AAPS PharmSciTech, Vol. 15, No.

1, February 2014 ( # 2013)

DOI: 10.1208/s12249-013-0047-x

Meeting Report

Recent Trends in Product Development and Regulatory Issues on Impurities

in Active Pharmaceutical Ingredient (API) and Drug Products. Part 1: Predicting
Degradation Related Impurities and Impurity Considerations for Pharmaceutical
Dosage Forms

Karen M. Alsante,1 Kim Huynh-Ba,2 Steven W. Baertschi,3 Robert A. Reed,4 Margaret S. Landis,1,10
Mark H. Kleinman,5 Christopher Foti,1 Venkatramana M. Rao,6 Paul Meers,7 Andreas Abend,8
Daniel W. Reynolds,9 and Biren K. Joshi9

Received 6 October 2013; accepted 22 October 2013; published online 27 November 2013

INTRODUCTION areas. Taken together, this comprehensive understanding is

used to achieve a more robust development process that en-
The American Association for Pharmaceutical Scientists ables predictability with a concomitant assurance of safety and
(AAPS) Workshop on Predicting and Monitoring Impurities in efficacy. Innovative methodologies for development of effec-
API and Drug Products: Product Development and Regulatory tive stability control strategies were also presented.
Issues was held on October 13–14, 2012 at the McCormick Place This article summarizes Sessions 1 and 2 of the American
in Chicago, IL, USA. The goal of the workshop was to discuss Association for Pharmaceutical Scientists (AAPS) Workshop
control strategies of chemical and physical changes of active phar- on Predicting and Monitoring Impurities in API and Drug
maceutical ingredients (API) and drug products in the drug devel- Products: Product Development and Regulatory Issues and
opment process. These changes can affect both the safety and addresses of predicting degradation related impurities and
efficacy of drugs; therefore, the ability to rapidly predict and assess impurity considerations for pharmaceutical dosage forms.
the potential for drug product performance changes for impurity Sessions 3 and 4 of the American Association for
formation and the associated safety concerns are important parts Pharmaceutical Scientists (AAPS) Workshop on Predicting
of speeding the development of innovative drug therapies. and Monitoring Impurities in API and Drug Products:
The workshop consisted of four different sessions. Each Product Development and Regulatory Issues are summarized
session focused on separate fundamental issues to build a in Recent Trends in Product Development and Regulatory
comprehensive understanding of the physical and chemical Issues on Impurities in Active Pharmaceutical Ingredient
processes that impact drug degradation, the control of impu- (API) and Drug Products Part 2: Safety Considerations of
rities and the impact of these factors on safety and regulatory Impurities in Pharmaceutical Products and Surveying the
Impurity Landscape published separately.

Dedication This manuscript is dedicated to the memory of the late SESSION I: STRESS TESTING: PREDICTING
Karen Russo, Ph.D. (United States Pharmacopeial Convention) DEGRADATION RELATED IMPURITIES
1
Pfizer Global Research and Development, Groton, Connecticut,
USA. The first session of the workshop consisted of four talks
2
Pharmalytik, Newark, Delaware, USA. specifically focused on various aspects of predicting degrada-
3
Eli Lilly and Company Lilly Research Laboratories, Indianapolis, tion pathways using stress testing with and without in silico/
Indiana, USA. computational input. Prediction of drug degradation (which
4
Celsion Corporation, Lawrenceville, New Jersey, USA.
5 encompasses both pathways and kinetics) is an area with a
GlaxoSmithKline, King of Prussia, Pennsylvania, USA.
6 great deal of interest, especially considering the evolving phar-
Bristol-Myers Squibb, New Brunswick, New Jersey, USA.
7
Rutgers University, New Brunswick, New Jersey, USA. maceutical interest in Quality-by-Design (QbD) approaches.
8
Merck, West Point, Pennsylvania, USA. The f ir st talk was g iven by Mar k Kleinman
9
GlaxoSmithKline, Research Triangle Park, North Carolina, USA. (GlaxoSmithKline) on the topic of using stress testing as a
10
To whom correspondence should be addressed. (e-mail: predictive tool. This talk highlighted the current realities (both
[email protected]) power and limitations) of in silico tools to predict “real world”

1530-9932/14/0100-0198/0 # 2013 American Association of Pharmaceutical Scientists 198

Predicting Degradation Impurities and Impurity Considerations 199

degradation products of drugs and formulated products. The scenario, the actual degradation products are completely pre-
talk also highlighted stress testing conditions that have been dictable and observed in the stress tests—a true subset.
proven to be effective in discovering degradation pathways Realistically, while almost all actual degradation products are
available to a particular drug molecule, with emphasis on the observed in stress tests, there are still a significant number that
more complex areas of oxidation and photodegradation path- are not predicted (Fig. 1).
ways. Finally, Dr. Kleinman discussed practical approaches to In essence, relative to in silico predictions, stress testing is
predictions of the kinetics of drug degradation using empirical a better predictor and yields greater specificity to actual deg-
data from stress testing. radation results. Zeneth® is a relatively new predictive tool
The second talk was given by Chris Foti (Pfizer), focusing from Lhasa Ltd. (developers of METEOR® and DEREK®)
on the analytical considerations when conducting stress testing (2). Currently, one of the main uses of Zeneth is to expand the
studies. This talk provided a comprehensive look at carrying number of hypothetical degradation products for consider-
out a stress testing protocol and the accompanying analytical ation. This allows for researchers to gain a view of the poten-
methodology. A detailed case study illustrated the process and tial degradation products, help to define stress conditions and
provided practical recommendations for interpreting the elucidate the tentative structures by matching mass/charge
results. ratios in mass spectrometry studies.
In the third talk, Steve Baertschi (Lilly) provided an in- The field of pharmaceutical stress testing has recently
depth look at mass balance, including the importance of mass come to the forefront in many areas (3–6). The design of
balance in developing stability-indicating methods and the successful and efficient stress testing studies is difficult due to
hidden complexities involved in assessing mass balance in a the fact that “one size does not fit all”. Even so, a typical set of
formulated product. Particular emphasis was given to the starting conditions for solution stress tests (e.g., acid and base)
importance of understanding the degradation reactions in- are generally accepted. It is noted that oxidative and pho-
volved, the changes in molecular weight (especially when tolytic conditions require special considerations. There are
excipient adducts are formed) and the stoichiometry of deg- a plethora of oxidative mechanisms (e.g., auto oxidation,
radation reactions. peroxidation, electron transfer, and photo-oxidation). A suc-
The fourth and final talk was provided by Karen Alsante cessful oxidative stress test depends on rendering the actual
(Pfizer), looking at recent advances in the understanding of degradation products. Therefore, appropriate oxidative condi-
drug degradation chemistry. In this presentation, the major tions should be utilized. One relatively new condition uses N-
mechanisms of chemical decomposition were examined in the methylpyrrolidinone (NMP) which affords a wide range
context of common functional groups. Of particular interest of oxidative products (7). Similarly, pharmaceutical
was the analysis of the frequency of various molecular weight photodegradation is more difficult to predict due to the
changes resulting from known drug degradation pathways, relative lack of expertise in the area compared to ther-
providing insight into which pathways are most common as mal processes. Photostability has become even more im-
well as those pathways that are rare or complex (e.g., portant to understand due to the relationship between
involving multiple steps). The utility and continuing re- photostability and phototoxicity (4,5,8,9). It is recom-
finement of the chemical degradation prediction software mended that in addition to ICH Q1B confirmatory test-
package Zeneth® (www.lhasalimited.org/products/zeneth.htm) ing with solid API, a stress test with two to five times ICH
was also described. Q1B light exposure is carried out with the solid. In addition, it is
suggested to more fully understand the mechanism of
STRESS TESTING: A PREDICTIVE TOOL photodegradation and the potential for phototoxicity by
(CONTRIBUTED BY MARK KLEINMAN, exposing a solution of API to a light exposure similar to
GLAXOSMITHKLINE) that in the 3T3 Neutral Red Uptake model (10). This will
help evaluate any potential photo-liabilities as well as explore
The stability of organic molecules and small pharmaceu- the link to phototoxicity, assuming that the API absorbs signif-
tical entities follows many rules defined by classic organic icantly (>1,000 M−1 cm−1) at wavelengths greater than 290 nm
reaction mechanisms. Each rule governs a single transforma- (11).
tion. Many small-molecule active pharmaceutical ingredients The other key topic explored is the assessment of kinetic
are inherently complex and have diverse functional groups equivalence. Understanding kinetic equivalence is paramount
that can undergo multiple reactions either simultaneously or to determining how long the stress should be applied to the
in sequence. Thus, the degradation of pharmaceuticals is solid or solution. The rate of degradation depends on the
an area that is complex. The goal is to discuss the possi- energy of activation of the reaction. Using the Arrhenius
bility of predicting degradation products that may form principle for solution or a modified Arrhenius treatment in
from small-molecule organic pharmaceuticals and to offer solids (12), one can show the expected duration of stress
insights into the design of forced degradation (stress testing) relative to 6 months at 40°C for any activation energy (12–
studies. 30 kcal/mol are typically reported for pharmaceuticals (13)). It
The prediction of drug degradation is still in its infancy. is important to consider the kinetic equivalence as it affords a
Hypothetical degradation products are defined as those that scientific rationale to complete a stress study even in the
are predicted in silico or through literature searches. absence of obtaining the recommended 5–20% degradation.
Additionally, potential degradation products are those that In summary, key areas for growth in stress testing are in silico
are observed in stress tests; while actual degradation products predictivity/specificity, oxidative stress testing and a better
are those formed under real-time (e.g., 30°C/65%RH) or ac- mechanistic understanding of the photodegradation of active
celerated stability studies (40°C/75%RH) (1). In an ideal pharmaceutical ingredients.
200 Alsante et al.

