In human health risk assessment of chemicals and pharmaceuticals, identification of genotoxicity hazard usually starts with a standard battery of in vitro genotoxicity tests, which is needed to cover all genotoxicity endpoints. The individual tests included in the battery are not designed to pick up all endpoints. This explains why resulting data can appear contradictory, thereby complicating accurate interpretation of the findings. Such interpretation could be improved through application of mathematical modeling. One of the advantages of mathematical modeling is that the strengths and weaknesses of each test are taken into account. Furthermore, the generated predictions are objective and convey the associated uncertainties. This approach was explored by the working group "Predictivity of In Vitro Genotoxicity Testing," convened in the context of the 8th International Workshop on Genotoxicity Testing (IWGT). Specifically, we applied mathematical modeling to a database with publicly available in vitro and in vivo data for genotoxicity. The results indicate that a mammalian in vitro clastogenicity test and a mammalian cell gene mutation test together provide strong predictive weight-of-evidence for evaluating genotoxic hazard of a substance, although they are better in predicting absence of genotoxic potential than in predicting presence of genotoxic potential. Remarkably, the bacterial reverse mutation (Ames) test did not significantly change these predictions when used in combination with in vitro mutagenicity and clastogenicity tests using cells of mammalian origin. However, in case only data from a bacterial reverse mutation test are available for the assessment of genotoxic potential, these do bear weight of evidence and thus can be used. Genotoxicity assays are generally executed in tiers, in which the bacterial reverse mutation test often is the starting point. Thus, it is reasonable to suspect that early in development test results from the bacterial reverse mutation test have influenced the composition of the database studied here. We performed several tests on the robustness of the database used for the analyses presented here, and the forthcoming results do not indicate a strong bias. Further research comparing in vitro genotoxicity data with in vivo data for additional compounds will provide more insights whether it is indeed time to reconsider the composition of the standard in vitro genotoxicity battery.

BACKGROUND:Currently, revisions to the ICH S1 guidance on rodent carcinogenicity testing are being proposed. Application of this approach would reduce the use of animals in accordance with the 3Rs principles (reduce/refine/replace). The method would also shift resources to focus on more scientific mechanism-based carcinogenicity assessments and promote safe and ethical development of new small molecule pharmaceuticals. In the revised draft, findings such as cellular hypertrophy, diffuse and/or focal cellular hyperplasia, persistent tissue injury and/or chronic inflammation, preneoplastic changes, and tumors are listed as histopathology findings of particular interest for identifying carcinogenic potential. In order to predict hepatocarcinogenicity of test chemicals based on the results from 2- or 4-week repeated dose studies, we retrospectively reanalyzed the results of a previous collaborative study on the liver micronucleus assay. We focused on liver micronucleus induction in combination with histopathological changes including hypertrophy, proliferation of oval cells or bile duct epithelial cells, tissue injuries, regenerative changes, and inflammatory changes as the early responses of hepatocarcinogenesis. For these early responses, A total of 20 carcinogens, including 14 genotoxic hepatocarcinogens (Group A) and 6 non-liver-targeted genotoxic carcinogens (Group B) were evaluated.RESULTS:In the Group A chemicals, 5 chemicals (NPYR, MDA, NDPA, 2,6-DNT, and NMOR) showed all of the 6 early responses in hepatocarcinogenesis. Five chemicals (DMN, 2,4-DNT, QUN, 2-AAF, and TAA) showed 4 responses, and 4 chemicals (DAB, 2-NP, MCT, and Sudan I) showed 3 responses. All chemicals exhibited at least 3 early responses. Contrarily, in the Group B chemicals (6 chemicals), 3 of the 6 early responses were observed in 1 chemical (MNNG). No more than two responses were observed in 3 chemicals (MMC, MMS, and KA), and no responses were observed in 2 chemicals (CP and KBrO3).CONCLUSION:Evaluation of liver micronucleus induction in combination with histopathological examination is useful for detecting hepatocarcinogens. This assay takes much less time than routine long-term carcinogenicity studies.

