ABSTRACT
Pharmaceutical impurities are any components of a new drug substance or a new drug product that are not the drug substance itself or excipients in the drug product [1,2] Genotoxic impurities are a subset of impurities that are mutagenic or carcinogenic To differentiate from genotoxic impurities, nongenotoxic impurities are often referred to as regular impurities The thresholds for reporting, identifying, and qualifying regular impurities are recommended by International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) Q3A [1] and Q3B [2] and are routinely applied during latestage development and at registration in the pharmaceutical industry For control of genotoxic impurities, the generic thresholds threshold of toxicological concern (TTC) and staged threshold of toxicological concern (sTTC) are recommended in the European Medicines Agency (EMA) [3] and Food and Drug Administration (FDA) [4] genotoxic impurity guidelines and the EMA questions and answers document [5]
151 Introduction 427 152 Risk of Genotoxic Degradation Products 429 153 General Strategies for Control of Genotoxic Degradation Products 430 154 Identification of Degradation Products for Genotoxity Risk Assessment 433
1541 Identification of Actual Degradation Products 434 1542 Identification of Potential Degradation Products 435
15421 Identification of Potential Degradation Products via Accelerated Stability Testing 435
15422 Identification of Potential Degradation Products via Stress Testing 436
15423 Prediction of Potential Degradation Products Using Chemical Knowledge 442
Conclusion 444 Acknowledgment 444 References 444
Although the ICH Q3A/B thresholds (eg, identification threshold [IT]) are intended for regular impurities and not for genotoxic impurities, in practice they still play an irreplaceable role in the identification and risk assessment of genotoxic impurities Thus, it is important to understand the relationships between the ICH thresholds and the genotoxic limits/thresholds, which can be illustrated using the “impurity iceberg,” as shown in Figure 151
Per ICH Q3A/B recommendations, regular impurities at a level above the reporting threshold are determined and accounted for in the total impurities and reported to the regulatory agencies In this sense, the reporting threshold is the “sea level” (see Figure 151) and impurities at the level above it are the “visible” ones and form the tip of the impurity iceberg In contrast, impurities at levels below the reporting threshold, although they can be significant in number and form the main body of the impurity iceberg, are “invisible” from a regulatory point of view Because genotoxic impurities are considered to be unusually potent, not only the visible impurities but also many of the invisible impurities must be considered in the identification and assessment of genotoxic impurities As a matter of fact, impurities below the reporting threshold but above the sTTC or TTC are the main objects of genotoxic impurity risk assessment
However, because the sTTC/TTC can be orders of magnitude lower than the ICH Q3A/B reporting threshold, depending on the maximum daily dose (MDD), identification and assessment of these impurities for their genotoxic potential is a great challenge
The concepts of actual impurity and potential impurity are of great importance in the pharmaceutical assessment of genotoxic impurities Although mentioned in ICH Q3A and routinely used in the industry, the concept of actual impurities is not clearly defined in regulatory guidelines In practice, actual impurities are typically considered to be those that are exceeding the reporting threshold because they are determined (hence, actual) and reported to regulatory agencies This concept of actual impurity is now included in ICH M7 step 2 document [6] ICH Q3A defines potential impurities as those “that theoretically can arise during manufacture or storage It may or may not actually appear in the new drug substance” In routine application, this definition is qualitatively meaningful but lacks clarity in the quantitative aspect on “to what level the impurity can theoretically arise” One school of thought or
perhaps the mainstream thought is that potential impurities are those that can theoretically or potentially arise in a drug at levels exceeding the reporting threshold, that is, potential impurities are those predicted to have the potential to be actual impurities before actual impurity data are available As such, often after the actual data are available a predicted potential impurity that is not actually observed above the reporting threshold can be no longer considered a potential impurity This concept makes a lot of sense in the ICH Q3A/B paradigm, because for a regular impurity if it does not exceed the reporting threshold it is not even considered an impurity Quite obviously, this concept is no longer appropriate for a genotoxic impurity because the threshold of concern for the genotoxic impurity is no longer the reporting threshold; rather, it is the TTC or sTTC So, for a thorough assessment of genotoxic impurities a potential impurity should be reasonably expected to be at a level exceeding the TTC or sTTC This logic is illustrated in Figure 151, where the impurities between (s) TTC and IT are labeled as “Assessed as potential impurities”
Based on their origin, two types of impurities, process-related impurities and degradation products, are differentiated and typically separately assessed for their genotoxic potential The former refers to those that are introduced or formed during the drug substance synthetic process (eg, starting materials, intermediates, reagents, and by-products) and is out of the scope of this chapter The latter refers to the impurities that are formed due to the chemical degradation of the drug substance and drug product during manufacture, in use, and during storage The purpose of this chapter is to provide an overview of the challenges and best practices in the control of genotoxic degradation products with a focus on the identification of actual and potential degradation products for assessing their genotoxic potential
