Preclinical strategy to reduce clinical hepatotoxicity using in vitro bioactivation data for >200 compounds

Considerations from an International Regulatory and Pharmaceutical Industry (IQ MPS Affiliate) Workshop on the Standardization of Complex In Vitro Models in Drug Development

In May 2022, there is an International Regulatory and Pharmaceutical Industry (Innovation and Quality [IQ] Microphysiological Systems [MPS] Affiliate) Workshop on the standardization of complex in vitro models (CIVMs) in drug development. This manuscript summarizes the discussions and conclusions of this joint workshop organized and executed by the IQ MPS Affiliate and the United States Food and Drug Administration (FDA). A key objective of the workshop is to facilitate discussions around opportunities and/or needs for standardization of MPS and chart potential pathways to increase model utilization in the context of regulatory decision making. Participation in the workshop included 200 attendees from the FDA, IQ MPS Affiliate, and 26 global regulatory organizations and affiliated parties representing Europe, Japan, and Canada. It is agreed that understanding global perspectives regarding the readiness of CIVM/MPS models for regulatory decision making and potential pathways to gaining acceptance is useful to align on globally. The obstacles are currently too great to develop standards for every context of use (COU). Instead, it is suggested that a more tractable approach may be to think of broadly applicable standards that can be applied regardless of COU and/or organ system. Considerations and next steps for this effort are described.

Cyanide Trapping of Iminium Ion Reactive Metabolites: Implications for Clinical Hepatotoxicity

Reactive metabolite formation is a major mechanism of hepatotoxicity. Although reactive electrophiles can be soft or hard in nature, screening strategies have generally focused on the use of glutathione trapping assays to screen for soft electrophiles, with many data sets available to support their use. The use of a similar assay for hard electrophiles using cyanide as the trapping agent is far less common, and there is a lack of studies with sufficient supporting data. Using a set of 260 compounds with a defined hepatotoxicity status by the FDA, a comprehensive literature search yielded cyanide trapping data on an unbalanced set of 20 compounds that were all clinically hepatotoxic. Thus, a further set of 19 compounds was selected to generate cyanide trapping data, resulting in a more balanced data set of 39 compounds. Analysis of the data demonstrated that the cyanide trapping assay had high specificity (92%) and a positive predictive value (83%) such that hepatotoxic compounds would be confidently flagged. Structural analysis of the adducts formed revealed artifactual methylated cyanide adducts to also occur, highlighting the importance of full structural identification to confirm the nature of the adduct formed. The assay was demonstrated to add the most value for compounds containing typical structural alerts for hard electrophile formation: half of the severe hepatotoxins with these structural alerts formed cyanide adducts, while none of the severe hepatotoxins with no relevant structural alerts formed adducts. The assay conditions used included cytosolic enzymes (e.g., aldehyde oxidase) and an optimized cyanide concentration to minimize the inhibition of cytochrome P450 enzymes by cyanide. Based on the demonstrated added value of this assay, it is to be initiated for use at GSK as part of the integrated hepatotoxicity strategy, with its performance being reviewed periodically as more data is generated.

3D human tissue models and microphysiological systems for HIV and related comorbidities

Three-dimensional (3D) human tissue models/microphysiological systems (e.g., organs-on-chips, organoids, and tissue explants) model HIV and related comorbidities and have potential to address critical questions, including characterization of viral reservoirs, insufficient innate and adaptive immune responses, biomarker discovery and evaluation, medical complexity with comorbidities (e.g., tuberculosis and SARS-CoV-2), and protection and transmission during pregnancy and birth. Composed of multiple primary or stem cell-derived cell types organized in a dedicated 3D space, these systems hold unique promise for better reproducing human physiology, advancing therapeutic development, and bridging the human-animal model translational gap. Here, we discuss the promises and achievements with 3D human tissue models in HIV and comorbidity research, along with remaining barriers with respect to cell biology, virology, immunology, and regulatory issues.

Prediction of cytochrome P450-mediated bioactivation using machine learning models and in vitro validation

Cytochrome P450 (P450)-mediated bioactivation, which can lead to the hepatotoxicity through the formation of reactive metabolites (RMs), has been regarded as the major problem of drug failures. Herein, we purposed to establish machine learning models to predict the bioactivation of P450. On the basis of the literature-derived bioactivation dataset, models for Benzene ring, Nitrogen heterocycle and Sulfur heterocycle were developed with machine learning methods, i.e., Random Forest, Random Subspace, SVM and Naïve Bayes. The models were assessed by metrics like "Precision", "Recall", "F-Measure", "AUC" (Area Under the Curve), etc. Random Forest algorithms illustrated the best predictability, with nice AUC values of 0.949, 0.973 and 0.958 for the test sets of Benzene ring, Nitrogen heterocycle and Sulfur heterocycle models, respectively. 2D descriptors like topological indices, 2D autocorrelations and Burden eigenvalues, etc. contributed most to the models. Furthermore, the models were applied to predict the occurrence of bioactivation of an external verification set. Drugs like selpercatinib, glafenine, encorafenib, etc. were predicted to undergo bioactivation into toxic RMs. In vitro, IC50 shift experiment was performed to assess the potential of bioactivation to validate the prediction. Encorafenib and tirbanibulin were observed of bioactivation potential with shifts of 3-6 folds or so. Overall, this study provided a reliable and robust strategy to predict the P450-mediated bioactivation, which will be helpful to the assessment of adverse drug reactions (ADRs) in clinic and the design of new candidates with lower toxicities.

Evaluation of the Drug-Induced Liver Injury Potential of Saxagliptin through Reactive Metabolite Identification in Rats

A liver injury was recently reported for saxagliptin, which is a dipeptidyl peptidase-4 (DPP-4) inhibitor. However, the underlying mechanisms of saxagliptin-induced liver injury remain unknown. This study aimed to evaluate whether saxagliptin, a potent and selective DPP-4 inhibitor that is globally used for treating type 2 diabetes mellitus, binds to the nucleophiles in vitro. Four DPP-4 inhibitors, including vildagliptin, were evaluated for comparison. Only saxagliptin and vildagliptin, which both contain a cyanopyrrolidine group, quickly reacted with L-cysteine to enzyme-independently produce thiazolinic acid metabolites. This saxagliptin-cysteine adduct was also found in saxagliptin-administered male Sprague-Dawley rats. In addition, this study newly identified cysteinyl glycine conjugates of saxagliptin and 5-hydroxysaxagliptin. The observed metabolic pathways were hydroxylation and conjugation with cysteine, glutathione, sulfate, and glucuronide. In summary, we determined four new thiazoline-containing thiol metabolites (cysteine and cysteinylglycine conjugates of saxagliptin and 5-hydroxysaxagliptin) in saxagliptin-administered male rats. Our results reveal that saxagliptin can covalently bind to the thiol groups of cysteine residues of endogenous proteins in vivo, indicating the potential for saxagliptin to cause drug-induced liver injury.

