Decisions to approve, prescribe and consume medicines involve risk/benefit assessments by regulatory agencies, health care professionals and consumers. For serious or life threatening conditions, drugs with higher risks for adverse effects or for serious adverse effects are sometimes acceptable. For example, some life-saving cancer chemotherapies are known human carcinogens. However, if one is suffering from a life threatening tumor, a 5% risk of a secondary, treatment-related tumor is generally considered acceptable. Arguably, the same is not true for impurities found in drug substances and drug products; impurities convey only risk with no associated benefit. Drug impurities might be viewed as "pollutants" in the pharmaceutical world. Much like pollutants in the environment, few people believe that they can be entirely eliminated. The challenge for regulatory agencies is to promulgate standards that assure that unavoidable drug impurities impart no or acceptable levels of risk.
Residual impurities resulting from manufacturing and formulation, or from degradation of the active pharmaceutical ingredient (API) and excipients, may be present in pharmaceutical products. A subset of these impurities may present a potential for genotoxicity and therefore pose an additional safety concern to clinical subjects and patients. The pharmaceutical industry and those that regulate it recognize their respective obligation to limit genotoxic impurities. Therefore, substantial efforts are made during development to control all impurities at safe concentrations. However, the effort made to limit impurities must be commensurate with the risk assessed at each phase of clinical development, taking into account the extent of the hazard, the disease indication, the size and characteristics of the exposed population, and the duration of that exposure, as well as the likely delay in the availability of beneficial medicines if the burden of limiting or controlling impurity levels is disproportionate. A balance of these considerations can be described best as the ''as low as reasonably practicable'' (ALARP) principle. It follows that the presence of impurities with genotoxic (mutagenic) potential may be unavoidable in clinical trial and ultimately in approved and marketed materials. Control of impurities in the drug substance and degradants in drug product are addressed in ICH Quality Guidelines Q3A(R) and Q3B(R), respectively, and the Q3C guideline that deals with residual solvents. However, no specific guidance for determining acceptable levels for genotoxic impurities is provided in these documents other than to recognize the fact that unusually toxic impurities may require tighter limits of control.
A major change in impurity testing involves a new regulatory guidance designed to tightly limit substances possessing potential for genotoxicity. The European Agency for the Evaluation of Medicinal Products (EMEA) guidance on genotoxic impurities, which became effective on Jan. 1, 2007, now applies a 1.5 mg daily exposure limit for such substances in most pharmaceuticals based on a precedent application of the threshold of toxicological concern (TTC) concept to food additives and food contact materials. Before the EMEA draft guidance, genotoxic impurities had been addressed only as a footnote in ICH Q3A (R2) "Impurities in New Drug Substances".
In 2004, the Pharmaceutical Research and Manufacturers of America (PhRMA) formed a task force to discuss genotoxic impurity limits. Concerned the 1.5 mg/day limit would be applied to drugs synthesized in the United States, even while these drugs were still in clinical development, the PhRMA group proposed a staged TTC approach that ties permissible impurity levels to the stage of development. Because clinical studies are conducted with limited duration of dosing, the group reasoned that total exposure is very low, and thus higher intake levels should be allowable during early clinical studies without a net increase in risk. PhRMA's staged TTC approach applies to all clinical routes and to compounds at all stages of development, for identified and predicted impurities. The TTC limits would not apply to already marketed products.
US FDA has considered both the guidelines in setting limits on genotoxic impurities.
The European Medicines Agency's (EMEA) Committee for Medicinal Products for Human Use (CHMP) published a guideline on the limits of genotoxic impurities. This guideline recommends dichotomizing genotoxic impurities into those for which there is "sufficient (experimental) evidence for a threshold-related mechanism" and those "without sufficient (experimental) evidence for a threshold-related mechanism.
" Those genotoxic compounds with sufficient evidence would be regulated using methods outlined in ICH Q3C, for class 2 solvents. This approach calculates a "permitted daily exposure" (PDE) which is calculated using the NOEL or LOEL from the most relevant animal study plus incorporation of safety factors. Examples of genotoxins that may fall into this class include chemicals that induce aneuploidy by interfering with the mitotic spindle, chemicals interfering with activity of topoisomerase or chemicals that inhibit DNA synthesis.