Fig. 1. Ideal versus realistic scenarios for degradation product prediction

ANALYTICAL CONSIDERATIONS FOR STRESS extreme pH values, with greater than 4 mg/mL solubility in
TESTING (CONTRIBUTED BY CHRIS FOTI, PFIZER water and acetonitrile/water mixtures. Prediction of degrada-
INC.) tion products using tools such as Lhasa’s in silico Zeneth®
software (16) combined with an assessment based on chemis-
Stress testing plays an important role in the drug devel- try knowledge (in cerebro) can also be an important part of
opment process by providing an understanding of the chemis- data gathering. For this API, the potential sites of reactivity
try of the drug substance and drug product and facilitates the are: (1) oxidation of aryl fused pyrrole, (2) hydrolysis of the
development of stability-indicating analytical methodology. amide, (3) oxidation of benzylic carbon and oxidation of the
Although some recommendations for stress testing are given amine functional groups. Stress testing studies were conducted
in the ICH [Q1A(R2)] and Q1B guidelines (1,14) on the to confirm these predicted stability liabilities. Solid API sam-
stability testing of drug substances and drug products, the ples were subjected to thermal and thermal/humidity and light
guidance given is very general concerning scope and timing exposure (Options 1 and 2). API was stressed in solution using
and is not particularly useful. This paper is a short summary 0.1 N NaOH or 0.1 N HCl conditions, solution-based
from the presentation on Analytical Considerations for Stress azobisisobutyronitrile (AIBN) radical oxidation (to simulate
Testing given as part of the 2012 AAPS Workshop on autoxidation), solution-based peroxide oxidation and solution-
Predicting and Monitoring Impurities in API and Drug based light exposure (Options 1 and 2). These stress testing
Products: Product Development and Regulatory Issues confer- studies were conducted in volumetric glassware to facilitate
ence to provide practical guidance concerning the design, quantitative assessments as needed. Low reactivity was ob-
setup and analytical aspects (i.e., data gathering, sample prep- served for the API under all of the solid-state stress conditions
aration) of carrying out stress testing studies for an API in late (data not shown) in contrast to the solution studies, which
stage development. A specific focus was given to hydrolysis showed degradation of the parent from 2–21% (Table I).
conditions at various pH values, oxidative reagents, and High reactivity was observed for the API in solution
photostress (15). These concepts were illustrated through a under hydrolysis conditions at the extreme pH values using
stress testing case study of the Eli Lilly LY334370 API shown 0.1 N HCl and 0.1 N NaOH. Moderate reactivity was observed
below in Fig. 2. Also, chromatographic method development in solutions containing hydrogen peroxide or a radical initia-
to measure the loss of parent compound as well as the levels of tor, suggesting that LY334370 may be susceptible to oxidative
degradation products or impurities formed under the stress degradation. The API in solution is also very sensitive to light-
conditions was discussed. catalyzed degradation. More details on this case study and
The first part of data gathering is to obtain information guidance for the design and setup of stress testing studies
on solubility and to identify co-solvents that are suitable and can be found in Pharmaceutical Stress Testing (3) and other
compatible with the API through a simple semi-quantitative references (17,18)
screen. It is especially important to select a dissolving solvent There are a variety of separation techniques and detec-
that is compatible with the chromatographic conditions, if tors available to measure potency and purity of stressed sam-
samples are to be injected directly from the stressed solutions. ples. However, within the pharmaceutical industry, the most
Results for the LY334370 API indicate that an acetonitrile co- often utilized methodology is LC with UV detection for quan-
solvent was needed to achieve solution solubility at the titation. Specifically, method development is facile, with selec-
tivity being impacted by column stationary phase, buffer, and
pH and these methods are typically rugged with the equip-
ment being readily available. In addition, the introduction of
ultra high performance liquid chromatography (UHPLC) al-
lows the analytical scientist to optimize speed and resolution.
The degradation conditions in Table I represent the key deg-
radation sample set that was used to evaluate the stability-
indicating nature of the analytical methodology.
This case study provides practical guidance for the design
and execution of an API stress testing with highlighted ana-
lytical considerations. It is also expected that this field will
continue to expand with different ways to conduct API and
drug product stress testing. In the next 3 to 5 years, degrada-
tion prediction could impact the experimental design of stress
Fig. 2. The chemical structure of LY334370 testing studies with both API and drug product.
Predicting Degradation Impurities and Impurity Considerations 201

Table I. Overall Solution Stress Testing Results for LY334370

AIBN radical H2O2 Oxidation 0.1 N HCl 0.1 N Photodegradation Photodegradation

Degradation oxidation(40°C, (room temperature, (70°C, NaOH (70°C, option 1 (1.3× ICH Vis option 2
condition 7 days) 7 days) 3 days) 3 days) and 4× ICH UV) (4.5× ICH Vis)

High reactivity 18% 21% 19%

(>10%)
Moderate reactivity 2% 6% 5%
(1–10%)

Values represent percent of degradation by loss of parent based on peak area percent by LC

REVISITING OLD CONCEPTS: NEW INSIGHTS INTO mass of the lactose adduct detected by the CAD should be
THE CONCEPT OF MASS BALANCE IN DRUG approximately three times that of the pregabalin reacted
PRODUCTS (CONTRIBUTED BY STEVE BAERTSCHI, (481/157∼3). The peak area of the degradation product would
ELI LILLY AND COMPANY) need to be 15%, not 5%, in order to have complete mass
balance. Since only one third of the mass needed for mass
The concept of “mass balance” with regard to stability- balance is being detected, there is a significant mass balance
indicating analytical methodologies is an old topic in pharma- problem, with two thirds of the lost mass from pregabalin unac-
ceutical analysis (19–21), as well as in synthetic chemistry. counted for in the analysis (see Fig. 4b). In order to have
During the process of conducting stability studies, the impor- complete confidence in mass balance calculations, a complete
tant question to address is: Can the analytical method(s) used understanding of the degradation pathways, including corre-
account for the entire parent drug “mass” that was present at the sponding degradation product structures, is required.
initial time point in the stability studies? There are at least seven analytical causes of “mass imbal-
The International Conference on Harmonisation of ance” when using HPLC and there are some practical ways to
Technical Requirements for Registration of Pharmaceuticals investigate or address these issues. The seven causes are
for Human Use (ICH) defines mass balance as “the process outlined below in Table II, along with practical suggestions
of adding together the assay value and levels of degradation or concepts to guide investigations.
products to see how closely these add up to 100% of the initial The topic of mass balance in relation to stability-indicating
value, with due consideration of the margin of analytical pre- analytical methods is underappreciated and can be more com-
cision” (19). In a balanced reaction, the stoichiometry is bal- plex than might be expected. An understanding of both reactant
anced such that the moles of reactants are accounted for in the and products structures is essential to developing a reliable mass
moles of products and the stoichiometry allows direct transla- balance assessment when drug degradation occurs.
tion into mass balance. In the case of drug product degrada-
tion, while the parent drug structure is known and the mass REVIEWING ADVANCES IN CHEMISTRY OF DRUG
(the amount present) can be readily measured, the structure DEGRADATION (CONTRIBUTED BY KAREN
and amounts of the reactants and degradation products typi- ALSANTE, PFIZER INC.)
cally is not known a priori. Common reactants are water,
peroxide, molecular oxygen, excipients or excipient impuri- Stability has long been recognized as critically important in
ties. If a reactant adds to the parent drug, there is an increase the drug development process, affecting both the safety and
in mass associated with that degradation product, which can efficacy of drugs. The ability to rapidly predict and assess the
alter the calculation of mass balance. potential for stability and safety concerns is an important part of
Consider the example illustrated in Fig. 3, when speeding the development of innovative drug therapies.
pregabalin is formulated with lactose as an excipient. The Degradation prediction enables understanding of labile function-
combination may form a condensation degradation product, alities critical in designing less reactive, more stable analogs. With
as is common with amines and reducing sugars, and is repre- efforts to reduce time and cost to market, the potential for stability
sentative of the extensively researched Maillard Reaction issues increases dramatically. Degradation studies conducted by a
(22). Using a detector that responds approximately equally chemistry-guided predictive stability approach enable analysts to
to mass (e.g., the charged aerosol detector or CAD), the deliver stability-indicating methodology more efficiently.
parent drug shows a loss of 5% of the potency value and a In this presentation, the major mechanisms of chemical
degradation product (the drug–excipient adduct) correspond- decomposition of pharmaceuticals in the context of common
ing to a peak area of 5% by area. As shown in Fig. 4a, the functional groups were examined. The major mechanisms of
conclusion would be that mass balance has been achieved chemical decomposition of pharmaceuticals include hydrolysis,
(95%+5%=100% on a mass basis). dehydration, oxidation, isomerization/epimerization, decarbox-
However, once the structure of the degradation product is ylation, dimerization, polymerization, and photolysis and trans-
determined, it is apparent that there is added mass from the formation products involving reaction with excipients/salt forms.
lactose addition to pregabalin. Using the balanced equation in While many pathways of degradation are obvious from basic
the bottom of Fig. 2, a 5% loss of pregabalin should corre- organic chemistry principles, it is not uncommon to find surpris-
spond to 0.157 mg of pregabalin reacted, 0.342 mg of lactose ing degradation chemistry leading to unexpected degradation
consumed and 0.481 mg of lactose adduct formed. Thus, the products and pathways (13,24–27).
202 Alsante et al.

