[Progress in novel biomarkers of anti-tuberculosis drug-induced liver injury].

Anti-tuberculosis drug-induced liver injury(ATB-DILI) is the most common adverse reaction during anti-tuberculosis therapy in tuberculosis patients. At present, the diagnosis of ATB-DILI is mainly based on traditional biomarkers such as transaminases, but these indicators have low specificity for liver toxicity, they cannot explain the mechanism of liver injury and the early onset of ATB-DILI. Based on the prediction of disease severity, treatment and prevention, this paper described the current potential biomarkers of ATB-DILI.

Development of a Novel In Silico Classification Model to Assess Reactive Metabolite Formation in the Cysteine Trapping Assay and Investigation of Important Substructures.

Predicting whether a compound can cause drug-induced liver injury (DILI) is difficult due to the complexity of drug mechanism. The cysteine trapping assay is a method for detecting reactive metabolites that bind to microsomes covalently. However, it is cumbersome to use 35S isotope-labeled cysteine for this assay. Therefore, we constructed an in silico classification model for predicting a positive/negative outcome in the cysteine trapping assay. We collected 475 compounds (436 in-house compounds and 39 publicly available drugs) based on experimental data performed in this study, and the composition of the results showed 248 positives and 227 negatives. Using a Message Passing Neural Network (MPNN) and Random Forest (RF) with extended connectivity fingerprint (ECFP) 4, we built machine learning models to predict the covalent binding risk of compounds. In the time-split dataset, AUC-ROC of MPNN and RF were 0.625 and 0.559 in the hold-out test, restrictively. This result suggests that the MPNN model has a higher predictivity than RF in the time-split dataset. Hence, we conclude that the in silico MPNN classification model for the cysteine trapping assay has a better predictive power. Furthermore, most of the substructures that contributed positively to the cysteine trapping assay were consistent with previous results.

[Interpretation of the Chinese guideline for diagnosis and management of drug-induced liver injury (2023 version)].

Drug can cause almost all known types of acute, subacute, and chronic liver injuries. Drug-induced liver injury (DILI) is an important cause of unexplained liver injury in clinical practice. Correct diagnosis of DILI is challenging due to lack of specific diagnostic biomarkers, especially in patients with pre-existing liver disease and multiple concomitant drugs. A comprehensive understanding of the risk factors, clinical features, and prognosis of liver injury caused by different drugs will help physicians to recognize, diagnose, and manage it timely. Although the guideline was developed based on evidence-based medicine provided by the latest studies, there is limited high-quality evidence in the field of DILI. Therefore, this guideline should be interpreted with caution, and physicians should adopt an optimal diagnostic and therapeutic strategy for individual patients within the framework of the guideline.

Phenotyping Hepatic Immune-Related Adverse Events in the Setting of Immune Checkpoint Inhibitor Therapy.

PURPOSE: We present and validate a rule-based algorithm for the detection of moderate to severe liver-related immune-related adverse events (irAEs) in a real-world patient cohort. The algorithm can be applied to studies of irAEs in large data sets. METHODS: We developed a set of criteria to define hepatic irAEs. The criteria include: the temporality of elevated laboratory measurements in the first 2-14 weeks of immune checkpoint inhibitor (ICI) treatment, steroid intervention within 2 weeks of the onset of elevated laboratory measurements, and intervention with a duration of at least 2 weeks. These criteria are based on the kinetics of patients who experienced moderate to severe hepatotoxicity (Common Terminology Criteria for Adverse Events grades 2-4). We applied these criteria to a retrospective cohort of 682 patients diagnosed with hepatocellular carcinoma and treated with ICI. All patients were required to have baseline laboratory measurements before and after the initiation of ICI. RESULTS: A set of 63 equally sampled patients were reviewed by two blinded, clinical adjudicators. Disagreements were reviewed and consensus was taken to be the ground truth. Of these, 25 patients with irAEs were identified, 16 were determined to be hepatic irAEs, 36 patients were nonadverse events, and two patients were of indeterminant status. Reviewers agreed in 44 of 63 patients, including 19 patients with irAEs (0.70 concordance, Fleiss' kappa: 0.43). By comparison, the algorithm achieved a sensitivity and specificity of identifying hepatic irAEs of 0.63 and 0.81, respectively, with a test efficiency (percent correctly classified) of 0.78 and outcome-weighted F1 score of 0.74. CONCLUSION: The algorithm achieves greater concordance with the ground truth than either individual clinical adjudicator for the detection of irAEs.