The bacterial reverse mutation test (Ames test) is the most commonly used genotoxicity test; it is a primary component of the chemical safety assessment data required by regulatory agencies worldwide. Within the current accepted in vitro genotoxicity test battery, it is considered capable of revealing DNA reactivity, and identifying substances that can produce gene mutations via different mechanisms. The previously published consolidated EURL ECVAM Genotoxicity and Carcinogenicity Database, which includes substances that elicited a positive response in the Ames test, constitutes a collection of data that serves as a reference for a number of regulatory activities in the area of genotoxicity testing. Consequently, we considered it important to expand the database to include substances that fail to elicit a positive response in the Ames test, i.e., Ames negative substances. Here, we describe a curated collection of 211 Ames negative substances, with a summary of complementary data available for other genotoxicity endpoints in vitro and in vivo, plus available carcinogenicity data. A descriptive analysis of the data is presented. This includes a representation of the chemical space formed by the Ames-negative database with respect to other substances (e.g. REACH registered substances, approved drugs, pesticides, etc.) and a description of the organic functional groups found in the database. We also provide some suggestions on further analyses that could be made.

Genotoxicity evaluation has been widely used to estimate the carcinogenicity of test substances during safety evaluation. However, the latest strategies using genotoxicity tests give more weight to sensitivity; therefore, their accuracy has been very low. For precise carcinogenicity evaluation, we attempted to establish an integrated testing strategy for the tailor-made carcinogenicity evaluation of test materials, considering the relationships among genotoxicity test results (Ames, in vitro mammalian genotoxicity and in vivo micronucleus), carcinogenicity test results and chemical properties (molecular weight, logKow and 179 organic functional groups). By analyzing the toxicological information and chemical properties of 230 chemicals, including 184 carcinogens in the Carcinogenicity Genotoxicity eXperience database, a decision tree for carcinogenicity evaluation was optimised statistically. A decision forest model was generated using a machine-learning methodrandom forestwhich comprises thousands of decision trees. As a result, balanced accuracies in cross-validation of the optimised decision tree and decision forest model, considering chemical space (71.5% and 75.5%, respectively), were higher than balanced accuracy of an example regulatory decision tree (54.1%). Moreover, the statistical optimisation of tree-based models revealed significant organic functional groups that would cause false prediction in standard genotoxicity tests and non-genotoxic carcinogenicity (e.g., organic amide and thioamide, saturated heterocyclic fragment and aryl halide). In vitro genotoxicity tests were the most important parameters in all models, even when in silico parameters were integrated. Although external validation is required, the findings of the integrated testing strategies established herein will contribute to precise carcinogenicity evaluation and to determine new mechanistic hypotheses of carcinogenicity.

The working group reached complete or majority agreement on many issues. Results from TGR and in vivo comet assays for 91 chemicals showed they have similar ability to detect in vivo genotoxicity per se with bacterial mutagens and Ames-positive carcinogens. TGR and comet assay results were not significantly different when compared with IARC Group 1, 2 A, and unclassified carcinogens. There were significantly more comet assay positive responses for Group 2B chemicals, and for IARC classified and unclassified carcinogens combined, which may be expected since mutation is a subset of genotoxicity. A liver comet assay combined with the bone marrow/blood micronucleus (MNviv) test would detect in vivo genotoxins that do not exhibit tissue-specific or site-of-contact effects, and is appropriate for routine in vivo genotoxicity testing. Generally for orally administered substances, a comet assay at only one site-of-contact GI tract tissue (stomach or duodenum/jejunum) is required. In MNviv tests, evidence of target tissue exposure can be obtained in a number of different ways, as recommended by ICH S2(R1) and EFSA (Hardy et al., 2017). Except for special cases the i.p. route is inappropriate for in vivo testing; for risk evaluations more weight should be given to data from a physiologically relevant administration route. The liver MN test is sufficiently validated for the development of an OECD guideline. However, the impact of dosing animals > 6 weeks of age needs to be evaluated. The GI tract MN test shows promise but needs more validation for an OECD guideline. The Pig-a assay detects systemically available mutagens and is a valuable follow-up to in vitro positive results. A new freeze-thaw protocol provides more flexibility. Mutant reticulocyte and erythrocyte frequencies should both be determined. Preliminary data are available for the Pig a assay in male rat germ cells which require validation including germ cell DNA mutation orgin.

