MSc Chemistry
Analytical Sciences
Literature Thesis
Relevant Forced Degradation Conditions to Develop and
Validate Stability-Indicating Analytical Methods
by
Willem-Jan Spaans
11808543
September 2019
12 EC
June 2019 – October 2019
Daily Supervisor:
Supervisor/Examiner:
Second reviewer:
ii
Abstract
An essential part of pharmaceutical development is to develop safe and effective drugs.
Pharmaceutical companies should develop drugs that are stable during their shelf-life, to ensure the products are still safe and effective. Stability indicating analytical methods are developed to measure a potential decrease in active pharmaceutical ingredients. Such methods are used to prove the stability of a drug product during stability studies of a drug substance or product. During the development of stability-indicating methods, samples are stressed to mimic stability study samples. The process is either called forced degradation or stress testing.
Regulatory agencies require stress testing for new drug substance and product applications. The most well know agencies are the ICH, WHO, USFDA and the EMA. The ICH describes that stress testing should include the effect of temperature, humidity, oxidation, photolysis, and hydrolysis across a wide range of pH values. The WHO describes that relevant stress conditions result in 10 – 30 % decrease in the main component. However, not details on conditions are generally lacking. The lack of details has led to a wide variety of strategies applied by pharmaceutical companies. Stress testing of drug products is carried out for developing stability-indicating methods. Those methods are usually reversed-phase high-performance liquid chromatography methods coupled to UV detection. Before experimental work, hypothetical degradation products are determined using organic chemistry theory, literature or in-silico tools like Zeneth. Subsequently, ‘potential’
degradation products are obtained by stress testing. At last, ‘actual’ degradation products are assessed through stability studies.
Multiple case studies were assessed in this literature thesis. Those case studies contained both results from stress testing studies and stability studies. The results of those studies were compared to determine which stress conditions were relevant. In most of the studied cases, a high number of non-relevant degradation products were obtained. The high number of non-relevant degradation products can be devoted to the harsh conditions used for stress testing. When a large number of non-relevant degradation products are used for developing the stability-indicating method, the method development can take longer since it is more challenging to separate more compounds that are closely related to each other. Moreover, the obtained method can be more difficult to run. In a study performed by Klick et al., it was proven that milder stress conditions lead to less non-relevant degradation, while all non-relevant degradation products were still obtained. That seems to be a promising approach for generating degradation products for stability-indicating method
development. Unfortunately, still non-relevant degradation products were obtained using milder stress conditions.
Automation can be of aid during stress testing studies because less manual work needs to be carried out. Automation of stress testing can be done by interfacing reaction vessels to chromatography systems. Using such an approach multiple different conditions can be tested at the same time. Microwave-assisted stress testing is another approach to decrease the time needed for a stress testing study.
It is difficult to find relevant stress testing conditions that can be applied to different types of drug molecules. The approach by Klick et al. seems like a good starting point for stress testing. However, the time it will take to carry out a study with those conditions is relatively long.
iii
List of abbreviations
ACN Acetonitrile
ACVA 4,4′-Azobis(4-cyanovaleric acid) AIBN Azobisisobutyronitrile
AL Alizapride
API Active Pharmaceutical Ingredient CAD Charged Aerosol Detection CBS Clopidogrel Bisulfate DAD Diode Array Detection DOX Doxorubicin
DSPE 1,2-distearoyl-sn-glycero-3-phosphoethanolamine sodium salt ELSD Evaporative Light Scatter Detection
EMA European Medicines Agency FDC Fixed-dose combinations FPP Finished Pharmaceutical Product
HPLC High-Performance Liquid Chromatography HSPC Hydrogenated Soy Phosphatidylcholine ICE iChemExplorer
ICH The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use
MMPP Magnesium monoperoxyphthalate MPEG Methoxy Polyethylene Glycol
MS Mass spectrometry
NA Not Applicable NMP N-Methylpyrrolidone
NMR Nuclear Magnetic Resonance OFAT One Factor at a Time
PVP Polyvinylpyrrolidone RH Relative Humidity RI Refractive Index
RRF Relative Response Factor SIM Stability Indicating Method
UHPLC Ultra-High Performance Liquid Chromatography USFDA United States Food & Drug Authority
USP United States Pharmacoepiea UV Ultra Violet
iv
Table of contents
1. Introduction ... 1
2. Reasons for stress testing ... 2
3. Regulatory guidance ... 3
4. Stress testing strategy for stability-indicating method development ... 6
4.1. ‘Hypothetical’ degradation products ... 7
4.2. ‘Potential’ degradation products ... 10
4.3. ‘Potential’ vs. ‘actual’ degradation products ... 18
5. A generic approach to stress testing ... 34
6. Automation... 36
6.1. Automated forced degradation ... 36
6.2. High throughput ... 36
6.3. Microwave-assisted stress testing ... 37
6.4. Design of experiments ... 38
7. Doxil example ... 38
8. Conclusion ... 39
1
1. Introduction
Pharmaceutical companies should ensure the quality of drugs. One of the aspects of drug quality is the presence of impurities in the formulation. Impurities can have a negative effect on the efficacy and safety of the drug product [1]. To make sure a drug is safe, manufacturers need to perform stability studies during drug development [2,3]. These stabilities studies help in predicting
degradation behavior during shipping and storage conditions. Degradation products have a negative impact on safety and efficacy. During stability studies, the drug product is tested on the formation of degradation products. Stability-indicating methods are developed to qualitatively and quantitatively determine the decrease of the drug and the related degradation products.
Degradation of drug molecules is a complex area [3]. There can be three types of degradation products defined during drug development. The first type is the hypothetical degradation product. Those are predicted using in silico tools or found through literature searches. The second group is potential degradation products observed during forced degradation/stress testing studies. The third group is the actual degradation products formed under accelerated and long-term stability studies. In an ideal world, the actual degradation products are a selection of potential degradation products. Furthermore, the potential degradation products are a subset of the hypothetical degradation products. However, this is not always the case in real-world situations.
A stability-indicating method is used to determine the decrease of the active pharmaceutical ingredient (API) due to degradations [4]. Such a method can precisely and accurately measure the amount of API in a drug substance or drug product without interference of excipient, impurities, and degradation products. The selectivity of the method, amongst other parameters, is validated to prove the stability-indicating nature of the method. Forced degradation studies aid in the validation of the selectivity [5].
During forced degradation studies, the drug substance or drug product is stressed under conditions harsher than the recommended conditions for accelerated stability studies. For accelerated stability studies, the prescribed conditions are 40 °C and 75% relative humidity. The ICH recommends performing forced degradations or stress testing in steps 10 °C above those accelerate conditions. Next to that, the degradations studies should consist of acidic, alkaline, neutral, oxidative, photolytic, and thermal stress testing [6]. However, there is little to no guidance on the details of such
conditions. Since there is little guidance, several strategies are developed to obtained suitable degradation during forced degradations studies to prove the stability indication power of an analytical procedure.
The goal of this literature thesis is to obtain an overview of the current approaches of forced degradation studies used for the development of stability indicating analytical procedures. First, regulatory guidance is discussed. Secondly, the literature describing how to conduct forced degradation studies is reviewed. Thirdly, the literature describing both forced degradation and stability studies of drug substances or drug products is discussed. That is done to compare the approach with regulatory guidance and the best practices in the literature. Fourthly, alternative and innovative approaches are discussed. The fifth part is a description of how the found knowledge about forced degradation study design can be applied to the drug product Doxil. The description is how I would approach stress testing based on the gained knowledge found in the literature. In the end, recommendations on forced degradation studies will be given and some suggestions for further research.
