Stress Testing of Drug Substances and Drug Products

Document Sample
Stress Testing of Drug Substances and Drug Products Powered By Docstoc
					Data and Review

                 Toward a Generic Approach for
                 Stress Testing of Drug
                 Substances and Drug Products
                 Silke Klick,* Pim G. Muijselaar, Joop Waterval,* Thomas Eichinger, Christian Korn, Thijs K. Gerding, Alexander J.
                 Debets, Cari Sänger-van de Griend, Cas van den Beld, Govert W. Somsen, and Gerhardus J. De Jong

The Impurity Profiling Group has developed a                                tress testing is defined as the stability testing of drug sub-
generic approach for conducting stress testing                              stances and drug products under conditions exceeding
on drug substances and drug products.The                                    those used for accelerated testing. Although it is an inte-
                                                                            gral part of the information provided to regulatory au-
proposed strategy is evaluated and verified                          thorities in registration application dossiers (ICH Q1A[R2])
with historical data and new experiments.                            (1), details regarding the design of such studies are not covered
Results demonstrate that the proposed                                by regulatory guidance documents (see Reynolds et al. for a re-
approach is reasonable and generates                                 view of regulatory guidance on stress testing [2]).
relevant, generally predictive results for the                          Pharmaceutical companies perform stress testing (also called
development of a stability-indicating method.                        forced-degradation studies) during preformulation to help se-
                                                                     lect compounds and excipients for further development, to fa-
                                                                     cilitate salt selection or formulation optimization, and to pro-
                                                                     duce samples for developing stability-indicating analytical
                                                                     methods. Stress testing provides information about degrada-
                                                                     tion mechanisms and potential degradation products. This in-
                                                                     formation then can be used to develop manufacturing processes
                                                                     or to select proper packaging. It may also help in preparing ref-
                                                                     erence material of identified degradation products. Although
                 S. Klick, PhD, is a team manager, Analytical        preformulation work is part of early-phase drug development,
                 Development, at AstraZeneca R&D (Mölndal
                 Sweden), tel. 46 31 7761758, fax 46 31
                                                                     stress testing often is repeated when manufacturing processes,
                 7767337, P.G.          product composition, and analytical procedures are refined and
                 Muijselaar, PhD, is a senior analytical scientist   reach a more final state (3). This is usually the case when a com-
                 and T.K. Gerding, PhD, is a manager, Sector         pound enters clinical Phase 3 studies. According to various reg-
                 Chemical & Pharmaceutical Development, at Solvay    ulatory guidance documents, results from stress testing should
                 Pharmaceuticals (The Netherlands). J. Waterval,
                 PhD, is a group leader, Department of Pharma-
                                                                     be included in the dossier when a registration application is
                 ceutics at Akzo Nobel, N.V. Organon (The            submitted to regulatory authorities.
                 Netherlands), tel. 31 412662070, fax 31                This article discusses stress testing according to the regula-
                 412662524,               tory guidance documents, with emphasis on what should be
                 T. Eichinger, PhD, is a deputy global director,     considered for late-clinical phases and for registration applica-
                 Analytical Sciences Department and C. Korn,
                 PhD, is a manager, Analytical Development,
                                                                     tion dossiers (e.g., marketing authorization applications or new
                 Sanofi-Aventis Frankfurt (Germany). A.J. Debets,    drug applications). (During early development, other aspects
                 PhD, is a general manager of Organon                of stress testing may be relevant for various screening purposes,
                 Development GmbH, Akzo Nobel (Germany). C.E.        but these are not the primary subject of this article.)
                 Sänger-van de Griend, PhD, is an associate             Many pharmaceutical companies have policies such as stan-
                 principal scientist, Analytical Development, at
                 AstraZeneca (Södertälje, Sweden). C. van den
                                                                     dard operating procedures or internal guidelines for stress test-
                 Beld, PhD, is the section head, Pharmaceutical      ing. There is no consensus, however, about how much stress is
                 Development Section/PDD at Yamanouchi Europe        adequate. This fact was recently confirmed in an extensive bench-
                 BV (The Netherlands). G.W. Somsen, PhD, is          marking study in which information about the design, condi-
                 an associate professor and G.J. De Jong, PhD,       tions, procedures, and the organization to oversee stress test-
                 is a professor, Pharmaceutical Analysis, Utrecht
                 University (The Netherlands).
                                                                     ing activities was requested from 20 companies (3). The authors
                                                                     concluded that the pharmaceutical companies were applying
                 *To whom all correspondence should be addressed.
                                                                     significantly diversified approaches.
48   Pharmaceutical Technology   FEBRUARY 2005                                                                     www.phar
Data and Review
                                                                                                                           Almost no literature can be found in
 Overview of stress-testing guidance documents                                                                          which stress conditions were investigated
  Standard            Title and reference                                                    Status Date                on their ability to predict the degrada-
  ICH Q1A(R2)        Stability Testing of New Drug Substances and Products (1)               Step 4 February 2003       tion products formed during long-term
  ICH Q1B            Photostability Testing of New Drug Substances and Products (12) Step 4 November 1996               stability studies of marketed formula-
  ICH Q2B            Validation of Analytical Procedures:Methodology (14)                    Step 4 November 1996       tions. Some direction regarding the prac-
  ICH Q3A(R)         Impurities in New Drug Substances (15)                                  Step 4 February 2002       tical conduct and issues related to stress
  ICH Q3B(R)         Impurities in New Drug Products (16)                                    Step 4 February 2003       testing under various ICH prescribed
  FDA Guidance Submitting Documentation for the Stability of Human Drugs                              February 1987     conditions has been published (5). This
                     and Biologics (6)                                                                                  reference described a classification scheme
  FDA Guideline Submitting Samples and Analytical Data for Methods Validation (7)                     February 1987     and included decision trees to help select
  FDA Reviewer Validation of Chromatographic Methods (8)                                              November 1994     the right type of stress conditions. Al-
    Guidance                                                                                                            though this approach is systematic, the
  FDA Guidance Stability testing of Drug Substances and Drug Products (9)                    Draft June 1998            suggested stress conditions seem rather
  FDA Guidance Analytical Procedures and Methods Validation (11)                             Draft August 2000          harsh. For example, refluxing the drug in
  FDA Guidance INDs for Phase 2 and Phase 3 Studies,Chemistry,Manufacturing,                          May 2003          0.1 N HCl /0.1 N NaOH for 8 h was sug-
                     and Controls Information (10)                                                                      gested as a starting condition, which in
  ICH Q1A(R2): Stability Testing of New Drug Substances and Products                                                    our opinion would generate irrelevant
  Drug substance. Stress testing a drug substance can help identify the likely degradation products,which can in        degradation in many cases.