Background: The repeated-dose liver micronucleus (RDLMN) assay is an effective and important in vivo test for detecting genotoxic compounds, particularly for those that require metabolic activation to show genotoxicity. In a collaborative study by the Collaborative Study Group for the Micronucleus Test (CSGMT)/The Japanese Environmental Mutagen Society (JEMS) - Mammalian Mutagenicity Study Group (MMS), micronucleus induction of 22 chemicals with the RDLMN assay employing the collagenase digestion method was examined and reported on. Recently, we have developed a method which enables retrospective evaluation of micronucleus induction in formalin-fixed liver tissues (the formalin-fixed method) obtained in general toxicity studies completed in the past Using this method, we were able to easily evaluate clastogenic potential of chemicals from the formalin-fixed tissues obtained in the general toxicity studies. In this study, to evaluate the usefulness of the formalin-fixed method, we have conducted a liver micronucleus assay using the formalin-fixed liver samples obtained from the above collaborative study (18 of 22 test chemicals) and carried out a comparison with the results obtained by the collagenase digestion method. Results: Comparison of the collagenase digestion and formalin-fixed methods was conducted using the results of the micronucleus assays with a total of 18 test chemicals which included 12 genotoxic hepatocarcinogens (Group A), 4 genotoxic carcinogens but not liver targeted (Group B), and 2 nongenotoxic hepatocarcinogens (Group C). The formalin-fixed method obtained the similar results as the collagenase digestion method in 10 out of the 12 chemicals of Group A, and all chemicals of Group B and Group C. Although the results were statistically contradictive due to different levels of concurrent negative control, the 2 other chemicals of Group A showed comparable responses between the two methods. Conclusion: The present study shows that the formalin-fixed method is capable of detecting liver carcinogens with sensitivity equal to or higher than that of the collagenase digestion method. We recommend use of the formalin-fixed method because of its capability of enabling retrospective evaluation of micronucleus induction in the formalin-fixed liver tissues obtained in general toxicity studies completed in the past.

As part of the hazard and risk assessment of chemicals in man, it is important to assess the ability of a chemical to induce mutations in vivo. Because of the commonalities in the molecular initiating event, mutagenicity in vitro can correlate well to the in vivo endpoint for certain compound classes; however, the difficulty lies in identifying when this correlation holds true. In silico alerts for in vitro mutagenicity may therefore be used as the basis for alerts for mutagenicity in vivo where an expert assessment is carried out to establish the relevance of the correlation. Taking this into account, a data set of publicly available transgenic rodent gene mutation assay data, provided by the National Institute of Health Sciences of Japan, was processed in the expert system Derek Nexus against the in vitro mutagenicity endpoint. The resulting predictivity was expertly reviewed to assess the validity of the observed correlations in activity and mechanism of action between the two endpoints to identify suitable in vitro alerts for extension to the in vivo endpoint. In total, 20 alerts were extended to predict in vivo mutagenicity, which has significantly improved the coverage of this endpoint in Derek Nexus against the data set provided. Updating the Derek Nexus knowledge base in this way led to an increase in sensitivity for this data set against this endpoint from 9% to 66% while maintaining a good specificity of 89%.

Two major endpoints for genotoxicity tests are gene mutation and chromosome damage (CD), which includes clastogenicity and aneugenicity detected by chromosomal aberration (CA) test or micronucleus (MN) test. Many in silico prediction systems for bacterial mutagenicity (i.e. Ames test results) have been developed and marketed. They show good performance for prediction of Ames mutagenicity. On the other hand, it seems that in silico prediction of CD does not progress as much as Ames prediction. Reasons for this include different mechanisms and detection methods, many false positives and conflicting test results. However, some (quantitative) structure-activity relationship ((Q)SAR) models (e.g. Derek Nexus [Derek], ADMEWorks [AWorks] and CASE Ultra [MCase]) can predict CA test results. Therefore, performances of the three (Q)SAR models were compared using the expanded Carcinogenicity Genotoxicity eXperience (CGX) dataset for understanding current situations and future development. The constructed dataset contained 440 chemicals (325 carcinogens and 115 non-carcinogens). Sensitivity, specificity, accuracy or applicability of each model were 56.0, 86.9, 68.6 or 89.1% in Derek, 67.7, 61.5, 65.2 or 99.3% in AWorks, and 91.0, 64.9, 80.5 or 97.7% in MCase, respectively. The performances (sensitivity and accuracy) of MCase were higher than those of Derek or AWorks. Analysis of predictivity of (Q)SAR models of certain chemical classes revealed no remarkable differences among the models. The tendency of positive prediction by (Q)SAR models was observed in alkylating agents, aromatic amines or amides, aromatic nitro compounds, epoxides, halides and N-nitro or N-nitroso compounds. In an additional investigation, high sensitivity but low specificity was noted in in vivo MN prediction by MCase. Refinement of test data to be used for in silico system (e.g. consideration of cytotoxicity or re-evaluation of conflicting test results) will be needed to improve performance of CD prediction.