2
2. Reasons for stress testing
Degradation products are formed in drug substances and drug products during the manufacturing process or storage. Degradation occurs in the presence of moisture, by exposure to light, by oxidation or heat. [7] Stress testing can help to understand the degradation of drug substances and drug products. By stress testing, ‘potential’ degradation products can be obtained. There are multiple reasons to perform stress testing of drug substances or drug products. The specificity of an analytical method is established using stress testing, also known as forced degradation studies. Hence, it can be called a stability-indicating method (SIM). The method should be able to quantitate the decrease of the main compound without interference from degradation products, impurities, and excipients. [8] Another reason for stress testing is determining the intrinsic stability and to elucidate the
degradation pathways, and to identify the degradants by their structure elucidation. The causes of degradation can be useful during formulation development, manufacturing, and packaging. It can aid in the discrimination between drug and non-drug related degradation products. [7,9–11]
Figure 1. Reasons to perform stress testing. One repsonse as primary reason for each respondent. Multiple responses for secondary reason for each respondent.
In a survey of over 20 companies about stress testing carried out in 2003, it was found that stress testing is a crucial subject in pharmaceutical product development. [12] According to the survey results, the most common reason for stress testing is method development [12]. The results of the question on reasons for stress testing are shown in Figure 1. However, still only seven companies gave that as a primary reason. Other reasons for stress testing were method validation, stability support and distribution, pre-formulation and excipient compatibility, and regulatory compliance. In case of stability support & distribution, stress testing is performed due to observed unknown degradation products after storing or shipping drug substances or products. In case of pre-formulation and excipient compatibility, stress testing is performed to gain information on the stability of an API in potential drug products.
3
3. Regulatory guidance
Multiple reviews describe the regulatory landscape concerning forced degradation studies and impurities in drug products [2,5,8,10,12–19]. Here, a summary of the essential guidance on forced degradation studies for stability-indicating method development is given. Only guidance from the primary government and international agencies is described in this review. These are the
International Conference on Harmonization (ICH), the World Health Organization (WHO), the United States Food and Drug Authority (USFDA), and the European Medicines Agency (EMA). Since most regulatory agencies are increasingly aligned, the essential guidance on this topic can be found in the ICH guidelines. Hence, this review covers mostly ICH guidelines.
In the guideline ICH Q1A(R2) ‘Stability Testing of New Drug Substances and Products’ [6], the ICH describes the purpose of forced degradation of a new drug substance. The use of stress testing is to identify degradation products and degradation pathways, to determine the intrinsic stability of the drug molecule and to validate the stability-indicating nature of the analytical methodology. Stress testing is performed on a single batch of a drug substance. The conditions for stress testing should include the effect of temperature, humidity, oxidation, photolysis, and hydrolysis across a wide range of pH values. In addition to stressing of the drug substance, stability studies, carried out on drug product outside its immediate container or in another packaging, can be added to the stress testing. These studies might provide useful information or can be used as supporting information. Certain drug product dosage forms (e.g., metered dose inhalers, creams, emulsions, refrigerated aqueous liquid products) might require specific stress testing. An important point in the guideline is that it might not be necessary to examine degradation products that are not found during long-term or accelerated stability studies. When more degradation products need separation from the main peak and each other, the stability-indicating method can become more and more complex.
An integral part of stress testing is photolysis, which is described by the ICH in Q1B [21] ‘Stability Testing: Photostability Testing of New Drug Substances and Products.’ According to this guideline, the photostability testing of a drug substance consists of two parts. The first part is the forced degradation or stress testing and the second part is the confirmatory testing. For confirmatory studies, the sample should be exposed to an overall illumination of not less than 1.2*106 lux*h fluorescent VIS light and integrated near-ultraviolet (UV) of not less than 200 W*h/m2. During forced degradation studies, the drug should be in transparent chemically inert containers. A variety of conditions may be used depending on the photosensitivity of the drug substance and the intensity of the light sources. The applicant can choose the exposure levels used; however, these should be justified. When degradation products observed under stress testing conditions are unlikely to be formed during confirmatory conditions, these products do not need further examination. Forced conditions should be used to develop and verify test methods for confirmatory studies.
The ICH Q2(R1) ‘Validation of Analytical Procedures: Text and Methodology’ [22] stress testing is mentioned to prove the specificity of an analytical procedure when no degradation product standards are available. Samples should be stored under light, heat, humidity, acid/base, and oxidation stress conditions. However, no details on the conditions are mentioned. Currently, the ICH Q2(R1) is under revision, and at the same time, Q14, which describes guidelines for analytical method development, is under construction. Although it seems there are no plans to give more specific stress testing conditions.
Other ICH guidelines mentioning stress testing are Q3A(R2), Q3B(R2), Q5C, Q5E, and M4Q(R1). Q3A(R2) is about impurities in new drug substances in registration applications. Stress testing must part of the application for a new drug product to identify potential degradation products. [23]
4 Likewise, Q3B(R2) is about registration applications of new drug substances. That guidance mentions that the application should include evidence that the analytical method is specific and for specified and unspecified degradation products. The validation should be performed using samples stored under relevant stress conditions. [24] Guidance Q5C and Q5E are explicitly written for
biotechnological/biological products in contrast to the previous documents which are written for small molecule drug substances or products. Stress testing of biological/biotechnological products might be useful to determine the effect of accidental exposures to other than the proposed
reasonable storage condition. In addition, it might be useful to perform stress testing after changing the manufacturing process to determine whether the degradation profile changed. [25,26] M4Q is guidance about the documents needed for registration of a drug product. In those documents results from forced degradation studies must be included and the guidance refers to the above discussed ICH guidance documents.[27]
The WHO states that stress testing conditions differ depending on individual drug substances or products. The guideline ‘Stability testing of APIs and finished pharmaceutical products’ [28] gives the same type of conditions as the ICH guidelines. However, the WHO describes that the extent of degradation should be small. The typical loss should be 10-30% of API as determined by assay. The conditions should be as such that no secondary products form. The API is considered stable under a particular stress condition when, after 10 days, there is a total absence of degradation products. A total absence of degradation products is not clearly defined in the specific guideline. However, it notes that the employed conditions should be justified. If degradation products are proven not to form under long-term or accelerated storage conditions, there is no need for further examination of those degradation products. Another addition is the description of typical stress conditions for Fixed-dose combinations [29]. These typical conditions are shown in Table 1.
Table 1. Typical conditions for stress studies of fixed-dose combination products during pre-formulation studies. Reproduced from ref. [29].
Stress factor Conditions Concentration of APIs1 Time
Heat 60 °C 1:1 (w/w) with
diluent2
1 – 10 days
Humidity 75% RH or greater Solid-state 1 – 10 days
Acid 0.1 M HCl 2:1 (w/v) in 0.1 M HCl 1 – 10 days
Base 0.1 M NaOH 2:1 (w/v) in 0.1 M
NaOH
1 – 10 days
Oxidation 3% H2O2 1:1 (w/v) in 3% H2O2 1 – 3 hours
Photolysis Metal halide, mercury, xenon or UV-B
fluorescent lamp
1:1 (w/w) with diluents2
1 – 10 days
Metal ions (optional) 0.05 M Fe2+ or Cu2+ 1:1 (w/w) with solution of metal ions
1 – 10 days 1When testing the degradability of APIs in combination, the APIs should be in the same ratio as in the FDC-FPP.
