MARCEL DEKKER, INC.

• 270 MADISON AVENUE • NEW YORK, NY 10016

©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

PHARMACEUTICAL DEVELOPMENT AND TECHNOLOGY

Vol. 8, No. 2, pp. 119–126, 2003

A New Approach to Accelerated Drug-Excipient Compatibility Testing

Jonathan L. Sims,* Judith A. Carreira, Daniel J. Carrier, Simon R. Crabtree,

Lynne Easton, Stephen A. Hancock, and Carol E. Simcox

GlaxoSmithKline R & D, New Frontiers Science Park, Harlow, Essex, UK

ABSTRACT

The purpose of this study was to develop a method of qualitatively predicting the most likely
degradants in a formulation or probing specific drug-excipient interactions in a significantly
shorter time frame than the typical 1 month storage testing. In the example studied,
accelerated storage testing of a solid dosage form at 508C, the drug substance SB-243213-A
degraded via the formation of two oxidative impurities. These impurities reached a level of
1% PAR after 3 months. Various stressing methods were examined to try to recreate this
degradation and in doing so provide a practical and reliable method capable of predicting
drug-excipient interactions. The technique developed was able to mimic the 1-month’s
accelerated degradation in just 1 hr. The method was suitable for automated analysis, capable
of multisample stressing, and ideal for use in drug-excipient compatibility screening.

Key Words: Drug-excipient interactions; Degradation studies; Automation; Prediction of

impurities.

INTRODUCTION Typically, degradation processes are studied for both the

During the preparation and storage of drug products,
new impurities often form as a result of an interaction
between the drug substance and species introduced by
formulation.[1 – 3] In extreme cases, this can lead to shelf life
issues which necessitate reformulation. The current
International Conference on Harmonisation (ICH) guide-
line; Stability Testing of New Drugs and Products,
Q1A(R)[4] requires stress testing of drug substance
to help establish the likely degradation pathways.
drug substance and drug product using extremes of pH,
heat, light, and oxidizing agents.[5 – 7] These experiments
are also used to develop stability indicating impurities
methods for the drug substance and drug product.[8 – 10]
However, many of the degradation products observed in
these processes do not occur in the drug substance or
product when stored under normal conditions and are not
observed in the stability trials. The presence of unnecessary
impurities in stressed samples can cause the development
of overly complicated methods leading to problems in

*Correspondence: Jonathan L. Sims, GlaxoSmithKline R & D, New Frontiers Science Park, Third Avenue, Harlow, Essex,
CM19 5AW, UK; Fax: 01279 622380; E-mail: jon_sims-1@gsk.com.

119

DOI: 10.1081/PDT-120018476 1083-7450 (Print); 1097-9867 (Online)

Copyright q 2003 by Marcel Dekker, Inc. www.dekker.com
 MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

