ABSTRACT
Cancer risk assessment consists of hazard identification, dose-response, exposure assessment, and finally risk characterization In terms of genotoxic impurities in pharmaceuticals, hazard identification is done using in silico analysis for the prediction of mutagenicity or an Ames test A hazard is given in case of a positive result A positive in silico prediction can be overruled by a negative Ames test, that
41 Cancer Risk Assessment 177 42 In Silico Testing 180 43 In Vitro Testing 181
431 Ames Test 182 432 Mammalian Gene Mutation Assay 183
44 In Vivo Testing 183 441 Comet/Micronucleus 185 442 Unscheduled DNA Synthesis in Liver 186 443 In Vivo Mutation Assay 186 444 Pig-a Mutation Test 187 445 Emerging Models 187
45 Regulatory Aspects 187 References 188
is, a negative result is accepted such that the test article is classified as not mutagenic (or presently nongenotoxic with respect to impurities) A positive Ames test can be overruled by a negative outcome in a relevant in vivo mutagenicity follow-up study The exposure assessment for genotoxic impurities in pharmaceuticals is simple; since 1997, impurities should be identified at 005% and qualified above 1 mg/day according to the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) quality guidelines on impurities in drug substances and drug products [1,2] Since 2007, genotoxic impurities should be controlled to the threshold of toxicological concern (TTC) limit of 15 µg/day at the cancer risk level of 1 in 100,000 according to the European Guideline on genotoxic impurities [3,4] and the respective Food and Drug Administration (FDA) draft guidance [5] The risk characterization for genotoxic impurities is predefined to be a mutagenic, very potent carcinogen unless data can prove otherwise and, therefore, they are classified as unusually toxic
Generally speaking, genotoxicity means damage to the DNA, the macromolecule that contains the genetic code transcribed in genes and when condensed to chromosomes allows the transfer of genetic information to daughter cells or the next generation Genotoxicity as it is understood presently includes mutations in the DNA (base pair or frameshift/deletion mutations, ie, the genetic information has a different sequence compared to the origin), clastogenicity (damage to chromosomes), and aneugenicity (numerical change in the number of chromosomes) Genotoxic impurities that must be controlled to 15 µg/day are by definition those that induce mutations, that is, mutagenic impurities Mutagenic compounds react with DNA, causing DNA damage, which can be fixed as mutation once the DNA is duplicated Mutations are most frequently detected in the bacterial reverse mutation assay, the Ames test Clastogenicity and aneugenicity can be detected in the chromosomal aberration test and/or the in vitro and in vivo micronucleus test The chromosomal aberration test is part of the genotoxicity package for impurities above the qualification threshold in ICH quality guidelines Nevertheless, it is only mutagenicity that is of concern, since the dose response for DNA reactivity and mutagenicity can span several orders of magnitude while clastogenic or aneugenic events have a threshold effect and therefore do not require extrapolation to the very low dose of 15 µg/day On the other hand, the rate of fixation is dependent on the rate of DNA damage formation, DNA repair, and kinetics of cell replication; therefore, a threshold may also exist for mutagenic compounds, but this conclusion needs to be supported by data
Accumulation of DNA damage in specific genes in a cell can lead to cancer, a process that develops in the course of several years [6] Genotoxicity studies are short-term studies in which DNA damage or mutations are investigated They are part of the clinical phase I-enabling preclinical safety package for the active pharmaceutical ingredient (API)
Carcinogenicity studies are performed in mice and rats for a treatment period of 18-24 months, that is, the average lifetime of these rodents These studies are performed late in development mainly in parallel to clinical phase III trials with the results being available at the time of submission A few dog and monkey studies were also performed with a few animals and 5-15 years of exposure All species are regarded to be as sensitive as humans, and if available the most sensitive species that