  turn help establish the degradation pathways and the intrinsic stability of the molecule and validate the                Representatives from several pharma-
  stability-indicating power of the analytical procedures used.The nature of the stress testing will depend on the      ceutical companies and Utrecht Univer-
  individual drug substance and the type of drug product.                                                               sity (The Netherlands) have collaborated
    Stress testing is likely to be carried out on a single batch of the drug substance.It should include the effect of  to form the Impurity Profiling Group
  temperatures (in 10 C increments [e.g., 50 C,60 C] above that for accelerated testing),humidity (e.g., 75% RH         (IPG). For approximately three years, this
  or greater) where appropriate,oxidation,and photolysis on the drug substance.Testing should evaluate the              group has examined the issues relevant to
  susceptibility of the drug substance to hydrolysis across a wide range of pH values when in solution or               stress testing of low molecular weight com-
  suspension.Photostability testing should be an integral part of stress testing.                                       pounds. The main focus of the group has
    Examining degradation products under stress conditions is useful for establishing degradation pathways and          been to generate representative degrada-
  developing and validating suitable analytical procedures.It may not be necessary,however,to examine for               tion samples for developing stability-in-
  specific degradation products if previous studies have demonstrated that these products are not formed under          dicating analytical methods for both drug
  accelerated or long-term storage conditions.                                                                          substances and drug products. The pre-
    Drug product. The design of formal stability studies for a drug product should be based on the behavior and         dictive value of the degraded samples for
  properties of the drug substance,the results from stability studies on the drug substance,and the experience          real-time storage of the final product
  gained from clinical formulation studies.The likely changes to storage conditions and the rationale for the           under long-term and accelerated storage
  selection of attributes to be tested in the formal stability studies should be stated.Photostability testing should   conditions was defined as a measure of the
  be conducted on at least one primary batch of the drug product if appropriate.Standard conditions for                 balance between “purposeful degradation”
  photostability testing are described in ICH Q1B.                                                                      and the formation of irrelevant artifacts
    Any evaluation should take into account not only the assay but also the degradation products and other              that may lead to misinterpretations (e.g.,
  appropriate attributes.Where appropriate,attention should be paid to reviewing the adequacy of the mass               secondary degradation products).
  balance and stability and degradation performance.                                                                       This article reviews the regulatory
                                                                                      —sidebar continued on page 52     guidances’ exact text related to stress test-
                                                                                                                        ing and the current practices in partici-
                                                                                                                        pating companies and presents a proposal
    Many researchers have developed stability-indicating meth- and verification with practical results of the feasibility of a
ods that involve applying various stress regimes to generate de- generic approach toward stress testing.
graded samples. Bakshi and Singh evaluated these studies and
concluded that the majority would fall short of meeting ICH Overview of regulatory guidance documents
and FDA requirements (4). The main issue was that the required Several regulatory guidance documents mention forced degrada-
stress studies were not performed.                                                            tion or stress testing of drug substances and/or drug products (1,
    Although the development of a broad range of pharmaceu- 6–16). A summary of these has been published (2). To enable a
tical compounds requires flexibility in how stress testing is good interpretation of the available guidance documents and to
conducted, many companies apply extreme stress conditions. provide the reader with a comprehensive overview, the text of the
Typically, stress conditions are determined on the basis of the relevant paragraphs of these guidance documents is provided in
senior scientist’s experience or are copied from previously ap- the sidebar, “Overview of stress-testing guidance documents.”
plied example studies on other products. Rationales explain-                                      The ICH stability guideline Q1A(R2) defines stress testing
ing why certain stress conditions are applied typically are not for drug substances and drug products. For drug substances,
provided.                                                                                     these tests are studies undertaken to elucidate the intrinsic stabil-
50     Pharmaceutical Technology   FEBRUARY 2005                                                                                          www.phar
Data and Review
                                                                                                           ity of the drug substance. Such testing is part of the
Overview of stress-testing guidance documents (continued)                                                  development strategy and is normally carried out
ICH Q1B: Stability Testing: Photostability Testing of New Drug Substances and Products                     under more severe conditions than those used for ac-
Drug substance. Photostability testing should consist of two parts:forced-degradation testing and          celerated testing. For drug products, stress tests are
confirmatory testing.The purpose of forced-degradation testing is to evaluate the overall                  studies undertaken to assess the effect of severe con-
photosensitivity of the material for method-development purposes and/or degradation pathway                ditions on the drug product. Such studies include pho-
elucidation.This testing may involve the drug substance alone and/or the substance in simple               tostability testing (see ICH Q1B) and specific testing
solutions and suspensions to validate the analytical procedures.In these studies,the samples should        on certain products (e.g., metered-dose inhalers,
be in chemically inert and transparent containers.For forced-degradation studies,various exposure          creams, emulsions, and refrigerated aqueous liquid
conditions may be used,depending on the photosensitivity of the drug substance and the intensity of        products).