Due to the course of registration of pesticides in Japan, the Food Safety Commission (FSC) has the responsibility to make a risk assessment of residual pesticides and related chemicals through foods. Among the set of safety evaluations for pesticides, genotoxicity assay data are mandatory. The standard test battery for this evaluation consists of a bacterial gene mutation assay, in vitro mammalian chromosomal aberrations and/or other chromosome damage assay, and in vivo rodent micronucleus assay. These assay outcomes are used for mechanistic consideration of carcinogenicity, if any. As a rule, if a certain substance is carcinogenic and the mechanism of it includes genotoxicity, the FSC might decide it is not possible to establish the acceptable daily intake of that pesticide. Therefore, the information about genotoxicity is critical for potentially carcinogenic chemicals, whether the applied substance will be adopted and permitted for use or not as pesticides. It is important to assess fairly, carefully, and transparently, but feasible, rapid, and efficient assessment also should be taken into account. Therefore, needless to say, the assay(s) should have the sensitivity to detect potent mutagens. It is also important to be aware that the required data set should be consisted of reliable assays without certain assay(s) that give(s) false positive information or offer less of a contribution for the safety assessment.

Enzyme-based amperometric biosensors measure target analyte concentrations through the transduction of enzymatically generated electrons into a detectable current at an electrode surface. To achieve efficient electron transfer, many such systems rely on the use of diffusive electron mediators or intricate, nanostructured electrode materials, which can limit applicability. Herein, we report on the development and characterization of enzyme-based glucose biosensors operating using a combination of intramolecular electron transfer and electron self-exchange through ferrocene-containing redox mediators covalently coupled directly to glucose oxidase (GOX). To accomplish this, we applied polymer-based protein engineering to grow poly(N-(3-dimethyl(ferrocenyl)methylammonium bromide)propyl acrylamide (pFcAc) via atom-transfer radical polymerization from initiator molecules attached to the GOX surface. The use of these GOX-pFcAc conjugates allowed the fabrication of simple, paper-based glucose biosensors through crosslinking with the carrier protein human serum albumin (HSA) within a viscous chitosan solution and drop casting onto carbon paper strips. Chit-GOX-pFcAc-HSA-carbon paper biosensors yielded a glucose sensitivity of 0.33 ± 0.01 μA mM−1 cm−2 and exhibited selectivity when challenged with other sugars. The ease of sensor fabrication further afforded a tailorable response through variation of GOX-pFcAc concentration upon crosslinking and drop casting. The reported approach shows the glucose sensing capabilities of a system that merges the characteristics of bioelectronics fabricated using redox-containing polymer networks with those based on the surface modification of enzymes with electron relays.

The CEFIC-LRI-AIMT6-VITO CON4EI project (CONsortium for in vitro Eye Irritation testing strategy; 2015-2016) assessed the reliability of 7 in vitro test methods and computational models as well as established tiered-testing strategies. BCOP-LLBO is incorporated in one of the testing strategies together with SkinEthicTM HCE EIT or EpiOcularTM EIT to address the whole spectra of eye irritation potential. Currently, a validation study for the BCOP-LLBO method (sponsored by CEFIC, 2016) is finalized to demonstrate the utility and the applicability of this device to be integrated in OECD Test Guideline 437. Furthermore, in the LNE-sponsored ALT4EI project (ALTernatives for Eye Irritation; 2017) remaining data gaps for the BCOP-LLBO method will be filled to strengthen the data set. The BCOP-LLBO results of these projects will be presented. III-113