Fig. 3. Hypothetical reaction of pregabalin, a primary amine, with lactose, a reducing sugar
to form the Amadori product (pregabalin-lactose product) and water

The chemistry explored in this presentation was extracted degradation product, and the conditions of the degradation
from actual degradation examples available in an online and publication reference are included in the database. The
structure searchable drug degradation database tool, Pharma database also allows for searching by molecular weight
D3. This chemical structure searchable database was started (MW) change, that is, the difference between the MW of
by Dr. Alsante and Dr. Baertschi, in collaboration with the parent and the degradation products. As this compilation
Cambridgesoft™ (maker of ChemDraw™, Cambridge, MA) in grows, the data should provide a useful tool for the field of
2005. The intent of the database is to be populated with drug degradation chemistry, enabling searches of specific drugs
degradation examples published either in the scientific literature and molecular scaffolds as well as uncovering patterns of
or presented at scientific conferences. The database allows struc- degradation of specific functional groups and of drugs in
ture and name-based searching of the parent drug or the general.

a b
Fig. 4. a Apparent mass balance in the case of a 5% loss of parent and a 5% increase in degradation product peak
area, using a charged aerosol detector (CAD). b Based on the added mass from lactose in the lactose adduct, a true
mass balance using the charged aerosol detector would require a total peak area corresponding to 110% of the peak
area of the initial (undegraded) pregabalin timepoint
Predicting Degradation Impurities and Impurity Considerations 203

Table II. Common Analytical Causes of Mass Imbalance and Suggested Actions

Potential analytical cause of mass imbalance Practical suggestions to guide mass imbalance investigations

Impurities are not eluted from the HPLC • Use gradient HPLC with wide polarity range and longer hold time with strongest mobile
column phase condition
• Analyze sample using reverse phase or normal-phase thin-layer chromatography (TLC)
• Analyze sample(s) using hydrophilic interaction liquid chromatography (HILIC) or
normal-phase HPLC
• Analyze sample(s) using capillary electrophoresis (CE)
Impurities are poorly separated and are • Use gradient HPLC with a steep gradient
“missed”(20) • Use isocratic HPLC with very strong mobile phase condition, with and without a column
in place
• Use UV spectrophotometry without any analytical separation
Impurities are co-eluting with the parent • Change HPLC stationary phases, solvent, or gradient
compound • Use an orthogonal (different) separation method
• Look for peak purity using a PDA-UV detector with a UV homogeneity algorithm
• Look for peak purity using LC/MS techniques
Impurities are not detected by the detector • Use PDA-UV (200–400 nm) detection to increase “universality”; monitor at low wavelength
(e.g., 205–210 nm) (20)
• Use UV-transparent solvents and buffers
• Consider alternative detector e.g., evaporative light scattering detector (ELSD), mass
spectrometry (MS), chemiluminescence nitrogen detector (CLND), corona, charged aerosol
detector (CAD), refractive index or flame ionization detector (FID)
• Analyze sample using alternate/orthogonal detection method
• Use reverse phase (RP) or normal-phase (NP) thin-layer chromatography (TLC)
–Use different options / chemistries for developing TLC spots or fluorescent-impregnated
TLC plates
• Use CE (often can look as low as 190 nm, more universal wavelength)
There is poor analytical recovery of the • Consider insolubility of impurities in analytical phases
impurities • Careful visual observation
There is poor analytical recovery of the parent • Consider different solvents for sample preparation
• Isolate solid material and analyze using other technique (e.g., probe-MS)
• Consider possibility of reactions with insoluble excipients
• Considered volatility (20)
• Consider other analytical techniques (e.g., GC-headspace)
• Consider adsorption losses
• Compare results using different containers (e.g., glass and polypropylene)
• Change sample/extraction solvent (e.g., different pH, different solvent)
• Consider possibility of instability during the analytical preparation or workup
There is inaccurate quantification due to • Examine UV spectra of detected impurities (PDA detector)
differences in response factors. • Consider alternative detector—evaporative light scattering, MS, Corona CAD,
chemiluminescence nitrogen detector (CLND), FID or LC/NMR
• Determine response factors (20,21)
• Isolate, purify, and determine using conventional means
• Use CLND(23) or CAD (without isolation of impurities)
• Use quantitative nuclear magnetic resonance (qNMR) (20)

Mass balance can be measured and expressed in a variety of ways, but the concepts of Absolute and Relative Mass Balance has been advanced
and discussed in detail by Nussbaum et al. (23). Absolute mass balance deficits (AMBD) can be expressed as the difference between the mass of
parent drug consumed and mass of the parent contained in the degradation products recovered. Relative mass balance deficit (RMBD) can be
expressed as the AMBD divided by the mass of the parent consumed and is expressed as a percentage

As an example of the power of such an exercise, a search changes are not surprising to the seasoned degradation
was conducted to determine the most common/frequent deg- scientist.
radation pathways as a function of changes in MW from the For example, changes in the MW of +16 and +32 amu
parent to the degradation products. Thus, the database (28) occur frequently, corresponding to the addition of 1 and 2
was searched for all examples of degradation products that oxygen atoms, respectively. Likewise, a change in the MW
show MW changes from the parent of −60 to +60 amu, in of +18 or −18 amu can readily be explained by the addition or
1 amu increments (29). The results of this effort are captured loss of water. The most common instance is where there is
in Fig. 5, where the number of degradation products is plotted no MW alteration, net change 0 amu, and this instance is
versus MW change. The plot shown reveals that there are represented by more than 60 examples found in the data-
patterns in degradation pathways, with certain MW changes base, resulting from epimerization and rearrangements.
occurring more frequently than others and many of the MW The presentation then transitioned to understanding the
204 Alsante et al.