Salivary metabolites are promising noninvasive biomarkers of drug-induced liver injury.

BACKGROUND: Drug-induced liver injury (DILI) is one of the most common adverse events of medication use, and its incidence is increasing. However, early detection of DILI is a crucial challenge due to a lack of biomarkers and noninvasive tests. AIM: To identify salivary metabolic biomarkers of DILI for the future development of noninvasive diagnostic tools. METHODS: Saliva samples from 31 DILI patients and 35 healthy controls (HCs) were subjected to untargeted metabolomics using ultrahigh-pressure liquid chromatography coupled with tandem mass spectrometry. Subsequent analyses, including partial least squares-discriminant analysis modeling, t tests and weighted metabolite coexpression network analysis (WMCNA), were conducted to identify key differentially expressed metabolites (DEMs) and metabolite sets. Furthermore, we utilized least absolute shrinkage and selection operato and random fores analyses for biomarker prediction. The use of each metabolite and metabolite set to detect DILI was evaluated with area under the receiver operating characteristic curves. RESULTS: We found 247 differentially expressed salivary metabolites between the DILI group and the HC group. Using WMCNA, we identified a set of 8 DEMs closely related to liver injury for further prediction testing. Interestingly, the distinct separation of DILI patients and HCs was achieved with five metabolites, namely, 12-hydroxydodecanoic acid, 3-hydroxydecanoic acid, tetradecanedioic acid, hypoxanthine, and inosine (area under the curve: 0.733-1). CONCLUSION: Salivary metabolomics revealed previously unreported metabolic alterations and diagnostic biomarkers in the saliva of DILI patients. Our study may provide a potentially feasible and noninvasive diagnostic method for DILI, but further validation is needed.

Novel clinical phenotypes, drug categorization, and outcome prediction in drug-induced cholestasis: Analysis of a database of 432 patients developed by literature review and machine learning support.

BACKGROUND: Serum transaminases, alkaline phosphatase and bilirubin are common parameters used for DILI diagnosis, classification, and prognosis. However, the relevance of clinical examination, histopathology and drug chemical properties have not been fully investigated. As cholestasis is a frequent and complex DILI manifestation, our goal was to investigate the relevance of clinical features and drug properties to stratify drug-induced cholestasis (DIC) patients, and to develop a prognosis model to identify patients at risk and high-concern drugs. METHODS: DIC-related articles were searched by keywords and Boolean operators in seven databases. Relevant articles were uploaded onto Sysrev, a machine-learning based platform for article review and data extraction. Demographic, clinical, biochemical, and liver histopathological data were collected. Drug properties were obtained from databases or QSAR modelling. Statistical analyses and logistic regressions were performed. RESULTS: Data from 432 DIC patients associated with 52 drugs were collected. Fibrosis strongly associated with fatality, whereas canalicular paucity and ALP associated with chronicity. Drugs causing cholestasis clustered in three major groups. The pure cholestatic pattern divided into two subphenotypes with differences in prognosis, canalicular paucity, fibrosis, ALP and bilirubin. A predictive model of DIC outcome based on non-invasive parameters and drug properties was developed. Results demonstrate that physicochemical (pKa-a) and pharmacokinetic (bioavailability, CYP2C9) attributes impinged on the DIC phenotype and allowed the identification of high-concern drugs. CONCLUSIONS: We identified novel associations among DIC manifestations and disclosed novel DIC subphenotypes with specific clinical and chemical traits. The developed predictive DIC outcome model could facilitate DIC prognosis in clinical practice and drug categorization.

Rechallenge in idiosyncratic drug-induced liver injury: An analysis of cases in two large prospective registries according to existing definitions.

