacy
15 ( 13 ) , 2083-2103 ( 1989 )
Britain Harry G .. Pharmaceutical Research
7 ( 7 ) :683 -690 (1990)
Byron PR and Notari RE
Journal of Pharmaceutical Sciences 65 ( 8 ), 1140-1 144 , 1976
Campbell
DB,
Europeen
Journal
of
Drug
Metabolism
Pharmacokinetics 5 ( 2) pp 109-125. ( 1990 )
Carstensen JE Journal of Pharmaceutical Sciences 65:992 - 997.
( 1977 )
Cartwright, AC
Drug Information Association Journal
Vol 24 , pp 115 - 116,
( 1990 )
259
Celik M and Marshall K. Drug Development & Industrial Pharmacy
15(2 ) . 758-800. ( 1989)
Chan HK and Doekler EK
Drug Development Industrial Pharmacy 11 ( 2&3 ) 315 - 332. ( 1985 )
Chilamkurti RN. Schwartz JB and Rhodes CT. Pharmaceutica Acta
Helvetica 58:253 ( 1983 )
Chowan ZT , Journal of Pharmacy & Pharmacology
32:1 0- 14 ( Jan ) 1980
Chowan ZT and Amaro AA
Drug Development & Industrial Pharmacy 8 ( 14 ) :1079- 1106 (1988 )
Chowan ZT and Chow YT
International Journal of Pharmaceutical
Technology & Product Manufacture 2 ( 1 ) :29-34 ( 1981 )
Chowan ZT and Palagyi LH
Journal of Pharmaceutical Sciences. 67(10 ) , 1385 - 1389 (1978 )
Cox SR . February (1988)
Clinical Pharmacology and Therapeutics 43 : 146 (1988).
Cox SR. Brown MA , Squires DJ . Murrill EA. Lednicer D, Knuth
Biopharmaceutics & Drug Disposition 9:539-549. ( 1988 ) .
260
DW.
Cruaud 0. Duche ne D. Puisieux F. Chauvet A. Masse J
Jour n al d e Pharmacie de Belgique 36(1 ) . pp 15-20. (1981 )
Edmonson IC
"Advances in Pharmaceutical Sciences· Vol.2
Edited by H.S. Bean, J.E . Careless and A.M. Beckett
Ac ademic Press. London 1967 pp-95
Formi A. Moretti I and Torre G, Chemical Society & Perkin
Transcripts 2. pp 791-797, 1984
Franz , Robert. M.
"Stable High Dose. High Bulk Density Ibuprofen Granulations
for Tablet and Capsule Manufacturing·
U.S. Patent 4.609 . 675 Sept. 2. 1986
P.E . Gordon , C.L . VanKoevering , and D.J. Reits
International Journal of Pharmaceutics 21 : 99 - 105.
( 1984 )
Haleblian , John K.
Journal of Pharmaceutical Sciences 64 ( 8 ), pp 1269 - 1288 , ( 1975 )
Grant DJW , Medhizadeh M.
AH and Fairbrother JH
Cho~
International Journal of Pharmaceutics 18 , 25 - 38 , ( 1984 )
Greenblat t.DJ
and
Shader , RI
261
Jou r nal of Cl inical Pha rmacology ( 27 ) . 85-86. ( 198 7)
Gesslinger G. Stock KP. Bach GL, Loew D. Brune K. Agen ts and
Ac ti on s . 27(3 4 ) :455-457. (1 989)
Haynes , J oh n D. Journal of Pharmaceutical Sc iences
( 70 ). 6. 673-675, (198 1 )
Hempel A, Camerman N and Camme r man A,
Journal American
Chemical Society . 2, 3453-56, ( 1982 )
Herzfeldt HR a nd Kummel R , Drug Development & Industrial
Pharmacy 9(5 ). 767-793 . 1983
Hess . Hans
Pharmaceutical Technology pp 36 - 50. Sept . (1 978 )
Hiestand EN, Amidon GE. Smith DP. and Tiffany BD
International
& Powder
& Bulk
Solids
Handling
Rosemo n t. I L (1981)
Kwon HK . Fran ck MJ. and Henderson NL
Journal of Pharmaceutical Sciences 74(3 ) , pp 283 -289 ,
March ( 1985)
A.J . Hutt and J. Caldwell
262
Procedures,
Journal of Pharmacy & Pharmacolology 35:693-704. ( 1983 )
Jacques J . Collet and Wilen SW
Enantiomers. Racernates and Resolutions J. Wiley 1981
a) pp
32-43: b) pp 88-104
Jarnal i F. Sing NN. Pasutto FM. Russell AS. Coutts RT .