excipients, degradation of the API via reactions with excipi-

ents and impurities in excipients as a few examples. This is a
critical area of importance in the pharmaceutical industry as
scientists must develop products that have acceptable chemi-
cal and physical stability over the course of the product shelf-
life. It is quite common that the API will show stability issues
when blended with excipients. There are often significant
challenges with surprise degradation chemistry due to un-
known impurities in excipients that cause the formation of
degradants at levels of concern as per ICH Guidelines for
Drug Products (ICH Q3B).
Margaret Landis (Pfizer) kicked off the session with the
fundamentals and latest approaches on excipient compatibility
testing. This is a critical area of understanding as knowledge of
reactivity changes when the API is mixed with excipients is
critical to formulation development. Since gathering real-time
Fig. 5. A plot of the frequency of specific MW changes from parent
stability data on pharmaceutical dosage forms is impractical,
observed in degradation products contained in the Pharma D3 data-
base in 2009. The height of the specific bar graphs indicate the most the accelerated testing outlined is important for the under-
frequent MW changes from parent as a result of drug degradation standing the stability of the drug product and for development
processes of stability-indicating methodology.
Venkatramana Rao (Bristol-Myers Squibb) continued
the focus on reactive impurities in pharmaceutical excipients
and their impact on product robustness. These impurities
include aldehydes/reducing sugars, peroxides, nitrates, ni-
impact of Zeneth® in silico software. Zeneth® is the only
trites, metals and solvents present at trace yet variable levels.
commercially available program designed to predict degra-
However, control strategies using compendial methods do
dation pathways of pharmaceutical compounds and is a key
not appear feasible. A risk-based strategy involving early
tool to aid in degradation protocol design and structure
identification of incompatibilities between reactive excipient
elucidation. Zeneth® uses a high quality knowledge base
impurities and the drug followed by designing a drug product
and reasoning engine (16) to produce detailed tree depic-
that can withstand the variability in excipients is provided as
tions, showing chemical degradation pathways that include
guidance.
a level of likelihood, chemical formula, exact mass, and
The session transitioned to a discussion of complex dos-
degradation pathway description.
age formulations, including nanoscale technology by Paul
The presentation concluded with Zeneth® benchmarking
Meers (Rutgers University). Liposomes are one of the first
studies based on disguised industry examples. Zeneth® pre-
“nanoscale” technologies used in drug delivery and have be-
diction captured to date demonstrated we are progressing in
come an important formulation option in the field as lipo-
our ability to predict degradation products with this in silico
somes can aid targeting and optimizing the pharmacokinetic
tool. However, the demonstration revealed there is room for
profile.
improvement, especially with respect to excipient-related deg-
The session closed with a presentation on combination
radation prediction capabilities. Obviously, there is still a need
therapies by Dan Reynolds (GlaxoSmithKline). Combination
for wet chemistry in degradation related studies, but the
products constitute dosage forms that contain more than one
power of prediction to aid in focusing stress testing pro-
API. Since reactions between the APIs becomes of particular
tocols was clearly demonstrated. While many pathways of
concern with these products, these reactions should be inves-
degradation are obvious from basic organic chemistry
tigated during product development. Most combination prod-
principles, it is not uncommon to find surprising degrada-
ucts contain APIs that are already in existing marketed
tion chemistry leading to unexpected degradation products
products. Consequently, the degradation of the individual
and pathways. As the field of degradation chemistry ma-
APIs is typically well understood but the interactions between
tures, and degradation pathways of various drug scaffolds,
the APIs represent potential novel drug–excipient and drug–
ring systems, and functional group arrangements are doc-
drug reactions that require further study.
umented in the literature, the predictability of degradation
pathways will dramatically improve and patterns will begin
to emerge. The concluding life cycle schematic (Fig. 6) SOLID STATE EXCIPIENT COMPATIBILITY TESTING
depicts the evolution of the degradation workflow lifecycle (CONTRIBUTED BY MARGARET LANDIS, PFIZER
and where we will continue to grow in our degradation INC.)
knowledge.
Excipient compatibility is a broad term utilized for a
SESSION 2: IMPURITY CONSIDERATIONS variety of studies at many different stages of drug develop-
FOR PHARMACEUTICAL DOSAGE FORMS ment. Compatibility screening with excipients can encompass
early, simple studies aimed at evaluating compatibility of a
The next session of the workshop transitioned to the drug with excipient to support initial biological efficacy stud-
pharmaceutical dosage form space expanding the complexity ies, screening to ensure compatible and robust formulations
beyond impurities in the API alone to impurities in the for pre-clinical toxicology, short term compatibility studies to
Predicting Degradation Impurities and Impurity Considerations 205

Fig. 6. Degradation workflow lifecycle to improve process and incorporate lessons learned

support early clinical dosage forms for Phase 1 or more com- The four core aspects of excipient compatibility studies
plex compatibility studies needed to support larger, longer include sample design and preparation, sample composition,
clinical studies, such as Phase 2A/2B and Phase 3 clinical for- storage and stress conditions and methods of stability analysis.
mulations. Extensive and definitive compatibility studies are Sample design and preparation include aspects such as levels of
needed to support commercial formulation development and API loading, blend preparation, compaction options, use and
product launch. Finally, additional, specialized compatibility extent of mechanical aggravation utilized. These parameters can
studies may be initialed to investigate alternative dosage formu- be chosen as conditions that represent a realistic state of the
lations, such as controlled release and combination products. dosage form or the most challenged state (i.e., a worst-case
Design of excipient compatibility studies will be unique to scenario). Sample composition aspects include choices between
each molecule and drug therapy in question. It should be the use of binary versus multi-component compatibility samples,
carefully considered how the excipient compatibility data will spiking of samples with reactive impurities or water to induce or
be used and clearly define the scope and limits of the compat- accelerate degradation, the use of statistical design or design of
ibility information that will be gained from each study. experiments for sample composition and the use of various
Different stages of dosage form development will require forms of the drug (salts, free forms, polymorphic, or amorphous
varying size and types of compatibility studies. Prior to initia- forms). Storage and stress conditions include the use of real-time
tion of any compatibility studies, a thorough review of rele- or accelerated stress conditions involving humidity, heat and
vant drug substance information available at the time of the light stress conditions, the incorporation of open or special
compatibility studies is extremely important. This evaluation packaging (use of desiccants, blister packing, etc.) and the time
should include a review of structural understanding of the course of the study. Methods of stability analysis need to address
molecular scaffold and sites of known reactivity, a detailed both physical and chemical stability during compatibility studies.
review of the synthetic route, a review of the API solution state Aspects to consider include the employment of destructive and/
stability data (pH, thermal, and photostability challenges), any or non-destructive analysis, the choice of most applicable spec-
and all forced degradation data, metabolite formation informa- troscopic techniques, the number of samples needed per sam-
tion and output from predictive models of degradation pling timepoint and sensitivity of the methods employed. The
(Zeneth®, etc.). stability data gathered during compatibility analysis may be used
In addition, a strong recommendation for compatibility to dictate optimal excipient composition of a dosage form or
studies includes generating a pre-compatibility profile of the generate stability–time profiles. Stability information gathered
drug candidate and the excipients. A detailed list of potential from compatibility studies serves to confirm real-time stability of
parameters to be evaluated and tracked is described in a dosage form or to predict longer term shelf-life performance,
Table III below. such as via the use of the isoconversion-based accelerated sta-
The lot-specific parameters listed above should be consid- bility assessment paradigm (ASAP) (36).
ered and potentially evaluated for every drug substance and New frontiers for the science of excipient compatibility
excipient evaluated in compatibility studies, especially the as- testing focus on the ability to acquire accurate compatibility
sessment of the levels of potentially reactive impurities present information faster, utilizing less materials, experimentation
in common tableting excipients. The reactive impurities, includ- and resources. Technological advances in the area will focus
ing peroxides (31,32) and aldehydes (33,34), are known to cause on the efficient automation of compatibility studies, the ability
chemical instability in dosages forms (35) to predict and model degradation of drug substances in
206 Alsante et al.

Table III. Important API and Excipient Attributes Relevant to Excipient Compatibility Studies

Important API Attributes relevant to Excipient Important Excipient Attributes relevant to Excipient
Compatibility Studies Compatibility Studies