the light sources.For development and validation purposes,it is appropriate to limit exposure and end          In addition to investigating drug substance or drug
the studies if extensive decomposition occurs.For photostable materials,studies may be terminated          product stability, stress testing studies may also pro-
after an appropriate exposure level has been used.The design of these experiments is left to the           vide information about degradation pathways and
applicant’s discretion although the exposure levels used should be justified.                              the selectivity of the applied analytical methods. Ac-
  Under forcing conditions,decomposition products may be observed that are unlikely to be formed           cording to the ICH and FDA guidance documents,
under the conditions used for confirmatory studies.This information may be useful in developing and        stress testing is conducted to fulfill three main pur-
validating suitable analytical methods.If in practice it has been demonstrated that they are not           poses: to provide a stability assessment of the drug
formed in the confirmatory studies,these degradation products need not be further examined.                substance or the drug product; to elucidate the pos-
ICH Q2B: Validation of Analytical Procedures: Methodology                                                  sible degradation pathways of the drug substance or
Drug substance/drug product. If impurity or degradation product standards are unavailable,                 the active pharmaceutical ingredient in the drug prod-
specificity may be demonstrated by comparing the test results of samples containing impurities or          uct; and to investigate the stability-indicating power
degradation products with a second well-characterized procedure;e.g., pharmacopeial method or              of the analytical procedures applied for the drug sub-
other validated analytical procedure (independent procedure).As appropriate,this should include            stance and the drug product.
samples stored under relevant stress conditions (light,heat,humidity,acid–base hydrolysis,                     Although the regulatory guidance documents
oxidation).                                                                                                listed in the sidebars define the concept of stress
                                                                                                           testing, they do not provide detailed information
ICH Q3A(R): Impurities in New Drug Substances
                                                                                                           about a stress testing strategy. The experimental
Drug substance. The applicant should summarize the actual and potential impurities most likely to
                                                                                                           conditions to conduct stress tests are described in a
arise during the synthesis,purification,and storage of the new drug substance.This summary should
                                                                                                           general way and the exact stress conditions to be ap-
be based on sound scientific appraisal of the chemical reactions involved in the synthesis,impurities
                                                                                                           plied are not described. Some guidelines emphasize
associated with raw materials that could contribute to the impurity profile of the new drug
                                                                                                           that the experimental conditions of stress tests de-
substance,and possible degradation products.This discussion can be limited to those impurities that
                                                                                                           pend on the nature of the drug substance and the
might reasonably be expected based on knowledge of the chemical reactions and conditions
                                                                                                           drug product (1, 9).
involved.In addition,the applicant should summarize the laboratory studies conducted to detect
                                                                                                               The available guidance documents also do not state
impurities in the new drug substance.This summary should include test results of batches
                                                                                                           the extent to which stress tests should be carried out—
manufactured during the development process and batches from the proposed commercial process,
                                                                                                           that is, how much stress should be applied or how
as well as the results of stress testing (see ICH Guideline Q1A on Stability) used to identify potential
                                                                                                           much degradation should be aimed for. It is recog-
impurities arising during storage.The impurity profile of the drug substance batches intended for
                                                                                                           nized, however, that during stress testing, degrada-
marketing should be compared with those used in development and any differences discussed.
                                                                                                           tion products can be observed that are not formed
ICH Q3B(R): Impurities in New Drug Products                                                                during accelerated or long-term stability studies.
Drug product. Analytical procedures should be validated to demonstrate specificity for the specified       Hence, these degradation products need not always
and unspecified degradation products.As appropriate,this validation should include samples stored          to be examined (1, 9).
under relevant stress conditions:light,heat,humidity,acid/base hydrolysis,and oxidation.                       It can be concluded that the available guidance doc-
FDA Guidance for Industry: Submitting Documentation for the Stability of Human Drugs                       uments allow for performing stress tests on a sound
  and Biologics                                                                                            scientific basis. The experimental conditions should
Drug substance. A program for the stability assessment might include storage at ambient                    be realistic and lead to “purposeful degradation.” That
temperature and under stress conditions.Stress testing conditions ordinarily include temperature           is, stress tests should generate representative samples
(e.g.,5,50,and 75 C),humidity,where appropriate (e.g.,75% or greater) and exposure to various              to assess drug substance and drug product stability,
wavelengths of electromagnetic radiation (e.g., 190–780 nm UV and visible ranges),preferably in            provide information about possible degradation path-
open containers,where applicable.It is also suggested that the following conditions be evaluated in        ways, and demonstrate the stability-indicating power
stability studies on solutions or suspensions of the bulk-drug substance:acidic and alkaline pH,high-      of the analytical procedures applied.
oxygen atmosphere,and the presence of added substances under consideration for product
formulation.                                                                                               Current practice in participating companies
  Drug product. Stress testing the drug product helps identify potential problems during storage and       As a starting point for the IPG’s discussion about
transportation and provides an estimate of the expiration-dating period.                                   stress testing, participants shared existing strategies
                                                                      —sidebar concluded on page 54        for stress testing applied by the participating com-
                                                                                                           panies. The approaches were very different. Some
52    Pharmaceutical Technology   FEBRUARY 2005                                                                                             www.phar
Data and Review
                                                                                                                Out of
Overview of stress-testing guidance documents (concluded)                                                     proportion                   False
FDA Guideline for Submitting Samples and Analytical Data for Methods Validation                                  10%                       34%
Drug product. Information supporting the suitability of the methodology for the dosage form.
Generally,this should include the following:A degradation schematic for the active ingredient in the
dosage form,where possible (e.g., products of acid–base hydrolysis,temperature degradation,
photolysis,and oxidation).                                                                                                          29%
FDA Reviewer Guidance: Validation of Chromatographic Methods
Drug substance. Submission of data from stress testing the drug substance using acid and base
hydrolysis,temperature,photolysis and oxidation according to the Guideline for Submitting Samples         Figure 1: The effect of a fixed set of fast and severe
and Analytical Data for Methods Validation is recommended.                                                stress conditions on relevant degradation.