A blinded in vitro analysis of the intrinsic immunogenicity of hepatotoxic drugs: implications for preclinical risk assessment

In vitro preclinical drug-induced liver injury (DILI) risk assessment relies largely on the use of hepatocytes to measure drug-specific changes in cell function or viability. Unfortunately, this does not provide indications toward the immunogenicity of drugs and/or the likelihood of idiosyncratic reactions in the clinic. This is because the molecular initiating event in immune DILI is an interaction of the drug-derived antigen with MHC proteins and the T-cell receptor. This study utilized immune cells from drug-naïve donors, recently established immune cell coculture systems and blinded compounds with and without DILI liabilities to determine whether these new methods offer an improvement over established assessment methods for the prediction of immune-mediated DILI. Ten blinded test compounds (6 with known DILI liabilities; 4 with lower DILI liabilities) and 5 training compounds, with known T-cell-mediated immune reactions in patients, were investigated. Naïve T-cells were activated with 4/5 of the training compounds (nitroso sulfamethoxazole, vancomycin, Bandrowski's base, and carbamazepine) and clones derived from the priming assays were activated with drug in a dose-dependent manner. The test compounds with DILI liabilities did not stimulate T-cell proliferative responses during dendritic cell-T-cell coculture; however, CD4+ clones displaying reactivity were detected toward 2 compounds (ciprofloxacin and erythromycin) with known liabilities. Drug-responsive T-cells were not detected with the compounds with lower DILI liabilities. This study provides compelling evidence that assessment of intrinsic drug immunogenicity, although complex, can provide valuable information regarding immune liabilities of some compounds prior to clinical studies or when immune reactions are observed in patients.

Machine Learning Enables Accurate Prediction of Quinone Formation during Drug Metabolism

Metabolism helps in the elimination of drugs from the human body by making them more hydrophilic. Sometimes, drugs can be bioactivated to highly reactive metabolites or intermediates during metabolism. These reactive metabolites are often responsible for the toxicities associated with the drugs. Identification of reactive metabolites of drug candidates can be very helpful in the initial stages of drug discovery. Quinones are soft electrophiles that are generated as reactive intermediates during metabolism. Quinones make up more than 40% of the reactive metabolites. In this work, a reliable data set of 510 molecules was used to develop machine learning and deep learning-based predictive models to predict the formation of quinone-type metabolites. For representing molecules, two-dimensional (2D) descriptors, PubChem fingerprints, electro-topological state (E-state) fingerprints, and metabolic reactivity-based descriptors were used. Developed models were compared to the existing Xenosite web server using the untouched test set of 102 molecules. The best model achieved an accuracy of 86.27%, while the Xenosite server could achieve an accuracy of only 52.94% on the test set. Descriptor analysis revealed that the presence of greater numbers of polar moieties in a molecule can prevent the formation of quinone-type metabolites. In addition, the presence of a nitrogen atom in an aromatic ring and the presence of metabolophores V51, V52, and V53 (SMARTCyp descriptors) decrease the probability of quinone formation. Finally, a tool based on the best machine learning models was developed, which is accessible at http://14.139.57.41/quinonepred/.

Bioactivation and reactivity research advances - 2022 year in review‡

With the 50th year mark since the launch of Drug Metabolism and Disposition journal, the field of drug metabolism and bioactivation has advanced exponentially in the past decades (Guengerich 2023).This has, in a major part, been due to the continued advances across the whole spectrum of applied technologies in hardware, software, machine learning (ML), and artificial intelligence (AI). LC-MS platforms continue to evolve to support key applications in the field, and automation is also improving the accuracy, precision, and throughput of these supporting assays. In addition, sample generation and processing is being aided by increased diversity and quality of reagents and bio-matrices so that what is being analyzed is more relevant and translatable. The application of in silico platforms (applied software, ML, and AI) is also making great strides, and in tandem with the more traditional approaches mentioned previously, is significantly advancing our understanding of bioactivation pathways and how these play a role in toxicity. All of this continues to allow the area of bioactivation to evolve in parallel with associated fields to help bring novel or improved medicines to patients with urgent or unmet needs.Shuai Wang and Cyrus Khojasteh, on behalf of the authors.

Gelatin-Based Hybrid Hydrogel Scaffolds: Toward Physicochemical Liver Mimicry

There exists a clear need to develop novel materials that could serve liver tissue engineering purposes. Those materials need to be researched for the development of bioengineered liver tissue as an alternative to donor livers, as well as for materials that could be applied for scaffolds to develop an in vitro model for drug-induced liver injury (DILI) detection . In this paper, the hydrogels oxidized dextran-gelatin (Dexox-Gel) and norbornene-modified dextran-thiolated gelatin (DexNB-GelSH) were developed, and their feasibility toward processing via indirect 3D-printing was investigated with the aim to develop hydrogel scaffolds that physicochemically mimic native liver tissue. Furthermore, their in vitro biocompatibility was assessed using preliminary biological tests using HepG2 cells. Both materials were thoroughly physicochemically characterized and benchmarked to the methacrylated gelatin (GelMA) reference material. Due to inferior properties, Dexox-gel was not further processed into 3D-hydrogel scaffolds. This research revealed that DexNB-GelSH exhibited physicochemical properties that were in excellent agreement with the properties of natural liver tissue in contrast to GelMA. In combination with an equally good biological evaluation of DexNB-GelSH in comparison with GelMA based on an MTS proliferation assay and an albumin quantification assay, DexNB-GelSH can be considered promising in the field of liver tissue engineering.