responds at a lower dose is taken as the basis for risk assessment The doses are chosen based on the maximum tolerated dose studies in rats and mice of generally 13-week duration The top dose should cause some toxicity like body weight gain reduction by maximally 20% for chemicals or a sufficient high systemic exposure of 25-to 50-fold for an API compared to human exposure established in the clinical phase II trials Two lower doses are included to see whether there is a dose-response relationship In general, there are 50 males and 50 females/group The majority of all 18-or 24-month carcinogenicity studies are performed with dose levels in the milligram per kilogram range or at a limit dose of 1500 mg/kg for an API In contrast, the human exposure of a mutagenic impurity would be in the range of 25 ng/kg for a 60 kg patient at the TTC limit In a carcinogenicity study, the so-called ED01 megamouse study with 2-aminofluorene, 24,192 mice were tested in seven dose groups (and different durations of treatment) with 1,728 mice each in the lowest-dose groups with serial sacrifice following 18 and 24 months and natural death This high number of animals was required for statistical reasons to adjust for the low incidence This study showed a rather linear dose-response relationship for liver tumors but clearly not for bladder tumors [7] The DNA adducts study using nine dose groups also showed a linear dose-response relationship in the liver and bladder [8], indicating that additional tissue-specific factors influenced the nonlinearity for bladder tumors but not for liver tumors In the absence of such carcinogenicity data, it is assumed that there is a linear dose-response relationship for all tumor types from the milligram per kilogram range tested in carcinogenicity studies down to the nanogram per kilogram, which is acceptable for humans The correct point of departure for linear extrapolation from the high doses in the carcinogenicity studies down to the low levels that people are exposed to is a matter of wide debate, that is, whether the TD50 value (the chronic daily dose that induces tumors in 50% of the animals), the TD10, the TD25, or other points of departure for linear extrapolation can be used The TD50 value could be either the harmonic mean of several studies as given in the Carcinogenic Potency Database (https:// potency berkeleyedu/) or the lowest TD50 value in one species, strain, and gender [9] Acceptable levels, also called virtual safe doses (VSDs) are calculated by linear extrapolation from the rodent TD50 value (risk level of 05) by division with 50,000 down to the cancer risk level of 1 in 105 (000001) for mutagenic and carcinogenic impurities in pharmaceuticals A VSD of 15 µg/day of TTC corresponds to a highly carcinogenic potent compound with a TD50 value of 125 mg/kg/day (15 µg/day times 50,000/60 kg patient) The risk level of 1 in 105 is considered appropriate for pharmaceuticals since there is a benefit for patients taking the drug The risk level for the general population exposed to contaminants that are without a benefit is 1 in 106 It has to be noted that animals also form tumors spontaneously, and the same is true for humans It is estimated that at least 10% of all people will have a tumor by the end of their lifetime, that is, the overall spontaneous risk level for cancer in humans is 01
There is human experience with chemicals, primarily those used in working places like polycyclic aromatic hydrocarbons (PAHs), aromatic amines, reactive chemicals, and alkylating anticancer drugs, which are known to be mutagenic, human carcinogens The discovery by Percivall Pott in the eighteenth century that chimney sweeps had higher occurrences of testicular cancer than any other group and the recognition that these tumors were due to the extensive exposure of this group to soot initiated the
research on human carcinogens The PAHs present in soot were the first identified human carcinogens PAHs are in general not used as intermediates in the synthesis of APIs and, therefore, mutagenic impurities in APIs do not belong to the class of PAHs However, aromatic amines and alkylating agents are frequently used in the synthesis of APIs Workers exposed to aromatic amines, for example, 2-naphthylamine, benzidine, and 4-aminobiphenyl, have a higher risk of developing bladder cancer [10,11] One study suggests an association between exposure to benzidine/β-naphthylamine and cancers of the liver, gallbladder, bile duct, large intestines, and lungs in humans [12], although the confounding factors like food and smoking habits are very difficult to assess The three human carcinogens (naphthylamine, benzidine, and 4-aminobiphenyl) as well as o-toluidine were either banned or strict control of workplace exposures was introduced Epidemiology data from workers exposed to alkylating agents are limited to α-chlorinated toluenes and benzoyl chloride, and the cohorts are small Nevertheless, the excess lung cancer risk is estimated to be approximately threefold compared to control [13] Intermediates can be impurities, and since the most potent carcinogens are banned from workplaces it is less likely that highly potent carcinogenic impurities are introduced in the synthesis of pharmaceuticals