FDA Guidance for Industry: Stability Testing of Drug Substances and Drug Products
                                                                                                          companies had standard operating procedures in
Drug substance/drug product. Stress testing helps determine the intrinsic stability characteristics of
                                                                                                          place in which conditions for stress testing stud-
a molecule by establishing degradation pathways to identify the likely degradation products and to
                                                                                                          ies were described in detail. Other companies were
validate the stability-indicating power of the analytical procedures used.Stress testing provides data
                                                                                                          applying case-by-case approaches that depended
about the forced-decomposition products and decomposition mechanisms of the drug substance.The
                                                                                                          on the chemistry of the compound to be evalu-
severe conditions that may be encountered during distribution can be covered by stress testing
                                                                                                          ated or the dosage form to be developed.
definitive batches of the drug substance.These studies should establish the inherent stability
                                                                                                             After thoroughly discussing the various pro-
characteristics of the molecule such as the degradation pathways and lead to identification of
                                                                                                          cedures, members of the IPG agreed that the com-
degradation products and hence support the suitability of the proposed analytical procedures.The
                                                                                                          panies were applying conditions that in many
detailed nature of the studies will depend on the particular drug substance and type of drug product.
                                                                                                          cases were too drastic and were causing irrele-
  This testing is likely to be carried out on a single batch of a drug substance.Testing should include
                                                                                                          vant degradation. Typical conditions for stress
the effects of temperatures in 10 C increments above the accelerated temperature test condition
                                                                                                          testing in solution included various pH (1–12),
(e.g., 50 C,60 C) and humidity,where appropriate (e.g., 75% or greater).In addition,oxidation and
                                                                                                          high temperatures ( 70 C), extreme light expo-
photolysis on the drug substance plus its susceptibility to hydrolysis across a wide range of pH values
                                                                                                          sure, or the presence of oxidative or reductive
when in solution or suspension should be evaluated.Results from these studies will form an integral
                                                                                                          agents. For stress testing in the solid state, applied
part of the information provided to regulatory authorities.Light testing should be an integral part of
                                                                                                          conditions included high temperatures ( 70 C)
stress testing.
                                                                                                          and relative humidity (RH) ( 70%). These con-
  Some degradation pathways can be complex and under forced conditions,decomposition products
                                                                                                          ditions frequently are set on the basis of tradi-
may be observed that are unlikely to be formed under accelerated or long-term testing.This
                                                                                                          tion rather than scientific rationale.
information may be useful in developing and validating suitable analytical methods,but it may not
                                                                                                             To investigate the effectiveness and possible
always be necessary to examine specifically for all degradation products if,in practice,it has been
                                                                                                          shortcomings of typical fast and severe stress con-
demonstrated that these are not formed.
                                                                                                          ditions, a screening study was set up by one of the
FDA Guidance for Industry: Analytical Procedures and Methods Validation                                   participating companies. Fifteen low molecular
Drug substance/drug product. Degradation information obtained from stress studies (e.g., products         weight (i.e., non–biologically derived) compounds
of acid and base hydrolysis,thermal degradation,photolysis,oxidation) for the drug substance and          were studied using a fixed set of 11 commonly ap-
for the active ingredient in the drug product should be provided to demonstrate the specificity of the    plied stress conditions (see Table I). To save time,
assay and analytical procedures for impurities.The stress studies should demonstrate that impurities      fast (drastic) conditions were applied (e.g., a max-
and degradants from the active ingredient and drug product excipients do not interfere with the           imum two weeks at high temperature for solid
quantitation of the active ingredient.                                                                    stressing and a maximum one day for stress in so-
FDA Guidance for Industry: INDs for Phase 2 and Phase 3 Studies, Chemistry                                lution). Because no long-term stability data were
  Manufacturing, and Controls Information                                                                 available for the test compounds, accelerated stor-
Drug substance. Performance of stability-stress studies with the drug substance early in drug             age conditions (six months at 40 C and 75% RH)
development is encouraged because these studies provide information crucial to selecting stability-       provided the relevant degradation products.
indicating analytical procedures for real-time studies.If not performed earlier,stress studies should        The screening study did not include stable
be conducted during Phase 3 to demonstrate the inherent stability of the drug substance,potential         compounds in which no degradation products
degradation pathways,and the capability and suitability of proposed analytical procedures.Stress          formed under accelerated conditions. The ob-
studies should assess the stability of the drug substance in different pH solutions,in the presence of    served degradation was classified into five cate-
oxygen and light,and at elevated levels of temperatures and humidity.These one-time stress studies        gories (see Table II).
on a single batch are not considered part of the formal stability program.                                   The category “adequate” is considered optimal
  Drug product. For certain drug products,one-time stress testing can be warranted to assess the          and “out of proportion” is considered acceptable.
potential for changes in the physical (e.g., phase separation,precipitation,aggregation,changes in        All the other categories are of no additional value
particular-size distribution) and/or chemical (e.g., degradation and/or interaction of components)        and may even be misleading if taken too seriously.
characteristics of the drug product.The studies could include testing to assess the effect of high        The combined results for the set of 11 typical
temperature,humidity,oxidation,photolysis and/or thermal cycling.                                         stress conditions are shown in the summary dia-
                                                                                                          gram (see Figure 1). The extraordinary observa-
54    Pharmaceutical Technology   FEBRUARY 2005                                                                                            www.phar
Data and Review
tion is that only in a rather small fraction of cases does “ad-
equate” degradation occur with the applied typical stress con-
                                                                     Table I: Investigated set of fixed stress conditions.
ditions. The results were combined in one chart because gen-         Temperature      Temperature                                     Acid/base/
erally the same pie-chart distribution was obtained for each                          and moisture                  Light              oxidative
individual stress condition. It was frequently observed that         30 min,         1 week                   1 h xenon light        2 h 1 M HCl
although one stress condition was too soft for one compound,         121 °C          70 °C, ambient           (70–90 klx)
it led to massive degradation for another compound. Appar-                           2 weeks                  2 h xenon light        2 h NaOH
ently, it is still difficult to generate the relevant degradation                    70°C, ambient            (70–90 klx)
products with a fixed set of stress conditions. Results indi-                        1 week                   35 h UV light          2 h 3% H2O2
cate that some fine-tuning of applied stress conditions or                           70 °C, 100% RH           ( 210 W h/m2)
stress duration will be necessary and that no universal set of                       2 weeks
stress conditions exists. The large proportion of cases in the                       70 °C, 100% RH
“false” category is of particular concern because such stud-
ies may give a false idea of achieving relevant degradation.