Pharmaceutical Research 5(1):40-47. ( 1989 )
Jones. TM
Journal of Pharmacy & Pharmacology 31. 17 -23 ( 1980 )
Jordan RP and Rhodes CT
Drug Development & Industrial Pharmacy 5, 151 - 171 (1 979 )
Jusko WJ and Gibaldi M
Journal of Pharmaceutical Sciences ( 61 ) .8 , 1270-1273, (1 972 )
Nahoko Kaniwa , Hiroyasu Ogata , Nobuo Aoyagi , Yasushi Takeda
and Mitsuru Uchimaya Journal of Pharmaceutical Sciences ( 78 )
12 , 1020-1024 , (1989 )
Khan KA and Rhodes CT
Journal of Pharmaceutical Sciences 65 ( 12), 1835 - 1837 ( 1976 )
Khan KA and Rhodes CT
263
Pharmaceutica Ac ta Helvetica 50 ( 6 ) . 20 - 24. ( 1975 )
Knihini cki RD. Day RO . Graham GG and Wi lliams KM.
Chirality 2:134 - 140. ( 1990 ) .
Kopp S , Beyer C. Graf E. Kubel F and Doelker EK
Journal of Pharmacy & Pharmacology 41:79 - 82 ( 1989 )
Kr yce I. Pope DG, Hersey JA
International Journal of Pharmaceutics 12:113- 134 ( 198 2)
Lamy , PE, Journal of Clinical Pharmacolology ( 26 ) . 309- 316 ,
( 1986 )
Lee EJD. Williams K. Day R. Graham G, Champion D.
British Journal of Clinical Pharmacology 19:669-674, ( 1985 )
Lefebvre C , Bouge R, Ringard J and Guyot - Herman AM
"Modif ication
actif
imputable
de
l ' aptitude
a
a
la
l ' augmentation
granulation d ' un principe
de
la
solubilite
de
surface de ses particules. "
Presented at Agpi Paris (June 1989 )
Lefebvre C , Guyo t - Herman AM, Guyo t JC, Bouchet R and Ringard J
Drug Development & Industrial Pharmacy 13 ( 9-11 ) , 224 - 235 ( 1988 )
264
Lefevbre C. Guyot - Hermann AM. Draguet M. Bonde R and Guyot JC
Drug Developme nt & I ndu s t rial Pharmacy 12 ( 11g13 ) . pp 1913 - 1927.
( 1986 )
Leuenberger . H . Internat ional Journal of Pharmace ut ics
12 : 41 -55 (1 982 )
Leuenberger H, Bier HP and Sucker HB
Pharmaceutical Technology 60 - 66 , June 1979
Levine Stephen Hippocrates Ma rch April 1989
Levy G
Journal of Pharmaceutical Sciences
52 : 1047 (1963)
Lin Jue Chen Drug Development & Industrial Pharmacy
13(12).
2087-2110 (1 987 )
Lindberg NO , Johnsson C a nd Holmquist B
Drug Development & Industrial Pharmacy , 11 ( 4 ), 917-930 (1 985 )
Liu ST and Hur v i tz A,
Journal of Pharmaceutical Sciences , Vo l 67, 5, pp 636-638. 1978
Lovering EG , McGilveray IJ. McMillan I and Tostowaryk W
Journal of Pharmac eutical Sci ences ( 64 ), 9 , 1521 -1 524. 1975
265
Lovering EG. McGilveray IJ. McMillan I. Tostowaryk W.
Matula T and Marier G
Canadi an Journal of Pharmaceut ical Scien c e s
(10). 2. 36-39.1975
Mackuch RW , Johnson MF , Ar c hive of Int e rnal Me d icine (1 46 ) .
986-989,
1986
Mahon WA. Leeder JS. Brill - Edwards MM, Correia J and
MacLeod SM Clinical Pharmacoki neti cs ( 13 ), 118-124.
1987
Mayer . JM , Ac ta Pharmaceutic a Nordica 2 ( 3 ). 196 -216,
1990
McConnell . JF , Crystal Structure Communication s 3:73 -75.
1974
Mehata A. Adams KS. Zoglio MA , Carstensen JT
Journal of Pharmaceutical Sciences 66(10 ) :1462-1464 ( 1977 )
Mehvar Rand Jamal i F, Pharmaceutical Research , 5:76 - 79. 1988.
Metzler. Carl M. ,
Journal of Clinical Pharmacology (2 9 ),
289-292 , 1989
Menard FA , Dehdiya MG. and Rhodes CT
Drug Development & Industrial Pharmacy 14 ( 11 ), 1452 (1 988 )
Midha KK. Chakraborty BS , Schewde R, Hawes EW ,
266
McKay G. Hubbard J W and Moore M
Journal of Pharmaceutical Sciences ( 79 ) . 1. 3-8. 1990
Moss JN and Rhodes CT. Canadian Journal of Pharmaceutical
Sciences (9),
1. 30-33, 1974
Mura P , Liguori A, Bramanti G. and Poggi L
Il Farmaco- Ed. Pr .. Vol. 42 , Fasc. 6 , pp 157 -164 . 1987
Nakai Y Drug Developemnt & Industrial Pharmacy 12(7 ). 1017-1039 ,
(1 986 )
Newitt JN and Conway-Jones ,
Transcript of International
Chemical Engineering 36:422, ( 1958 )
Nightingale SL , Morrison JM. Journal of the
American Medical
Association ( 258 ) . 9 , 1200 - 1204. 1987
Nyqvist . Hakan
Drug Development & Industrial Pharmacy 15 ( 16&17 ), 957 - 964 ( 1989 )
Olson SC, Eldon MA , Toothaker RD, Ferry JJ and Colburn WA
Journal of Clinical Pharmacology (27), 342-345, 1987
Perrier , Donald D and Gibaldi M
Journal of Pharmaceutical Sciences ( 62 ), 2 , 225 - 228 , 1973
267
PMA
Ad
Hoc
Committee
on
Racemic
Mixtures .