INTRODUCTION: Data on positive rechallenge in idiosyncratic drug-induced liver injury (DILI) are scarce. We aim to analyse the clinical presentation, outcome and drugs associated with positive rechallenge in two DILI registries. METHODS: Cases from the Spanish and Latin American DILI registries were included. Demographics, clinical characteristics and outcome of cases with positive rechallenge according to CIOMS/RUCAM and current definitions were analysed. RESULTS: Of 1418 patients with idiosyncratic DILI, 58 cases had positive rechallenge (4.1%). Patients with positive rechallenge had shorter duration of therapy (p=0.001) and latency (p=0.003). In patients with rechallenge, aspartate transaminase levels were increased (p=0.026) and showed a prolonged time to recovery (p=0.020), albeit no differences were seen in terms of fatal outcomes. The main drug implicated in rechallenge was amoxicillin-clavulanate (17%). The majority of re-exposure events were unintentional (71%). Using both existing definitions of positive rechallenge, there were four cases which exclusively fulfilled the current criteria and five which only meet the historical definition. All cases of positive rechallenge, irrespective of the pattern of damage, fulfilled the criteria of either alanine transaminase (ALT) ≥3 times the upper limit of normal (ULN) and/or alkaline phosphatase (ALP) ≥2 times ULN. CONCLUSIONS: Episodes of rechallenge were characterised by shorter duration of therapy and latency, and longer time to resolution, but did not show an increased incidence of fatal outcome. Based on our findings, ALT ≥3 times ULN and/or ALP ≥2 times ULN, regardless of the pattern of damage, is proposed as a new definition of rechallenge in DILI.

Evaluation of capillary miR-122 as a prognostic biomarker of paracetamol-induced liver toxicity.

INTRODUCTION: Paracetamol (acetaminophen) overdose is a leading cause of acute liver failure in many Western countries. Diagnostic tools for this poisoning may be suboptimal in some cases and new biomarkers have been investigated. We investigated the role of capillary microRNA-122 (miR-122) as a prognostic biomarker of liver injury in the clinical management of patients with paracetamol overdose. METHODS: In a paracetamol overdose patient cohort, miR-122 was measured by quantitative polymerase chain reaction in a blood drop obtained by a finger prick at the end of an antidote cycle treatment with N-acetylcysteine treatment (12 h). Liver injury was defined as serum alanine aminotransferase (ALT) activity > 100 IU/L collected at 10 or 20 h after the start of treatment. Pearson's correlation analyses were performed. RESULTS: In patients with paracetamol overdose, capillary miR-122 was positively correlated with ALT measured at 10 h and at 20 h (r = 0.83, P < 0.0001; r = 0.96, P < 0.0001, respectively). CONCLUSION: This work supports the potential use of capillary miR-122 as a prognostic biomarker of liver injury throughout clinical management of patients with paracetamol overdose. Capillary miR-122 can be measured in a blood drop collected by a finger prick, a minimally invasive diagnostic test for patient stratification.

Interpretable machine learning in predicting drug-induced liver injury among tuberculosis patients: model development and validation study.

BACKGROUND: The objective of this research was to create and validate an interpretable prediction model for drug-induced liver injury (DILI) during tuberculosis (TB) treatment. METHODS: A dataset of TB patients from Ningbo City was used to develop models employing the eXtreme Gradient Boosting (XGBoost), random forest (RF), and the least absolute shrinkage and selection operator (LASSO) logistic algorithms. The model's performance was evaluated through various metrics, including the area under the receiver operating characteristic curve (AUROC) and the area under the precision recall curve (AUPR) alongside the decision curve. The Shapley Additive exPlanations (SHAP) method was used to interpret the variable contributions of the superior model. RESULTS: A total of 7,071 TB patients were identified from the regional healthcare dataset. The study cohort consisted of individuals with a median age of 47 years, 68.0% of whom were male, and 16.3% developed DILI. We utilized part of the high dimensional propensity score (HDPS) method to identify relevant variables and obtained a total of 424 variables. From these, 37 variables were selected for inclusion in a logistic model using LASSO. The dataset was then split into training and validation sets according to a 7:3 ratio. In the validation dataset, the XGBoost model displayed improved overall performance, with an AUROC of 0.89, an AUPR of 0.75, an F1 score of 0.57, and a Brier score of 0.07. Both SHAP analysis and XGBoost model highlighted the contribution of baseline liver-related ailments such as DILI, drug-induced hepatitis (DIH), and fatty liver disease (FLD). Age, alanine transaminase (ALT), and total bilirubin (Tbil) were also linked to DILI status. CONCLUSION: XGBoost demonstrates improved predictive performance compared to RF and LASSO logistic in this study. Moreover, the introduction of the SHAP method enhances the clinical understanding and potential application of the model. For further research, external validation and more detailed feature integration are necessary.