2In each case, the diluent is either an excipient or all excipients in the formulation in the same ratios as in the formulation.
Other ratios of diluent may also be appropriate, for example, the approximate ratio in which the drug and excipients will be used in a formulation.
The USFDA mentions in the guidance on ‘Validation of Chromatographic Methods’ [30] that
extraneous peaks should be baseline resolved from the parent analyte. These extraneous peaks may be obtained by the addition of know compounds or by stress testing. Stress testing should be done using acid and base hydrolysis, temperature, photolysis, and oxidations.
5 In the guidance for industry, ‘Analytical Procedures and Methods Validation for Drugs and Biologics,’ recommendations are described to prove the specificity of a stability-indicating method (SIM) [31]. The SIM should undergo a combination of challenges. Some of these challenges are:
1. Samples spiked with target analytes and all known interferences 2. Samples that have undergone various laboratory stress conditions
3. Actual product samples that are either aged or have been stored under accelerated conditions. ‘Stability Testing of Existing Active Substances and Related Finished Products’ of the EMA adds that stress tests are unnecessary for herbal drugs in most cases [32]. Stress testing on active substances should be carried out when no data on degradation products is available in the scientific literature and official pharmacopoeias. The conditions for stress testing, including photostability stress testing, are according to the ICH.
United states pharmacopeia (USP) describes stress testing in several documents [33–35]. In the context of this review, the ‘Validation of Compendial Procedures’ gives the most crucial information about stress testing [33]. Namely, stress testing should be done to prove the specificity of an analytical procedure when impurity or degradation product standards are not available, which is in line with the previously described guidelines.
The Chinese Pharmacopoeia (ChP) describes under Appendix XXC affecting factors testing for drug substance [13]. In Table 2, the conditions recommended by the ChP are given.
Table 2. Conditions of stress testing as per ChP under Appendix XXC.
Type Condition Duration Sampling
time points Limit for additional testing Additional testing conditions High temperature 60 °C 10 days 5, 10 days > 5% potency loss 40 °C High humidity 25 °C
90±5%
10 days 5, 10 days Weight increase > 5% of its weight
25 °C 75±5% Photo stability 4500 ±
500 lx
10 days 5, 10 days
There are no details for strategies and principles of forced degradation studies in the reviewed regulatory guidelines, except for the stress conditions of the WHO for FDC-FPP and the conditions given by the ChP. Also, for guidelines specific problems such as drugs with low/poor solubility or very stable drug molecules, no details are given. There are no recommendations on how much
degradation to obtain for relevant degradation product profiles. Only the WHO states that a 10 – 30% decrease of the main compound is needed; however, it is not clear if this is required to prove the stability-indicating nature of an analytical method. Too much stress might result in irrelevant degradation profiles. When degradation products not forming during stability studies might be irrelevant for SIM development [9] because they can overcomplicate method development. A sound forced degradation study helps to evaluate catalytic conditions under which a drug is unstable. Moreover, it assists in establishing the intrinsic stability of a drug. Next to that is aids in the development and validation of a SIM and to identify and characterize the degradation products and to establish degradation pathways. [13]
6
4. Stress testing strategy for stability-indicating method development
Stress testing is an essential part of stability-indicating method development. Stability-indicating methods are developed to analyze samples generated during stability studies. Stability studies are used to determine the shelf life and storage conditions of a drug substance or drug product.
Degradation products obtained during stability studies can be named ‘actual’ degradation products. Stress testing is used to predict the ‘potential’ degradation products. ‘Potential’ degradation
products are obtained in stress testing studies. Additionally, ‘hypothetical’ degradation products can be predicted before performing experiments to help identify stress testing conditions. In an ideal situation, ‘potential’ degradation products are a subset of ‘hypothetical’ degradation products. In turn, the ‘actual’ degradation products are a subset of the ‘potential’ degradation products. That situation is depicted in Figure 2. However, in realistic situations, not all degradation products might be predicted by theory. [3]
Figure 2. The ideal and realistic situation in degradation product prediction. Reproduced from ref [3].
In Figure 3, a simplified overall approach is presented. This strategy is used to identify likely degradation products by stress-testing the API or drug substance and the drug product. This identification is performed using highly-resolving methods. A highly-resolving method is a generic ‘screening’ method which is not optimized for a particular compound. Such a method can be used for a broad selection of compounds. To gain more confidence in finding all degradation products, a second orthogonal technique can be used. According to Baertschi, reversed-phase high-performance liquid chromatography (RP-HPLC) coupled with UV detection, is mostly used in the pharmaceutical industry. Since compounds without a chromophore are not detected using UV the might be a chance that not all degradation products are found. To detect degradation products without chromophore, other detectors might prove useful during screening. Evaporative light scatter detection (ELSD), refractive index (RI) detection, nitrogen chemiluminescence nitrogen (CLN) detection, charged aerosol detection (CAD), mass spectrometry (MS) and nuclear magnetic resonance (NMR) can be used, where the last two can also aid in structure elucidation.
After obtaining the potential degradation products, relevant degradation products can be identified using accelerated and long term stability studies. Using the results obtained in accelerated stability studies, ‘focused’ methods can be developed that can detect degradation products that are relevant to real-word handling and storage conditions. Finally, degradation product levels can be determined via long-term stability studies from which specifications, as well as storage conditions and shelf life, can be set. [10]
7
Figure 3. The overall strategy for the development of stability-indicating methods. Reproduced from ref [10].
4.1. ‘Hypothetical’ degradation products
The degradation of small-molecule drugs follows many rules defined by classic organic chemistry. Many APIs have several functional groups that can undergo multiple reactions [3]. Therefore, the first step in the development of a SIM is to assess the functional groups of the molecule in question. This identification might help to identify probable degradation pathways [17]. Gathering information on the basic physicochemical properties of a drug can aid in designing forced degradation studies. Properties such as pKa, log P, and solubility in different solvents can help to determine the conditions used in stress testing. Predicted and experimental values can both be used [13]. During this first step, it is also recommended to look at reports on the degradation of drug substances with a similar structure. Besides that, historical data on functional groups can be useful [19]. Since new
degradation pathways are possible, caution is required when predicting degradation products. The initial discussions should be done in a team-based environment with analysts, process chemists, formulators, and discovery representatives present [19].
In addition to in-cerebral approaches, in-silico tools can aid in degradation product prediction. An example of such a tool is Zeneth [36–38], which can predict degradation products based on
knowledge of organic reactions. Others are CAMEO [39], Pharma D3, and DELPHI [40], but the former two were discontinued, and the latter one is not commercially available. Additionally, ASAPprime [41,42] SciFinder is a search database that might be useful since it searches for reactions from molecular structures [43].
8
4.1.1. Zeneth
Zeneth is a computation system to predict forced degradation of ‘small’ organic molecules. [36] It is an information-oriented system, also called an expert system. An expert system emulates the decision making of a human expert. Therefore, the successful prediction of potential forced
degradation reactions of a molecule is limited by the content of the library. Such a library is called the ‘knowledge base’. The knowledge base consists of information from published sources and data from member organizations. The reasoning rules are based on empirical observations. Those observations are supported by the understanding of the transformation mechanism or otherwise by rigorous control.