120 Sims et al.

the analytical support of late-stage products. Therefore, we ability to quickly screen potential excipients for
set out to develop a procedure that would enable us to focus incompatibilities, although other features such as low
on the most likely degradants to arise from a given tablet sample consumption make it an attractive method.
formulation by applying stressing techniques to the Although DSC is unquestionably a valuable technique,
solid state. interpretation of the data is not without problems.
The formulation of a drug substance involves it being Similarly, isothermal microcalorimetry is becoming
blended with a combination of different excipients to popular as a method of detecting changes in the solid
maximize the products ability to administer the dosage state of drug-excipient mixtures through heat changes.[19]
effectively. Ideally, the excipients used in the formulation The results from these methods can be indicative of
should not interact with the drug substance or introduce whether formulations are stable, but thermal techniques
species capable of accelerating the formation of new reveal no information concerning the cause or nature of
impurities. Certain classes of compounds are known to be any incompatibility. Techniques such as hot stage
incompatible with given excipients[11]; thus, knowledge of microscopy and scanning electron microscopy can be
the chemistry of the drug substance can often minimize used in conjunction with DSC to determine the nature of
formulation surprises. Heat and water are the primary an apparent incompatibility.[20] These techniques study
source of drug product incompatibilities and play a critical the morphology of the drug substance and can determine
role in the stability of a drug substance to degradation.[12] the nature of physical transformations, thus indicating the
Many different degradation mechanisms exist but those type of incompatibility that has occurred.
mediated by surface moisture appear to be the most
common,[13] and as a result, it is important that stressing
methods incorporate water to allow for the formation of all
possible impurities. The way in which water facilitates MATERIALS AND METHODS
degradation is not fully understood, but the work carried
out by Kontny et al.[13] has confirmed its importance. Manufacture of Tablets
Degradation problems can, therefore, be difficult to avoid
because water is often trapped inside drug products on Tablets were manufactured by using a wet granula-
formulation. Many excipients are hygroscopic tion process. The active, povidone, sodium lauryl sulfate
materials[14,15] and absorb water during formulation and microcrystalline cellulose were granulated with
processes such as wet granulation. Excipients generally water. Further microcrystalline cellulose, croscarmellose
contain more free moisture than the drug substance,[14] and sodium, and magnesium stearate were added extragra-
in an attempt to obtain the most thermodynamically stable nularly, and the mixture was compressed into tablets.
state, water is able to equilibrate between the tablet Tablets were then coated in a fluid bed dryer.
components.[16] Formulation can, therefore, potentially
expose the drug substance to higher levels of moisture than
normal, which greatly increases the susceptibility of even
Drug-Excipient Mixtures
the most stable compound to degradation. Another
common source of formulation problems is the interactions
The basic formulation contains active plus micro-
that can occur between residues found in excipients and the
crystalline cellulose, crosscarmellose sodium, povidone,
drug substance.[3] Excipients are often isolated from
magnesium stearate, sodium lauryl sulphate, titanium
natural products or polymerization processes and may
dioxide, hydroxypropyl methylcellulose, and polyethy-
contain low-level process impurities, particularly in the
lene glycol 400. Experiments where the drug substance
case of polymeric excipients such as the polyethylene
was stressed with either single or binary excipient
glycols. These residues have the potential to react with the
drug substance; therefore, it is important to know the purity
and composition of the excipient prior to formulation.
These incompatibilities are often difficult to predict but
represent a very real source of formulation instability.
Thermal techniques are a rapid tool that can be used
to examine for incompatibilities between a drug substance
and excipients.[17,18] Differential scanning calorimetry
(DSC) is currently the leading technique in this field. The
main benefit of DSC over stressed storage methods is its Figure 1. Structure of SB-243213-A.
 MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Drug-Excipient Compatibility Testing 121

Figure 2. Original stressed tablet impurity profile.

mixtures used 10 mg of each material. All mixtures
prepared from dry powders were ground together by
using a pestle and mortar.
Tablet Stability Studies
Tablets were stored in high-density polyethylene
bottles with heat induction foil seals at 508C, 408C/75%
relative humidity, 258C/60% relative humidity and 58C.
HPLC Method
ammonium formate buffer adjusted to pH 3.3 with formic
acid. The initial mobile phase composition of 25%
acetonitrile was increased to 60% linearly over 25 min,
followed by an increase to the final composition of 80%
linearly over 5 min and held at this composition for 5 min.
Degradants were routinely monitored by using a UV
detector at a wavelength of 266 nm with identity
confirmed by using mass spectral detection. The sample
and controls were dissolved in a 50/50 vol/vol mixture of
water and HPLC grade acetonitrile and were passed
through a 0.2 mm filter to remove any undissolved
excipients.

HPLC analysis was performed by using a 5-mm, SK233/STEM Automated Analysis System
150 £ 3.9 mm i.d. symmetry C18 column (Waters, Herts,
UK) at 408C with a mobile phase flow rate of 1 mL/min. This instrument consists of an Anachem SK233 XL
The gradient elution used acetonitrile and a 50 mM autosampler with 183-mm piercing probe, Gilson 402

Figure 3. Structures of impurities.

 MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
122
Figure 4.
dilutor, and STEM Corporation model RS1000 reacto
station controlled by SK233 Workstatione software
(Anachem, Luton, Beds., UK) coupled to an Agilent
1100 high-pressure gradient HPLC system with diode
array UV detection (DAD) and electrospray ionization
mass spectrometric detection (MSD) controlled by
Agilent Chemstation software version Rev A7.01
(Agilent Technologies, California, USA).
Figure 5. Glass weighing boat in vial for oven experiments.
Sims et al.
Degradation of SB-243213-A.
RESULTS AND DISCUSSION
The series of experiments described in this article were
designed to try and recreate the degradation observed
following the 3-month storage at 508C of the initial clinical
formulation for SB-243213-A, a 5HT2c receptor antagon-
ist under development for the treatment of anxiety and
depression (Fig. 1). Figure 2 shows a comparison of the
impurity profiles for the degraded product, drug substance
control, placebo, and blank; the degradation produces two
impurities at a level of approximately 1% PAR designated
as impurity A and impurity B (Fig. 3). The structures of
these impurities had been elucidated previously during
forced degradation experiments on batches of drug
substance, and the major degradation products are shown
in Fig. 4. However, impurities A and B are only detected as
major components when SB-243213-A is stressed with
light and oxidizing agents, not from heat and moisture
stressing.
The drug-excipient compatibility testing method
adopted by Serajuddin et al.[21] involved the storage of
formulated samples with 20% vol/wt added water at
508C for 1 – 3 weeks. We desired a more rapid screening
system; therefore, samples were prepared with 20%
vol/wt of water, placed in closed vials, and stored in an
 MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Drug-Excipient Compatibility Testing 123

Figure 6. Oven experiment results.

oven at 1008C for 1 and 3 days, with relevant controls.
The results from the oven experiments revealed that
although the desired degradants were being generated,
the major degradation process differed from that seen in
the original study to such an extent that impurities A and
B would have been considered insignificant. It is of
interest that the drug-excipient sample stressed in the
presence of water formed impurities A and B at increased
levels in comparison with the same sample stressed
without water. This confirmed the importance of water as
a promoter of the observed degradation process.
Following the results obtained from the solid-state
oven experiments, we decided to modify the experimen-
tal design to produce a more controlled and humid
atmosphere on the basis that stressing the sample in the
presence of moisture rather than having the sample in
direct contact with water might provide a more predictive
result. The method required the sample to be held in
a glass weighing boat inside a closed vial with water
around the outside of the boat (Fig. 5). The initial
experiment stressed the sample for 24 hr at 1008C, and
this resulted in the formation of the oxidative impurities
at low levels (A and B forming at 0.08% and 0.67% PAR,
respectively), alongside the major degradation products
observed in the previous heating experiment. The level of
degradation found, in comparison with the stored drug
product, was considerable and significantly different
from that shown in Fig. 2. The experiment was repeated
by using a stressing time of 1 hr, and the overall
degradation was reduced. Furthermore, it was believed
that 1008C might be too high a temperature to mimic
the drug product result, and a comparison was performed
at temperatures of 60, 80, and 1008C using three
subsamples from a single preparation of the formulation.
At 1008C, a large number of degradants not found in the
drug product predominated, whereas in the 60 –808C
range, the oxidative impurities were the predominant
result (Fig. 6). To evaluate the reproducibility of the
system, a drug-excipient mixture was prepared, and three
subsamples were stressed at 808C for 1 hr. The degradants

Table 1. Reproducibility of stressing in oven.

Impurity A %PAR Impurity B %PAR

Sample 1 0.33 0.36
Sample 2 0.29 0.24
Sample 3 0.29 0.23
Figure 7. STEM Block with condenser array.
 MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

124 Sims et al.

Figure 8. Chromatograms from standard STEM stressing experiments.