Drug-induced liver injury as assessed by the updated Roussel Uclaf Causality Assessment Method following mild COVID-19 in a patient under anastrozole therapy-A case report.

BACKGROUND: Anastrozole is a selective aromatase inhibitor used for the treatment of postmenopausal hormone-sensitive breast cancer. The major side effects include osteoporosis, hypercholesterolemia, and musculoskeletal events, such as arthralgia and myalgia. Other adverse events are rare, including symptoms of acne, masculinization, and drug-induced liver injury, with the latter reported in a few cases only. CASE: Here, we report on a patient under anastrozole therapy who developed drug-induced liver injury as assessed by the updated Roussel Uclaf Causality Assessment Method 5 weeks after a mild SARS-CoV-2 infection, which is, to the best of our knowledge, the first report of its kind involving anastrozole. Discontinuation of anastrozole resulted in a marked improvement of the alanine aminotransaminase, and aspartate aminotransaminase as well as normalized lactate dehydrogenase serum levels already seen after 26 days. Surprisingly, however, the cholestatic serum markers gamma-glutamyl transpeptidase and alkaline phosphatase showed a further rise, and took another 4 weeks to drop significantly. CONCLUSION: The presentation of this case is meant to alert physicians to a potential drug-induced liver injury following mild SARS-CoV-2 infection in patients under anastrozole medication.

IL-2-mediated hepatotoxicity: knowledge gap identification based on the irAOP concept.

Drug-induced hepatotoxicity constitutes a major reason for non-approval and post-marketing withdrawal of pharmaceuticals. In many cases, preclinical models lack predictive capacity for hepatic damage in humans. A vital concern is the integration of immune system effects in preclinical safety assessment. The immune-related Adverse Outcome Pathway (irAOP) approach, which is applied within the Immune Safety Avatar (imSAVAR) consortium, presents a novel method to understand and predict immune-mediated adverse events elicited by pharmaceuticals and thus targets this issue. It aims to dissect the molecular mechanisms involved and identify key players in drug-induced side effects. As irAOPs are still in their infancy, there is a need for a model irAOP to validate the suitability of this tool. For this purpose, we developed a hepatotoxicity-based model irAOP for recombinant human IL-2 (aldesleukin). Besides producing durable therapeutic responses against renal cell carcinoma and metastatic melanoma, the boosted immune activation upon IL-2 treatment elicits liver damage. The availability of extensive data regarding IL-2 allows both the generation of a comprehensive putative irAOP and to validate the predictability of the irAOP with clinical data. Moreover, IL-2, as one of the first cancer immunotherapeutics on the market, is a blueprint for various biological and novel treatment regimens that are under investigation today. This review provides a guideline for further irAOP-directed research in immune-mediated hepatotoxicity.

Acute grade IV toxic hepatitis due to the e-cigarette.

A 46-year-old woman presented at the emergency department because of acute hepatitis with jaundice. After hepatological work-up including liver biopsy, drug induced liver disease (DILI) was suspected. Patient recovered completely within a few months. One year later she presented again with jaundice due to acute hepatitis. Vaping was the only agent that could be identified as causative agent for DILI. After VAPING cessation, the hepatitis resolved completely. Calculated RUCAM score was 10, making the diagnosis of toxic hepatitis very likely. During follow-up liver tests remained normal. This is the first report of severe DILI secondary to the use of e-cigarettes. In future vaping can be included in the differential diagnosis of DILI.