    On the basis of the degree of degradation, one can assume        Table II: Type of degradation observed with a fixed set of
that typically the right amount of stress was applied and, there-    fast and severe stress conditions.
fore, relevant degradation products could be expected. This          Category   Explanation
was not the case, however. The most probable cause for the           Soft       No significant degradation and therefore
high percentage of irrelevant degradation may be the signifi-                   no relevant degradation products observed
cant difference in the thermodynamics of the applied stress          False      Fair amount of degradation ( 15%), however
conditions, resulting in degradation pathways other than what                   no relevant degradation product(s) observed
is observed during shelf-life stability-testing conditions. Ac-      Adequate Fair amount of degradation ( 15%) and at least
cording to the IPG, these results demonstrate that degradation                  one or all relevant degradation product(s) observed
from fast stress-testing conditions may be misleading and there-     Out of     Between 15 and 100% degradation and at least
fore should be interpreted with precaution.                          proportion one relevant degradation product observed
    Results from the screening test show that the typically ap-      Overdone Between 15 and 100% degradation, however, no
plied conditions usually generate a lot of irrelevant degrada-                  relevant degradation products are observed
tion products, thereby leading to complicated results and a
potential for not producing relevant degradation products. In
addition, evaluating large numbers of irrelevant degradation            Requirements for relevant stress conditions
products may generate high amounts of workload. The screen-             q Should lead to the degradation of the main compound,but not more than 5–15%
ing study by the IPG, therefore, demonstrates the limited value         q Should lead to a good predictability of degradation pathways (i.e., a low
of using fast and severe stress-testing conditions. Our hypoth-           probability of “drastic”or “false”degradation)
esis is that for successful stress testing, mild conditions should      q Should be conducted for no longer than three months

be used that may result in storage times that are longer than in
usual stress testing.                                                 when this percentage of degradation is achieved. It is not de-
                                                                      sirable to generate samples with extensive degradation because
Strategy toward a generic approach                                    of their limited relevance and the formation of secondary degra-
The IPG tried to elaborate an applicable generic approach for         dation products, which would lead to complicated degradation
stress testing to produce representative samples for developing       patterns.
stability-indicating methods for drug substances and drug prod-          One issue to address is how long a stress test should be con-
ucts. The initial question was how much stress could be regarded      tinued if the target is not achieved. The IPG proposes a maxi-
as adequate to generate such samples. The IPG group agreed            mum of 14 days for stress testing in solution (a maximum of
that the amount of stress applied to a sample should be selected      24 h for oxidative tests) to provide relevant stressed samples for
according to what is necessary to achieve “purposeful degrada-        methods development. The group also proposes a three-month
tion.” An optimal degradation pattern generated during stress         period for drug substances in the solid state and drug products,
testing would show only those degradation products observed           which may be too long for preformulation during early devel-
at the end of shelf life in regulatory stability studies and those    opment (e.g., early development screening or salt screening).
that might appear if the drug substance or drug product is not        In such cases, more drastic conditions can be applied to gener-
handled or packed properly (see sidebar, “Requirements for rel-       ate results on a shorter course but at a risk that the degradation
evant stress conditions”).                                            patterns may not be predictive. In general, to evaluate their rel-
   Chromatograms thus obtained will be representative and not         evance, results from stress tests should be compared with re-
too complicated to evaluate, which may be the case if drastic         sults obtained from long-term and accelerated stability studies
conditions are applied and many second- and third-generation          as soon as they are available.
degradation products are formed.                                         The IPG developed a generic protocol for stress testing on
   Stress testing should induce not more than 5–15% degrada-          the basis of the experience and scientific expertise of its partic-
tion of the main compound. A stress test should be stopped            ipating companies and by taking into consideration the exist-
56    Pharmaceutical Technology   FEBRUARY 2005                                                                              www.phar
Data and Review
ing regulatory guidance documents. The proposed conditions
should be sufficient and should normally induce realistic, pre-
                                                                        Table III: Storage conditions for stress testing of drug
dictive degradation. Predominant degradation products ob-               substance in solid state and drug product.
tained under these conditions should be regarded as relevant            Storage conditions                            Testing period*
for developing a stability-indicating method and will provide           40 C, 75 % RH; open storage**                 3 months
valuable information about relevant degradation pathways. For           50–60 C, ambient RH; open storage             3 months
this purpose, it is essential to identify the observed degradation      Photostability; according to ICH              according to ICH(67)
products by structure elucidation, especially if they are formed         *3 months or 5–15% degradation, whatever comes first.
in more than one of the experiments included in the generic             **For drug substances, typically a thin layer of material is spread in
protocol. However, common sense should be used when select-               a petri dish. Open storage is recommended if possible.
ing degradation products for identification from a complicated
pattern. Identification work should not be mandatory unless a
particular degradation product has been shown to be relevant            Table IV: Conditions for stress testing of drug substance
under long-term or accelerated storage conditions.                      in solution.
   Finally, the IPG decided that the agreed generic protocol for        Storage conditions                             Testing period*
stress testing should be developed using examples from the par-         pH 2, room temperature                         2 weeks
ticipating companies, with evaluation according to both histor-         pH 7, room temperature                         2 weeks
ical data from marketed products and, in some cases, results of         pH 10–12, room temperature                     2 weeks
new experiments.                                                        H2O2, 0.1–2% at neutral pH,                    24 h
   Program for stress testing drug substances in the solid state and    room temperature
drug product. The IPG established conditions for stress testing
                                                                        *Storage times given or 5–15% degradation, whatever comes first.
drug substances in the solid state and drug product (see Table
III). These conditions are regarded as sufficient to expose a solid-
state sample to temperature, humidity, and oxidation (oxida-           Such an approach increases the probability of generating irrel-
tion by open storage) stress. Testing periods are upper limits         evant degradation products, especially secondary degradation
and studies should be stopped when 5–15% degradation is                products, which was confirmed in the IPG’s screening study.
achieved.                                                              After extensive discussion, the IPG proposed not to use elevated
   Temperatures higher than those proposed in Table III (e.g.,         temperatures as a means to accelerate degradation.