For genotoxic compounds without sufficient evidence for a threshold-related mechanism, the guideline proposes a policy of controlling levels to "as low as reasonably practicable" (ALARP principal). This approach specifies that every effort should be made to prevent the formation of such compounds during drug substance synthesis and, if not possible, efforts should be made to reduce such impurities through technical efforts (e.g. purification steps). Compounds falling into this class are generally those that interact with DNA either directly or indirectly such as alkylating agents, intercalating agents or agents generating free radicals. Since all exposures to such agents theoretically convey some level of carcinogenic risk, regulatory agencies generally perform quantitative risk assessments to calculate the increased levels of adverse events, such as cancers, that result from particular exposures and set exposure levels which result in "acceptable" risks; often 1 in 105 or 1 in 106 additional cancers from lifetime exposures.
While the approach described above has sound scientific support, in most instances sufficient mechanistic data will be lacking with which to decide whether a threshold mechanism is applicable for genotoxic impurities. Furthermore, it is also unlikely that data will exist on which quantitative risk assessments can be performed. The guideline recognizes these limitations and therefore proposes the use of a "threshold of toxicological concern" (TTC) for genotoxic impurities. The TTC refers to a threshold exposure level to compounds that will not pose a significant risk of carcinogenicity or other toxic effects and was originally developed as a "threshold of regulation" for food contact materials by the FDA. The draft guideline proposes a TTC of 1.5 ug/day. This threshold corresponds to a 10-5 lifetime risk of cancer, a risk level that the EMEA considers justified due to the benefits derived from pharmaceuticals. Importantly, however, this draft guideline only addresses levels of genotoxic impurities in marketed products; the guideline is silent on what might constitute acceptable TTCs for drugs during development, especially for trials of short duration.
Examples of mechanisms of genotoxicity that may be demonstrated to lead to non-linear or thresholded dose-response relationships include interaction with the spindle apparatus of cell division leading to aneuploidy, topoisomerase inhibition, inhibition of DNA synthesis, overloading of defence mechanisms, metabolic overload and physiological perturbations (e.g. induction of erythropoeisis, hyper- or hypothermia). For (classes of) compounds with clear evidence for a thresholded genotoxicity, exposure levels which are without appreciable risk of genotoxicity can be established according to the procedure as outlined for class 2 solvents in the Q3C Note for Guidance on Impurities: Residual Solvents. This approach calculates a "Permitted Daily Exposure" (PDE), which is derived from the NOEL, or the lowest observed effect level (LOEL) in the most relevant (animal) study using "uncertainty factors" (UF).
The assessment of acceptability of genotoxic impurities for which no threshold mechanisms are identified should include both pharmaceutical and toxicological evaluations. In general, pharmaceutical measurements should be guided by a policy of controlling levels to "as low as reasonably practicable" (ALARP principle), where avoiding is not possible. Levels considered being consistent with the ALARP principle following pharmaceutical assessment should be assessed for acceptability from a toxicological point of view.
A rationale of the proposed formulation/manufacturing strategy should be provided based on available formulation options and technologies. The applicant should highlight, within the chemical process and impurity profile of active substance, all chemical substances, used as reagents or present as intermediates, or side-products, known as genotoxic and/or carcinogenic (e.g. alkylating agents). More generally, reacting substances and substances which show "alerting structure" in terms of genotoxicity which are not shared with the active substance should be considered. Potential alternatives which do not lead to genotoxic residues in the final product should be used if available. A justification needs to be provided that no viable alternative exists, including alternative routes of synthesis or formulations, different starting materials. This might for instance include cases where the structure, which is responsible for the genotoxic and/or carcinogenic potential, is equivalent to that needed in chemical synthesis (e.g. alkylation reactions).
If a genotoxic impurity is considered to be unavoidable in a drug substance, technical efforts (e.g. purification steps) should be undertaken to reduce the content of the genotoxic residues in the final product in compliance with safety needs or to a level as low as reasonably practicable. Data on chemical stability of reactive intermediates, reactants, and other components should be included in this assessment. Detection and/or quantification of these residues should be done by state-of-the-art analytical techniques.