Figure 4 shows the workflow of Zeneth. [38]
Since Zeneth is an expert system, it uses reasoning rules for its predictions. Also, transformation patterns are available in the knowledge base. A transformation pattern is the smallest part of an organic molecule which undergoes the degradation reaction, also named ‘degradophore’. The system matches the molecule input of the user with transformation patterns available in the
knowledge base. The knowledge base does not contain full structures, only generalized structures by using R groups. For the transformation of a molecule, the software takes into account transformation rules. These rules are made up of reaction mechanisms, geometric considerations and, general reactivity.
The next step in the process is to predict the likelihood of a degradation reaction. Zeneth predicts degradation products using reasoning rules. The system contains two types of reasoning rules, absolute and relative. Absolute reasoning rules are used to predict if the reaction conditions generate a transformation. Relative reasoning rules are used to predict the relative reactivity of competing degradation reactions.
9
Figure 4. The logic flow of Zeneth showing the processing of a starting material into its degradation products. Reproduced from ref [36].
Thermolytic, hydrolytic, oxidative, and photolytic reaction conditions can be assessed. The software uses information on temperature, pH, presence or absence of water, oxygen, metal, radical initiator, peroxide, light entered by the user to predict oxidation, condensation, addition, elimination,
substitution, isomerization, rearrangement, and photochemical reactions. An additional feature is the prediction of intermolecular reactions. For instance, Zeneth can predict dimerization reactions and drug excipient reactions. The system contains a list of common excipients. The user can add excipients or other molecules to obtain predictions on intermolecular reactions.
Kleinman et al. (2014) studied the predictive power of Zeneth compared to stability studies.
Additionally, the authors assessed the growth of the knowledge base and gave guidance for further improvement. They used knowledge-base versions between 2010 and 2012 for their assessment. Also, different software versions were used, starting with version 1 up to version 5. [37]
10 The predictive power was assessed using 27 compounds from 5 different pharmaceutical companies. Those companies shared confidential experimental data from the Zeneth predictions together with the stress testing, accelerated and long-term stability studies. In those studies a total of 191 actual degradation products were observed. The techniques used to determine the degradation products were not disclosed. The predicted degradation products, from Zeneth, were compared with
degradation products obtained during forced degradation studies and stability studies. Using the first knowledge base and first software version, Zeneth was able to predict 31%. That went up to 54% with the newest software and knowledgebase version.
The authors observed some gaps in the software. The prediction was incomplete because the scope of some transformations was too narrow. In some cases, multiple steps can be joined in on a single transformation to improve first-generation degradation products more effectively. Additionally, the knowledge base generates a large number of false-positive degradation products. The reason for this is that the development of the software is focussed on sensitivity, which means that all
experimentally observed degradation products are predicted. It does not focus on specificity, which means that all predicted degradation products are observed experimentally.
4.2. ‘Potential’ degradation products
Several different stress conditions should be tested on drug substances and drug products to identify ‘potential’ degradation products. Figure 5 shows a schematic representation of a complete
degradation study for drug substances as well as drug products. A compound may not necessarily degrade under a given stress condition [19].
Thermal, humidity, photostability, and oxidation stress testing should be performed when the drug product is in the solid-state. If the drug product is a solution or a suspension, hydrolytic degradation should be additionally assessed. Comparing stressed to unstressed samples, and unstressed to stressed placebo samples might rule out degradation products originating from excipients or diluents. [19]
Figure 5. Schematic drawing of stress testing study. Reproduced from ref.[7]
Selecting conditions for a forced degradation study can be difficult, especially for new drug
substances, since there is limited knowledge of the intrinsic stability of the molecule. It makes sense to use a set of standard conditions in this case. Such a set was reported by Jitendra Kumar et al. (2013) and shown in Table 3. Based on the outcome of a first study or an early stability study, the conditions may be adjusted. If the conditions are satisfactory, these may be used throughout the
11 development of a drug product to maintain a consistent approach and to be able to compare the results. [11]
Table 3. Standard stress testing conditions. Reproduced from ref. [11].
Degradation type Experimental conditions Storage conditions
Sampling time (days) Hydrolysis Control API (no acid or base)
0.1 M HCl 0.1 M NaOH
Acid control (no API) Base control (no API) pH: 2,4,6,8 40 °C, 60 °C 40 °C, 60 °C 40 °C, 60 °C 40 °C, 60 °C 40 °C, 60 °C 40 °C, 60 °C 1,3,5 1,3,5 1,3,5 1,3,5 1,3,5 1,3,5 Oxidation 3% H2O2 Peroxide control AIBN AIBN control 25 °C, 60 °C 25 °C, 60 °C 40 °C, 60 °C 40 °C, 60 °C 1,3,5 1,3,5 1,3,5 1,3,5
Photolytic Light 1 × ICH
Light 3 × ICH Light control NA NA NA 1,3,5 1,3,5 1,3,5
Thermal Heat chamber
Heat chamber Heat chamber Heat chamber Heat control 60 °C 60 °C/75%RH 80 °C 80 °C/75%RH Room temp. 1,3,5 1,3,5 1,3,5 1,3,5 1,3,5
4.2.1. Acid/Base
Because of the lack of details on stress testing conditions, Singh & Bakshi (2000) were the first to publish a decision tree system to search for optimal stress conditions of new drugs substances [15]. In Figure 6, such decision trees are shown for hydrolytic degradation under neutral and
acidic/alkaline conditions. Depending on the result, it could be decided to alter the strength of the conditions. The decision trees are limited to drug substances only. The use of drug products instead of drug substances in stress testing might be reasonable for manufacturers that only produce drug products. However, it can give additional information to drug substance stress testing. For instance, the influence of excipients, additives, and packaging on the degradation behavior can be studied [17]. The stress testing of drug products to obtain additional information was appreciated by Reynolds at al. after reviewing the guidance for stress testing studies [9]. A general design for stress testing was published and shown in Figure 6.
12
Figure 6. Decision trees to investigate stress conditions for (A) neutral and (B) acidic or alkaline degradation. Reproduced from ref [15].
Stress testing of drug substance at acidic and alkaline conditions in solution is preferred. The most popular acid is hydrochloric acid (HCl), although, sulfuric acid (H2SO4), could also be used. The recommended concentration of both acids is in the range of 0.1-1 mol/L [11]. Heating might be applied to accelerate the degradation reaction. When the drug is insoluble in water, an appropriate co-solvent may be used, or otherwise the stress testing might be performed in suspension. Alcohols should not be used as co-solvent during acid stress testing. DMSO, acetic acid, and propionic acid might prove useful. [19]
For alkaline stressing, usually NaOH, LiOH, or KOH (0.1 – 1 mol/L) are used [11]. Stressing is
preferably performed in the solution state. However, when the drug is insoluble, stressing could be performed in a suspension. Otherwise, an appropriate co-solvent could be used to dissolve the drug. Glyme(dimethoxymethane) and 1,4-dioxane both facilitate reactions in alkaline conditions. Heating can be used to accelerate the degradation reaction [19].