4-Methylimidazole (4-MI) is formed as a result of the Maillard reaction process, and therefore is found in many foods and beverages. It is also found in soft drinks (i.e., cola) as a by-product in the production of some caramel colors. NTP bioassays revealed clear evidence of lung carcinogenicity of 4-MI in male and female mice, but not in rats and then IARC classified 4-MI as group 2B carcinogen. Genotoxicity studies with 4-MI were negative in the Ames tests and in the erythrocyte micronucleus tests with mice or rats. US California EPA (CEPA) evaluated the testing has not been adequately comprehensive to rule out a genotoxic mode of action; as target tissue of the carcinogenicity of 4-MI was lung, the lung should be used as a source tissue for in vitro metabolic activation system. Thus, CEPA defined the No Significant Risk Level (NSRL) for 10−5 lifetime risk level of cancer by 4-MI as 29 μg/day based on the non-threshold approach. As higher levels of 4-MI than the NSRL were identified in some kinds of cola, health concerns of 4-MI were drawn the attention. On the other hand, other regulatory bodies (e.g., European Food Safety Authority, EFSA) showed no concerns of 4-MI from the use of caramel colors in food. EFSA evaluated 4-MI is not genotoxic, so, non-observed adverse effect level of 4-MI was considered to be 80 mg/kg/day. In this paper, genotoxic assessments of 4-MI in different regulatory bodies are presented and the risk evaluation of 4-MI is discussed based on new genotoxicity data.

While the in vivo genotoxicity of a compound may not always correlate well with its activity in in vitro test systems, for certain compound classes a good overlap may exist between the two endpoints. The difficulty, however, lies in establishing the cases where this relationship holds true and selecting the most appropriate protocol to highlight any potential in vivo hazard. With this in mind, a project was initiated in which existing structural alerts for in vitro chromosome damage in the expert system Derek Nexus were assessed for their relevance to in vivo activity by assessing their predictivity against an in vivo chromosome damage data set. An expert assessment was then made of selected alerts. Information regarding the findings from specific in vivo tests was added to the alert along with any significant correlations between activity and test protocol or mechanism. A total of 32 in vitro alerts were updated using this method resulting in a significant improvement in the coverage of in vivo chromosome damage in Derek Nexus against a data set compiled by the mammalian mutagenicity study group of Japan. The detailed information relating to in vivo activity and protocol added to the alerts in combination with the mechanistic information provided will prove useful in directing the further testing of compounds of interest.

New OECD test guidelines have been issued, in which the cytotoxicity index relative cell count (RCC) is replaced with a new index, RICC or RPD (relative increase in cell count/relative population doubling), with the goal of reducing the high proportion of false positive results in in vitro chromosomal aberration tests. Using a mathematical approach to estimate new indices from the RCC, we constructed an evaluation flow that quantitatively estimates how often the previous test conclusions change when applying the updated cytotoxicity criteria. The new evaluation flow was applied to a retrospective evaluation of 285 chemicals in two databases. The effects of the employment of new cytotoxicity indices are investigated at a large scale. Using the new evaluation flow, 90 chemicals were estimated as positive, 39 were designated as estimated negative (13 probably negative and 26 possibly negative), and 140 were designated as negative. Moreover, we also applied a prioritization index to indicate the likelihood of a chemical being re-evaluated as negative and assigned priorities for testing. Most of the chemicals that were designated as estimated negative and had negative results in the in vivo micronucleus tests were considered as false-positives that would be correctly judged under the new test guideline. Furthermore, statistical analysis of the frequency of estimated negatives revealed that the results for Ames-positive chemicals, especially those with a strong response, are unlikely to change. Therefore, we concluded that the new indices would likely reduce the proportion of false positive results and not increase the proportion of false negative results. This study is the first report of a comprehensive re-evaluation of test results in terms of new cytotoxicity indices. The evaluation flow we have developed facilitates efficient retrospective evaluation of genotoxicity.

The in vivo mutation assay using the X-linked phosphatidylinositol glycan class A gene (Pig-a in rodents, PIG-A in humans) is a promising tool for evaluating the mutagenicity of chemicals. Approaches for measuring Pig-a mutant cells have focused on peripheral red blood cells (RBCs) and reticulocytes (RETs) from rodents. The recently developed PIGRET assay is capable of screening >1 × 106 RETs for Pig-a mutants by concentrating RETs in whole blood prior to flow cytometric analysis. Additionally, due to the characteristics of erythropoiesis, the PIGRET assay can potentially detect increases in Pig-a mutant frequency (MF) sooner after exposure compared with a Pig-a assay targeting total RBCs (RBC Pig-a assay). In order to test the merits and limitations of the PIGRET assay as a short-term genotoxicity test, an interlaboratory trial involving 16 laboratories was organized by the Mammalian Mutagenicity Study Group of the Japanese Environmental Mutagenicity Society (MMS/JEMS). First, the technical proficiency of the laboratories and transferability of the assay were confirmed by performing both the PIGRET and RBC Pig-a assays on rats treated with single doses of N-nitroso-N-ethylurea. Next, the collaborating laboratories used the PIGRET and RBC Pig-a assays to assess the mutagenicity of a total of 24 chemicals in rats, using a single treatment design and mutant analysis at 1, 2, and 4 weeks after the treatment. Thirteen chemicals produced positive responses in the PIGRET assay; three of these chemicals were not detected in the RBC Pig-a assay. Twelve chemicals induced an increase in RET Pig-a MF beginning 1 week after dosing, while only 3 chemicals positive for RBC Pig-a MF produced positive responses 1 week after dosing. Based on these results, we conclude that the PIGRET assay is useful as a short-term test for in vivo mutation using a single-dose protocol.