found for the three samples gave good agreement,
confirming the method to be reliable; the results are
summarized in Table 1.
We had previously successfully automated solution
stressing of drug substance by adapting the approach of
Figure 9. Reaction vessel.
Armitage et al.[22] using the SK-233 and STEM
system,[23] which has 10 reaction tubes with condensers
(Fig. 7). This equipment was, therefore, used to try to
develop a method capable of on-line stressing and
analysis for solids. The drug and excipients were
prepared as described earlier and were stressed in the
standard STEM tubes with 20% vol/wt added water at
808C. The STEM results (Fig. 8) were very similar to
those obtained for the oven solid-state experiments.
These results confirmed that keeping the sample and
water separate is a necessity for a predictive method;
therefore, we designed and built a novel reaction vial
(Fig. 9) for the STEM block system. The new vial is only
5 cm at the shoulder (equivalent to the depth of the hole
in the STEM block) and, therefore, only the cap
protrudes above the heated block. It has the sample in an
insert inside the vessel, water is contained around the
outside of the insert, and the vial sits in the stem block.
Samples of the formulation and drug substance were
heated in this apparatus at 808C for 1 hr. For the
formulation, this experiment generated impurities A and
B at a level of 0.36% and 0.53% PAR, respectively,
giving a result consistent with that found in the original
storage study (Fig. 10). In the STEM reactor, the drug
substance can also be seen to undergo the same
degradation process as the formulation although to a
lesser extent. As in the oven experiments, we performed
a reproducibility study for the STEM vial using four
subsamples from a prepared mixture which produced
 MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Drug-Excipient Compatibility Testing
Figure 10. Chromatograms from STEM stressing experiments using new reaction vial.
results consistent with the impurities A and B being the
most significant degradants for this formulation (Table 2).
The experiments described in this article prove the
viability of performing rapid stressing experiments on
drug products; however, the exposure time may vary
dramatically depending on the nature of the drug
substance. SB-243213-A is reasonably easy to degrade,
hence, the optimum stressing time of 1 hr. Other
formulations of SB-243213-A could be stabilized by the
excipient combination[12] and require longer exposure;
therefore, we do not believe that a generic method can be
proposed for this application. Because the STEM block
has a capacity of 10 vials, it is possible to have samples
taken at different time points to establish the optimum
exposure time; clearly, if the required exposure is in the
order of weeks, then there is no need to automate.
The STEM block plus 10 vials is relatively
inexpensive as a stand-alone option and can be used for
screening of different formulations for variation in
degradation or for the potential to stabilize a given
formulation through selected additives.[12] With 10 vials
exposed to identical heating, the apparatus is suited to
structured experiments exploring binary or tertiary
Table 2. Reproducibility of stressing in STEM vial.
Impurity A %PAR Impurity B %PAR
Sample 1 0.36 0.53
Sample 2 0.33 0.36
Sample 3 0.29 0.24
Sample 4 0.29 0.23
125
mixtures to identify specific degradation drivers.[21] The
vial design is also suitable for stressing of whole tablets.
CONCLUSION
A forcing method has been developed, which is
capable of reproducibly producing degradants in a
formulated product which were previously revealed
during a formal accelerated stressing study. The new
method offers a significant reduction in the time taken to
discover the most likely degradants for a formulation and
can also be applied to comparative studies of potential
formulations.
REFERENCES
1. Buckley, D.; Haas, M.J.; Scholtz, T.H.; Sigvardson, K.W.
Isolation and identification of a degradation product in a
capsule formulation containing the elastase inhibitor
DMP 777. J. Pharm. Biomed. Anal. 2001, 24, 651– 657.
2. Chen, J.G.; Markovitz, D.A.; Yang, A.Y.; Rabel, S.R.;
Pang, J.; Dolinsky, O.; Wu, L.-S.; Alasandro, M.
Degradation of a fluoropyridinyl drug in capsule
formulation: degradant identification, proposed degra-