In principle, electrophilic chemicals can react with DNA Relevant DNA-reactive chemical groups were described [14,15], including those that require first metabolic activation to DNA-reactive metabolites, as described for aromatic amines The initial in silico systems for the prediction of mutagenicity were based on the Ashby and Tennant alerts All in silico systems for the prediction of mutagenicity are either based on knowledge or based on quantitative, statistical structural activity relationships These systems predict mutagenicity as seen in the bacterial reverse mutation assay (the Ames test) They are continuously updated with the goal to reduce the false negative predictivity without increasing the false positive predictivity The false negative rate needs to be low since such impurities are currently controlled at ICH Q3A/B levels (666-fold higher than the TTC with a calculated risk of 666 × 105 = 7 in 103 under the assumption that the impurity is mutagenic and highly carcinogenic) The negative predictivity for the Ames test is in a good case around 90%, that is, if all in silico negative compounds would be tested in the Ames test, 90% are expected to be negative The predictivity depends on the variety of chemical spaces The better the prediction, the closer the chemical space and the higher the coverage, that is, more Ames data are available to create and/or optimize a rule for structural alert Nevertheless, some classes are difficult to predict, like aromatic amines The false positive predictivity is in the range of 30%–60% The false positive predictivity triggers an unnecessary burden on chemistry and toxicology since efforts are undertaken to reduce potential genotoxic impurity to the TTC level or the impurity is synthesized and tested in the Ames test All predictivity models are limited by the fact that the Ames test can be tested only up to bacteriotoxic and/or insoluble concentrations (or 5000 µg/plate), that is, insoluble, bacteriotoxic compounds do not exert their mutagenic properties under the test conditions
There are numerous in vitro mutagenicity assays; however, only the following tests are described in the Organisation for Economic Co-operation and Development (OECD) guidelines and are regulatorily well accepted In general, only high-quality, reproducible, and well-described mutagenicity tests should be evaluated (Table 41) In the published literature, at least a detailed description of the model used, test article identity (CAS number), batch, and purity should be provided as basic information The maximum concentrations should be up to 10 mmol/L for chemicals (1 mmol/L for pharmaceuticals), that is, concentrations that are in general far above the systemic exposures achieved in patients It is suggested to use a conversion factor and calculate the concentrations for the base if a salt form of the test article is used The salt form of an intermediate or impurity may influence the solubility, and solubility is a limiting factor in all in vitro tests The solvents used in in vitro tests are water, phosphate buffer, and dimethylsulfoxide Other solvents like ethylene glycol dimethyl ether, formamide, dimethylformamide, 001 M NaOH, acetone, or chloroform need to be tested first for bacterio-/cytotoxicity It is assumed that the impurities are stable under the test conditions (usually for 1-6 hours at room temperature) However, in specific cases the stability in the vehicle needs to be assessed All in vitro tests are performed in absence and presence of a metabolic activation system (S9 mix) S9 is prepared from rats treated with Aroclor 1254® or phenobarbital/β-naphthoflavone that induce enzymes like cytochromes P450 in the liver S9 is an organ (usually the liver) homogenate isolated as supernatant following centrifugation at 9000g for 20 minutes containing cytosol and microsomes The S9 mix is the organ fraction and an NAPDH-generating system The metabolic activation system is in a first tier with rat liver S9 mix In a second tier, S9 mix from other species like hamsters or humans or other enzyme inducers may be used In general, the cytochromes in human livers
TABLE 4.1 List of In Vitro Mutagenicity Tests Relevant for Impurities in Pharmaceuticals
