Although amidopropyl betaine (APB) is a major amphoteric surfactant, risk assessment is not enough because of lacking detailed usage information of consumer products and environmental fate data.

In this study, we investigated usage scenario and concentration in products. And we also studied a half-life of biodegradation in surface river water, key data for environmental fate. Utilizing those information, human exposure level and predicted environmental concentration were calculated less than the no-effect level. Therefore, we concluded that APB is safe as currently used in consumer products.

In a screening assessment under the Japanese Chemical Substances Control Law, the validity of analogue substances is required when conducting read-across. In the present study, we established a method for collecting and selecting analogue substance candidates, based on the quantitative similarity and statistical clustering using multiple databases and software. We were able to obtain candidates for analogue substances for 1-nonanol, cyclamen aldehyde and tetramethyl ammonium hydroxide using this method, indicating that this established method is appropriate for these three chemicals.

In a screening assessment under the Japanese Chemical Substances Control Law, the toxicity of nonan-1-ol (C₉H₂₀O; C9) has not been evaluated due to lack of data. In this study, we assessed the toxicity of C9 by read-across using the data of analogous substances. The ranges of NOAELs for repeated dose toxicity and reproductive/developmental toxicity were 370-2000 mg/kg/day and 130-2000 mg/kg/day, respectively. Results of all genotoxicity tests for the analogue substances were negative. Based on these results, hazard classes of the repeated dose toxicity and the genotoxicity of C9 were evaluated as “not classified”, and the reproductive/developmental toxicity was classified into “class 4”.

Sources of water supply can be influenced by chemical accidents, and the concentration of contaminants may transiently exceed the level of the guidance values (GVs) such as Drinking Water Quality Standards (DWQS) or non-legally binding target values. In such cases, the subacute guidance values (SGVs), i.e. concentrations of drinking-water contaminants that will not affect human health for subacute exposure duration, will be useful for risk management. We previously proposed SGVs of 43 organic items that can be contaminated in drinking-water, and most SVGs were calculated to be 4 to >10 times higher than the corresponding GVs. In this study, we attempted to derive SGVs for 6 inorganic metals additionally. The NOAEL or BMDL10 of a subacute study was basically used with a total uncertainty factor of 100 to establish Subacute Reference Dose (saRfD). SGVs were calculated by using the saRfDs and allocation factor of 100%. SGVs of cadmium and selenium whose GVs were derived by human epidemiological studies were also calculated as 3-4 times higher than GVs. On the other hand, SGVs for a few items such as arsenic and chromium (VI) were no alternative but to be evaluated as same as GVs. For these contaminants, immediate action will be required. Taken together, the SGVs we proposed in this and previous studies are considered useful for risk management of the contaminants in drinking-water. ACKNOWLEGMENT: This study was supported by a Health and Labour Sciences Research Grant (19LA1005) from the Ministry of Health, Labour and Welfare, Japan.

Assessment of mutagenicity for substances permitted by the positive list (PL) for food contact materials is ongoing in Japan. To demonstrate what is necessary for decreasing uncertainty as regards risk assessment of PL for nanomaterials, here we focused on genotoxic responses of both bulk and nano-sized titanium dioxide. The results of in vitro and in vivo studies show that bulk titanium dioxide has no mutagenicity. It is suggested that oxidative stress in cells induced by nano-sized titanium dioxide can be responsible for indirect genotoxic effects. In general, further consideration will be necessary to evaluate nanomaterials under PL system.

The computerized system has led to increases in efficiency and reliability in Good Laboratory Practice (GLP) facilities. On the other hand, digitalized data and records produced new problems in quality assurance such as difficulties in detecting the falsified signs, distinguishing between originals and duplications, and insufficient duration of electronic recording media; however, such problems have been solved by proper CSV implementation and operational procedure setting. The computerized system, which allows more objective and rigid management of data and records than paper, has become a central core in GLP quality assurance.

Today’s innovative IT technologies have made remarkable progress and have been creating the products and services we could not have previously imagined. New IT technologies have permeated our daily lives as common services within a short period and are also expected to bring more efficiency within GLP facilities. However, the introduction of innovative IT technologies has been hesitant in GLP facilities and they are hardly pervasive. This is likely to be caused by difficulty in establishing the new method for quality assurance because there is a case in which those IT technologies are hard to apply to the conventional CSV approaches. In order to overcome this, it is important that the industry and regulatory authorities share the GLP’s future image brought about by innovative IT technologies to seek problems and the counter-measures.