Pharmaceutical
Technology pp 46 - 52. May 1990
Powell ML. Wei sberger M. Gural R. Chung M. Patrick JA. Radwanski
E and Symchowicz SS Journal of Pharmaceutical
Sciences ( 74 ) .2 , 217-219 , 1985
Rees JE. Bulletin of Chemistry & Pharmacy , 112. 216-220 . 1973
Remon JP and Schwartz JB
Drug Development & Industrial
Pharmacy 13 (1), 1-14 ( 1984 )
Repta AJ , Ealtezor MJ and Bansal PC
Journal of Pharmaceutical Sciences , 65 (2): 238-242. 1976
Riad LE , Chan KK , Wagner WE and Sawshuk RJ
Journal of Pharmaceutical Sciences (75), 9, 897-900, 1986
Richmond B , vescuso P.
STELLA.
Hight
Peterson
Performance
S,
Systems,
A
Business
User ' s
Guide
Inc. Lyme, New Hampshire .
1987.
Ringhand HE, Ritschel WA , Meyer MC, Straughn AB and
Cabana BE
Journal of Pharmaceutical Sciences ( 72 ), 12 , 1983
Romero AJ , Grady LT
and Rhode s CT
268
Drug Development & I ndu strial Pharmacy ,
14 (11).
pp
1429 - 1457.
Acta
Helvetica ,
1988
Romer o AJ. Lukas G.
Rhodes
CT
Pharmaceutica
66 (2). 34-43. 1991
Romer o AJ , Rackley R and Rhodes CT
Chirality . Accepted for Publication. III ( 5 ) . 1991
Romer o AJ and Rhodes CT accepted for publicati on in
Drug Development & Industrial Pharmacy . 17 (5),
1991
Romero AJ. Bon C., Johnson EF , Rosembaum SL and Rhodes CT
Clinical
Research
Practices & Regulatory Affairs 8 ( 2 ): 123 -15 1,
( 1990)
Rosenbaum SE, Bon C. Rhodes CT
Drug Development & Industrial Pharmacy ( 16 ),
1. 157 - 163. 1990
Rumpf H. Chemical Engineering & Technology , 30:144- 158 ( 1958 )
Schulman. Robert A Journal of the American
Medical
Association
(26 1 ), 17 , 2499 , 1989
Stamm A and Paris L
Drug Development & Industrial Pharmacy 11 (2&3 ), 333 - 360 ( 1987)
269
St o ffer SS
Szpuner WE Journa l of t h e American Me di c al
Association 244 ( 15 ) . 1704 - 1705. 1980
Summers SE . Enever RP and Carless JE
Journal
o f Pharmaceutical Sc ienc e s 66 ( 8 ), 1172-1175.
( 1977 )
Wyllie E. Pippenger CE , Rother AD
Journal of the American Medical Association ( 258 ) . 9 ,
1216-1217 , 1987
York P and Grant DJW
Ac ta Pharmac eutica Suecica . Symposium on Tablet Technology
February 24 - 27 , pp 63-64, 1987
Udupa N. Drug Development & Indus trial
Pharmacy
13(15).
2749-
2769 ( 1987 )
Unvala HM. Schwartz JB and Schn aare RL
Dru g Developemnt & Industrial Pharmacy 14(10) , 1327 - 134 9 (1988 )
Vromans H. DeBoer AM, Bolhuis GK a nd Lerk CF
Acta Pharmaceutica Suecica 22, 163-172 (1985)
Wagner. John G "Phar macoki ne tics:
Notes"
Published by J.M . Richards Laboratory 1969
270
pp
132 - 138,
161-166
Wagner, John G.
J ournal of Pharmac eutical Scien ces ( 72 ) . 7.
838 - 884. 1983
Wagner JG and Nelson E
J ournal
of Phar mac eutic al Sc ienc es (52). 6 , 610-611 , 1963
Walds PD , Galleta G, Braden NJ and Alexander L
Ac ta Pharmacologica & Toxicologica 59 ( Supp.V ) . 155 (1986)
Wallace JW. Capozzi JT. Shangraw RF,
Pharmaceutical Technology Sept 1983.
Wood JH, Syarto JE and letterman H,
Sciences
Journal
of
Pharmac eutical
54, 1068-1075, 1965
Yalkowsky SH and Valvani SE Journal of
Pharmaceutical
Sciences
69 (8) , 912-921. August 1980
Zog l io MA.
Journal
of
Pharmaceutical
Sciences
65:1205 - 1208 ,
(1978)
Division
of
Bioequivalence
Bioequi v ale nc e ,
and
In
"Gu i d a n c e
Vitro
I ndomethacin " FDA 2 1 CFR 10.90 ( 1988)
271
f or
Di sso l ution
In
St udies
Vivo
for