API Impurity Profile (lot specific): water, solvent, metals and Excipient impurity profile (lot specific): water, solvent, metals,
amorphous content, acidic/basic impurities, process-related amorphous content, acidic/basic impurities, reactive impurities
impurities, alternate forms of the API present in small (peroxides, aldehydes, organic acids)
quantities (free forms or higher energy polymorphic forms)
Thermal and thermal/humidity solid-state stability Thermal and thermal/humidity solid-state stability (chemical
(chemical and physical) and physical)
Hygroscopicity profile Equilibrium moisture content and hygroscopicity profile (1)
Particle attributes: size, shape, distribution, surface area Particle attributes: size, shape, distribution, surface area
Effects of mechanical aggravation and processing Effects of mechanical aggravation and processing
Crystal packing information (actual or predicted) Details of excipient manufacture and processing
Effective pH in water (30)
Spectroscopic properties
Age and storage history of excipients (lot specific)

pharmaceutical dosage forms, the development of more effi- degradation pathways the drug candidate can undergo. This
cient analytical techniques that can evaluate very low level may be accomplished with prior knowledge, predictive model-
physical and chemical changes of the API in excipient com- ing tools or experimental studies. Knowledge of reactive im-
patibility matrices. Future challenges for the field will come purities that may be present in the specific excipients that are
from the emergence of formidable new chemical entities, such being considered in the product design is also essential.
as complex drug conjugates (i.e., antibody–drug conjugates, The risk assessment approach involves combining the
ADCs) and biopharmaceuticals. These systems will require knowledge of reactive impurities in excipients along with an
new technology and paradigms for assessing compatibility in understanding of drug degradation pathways. Other factors
pharmaceutical dosage forms. such as drug to excipient ratio, crystal form of the API, envi-
ronmental conditions, surface acidity or microenvironmental
pH must also be considered during the assessment and miti-
REACTIVE IMPURITIES IN PHARMACEUTICAL
gation of risk. The mitigation strategies include “designing
EXCIPIENTS AND THEIR IMPACT ON PRODUCT
out” the incompatibilities through formulation design, pack-
ROBUSTNESS (CONTRIBUTED BY VENKATRAMANA
aging configurations or putting in place a control strategy
M. RAO, BRISTOL-MYERS SQUIBB COMPANY)
beyond compendial testing. Setting specific acceptance criteria
for the excipients on a particular dosage form requires a
A robust drug product must accommodate typical varia-
strong and transparent relationship between excipient vendor
tions in raw materials, i.e. API and excipients, operational
and the drug product manufacturer. Arriving at the specifica-
elements (processing, equipment, etc.) and ambient condi-
tion limits also requires an understanding of quantitative re-
tions.(37) Excipients that directly react with drugs are elimi-
lationship between levels of impurities in the excipients and
nated during the excipient compatibility or stability studies,
product stability. This may be challenging as procuring sam-
i.e. prior to the drug product design. That is why a large
ples of excipients with varied levels of impurities may require
portion of reported excipient incompatibilities with drugs is
the excipient vendor to produce batches that are not “typical”
due to reactions between drugs and “reactive” components
and spiking of the “reactive impurity” into the excipient by the
within excipients (38). The main challenge for product robust-
end-user may not be practical. Early identification of incom-
ness arises because these impurities are at trace yet variable
patibilities between reactive excipient impurities and the drug
levels and a control strategy based solely on compendial re-
followed by designing a drug product that can withstand the
quirements (USP/NF, Ph.Eur., J.P.) is often not adequate.
variability in excipients is the best approach to avoid undesir-
Additionally, very little information is available about these
able surprises in subsequent stages of development and com-
“reactive impurities” as manufacturing processes of excipients
mercialization of the drug product.
are trade secrets and not public information. Some of the most
common reactive impurities include aldehydes/reducing
sugars, peroxides, nitrates, nitrites, metals and solvents. LIPOSOMAL DEGRADATION (CONTRIBUTED
These impurities could be introduced during the manufactur- BY PAUL MEERS, RUTGERS UNIVERSITY)
ing processes of excipients or generated during storage or use.
The analytical methods to quantify these trace level impurities Liposomes represent one of the first nanoscale technologies
need to be sensitive to detect such low levels and selective to used in drug delivery and have become an increasingly important
differentiate between different “forms” of the impurities such formulation choice to address a number of pharmaceutical
as hydroperoxide versus hydrogen peroxide, etc. as the reac- needs. As nanoscale to microscale supramolecular assemblies
tivity between drugs and these species may vary. of lipids that encapsulate or associate with an active pharmaceu-
An approach to assess and mitigate the risks posed by the tical ingredient, liposomes can aid targeting and optimize the
interactions between the impurities in excipients and drugs is pharmacokinetic profile. Truly liposomal formulations comprise
proposed to be able to predict and/or determine the potential one or more lipid bilayers that completely enclose or encapsulate
Predicting Degradation Impurities and Impurity Considerations 207

an aqueous space (Fig. 6). Phospholipids, cholesterol, polymer- phospholipid acyl chains directly affects parameters such as
grafted lipids and cationic lipids are the major structural compo- the leakage of encapsulated substances (45), the liposomal
nents of many liposomal formulations in development, and ap- surface charge can affect the rate of these hydrolytic processes
propriate stress conditions and tests for chemical changes need to under certain conditions (46).
be identified (Fig. 7). Design of pharmaceutical liposomes should be guided by
The primary degradation products for the many common some of the known degradation pathways to yield robust,
liposomal components result from well-known reactions. For stable formulations that can be easily characterized physically
example, hydrolytic degradation products of phospholipids and chemically. The impetus to apply liposomal strategies to
often include lyso-lipids that are produced by the cleavage of more complex delivery problems will lead to more challenges
an acyl chain at the ester linkage of the sn-2 position, which is in performing appropriate stress tests. New chemical stress
catalyzed by acid or base (39,40). Oxidation products can test protocols have become necessary as non-phospholipid
occur near the double bonds of unsaturated or especially constituents such as cationic lipids have become more preva-
polyunsaturated phospholipid acyl chains, or the 7-carbon lent, particularly for delivery of novel biopharmaceuticals
position of cholesterol (41) (Fig. 1). Standard stress protocols (47). Furthermore, new devices for alternative routes of admin-
include non-neutral pH and/or heat for hydrolysis, and free istration may have important effects on the physical parameters
radical initiators and propagators for oxidation (42). of the formulation. For instance, the development of liposomes
Because liposomes are particulate, it is the physical char- for inhalation leads to unique parameters for stress testing in-
acteristics of the assembly that primarily determine the phar- volving analysis of the effects of nebulization on the aerosol
macological delivery activity. In this sense, “impurities” distribution of resulting physical degradation products (48).
include not only the chemical degradation products of the Addressing stress protocols for any particular formulation
constituent lipids, but also alternate physical organizations of will require an informed and specific conglomeration of appro-
these lipids. Regulatory guidance therefore strongly recom- priate chemical and physical tests. The knowledge base in this
mends characterization of a number of physical properties of relatively new field continues to develop and can be expected to
the liposomes that include surface charge, “leakiness”, size, significantly improve over the coming years as more drug deliv-
and lamellarity (number of bilayers)(43) (Table IV). ery issues are addressed by products utilizing liposomes.
Importantly, the relevant chemical degradation products
can affect the physical parameters of the liposomal assembly PRACTICAL ASPECTS OF STRESS TESTING
and vice versa. The tendency of lipids to organize in bilayers is ON SMALL MOLECULE PARENTERAL PRODUCTS
highly dependent on their detailed chemistry, and even small (CONTRIBUTED BY ANDREAS ABEND, MERCK
chemical changes in the lipids can cause a change in the & CO., INC.)
liposomal physical properties. Conversely, the physical and
structural properties of the liposome can dictate the stability Small molecule injectables are typically administered to
of the lipid components. For instance, while hydrolysis of patients by a physician and they are available as sterile

1,2-dipalmitoyl-sn-
Liposome glycero-3-phosphocholine
Cutaway view (DPPC)
bilayer
H3 C
+ CH3
H3 C N

O
O P O
O

O Hydrolysis
O
O target
O
Typical
Components
Phospholipids Oxidation
Cholesterol target
Other polar lipids (if there are
conjugated
double
bonds)
sn-1
sn-2

Fig. 7. Schematic representation of a liposome. A liposome composed of the lipid 1,2-dipalmitoyl-

sn-glycero-3-phosphocholine (DPPC) with a diameter of approximately 100 nm and a membrane
thickness (phosphorus to phosphorus) of approximately 3.8 nm is shown. Top inset space filling
models of DPPC bilayer arranged in a typical bilayer organization. Far right inset chemical structure
of DPPC with stereo-specific numbering (sn) for the glycerol backbone attachment of acyl chains.
Arrows show site of possible hydrolysis and oxidation sites
208 Alsante et al.