Therefore, the study group of Japan Society of Quality Assurance (JSQA) GLP Subcommittee 3 started the investigation by envisioning the pictures of near-future GLP facilities which had introduced the latest IT technologies. From considerations, we outlined “Prospects of innovation for GLP data storage based on blockchain technology” and “Basic approach to quality assurance for AI products” on the poster.

Submission of toxicity study data in SEND format was mandated by the FDA. SEND data can give new insights into toxicology research through visualization and analysis. TIBCO Spotfire is a widely used visualization tool. CNUV is a community for non-clinical data utilization and visualization using Spotfire SEND Viewer (SSV). Here, we introduce data visualization examples and present issues of SEND data creation in consideration of more effective data utilization. We hope that visualization of SEND data using SSV will contribute to further development of toxicology research.

As part of the quality assurance of SEND data, QC work needs to be performed with the corresponding part of the final report. However, searching for the relevant part in the final report every time when performing QC may cause variations depending on the experience of the personnel. Therefore, TF4 prepared a “correspondence table” between items described in the final report and SEND data. Our “correspondence table” is expected to be a useful tool that performs QC efficiently; the items in the SEND data to be checked in QC are clarified, and the variations between personnel should be reduced.

We have conducted a qualitative analysis of SEND domains related to study design. The outcome of the analysis revealed wide variations in how facilities populate each variable and whether or not facilities populate certain variables. Our group conduced similar analysis on SEND domains related to study data (EX, DS, BW, CL, DD, FW, LB, MA, MI, OM, PC, and PP). In general, we found similar variability in these domains among facilities. We will report the details of such variations as well as some examples that are considered critical when we accumulate SEND data for future data utilization.

SEND is the standard for electric data of nonclinical studies developed by CDISC. Single/repeated dose toxicity studies, carcinogenicity study, and cardiovascular and respiratory safety pharmacology studies are the scope of SEND. For genotoxicity studies, SEND implementation guide (IG) is under development by CDISC. In this presentation, we will introduce the points to be considered for the future SEND implementation of genotoxicity studies, obtained from our activities such as mock dataset preparation according to the latest draft genotoxicity IG.

The FDA has mandated submission of standardized electronic data in compliance with CDISC SEND. However, SEND-specific personnel mainly create SEND data and involvement of toxicologists is often limited because the creation requires specialized knowledge and skill such as SENDIG and use of xpt files. CJUG SEND team has created a new check list summarizing points to be considered when creating SEND data. In this presentation, importance of involvement of toxicologists in creating SEND data and in solving issues in the creation process with efficient use of the check list will be introduced.

The FDA has mandated submission of standardized electronic data in compliance with CDISC SEND. In SEND creation process, the creators and their sponsors have often suffered the variation of SEND data specification, which would cause the difficulties in using the data for the integrated study analysis and force the creators to inefficient QC tasks. Define-XML is a potential tool to resolve this issue. In this presentation, we will introduce a mock-up template for preparing a Define-XML, and propose its effective use for communicating the specification between the creators and sponsors.

SEND to be submitted to the U.S. FDA must be created in conformance with SENDIG v.3.1 from 2019 (for studies started after May 15, 2019 for NDA and those started after May 15, 2020 for IND). In addition to the currently-supported study types (single- and repeated-dose general toxicity and carcinogenicity studies), SENDIG v3.1 adds support for safety pharmacology studies (cardiovascular and respiratory data). Although there are only 2 domains [Cardiovascular Test Results (CV) domain and Respiratory Test Results (RE) domain] newly added to SENDIG v3.1, some data can be difficult to handle in creation of SEND datasets. For instance, when we select domains based on the controlled terminologies, the newly added domains (CV and RE) make it difficult for us to decide which domain to use to populate blood pressure, heart rate, respiratory rate and blood gas data. For another example, [--NOMDY], one of the newly-added variables in SENDIG v3.1, should be filled with a study day counted relative to the start date of dosing. However, in a study design such as a Latin square design where each animal is treated with multiple dose levels, it can be difficult to judge which study day of the dose to use to count a relative day from. This presentation will introduce points of SEND dataset creation for the safety pharmacology studies based upon the experience obtained at Ina Research Inc. and the feedback from FDA in SENDIG v3.1 Fit for Use Pilot.

New drug applications in the United States America are required to submit electronic data in the CDISC standard, and pharmaceutical companies are preparing datasets in accordance with SEND that a non-clinical trial data standard. SEND is expected to be used for visualizing study data. However, there are points to consider when visualizing study data. SEND cannot completely standardize the data. It is known that the data storage method differs depending on the creator. In this presentation, we will introduce examples of visualization and their points.