dation mechanism, and formulation optimization. Pharm.
Dev. Technol. 2000, 5 (4), 561– 570.
3. Hartauer, K.J.; Arbuthnot, G.N.; Baertschi, S.W.;
Johnson, R.A.; Luke, W.D.; Pearson, N.G.; Rickard,
E.C.; Tingle, C.A.; Tsang, P.K.S.; Wiens, R.E. Influence
of peroxide impurities in povidone and crospovidone on
the stability of raloxifene hydrochloride in tablets:
 MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
126
identification and control of an oxidative degradation
product. Pharm. Dev. Technol. 2000, 5 (3), 303–310.
4. International Conference on Harmonisation. Stability
Testing of New Drugs and Products (Revised guideline);
Adopted by CPMP, November 2000, issued as
CPMP/ICH/2736/99.
5. Tipre, D.N.; Vavia, P.R. Oxidative degradation study of
nitrendipine using stability indicating, HPLC, HPTLC
and spectrophotometric method. J. Pharm. Biomed. Anal.
2001, 24, 705– 714.
6. Zhao, Z.; Wang, Q.; Tsai, E.W.; Qin, X.-Z.; Ip, D.
Identification of losartan degradates in stressed tablets by
LC-MS and LC-MS/MS. J. Pharm. Biomed. Anal. 1999,
20, 129– 136.
7. Segarra, V.; Carrera, F.; Fabregas, J.-L.; Claramunt, J.
Degradation profile and identification of the major
degradation products of dobupride under several con-
ditions by GC/MS and HPLC-particle beam/MS. J. Pharm.
Biomed. Anal. 1995, 13, 987– 993.
8. Hou, S.; Hindle, M.; Byron, P.R. A stability indicating
HPLC assay method for budenoside. J. Pharm. Biomed.
Anal. 2001, 24, 371– 380.
9. Brightman, K.; Finlay, G.; Jarvis, I.; Knowlton, T.;
Manktelow, C.T. A stability-indicating method for the
determination of melphalan and related impurity content
by gradient HPLC. J. Pharm. Biomed. Anal. 1999, 20,
439– 447.
10. Abdel-Moety, E.M.; Al-Rashood, K.A.; Rauf, A.;
Khattab, N.A. Photostability-indicating HPLC method
for determination of trifluoperazine in bulk form and
pharmaceutical formulations. J. Pharm. Biomed. Anal.
1996, 14, 1639 –1644.
11. Monkhouse, D.C.; Maderich, A. Whither compatibility
testing? Drug Dev. Ind. Pharm. 1989, 15 (13),
2115– 2130.
12. Crowley, P.J. Excipients as stabilizers. Pharm. Sci.
Technol. Today 1999, 2 (6), 237– 243.
13. Kontny, M.J.; Grandolfi, G.P.; Zografi, G. Water vapor
sorption of water-soluble substances: studies of crystal-
Sims et al.
line solids below their critical relative humidities. Pharm.
Res. 1987, 4 (2), 104– 112.
14. Callahan, J.C.; Cleary, G.W.; Elefant, M.; Kaplan, G.;
Kensler, T.; Nash, R.A. Equilibrium moisture content of
pharmaceutical excipients. Drug Dev. Ind. Pharm. 1982,
8, 355.
15. Hancock, B.C.; Shamblin, S.L. Water vapor sorption by
pharmaceutical sugars. Pharm. Sci. Technol. Today 1998,
1 (8), 345–351.
16. Kontny, M.J. Distribution of water in solid pharmaceutical
systems. Drug Dev. Ind. Pharm. 1988, 14 (14),
1991–2027.
17. Clas, S.-D.; Dalton, C.R.; Hancock, B.C. Differential
scanning calorimetry: applications in drug development.
Pharm. Sci. Technol. Today 1999, 2 (8), 311– 320.
18. Giron, D. Contribution of thermal methods and related
techniques to the rational development of pharma-
ceuticals—Part 2. Pharm. Sci. Technol. Today 1998, 1
(6), 262– 268.
19. Phipps, M.A.; Mackin, L.A. Application of isothermal
microcalorimetry in solid state drug development. Pharm.
Sci. Technol. Today 2000, 3 (1), 9 – 17.
20. Mura, P.; Faucci, M.T.; Manderioli, A.; Bramanti, G.;
Ceccarelli, L. Compatibility study between ibuproxam
and pharmaceutical excipients using differential scanning
calorimetry, hot-stage microscopy and scanning electron
microscopy. J. Pharm. Biomed. Anal. 1998, 18 (1,2),
151–163.
21. Serajuddin, A.T.M.; Thakur, A.B.; Ghoshal, R.N.; Fakes,
M.G.; Ranadive, S.A.; Morris, K.R.; Varia, S.A. Selection
of solid dosage form composition through drug-excipient
compatibility testing. J. Pharm. Sci. 1999, 88 (7), 696– 704.
22. Armitage, M.A.; Smith, G.E.; Veal, K.T. A versatile and
cost-effective approach to automated laboratory organic
synthesis. Org. Process Res. Dev. 1999, 3, 189– 195.
23. Sims, J.L.; Roberts, J.K.; Bateman, A.G.; Carreira, J.A.;
Hardy, M.J. An automated workstation for forced
degradation of active pharmaceutical ingredients.
J. Pharm. Sci. 2002, 91 (3), 884– 892.
Copyright of Pharmaceutical Development & Technology is the property of Taylor & Francis Ltd and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.