Figure 7. pH ranges for acid-base-solution stress testing of (A) drug substance and (B) drug product from 19 companies out of 20 surveyed. Reproduced from ref. [12]
In a survey amongst twenty companies carried out in 2013, acidic and alkaline stress testing was performed by nineteen companies [12]. All of these nineteen companies performed acid and base stress testing on drug substances. Several pH ranges were used, as showed in Figure 7. The
specification of pH seems to be a different approach than the earlier discussed decision trees, where no pH is specified; only a certain concentration of acid or base was described. Final pharmaceutical products are acid-base-solution stress-tested by 13 out of 20 surveyed companies. 30% cover pH from 1-13, 30% use low (0-2), mid (5-9) and high (12 – 13) ranges, 25% only middle range and 15% only outer ranges.
13 Hydrolytic reactions can be accelerated using heating. In the previously mentioned survey, most companies indicated they use a range between ambient and 70 °C. However, higher temperatures up to and above 90 °C were used by multiple companies as well. Most companies used co-solvents, of which acetonitrile (ACN) and methanol were mostly used. The duration of these stress testing studies may go up to 25 days or even longer.
4.2.2. Oxidation
Oxidation is one of the major pathways in drug degradation. Multiple mechanisms can oxidize drug molecules. Auto oxidation, peroxidation, electron transfer, and photo-oxidation are a few examples. These types of oxidation are mediated by free radicals, peroxides, transition metals, and singlet oxidation, respectively. Autoxidation occurs in both solution and solid-state and can initiate a chain process that consists of 3 phases: initiation, propagation, and termination. Different initiators were described by Hovorka & Schöneich [44].
There are several different approaches to perform oxidation stress testing. Peroxides, oxygen pressure, free radical initiators, metal ions, electrooxidation, potassium permanganate, Tween 80 with Fe(III), Glyme with Fe(II), singlet oxygen and N-Methylpyrrolidone are known in oxidation stress testing [45]. As in most types of degradation studies, a co-solvent may be used when the drug is not soluble in water. However, the use of ACN should be avoided since it can lead to unrealistic
degradation products due to the formation of peroxycarboximidic acid [46]. Additionally, heating can accelerate the degradation reaction.
Of 20 companies surveyed, 65% perform oxidation stress testing [12]. Besides the use of peroxide, five companies used radical initiators, three used transition metals, three used pressured oxygen, and two used bubbled oxygen. The studies were carried out using temperatures ranging between ambient and 30 °C for peroxides and bubbled oxygen, between 31 °C and 40 °C for radical initiators and transition metals and > 50 °C for pressurized oxygen. The duration of the studies ranged between 1 and 14 days.
Stress testing using an oxygen atmosphere can be carried out under a pressure of 50 – 300 psi. Solid-state stress testing can be done using this condition, one closed container with an oxygen headspace compared to a closed container with a headspace of argon or nitrogen [19].
4.2.2.1. Peroxides
Hydrogen peroxide the mostly used to apply oxidative stress to samples. Peroxide stress testing is useful in drug product stress studies if H2O2 is an impurity. Similarly to the acid and base stress testing, a decision tree for oxidative stress was developed [15]. In that decision tree, the starting point for oxidative stress is the use of 3% hydrogen peroxide for 6 hours at room temperature. Based on the outcome, the harshness of the stress could be increased or decreased. Typically, hydrogen peroxide can is used at a concentration between 1 – 3% [11–13,19]. According to other approaches, up to 30% of hydrogen peroxide can be used [15,17].
Although the widespread use of peroxides in oxidative stress testing, there are some limitations. For example, at ambient temperature, there are no free radicals present. Consequently, no autoxidation can be predicted using peroxides. Additionally, peroxides decompose in hydroxy radicals when heated. Hydroxy radicals are more reactive than hydroperoxyl radicals and can be able to generate irrelevant degradation products [45].
4.2.2.2. Radical initiators
Another type of oxidative condition is a free radical initiator. Azobisisobutyronitrile (AIBN) is the most commonly used radical initiator [11]. Typical predictive conditions are 1-5 mM AIBN at 40 °C under
14 atmospheric oxygen atmosphere [47]. Under heating, two alkyl free radicals are generated. These free radicals react rapidly with oxygen to form peroxy radicals. This condition mimics the
autoxidation pathway of drug molecules. Hence, AIBN can be used to predict autoxidation in drug products. Another well-known free radical initiator is 4,4′-Azobis(4-cyanovaleric acid) (ACVA). Complementary to AIBN as a radical initiator for oxidation stress studies is the use of a combination of Tween 80 and Fe(III) [48]. Tween 80 contains impurity organic hydroperoxide groups, which can be readily oxidized by Fe(III). The obtained peroxy radicals react with the drug molecules. In a
comparative study, 18 molecules were oxidize using AIBN and Tween 80/Fe(III). The results of both approaches were in agreement, showing differences in the extent of degradation.
Most of the oxidation stress studies are performed in the solution-state. These studies are potentially non-predictive for solid dosage forms. Therefore, the Tween 80/Fe(III) was used as a solid-state forced oxidation system. A homogeneous mixture is prepared by dissolving the drug substance in an organic solvent in combination with Tween 80 and Fe(III). Subsequently, the solvent is removed by evaporation. The solid samples are exposed to long-term stability conditions and tested at several time points by, for instance, HPLC to determine the chemical stability. The forced degradation approach generated predictive oxidation products in solid dosage forms for long-term stability studies, proven by multiple model compounds. [49]
4.2.2.3. Transition metals
Metal ions can indicate whether the API is catalytically oxidized or can act as radical initiators of autoxidation [19,45]. As well, metals can react with peroxides to form hydroxyl radicals in a Fenton-type reaction.
KMnO4 is not widely used; however, it might be useful in some cases. It is known that KMnO4 can convert aldehydes and aromatic methyl groups to acids, secondary alcohols to ketones, and olefins to vicinal diols. Singlet oxygen is an excited state of oxygen which forms thermally or by
photochemical sensitization of the drug molecule. Forced degradation studies using singlet oxygen are usually performed to confirm reaction mechanisms. [45]
4.2.2.4. Electrooxidation
The electrochemical oxidation was carried out by pumping a solution of 0.086 mg/mL of the model compound in 10 mM ammonium acetate through a thin-layer electrochemical cell, fitted with a glassy carbon electrode. Three different pH were tested. The applied potential on the electrode was increased from 0 to +2 V in 100 mV steps. Samples of the effluent were collected and analyzed offline by HPLC-UV-MS. [50]
Oxidative forced degradation was performed with a concentration of 0.86 mg/mL. Both hydrogen peroxide at 0.3 % in 1:1 (v/v) methanol: water and AIBN at 5 mM in (v/v) ACN: water were used to stress the drug. Zeneth 6.0 was used for in-silico predictions. The oxidation products obtained with chemical oxidation and electrooxidation were compared to the results of a previous accelerated stability study and in-silico predictions.
Electrooxidation of the drug substance resulted in a total of 14 oxidation products. From these oxidation products, four were predicted by Zeneth, three were obtained in an accelerated stability study and two were observed using chemical oxidation. This shows that there is a possibility that a significant number of non-relevant oxidation products can be obtained by electrooxidation. Developing a stability-indicating LC-UV method can be more difficult for 14 degradation products than for 5. Optimizing a method to baseline separate 15 compounds can take longer than when only 5 compounds need to be separated on an RP-HPLC column. Next to that the method may have a
15 longer runtime, which costs more time for quality control (QC) laboratories and may be more
expensive to run. On the other hand, some relevant impurities found by electrooxidation were not found using chemical oxidation. Therefore, the electrooxidation approach might be beneficial for other studies such as structure elucidation studies.