2,2′,3,3′,4,4′,5,5′,6,6′-Decabrominated diphenyl ether (PBDE 209) is the second most used brominated flame retardant (BFRs) in constructed materials because it is considered less toxic than others, though other fire retardants, some congeners of PBDE 209, are reported to be toxic. This combined the fact that PBDE 209 has been found in high levels in human milk, blood, indoor environments as well as in foodstuffs has led us in this study attempt to find out whether PBDE 209, also known as decaBDE and decabrominated diphenyl oxide (DBDPO), has an adverse effect on this histology of testes and sperm in CD-1 male mice. The mice we studied were divided into groups and gavaged with 10, 100, 500 and 1500 mg/kg PBDE 209 in corn oil per day between postnatal Days 21 and 70. On Day 71, the mice were anesthetized and sperm function, testis DNA content, and histopathology were studied. We found in the 500- and 1500-mg/kg/day groups that neonatal exposure to PBDE 209 reduced sperm epididymal sperm mitochondrial membrane potential (MMP), reduced amplitude of the lateral head displacement (ALH) and induced the generation of hydrogen peroxide (H2O2) in the sperm of sexually mature male mice, without affecting the sperm count, motility, morphology, curvilinear velocity (VCL), angular progressive velocity (VAP), straight-line velocity (VSL), beat-cross frequency (BCF), sperm chromatin structure assay (SCSA), superoxide anion (O2−) generation, DNA content in testis cells, or testicular histopathology. ALH was positively associated with an increase in MMP and negatively associated with generation of sperm H2O2. The reduction of MMP was negatively associated with an increase in generation of sperm H2O2. The presence of the relationships between sperm ALH, MMP, and generation of H2O2 indicate toxic action possibly resulting from PBDE 209-induced oxidative stress. In conclusion, this is the first study to report the lowest-observed-adverse-effect level (LOAEL) for sperm function to be 500 mg/kg of PBDE 209 in male mice. Decreased epididymal sperm MMP and ALH as well as induced generation of sperm H2O2 were some of the most serious effects of postnatal PBDE 209 exposure. Future investigations should be performed to study the effects of prenatal exposure of PBDE 209 and the mechanism behind PBDE 209-related oxidative stress in the fetal and pubertal stages of development.

We investigated the performance of an integrated approach to testing and assessment (IATA), designed to cover different genotoxic mechanisms causing cancer and to replicate measured carcinogenicity data included in a new consolidated database. Genotoxic carcinogenicity was predicted based on positive results from at least two genotoxicity tests: one in vitro and one in vivo (which were associated with mutagenicity categories according to the Globally Harmonized System classification). Substances belonging to double positives mutagenicity categories were assigned to be genotoxic carcinogens. In turn, substances that were positive only in a single mutagenicity test were assigned to be mutagens. Chemicals not classified by the selected genotoxicity endpoints were assigned to be negative genotoxic carcinogens and subsequently evaluated for their capability to elicit non‐genotoxic carcinogenicity. However, non‐genotoxic carcinogenicity mechanisms were not currently included in the developed IATA. The IATA is docked to the OECD Toolbox and uses measured data for different genotoxicity endpoints when available. Alternatively, the system automatically provides predictions by SAR genotoxicity models using the OASIS Tissue Metabolism Simulator platform. When the developed IATA was tested against the consolidated database, its performance was found to be high, with sensitivity of 74% and specificity of 83%, when measured carcinogenicity data were used along with predictions falling within the models' applicability domains. Performance of the IATA would be slightly changed to a sensitivity of 80% and specificity of 72% when the evaluation by non‐genotoxic carcinogenicity mechanisms was taken into account. Copyright © 2016 John Wiley & Sons, Ltd.