Elucidation of the relationship between evodiamine-induced liver injury and CYP3A4-mediated metabolic activation by UPLC-MS/MS analysis

Evodiamine (EVD), which has been reported to cause liver damage, is the main constituent of Evodia rutaecarpa (Juss.) Benth and may be bioactivated into reactive metabolites mediated by cytochrome P450. However, the relationships between bioactivation and EVD-induced hepatotoxicity remain unknown. In this study, comprehensive hepatotoxicity evaluation was explored, which demonstrated that EVD caused hepatotoxicity in both time- and dose-dependent manners in mice. By application of UPLC-Q/TOF-MS/MS, two GSH conjugates (GM1 and GM2) derived from reactive metabolites of EVD were identified, in microsomal incubation systems exposed to EVD with glutathione (GSH) as trapping agents. CYP3A4 was proved to be the main metabolic enzyme. Correspondingly, the N-acetyl-L-cysteine conjugate derived from the degradation of GM2 was detected in the urine of mice after exposure to EVD. For the first time, the iminoquinone intermediate was found in EVD-pretreated rat bile by the high-resolution MS platform. Pretreatment with ketoconazole protected the animals from hepatotoxicity, decreased the protein expression of cleaved caspase-1 and -3, but increased the area under the serum-concentration-time curve of EVD in blood determined by UPLC-QQQ-MS/MS. Depletion of GSH by buthionine sulfoximine exacerbated EVD-induced hepatotoxicity. These results implicated that the CYP3A4-mediated metabolic activation was responsible for the observed hepatotoxicity induced by EVD.

Drug-induced liver injury: a comprehensive review

Drug-induced liver injury (DILI) remains a challenge in clinical practice and is still a diagnosis of exclusion. Although it has a low incidence amongst the general population, DILI accounts for most cases of acute liver failure with a fatality rate of up to 50%. While multiple mechanisms of DILI have been postulated, there is no clear causal relationship between drugs, risk factors and mechanisms of DILI. Current best practice relies on a combination of high clinical suspicion, thorough clinical history of risk factors and timeline, and extensive hepatological investigations as supported by the international Roussel Uclaf Causality Assessment Method criteria, the latter considered a key diagnostic algorithm for DILI. This review focuses on DILI classification, risk factors, clinical evaluation, future biomarkers and management, with the aim of facilitating physicians to correctly identify DILI early in presentation.

MiRNA Differences Related to Treatment-Resistant Schizophrenia

Schizophrenia (SZ) is a serious mental disorder that is typically treated with antipsychotic medication. Treatment-resistant schizophrenia (TRS) is the condition where symptoms remain after pharmacological intervention, resulting in long-lasting functional and social impairments. As the identification and treatment of a TRS patient requires previous failed treatments, early mechanisms of detection are needed in order to quicken the access to effective therapy, as well as improve treatment adherence. In this study, we aim to find a microRNA (miRNA) signature for TRS, as well as to shed some light on the molecular pathways potentially involved in this severe condition. To do this, we compared the blood miRNAs of schizophrenia patients that respond to medication and TRS patients, thus obtaining a 16-miRNA TRS profile. Then, we assessed the ability of this signature to separate responders and TRS patients using hierarchical clustering, observing that most of them are grouped correctly (~70% accuracy). We also conducted a network, pathway analysis, and bibliography search to spot molecular pathways potentially altered in TRS. We found that the response to stress seems to be a key factor in TRS and that proteins p53, SIRT1, MDM2, and TRIM28 could be the potential mediators of such responses. Finally, we suggest a molecular pathway potentially regulated by the miRNAs of the TRS profile.

In vitro cell-based models of drug-induced hepatotoxicity screening: progress and limitation

Drug-induced liver injury (DILI) is one of the major causes of post-approval withdrawal of therapeutics. As a result, there is an increasing need for accurate predictive in vitro assays that reliably detect hepatotoxic drug candidates while reducing drug discovery time, costs, and the number of animal experiments. In vitro hepatocyte-based research has led to an improved comprehension of the underlying mechanisms of chemical toxicity and can assist the prioritization of therapeutic choices with low hepatotoxicity risk. Therefore, several in vitro systems have been generated over the last few decades. This review aims to comprehensively present the development and validation of 2D (two-dimensional) and 3D (three-dimensional) culture approaches on hepatotoxicity screening of compounds and highlight the main factors affecting predictive power of experiments. To this end, we first summarize some of the recognized hepatotoxicity mechanisms and related assays used to appraise DILI mechanisms and then discuss the challenges and limitations of in vitro models.

Relationships between the Structure and Severe Drug-Induced Liver Injury for Low, Medium, and High Doses of Drugs

Assessment of structure-activity relationships (SARs) for predicting severe drug-induced liver injury (DILI) is essential since in vivo and in vitro preclinical methods cannot detect many druglike compounds disrupting liver functions. To date, plenty of SAR models for the prediction of DILI have been developed; however, none of them considered the route of drug administration and daily dose, which may introduce significant bias into prediction results. We have created a dataset of 617 drugs with parenteral and oral administration routes and consistent information on DILI severity. We have found a clear relationship between route, dose, and DILI severity. According to SAR, nearly 40% of moderate- and non-DILI-causing drugs would cause severe DILI if they were administered at high oral doses. We have proposed the following approach to predict severe DILI. New compounds recommended to be used at low oral doses (<∼10 mg daily), or parenterally, can be considered not causing severe DILI. DILI for compounds administered at medium oral doses (∼10-100 mg daily; 22.2% of drugs under consideration) can be considered unpredictable because reasonable SAR models were not obtained due to the small size and heterogeneity of the corresponding dataset. The DILI potential of the compounds recommended to be used at high oral doses (more than ∼100 mg daily) can be estimated using SAR modeling. The balanced accuracy of the approach calculated by a 10-fold cross-validation procedure is 0.803. The developed approach can be used to estimate severe DILI for druglike compounds proposed to use at low and high oral doses or parenterally at the early stages of drug development.