70–90 C) can be applied to rapidly generate method develop-               In special cases, however, testing the drug substance at ele-
ment or preformulation samples. This approach, however, may            vated temperature in solution may be of interest (e.g., to pre-
lead to various degradation kinetics and irrelevant degradation        dict stability during autoclaving of a solution). Therefore, test-
products, as demonstrated in the screening experiment. Shorter         ing a drug substance in solution under elevated temperatures
storage times should be avoided to maintain realistic thermo-          should be considered on a case-by-case basis.
dynamics. Longer storage times are not recommended because                Stress testing in solution should be conducted on dissolved
such samples are hardly available early in development and             samples. Additives may be used to enhance the solubility of
should be considered only if no degradation is achieved after          compounds with limited aqueous solubility (e.g., cosolvents or
three months.                                                          cyclodextrins). Caution should be taken in these cases because
   During the evaluation of the generic stress program, a revi-        several additives are not inert agents. Cyclodextrins, for instance,
sion of ICH Q3B16 was published, in which light, heat, humid-          may protect potentially reactive moieties in a molecular struc-
ity, acid–base hydrolysis, and oxidation are mentioned as stress       ture, whereas cosolvents may affect degradation mechanisms
conditions to demonstrate the specificity of a stability indicat-      and even induce degradation (e.g., formation of reaction prod-
ing method for drug product. Light, heat, humidity and oxida-          ucts such as methyl or ethyl esters).
tion (open storage) are considered to be covered by the condi-            Suspensions are not recommended because degradation
tions listed in Table III. However, acid–base hydrolysis of the        might be influenced by the presence of particles, and degrada-
drug product is considered as a less relevant stress condition by      tion-sensitive moieties can be protected if a suspension is used.
the IPG members, provided that the information about the               Suspensions may sometimes be justified, however, especially if
degradation of the drug substance obtained under acidic and            the formulation is a suspension. The three pH regions given in
basic conditions is taken into consideration.                          Table IV may be chosen with flexibility, depending on the pH
   Program for stress testing of drug substance in solution. Condi-    of the potential formulation, anticipated degradation, solubil-
tions for stress testing of drug substance in solution are shown       ity, and so forth. Again, not more than 5–15 % degradation is
in Table IV. For many final products, especially most solid oral       the target value. If this amount of degradation is not achieved,
dosage forms, testing the drug substance in solution may be of         the samples may be stored for a longer period of time or at
limited relevance. However, stress testing in solution is gener-       higher temperatures on a case-by-case basis.
ally seen as relevant both for elucidation of degradation path-           Experiments should be performed in the laboratory without
ways and for specificity testing of the analytical method.             protection from daylight. Exposure of solutions in photostabil-
   Stress testing in solution or in suspension of the drug sub-        ity testing chambers will normally lead to much degradation
stance often is conducted at temperatures above ambient (3,5).         and an overestimation of photosensitivity. If degradation oc-
58    Pharmaceutical Technology   FEBRUARY 2005                                                                        www.phar
Data and Review

 Table V: Retrospective evaluation of the generic stress-testing protocol.
                                                                                      Degradation products observed
                                                            Company A                 Company B        Company C               Company D
                                                      #1    #2 #3 Remark               #1 Remark      #1 Remark               #1 #2 Remark
 Drug substance in solid state
 40 C, 75 % RH, 3 months open storage                                               –       N          –                        –
 50 C,ambient RH, 3 months open storage                                             –       N          –                   NP NP
 Photostability, according ICH                                                             L          –                    –    –       **
 Drug substance in solution
 pH 1, 3 days, room temperature                                                                           –       *
 pH 2, 2 weeks, room temperature                                                    –       N                               –           H
 pH 7, 2 weeks, room temperature                                                    –       N              3 days              –       H
 pH 10, 2 weeks, room temperature                                                                            6h
 pH 11, 2 weeks, room temperature                                                      –       N
 pH 12, 2 weeks, room temperature                                                                                           –    –       **
 0.1 % H2O2, 24 h, room temperature                                    H, L
 1 % H2O2, 24 h room temperature                                                              H
 2 % H2O2, 24 h, room temperature                                      H, L                                                –    –       **
 Fe2+ (0.05 mM), 2 weeks, room temperature.                                            –       N
 Cu2+ (0.05 mM), 2 weeks, room temperature                                                    L
 Drug product
 40 C, 75 % RH, 3 months open storage                NP     NP NP                     NP                                     NP NP
 50–60 C, ambient RH,                                NP     NP NP                     NP                 NP                   NP NP
   3 months open storage
 Photostability, according to ICH                    NP     NP NP                      –                  –                    –    –
 Real-time stability studies of drug product
 25 C, 60 % RH                                              –     –   18 months       36 months          24 months              – 5 years
 30 C, 70 % RH                                                                                                                      3 years
 40 C, 75 % RH                                                                         6 months           3 months                3 months
 40 C, 80 % RH                                                   18 months
 50 C, uncontrolled RH                                      –    – 18 months

  *pH 1 was chosen for harsher stress conditions because Compound 1 is stable in acidic media.
 **A known degradation product is formed under these stress conditions. This degradation product, however, is not formed during the real
   time stability studies. Key: “” denotes observed; “–” denotes not observed; “NP” denotes.not performed; “H” denotes additional degradation
   products observed at higher levels than the relevant degradation products mentioned in the table; “L” denotes additional degradation products
   observed at lower levels than the relevant degradation products mentioned in the table; “N” denotes no significant degradation. observed.