4.2.2.5. N-Methylpyrrolidone
N-Methylpyrrolidone mixed with water under oxygen or air at elevated temperatures can be used as oxidative stress testing conditions. NMP oxidizes in the presence of oxygen and heat to form a hydroperoxide and an imide. APIs can be dissolved in a mixture of NMP and water at a concentration of 1 – 10 mg/mL. The solution can be placed in stoppered vials and incubated at 60 – 80 °C. The headspace is flushed with oxygen or nitrogen in order to differentiate between hydrolytic and oxidation generated degradation products. [45]
Four case studies were presented whereby NMP was used to perform oxidative stress. In three out of four case studies, the oxidation degradation was predictive for accelerated or long-term conditions. For compound A, hydroperoxides were found during stress testing. These hydroperoxides were found later in bulk API and stability samples of the API. For compound B, three secondary degradation products were the main degradants in drug product stability. The formulation of compound B contained polyvinylpyrrolidone and crospovidone, which have similar oxidative
properties as NMP. For compound C a genotoxic hydroperoxide was found, which was found later on during stability studies at levels below toxicological concern. For compound D, the results of the stress test differed from the drug product API results. Hence, it was not predictive of the formulation. [45]
The advantages of using NMP are that hydrolytic and oxidative processes co-occur. These processes can predict if degradation products are forming in a formulation by both oxidation and hydrolysis in multiple steps. The comparison between nitrogen and air or oxygen helps in distinguishing between hydrolysis and oxidation pathways. NMP does not interfere with UV or MS detection in
chromatography. Solubility is enhanced. Various oxidation mechanisms can occur. NMP is not volatile, which makes mass balance studies possible even at elevated temperatures. Since NMP is structurally similar to polyvinylpyrrolidone (PVP) and crospovidone, it can be predictive when these two excipients are present in the formulation
There are also disadvantages. There will be an interference in the void volume peak using LC with UV detection at wavelengths below 280 nm. It will be challenging to determine whether degradation products elute in the void volume. Reactions of the API with NMP can cause another problem.
4.2.3. Photostability
Several functional groups are susceptible to photodegradation. These include carbonyl, nitroaromatic, N-oxide, alkenes, aryl chloride, weak C-H and O-H bonds, sulfide, and polyene functional groups [11].
Photostability should is tested by exposing the sample to 1.2 · 106 lux·h of visible light and 200 W·h/m2 of near-UV radiation (320 - 400 nm). The first condition mimics a continuous exposure of the sample to visible light without protective packaging for three months. This amount of light exposure to the drug could occur in the pharmacy, at the warehouse or home. The latter condition could occur in the same three months by solar radiation behind the window glass. However, 200 W·h/m2 of near-UV radiation is roughly the exposure of 1 – 2 days in the windowsill. For forced degradation studies, no requirements on the exposure are defined. There are two options for photostability testing. Using option 1, daylight, or indirect daylight filtered through window glass is mimicked. With option 2, an
16 attempt is made to mimic indoor lighting conditions. The different options of light sources might result in a difference in photostability results. [16]
The substance should at first be subjected to an illumination up to 1.2 · 106 lux·h, which is recommended by the ICH. The exposure can be increased up to 5 times. If, in that case, the degradation is negligible, the drug substance can be declared photostable [15]. Moreover, when applying 2 – 5 times the ICH recommended conditions, the mechanisms of photodegradation can be more fully understood [3]. In another article, the authors recommended that at least two times the ICH condition should be applied for drug substance stressing to ensure the sample is adequately exposed to light. [19]
Photostability conditions performed by companies are given in Table 4. 19 out of 20 companies perform photostability studies on the drug product. Of these companies, 14 perform photostability testing in solution. When light stress is tested in the solution state, four companies test at more than one pH when the molecule has ionizable groups. ICH option 1 is the most popular: 18 companies use that option. Two companies use both options for all studies. Sixteen companies perform both stress testing and confirmatory studies. The results show that photostability is an integral part of stress testing.
Table 4. Conditions used for photostability stress testing. Reproduced from ref.[12].
# companies # companies
ICH dose 1.2 * 106 lux h 200 W h/m2
ICH typical maximum typical maximum
1x 12 2 12 3
1 to 2x 2 5 1 3
2 to 5x 3 5 2 5
5 to 10x 1 5 2 5
> 10x 1 2 1 2
La Cruz et al. proposed an experimental and data-analysis workflow that results in a mathematical model to predict the photolytic degradation of solid materials. From this model, the degradation of a compound could be estimated against any light source. The model was applied to light source characteristics in manufacturing environments. From these results, a control strategy could be made in order to stay below a specified amount of degradation during manufacturing. [51]
4.2.4. Thermal and humidity stress conditions
It is recommended to perform thermal/humidity exposure above 50 °C and 75% RH. Using Arrhenius kinetics, temperature and duration can be estimated. For the calculation, an activation energy of 62 kJ/mol (50-75 kJ/mol) is typically used. When the temperature is 70 °C or above, deviation from Arrhenius kinetics may be expected [19]. It is useful as a scientific rationale to complete a forced degradation study even when 5 – 20% degradation is not reached [3].
The energy required can be estimated by making kinetic assumptions using the Arrhenius equation. 𝑘 = 𝐴𝑒−𝐸𝑎/𝑅𝑇
where k is the reaction rate constant in s-1 for a 1st order reaction or in M-1 s-1 for a 2nd order reaction, A is a frequency factor in s-1 for a 1st order reaction or in M-1 s-1 for a 2nd order reaction, E
a is the energy of activation in J mol-1, R is the gas constant (8.314 J mol−1 K−1 ), and T is the absolute temperature in K.
17
Table 5. Calculated days of storage for several temperatures to obtain the same amount of energy as for accelerated conditions.
Temperature (°C) Duration (days)
40 180 50 90 60 45 70 23 80 11 90 6 100 3
The activation energy of most drug substances ranges from 50-100 kJ/mol. A doubling of the reaction rate is achieved for every 10 °C, assuming 50 kJ/mol activation energy. The storage time at a specific temperature can be calculated, using that assumption, to reach the same amount of energy as for accelerated conditions. In Table 5, the duration of stress testing, to compare to accelerated stability, for several storage temperatures is listed.
4.2.5. Samples per condition
Per stress conditions, four samples should be prepared. These four samples consist of untreated blank, stressed blank, untreated sample, and stressed sample [15]. By comparing these samples, it can be determined whether observed peaks in a chromatogram are drug or non-drug related. It is recommended to compare stressed drug product with stressed placebo and stressed drug substance to identify drug-excipient reactions [9]. Additionally, during stress testing, samples can be withdrawn at multiple time points to determine the stability of the degradation products [17].
The concentration to perform stress testing should be about 1 mg/ml wherever possible [15,17]. At such a concentration, minor degradation products are easily detected. If the drug is insufficiently soluble, a co-solvent could be used to increase the solubility. Different amounts of methanol could be used [17]. However, each organic solvent might potentially react with the drug under a given stress condition [9]. Therefore, stress testing in suspension might be an option when the drug is not soluble.
When chromatography is used, the sample could be further diluted before sample injections. Another option is to neutralize the samples before injection. Neutralizing a sample for example can be achieved by adding an equal amount of acid to the amount of base present during stressing. However, this cannot be done quantitatively, and it could happen that dissolved components precipitate.