Evaluation of a Three-Dimensional Primary Human Hepatocyte Spheroid Model: Adoption and Industrialization for the Enhanced Detection of Drug-Induced Liver Injury

Drug-induced liver injury is a leading cause of compound attrition during both preclinical and clinical drug development, and early strategies are in place to tackle this recurring problem. Human-relevant in vitro models that are more predictive of hepatotoxicity hazard identification, and that could be employed earlier in the drug discovery process, would improve the quality of drug candidate selection and help reduce attrition. We present an evaluation of four human hepatocyte in vitro models of increasing culture complexity (i.e., two-dimensional (2D) HepG2 monolayers, hepatocyte sandwich cultures, three-dimensional (3D) hepatocyte spheroids, and precision-cut liver slices), using the same tool compounds, viability end points, and culture time points. Having established the improved prediction potential of the 3D hepatocyte spheroid model, we describe implementing this model into an industrial screening setting, where the challenge was matching the complexity of the culture system with the scale and throughput required. Following further qualification and miniaturization into a 384-well, high-throughput screening format, data was generated on 199 compounds. This clearly demonstrated the ability to capture a greater number of severe hepatotoxins versus the current routine 2D HepG2 monolayer assay while continuing to flag no false-positive compounds. The industrialization and miniaturization of the 3D hepatocyte spheroid complex in vitro model demonstrates a significant step toward reducing drug attrition and improving the quality and safety of drugs, while retaining the flexibility for future improvements, and has replaced the routine use of the 2D HepG2 monolayer assay at GlaxoSmithKline.

Development and Application of a Transcriptomic Signature of Bioactivation in an Advanced In Vitro Liver Model to Reduce Drug-induced Liver Injury Risk Early in the Pharmaceutical Pipeline

Early risk assessment of drug-induced liver injury (DILI) potential for drug candidates remains a major challenge for pharmaceutical development. We have previously developed a set of rat liver transcriptional biomarkers in short-term toxicity studies to inform the potential of drug candidates to generate a high burden of chemically reactive metabolites that presents higher risk for human DILI. Here, we describe translation of those NRF1-/NRF2-mediated liver tissue biomarkers to an in vitro assay using an advanced micropatterned coculture system (HEPATOPAC) with primary hepatocytes from male Wistar Han rats. A 9-day, resource-sparing and higher throughput approach designed to identify new chemical entities with lower reactive metabolite-forming potential was qualified for internal decision making using 93 DILI-positive and -negative drugs. This assay provides 81% sensitivity and 90% specificity in detecting hepatotoxicants when a positive test outcome is defined as the bioactivation signature score of a test drug exceeding the threshold value at an in vitro test concentration that falls within 3-fold of the estimated maximum drug concentration at the human liver inlet following highest recommended clinical dose administrations. Using paired examples of compounds from distinct chemical series and close structural analogs, we demonstrate that this assay can differentiate drugs with lower DILI risk. The utility of this in vitro transcriptomic approach was also examined using human HEPATOPAC from a single donor, yielding 68% sensitivity and 86% specificity when the aforementioned criteria are applied to the same 93-drug test set. Routine use of the rat model has been adopted with deployment of the human model as warranted on a case-by-case basis. This in vitro transcriptomic signature-based strategy can be used early in drug discovery to derisk DILI potential from chemically reactive metabolites by guiding structure-activity relationship hypotheses and candidate selection.

Predicting Antituberculosis Drug-Induced Liver Injury Using an Interpretable Machine Learning Method: Model Development and Validation Study

Background

Tuberculosis (TB) is a pandemic, being one of the top 10 causes of death and the main cause of death from a single source of infection. Drug-induced liver injury (DILI) is the most common and serious side effect during the treatment of TB.

Objective

We aim to predict the status of liver injury in patients with TB at the clinical treatment stage.

Methods

We designed an interpretable prediction model based on the XGBoost algorithm and identified the most robust and meaningful predictors of the risk of TB-DILI on the basis of clinical data extracted from the Hospital Information System of Shenzhen Nanshan Center for Chronic Disease Control from 2014 to 2019.

Results

In total, 757 patients were included, and 287 (38%) had developed TB-DILI. Based on values of relative importance and area under the receiver operating characteristic curve, machine learning tools selected patients’ most recent alanine transaminase levels, average rate of change of patients’ last 2 measures of alanine transaminase levels, cumulative dose of pyrazinamide, and cumulative dose of ethambutol as the best predictors for assessing the risk of TB-DILI. In the validation data set, the model had a precision of 90%, recall of 74%, classification accuracy of 76%, and balanced error rate of 77% in predicting cases of TB-DILI. The area under the receiver operating characteristic curve score upon 10-fold cross-validation was 0.912 (95% CI 0.890-0.935). In addition, the model provided warnings of high risk for patients in advance of DILI onset for a median of 15 (IQR 7.3-27.5) days.

Conclusions

Our model shows high accuracy and interpretability in predicting cases of TB-DILI, which can provide useful information to clinicians to adjust the medication regimen and avoid more serious liver injury in patients.

ALP-activated probe for diagnosis of liver injury by multispectral optoacoustic tomography

In this chapter, we highlight the advantages of multispectral optoacoustic tomography (MSOT) technique and the activatable photoacoustic probes in the biomedical field, and give a brief introduction to enzyme-activated probes for disease diagnosis and therapeutic outcome evaluation. We also present a detailed description of the procedures for the synthesis of an activatable small molecule probe C₁X-OR₁ and confirmation of its specific response to alkaline phosphatase in solution and cells. With MSOT, the liposomal C₁X-OR₁ can be utilized for detection of hepatic ALP as well as for in vivo diagnosis of drug-induced liver injury in a three-dimensional manner.

The Medicinal Chemistry of Caffeine

The purine alkaloid caffeine is the most widely consumed psychostimulant drug in the world and has multiple beneficial pharmacological activities, for example, in neurodegenerative diseases. However, despite being an extensively studied bioactive natural product, the mechanistic understanding of caffeine's pharmacological effects is incomplete. While several molecular targets of caffeine such as adenosine receptors and phosphodiesterases have been known for decades and inspired numerous medicinal chemistry programs, new protein interactions of the xanthine are continuously discovered providing potentially improved pharmacological understanding and a molecular basis for future medicinal chemistry. In this Perspective, we gather knowledge on the confirmed protein interactions, structure activity relationship, and chemical biology of caffeine on well-known and upcoming targets. The diversity of caffeine's molecular activities on receptors and enzymes, many of which are abundant in the CNS, indicates a complex interplay of several mechanisms contributing to neuroprotective effects and highlights new targets as attractive subjects for drug discovery.