In general, drug degradation is not considered the major source of genotoxic impurities As chemically reactive species are intentionally brought into the manufacturing process of a drug substance, most of the genotoxic impurities are process-related impurities Table 151 provides a snapshot of all the Ames-positive impurities that we (Boehringer Ingelheim Pharmaceuticals, Inc [BIPI, Ridgefield, Connecticut]) have encountered between 2005 and 2010 Among 37 structurally unique Amespositive impurities in various drug substances and products, 30 were introduced or
TABLE 15.1 Occurrence of Ames-Positive Impurities (between 2005 and 2010) at BIPI
formed during the chemical synthesis of drug substances and only 7 were degradation products (and a few of these were also intermediates) Since 2010, the overall ratio of Ames-positive degradation products was significantly decreased as a result of increased diligence in avoiding the use of Ames-positive building blocks that can potentially end up in the final drug substance and be subsequently generated as Ames-positive degradation products (eg, aromatic amines) As a matter of fact, no new genotoxic degradation product has been identified since 2010 among all the new BIPI new chemical entity (NCE) projects This observation demonstrates that it is possible to significantly minimize the risk of genotoxic degradation products using the quality-by-design principles (ie, quality should be built in by design) advocated in ICH Q8 [7] during the lead identification and optimization stages
Nonetheless, for the following reasons a risk assessment of genotoxic degradation products is still necessary First of all, due to the variety of drug molecules and the complexity of possible degradation pathways, it is not always possible to predict and avoid the risk of formation of genotoxic degradation products during lead identification and optimization Second, unlike a process impurity, a degradation product once formed cannot be purged or removed In other words, the consequence of a genotoxic degradation product can be severer than that of a genotoxic process impurity because the former can be much more challenging to control Further, the level of a process impurity does not increase and sometime decreases (eg, ethyl methanesulfonate [8]) during storage In contrast, the level of a degradation product can increase with time, which means that the level of a genotoxic degradation product needs to be controlled not only at release of the drug substance or product but also during the whole shelf life of the product Thus, understanding the chemistry and rate of formation of the degradation product is essential for its control
The most effective approach in controlling the risk of genotoxic degradation products is to prevent them from forming by engineering the chemical structure of the drug candidate during the lead identification and optimization stages This can be achieved by proactively identifying and subsequently avoiding the use of genotoxic building blocks that can potentially be generated as degradation products during manufacture or storage As shown in Figure 152, a building block (eg, an intermediate in the penultimate step of the chemical synthesis of a drug molecule) that is reasonably expected to be released in whole or in part as a degradation product through hydrolysis or some other predictable degradation pathway is screened for alerting structures using an appropriate in silico tool (eg, DEREK) If an alert structure is identified in the molecule, an Ames test is performed to further assess the genotoxic potential If the Ames test is positive, this building block should be replaced An example of the application of this approach is the screening of aromatic amine building blocks, as this type of molecule is potentially genotoxic and is a potential degradation product generated through the hydrolysis of the amide bond Unfortunately, the application of this approach to types of degradation pathways
other than hydrolysis can be tricky because the degradation pathways may not be obvious and may not be readily identified without a thorough stability study, which during the lead identification/optimization stages is not practicable
During clinical development or commercial stages the chemical structure of the drug molecule is fixed and, as such, modifying the chemical structure of the drug molecule to minimize the risk of generation of genotoxic degradation products is out of the question To understand and control the risk of genotoxic degradation products, the following questions, as shown in Figure 153, should be answered:
1 What is the degradation product? As the first step of risk assessment, the structures of all actual and
potential degradation products must be identified to allow further hazard assessment
2 Is it genotoxic? Once the structure of a degradation product is identified, an in silico assess-
ment should be performed to identify whether an alert structure is present by using an adequate in silico tool software (eg, DEREK) If an alert structure is not identified, the degradation product is considered a regular impurity and controlled per ICH Q3A/B in accordance with the stage of development On the other hand, if an alert structure is identified an Ames test is typically performed A negative Ames result can overrule the in silico result, and the degradation can be recategorized as a regular impurity A positive Ames result is sufficient to categorize the degradation product as genotoxic [3,4]
3 Is it controllable? A degradation product cannot be purged in a similar fashion as a process
impurity; therefore, it is important to thoroughly understand the chemistry and the rate of formation of the degradation product Because the drug molecule is fixed at this stage, the focus must be on understanding the conditions that can stabilize or destabilize the drug
Points of consideration for stabilization of the drug substance include but are not limited to the following:
• Salt form: a salt form can dramatically alter the chemical and physical properties of the drug and hence its stability For example, an HCl salt of a tertiary amine drug molecule will eliminate the susceptibility of the tertiary amine functional group toward oxidation