Validation of the revised electronic version of RUCAM for diagnosis of DILI in Chinese patients.

BACKGROUND AIMS: The Revised Electronic Causality Assessment Method (RECAM), a computerized update of the Roussel Uclaf Causality Assessment Methodology (RUCAM), was recently proposed. In this study, we validated and compared the utility of the RECAM and RUCAM in Chinese patients with a single conventional or herbal agent-induced liver injury. METHODS: In this retrospective multicenter cohort of well-established DILI and non-DILI patients from 5 centers in China, the diagnostic performance of the RUCAM and RECAM was compared by AUC analysis. The consistency was evaluated by weighted kappa. The major causes of discrepancy were explored. RESULTS: A total of 481 DILI and 100 non-DILI patients were included. In total, 62.6% of the DILI cases were induced by conventional agents, and 37.4% were induced by herbs. The RECAM had relatively higher AUC than RUCAM for overall [0.947 (0.926-0.964) vs. 0.867 (0.836-0.893), p=0.0016], conventional agents [0.923 (0.890-0.949) vs. 0.819 (0.775-0.858), p=0.0185], and herbs [0.972 (0.941-0.989) vs.0.911 (0.866-0.944), p=0.0199]. Latency, scores associated with hepatitis B, and hepatotoxicity information of the insulting drugs were the 3 main causes for the inconsistency between RECAM and RUCAM scores. CONCLUSIONS: The RECAM had relatively better diagnostic performance than RUCAM, with a higher AUC for Chinese DILI patients. Timely updates of the LiverTox category and refinement of serum markers to exclude hepatitis B activity would further improve the applicability of RECAM in areas where the use of herbs and resolution of past HBV infections are common.

Comparative histologic features among liver biopsies with biliary-pattern injury and confirmed clinical diagnoses.

Biliary-pattern injury in the liver (eg, duct injury, ductular reaction, cholestasis) can occur in several conditions, including primary biliary cholangitis (PBC), primary sclerosing cholangitis (PSC), large duct obstruction (LDO), and drug-induced liver injury (DILI). While the histologic changes in these conditions have been individually well described, distinguishing among them remains often challenging, particularly when biopsy samples are limited in size, robust clinical information is unavailable, and/or the pathologist does not feel confident in evaluating liver disease. This study evaluated histologic features that could aid the diagnosis of biliary-pattern injury on biopsy. We reviewed 121 liver biopsies from clinically confirmed cases of PBC, PSC, chronic LDO, or DILI for multiple clinical and histologic parameters. The rates of these histologic findings were then compared among different entities. Onion-skin fibrosis was seen in 14% of PSC in comparison to 0%, 5%, and 0% of PBC, DILI, and chronic LDO (P = 0.031). Florid duct lesions were identified in 21% of PBC compared to 2% of PSC and 0% of DILI and LDO (P = 0.0065). Similarly, 42% of PBC showed lobular granulomas, compared to 7% of PSC, 11% of DILI, and 33% of chronic LDO (P = 0.0001). Cholestasis was more commonly seen in DILI (42%) and chronic LDO (83%) than in PBC (4%) and PSC (16%) (P < 0.0001). Lobular chronic inflammation was found in a significantly higher percentage of PBC and LDO than of PSC and DILI (P = 0.0009). There were significantly fewer cases of PBC showing neutrophils in ductular reaction than PSC, DILI, and LDO (P = 0.0063). Histologic findings that can help suggest a diagnosis in liver biopsies with biliary-pattern injury include florid duct lesions, lobular granulomas, lack of neutrophils in ductular reaction, and lobular chronic inflammation in PBC; onion-skin fibrosis in PSC; cholestasis and feathery degeneration in DILI; and lobular granulomas, lobular chronic inflammation, cholestasis, and feathery degeneration in chronic LDO. These findings are likely most helpful when complicating factors interfere with biopsy interpretation.