vary and are in general lower than those in induced animals Rat S9 mix apparently does not permit cytochrome P450 2E1-mediated metabolic activation as required for urethane-like structures Therefore, this chemical class is usually negative in the Ames test Otherwise, there is a very good relationship between rat S9 mix and all important human primary metabolic pathways [16]
Good descriptions for the Ames and mouse lymphoma assays are given in various books and publications and in Redbook 2000 IVC1a-d from the FDA The target molecule, DNA, is ubiquitously present in all species and, therefore, even the data from bacteria are representative of events in mammalian systems
The bacterial reverse mutation assay Ames test was named after Bruce Ames, who created this test in the early 1970s The Ames test is performed according to OECD Guideline No 471, and it is recommended to perform it under good laboratory practice conditions to reach regulatory acceptability At least five different bacterial strains of Salmonella typhimurium and/or specific strains of Escherichia coli from the list in the OECD guideline should be tested with and without a metabolic activation system (S9 mix) The bacterial strains are mutated and therefore not able to synthesize histidine (tryptophan in the case of E. coli) by themselves, and thus bacterial growth is not possible The bacterial strains, which are reverse mutated by the test article to the wild type or with restored histidine/tryptophan-producing pathways, can grow to visible colonies within 2 to 3 days The Ames test is positive if the number of revertants in treated cultures increased twofold to threefold above those of the negative control The judgment of an Ames positive response is based on a concentration-dependent, reproducible increase in one of the tester strains with or without S9 mix compared to control Equivocal responses are sometimes seen with test articles that induce only weak (less than twofold to threefold) but reproducible increases Whenever possible the isolated impurity should be tested, and for critical classes only test articles of the highest purity (> 999%, eg, for aromatic amines) should be used or a repurification should be taken into consideration The maximum recommended concentration in the Ames test is either 5000 µg/ plate or a bacteriotoxic or insoluble concentration Under specific circumstances, an impurity like a degradation product may be tested in the presence of the drug substance/product as long as the impurity levels reach 250 µg/plate In this case, a chemical rationale should be provided that it is technically not feasible to synthesize the impurity Besides the major challenges of purity and stability of the test article, it is recognized that histidine-like structures may be responsible for false positive responses
A positive genotoxic outcome indicates a high probability for a positive outcome in the carcinogenicity study and, therefore, there is a tendency in drug development to use Ames data like a switch from negative = GO to positive = avoid in view of the higher risk for cancer The sensitivity of the Ames test to correctly identify rodent carcinogens is 588%, and the specificity to correctly recognize rodent noncarcinogens is 739% [17] This means that at least 261% of the noncarcinogens are Ames positive For example, p-anisidine hydrochloride, a potential impurity, is
Ames positive but negative in the carcinogenicity studies in mice and rats Therefore, in specific cases with low exposure levels it is worthwhile to assess the biological relevance of the Ames test in further relevant in vivo genotoxicity studies
In general, mutagenic impurities below the ICH Q3A/B qualification threshold do not need to be tested in the mouse lymphoma thymidine kinase (ML-TK) assay However, if mouse lymphoma data are already available, for example, for worker safety reasons according to in-house environmental health protection rules, then the risk assessment should be done on the mutagenic outcome, that is, an increase in mutation frequency of the large clones Various cell lines are listed in OECD Guideline No 476, but the L5178Y+/– cell line is used for pharmaceuticals according to the ICH S2(R1) guideline [18] The cell line is heterozygotic for the mutated thymidine kinase gene, that is, only one additional mutation is necessary to deactivate the thymidine kinase from the second chromosome Cells with inactivated thymidine kinase can grow in the presence of trifluorothymidine, whereas the nonmutated cells die The reason is that thymidine is synthesized by the mutated cells via a de novo synthesis