The impossibility of defining a safe exposure level (zero risk concept) for genotoxic carcinogens without a threshold and the realization that complete elimination of genotoxic impurities from drug substances is often unachievable, requires implementation of a concept of an acceptable risk level, i.e. an estimate of daily human exposure at and below which there is a negligible risk to human health. However, these approaches require availability of adequate data from long-term carcinogenicity studies. In most cases of toxicological assessment of genotoxic impurities only limited data from in vitro studies with the impurity (e.g. Ames test, chromosomal aberration test) are available and thus established approaches to determine acceptable intake levels cannot be applied. Calculation of "safety multiples" from in vitro data (e.g. Ames test) are considered inappropriate for justification of acceptable limits. Moreover, negative carcinogenicity and genotoxicity data with the drug substance containing the impurity at low ppm levels do not provide sufficient assurance for setting acceptable limits for the impurity due to the lack of sensitivity of this testing approach. Even potent mutagens and carcinogens are most likely to remain undetected when tested as part of the drug substance, i.e. at very low exposure levels. A pragmatic approach is therefore needed which recognizes that the presence of very low levels of genotoxic impurities is not associated with an unacceptable risk.
The establishment of more widely accepted TTC values would benefit consumers, industry and regulators. By avoiding unnecessary extensive toxicity testing and safety evaluations when human intakes are below the relevant TTC value, it will allow limited resources of time, animal use, cost and expertise to be devoted to the testing and evaluation of those substances with greater potential to pose risks to human health. In consequence, its application will contribute to a considerable reduction in the number of animals used for safety testing.
The general framework for genotoxicity testing of pharmaceuticals is given in two internationally agreed ICH safety guidelines (ICH S2A, ICH S2B). One of these guidelines (ICH S2B) describes the standard battery of tests for genotoxicity for drug substance, which consists of:
(i)A test for gene mutation in bacteria.
(ii)An in vitro test with cytogenetic evaluation of chromosomal damage in mammalian cells
(iii)An in vivo test for chromosomal damage in rodent hematopoietic cells.
The ICH safety guidelines (S2A and S2B) state: ''for compounds giving negative results, the completion of this 3-test battery, performed and evaluated in accordance with current recommendations, will usually provide a sufficient level of safety to demonstrate the absence of genotoxic activity.'' In this context, genotoxicity is a broad term encompassing effects from mutagenicity through DNA reactivity, DNA damage, and chromosomal damage, both structural chromosome breakage and aneuploidy. Any compound that produce a positive result in one or more assays in the standard battery has historically been regarded as genotoxic, which may require further testing for risk assessment. Thus, the standard battery of genotoxicity assays used for testing the API provides important information about a diversity of mechanisms of genotoxicity, both directly and indirectly associated with effects on DNA. Genotoxicants that do not act directly on DNA are typically associated with threshold-related mechanisms, while those that directly target DNA (typically detected in assays measuring the reverse or forward mutations in a specific gene with a selection agent) are considered by regulatory authorities not to have threshold- related mechanisms. Requirements for control of genotoxic impurities in pharmaceutical products are different depending upon whether or not there is evidence for a threshold-related mechanism. DNA-reactive genotoxic impurities for which there is no evidence of a threshold-related mechanism are regarded to be potentially trans-species and multi-organ carcinogens that may require control at relatively low levels. In contrast, it is accepted that impurities acting via threshold-related mechanisms do not require control at similarly low levels. Since the main concern that should drive control of impurities to relatively low levels is direct DNA reactivity, the primary endpoint of relevance for genotoxic impurities is mutagenicity. DNA-reactive carcinogens can be identified with a low incidence of false negative results by a procedure that combines the assessment of chemical structural features that infer DNA reactivity (such as electrophilicity) with a single biological hazard identification test such as a bacterial reverse mutation test, known as the ''Ames test''. A flexible use of this approach is sometimes advisable since genotoxicity assessment of impurities in mammalian cells may be needed for specific structural groups, such as carbamates, which are known carcinogens and that are known to be inefficiently detected in bacterial genotoxicity tests. A clearly negative result in an appropriate genotoxicity test (i.e., a bacterial reverse mutation test or mammalian cell assay) usually indicates a sufficient level of safety to conclude the absence of genotoxicity for the purpose of controlling impurities.
It is proposed here that impurities be classified into one of five classes using data (either published in the literature or from genotoxicity testing) and comparative structural analysis to identify chemical functional moieties correlated with mutagenicity. The five classes are:
This group includes known animal carcinogens with reliable data for a genotoxic mechanism and human carcinogens. Published data on the chemical structure exist demonstrating the genotoxic nature of the impurity.