• Polymorphism: physical forms affect chemical stability Amorphous materials are typically less stable than crystalline materials Different forms of crystalline materials may also exhibit different chemical stabilities
• Particle size: not always but can potentially be a stability attribute • Impurities: reactive impurities in the drug substance can be a critical
stability attribute For example, loose water (noncrystalline) is always a concern if the drug is susceptible to hydrolysis Also, transition metals may mediate hydrolysis or oxidation Generally, however, impurities are not a major mediator of degradation [9,10]
• Light protection: the light sensitivity of the drug substance must be understood so that protection measures can be in place during manufacture, handling, and storage
• Packaging: conventional drug substance packaging comprises polyethylene plastic bags in a fiber drum If necessary, more moisture or oxygen protective packaging can be used (eg, aluminum pouch)
When a drug substance is formulated into a drug product, the stability of the drug can be decreased Hence, assessment of degradation and stabilization of a formulated product can be more critical than that of the drug substance In addition to the points of consideration for the drug substance, stabilization of a formulated product includes but is not limited to the following:
• Excipients: the impact of excipients on drug stability is multifaceted On the one hand, certain excipients (eg, antioxidants) can retard drug degradation; on the other hand, excipients may directly or indirectly (eg, via trace-level reactive impurities in the excipients) interact with the drug
• Coating: an appropriate coating can protect the drug from exposure to a destructive environment (eg, moisture or oxygen)
• Container/closure/packaging systems: more complex container/closure/ packaging systems can be considered for the drug product (Note that leachables from container closure systems are drug product impurities, some of which can be potentially genotoxic However, these are not drug degradation products and therefore are out of the scope of this chapter)
• Long-term storage conditions: chemical degradation typically decreases with a decrease in temperature For example, changing the storage conditions from a controlled room temperature to refrigeration should typically significantly decrease the rate of formation of the degradation product, although the effect of temperature can vary with the mechanism of degradation
4 With a thorough understanding of the mechanism of formation of the genotoxic degradation product and the possible chemical and pharmaceutical stabilization measures, an analytical control strategy can be devised A specification should be established if necessary
The first step in the risk assessment of genotoxic degradation products is the identification of chemical structures of actual and potential degradation products All degradation products above the sTTC or TTC in the impurity iceberg should be considered However, because the sTTC or TTC can be at parts-per-million levels, experimental elucidation of all those degradation products not only is impractical from a technology perspective but also contradicts the ICH Q3A/B guidelines, as these guidelines clearly state that only those degradation products above the ITs need to be identified
The strategy for identification of a degradation product can be significantly impacted by the MDD of the product This is because the IT is largely a constant value across a large range of MDDs, whereas the sTTC or TTC changes linearly with the MDD, as shown in Table 152 for TTC
TABLE 15.2 Impact of MDD on the Concentration at TTC and ICH ITs
When the MDD is 15 mg/day, the concentration limit (01%) at the TTC level (15 µg) is the same as the ICH IT for drug substances Also, when the MDD is at or lower than 15 mg/day, the ICH Q3A ITs are the same as or even stricter (ie, lower) than the concentration limit at the TTC Therefore, the risk assessment for genotoxic degradation products can fully rely on the assessment of identified actual degradation products observed at release and during storage under recommended conditions
With the increase of MDD, the concentration limit at TTC level becomes lower and can be orders of magnitude lower than the ICH ITs Consequently, identification of only those degradation products that exceed the ICH ITs will not warrant a thorough genotoxic risk assessment of all actual and potential degradation products at release and during storage In this situation, a chemistry-and risk-based strategy for the identification of potential degradation products should be applied
A typical workflow for the identification and assessment of actual degradation products for genotoxic potential is described in Figure 154 Actual degradation products are typically experimentally determined to be present by high-performance liquid chromatography (HPLC)-based techniques in a drug substance or drug product at release and during long-term storage, so technically it is possible to elucidate the chemical structures by using advanced online structural analytical means (eg, HPLC-MS)
With additional efforts, adequate amounts of the degradant of interest can be isolated for further elucidation (eg, by nuclear magnetic resonance) However, the efforts should be appropriate according to the stage of development
During late-stage clinical development or for commercial products, ICH ITs should be applied As stated in the EMA questions and answers document #4: Question: “What would be an appropriate strategy to qualify a new impurity that
arises during Phase III or with a commercial product?” Answer: “In line with the ICH guideline, no action is generally required for a
new unidentified impurity found at levels below the ICH identification threshold”
During early clinical development stages the appropriate IT is not defined by regulatory agencies, and thus industry practices vary from company to company Recently, a proposal put forth by an industry working group under the IQ Consortium on Early Development GMPs advocated a threefold ICH IT during early development stages (eg, phase 1 to 2a) [11] For example, if the IT intended for a drug at the registration stage is 01%, then 03% would be an appropriate threshold for the identification of actual unidentified impurities observed at release or during longterm stability studies during the early clinical development stages