Table IV. Methods to Test Important Liposome Physical Properties

Test parameter/(relevance) Analytical methods Stress methods

Captured volume • Electron Spin Resonance (ESR) with Chemical degradation; heat, detergents,
(encapsulation efficiency, CAT1 spin probe (volume exclusion method) (44) osmolarity change
release rate) • Encapsulation of water soluble probe
Drug encapsulation • Field flow fractionation Chemical degradation; heat (phase transition
(encapsulation efficiency, dose) • Filtration temps), detergents, mechanical stress,
• Size exclusion chromatography sonication, osmolarity change
• Sedimentation
• Fluorescent probes
• NMR probes
Liposome size (tissue targeting) • Photon correlation spectroscopy (PCS) Chemical degradation; heat, detergents,
• Size exclusion chromatography mechanical stress, sonication, osmolarity
• Field flow fractionation change, ionic milieu
• Freeze fracture electron microscopy
• Cryo-electron microscopy
Lamellarity (release rate) • 31P NMR Chemical degradation; heat, detergents,
• PCS w/captured volume mechanical stress, sonication
• Fluorescent probes
• Chemical surface labeling
Surface charge (tissue targeting) • Electrophoretic mobility Chemical degradation; ionic milieu

powders lyophilized for injection or as a sterile liquid for stability studies with stoppers dried over a period of time and
injection. Compared to solid oral dosage forms, these formu- then placed on product vials can be used to gauge stopper
lations demand additional understanding of their stability drying effectiveness with much higher sensitivity. The relative
since they may require reconstitution or dilution with a variety humidity inside the stoppered vials can be measured by field
of diluents prior to their use. The product may be adminis- modulated IR spectroscopy (51). Assessing the headspace
tered directly with a syringe or injected into a drip-bag and moisture in vials with product as function of stopper drying
slowly infused into the vein of a patient. In addition, extract- time has shown to be predictive of physical stability of the
able and leachable studies are necessary to ensure no impuri- lyophilized product, whereas KF measurements on stoppers
ties are introduced into the product upon storage or during the alone was not (52). Once the critical headspace moisture in
final patient administration process (49). the vial has been established, these experiments can be used
Lyophilization, or “freeze-drying”, is the process of mak- for example to evaluate the effectiveness of the stopper drying
ing sterile formulations for injection. This formulation ap- process during process scale-up development. A disadvantage
proach is used to stabilize drugs that are otherwise sensitive of the headspace moisture experiments is the time it takes for
to hydrolysis. The impact of moisture on the stability of the system to reach equilibrium. On the other hand, these
freeze-dried products was presented in two case studies experiments are still short in comparison to performing long
(Cases 1 and 2). term studies each time the stopper drying process is changed
The formulation presented in Case 1 showed hydrolysis or the stopper composition or vendor is changed.
of the drug substance stemming from residual moisture in the Case studies 3 and 4 discussed the impact of light on
freeze-dried product upon storage. The chemical nature of the products in solution. ICH (Q1B) photostability studies are
degradation product was not a concern, but its limited solu- usually part of the battery of stress testing performed during
bility in commonly used diluents and the levels anticipated formulation development and part of the stability data pro-
based on the available kinetic data at the end of the proposed vided for regulatory submissions. Case study 3 showed that for
product shelf-life may have posed a patient safety risk. The a drug substance that is stable when exposed to light, the
rate of hydrolysis was ultimately controlled by the excipients, sensitivity of the formulation towards photodegradation in-
the residual cake moisture at the end of the lyophilization creased over time (53). The root cause of this increased
process and the storage temperature. photodegradation was the amount of Fe3+ ions that had
The second case study highlighted a physical stability risk leached from the glass vial. The increased amounts of iron
in a freeze-dried formulation. Here, controlling the amount of ions leaching into the product was promoted by the presence
water released from the rubber closure turned out to be a of chelators in the formulation. Most soluble organic Fe3+
significant risk during product development. Karl Fischer chelated complexes absorb ultraviolet-A and visible light.
(KF) titration is normally the method of choice for determi- The actual formulation may not show a noticeable UV–
nation of moisture levels in stoppers because of its simplicity Visible light absorption profile due to the very low levels of
and fast turnaround. KF is usually sensitive and suitable these metal complexes. However, the presence of dissolved
enough for the development of a stopper drying process for oxygen and exposure to light give rise to formation of
most lyophilized products. However, this method may not be hydrogen peroxide and Fe2+(54). The hydrogen peroxide is
sensitive enough to detect small differences in residual stopper subsequently reduced by Fe2+ ions to hydroxyl radicals (55).
moisture that may cause unacceptable changes upon storage The hydroxyl radicals then account for the observed
in products with a relatively small mass (50). Relatively short degradation when the product is exposed to light.
Predicting Degradation Impurities and Impurity Considerations 209

Finally, Case 4 discussed the assessment of photodegradation of stability-indicating methods (SIMs), adequate packaging and
during patient treatment, or in-use photodegradation. For drug shelf-life, etc.
products that are considered photosensitive (based on the A survey of the literature prior to 2002 showed that most
photostability assessment outlined in ICH Q1B), the risk of un- investigators developing SIMs for combination therapies did
acceptable degradation during product administration is consid- not investigate the possibility of drug–drug reactions.
ered very low. This is due to the fact that the time of actual However, a survey in 2010 showed the possibility of such
exposure to light is very short in comparison to the photostress drug–drug reactions were being taken into account by a ma-
study. Case 4 highlighted the level of additional product under- jority albeit in various and inconsistent ways.
standing one might consider when developing a highly light sen- Little has been reported in the literature concerning the
sitive product especially with respect to patient in-use. The drug chemistry of drug–drug reactions in combination therapies.
product can be routinely manufactured with virtually no degra- Examples are illustrated in Schemes 1, 2, and 3 below and
dation products and product storage in secondary package com- include: the effect of Timolol Maleate on the stability of
fortably supports a 2-year shelf-life. Once removed from the pilocarpine (56,57) (Scheme 1), the reaction of isoniazid with
protective secondary package and reconstituted, the product a degradation product of rifampicin (58) (Scheme 2) and
slowly undergoes photodegradation inside the primary package. transacetylation reactions of aspirin (59) with codeine and
The main issue with the drug is that when exposed to ambient sulfadiazine (Scheme 3).
light in a normal clear glass syringe, degradation is fast, even if the GlaxoSmithKline (GSK) developed a triple combination
drug is administered right away. Experiments to gauge the extent tablet of the HIV therapies abacavir sulfate, lamivudine, and
of photodegradation included performing light measurements in zidovudine. Prior to the NDA submission, the FDA requested
hospital emergency rooms under various scenarios under which in a letter that GSK stress the three APIs together in solution
the drug may be administered. With an understanding of the (acid/base, oxidation) and the solid-state (heat, heat/humidity,
conditions under which the drug would be administered, studies light). GSK complied with that request and also stressed the
were performed to assess the amount of degradation during the product tablets. The results of the degradation studies were
administration process. These studies showed that the drug prod- reported to the FDA; no questions concerning degradation
uct can be safely administered to patients but requires clear chemistry were asked. This approach has also been used for
instructions about the exposure time and tight measures to pro- Advair (fluticasone propionate, salmeterol xinafoate) MDI,
tect the product from direct light exposure. Combivir (lamivudine, zidovudine) tablets, and Treximet
(naproxen sodium, sumatriptan) tablets. These studies were
STRESS TESTING OF COMBINATION THERAPIES also reported in the respective NDAs without questions from
(CONTRIBUTED BY DAN REYNOLDS, regulators.
GLAXOSMITHKLINE) A recommended experimental approach to combination
therapies includes stressing the combined APIs (1:1 mole
Combination therapies contain more than one API. Several ratio) in 0.1 N HCl; NMP/water (7) under N2, air, and O2; in
regulatory entities including International Conference on 0.1 N NaOH; in the solid-state with ambient and 75% RH; and
Harmonisation of Technical Requirements for Registration of in the solid-state with >2× ICH light storage conditions. All
Pharmaceuticals for Human Use (ICH), Food and Drug samples (except light storage) may be stored at 60°C for
Administration (FDA), and the World Health Organization 7 weeks and/or 80°C for 2 weeks (kinetic equivalent) or until
(WHO) maintain that the possibility of reactions between APIs the most labile API has degraded 10%, whichever comes first.
in combination therapies should be investigated during product This protocol does not apply to combination therapies where
development. Most combination therapies contain APIs that are the APIs are not in physical contact with each other. In that
already in existing marketed products. Consequently, what usu- case, drug–drug reactions are not considered an issue.
ally remains to be understood are potential novel drug–excipient Results of API degradation studies should be reported in
and drug–drug reactions. Associated concerns are development section 3.2.S.7.3 of the Common Technical Document (CTD)

N N
O
O HOOC
O N HO N N N
Pilocarpine Pilocarpic Acid S
N O N
H
O O OH
HO OH

N N Timolol Maleate
HOOC
O O N HO N
Isopilocarpine Isopilocarpic Acid
Scheme 1. Degradation pathways of pilocarpine accelerated in the presence of timolol maleate
210 Alsante et al.