This group includes impurities with demonstrated mutagenicity based on testing of the impurity in conventional genotoxicity tests, but with unknown carcinogenic potential.
This group includes impurities with functional moieties that can be linked to genotoxicity based on structure, but which have not been tested as isolated compounds. They are identified based on chemistry and using knowledge based expert systems for structure-activity relationships. The alerting functional moiety is not present in the structure of the parent API. Some widely recognized alerts for DNA reactivity, i.e., mutagenic activity, are depicted in Fig. 9.1.
Some generic rule-based alerts may be quite unspecific (e.g., the general alerts for aromatic amines; and further consideration must be given to chemical structural constraints, chemical environment, or experimental data in the assessment of potential genotoxicity. Due to the uncertain relevance of structural alerts, regulatory action should not be based solely on the presence of a particular functional group; rather the accuracy for predicted genotoxicity should be evaluated case-by-case based on the available scientific literature, additional unpublished (proprietary) data on the chemical class and further available (genotoxicity) test results on closely related structures.
This group includes impurities that contain an alerting functional moiety that is shared with the parent structure. The genotoxicity of the isolated impurity is unknown, but the genotoxicity of the active principle has been characterized through conventional genotoxicity testing. Similar chemical constraints and chemical environment exist for the alerting substructure in the impurity and the API.
Alkyl, Aryl, or H
Halogen = F, Cl, Br, I
EWG = Electron withdrawing group (CN, C=O, ester, etc)
Fig. 9.1 some examples of structurally alerting functional groups that are known to be involved in reactions with DNA (this list is not exhaustive) (Muller et al., 2006).
This group would be adequately covered by existing ICH Q3A(R), Q3B(R), and Q3C guidelines. It has to be emphasized that this classification system would be used solely for the purpose to decide whether an impurity possesses a high level of risk and is therefore to be controlled at very low levels of daily intake. Hence, this classification is not a general classification of genotoxicity.
The Pharmaceutical Research and Manufacturing Association (PhRMA) established a Genotoxic Impurity Task Force which developed a White Paper (document outlining the proposal is currently in press) and presented their proposal at various public meetings. The document outlines a procedure for testing, classification, qualification and toxicological risk assessment of potentially genotoxic impurities in pharmaceutical products. The Task Force proposed that all identified or predicted impurities should be classified into one of five classes: those known to be genotoxic (mutagenic) and carcinogenic, those known to be genotoxic (mutagenic) but with unknown carcinogenic potential, those with a unique alerting structure and of unknown genotoxic (mutagenic) potential, those with an alerting structure related to the parent active pharmaceutical ingredient, and those with no structural alert.
The Task Force proposal for addressing genotoxic impurities for marketing applications is similar to that of the previously described EMEA draft guideline. However, the Task Force recognizes that the TTC established a limit for daily human exposure to genotoxic impurities for lifelong treatment while most medicines are given for limited time spans, especially in early clinical development. Therefore, the Task Force proposes a staged TTC approach with adjusted limits for shorter duration clinical trials. The adjusted limits are derived from a linear extrapolation of the TTC for a lifetime daily exposure to short term daily exposures. The proposed limits for short-term exposures (?12 months) are based on a 10-6 risk because of the common inclusion in clinical studies of normal volunteers, for whom there is assumed to be no pharmacological benefit. The proposed limit for exposures greater than 12 months in duration is based on a 10-5 risk, because individuals in longer-term clinical studies have the target indication and may benefit from treatment, an approach consistent with the EMEA guideline.
ICH guidances do not provide clear recommendations for handling these types of impurities. The Center for Drug Evaluation and Research (CDER) of the USFDA is developing guidance to address issues with genotoxic impurities in pharmaceutical products. The CDER is considering the proposals of the EMEA and PhRMA in developing its guidance. The presence of genotoxic impurities should be avoided if possible. However, it is recognized that complete removal is often not possible. In these cases, the amounts of genotoxic impurity present should be limited to a level that represents an insignificant increase in risk to clinical trial subjects or patients.