Reaction monitoring can be of aid in establishing stability-indicating methods to be used in stability studies under accelerated and long term conditions. Kinetic time points to determine primary degradants and to understand the pathway of degradation. [19]
4.2.6. Mass balance
Mass balance is the correlation of the measured loss in the amount of the drug with the measured increase in amount of degradation products. Mass balance can be determined by comparing the total area of all peaks obtained in a stressed sample to the total area of an unstressed reference sample. The total peak area in a stressed sample should be close to the total peak area in a non-stressed sample. For instance, an unstressed sample results in one chromatographic peak with a peak area of 90. If the stressed sample, with the same original concentration contains 3 peaks with peak areas of
18 12, 20 and 55, the mass balance can be calculated as follows: (12+20+55)/90*100 % = 96.7 %. In industry, the mass balance has much attention. However, it is not an explicit requirement in the guidelines. [13].
Assessing the mass balance can help in determining the adequacy of the analytical method. The analyst should investigate a substantial mass imbalance. Not in all cases, mass balance can be obtained for the degradation sample, and therefore, the emphasis can be put on the thoroughness to determine the specificity of the analytical method and the degradation pathways. The seven main causes for failing mass balance given by Nussbaum et al. are [52]:
Degradation products do not elute from the HPLC column The detector does not detect all degradation products
Degradation products are lost from the sample matrix (not recovered/extracted or due to volatility)
The parent compound is lost from the sample matrix Degradation products co-elute with the parent compound
Degradation products not observed in the chromatogram due to poor chromatography. That can be seen as a drifting baseline rather than separate peaks.
Difference in response factor between the degradation products and the parent compound. For these reasons, mass balance in SIM development can be complicated. Understanding of the structures of both the reactant and the degradation products is essential to develop reliable SIM [3]. Use of ELSD, CAD, RI, MS, and NMR detection hyphenated to LC for evaluation of mass-balance and peak purity can aid in assessing the mass balance.
When there is a mass-balance deficit, relative response factors (RRF) can be determined for
degradation products to obtain mass balance. The determination of RRFs might be difficult when the degradation products are instable. Multivariate regression can aid in the RRF determination of unstable degradation products [53]. Campbell et al. determined the RRF using multivariate regression of an unstable degradation product of Molibresib generated by acid stressing. The degradation reaction is readily reversible upon neutralization of the sample, making it challenging to isolate the degradation product. It was possible to fit a multiple linear regression model to a dataset containing several levels of the degradation product, to estimate the RRF. The mass balance was close to 100% after the RRF correction of the degradation products.
4.3. ‘Potential’ vs. ‘actual’ degradation products
In this section, relevant forced degradation conditions are sought. This search was done using case studies from literature. Especially studies describing forced degradation and stability study
experiments were used. To determine whether forced degradation conditions are relevant, the results from those studies are compared to results obtained during stability studies, either accelerated or long-term. When degradation products observed at a specific forced degradation condition and not observed during stability studies, such a forced degradation condition is likely not relevant. On the other hand, when all forced degradation conditions are unable to generate a degradation product observed during a stability study, the forced degradation condition can be partly irrelevant as well.
4.3.1. Enalapril maleate
Bhardwaj & Singh developed a validated SIM for enalapril maleate using stress testing studies according to the ICH requirements [54]. The drug was hydrolytically stressed at a concentration of 2 mg/mL at 80 °C. The concentrations of acid and base were 0.1 mol/L HCl and 0.1 mol/L NaOH.
19 Stressing was terminated when about 20% of degradation was obtained. Oxidative stress was carried out at a concentration of 2 mg/mL with 3% H2O2 for eight days at room temperature. Photolytic stress testing of enalapril maleate was done both in the solution state and the solid state. For the solution state, the drug was dissolved in 0.1 mol/L HCl or NaOH at 2 mg/mL. The samples were subjected to 1.25 x 106 lux*h of visible light and 210 W*h/m2 of near-UV light. A 1 mm layer of the solid sample in a petri dish was subjected to the ICH light dose at 40 °C and 75% relative humidity. The solid drug was also stressed using dry heat at 50 °C for 60 days. Accelerated stability testing was performed for three months using the recommended ICH conditions of 40 °C and 75% relative humidity
Table 6. Overview of stress testing and accelerated stability conditions for enalapril maleate combined with results obtained.
Type Solution
and/or solid
Condition Degradation
Acidic Solution 0.1 mol/L HCl at 80 °C ~20 % in 18 h
5 degradation products (I-V) Alkaline Solution 0.1 mol/L NaOH at 80 °C 100 % in 3 h
1 degradation product (II)
Neutral Solution Water at 80 °C ~20 % in 2 days
4 degradation products (II – V) Oxidative Solution 3% H2O2 for 8 days and
30% H2O2 for 5 days
No degradation observed Photolytic Solution
Solid
1.25 * 106 lux*h of visible light and 210 W*h/m2 at 40 °C in 0.1 mol/L HCL and,
1.25 * 106 lux*h of visible light and 210 W*h/m2 at 40 °C in 0.1 mol/L NaOH
1.2 * 106 lux*h of visible light and 200 W*h/m2 at 40 °C and 75% RH
Unknown %
2 degradation products (II and V)
No degradation observed
No degradation observed
Thermal Solid 50 °C No degradation observed
Accelerated stability
Solid 40 °C and 75% RH for 3 months Slight degradation
2 degradation products (II and V) A total of five degradation products was obtained from the forced degradation studies. Enalapril and the degradation products were determined using RP-HPLC using a PDA detector. The separation was achieved using a C18 column (250 mm × 4.6 mm, 5µm) with a gradient of a phosphate buffer (0.01M, pH3) and ACN. The characterization of the compounds was done using LC-MS using the same column and mobile phase gradient where phosphate buffer was replaced by water adjusted to pH 3 with formic acid. Figure 8 shows a chromatogram of enalapril and its degradation products.
20
Figure 8. Chromatogram of enalapril and its degradation products. Reproduced from ref. [54].
Acidic conditions resulted in about 20 % degradation after 18 h. Five degradation products were obtained. Using LC-MS, the authors proposed the structures and degradation pathways. Two of the five degradation products turned out to be secondary degradation products. These two secondary degradation products and one primary degradation product were not observed during the
accelerated stability study.
The acidic and alkaline stress conditions might have been too harsh. Because irrelevant degradation products were obtained for the alkaline condition, and total degradation occurred using the alkaline condition. On the other hand, the accelerated stability study was done up to 3 months, whereas the ICH recommends six months. In the additional three months, more degradation products could have occurred. Also, stress testing was not carried out at steps of 10 °C above the ICH accelerated stress testing conditions. Besides, only peroxide was used by the authors as an oxidative stress condition. Subsequently, no autooxidation products will be obtained with oxidative stress testing.
21
Table 7. Result of forced degradation studies performed by Toporišič et al. Reproduced from ref [55].