The Promise of AI for DILI Prediction

Drug-induced liver injury (DILI) is a common reason for the withdrawal of a drug from the market. Early assessment of DILI risk is an essential part of drug development, but it is rendered challenging prior to clinical trials by the complex factors that give rise to liver damage. Artificial intelligence (AI) approaches, particularly those building on machine learning, range from random forests to more recent techniques such as deep learning, and provide tools that can analyze chemical compounds and accurately predict some of their properties based purely on their structure. This article reviews existing AI approaches to predicting DILI and elaborates on the challenges that arise from the as yet limited availability of data. Future directions are discussed focusing on rich data modalities, such as 3D spheroids, and the slow but steady increase in drugs annotated with DILI risk labels.

Stem Cells for Next Level Toxicity Testing in the 21st Century

The call for a paradigm change in toxicology from the United States National Research Council in 2007 initiates awareness for the invention and use of human-relevant alternative methods for toxicological hazard assessment. Simple 2D in vitro systems may serve as first screening tools, however, recent developments infer the need for more complex, multicellular organotypic models, which are superior in mimicking the complexity of human organs. In this review article most critical organs for toxicity assessment, i.e., skin, brain, thyroid system, lung, heart, liver, kidney, and intestine are discussed with regards to their functions in health and disease. Embracing the manifold modes-of-action how xenobiotic compounds can interfere with physiological organ functions and cause toxicity, the need for translation of such multifaceted organ features into the dish seems obvious. Currently used in vitro methods for toxicological applications and ongoing developments not yet arrived in toxicity testing are discussed, especially highlighting the potential of models based on embryonic stem cells and induced pluripotent stem cells of human origin. Finally, the application of innovative technologies like organs-on-a-chip and genome editing point toward a toxicological paradigm change moves into action.

Oxidative Stress in Drug-Induced Liver Injury (DILI): From Mechanisms to Biomarkers for Use in Clinical Practice

Idiosyncratic drug-induced liver injury (DILI) is a type of hepatic injury caused by an uncommon drug adverse reaction that can develop to conditions spanning from asymptomatic liver laboratory abnormalities to acute liver failure (ALF) and death. The cellular and molecular mechanisms involved in DILI are poorly understood. Hepatocyte damage can be caused by the metabolic activation of chemically active intermediate metabolites that covalently bind to macromolecules (e.g., proteins, DNA), forming protein adducts-neoantigens-that lead to the generation of oxidative stress, mitochondrial dysfunction, and endoplasmic reticulum (ER) stress, which can eventually lead to cell death. In parallel, damage-associated molecular patterns (DAMPs) stimulate the immune response, whereby inflammasomes play a pivotal role, and neoantigen presentation on specific human leukocyte antigen (HLA) molecules trigger the adaptive immune response. A wide array of antioxidant mechanisms exists to counterbalance the effect of oxidants, including glutathione (GSH), superoxide dismutase (SOD), catalase, and glutathione peroxidase (GPX), which are pivotal in detoxification. These get compromised during DILI, triggering an imbalance between oxidants and antioxidants defense systems, generating oxidative stress. As a result of exacerbated oxidative stress, several danger signals, including mitochondrial damage, cell death, and inflammatory markers, and microRNAs (miRNAs) related to extracellular vesicles (EVs) have already been reported as mechanistic biomarkers. Here, the status quo and the future directions in DILI are thoroughly discussed, with a special focus on the role of oxidative stress and the development of new biomarkers.

Deep Graph Learning with Property Augmentation for Predicting Drug-Induced Liver Injury

Drug-induced liver injury (DILI) is a crucial factor in determining the qualification of potential drugs. However, the DILI property is excessively difficult to obtain due to the complex testing process. Consequently, an in silico screening in the early stage of drug discovery would help to reduce the total development cost by filtering those drug candidates with a high risk to cause DILI. To serve the screening goal, we apply several computational techniques to predict the DILI property, including traditional machine learning methods and graph-based deep learning techniques. While deep learning models require large training data to tune huge model parameters, the DILI data set only contains a few hundred annotated molecules. To alleviate the data scarcity problem, we propose a property augmentation strategy to include massive training data with other property information. Extensive experiments demonstrate that our proposed method significantly outperforms all existing baselines on the DILI data set by obtaining a 81.4% accuracy using cross-validation with random splitting, 78.7% using leave-one-out cross-validation, and 76.5% using cross-validation with scaffold splitting.

[The trends in predicting drug-induced liver injury]

Drug-induced liver injury (DILI) is the major reason for the discontinuation of new drug development and the withdrawal of drugs from the market. Hence, the evaluation systems which predict the onset of DILI in the pre-clinical stage are needed. To date, many researchers have conducted the mechanism of DILI, but the DILI prediction is poor because of the complexity of DILI. In this regard, based on the information obtained from basic research and clinical case, several pharmaceutical companies have been developed DILI prediction methods with high sensitivity and specificity by combining multiple targets. Another reason for low predictability is derived from the conventional culture method which causes a rapid decrease in hepatocyte function. To overcome these problems, the construction of a high-level in vitro evaluation system has been developed and applied to DILI evaluation. On the other hand, these in vitro evaluation methods require a lot of labor and cost so, in silico prediction methods have also been constructed in recent years. Based on this point, this article reviews the trends in DILI prediction systems in the non-clinical stage.

Evaluation of covalent binding of flutamide and its risk assessment using 19F-NMR

The formation of reactive metabolites (RMs) is a problem in drug development that sometimes results in severe hepatotoxicity. As detecting RMs themselves is difficult, a covalent binding assay using expensive radiolabelled tracers is usually performed for candidate selection. This study aimed to provide a practical approach toward the risk assessment of hepatotoxicity induced by covalent binding before candidate selection. We focused on flutamide because it contains a trifluoromethyl group that shows a strong singlet peak by ¹⁹F nuclear magnetic resonance (NMR) spectrometry. The covalent binding of flutamide was evaluated using quantitative NMR and its risk for hepatotoxicity was assessed by estimating the RM burden, an index that reflects the body burden associated with RM exposure by determining the extent of covalent binding, clinical dose and in vivo clearance. The extent of covalent binding and RM burden was 296 pmol/mg/h and 37.9 mg/day, respectively. Flutamide was categorised as high risk with an RM burden >10 mg/day consistent with its clinical hepatotoxicity. These results indicate that a combination of covalent binding assay using ¹⁹F-NMR and RM burden is useful for the risk assessment of RMs without using radiolabelled compounds.