R
N N N

Rifampicin

H+/H2O

R
O + H2 N N N
3-Formyl Rifampicin

H
O N NH2

HO
N
Isoniazid O OH O
O OH OH
H R NH
O N N R=

O
O OH
N O
Scheme 2. Degradation of combination therapy containing rifampicin and isoniazid

while drug product studies are reported in section 3.2.P.8.3. each significant degradation product (conditions, mechanism),
Suggested contents for the API module include a description and a summary of peak homogeneity experiments on each
of stress conditions, a scheme of the degradation pathways for API. It is recommended to dismiss any insignificant peaks
each API, quantitative results (table format) for solution and observed in stress studies that are below Q3A identification
solid-state samples (mass balance), results of chiral testing thresholds in formal stability studies. The same information is
results (may refer to previous studies), chromatograms from suggested for the drug product module plus the description of
HPLC testing on key samples, a discussion of the formation of the formulation and discernment between drug and excipient-

O O O

OH
O + O
O
N O N
O
HO O
codeine aspirin acetylcodeine

O
O N O
OH O N
+ H2 N S N
H N S N
O O N H H
O N
O Sulfadiazine acetylsulfadiazine
aspirin
Scheme 3. Transesterification degradation reactions of aspirin in combination products with codeine and sulfadiazine
Predicting Degradation Impurities and Impurity Considerations 211

related peaks, as Q3B also dictates identification thresholds Use (ICH). Photosafety Evaluation of Pharmaceuticals S10:
for drug products. When the degradation products observed in Draft ICH Consensus Guideline. ICH Harmonized Tripartite
Guidelines. 2013.
the drug product on stability are the same as seen in stress 6. International Conference on Harmonisation of Technical
studies of the APIs, it may not be necessary to file a drug Requirements for Registration of Pharmaceuticals for Human
product stress testing module (60). Use (ICH). Guideline on assessment and control of dna reactive
(mutagenic) impurities in pharmaceuticals to limit potential car-
CONCLUSION cinogenic risk M7: Draft Consensus Guideline. ICH Harmonized
Tripartite Guidelines. 2013.
7. Reynolds DW, Galvani M, Hicks SR, Joshi BJ, Kennedy-Gabb
The conference sessions summarized in this white paper SA, Kleinman MH, et al. The use of N-methylpyrrolidone as a
(Part 1 of 2) covered a variety of important aspects involved in cosolvent and oxidant in pharmaceutical stress testing. J Pharm
the study of drug degradation, both as the drug substance and Sci. 2012;101:761–76.
in the drug product. Important advances in technology to 8. Kleinman MH, Smith MD, Kurali E, Kleinpeter S, Jiang K,
Zhang Y, et al. An evaluation of chemical photoreactivity and
predict drug degradation were investigated and reviewed. the relationship to phototoxicity. Regul Toxicol Pharmacol.
Focus was drawn to specific analytical and mass balance con- 2010;58:224–32.
siderations involved in conducting successful and informative 9. Onoue S, Hosoi K, Wakuri S, Iwase Y, Yamamoto T, Matsuoka
forced degradation studies of drug substances. Investigation of N, et al. Establishment and intra-/inter-laboratory validation of a
standard protocol of reactive oxygen species assay for chemical
degradation in pharmaceutical dosage forms was shown to
photosafety evaluation. J Appl Toxicol. 2012;13(10).
include a consideration of both the strategic design of solid- 10. Spielmann H, Balls M, Brand M, Doring B, Holzhutter HG,
state compatibility studies and the need to understand reactive Kalweit S, et al. EEC/COLIPA project on in vitro phototoxicity
impurities present in pharmaceutical excipients. Stress testing testing: first results obtained with a Balb/c 3T3 cell phototoxicity
of liposomal and small-molecule parenteral products requires assay. Toxicol In Vitro. 1994;8(4):793–6.
11. Henry B, Foti C, Alsante K. Can light absorption and
in-depth understanding of these complex systems and need to photostability data be used to assess the photosafety risks in
be approached with concerns of both chemical and physical patients for a new drug molecule? J Photochem Photobiol B.
stability. Finally, a timely discussion of recommendations for 2009;96(1):57–62.
stress testing of combination drug products was presented. 12. Waterman KC, Carella AJ, Gumkowski MJ, Lukulay P,
MacDonald BC, Roy MC, et al. Improved protocol and data
Overall, the sessions served to highlight the importance of
analysis for accelerated shelf-life estimation of solid dosage
having the ability to rapidly predict and assess the potential forms. Pharm Res. 2007;24(4):780–90.
for impurity formation in drug products, which can lead to 13. Baertschi SW, Jansen PJ, Alsante KM. Stress testing: a predictive
performance, regulatory, efficacy, and safety concerns. tool (Chapter 2). In: Baertschi SW, Alsante KM, Reed RA,
editors. Pharmaceutical Stress Testing: Predicting Drug
Degradation: Informa Life Sciences; 2011. 10-48.
ACKNOWLEDGMENTS 14. Baertschi SW, Alsante KM, Tonnesen HH. A critical assessment
of the ICH guideline on photostability testing of new drug sub-
The organizing committee would like to thank all the stances and products (Q1B): Recommendation for revision. J
Pharm Sci. 2010;99(7):2934–40.
speakers for sharing their work at this conference and through 15. Jansen PJ, Smith WK, Baertschi SW. Stress testing: Analytical
this publication. considerations (Chapter 4). In: Baertschi S, Alsante K, Reed R,
editors. Pharmaceutical Stress Testing. New York, NY: Informa
Disclaimer The views and opinions in this article are only those of Healthcare; 2011.
16. Lhasa Limited. Zeneth. Leeds UK: Lhasa Limited,; 2013; Version
the authors and do not necessarily reflect the views or policies of 5:(Available from: www.lhasalimited.org/zeneth/.
the companies, organizations or agencies cited herein. 17. Albini A, Anderson NH, Baertschi S, Boxhammer J, Byard SJ,
Carter PL, et al. In: Tonnesen H, editor. Photostability of drugs
and drug formulations. 2nd ed. Boca Raton, Florida: CRC Press;
2004. 448 p.
18. Piechocki JT, Tonnesen H, Allen JM, Allen SK, Gauglitz G,
REFERENCES Hubig SM, et al. In: Piechocki JT, Thoma K, editors.
Pharmaceutical photostability and stabilization technology. 1st
1. International Conference on Harmonisation of Technical ed. New York, NY: Informa Healthcare; 2007. 445 p.
Requirements for Registration of Pharmaceuticals for Human 19. Lukulay P, Hokanson G. Reconciling mass balance in forced
Use (ICH). Stability Testing of New Drug Substances and degradation studies. Pharm Tech. 2005; 106-12.
Products Q1A (R2). ICH Harmonized Tripartite Guidelines. 20. Baertschi SW. Analytical methodologies for discovering and pro-
2003. filing degradation-related impurities. TrAC Trends Anal Chem.
2. Parenty ADC, Button WG, Ott MA. An expert system to predict 2006;25(8):758–67.
the forced degradation of organic molecules. Molecular 21. Nussbaum MA, Kaerner A, Jansen PJ, Baertschi SW. Role of
Pharmaceutics. 2013;in press. "mass balance" in pharmaceutical stress testing (Chapter 9). In:
3. Baertschi SW, Reynolds DW, Jansen P, Alsante K, Santafianos D, Baertschi SW, Alsante KM, Reed RA, editors. Pharmaceutical
Kimmer Smith W, et al. Pharmaceutical stress testing: predicting Stress Testing: Predicting Drug Degradation (Drugs and the
drug degradation. Second ed. Baertschi SW, Alsante KM, Reed Pharmaceutical (Sciences). Second Edition ed: Informa Life
RA, editors: InformaHealth Sciences; 2011. 624 p. Sciences; 2011. p. 233-53.
4. International Conference on Harmonisation of Technical 22. Nursten HE, Royal Society of Chemistry. The Maillard Reaction:
Requirements for Registration of Pharmaceuticals for Human Chemistry, Biochemistry, and Implications: Royal Society of
Use (ICH). Guideline on Nonclinical Safety Studies for the Chemistry; 2005.
Conduct of Human Clinical Trials and Marketing Authorization 23. Nussbaum MA, Baertschi SW, Jansen PJ. Determination of relative
for Pharmaceuticals ICH M3 (R2). ICH Harmonized Tripartite UV response factors for HPLC by use of a chemiluminescent nitro-
Guidelines. 2009. gen-specific detector. J Pharm Biomed Anal. 2002;27(6):983–93.
5. International Conference on Harmonisation of Technical 24. Baertschi SW, Alsante KM, Santafianos D. The chemistry of drug
Requirements for Registration of Pharmaceuticals for Human degradation (Chapter 3). In: Baertschi SW, Alsante KM, Reed
212 Alsante et al.