curs in samples exposed to ambient light, a dark control may               Results
be useful to distinguish between photodegradation and other                The proposed stress testing conditions of the generic approach
mechanisms.                                                                were evaluated retrospectively by comparing the degradation
   To test susceptibility to oxidation, 0.1–2 % of H2O2 at neu-            patterns obtained during real-time storage in regulatory stabil-
tral pH is the most predictive. A non-neutral pH may be help-              ity studies of marketed products with the degradation patterns
ful in cases of limited solubility or to distinguish between ox-           obtained by applying the conditions proposed by the generic
idative and other degradation.                                             protocol. Four products developed by their respective manu-
   Optional stress conditions for drug substance in solution. Routine      facturers (companies A, B, C, and D) were compared (see Table
testing of radical initiators or transition metals such as Fe3 and         V). These compounds represent a range of chemical character-
Cu2 as initiators or catalysts for oxidative degradation are not           istics that to some extent can be considered to be representa-
generally regarded as relevant, and a case-by-case approach is             tive of low molecular weight compounds in the pharmaceuti-
recommended. This means that the influence of radical initia-              cal industry.
tors or metal ions should be tested if expected to be relevant                The selected test compound from Company A had under-
(e.g., when these constituents are expected to be present in the           gone previous regulatory stability studies and all of its stability
potential formulation or if a special susceptibility can be antic-         data, including degradation pathways, were known and avail-
ipated). Testing Na2S2O3 as a reducing agent may be included               able. Results indicated that all degradation products that were
in stress testing studies in special cases if considered relevant.         observed in the regulatory stability study were readily detected
60    Pharmaceutical Technology   FEBRUARY 2005                                                                             www.phar
Data and Review
under the stress conditions of the pro-           described in a monograph in the European       eral tests in the generic stress-testing pro-
posed generic stress testing protocol (see        Pharmacopoeia. The test compound is sus-       tocol (see Table V). The most relevant con-
Table V). Degradation products 1, 2, and          ceptible to oxidation, and the resulting ox-   dition was the forced photostability study,
3 were relevant. Although some irrelevant         idation product is the only relevant degra-    which is required by ICH Q1B (12).
degradation products were observed, for           dation product for drug product stability.     Degradation product 1 was also generated
example under the oxidative conditions,           It is also included as a named degradation     in the presence of copper ions and by hy-
the number of degradation products was            product in the pharmacopeial monograph         drogen peroxide. Under these conditions
still easy to handle.                             for the drug substance. Results show that      the degradation product was not as clearly
   The test compound from Company B               the relevant degradation product is read-      dominating as in photostability testing,
is registered in most countries and also          ily formed in the drug substance by sev-       and in the case of hydrogen peroxide, it
                                                                                                 was not even the largest degradation prod-
                                                                                                 uct. The oxidation of this compound is a
                                                                                                 well-known feature of the substance class,
                                                                                                 and hence, the degradation product would
                                                                                                 be considered as a likely degradation prod-
                                                                                                 uct, even without performing any stress
                                                                                                    The drug product of this compound
                                                                                                 (extended-release tablets) was not tested
                                                                                                 according to the generic stress-testing pro-
                                                                                                 tocol. A review of older data indicated that
                                                                                                 the proposed conditions would not nec-
                                                                                                 essarily generate the degradation product.
                                                                                                 In particular, the tablet coating provides
                                                                                                 an effective protection against photochem-
                                                                                                 ical degradation. However, the degrada-
                                                                                                 tion product was formed in the drug sub-
                                                                                                 stance upon stress testing, and the
                                                                                                 oxidation is well-known for the substance
                                                                                                 class. Hence, the degradation product has
                                                                                                 been considered as a most likely degrada-
                                                                                                 tion product from early development of
                                                                                                 this product and onward.
                                                                                                    A marketed product from Company C
                                                                                                 was selected as a test compound. Several
                                                                                                 stability studies have been performed on
                                                                                                 this compound, and the degradation
                                                                                                 pathways are known. For the drug sub-
                                                                                                 stance studies, degradation product 1,
                                                                                                 which is not formed in the crystalline
                                                                                                 drug substance during long-term and ac-
                                                                                                 celerated storage, is formed by stress test-
                                                                                                 ing in solution (neutral and alkaline pH),
                                                                                                 as described in the generic stress-testing
                                                                                                    For drug product studies, degradation
                                                                                                 product 1 is formed under the influence
                                                                                                 of humidity and even under long-term
                                                                                                 conditions when the ratio of active to ex-
                                                                                                 cipient is worse (data for another dosage
                                                                                                 strength were not given). In this example,
                                                                                                 the proposed stress humidity conditions
                                                                                                 were sufficient to induce degradation.
                                                                                                    Company D also chose a well-docu-
                                                                                                 mented compound with an extensively in-
                                                                                                 vestigated degradation mechanism. Two
                                                                                                 main degradation products are known for
62    Pharmaceutical Technology   FEBRUARY 2005                                                                        www.phar
Data and Review
the various strengths of the marketed prod-       stress conditions in solution at acidic pH.      ferent drug products were formed during
uct. Degradation product 1 is observed dur-       From these results, one can conclude that        the proposed generic stress-testing proto-
ing all long-term stability studies. This         both degradation products observed dur-          col (see Table V). In general, purposeful
degradation product is formed during 3            ing long-term stability studies were ob-         degradation that is relevant to the prod-
months open storage at 40 C and 75% RH            served with the drug substance in solu-          ucts’ stability was obtained. The generated
of the drug substance, as well as in solution     tion during the generic stress-testing           samples were suitable for the development
at pH 7. Degradation product 2 is ob-             protocol.                                        of stability-indicating methods.
served during the long-term stability stud-          Results demonstrated that all relevant
ies at higher temperatures (30 and 40 C).         degradation products observed during             Conclusion
This degradation product is formed under          real-time stability studies of these four dif-   Results from stress testing studies are used
                                                                                                   to assess drug substance and drug prod-
                                                                                                   uct stability, to provide information about
                                                                                                   possible degradation pathways, and to
                                                                                                   demonstrate the stability-indicating ca-
                                                                                                   pability of the analytical methods used.