Strategies to Mitigate the Bioactivation of Aryl Amines

The bioactivation of xenobiotics to yield reactive metabolites can lead to tolerability and toxicity concerns within a drug discovery program. Development of strategies for mitigating the metabolic liability of commonly encountered toxicophores, such as anilines, relies on an understanding of the relative tendency of these functionalities to undergo bioactivation. In this report, we present the first systematic study of the structure-activity relationships of the bioactivation of aryl amine fragments (molecular weight < 250 Da) using a glutathione (GSH) trapping assay in the presence of human liver microsomes and the reduced form of nicotinamide adenine dinucleotide phosphate. This study demonstrates that conversion of anilines to nitrogen-containing heteroarylamines results in a lower abundance of GSH conjugates in the order phenyl > pyrimidine ≈ pyridine > pyridazine. Introduction of electron-withdrawing functionality on the aromatic ring had a less pronounced effect on the extent of GSH conjugation. Examination of more drug-like compounds sourced from in-house drug discovery programs revealed similar trends in bioactivation between matched pairs containing (hetero)aryl amines. This study provides medicinal chemists with insights and qualitative guidance for the minimization of risks related to aryl amine metabolism.

Drug Induced Liver Injury (DILI). Mechanisms and Medicinal Chemistry Avoidance/Mitigation Strategies

Adverse drug reactions (ADRs) are a common cause of attrition in drug discovery and development and drug-induced liver injury (DILI) is a leading cause of preclinical and clinical drug terminations. This perspective outlines many of the known DILI mechanisms and assessment methods used to evaluate and mitigate DILI risk. Literature assessments and retrospective analyses using verified DILI-associated drugs from the Liver Tox Knowledge Base (LTKB) have been used to derive the predictive value of each end point, along with combination approaches of multiple methods. In vitro assays to assess inhibition of the bile salt export pump (BSEP), mitotoxicity, reactive metabolite (RM) formation, and hepatocyte cytolethality, along with physicochemical properties and clinical dose provide useful DILI predictivity. This Perspective also highlights some of the strategies used by medicinal chemists to reduce DILI risk during the optimization of drug candidates.

The evolution of strategies to minimise the risk of human drug-induced liver injury (DILI) in drug discovery and development

Early identification of toxicity associated with new chemical entities (NCEs) is critical in preventing late-stage drug development attrition. Liver injury remains a leading cause of drug failures in clinical trials and post-approval withdrawals reflecting the poor translation between traditional preclinical animal models and human clinical outcomes. For this reason, preclinical strategies have evolved over recent years to incorporate more sophisticated human in vitro cell-based models with multi-parametric endpoints. This review aims to highlight the evolution of the strategies adopted to improve human hepatotoxicity prediction in drug discovery and compares/contrasts these with recent activities in our lab. The key role of human exposure and hepatic drug uptake transporters (e.g. OATPs, OAT2) is also elaborated.

Predicting Drug-Induced Liver Injury with Bayesian Machine Learning

Drug induced liver injury (DILI) can require significant risk management in drug development and on occasion can cause morbidity or mortality, leading to drug attrition. Optimizing candidates preclinically can minimize hepatotoxicity risk, but it is difficult to predict due to multiple etiologies encompassing DILI, often with multifactorial and overlapping mechanisms. In addition to epidemiological risk factors, physicochemical properties, dose, disposition, lipophilicity, and hepatic metabolic function are also relevant for DILI risk. Better human-relevant, predictive models are required to improve hepatotoxicity risk assessment in drug discovery. Our hypothesis is that integrating mechanistically relevant hepatic safety assays with Bayesian machine learning will improve hepatic safety risk prediction. We present a quantitative and mechanistic risk assessment for candidate nomination using data from in vitro assays (hepatic spheroids, BSEP, mitochondrial toxicity, and bioactivation), together with physicochemical (cLogP) and exposure (Cmax_total) variables from a chemically diverse compound set (33 no/low-, 40 medium-, and 23 high-severity DILI compounds). The Bayesian model predicts the continuous underlying DILI severity and uses a data-driven prior distribution over the parameters to prevent overfitting. The model quantifies the probability that a compound falls into either no/low-, medium-, or high-severity categories, with a balanced accuracy of 63% on held-out samples, and a continuous prediction of DILI severity along with uncertainty in the prediction. For a binary yes/no DILI prediction, the model has a balanced accuracy of 86%, a sensitivity of 87%, a specificity of 85%, a positive predictive value of 92%, and a negative predictive value of 78%. Combining physiologically relevant assays, improved alignment with FDA recommendations, and optimal statistical integration of assay data leads to improved DILI risk prediction.

3D Primary Hepatocyte Culture Systems for Analyses of Liver Diseases, Drug Metabolism, and Toxicity: Emerging Culture Paradigms and Applications

Recent research has shown that the maintenance of relevant liver functions ex vivo requires models in which the cells exhibit an in vivo-like phenotype, often achieved by reconstitution of appropriate cellular interactions. Multiple different models have been presented that differ in the cells utilized, media, and culture conditions. Furthermore, several technologically different approaches have been presented including bioreactors, chips, and plate-based systems in fluidic or static media constituting of chemically diverse materials. Using such models, the ability to predict drug metabolism, drug toxicity, and liver functionality have increased tremendously as compared to conventional in vitro models in which cells are cultured as 2D monolayers. Here, the authors highlight important considerations for microphysiological systems for primary hepatocyte culture, review current culture paradigms, and discuss their opportunities for studies of drug metabolism, hepatotoxicity, liver biology, and disease.