Pediatric acute liver failure: Current perspective in etiology and management.

Pediatric acute liver failure (PALF) is a catastrophic clinical condition with very high morbidity and mortality without early detection and intervention. It is characterized by the acute onset of massive hepatocellular injury that releases circulating inflammatory mediators, resulting in metabolic disturbances, coagulopathy, hepatic encephalopathy and multi-organ failure. The etiological spectrum is dominated by hepatotropic viruses, drug-induced liver injury, metabolic and genetic disorders and immune-mediated diseases. Unlike adults, indeterminate causes for acute liver failure constitute a considerable proportion of cases of acute liver failure in children in the west. The heterogeneity of age and etiology in PALF has led to difficulties in developing prognostic scoring. The recent guidelines emphasize prompt identification of PALF, age-appropriate evaluation for hepatic encephalopathy and laboratory evaluation with careful monitoring. Current therapy focuses on supporting the failing liver and other organs, pending either spontaneous recovery or liver transplantation. Targeted therapy is available for a select group of etiologies. Liver transplantation can be lifesaving and a plan for the same should be organized, whenever indicated. The aim of this review is to define PALF, understand its etiopathogenesis, address the challenges encountered during the management and update the latest advances in liver transplantation and non-transplant treatment options in PALF.

Exposure to Synthetic Psychoactive Substances: A Potential Cause for Increased Human Hepatotoxicity Markers.

BACKGROUND: Approximately 30 million people worldwide consume new psychoactive substances (NPS), creating a serious public health issue due to their toxicity and potency. Drug-induced liver injury is the leading cause of liver disease, responsible for 4% of global deaths each year. CONTENT: A systematic literature search revealed 64 case reports, in vitro and in vivo studies on NPS hepatotoxicity. Maximum elevated concentrations of aspartate aminotransferase (136 to 15 632 U/L), alanine transaminase (121.5 to 9162 U/L), total bilirubin (0.7 to 702 mg/dL; 0.04 to 39.03 mmol/L), direct (0.2-15.1 mg/dL; 0.01-0.84 mmol/L) and indirect (5.3 mg/dL; 0.29 mmol/L) bilirubin, alkaline phosphatase (79-260 U/L), and gamma-glutamyltransferase (260 U/L) were observed as biochemical markers of liver damage, with acute and fulminant liver failure the major toxic effects described in the NPS case reports. In vitro laboratory studies and subsequent in vivo NPS exposure studies on rats and mice provide data on potential mechanisms of toxicity. Oxidative stress, plasma membrane stability, and cellular energy changes led to apoptosis and cell death. Experimental studies of human liver microsome incubation with synthetic NPS, with and without specific cytochrome P450 inhibitors, highlighted specific enzyme inhibitions and potential drug-drug interactions leading to hepatotoxicity. SUMMARY: Mild to severe hepatotoxic effects following synthetic NPS exposure were described in case reports. In diagnosing the etiology of liver damage, synthetic NPS exposure should be considered as part of the differential diagnosis. Identification of NPS toxicity is important for educating patients on the dangers of NPS consumption and to suggest promising treatments for observed hepatotoxicity.

Liver toxicity in oncology trials and beyond: a simplified concept for management of hepatocellular drug-induced liver injury in patients with abnormal baseline liver tests.

BACKGROUND: Management of side effects in clinical trials has to balance generation of meaningful data with risk for patients. A toxicity area requiring detailed management guidelines is drug-induced liver injury (DILI). In oncology trials, patients are often included despite baseline liver test abnormalities, for whom there is no consensus yet on levels of liver test changes that should trigger action, such as drug interruption or discontinuation. METHODS: We provide an innovative approach to manage hepatocellular DILI in oncology trials for patients with abnormal baseline alanine aminotransferase (ALT) levels. The algorithm proposed is based on mathematical derivation of action thresholds from those generally accepted for patients with normal baselines. RESULTS: The resulting algorithm is grouped by level of baseline abnormality and avoids calculation of baseline multiples. Suggested layered action levels are 4, 6, and 11 × Upper Limit of Normal (ULN) for patients with baseline ALT between 1.5 and 3 × ULN, and 6, 8, and 12 × ULN for patients with baseline ALT between 3 and 5 × ULN, respectively. CONCLUSIONS: Our concept and resulting algorithm are consistent, transparent, and easy to follow, and the method for derivation from consensus-based thresholds may also be applicable to other drug toxicity areas.