route and not via thymidine kinase Thymidine is one of the nucleotides of DNA and is necessary for replicating/growing cells Cells are treated with the test article for 3 to 4 hours with and without a metabolic activation system The 24-hour incubation experiment is not necessary for impurities since aneugens and clastogens are primarily detected under these conditions Four analyzable test concentrations should be used, the highest concentration reaching a survival of 10%–20% or up to 1 mmol/L for the API and large-molecular-weight impurities Small-molecular-weight impurities could be tested up to 10 mmol/L as recommended by the OECD guideline for testing of chemicals, whereas API-like impurities could be limited to 1 mmol/L Only the mutant frequencies for large clones need to be evaluated since large clones are due to mutations and small clones due to clastogenic events The sensitivity of the ML-TK assay (both mutagenic and clastogenic events) is 808%, and the specificity is 476% [17]
Impurities for which genetic toxicity data from the chromosomal aberration assay, in vitro micronucleus test, or mouse lymphoma assay (small clones and 24-hour incubation data) are available do not need to be controlled at the TTC level since they do not follow the linear extrapolation for cancer risk but have an apparent threshold effect In these cases, the risk assessment follows a different approach, which is not yet defined One approach could be to calculate permitted daily exposure levels using safety factors, as described in the ICH Q3C guideline [19]
The biological relevance of a positive result in the Ames test or the ML-TK may be assessed in a relevant in vivo follow-up study (Table 42) Several questions need to be addressed like the sensitivity of the in vivo model to assess the DNA reactivity of a specific chemical compound Furthermore, it needs to be considered whether the impurity should be spiked to the drug substance/product or the “pure” isolated
TA B
LE 4
.2 Li
st o
fI n
V iv
o M
ut ag
en ic
it y
Te st
s
impurity and finally which doses should be tested It is recommended to test the pure impurity to fully assess the mutagenic potential since the dose should be several multiples of 125 mg/kg (see Chapter 1, VSD of 15 µg/kg) Currently, the top dose for chemicals should be the maximum tolerated dose, that is, about 50%–66% of the lethal dose that induced death in 50% of the animals (LD50) but not higher than 2000 mg/kg for acute dosing or 1000 mg/kg for repeated daily treatment for 4 weeks
Alkaline single cell gel electrophoresis (SCGE), or comet assay, is a sensitive and reliable method for detecting DNA damage expressed as single-and double-strand breaks, alkaline-labile sites, DNA-DNA, and DNA-protein cross-links including incomplete excision repair sites in single cells of different organs [20] Rats or mice are treated with the test article and organs are isolated at the time of suspected highest systemic exposure (3-6 hours) The liver is a default organ with high metabolic activity The stomach, duodenum, and jejunum (first site of contact or organ with highest absorption) are taken in case the Ames positive result was obtained in the absence of the metabolic activation system However, in principle, any known target organ of toxicity can be included Single cells are prepared and embedded in agarose, the DNA unwinded at pH ≥ 13, and the single-strand fragments migrated by electrophoresis After neutralization, the DNA is stained with the fluorescent dye ethidium bromide and evaluated by microscopy The resulting images, which were subsequently named for their appearance as “comets,” are measured to determine the extent of DNA damage The light intensity of the cell nucleus “head” (normal DNA or large fragments) in relation to the “tail” (fragmented DNA) is quantified The comet assay is easily combined with the micronucleus tests since only the bone marrow or peripheral blood needs to be used Furthermore, the integration of further mutagenic end points, that is, gene expression profile, would be possible The treatment duration should be extended to three administrations of the test article in the combined comet/micronucleus assay since it takes more time to express micronuclei in the bone marrow The administration schedule is that the first two administrations are given within 24 hours and the last 3-6 hours prior to necropsy The micronucleus assay is designed to detect clastogens and aneugens; nevertheless, it was shown that it also detects mutagenic carcinogens when the doses were sufficiently high