Identification of Novel Metabolites of Vildagliptin in Rats: Thiazoline-Containing Thiol Adducts Formed via Cysteine or Glutathione Conjugation

Vildagliptin (VG), a dipeptidyl peptidase-4 inhibitor, is used for treating type 2 diabetes. On rare occasions, VG causes liver injury as an adverse reaction. One case report suggested the involvement of immune responses in the hepatotoxicity, but the underlying mechanisms are unknown. We recently reported that VG binds covalently in vitro to l-cysteine to produce a thiazoline acid metabolite, M407, implying that the covalent binding may trigger the immune-mediated hepatotoxicity. There was no evidence, however, that such a thiazoline acid metabolite was formed in vivo. In the present study, we administered a single oral dose of VG to male Sprague-Dawley rats, and detected M407 in plasma. The sum of urinary and fecal excretions of M407 reached approximately 2% of the dose 48 hours postdosing. Using bile duct-cannulated rats, we demonstrated that M407 was secreted into bile as a glucuronide, designated as M583. Another newly identified thiazoline metabolite of VG, the cysteinylglycine conjugate M464, was detected in urine, feces, and bile. The formation of M464 was confirmed by in vitro incubation of VG with glutathione even in the absence of metabolic enzymes. A glutathione adduct against the nitrile moiety M611 was also detected in vitro but not in vivo. In summary, we found three new thiazoline-containing thiol adduct metabolites in VG-administered rats. Nonenzymatic covalent binding of VG would likely occur in humans, and it may be relevant to predicting adverse reactions.

Prediction of Drug-Induced Hepatotoxicity Using Long-Term Stable Primary Hepatic 3D Spheroid Cultures in Chemically Defined Conditions

High failure rates of drug candidates in the clinics, restricted-use warnings as well as withdrawals of drugs in postmarketing stages are of substantial concern for the pharmaceutical industry and drug-induced liver injury (DILI) constitutes one of the most frequent reasons for such safety failures. Importantly, as DILI cannot be accurately predicted using animal models, animal safety tests are commonly complemented with assessments in human in vitro systems. 3D spheroid cultures of primary human hepatocytes in chemically defined conditions, hereafter termed CD-spheroids, have recently emerged as a microphysiological model system in which hepatocytes retain their molecular phenotypes and hepatic functions for multiple weeks in culture. However, their predictive power for the detection of hepatotoxic liabilities has not been systematically assessed. Therefore, we here evaluated the hepatotoxicity of 123 drugs with or without direct implication in clinical DILI events. Importantly, using ATP quantifications as the single endpoint, the model accurately distinguished between hepatotoxic and nontoxic structural analogues and exceeded both sensitivity and specificity of all previously published in vitro assays at substantially lower exposure levels, successfully detecting 69% of all hepatotoxic compounds without producing any false positive results (100% specificity). Furthermore, the platform supports the culture of spheroids of primary hepatocytes from preclinical animal models, thereby allowing the identification of animal-specific toxicity events. We anticipate that CD-spheroids represent a powerful and versatile tool in drug discovery and preclinical drug development that can reliably flag hepatotoxic drug candidates and provide guidance for the selection of the most suitable animal models.

A UPLC-MS/MS application for comparisons of the hepatotoxicity of raw and processed Xanthii Fructus by energy metabolites

The ripe fruit of Xanthium strumarium L. (Xanthii Fructus) cannot be widely used as a Chinese herbal medicine (CHM) owing to its hepatotoxicity. However, Xanthii Fructus (XF) can be used effectively and safely after correct processing based on traditional experience, although a high hepatotoxicity risk remains owing to improper usage. Therefore, the processing methods used must be clarified to ensure safety. The adenosine-5'-triphosphate (ATP) level in tissues is an important indicator reflecting the functional status of liver cells. Therefore, this study aims to evaluate the hepatotoxicity of XF using UPLC-MS/MS. The hepatotoxicity of raw XF (RXF) and XF processed by intermediary energy metabolites (PXF) is compared. The method is evaluated for its analytical performance and successfully applied to the quantification of ATP, adenosine-5'-diphosphate (ADP), adenosine-5'-monophosphate (AMP), atractyloside, and carboxyatractyloside in mouse liver. The hepatotoxicity results also indicate that the toxicity of XF is decreased after processing, perhaps due to the decrease in atractyloside and carboxyatractyloside contents. Importantly, the experimental evidence provides a rationale for the reduction in toxicity. These data show that mouse livers are damaged between the days 20 and 30 of RXF oral administration, and that the ATP level is decreased. Importantly, no significant difference is observed between the PXF treatment group and control group, while the RXF treatment group is significantly different. Therefore, processing can reduce the toxicity of XF.

Lighting up alkaline phosphatase in drug-induced liver injury using a new chemiluminescence resonance energy transfer nanoprobe

We present a new nanoprobe based on chemiluminescence resonance energy transfer for in vivo imaging of drug-induced alkaline phosphatase (ALP). This probe exhibits the perfect degrees of high sensitivity and specificity for real-time monitoring of ALP levels to directly evaluate drug-induced liver injury.

Identification of a novel metabolite of vildagliptin in humans: Cysteine targets the nitrile moiety to form a thiazoline ring

The dipeptidyl peptidase-4 (DPP-4) inhibitor vildagliptin (VG) is used to treat type 2 diabetes. In rare cases, VG-induced liver injury has been reported. One case report suggested that immune responses were involved in the hepatotoxicity. However, the underlying mechanisms of VG-induced hepatotoxicity are uncertain. In the present study, we investigated whether VG has the potential to covalently bind to macromolecules in cells, a process that could initiate immune-mediated hepatotoxicity. For comparison, M20.7, a major metabolite of VG, and other DPP-4 inhibitors were also evaluated. We found that VG and anagliptin (ANG), which both contain a cyanopyrrolidine moiety, rapidly reacted in non-enzymatic manners on co-incubation with l-cysteine. Both VG and ANG had half-lives of 20-30 min. In contrast, incubation with GSH, rather than l-cysteine, failed to decrease the concentrations of VG or ANG. M20.7, sitagliptin, linagliptin, and alogliptin, having no cyanopyrrolidine moiety, were stable on incubation with l-cysteine or GSH. Structural analysis of the VG- and ANG-cysteine adducts, designated M407 and M487, respectively, revealed that the nitrile moieties of VG and ANG were irreversibly converted to a thiazoline acid. In conclusion, we found that VG and ANG have the potential to covalently bind to a thiol residue of l-cysteine in proteins. Such binding may lead to unpredictable immune responses in humans. l-Cysteine, rather than GSH, would likely be useful to detect the potential for covalent binding that could initiate immune-mediated hepatotoxicity.