As discussed in Section 151, for risk assessment of genotoxic degradation products potential degradation products are considered that can theoretically and reasonably be expected to arise during manufacture or storage at levels exceeding the sTTC or TTC Since the identification of these degradation products at release or directly from the drug stored under long-term storage conditions is impractical, approaches that can predict these potential degradation products must be employed
15.4.2.1 Identification of Potential Degradation Products via Accelerated Stability Testing
Accelerated stability testing comprises “studies designed to increase the rate of chemical degradation or physical change of a drug substance or drug product by using exaggerated storage conditions as part of the formal stability studies” [12] Since accelerated testing is part of the formal stability studies, the storage conditions are not arbitrary; rather, they are defined in ICH Q1A [12] according to the intended longterm storage conditions (eg, 40°C/75% relative humidity [RH] is the accelerated condition for 25°C/60% RH long-term storage, and 25°C/60% RH is the accelerated condition for 5°C refrigerated long-term storage)
Because the accelerated storage condition is only 15°–20° higher in temperature than the long-term storage temperature, and is within the range of temperature excursion that might occur during shipment, the degradation pathways observed under accelerated conditions are typically considered highly relevant to those under the long-term storage conditions Chemical degradations under the accelerated conditions can be several-fold faster than those under the long-term storage conditions due to the temperature effect or a combination of both temperature and humidity The extent of acceleration of the degradation from the temperature effect is determined
by the activation energy of the degradation pathway Because of the wide variety of molecules, activation energies of drug degradation vary in a wide range from 12 to 40 kcal/mol [13-15] Assuming that the activation energy is the same as that used by the United States Pharmacopoeia (83144 kJ/mol = 198 kcal/mol) [16] to calculate the mean kinetic temperature, the rate of degradation at 40°C is about five times faster than that at 25°C [13] Therefore, typically, a 6-month accelerated stability study should be highly predictive of at least 2 to 3 years of storage under long-term storage conditions Degradation products actually observed during the accelerated stability studies are considered to be potential degradation products of the drug under the long-term storage conditions Those above the ICH ITs (as adjusted for early development, as discussed in Section 1541) should be identified and assessed for genotoxic potential
15.4.2.2 Identification of Potential Degradation Products via Stress Testing Stress testing is the “studies undertaken to elucidate the intrinsic stability of the drug substance” or “studies undertaken to assess the effect of severe conditions on the drug product” [12] and “is normally carried out under more severe conditions than those used for accelerated testing” [12] Unlike long-term stability or accelerated stability studies whose conditions (eg, temperature and RH) are clearly defined by ICH Q1A [12], stress testing covers a much broader range of conditions (eg, temperature, humidity, pH, light, and oxidants), and the details of the conditions are not clearly defined by regulators This lack of regulatory guidance on stress testing is challenging for the industry and, in general, the following principles should be considered:
• The purpose of stress testing is to predict the degradation/stability of the drug under long-term storage conditions, not for the sake of generating degradation products
• The severity of the conditions should be justified and relevant • The end point of the testing should be determined by the level of degrada-
tion desired or by a “maximum duration” A maximum duration should be applied to every condition Once the maximum duration is achieved, “no degradation” should be an acceptable outcome
Nonetheless, because stress testing conditions are significantly harsher than the typical long-term and accelerated storage conditions for drug substances or drug products, it is expected that some degradation products that may not be formed under long-term and accelerated storage conditions can be formed under those severe stress conditions For this reason, whether stress testing can be used for the prediction of potential degradation products remains a very controversial topic It is the author’s point of view that as a predicting tool it is unreasonable to expect that stress testing can accurately generate the exact degradation products as those formed under long-term stability conditions With that said, thoughtfully designed stress testing should be expected to generate degradation products that cover all or the majority of all actual or potential degradation products that can be formed under long-term storage conditions [13,17] Hence, a critical issue in using stress testing to predict potential degradation products is how to ensure that the focus is on those products
that are more likely to be formed under long-term storage and exclude those that are less likely to be formed under long-term storage based on a chemistry-based risk assessment In other words, the challenge is to avoid the overprediction of potential degradation products Two important concepts must be clarified and understood: “relevant stress conditions” and “major degradation products”
Relevant stress conditions are those that are thermodynamically and kinetically relevant to long-term-storage and in-use conditions (eg, 25°C/60% RH, indoor light) Several criteria can be used to assess the relevancy:
• What conditions most reveal the intrinsic stability of the drug? • Are the conditions thermodynamically or kinetically relatable to the long-
term storage conditions? • Are the conditions relevant to in-use or physiological conditions?