SEND datasets from certain types of nonclinical safety studies are required for new drug development in US. nSDRG is an important document to explain the summary of SEND datasets. PhUSE guideline and template are very informative in preparing the nSDRG; but it is not easy to understand the details due to its complexity. The nSDRG subteam of CJUG SEND team picked up the issues in preparing nSDRG and discussed how to deal with them. In this presentation, we will introduce such issues and provide some points to consider for solving them.

Promoting a preclinical small molecule candidate to clinical phases is a high-risk decision taken on the basis of a large amount of heterogenous in vitro and in vivo data that may, or may not, translate into humans. Collecting safety data along the path from preclinical, through clinical and all the way up to the post-marketing life of a drug provides a unique perspective of the correspondence, validity, evolution, and detection of drug safety signals.

Chemotargets CLARITY is an integrative drug discovery platform widely used across pharmaceutical companies, large academic institutions and non-for-profit organizations worldwide [1,2]. It uses artificial intelligence on carefully curated data to connect molecules, protein targets, safety events, and therapeutic indications with the help of state-of-the-art analytics tools. Beyond a knowledge-based source that can be interrogated from any discovery angle, the platform can be used to identify the probable primary and secondary targets of small molecules and their predicted metabolites by rapidly profiling through consensus ligand-based models of 4,799 proteins built from affinity data for almost 3 million small molecules. In addition, machine learning models of 1,278 safety endpoints were constructed from data extracted from almost 16 million spontaneous reports, including hundreds of safety-related mechanisms annotated with preclinical toxicity and clinical safety issues.

The new version of Chemotargets CLARITY (version 5) will offer a new translational drug safety dashboard that will allow users to analyze the evolution of safety issues for a given drug, from the initial preclinical stages in animals to the post-marketing exposure to humans.

[1] Ellis et al. (2020) Evaluating kratom alkaloids using PHASE. PLoS ONE 15, e0229646.

[2] Ellis et al. (2019) Assessing the structural and pharmacological similarity of newly identified drugs of abuse to controlled substances using Publich Health Assessment via Structural Elucidation. Clin. Pharmacol. Ther. 106, 116-122.

In silico analysis have been expected as an alternative for animal testing of toxicity. Recently, deep learning (DL) has a focus on image classification. Thus, a novel DL-based quantitative structure-activity relationship method, called DeepSnap-DL that applied molecular image into DL, has been developed in our laboratory. This study revealed that DeepSnap-DL shows high-prediction performance in the activation of the progesterone receptor that could mediate key initiating events on adverse outcome pathway, and outperforms conventional machine learning technics.

Marketed and withdrawn drugs provide a rich source of information on the clinical effects of pharmacological intervention which can be brought back in to the Pre-Clinical space to improve the translational and predictive power of non-clinical assays and models. Accessing adverse event report data (FAERS) and clinical PK in a structured manner, through PharmaPendium, enables improved understanding of the performance non-clinical assays and facilitates more detailed risk assessment for candidate drugs.

[Background, Objective] Gastrointestinal toxicity, such as the injury to the mucosal epithelial cells in the gastrointestinal tract, is a common adverse event among cytotoxic anticancer drugs and has become one of the dose-limiting factors for many drugs. Evaluation of gastrointestinal toxicity in humans is extremely important from the viewpoint of selecting compounds and predicting side effects. However, since the primary culture of gastrointestinal epithelial cells was very difficult, cell-based model system mimicking physiological condition that stably and universally evaluates gastrointestinal toxicity had not been established. Recently, an intestinal organoids culture method has been established that is capable of culture maintenance in a state close to an in vivo tissue using intestinal epithelial stem cells or iPS cells. In this study, we evaluated the in vitro gastrointestinal toxicity of cytotoxic anticancer drugs using intestinal organoids derived from mouse small intestine and human iPS cells.

[Methods] The mouse intestinal organoids were prepared from the small intestine of C57BL/6 mice, and the human intestinal organoids were differentiated from human iPS cells. SN-38 (active metabolite of irinotecan) and cisplatin were used as the evaluation compounds. On day 3 after seeding, intestinal organoids were treated with various concentrations of each compound. After treatment (2, 6, 24, 48 and 72 hours), cytotoxicity was evaluated by LDH assay.

[Results] SN-38 and cisplatin did not show any harmful effects after short time exposure, and showed a significant damaging effect as the exposure time became longer. In comparison of human and mouse injury responsiveness, cisplatin was similar in human and mouse intestinal organoids, but SN-38 showed a strong damaging effect in human intestinal organoids, and there were species differences in sensitivity.

[Conclusions] This method was considered to be useful as an in vitro system for evaluating the damaging effects and species differences in sensitivity of cytotoxic anticancer drugs on gastrointestinal toxicity.