Evaluating Drug-Induced Liver Injury and Its Remission via Discrimination and Imaging of HClO and H2S with a Two-Photon Fluorescent Probe

Drug-induced liver injury (DILI) has aroused wide concern. Finding new markers or indicators as well as detoxification molecules for DILI is of great significance and good application prospect, which can help develop effective preclinical screening methodology and corresponding treatment protocols. Herein, in this article, DILI caused by antidepressant drugs of duloxetine and fluoxetine and its remission were evaluated by a two-photon fluorescent probe, RPC-1, through discriminating and imaging HClO and H₂S simultaneously. By being applied both in vitro and in vivo, RPC-1 revealed slight up-regulation of HClO and negligible liver damage after administration of either of the two drugs. In contrast, an apparent up-regulation of HClO and obvious liver damage was observed after combined administration of the drugs. Meanwhile, the pretreatment of N-acetyl cysteine (NAC) resulted in the increasing of endogenous H₂S level, which contributed to the remission of DILI. The histological analysis and serological test both gave good consistency with the imaging results. These findings demonstrate that HClO may be an appropriate indicator of DILI, and H₂S plays an important role in the antidotal effect of NAC. We envision that RPC-1 can be used as a powerful tool to predict clinical DILI and study the effect of antidote, as well as explore the molecular mechanisms involved.

The protective effects of Cichorium glandulosum seed and cynarin against cyclophosphamide and its metabolite acrolein-induced hepatotoxicity in vivo and in vitro

Cyclophosphamide (CP) is a widely utilized chemotherapy drug. CP and its metabolite, acrolein, could induce hepatotoxicity. In this study, Cichorium glandulosum seed (CGS) effectively mitigated CP-induced hepatotoxicity in mice. Protection of cynarin, the major compound of CGS, against acrolein cytotoxicity in HepG2 cells was studied. Pretreatment with cynarin could improve cell survival against acrolein cytotoxicity. Cynarin restored the balance of glutathione (GSH) and reactive oxygen species (ROS), and inhibited mitochondrial depolarization. The kinetics of Nrf2 expression in cytosolic and nuclear fractions were observed after acrolein exposure. Intracellular Nrf2 expression was triggered within 6 h of exposure but did not translocate to the nucleus. Cynarin pretreatment ameliorated the expression and activity of GSH S-transferase and triggered Nrf2 nuclear translocation. In conclusion, treatment with CGS and cynarin protects liver injury against CP and acrolein hepatotoxicity via improvement of GSH activity and activation of the Nrf2 pathway.

Impact of Physicochemical Properties on Dose and Hepatotoxicity of Oral Drugs

A database containing maximum daily doses of 1841 marketed oral drugs was used to examine the influence of physicochemical properties on dose and hepatotoxicity (drug induced liver injury, DILI). Drugs in the highest ∼20% dose range had significantly reduced mean lipophilicity and molecular weight, increased fractional surface area, increased % of acids, and decreased % of bases versus drugs in the lower ∼60% dose range. Drugs in the ∼20-40% dose range had intermediate mean properties, similar to the mean values for the full drug set. Drugs that are both large and highly lipophilic almost invariably do not have doses in the upper ∼20% range. The results show that oral druglike physicochemical properties are different according to these dose ranges, and this is consistent with maintenance of acceptable safety profiles as efficacious exposure increases. Verified DILI annotations from a compilation of >1000 approved drugs (Chen, M.; et al. Drug Discov. Today, 2016, 21, 648 ) were used. The drugs classified as "No DILI" ( n = 163) had significantly lower dose and lipophilicity, and higher Fsp3 (fraction of carbon atoms that are sp³ hybridized) versus the "Most DILI" ( n = 163) drugs. The percentages of acids were reduced and bases increased in the "No DILI" versus the "Most DILI" groups. Drugs classified as "Less DILI" or "Ambiguous DILI" had intermediate mean values of dose, lipophilicity, Fsp3, and % acids and bases. The impact of lipophilicity and Fsp3 on DILI increases in the upper 20% versus the lower 80% dose range, and a simple decision tree model predicted "No DILI" versus "Most DILI" outcomes with 82% accuracy. The model correctly classified 19 of 22 drugs (86%) that failed in development due to human hepatotoxicity. Because many oral drugs lacking DILI annotations are predicted to be "Most DILI", the model is best used preclinically in conjunction with experimental DILI mitigation.

Drug rechallenge following drug-induced liver injury

Drug-induced hepatocellular injury is identified internationally by alanine aminotransferase (ALT) levels equal to or exceeding 5× the upper limit of normal (ULN) appearing within 3 months of drug initiation, after alternative causes are excluded. Upon withdrawing the suspect drug, ALT generally decrease by 50% or more. With drug readministration, a positive rechallenge has recently been defined by an ALT level of 3-5× ULN or greater. Nearly 50 drugs are associated with positive rechallenge after drug-induced liver injury (DILI): antimicrobials; and central nervous system, cardiovascular and oncology therapeutics. Drugs associated with high rates of positive rechallenge exhibit multiple risk factors: daily dose >50 mg, an increased incidence of ALT elevations in clinical trials, immunoallergic clinical injury, and mitochondrial impairment in vitro. These drug factors interact with personal genetic, immune, and metabolic factors to influence positive rechallenge rates and outcomes. Drug rechallenge following drug-induced liver injury is associated with up to 13% mortality in prospective series of all prescribed drugs. In recent oncology trials, standardized systems have enabled safer drug rechallenge with weekly liver chemistry monitoring during the high-risk period and exclusion of patients with hypersensitivity. However, high positive rechallenge rates with other innovative therapeutics suggest that caution should be taken with rechallenge of high-risk drugs.

CONCLUSION

For critical medicines, drug rechallenge may be appropriate when 1) no safer alternatives are available, 2) the objective benefit exceeds the risk, and 3) patients are fully informed and consent, can adhere to follow-up, and alert providers to hepatitis symptoms. To better understand rechallenge outcomes and identify key risk factors for positive rechallenge, additional data are needed from controlled clinical trials, prospective registries, and large health care databases. (Hepatology 2017;66:646-654).