                                                                                                   Drastic conditions tend to generate irrel-
                                                                                                   evant degradation in the sample and may
                                                                                                   lead to methods optimized for the sepa-
                                                                                                   ration of an impurity profile, which is not
                                                                                                   relevant for an end-of-shelf-life sample.
                                                                                                      The IPG developed a generic approach
                                                                                                   for stress testing of drug substances and
                                                                                                   drug products. Efforts were aimed to pre-
                                                                                                   vent too-drastic conditions and to achieve
                                                                                                   “purposeful degradation,” predictive for
                                                                                                   what is relevant under long-term and ac-
                                                                                                   celerated storage conditions. The proposed
                                                                                                   stress testing program, which is in line with
                                                                                                   requirements from the regulatory agencies,
                                                                                                   was applied to four products developed by
                                                                                                   the participating companies and evaluated
                                                                                                   both retrospectively by historic data and by
                                                                                                   new experiments. Results demonstrated that
                                                                                                   the proposed generic approach is reason-
                                                                                                   able and generates relevant, generally pre-
                                                                                                   dictive results for the development of a sta-
                                                                                                   bility-indicating method. Some of the
                                                                                                   conditions included, however, still induce
                                                                                                   irrelevant degradation.
                                                                                                      Stress tests for developing a stability-
                                                                                                   indicating method should always be de-
                                                                                                   signed and evaluated with common sense
                                                                                                   and chemical knowledge, keeping in mind
                                                                                                   the manufacturing process and the nature
                                                                                                   of the final drug product. The proposed
                                                                                                   generic approach can be used as a start-
                                                                                                   ing point to set up a stress testing study,
                                                                                                   but a case-by-case approach for stress test-
                                                                                                   ing is essential to allow flexibility. This is
                                                                                                   also recognized by the regulatory agencies
                                                                                                   because very detailed instructions about
                                                                                                   how to perform stress testing are not given
                                                                                                   in the available guidance documents.
                                                                                                      A general experience in the IPG was
                                                                                                   that the formal stability study under ac-
                                                                                                   celerated conditions is highly predictive
                                                                                                   for long-term storage and the impurity
64    Pharmaceutical Technology   FEBRUARY 2005                                                                           www.phar
Data and Review
profile at the end of shelf life. These results should be used as             6. Guidance for Industry: Submitting Documentation for the Stability of
soon as they become available to verify the relevance of the stress              Human Drugs and Biologics (FDA, Rockville, MD, 1987).
                                                                              7. Guideline for Submitting Samples and Analytical Data for Methods Val-
testing performed.                                                               idation (FDA, Rockville, MD, 1987).
                                                                              8. Reviewer Guidance: Validation of Chromatographic Methods (FDA,
Acknowledgment                                                                   Rockville, MD,1994).
We wish to thank Jenny Ottosson for performing the stress test-               9. Guidance for Industry (Draft): Stability Testing of Drug Substances and
ing experiments; Magnus Erickson (AstraZeneca R&D, Möl-                          Drug products (FDA, Rockville, MD, 1998).
                                                                             10. Guidance for Industry (Draft): CMC Content and Format, INDs for
ndal, Sweden) and Han Op't Land, Henriëtte Hamstra, and Piet                     Phase 2 and 3 Studies of Drugs, Including Specified Therapeutic Biotech-
Hoogkamer (all at Solvay Pharmaceuticals B.V., The Nether-                       nology-Derived Products(FDA, Rockville, MD, 1999).
lands) for their valuable comments on the manuscript; Laurent                11. Draft Guidance for Industry on Analytical Procedures and Methods Val-
Duhau (Aventis, France) for the inspiring discussions; Mark                      idation: Chemistry, Manufacturing, and Controls Documentation (FDA,
Vermunt and Gerben Wynia for performing the screening study                      Rockville, MD, 2000).
                                                                             12. International Conference on Harmonization, “ICH Q1B: Photosta-
experiments; and Eddy Ruyter (Akzo Nobel, N.V. Organon, The                      bility Testing of New Drug Substances and Products,” Step 5 (1996).
Netherlands) for his comments.                                               13. International Conference on Harmonization, “ICH Q2A: Validation
                                                                                 of Analytical Procedures: Methodology,” Step 5 (1994).
References                                                                   14. International Conference on Harmonization, “ICH Q2B: Validation
 1. International Conference on Harmonization, “ICH Q1A(R2): Stabil-             of Analytical Procedures: Terms and Definitions,” Step 5 (1996).
    ity Testing of New Drug Substances and Products,” Step 5 (2003).         15. International Conference on Harmonization, “ICH Q3A-(R): Impu-
 2. D.W. Reynolds et al., “Available Guidance and Best Practices for Con-        rities in New Drug Substances,” Step 5 (2002).
    ducting Forced Degradation Studies,” Pharm. Technol. 26 (2), 48–56       16. International Conference on Harmonization, “ICH Q3B-(R): Impu-
    (2002).                                                                      rities in New Drug Products,” Step 5 (2003). P   T
 3. K.M. Alsante, L. Martin, S.W. Baertschi, “A Stress Testing Benchmark-
    ing Study,” Pharm. Technol. 27 (2), 60–72(2003).                                      Please rate this article.
 4. M. Bakshi and S. Singh, “Development of Validated Stability-Indicat-                  On the Reader Service Card, circle a number:
    ing Assay Methods: Critical Review,” J. Pharm. Biomed. Anal. 28,                       345 Very useful and informative
    1011–1040 (2002).                                                                      346 Somewhat useful and informative
 5. S. Singh and M. Bakshi, “Guidance on Conduct of Stress Tests to De-                    347 Not useful or informative
    termine Inherent Stability of Drugs,” Pharm. Technol. 24, 1–14 (2000).                Your feedback is important to us.

66    Pharmaceutical Technology   FEBRUARY 2005                                                                                 www.phar