For aromatic nitro intermediates and impurities for which the bacterial nitroreductase renders this chemical class to be oversensitive in the Ames test, a combined in vivo comet/micronucleus test is considered suitable 2,6-Dinitrotoluene is Ames positive and has a high carcinogenic potency with a TD50 value of 029 mg/kg The comet assay in rat liver was positive when animals were treated with 25 mg/kg [21], thereby confirming the validity and sensitivity of this test to detect mutagenic carcinogens
Furthermore, the combined comet/micronucleus test is considered suitable if, for example, the impurity has the properties to react with the N7/C8 of the deoxyguanosine-forming alkaline-labile sites Deoxyguanosine is one of four nucleotides of the DNA and is the major acceptor for many electrophilic compounds [22]
Rodents are given a single dose of the test article Viable liver cells are harvested at about 2-4 hours postadministration (pa) and 12-16 hours pa by in situ collagenase perfusion The established hepatocyte cultures are incubated in the presence of 3H-thymidine for 18 hours The radiolabeled nucleotide is incorporated into the DNA in the course of DNA repair synthesis if the test article had caused DNA damage in vivo Quantification of the unscheduled DNA synthesis (UDS) response is done by silver grain counting (net nuclear grain count, ie, grains over the nucleus minus the background grains in plasma) using autoradiography The grains over the nucleus are of high density and the grains over the cytoplasma are of low density for positive compounds
The UDS test in liver has a low sensitivity for carcinogens [23] It was positive only for one-third of the Ames positive carcinogens Strong mutagenic liver carcinogens of the aryl amine class or nitro aromatic compounds were positive or equivocal in this test Therefore, this test may be used for specific chemical classes known to be metabolically activated in the liver and to have an Ames positive result in the presence of the rat liver metabolic activation system
Two different transgenic mouse strains are available for in vivo mutagenicity studies, the Big Blue® and the Muta™ mouse [24,25] These animals are transgenic for bacterial genes (λlacI and λlacZ, respectively) incorporated in all their organs and therefore suitable to detect mutations in any organ of interest Animals are treated daily for at least 28 days, and organs are sampled 3 days after the final treatment Treatment duration depends on the cell replication rate, that is, the rate of DNA synthesis It can be shorter for organs with high cell replication, for example, the hematopoietic system, skin, the gastrointestinal tract, and testes, and longer for organs with low cell replication, for example, the heart and kidneys Genomic DNA is carefully extracted from the organ of interest The DNA is packed into bacteriophages These packed bacteriophages infect a special bacterial strain of E. coli and lyse the bacterial lawn forming plaques The bacterial gene product of lacZ is β-galactosidase, which cleaves X-gal (5-bromo-4-chloro-3-indolyl-b-galactopyranoside) to galactose and an insoluble blue precipitate If a mutation occurred in the bacterial lacZ gene in the Muta mouse model, the gene product may be reduced or inactivated and, therefore, X-gal is not cleaved, that is, the mutant clear plaques are counted in a background of blue plaques The lacI gene product in the Big Blue model is a Lac repressor protein and, therefore, a mutation allows the subsequent lacZ gene product to become active The total number of mutants is determined by counting the blue (mutant) plaques among a background of clear plaques in the Big Blue model
In principle, all types of DNA-reactive chemicals that induce mutations in vivo should be detectable in these models It is possible to detect the type of mutagenic events, that is, whether transversion or transition occurred since DNA sequencing could be performed, and thus assign a specific molecular signature for specific chemical classes The spontaneous rate of mutations is age and organ dependent The isolation of high-molecular-weight and pure DNA from organs is a prerequisite for
good-quality results Necrosis or iron deposits or any other DNA-protein cross-links may interfere with the isolation process Therefore, the maximum dose should not cause severe morphological changes
The phosphatidylinositol glycan class A (pig-a) gene is located on the X chromosome and, therefore, a single mutation inactivates the gene product in all daughter cells The gene product, a catalytic subunit of N-acetylglucosaminyltransferase, is required for glycosylphosphatidyl inositol (GPI) anchor synthesis [26] Among all cell surface proteins, CD59 is anchored by GPI to the cell surface and can be easily detected in the blood using, for example, flow cytometric methods, that is, a decrease in CD59 is indicative for mutations in the pig-a gene