RA, editors. Pharmaceutical Stress Testing: Predicting Drug 43. Food and Drug Administration (FDA). Guidance for Industry:
Degradation: Informa Life Sciences; 2011. p. 49-141. Liposome Drug Products: Chemistry, Manufacturing, and
25. Alsante KM, Baertschi SW, editors. Reviewing Advances in Controls; Human Pharmacokinetics and Bioavailability; and
Kn owle dge of D rug De grad ati on Che mist ry. Forc ed Labeling Documentation (Draft Guidance). 2002.
Degradation for Pharmaceuticals : Conference Proceedings; 44. Moll KP, Stosser R, Herrmann W, Borchert HH, Utsumi H. In
2012 January 18-19, 2012; United Kingdom. vivo ESR studies on subcutaneously injected multilamellar lipo-
26. Alsante KM, Ando A, Brown R, Ensing J, Hatajik TD, Kong W, somes in living mice. Pharm Res. 2004;21(11):2017–24.
et al. The role of degradant profiling in active pharmaceutical 45. Zuidam NJ, Gouw HK, Barenholz Y, Crommelin DJ. Physical
ingredients and drug products. Adv Drug Del Rev. 2007;59:29–37. (in) stability of liposomes upon chemical hydrolysis: the role of
27. Reynolds DW, Facchine KL, Mullaney JF, Alsante KM, Hatajik TD, lysophospholipids and fatty acids. Biochim Biophys Acta.
Motto MG. Available guidance and best practices for conducting 1995;22(1):101–10.
forced degradation studies. Pharm Tech. 2002;26(2):48–54. 46. Grit M, Crommelin DJ. The effect of surface charge on the
28. Alsante KM, Snyder KD, Swartz M, Parks C. Applications de- hydrolysis kinetics of partially hydrogenated egg phosphatidyl-
velopment using out-of-the-box software: a structure searchable choline and egg phosphatidylglycerol in aqueous liposome dis-
degradation/impurity database. Scientific Computing and persions. Biochim Biophys Acta. 1993;17(1):49–55.
Instrumentation. June 2012:30-7. 47. MacLachlan I. Liposomal formulations for nucleic acid delivery
29. Boyd DB, Sharp TR. The Power of Computational Chemistry to (Chapter 9). In: Corooke ST, editor. Antisense drug technologies:
leverage Stress Testing of Pharmaceuticals. In: Baertschi SW, principles, strategies, and applications. 2nd ed. Boca Raton, FL:
Alsante KM, Reed RA, editors. Pharmaceutical stress testing. CRC; 2007. p. 237–69.
Predicting Drug Degradation: Informa Life Sciences; 2011. 48. Li Z, Zhang Y, Wurtz W, Lee JK, Malinin VS, Durwas-Krishnan
30. Handbook of Pharmaceutical Excipients. Seventh ed. Rowe RC, S, et al. Characterization of nebulized liposomal amikacin
Sheskey PJ, Cook WG, Fenton ME, editors: Pharmaceutical; 2012. (Arikace) as a function of droplet size. J Aerosol Med Pulm
31. Wasylaschuk W, Harmon P, Wagner G, Harman AB, Templeton Drug Deliv. 2008;21(3):245–54.
AC, Xu H, et al. Evaluation of hydroperoxides in common phar- 49. Abend A, Duersch B, Fiszlar K. Small molecule parenteral drugs:
maceutical excipients. J Pharm Sci. 2007;96(1):106–16. practical aspects of stress testing (Chapter 12). In: Baertschi SW,
32. Yue H, Bu X, Huang M-H, Young J, Raglione T. Quantitative Alsante KM, Reed RA, editors. Pharmaceutical Stress Testing:
determination of trace levels of hydrogen peroxide in Predicting Drug Degradation: Informa Life Sciences; 2011. p.
crospovidone and a pharmaceutical product using high perfor- 322-42.
mance liquid chromatography with coulometric detection. Int J 50. Abend A, Templeton A. The impact of stopper drying for lyoph-
Pharm. 2009;375:33–40. ilized drug products. Am Pharm Rev. 2010;25(6):201–6.
33. Li Z, Jacobus LK, Wuelfing WP, Golden M, Martin GP, Reed 51. Carlisle CB, Cooper DE. Tunable diode laser frequency modu-
RA. Detection and quantification of low-molecular-weight alde- lation spectroscopy through an optical fiber: high‐sensitivity de-
hydes in pharmaceutical excipients by headspace gas chromatog- tection of water vapor. Appl Phys Lett. 1990;56:805–7.
raphy. J Chromatogr A. 2006;1104(1–2):1–10. 52. Templeton AC, Placek J, Xu H, Mahajan R, Hunke WA, Reed
34. Li Z, Kozlowski BM, Chang EP. Analysis of aldehydes in excip- RA. Determination of the moisture content of bromobutyl rub-
ients used in liquid/semi-solid formulations by gas chromatogra- ber stoppers as a function of processing: implications for the
phy-negative chemical ionization mass spectrometry. J stability of lyophilized products. PDA J Pharm Sci Technol.
Chromatogr A. 2007;1160(1–2):299–305. 2003;57(2):75–87.
35. Hartauer KJ, Arbuthnot GN, Baertschi SW, Johnson RA, Luke 53. Reed RA, Harmon P, Manas D, Wasylaschuk W, Galli C, Biddell
WD, Pearson NG, et al. Influence of peroxide impurities in R, et al. The role of excipients and package components in the
povidone and crospovidone on the stability of raloxifene hydro- photostability of liquid formulations. PDA J Pharm Sci Technol.
chloride in tablets: identification and control of an oxidative 2003;57(5):351–68.
degradation product. Pharm Dev Technol. 2000;5(3):303–10. 54. Faust BC, Zepp RG. Photochemistry of aqueous iron(III)-
36. Waterman KC. The application of the Accelerated Stability polycarboxylate complexes: roles in the chemistry of atmospheric
Assessment Program (ASAP) to quality by design (QbD) for and surface waters. Environ Sci Technol. 1993;27(12):2517–22.
drug product stability. AAPS PharmSciTech. 2011;12(3):932–7. 55. Haber F, Weiss J. Uber die Katalyse des Hydroperoxydes.
37. Glodek M, Liebowitz S, McCarthy R, McNally G, Oksanen C, Naturwissenschaften. 1932;20(51):948–50.
Schultz T, et al. Product robustness—a PQRI white paper. Pharm 56. Pilatti C, Torre MC, Chiale C, Spinetto M. Stability of pilocarpine
Eng. 2006;26(6):1–11. ophthalmic solutions. Drug Dev Ind Pharm. 1999;25(6):801–5.
38. Wu Y, Levons J, Narang AS, Raghavan K, Rao VM. Reactive 57. Al-Badr AA, Aboul-Enein HY. Pilocarpine. In: Florey K,
impurities in excipients: profiling, identification and mitigation of editor. Analytical profiles of drug substances. Academic; 1983.
drug-excipient incompatibility. AAPS PharmSciTech. p. 385–432.
2011;12(4):1248–63. 58. Singh S, Mariappan TT, Sharda N, Kumar S, Chakraborti AK.
39. Kensil CR, Dennis EA. Alkaline hydrolysis of phospholipids in The reason for an increase in decomposition of rifampicin in the
model membranes and the dependence on their state of aggrega- presence of isoniazid under acid conditions. Pharm Pharmacol
tion. Biochemistry. 1981;20(21):6079–85. Commun. 2000;6(9):405–10.
40. Zuidam NJ, Crommelin DJ. Chemical hydrolysis of phospho- 59. Yoshioka S, Stella VJ. Stability of drugs and dosage forms. First
lipids. J Pharm Sci. 1995;84(9):1113–9. ed: Springer; 2000.
41. Schnitzer E, Pinchuk I, Lichtenberg D. Peroxidation of liposomal 60. Food and Drug Administration (FDA). Questions and Answers on
lipids. Eur Biophys J. 2007;36(4–5):499–515. Current Good Manufacturing Practices, Good Guidance Practices,
42. Zhang JA, Pawelchak J. Effect of pH, ionic strength and oxygen Level 2 Guidance—Laboratory Controls. Guidance, Compliance,
burden on the chemical stability of EPC/cholesterol liposomes & Regulatory Information [Internet]. 2011. Available from: http://
under accelerated conditions. Part 1: Lipid hydrolysis. Eur J www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/
Pharm Biopharm. 2000;50(3):357–64. Guidances/ucm124785.htm.