Management of Tacrolimus-Induced Toxicity With Normal Serum Levels After Liver Transplant.

Drug-induced liver injury after liver transplant occurs in 1.7% of patients. Tacrolimus is an effective immunosuppressant that is used to treat acute rejection. Although rare, it can cause toxicity, which is demonstrated by cholestatic liver injury. Here, we present a case of a young male patient who was diagnosed with Wilson disease, had penicillaminechelating therapy, and underwent living related liver transplant. Within 1 month posttransplant, he developed deranged, predominantly cholestatic pattern liver function tests. Laboratory parameters showed total bilirubin of 1.12 mg/ dL, alanine aminotransferase of 553 IU/L, gammaglutamyltransferase of 624 IU/L, and tacrolimus level of 10.2 ng/mL. After thorough evaluation, a liver biopsy was performed. Liver biopsy showed hepatocellular necrosis with centrilobular cholestasis without any evidence of graft rejection. However, with normal level of tacrolimus, the biopsy was suggestive of drug-induced liver injury. Thus, tacrolimus dose was reduced, resulting in improved liver function tests and patient discharge from the hospital. Tacrolimus is an effective immunosuppressant after liver transplant and has the ability to treat early acute rejection. The patient's liver biopsy showed hepatocellular necrosis with centrilobular cholestasis without any evidence of graft rejection. Cholestatic liver injury after tacrolimus usually resolves after dose reduction or by switching to another agent. With demonstrated tacrolimus-induced toxicity in liver transplant recipients, despite normal serum levels, transplant physicians should keep high index of suspicion regarding toxicity in the posttransplant setting.

Chinese guideline for the diagnosis and treatment of drug-induced liver injury: an update.

Drug-induced liver injury (DILI) is an important adverse drug reaction that can lead to acute liver failure or even death in severe cases. Currently, the diagnosis of DILI still follows the strategy of exclusion. Therefore, a detailed history taking and a thorough and careful exclusion of other potential causes of liver injury is the key to correct diagnosis. This guideline was developed based on evidence-based medicine provided by the latest research advances and aims to provide professional guidance to clinicians on how to identify suspected DILI timely and standardize the diagnosis and management in clinical practice. Based on the clinical settings in China, the guideline also specifically focused on DILI in chronic liver disease, drug-induced viral hepatitis reactivation, common causing agents of DILI (herbal and dietary supplements, anti-tuberculosis drugs, and antineoplastic drugs), and signal of DILI in clinical trials and its assessment.

Deep Learning Algorithm Based on Molecular Fingerprint for Prediction of Drug-Induced Liver Injury.

Drug-induced liver injury (DILI) is one the rare adverse drug reaction (ADR) and multifactorial endpoints. Current preclinical animal models struggle to anticipate it, and in silico methods have emerged as a way with significant potential for doing so. In this study, a high-quality dataset of 1573 compounds was assembled. The 48 classification models, which depended on six different molecular fingerprints, were built via deep neural network (DNN) and seven machine learning algorithms. Comparing the results of the DNN and machine learning models, the optional performing model was found as the one developed based on the DNN with ECFP_6 as input, which achieved the area under the receiver operating characteristic curve (AUC) of 0.713, balanced accuracy (BA) of 0.680, and F1 of 0.753. In addition, we used the SHapley Additive exPlanations (SHAP) algorithm to interpret the models, identified the crucial structural fragments related to DILI risk, and selected the top ten substructures with the highest contribution rankings to serve as warning indicators for subsequent drug hepatotoxicity screening studies. The study demonstrates that the DNN models developed based on molecular fingerprints can be a trustworthy and efficient tool for determining the risk of DILI during the pre-development of novel medications.