This level may be based on adequate compound-specific data to calculate an acceptable risk-specific dose or may be based on a toxicological threshold derived from a robust carcinogenicity database. A staged implementation of the threshold approach is considered acceptable for products that are under development. In applying qualification thresholds, consideration should be given to the product's stage of clinical development, the maximum duration of drug administration at that stage, and the proposed indication. In some cases, increases in the recommended thresholds may be supported in the presence of a potential pharmacological benefit to patients.
In general, impurities should be quantitated at levels ?0.03 or 0.05% by weight according to ICH guidelines. Genotoxic impurities or potential genotoxic impurities must be controlled at levels significantly lower than the 0.03-0.05% levels that are typically reported by an HPLC impurity assay. Typically, developing limit tests (e.g. <50 ppm) for highly toxic impurities is readily achievable, however it can be difficult to develop a test to control a particular genotoxic impurity at 1 ppm (0.0001% w/w).
"For genotoxic impurities we need very sensitive and selective methods. One needs higher selectivity to determine ppm - level impurities and selective methods to separate low levels of genotoxic impurities from base line noise and other organic impurities. The typical HPLC methods with a nonspecific detector (e.g. UV) that are used to measure organic impurities may not be appropriate to quantitate low ppm levels of genotoxic impurities.
The quantitation of low levels (in the range of ppms) of impurities is the challenging part, and using specific detectors such as MS or MS-MS with LC will significantly improve the method selectivity and the quantitation limit. The goal for scientists is to identify potential genotoxic impurities early in development, develop analytical methods to test for these impurities in the intermediates, and if possible, to demonstrate that the manufacturing process controls them before reaching the final drug substance. "If you eliminate them early enough, then your actual active drug substance is pure, free of genotoxic impurities".
Compounds that are both carcinogenic and genotoxic frequently give rise to difficulties for regulators and food businesses when they are present at low levels in foods. Among regulatory and advisory bodies, such as JECFA, EFSA and FDA, there is no international consensus on how to evaluate the potential risk of genotoxic carcinogens in food. In Europe , the Scientific Committee on Food (SCF) has addressed the topic of chemicals that may be genotoxic carcinogens in a written opinion. The SCF has evaluated genotoxic carcinogens in the diet (such as contaminants and natural toxicants) on a case by-case basis using a ''weight of evidence'' approach, whereas in the USA (e.g. US EPA), as well as in some European countries (e.g. Norway ), a quantitative risk characterization is commonly performed by mathematical low dose extrapolation of animal data. It is recognized that current risk analysis approaches to compounds in food that are genotoxic and carcinogenic in experimental animals may sometimes incur disproportionate or even unnecessary measures on the part of regulators and industry.
The risk assessment of genotoxic carcinogens has been considered recently by the European Food Safety Authority (EFSA) and by the WHO/FAO Joint Expert Committee of Food Additives (JECFA). In 2003, EFSA established a Working Group to consider how to improve advice given on the health risks arising from the presence in food of compounds that are both genotoxic and carcinogenic.
DNA-reactive carcinogens have long been known to be present in the human diet. Most problems with DNA-reactive carcinogens in food arise from chemicals that are either natural food constituents (such as ethyl carbamate) or contaminants (such as acrylamide, heterocyclic amines, polycyclic aromatic hydrocarbons and heterocyclic amines such as PhIP (2- amino-1-methyl-6-phenylimidazo [4,5-b] pyridine), which are formed during cooking processes or by fungal toxins (such as aflatoxins) because they cannot be completely eliminated from the human diet unless the food itself is banned. Compounds that are incorporated into food intentionally, either directly (e.g. additives) or indirectly (e.g. residues of processing aids, pesticides, veterinary drugs or migrants from food contact materials) are assessed for their genotoxic and carcinogenic potentials prior to marketing, and compounds that might be DNA-reactive carcinogens would not be permitted.
Risk characterization has been defined as ''the quantitative or semi-quantitative estimate, including attendant uncertainties, of the probability of occurrence and severity of adverse effect(s)/event(s) in a given population under defined conditions based on hazard identification, hazard characterization and exposure. At its best, a risk characterization ''synthesizes an overall conclusion about risk that is complete, informative and useful for decision makers''. The usual hazard characterization approach for compounds that cause cancer by non-DNA-reactive mechanisms is to calculate a health-based guidance value, such as a tolerable daily intake, using the no-observed adverse effect level and uncertainty factors. Such an approach is not used for compounds that cause cancer by DNA-reactive mechanisms. The decision-making environment also may dictate the nature and focus of the advice. Depending upon the circumstances, risk managers could be presented a risk characterization containing only a qualitative judgment of risk. Alternatively, the risk characterization may include one or more quantitative estimates of risk in addition to the qualitative judgment. Whichever method of risk characterization is adopted, the output is only as reliable as the quality of the data used. This applies to all data used, i.e. hazard identification, hazard characterization and dose-response analysis and exposure estimation.