Conditions RRTa nd labels of degradation products
0.29 0.6 0.72 0.78 0.82 0.86 1.08 1.33 1.78 2.12
I II IIIa IIIb IV IIIc IIId V VIa VIb
Room temperature Level of degradation products (%)
1 M NaOH 65.90 1 M HCl 0.14 3% H2O2 0.13 1.00 10 mM MMPP 0.70 2.23 0.06 1.58 0.30 0.14 Water 0.14 0.32 1.90 Solvent 0.14 0.32
60 minutes at 60 °C Level of degradation products (%)
1 M NaOH 74.60 1 M HCl 0.46 3% H2O2 0.15 0.34 1.05 10 mM MMPP 0.69 2.29 0.05 1.63 0.29 Water 0.24 0.30 1.00 Solvent 0.24 0.34 0.12
10 minutes at 100 °C Level of degradation products (%)
1 M NaOH 71.74 0.05 1 M HCl 0.65 0.32 0.71 3% H2O2 0.27 0.18 0.18 0.18 0.20 0.71 10 mM MMPP 1.47 0.08 0.09 0.05 0.55 0.14 6.21 0.30 0.17 0.43 Water 0.38 0.32 0.32 Solvent 0.30 0.32
aRetention time relative to the retention time of elanapril
In another stress testing study, five additional impurities were obtained [55]. Magnesium
monoperoxyphthalate (MMPP) was used in addition to H2O2 for oxidative stress. The results of that stress testing are shown in Table 7, together with other forced degradation tests. The degradation products and elanapril were determined using HPLC. The separation was carried out using a polymeric PLRP-S column (250 mm x 4.6 mm, 5 µm) with a gradient of ACN and 0.02 M phosphate buffer at pH 6.8. Detection was done using a UV detector at 215 nm. A chromatogram of enalapril and degradation products is shown in Figure 9.
22
Figure 9. Chromatogram of enalapril treated with MMPP. Known impurities are marked I, II, V, VIa and VIb. New impurities are marked IIIa, b, c, d and IV. Reproduced from ref [55].
MMPP was very effective in degrading the drug substance compared to the other conditions. However, much more impurities were generated compared to an accelerated stability study, where only three degradation products formed during accelerated stability. Therefore, it is not likely that MMPP will generate relevant degradation products. Hence, it would not be a suitable approach to use in SIM development.
4.3.2. Exemestane
Kumar et al. [56] performed stress testing on exemestane bulk samples. The stress samples were generated to aid in LC method development. The conditions and results are listed in Table 8. During stress testing, four major degradation products (imp-5, imp-6, imp-7, and imp-8) were obtained in addition to four process impurities (imp-1, imp-2, imp-3, and imp-4). Additionally, unknown degradation products were observed during stress testing.
Figure 10. Overlay of chromatograms of examestane and degradation products. x-axis represents retention time in minutes. (A) peak identification solution. (B) Acid degradation. (C) Oxidative degradation. (D) Base degradation. (E) Thermal
23 The degradation products were determined using HPLC. A C18 (150 mm x 4.6, 3µm) was used for the separation. A gradient was used with water and methanol to achieve separation. UV detection was used and the chromatograms were recorded at a wavelength of 247 nm. Chromatograms of several samples overlaid are shown in Figure 10. In addition to known degradation products, unknown degradation products were observed in stressed samples. Only major degradation products were marked.
Stress testing of exemestane using 2 M NaOH at 70 °C for 3 days resulted in a major degradation product (imp-5). Two major degradation products of exemestane were formed during oxidation stress testing using 3% H2O2 at 60 °C for 3 days (imp-6 and imp-8). One major degradation product was obtained by acid hydrolysis using 2 M HCl at 70 °C for 3 days. All those obtained degradation products were not obtained during an accelerated stability study for 3 months.
Table 8. Results of Exemestane stress testing. Reproduced from ref [56].
Stress condition Time Assay of API (%, w/w) Total impurities (%, w/w) Mass balance (%, w/w) Remarks Thermal (105 °C)
10 days 98.2 1.6 99.8 Imp-1 and Imp-2 were major
degradation products Acid (2 M
HCl, 70 °C)
3 days 95.5 4.0 99.5 Unknown degradation products
formed. Major degradation product designated as Imp-7 Base (2 M
NaOH, 70 °C)
3 days 97.1 2.6 99.7 Unknown degradation products
formed. Major degradation designated as Imp-5 Oxidation
(3 % H2O2, 60 2 M HCl, 70 °C)
3 days 95.2 4.4 99.6 Unknown degradation products
formed. Major degradation designated as Imp-6 and second major degradation product is Imp-8
Since none of the obtained degradation products obtained during stress testing were seen during accelerated stability testing, those degradation products might be irrelevant, although there might be a possibility that those impurities might be obtained during extended accelerated stability studies or long-term stability studies.
Since the obtained degradation products are not relevant, it can be concluded that the stress testing conditions were too harsh to be predictive for stability studies. Heating the sample in combination with H2O2 results in hydroxy radicals, which are known to cause irrelevant degradation products. Moreover, the thermal stress testing was done at 105 °C, whereas thermal stressing at 100 °C for 3 days is already similar in energy as with accelerated stability conditions for 6 months. Similarly, the concentration of acid and base in the hydrolysis samples is too high. The pH probably obtained using those concentrations are unrealistic.
Additionally, no photostability, neutral hydrolysis, and radical initiator oxidation were tested. Therefore it is difficult to determine whether the method is genuinely stability-indicating. At least, not all the requirements from the ICH are fulfilled.
24
4.3.3. Alizapride
A forced degradation study was performed to develop a SIM for Alizapride (AL) tablet and ampoules formulations [57]. Drug substance was stressed using acid, alkaline, and neutral hydrolysis, oxidation using hydrogen peroxide and metal ions, thermal stressing, and photolytic stressing. Details on those conditions are listed in Table 9. The stress testing was done for 1 mg/mL. The conditions were adjusted to obtain 10 – 25 % loss of the parent compound. The adjustments were made to avoid unrealistic degradation.
AL and degradation products were separated using a C18 column and a gradient of an aqueous sodium acetate buffer (20 mM, pH 4.0) and methanol. UV detection at a wavelength of 225 nm was used. A chromatogram of a standard including degradation products and a chromatogram of a long-term stability study sample are shown in Figure 11. Structure elucidation of the degradation products was done using LC-MS/MS and 1H and 13C NMR.
Table 9. Stress testing conditions and results of AL.
Stress condition Time
(hours)
Degradation
1 M HCl at 70 °C 72 Polar degradation product formed (AL-CA)
0.5 M NaOH at 70 °C 48 Polar degradation product formed (AL-CA)
Water at 70 °C 72 No results were disclosed for this condition
30 % H2O2 at room temperature 48 Two major oxidation products were observed.
1.5 mM Cu2+ at room temperat 120 No oxidation 1.5 mM Fe3+ at room temperat 120 No oxidation
Thermal 60 °C 48 No degradation
Thin layer of 50 mg exposed to daylight 72 No degradation Aqueous solution exposed to daylight 72 No degradation Thin layer of 50 mg exposed to UV-light 8 No degradation Aqueous solution exposed to UV-light 8 No degradation
From the forced degradation study, three degradation products were observed. Of these degradation products, one was formed during both acid and alkaline hydrolysis. The other two degradation products were obtained using hydrogen peroxide stress and were determined to be diastereomers. No degradation products were obtained in photostability, thermal, and metal ion conditions.
Thermal stress testing was done at 60 °C for 3 days. From the Arrhenius equation follows that stress testing at 60 °C for 45 days is comparable to accelerated stability conditions. Therefore, the used thermal stress conditions might not be predictive enough. On the other hand, no degradation products were obtained during a long-term stability study (t = 24 months) of AL tablets (Figure 11). Furthermore, the oxidative and hydrolytic stress conditions are not generating relevant degradation products, since the degradation products obtained in those conditions were not obtained during a stability study.