The method can be used in a variety of animal species including humans Furthermore, all types of DNA-reactive chemicals that induce mutations in vivo should be detectable in this model However, the detected mutation events are restricted to organs with high cell replication like the hematopoietic cell lineage Therefore, impurities that are metabolically activated, for example, in the liver or highly reactive first site of contact chemicals, are less likely to be mutagenic in the pig-a test A thorough validation will show whether the pig-a test can be used for all different types of chemical classes Furthermore, it requires repeated dosing to accumulate a sufficiently high number of mutagenic events and, therefore, the dose might be lower than what it is after acute or short-term treatment due to cumulative toxicity or intolerability
Gene expression profiles may be used to elucidate the mode of action of test articles, including genotoxicity Toxicogenomics (mRNA transcript profiling) allows the identification of genes that respond to toxic effects Investigations are ongoing whether it is possible to discriminate the gene expression profile of genotoxic carcinogens from that of nongenotoxic carcinogens and noncarcinogens [27]
In 1959, health authorities became aware that residues of the carcinogenic pesticide aminotriazole were present in cranberries They ensured that the complete cranberry harvest was analyzed just prior to Thanksgiving As a consequence, the US health authorities released the “Delaney Clause,” setting the zero cancer risk standard The Delaney Clause reads that “the FDA shall not approve for use in food any chemical additive found to induce cancer in man, or, after tests, found to induce cancer in animals”
Health authorities have the task to protect public health This includes identifying the major sources for cancer-causing agents and enforcing the avoidance or reduction of exposure to hazard-causing agents The International Agency for Cancer Registry (IARC) in France have identified repetitively that food, drinking, and lifestyle are the major risk factors that induce cancer It was acknowledged that many food components are damaging the DNA themselves However, it is the mixture of
nutrition, anticancer food components, and DNA-damaging components that needs to be balanced Furthermore, it was recognized that it is impossible to avoid carcinogens in food For pragmatic reasons, TTC concepts for different toxicological end points were established by the FDA for indirect food additives [9] The carcinogenic potency data from 709 Ames positive and therefore mutagenic rodent carcinogens were taken as the basis VSDs were calculated down to the risk level of 1 in 106 The logarithmic VSD values form a normal distribution curve in a semilogarithmic scale with the most likely potency corresponding to 05 ppb This value corresponds to an exposure of 015 µg/day assuming a daily consumption of 1500 g of food and 1500 g of fluids Therefore, if the daily lifetime exposure to a mutagen and presumptive carcinogen is kept at 015 µg/day, the cancer risk is limited to 1 excess cancer case per 1,000,000 lives In 2004, the TTC concept was published to set human exposure threshold values also for all chemicals with no toxicity data based on established toxicity profiles [28]
In the pharmaceutical industry, the toxicological qualification of impurities in drug substances and drug products is addressed in the ICH Q3A and Q3B guidelines [1,2] These guidelines ascertain that the exposure to a putative genotoxic impurity is not greater than 1 mg/day In the Safety Working Party of the European Medicines Agency (EMA), discussions on the limits of genotoxic impurities started in 2002 The EMA Guideline on the limits of genotoxic impurities [3] came into effect in January 2007 along with subsequent questions and answers on this guideline [4] In this guideline, a TTC value of 15 μg/day intake of a genotoxic impurity is considered to be associated with an acceptable risk (excess cancer risk of 1 in 100,000 over a lifetime) for most pharmaceuticals The FDA released a draft guidance [5] in December 2008 The ICH M7 guideline Assessment and control of DNA reactive (mutagenic) impurities in pharmaceuticals to limit potential carcinogenic risk is in preparation Until finalization, the respective local guidelines are active