The decision-making environment also may dictate the nature and focus of the advice. Depending upon the circumstances, risk managers could be presented a risk characterization containing only a qualitative judgment of risk. Alternatively, the risk characterization may include one or more quantitative estimates of risk in addition to the qualitative judgment. Whichever method of risk characterization is adopted, the output is only as reliable as the quality of the data used. This applies to all data used, i.e. hazard identification, hazard characterization and dose-response analysis and exposure estimation.
The only data essential for qualitative approaches, such as ALARA (as low as reasonably achievable), is identification of the compound as a genotoxic carcinogen. An important decision is that the genotoxicity arises via direct covalent binding to DNA, rather than via a mechanism that would show a threshold in the dose-response relationship. Often data are available from a variety of genotoxicity tests and the decision is based on a weight of evidence approach, which takes into account the quality of the available data. Although the ALARA principle is an easy to understand concept, it poses some major difficulties for the risk manager as it does not discriminate between very potent and very weak carcinogens and does not take human exposure into account. It does not give any guidance on the magnitude of any risk that might be associated with a given ''reasonably achievable'' low exposure level. Although these difficulties would point to the alternative approach of calculating the intake levels associated with ''acceptably small risks'', the mathematical models necessary may give widely divergent answers and do not provide a reliable basis for the formulation of realistic risk management advice.
The control of impurities bearing a genotoxic potential in pharmaceutical products and food products has received more and more attention over the past years. The inherent difficulties of true or hypothesized linear dose effect relationships have led to diverse strategies and risk calculations to achieve a rational level of control. Hence, a unified approach for product development and marketing would be useful. The ultimate risk concern for genotoxicants is carcinogenicity but carcinogenicity data are not available in most cases. Hence, a risk assessment based on surrogate data such as structure-activity relationships and limited genotoxicity testing in bacterial reverse mutation tests, knowledge about the relationship between genotoxicity and carcinogenicity, and a generic determination of virtually safe exposure levels for the world of genotoxic carcinogens is proposed. The risk assessment is based on Threshold of Toxicological Concern (TTC) approach for the intake of genotoxic impurities over various periods of exposure. This staged TTC is based on knowledge about tumorigenic potency of a wide range of genotoxic carcinogens and can be used for genotoxic compounds, for which cancer data are limited or not available. The delineated acceptable daily intake values of between ~1.5 mg/day for ~ lifetime intake and ~120mg/day for < 1 month are virtually safe doses. Based on sound scientific reasoning, these virtually safe intake values do not pose an unacceptable risk to either human volunteers or patients at any stage of clinical development and marketing of a pharmaceutical product. The intake levels are estimated to give an excess cancer risk of 1 in 100,000 to 1 in a million over a lifetime, and are extremely conservative given the current lifetime cancer risk in the population of over 1 in 4. This approach, together with a proactive process analysis of the chemistry behind the synthesis of the pharmaceutical product and matching analytical capabilities, ensures patient and volunteer safety and may not hinder inappropriately the fastest possible development of new medicines to improve patient health.
Apart from pharmaceuticals various difficulties are presented due to presence of low levels of food-borne DNA-reactive genotoxic carcinogens, some of which may be difficult to eliminate completely from the diet, thus a structured approach has been proposes for the evaluation of such compounds. While the ALARA approach is widely applicable to all substances in food that are both carcinogenic and genotoxic, it does not take carcinogenic potency into account and, therefore, does not permit prioritization based on potential risk or concern.
In the absence of carcinogenicity dose-response data, an assessment based on comparison with an appropriate threshold of toxicological concern may be possible. The above approaches were applied to selected food-borne genotoxic carcinogens. The proposed approach is applicable to all substances in food that are DNA reactive genotoxic carcinogens and enables the formulation of appropriate semi-quantitative advice to risk managers.